Technical report on critical concentrations for drug susceptibility testing of isoniazid and the rifamycins (rifampicin, rifabutin and rifapentine)

WHO Collaborating Centre for Tuberculosis Laboratory Strengthening and Diagnostic Technology Evaluation Technical Report on critical concentrations for drug susceptibility testing of isoniazid and the rifamycins (rifampicin, rifabutin and rifapentine) ISBN 978-92-4-001728-3 (electronic version) WHO Collaborating Centre for Tuberculosis Laboratory Strengthening and Diagnostic Technology Evaluation Technical Report on critical concentrations for drug susceptibility testing of isoniazid and the rifamycins (rifampicin, rifabutin and rifapentine) Technical report on critical concentrations for drug susceptibility testing of isoniazid and the rifamycins (rifampicin, rifabutin and rifapentine). ISBN 978-92-4-001728-3 (electronic version) ISBN 978-92-4-001729-0 (print version) © World Health Organization 2021 Some rights reserved. This work is available under the Creative Commons Attribution- NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo ). Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation: “This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition”. Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization (http://www.wipo.int/amc/en/mediation/rules/ ). Suggested citation. Technical report on critical concentrations for drug susceptibility testing of isoniazid and the rifamycins (rifampicin, rifabutin and rifapentine). Geneva: World Health Organization; 2021. Licence: CC BY-NC-SA 3.0 IGO. Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris . Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders . To submit requests for commercial use and queries on rights and licensing, see http://www.who.int/about/licensing . Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned component in the work rests solely with the user. General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by WHO to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising from its use. iii Contents Acknowledgements.................................................................................................................... vii Abbreviations .............................................................................................................................. x Glossary of terms ........................................................................................................................ xi Executive summary ................................................................................................................... xiv SECTION 1: Introduction ............................................................................................................... 1 1.0 Background ..................................................................................................................................... 2 1.1 Scope of the Technical Expert Consultation Meeting ..................................................................... 3 1.2 Systematic review ........................................................................................................................... 3 1.2.1 Search methodology ................................................................................................................ 3 1.2.2 Inclusion criteria ...................................................................................................................... 4 1.2.3 Studies identified through the systematic review .................................................................... 4 1.3 Data presentation ........................................................................................................................... 6 1.3.1 Format of this report................................................................................................................ 6 1.3.2 Format of MIC tables ............................................................................................................... 6 SECTION 2: Isoniazid .................................................................................................................... 9 2.0 INH resistance mechanisms .......................................................................................................... 10 2.1 Current statements and policies regarding genotypic markers of INH resistance ........................ 11 2.2 INH MIC data stratification and current breakpoints .................................................................... 12 2.A.1 INH MIC data on LJ..................................................................................................................... 14 2.A.1.1 INH MICs for pWT isolates on LJ ......................................................................................... 14 2.A.1.2 INH MICs for mutated isolates on LJ ................................................................................... 14 2.A.1.3 Conclusion for INH CC for LJ ................................................................................................ 17 2.A.2 INH MIC data on 7H10 ............................................................................................................... 18 2.A.2.1 INH MICs for pWT isolates on 7H10 .................................................................................... 18 2.A.2.2 INH MICs for mutated isolates on 7H10 ............................................................................. 18 2.A.2.3 Conclusion for INH CC for 7H10 .......................................................................................... 21 2.A.3 INH MIC data on 7H11 ............................................................................................................... 22 2.A.3.1 INH MICs for pWT isolates on 7H11 .................................................................................... 22 2.A.3.2 INH MICs for mutated isolates on 7H11 ............................................................................. 22 2.A.3.3 Conclusion for INH CC for 7H11 .......................................................................................... 23 2.A.4 INH MIC data in MGIT ................................................................................................................ 24 2.A.4.1 INH MICs for pWT isolates in MGIT .................................................................................... 24 2.A.4.2 INH MICs for mutated isolates in MGIT .............................................................................. 24 2.A.4.3 Conclusion for INH CC in MGIT ............................................................................................ 28 2.3 INH conclusions and comments .................................................................................................... 29 2.4 References for INH MIC studies .................................................................................................... 30 SECTION 3: Rifamycins ............................................................................................................... 35 3.0 Rifamcyin resistance mechanisms ................................................................................................ 36 3.1 Current statements and policies regarding genotypic markers of rifamycin resistance ............... 37 3.2 Rifamycin MIC data stratification and current breakpoints .......................................................... 41 3.A.1 RIF MIC data on LJ ...................................................................................................................... 43 3.A.1.1 RIF MICs for pWT isolates on LJ .......................................................................................... 43 3.A.1.2 RIF MICs for mutated isolates on LJ .................................................................................... 43 3.A.1.3 Conclusion for RIF CC for LJ ................................................................................................. 45 3.A.2 RIF MIC data on 7H10 ................................................................................................................ 47 3.A.2.1 RIF MICs for pWT isolates on 7H10 ..................................................................................... 47 3.A.2.2 RIF MICs for mutated isolates on 7H10 .............................................................................. 47 3.A.2.3 Conclusion for RIF CC for 7H10 ........................................................................................... 51 3.A.3 RIF MIC data on 7H11 ................................................................................................................ 52 iv 3.A.3.1 RIF MICs for pWT isolates on 7H11 ..................................................................................... 52 3.A.3.2 RIF MICs for mutated isolates on 7H11 .............................................................................. 52 3.A.3.3 Conclusion for RIF CC for 7H11 ........................................................................................... 53 3.A.4 RIF MIC data in MGIT ................................................................................................................. 54 3.A.4.1 RIF MICs for pWT isolates in MGIT ..................................................................................... 54 3.A.4.2 RIF MICs for mutated isolates in MGIT ............................................................................... 54 3.A.4.3 Conclusion for RIF CC for MGIT ........................................................................................... 58 3.A.5 References for RIF MIC studies .................................................................................................. 60 3.B.1 RFB MIC data on LJ ..................................................................................................................... 64 3.B.1.1 RFB MICs for pWT isolates on LJ ......................................................................................... 64 3.B.1.2 RFB MICs for mutated isolates on LJ ................................................................................... 64 3.B.1.3 Conclusion for RFB CC for LJ ................................................................................................ 64 3.B.2 RFB MIC data on 7H10 ............................................................................................................... 65 3.B.2.1 RFB MICs for pWT isolates on 7H10 .................................................................................... 65 3.B.2.2 RFB MICs for mutated isolates on 7H10 ............................................................................. 65 3.B.2.3 Conclusion for RFB CC for 7H10 .......................................................................................... 67 3.B.3 RFB MIC data on 7H11 ............................................................................................................... 68 3.B.3.1 RFB MICs for pWT isolates on 7H11 .................................................................................... 68 3.B.3.2 RFB MICs for mutated isolates on 7H11 ............................................................................. 68 3.B.3.3 Conclusion for RFB CC for 7H11 .......................................................................................... 69 3.B.4 RFB MIC data in MGIT ................................................................................................................ 70 3.B.4.1 RFB MICs for pWT isolates in MGIT .................................................................................... 70 3.B.4.2 RFB MICs for mutated isolates in MGIT .............................................................................. 70 3.B.4.3 Conclusion for RFB CC for MGIT .......................................................................................... 73 3.B.5 References for RFB MIC studies ................................................................................................. 74 3.C.1 RPT MIC data on LJ ..................................................................................................................... 76 3.C.1.1 RPT MICs for pWT isolates on LJ ......................................................................................... 76 3.C.1.2 RPT MICs for mutated isolates on LJ ................................................................................... 76 3.C.1.3 Conclusion for RPT CC for LJ ................................................................................................ 76 3.C.2 RPT MIC data on 7H10 ............................................................................................................... 77 3.C.2.1 RPT MICs for pWT isolates on 7H10 .................................................................................... 77 3.C.2.2 RPT MICs for mutated isolates on 7H10.............................................................................. 77 3.C.2.3 Conclusion for RPT CC for 7H10 .......................................................................................... 78 3.C.3 RPT MIC data on 7H11 ............................................................................................................... 79 3.C.3.1 RPT MICs for pWT isolates on 7H11 .................................................................................... 79 3.C.3.2 RPT MICs for mutated isolates on 7H11.............................................................................. 79 3.C.3.3 Conclusion for RPT CC for 7H11 .......................................................................................... 79 3.C.4 RPT MIC data in MGIT ................................................................................................................ 80 3.C.4.1 RPT MICs for pWT isolates in MGIT..................................................................................... 80 3.C.4.2 RPT MICs for mutated isolates in MGIT .............................................................................. 80 3.C.4.3 Conclusion for RPT CC for MGIT .......................................................................................... 80 3.C.5 References for RPT MIC studies ................................................................................................. 81 3.3 Rifamycin conclusions and comments .......................................................................................... 82 Rifampicin ....................................................................................................................................... 82 Rifabutin ......................................................................................................................................... 86 Rifapentine ..................................................................................................................................... 87 v List of figures FIGURE 1. PRISMA DIAGRAM FOR ISONIAZID SEARCH RESULTS AND EXCLUSION CRITERIA ....................................... 5 FIGURE 2. PRISMA DIAGRAM FOR RIFAMYCIN SEARCH RESULTS AND EXCLUSION CRITERIA ...................................... 6 List of tables TABLE 1. CRITICAL CONCENTRATIONS FOR INH AND THE RIFAMYCINS. ............................................................... XV TABLE 2. OVERVIEW OF MIC DATA PRESENTATION. ......................................................................................... 7 TABLE 3. PROBE-BINDING REGIONS AND KATG AND INHA PROMOTER MUTATION COVERAGE OF WHO-ENDORSED LPAS. ........................................................................................................................................... 11 TABLE 4. OVERVIEW OF CURRENT INH CCS.................................................................................................. 13 TABLE 5. INH MICS FOR PWT ISOLATES ON LJ. ............................................................................................ 14 TABLE 6. INH MICS FOR CLINICAL KATG S315 MUTANTS ON LJ. ..................................................................... 15 TABLE 7. INH MICS FOR CLINICAL INHA PROMOTER MUTANTS ON LJ. .............................................................. 15 TABLE 8. INH MICS FOR CLINICAL INHA CODING MUTANTS ON LJ. ................................................................... 16 TABLE 9. INH MICS FOR CLINICAL OXYR-AHPC INTERGENIC REGION MUTANTS ON LJ. ......................................... 16 TABLE 10. INH MICS FOR CLINICAL KATG S315, INHA PROMOTER AND INHA CODING DOUBLE MUTANTS ON LJ. ..... 17 TABLE 11. INH MICS FOR PWT ISOLATES ON 7H10. .................................................................................... 18 TABLE 12. INH MICS FOR CLINICAL KATG S315 MUTANTS ON 7H10. .............................................................. 19 TABLE 13. INH MICS FOR CLINICAL INHA PROMOTER MUTANTS ON 7H10. ....................................................... 19 TABLE 14. INH MICS FOR CLINICAL INHA CODING MUTANTS ON 7H10. ........................................................... 20 TABLE 15. INH MICS FOR CLINICAL KATG S315, INHA PROMOTER AND INHA CODING DOUBLE MUTANTS ON 7H10. 21 TABLE 16. INH MICS FOR PWT ISOLATES ON 7H11. .................................................................................... 22 TABLE 17. INH MICS FOR CLINICAL KATG S315 MUTANTS ON 7H11. .............................................................. 22 TABLE 18. INH MICS FOR CLINICAL INHA PROMOTER MUTANTS ON 7H11. ....................................................... 22 TABLE 19. INH MICS FOR CLINICAL KATG S315, INHA PROMOTER AND INHA CODING DOUBLE MUTANTS ON 7H11. 23 TABLE 20. INH MICS FOR PWT ISOLATES IN MGIT. ..................................................................................... 24 TABLE 21. INH MICS FOR CLINICAL KATG S315 MUTANTS IN MGIT. ............................................................... 25 TABLE 22. INH MICS FOR CLINICAL INHA PROMOTER MUTANTS IN MGIT. ........................................................ 26 TABLE 23. INH MICS FOR CLINICAL INHA CODING MUTANTS IN MGIT. ............................................................ 26 TABLE 24. INH MICS FOR CLINICAL OXYR-AHPC INTERGENIC REGION MUTANTS IN MGIT. ................................... 26 TABLE 25. INH MICS FOR CLINICAL KATG S315, INHA PROMOTER AND INHA CODING DOUBLE MUTANTS IN MGIT.. 27 TABLE 26. OVERVIEW OF CONFIDENCE-GRADED RPOB RIF RESISTANCE MUTATIONS, INCLUDING THE SEVEN BORDERLINE RESISTANCE MUTATIONS. ................................................................................................................. 39 TABLE 27. PROBE-BINDING REGIONS AND RPOB MUTATION COVERAGE OF WHO-ENDORSED GDST ASSAYS. ........... 40 TABLE 28. OVERVIEW OF CURRENT RIFAMYCIN CCS. ...................................................................................... 42 TABLE 29. RIF MICS FOR PWT ISOLATES ON LJ. ........................................................................................... 43 TABLE 30. RIF MICS FOR RPOB S450 (S531) MUTANTS ON LJ. ...................................................................... 43 TABLE 31. RIF MICS FOR RPOB BORDERLINE RRDR MUTANTS ON LJ. .............................................................. 44 TABLE 32. RIF MICS FOR OTHER RPOB RRDR MUTANTS ON LJ. ...................................................................... 45 TABLE 33. RIF MICS FOR RPOB MUTANTS OUTSIDE THE RRDR ON LJ. .............................................................. 45 TABLE 34. EFFECT OF INTRODUCING AN ATU FOR RIF DST ON LJ. ................................................................... 46 TABLE 35. RIF MICS FOR PWT ISOLATES ON 7H10. ..................................................................................... 47 TABLE 36. RIF MICS FOR RPOB S450 (S531) MUTANTS ON 7H10. ................................................................ 48 TABLE 37. RIF MICS FOR RPOB BORDERLINE RRDR MUTANTS ON 7H10. ......................................................... 49 TABLE 38. RIF MICS FOR OTHER RPOB RRDR MUTANTS ON 7H10. ................................................................ 50 TABLE 39. RIF MICS FOR RPOB MUTANTS OUTSIDE THE RRDR ON 7H10. ........................................................ 50 TABLE 40. EFFECT OF CHANGING THE CC FOR RIF DST ON 7H10 FOR DETECTION OF RPOB MUTANTS.................... 51 TABLE 41. RIF MICS FOR PWT ISOLATES ON 7H11. ..................................................................................... 52 TABLE 42. RIF MICS FOR RPOB S450 (S531) MUTANTS ON 7H11. ................................................................ 52 TABLE 43. RIF MICS FOR OTHER RPOB RRDR MUTANTS ON 7H11. ................................................................ 53 TABLE 44. RIF MICS FOR PWT ISOLATES IN MGIT. ...................................................................................... 54 vi TABLE 45. RIF MICS FOR RPOB S450 (S531) MUTANTS IN MGIT. ................................................................. 55 TABLE 46. RIF MICS FOR RPOB BORDERLINE RRDR MUTANTS IN MGIT. .......................................................... 56 TABLE 47. RIF MICS FOR OTHER RPOB RRDR MUTANTS IN MGIT. ................................................................. 57 TABLE 48. RIF MICS FOR RPOB MUTANTS OUTSIDE THE RRDR IN MGIT. ......................................................... 58 TABLE 49. EFFECT OF CHANGING THE CC FOR RIF DST ON MGIT FOR DETECTION OF RPOB MUTANTS. .................. 59 TABLE 50. RFB MICS FOR PWT ISOLATES ON 7H10. .................................................................................... 65 TABLE 51. RFB MICS FOR RPOB S450 (S531) MUTANTS ON 7H10. ............................................................... 65 TABLE 52. RFB MICS FOR RPOB BORDERLINE RRDR MUTANTS ON 7H10. ........................................................ 66 TABLE 53. RFB MICS FOR OTHER RPOB RRDR MUTANTS ON 7H10. ............................................................... 66 TABLE 54. RFB MICS FOR RPOB MUTANTS OUTSIDE THE RRDR ON 7H10. ....................................................... 67 TABLE 55. EFFECT OF ADOPTING A CC FOR RFB DST ON 7H10 FOR DETECTION OF RPOB MUTANTS. ..................... 67 TABLE 56. RFB MICS FOR PWT ISOLATES ON 7H11. .................................................................................... 68 TABLE 57. RFB MICS FOR RPOB S450 (S531) MUTANTS ON 7H11. ............................................................... 68 TABLE 58. RFB MICS FOR OTHER RPOB RRDR MUTANTS ON 7H11. ............................................................... 69 TABLE 59. RFB MICS FOR PWT ISOLATES IN MGIT. ..................................................................................... 70 TABLE 60. RFB MICS FOR RPOB S450 (S531) MUTANTS IN MGIT. ................................................................ 70 TABLE 61. RFB MICS FOR RPOB BORDERLINE RRDR MUTANTS IN MGIT. ......................................................... 71 TABLE 62. RFB MICS FOR OTHER RPOB RRDR MUTANTS IN MGIT. ................................................................ 72 TABLE 63. RFB MICS FOR RPOB MUTANTS OUTSIDE OF RRDR IN MGIT. ......................................................... 72 TABLE 64. EFFECT OF ADOPTING A CC FOR RFB DST ON MGIT FOR DETECTION OF RPOB MUTANTS. ..................... 73 TABLE 65. RPT MICS FOR PWT ISOLATES ON 7H10. .................................................................................... 77 TABLE 66. RPT MICS FOR RPOB S450 (S531) MUTANTS ON 7H10. ............................................................... 77 TABLE 67. RPT MICS FOR RPOB BORDERLINE RRDR MUTANTS ON 7H10. ........................................................ 77 TABLE 68. RPT MICS FOR OTHER RPOB RRDR MUTANTS ON 7H10. ............................................................... 78 TABLE 69. RPT MICS FOR PWT ISOLATES ON 7H11. .................................................................................... 79 vii Acknowledgements WHO Steering Committee The development of this document was led by Nazir Ismail (WHO Global TB Programme) with input from Alexei Korobitsyn, Matteo Zignol and Karin Weyer (WHO Global TB Programme), on the basis of a systematic review of critical concentrations written by Sophia Georghiou (FIND, Switzerland) and Claudio Köser (University of Cambridge, United Kingdom of Great Britain and Northern Ireland), and consensus agreed at a Technical Expert Group meeting convened by WHO on 24 February 2020, in Geneva, Switzerland. Members of the WHO Technical Expert Consultation Group Lynette Berkeley (US Food and Drug Administration, United States of America), Daniela Cirillo (San Raffaele Scientific Institute, Italy), Kelly Dooley (Johns Hopkins University School of Medicine, United States of America), Gerry Davies (Liverpool School of Tropical Medicine, United Kingdom of Great Britain and Northern Ireland), Bernard Fourie (University of Pretoria, South Africa), Christopher Gilpin (International Organization for Migration, Switzerland), Patricia Hall (Centers for Disease Control and Prevention, United States of America), Rumina Hasan (Aga Khan University, Pakistan), Nazir Ismail (National Institute of Communicable Diseases, South Africa), Marguerite Massinga Loembé (Africa Centres for Disease Control and Prevention, Gabon), Charles Peloquin (University of Florida, United States of America), Max Salfinger (University of South Florida, United States of America), Thomas Shinnick (Independent consultant, United States of America), Thomas Schön (Linköping University, Sweden), Sabira Tahseen (National Reference Laboratory, Pakistan), Gabriela Torrea (Prince Leopold Institute of Tropical Medicine, Belgium) and Armand Van Deun (Prince Leopold Institute of Tropical Medicine, Belgium). Systematic reviewers/technical consultants Sophia Georghiou and Claudio Köser. Observer Brigit Quinn (Becton Dickinson, United States of America). Additional literature reviewers Alexandra Aubry (National Reference Center for Mycobacteria, France), Ivan Barilar (Research Centre Borstel, Germany), Céline Cunen (University of Oslo, Norway), Victoria Furio (Institute of Biomedicine of Valencia, Spain), Hairong Huang (Beijing Tuberculosis and Thoracic Tumor Institute, China), Kiyohiko Izumi (Research Institute of Tuberculosis, Japan), Tomasz Jagielski (University of Warsaw, Poland), Mikashmi Kohli (McGill University, Canada), Alexei Korobitsyn, Paolo Miotto (San Raffaele Scientific Institute, Italy), Satoshi Mitarai (Research Institute of Tuberculosis, Japan), Vaidehi Nafade (McGill University, Canada), Vlad Nikolayevskyy (National Mycobacterium Reference Laboratory, United Kingdom of Great Britain and Northern Ireland), Thomas Schön, Natalia Shubladze (National Reference Mycobacteriology Laboratory, Georgia), Erik Svensson (Statens Serum Institut, Denmark), Akiko Takaki (Research Institute of Tuberculosis, Japan), Jakko van Ingen (Radboud University Medical Center, Netherlands), Nicolas Veziris (National Reference Center for Mycobacteria, France), Miguel Viveiros (Universidad de Nova de Lisboa, Portugal), Suan Wen (Beijing Tuberculosis and Thoracic Tumor Institute, China), Ferda Yılmaz (Ege University, Turkey) and Tingting Zhang (Beijing Tuberculosis and Thoracic Tumor Institute, China). viii Major data contributors Said Abbadi (Suez University, Egypt), David Alland (Rutgers New Jersey Medical School, United States of America), Sönke Andres (National Tuberculosis Reference Laboratory, Germany), Marie Ballif (University of Bern, Switzerland), Zenda Berrada (California Department of Public Health, United States of America), Henry Boom (Case Western Reserve University, United States of America), Emanuele Borroni (San Raffaele Scientific Institute, Italy), Erik Böttger (University of Zurich, Switzerland), Emmanuelle Cambau (National Reference Center for Mycobacteria and Antimycobacterial Resistance, France), Soumitesh Chakravorty (Cepheid, United States of America), Daniela Cirillo, Edward Desmond (California Department of Public Health, United States of America), Jurriaan de Steenwinkel (Erasmus University Medical Center, Netherlands), Reynaldo Dietze (Universidade Federal do Espirito Santo, Brazil), Matthias Egger (University of Bern, Switzerland), Maha Farhat (Harvard Medical School, United States of America), Lanfranco Fattorini (Istituto Superiore di Sanità, Italy), Ramona Groenheit (Public Health Agency of Sweden, Sweden), Scott Heysell (University of Virginia, United States of America), Eric Houpt (University of Virginia, United States of America), Tomasz Jagielski, John Johnson (Case Western Reserve University, United States of America), Eranga Karunaratne (University of Colombo, Sri Lanka), Peter Keller (University of Bern, Switzerland), Grace Lin (California Department of Public Health, United States of America), Diana Machado (Universidade NOVA de Lisboa, Portugal), Koleka Mlisana (National Health Laboratory Service, South Africa), Nomonde Mvelase (University of Kwazulu Natal, South Africa), Alphonse Okwera (Uganda- CWRU Research Collaboration, Uganda), Ralf Otto-Knapp (HELIOS Klinikum Emil von Behring, Germany), Juan Palacios (Hospital Universitario Central de Asturias, Spain), Jennifer Perera (University of Colombo, Sri Lanka), Catherine Pierre-Audigier (Institut Pasteur, France), Leen Rigouts (Prince Leopold Institute of Tropical Medicine, Belgium), Jaime Robledo (Corporación Para Investigaciones Biológicas, Colombia), James Posey (Centers for Disease Control and Prevention, United States of America), Neesha Rockwood (Imperial College, United Kingdom of Great Britain and Northern Ireland), Marcos Schechter (Emory University School of Medicine, United States of America), Thomas Schön, Meenu Sharma (National Reference Centre for Mycobacteriology, Canada), Frederick Sirgel (Stellenbosch University, South Africa), Elizabeth Streicher (Stellenbosch University, South Africa), Gabriela Torrea, Armand Van Deun, Jakko van Ingen, Annelies van Rie (University of Antwerp, Belgium), Dick van Soolingen (National Institute for Public Health and the Environment, Netherlands), Andrew Vernon (Centers for Disease Control and Prevention, United States of America), Miguel Viveiros, Rob Warren (Stellenbosch University, South Africa) and Jim Werngren (Public Health Agency of Sweden, Sweden). Acknowledgement of financial support This work was commissioned by and funded in part by WHO. Further funding was received from FIND. The United States Agency for International Development and the Federal Government of Germany covered a proportion of Claudio Köser’s salary for this report as part of his work for the European Laboratory Initiative, led by Soudeh Ehsani (WHO Regional Office for Europe, Denmark). Claudio Köser is research associate at Wolfson College and visiting scientist at the Department of Genetics, University of Cambridge. Moreover, we thank Gunnar Kahlmeter and Erika Matuschek (European Committee on Antimicrobial Susceptibility Testing Development Laboratory, Sweden) for their advice and the European Society of Clinical Microbiology and Infectious Diseases for support provided through an observership. Alexandra Aubry, Stefan Niemann (Research Center Borstel, Germany), Leen Rigouts and Nicolas Veziris paid for publications that were not otherwise accessible. ix Provision of publications The following individuals helped with obtaining papers for this review: Kanchan Ajbani (P.D. Hinduja Hospital and Medical Research Centre, India), Iñaki Comas (Institute of Biomedicine of Valencia, Spain), Chris Coulter (Queensland Mycobacterium Reference Laboratory, Australia), Ellis Kelly (University of Cambridge, United Kingdom of Great Britain and Northern Ireland), Laure Fournier Le Ray (Sorbonne Universités, France), Priti Kambli (P.D. Hinduja Hospital and Medical Research Centre, India), Margarita Kholina (Russian Academy of Sciences, Russian Federation), Matthias Merker (Research Centre Borstel, Germany), Lies Mertens (Prince Leopold Institute of Tropical Medicine, Belgium), Sheila Palmer (University of Cambridge, United Kingdom of Great Britain and Northern Ireland), Clare Pierard (International Journal of Tuberculosis and Lung Disease, France), Camilla Rodrigues (P.D. Hinduja Hospital and Medical Research Centre, India), Therdsak Prammananan (National Center for Genetic Engineering and Biotechnology, Thailand), Christina van Dijk (Prince Leopold Institute of Tropical Medicine, Belgium) and Danila Zimenkov (Russian Academy of Sciences, Russian Federation). Declaration and management of conflict of interest All the contributors completed a WHO Declaration of Interest form. All stated declarations of interest were evaluated by members of the Steering Group for the existence of any possible financial conflict of interest which might warrant exclusion from membership of the Technical Expert Consultation Group or from the discussions as part of the consensus process. Intellectual conflict of interest was not considered for exclusion from membership of the Group, as broader expertise on DST methods for MTBC was considered as criteria for the selection. In addition, the diversity and representation in the Groups was large enough to balance and overcome any potential intellectual conflict of interest. During the consensus development process and the meeting, any emergence of intellectual conflict of interest was monitored by the Chair and there was no perceived intellectual conflict of interest identified during the meeting. x Abbreviations 7H10 = Middlebrook 7H10 medium 7H11 = Middlebrook 7H11 medium ATU = area of technical uncertainty BACTEC = BACTEC™ 460 CB = clinical breakpoint CC = critical concentration CI = exact binomial confidence interval CLSI = Clinical & Laboratory Standards Institute DST = drug-susceptibility testing ECOFF = epidemiological cut-off value ELI = European Laboratory Initiative EUCAST = European Committee on Antimicrobial Susceptibility Testing FIND = Foundation for Innovative New Diagnostics gDST = genotypic drug susceptibility testing gNWT = genotypically non-wild type GLI = Global Laboratory Initiative GTB = Global TB Program gWT = genotypically wild type HLR = high-level resistance/resistant INH = isoniazid ISO = International Organization for Standardization LLR = low-level resistance/resistant LJ = Löwenstein-Jensen medium LPA = line probe assay MDR = multidrug-resistant MGIT = BACTEC™ Mycobacterial Growth Indicator Tube™ 960 MIC = minimum inhibitory concentration MTBC = Mycobacterium tuberculosis complex NGS = next generation sequencing pDST = phenotypic drug susceptibility testing pNWT = phenotypically non-wild type PK/PD = pharmacokinetic/pharmacodynamic PMID = PubMed ID pWT = phenotypically wild type R = resistance/resistant RFB = rifabutin RIF = rifampicin RPT = rifapentine RRDR = rifampicin resistance-determining region S = susceptible/susceptibility TB = tuberculosis TEG = Technical Expert Group WHO = World Health Organization xi Glossary of terms Antimicrobial susceptibility test interpretive category – a classification based on an in vitro response of an organism to an antimicrobial agent. For mycobacteria, two different categories, “critical concentration” and “minimum inhibitory concentration,” have been used to categorise the in vitro results. For isolates of Mycobacterium tuberculosis complex, when tested against the lower concentration of some agents, the “critical concentration” category is applied. Testing of an additional higher concentration (a clinical breakpoint concentration) may also be recommended for some agents. However, there is no “intermediate” interpretive category, even when testing is performed both at the critical concentration and the clinical breakpoint concentration. Critical concentration of an anti-tuberculous agent has been adopted and modified from international convention. The critical concentration is defined as the lowest concentration of an anti-TB agent in vitro that will inhibit the growth of 99% of phenotypically wild type isolates of M. tuberculosis complex. Clinical breakpoint – is the concentration or concentrations of an antimicrobial agent which defines an MIC above the critical concentration that separates isolates that will likely respond to treatment from those which will likely not respond to treatment. This concentration is determined by correlation with available clinical outcome data, MIC distributions, genetic markers, and pharmacokinetic/pharmacodynamic data including drug dose. A dose increase can be used to overcome resistance observed at lower dosing, up until the maximum tolerated dose, and therefore a higher clinical breakpoint above which the particular drug is not recommended for use. The clinical breakpoint is used to guide individual clinical decisions in patient treatment. The clinical breakpoint is not applicable for drug resistance surveillance purposes. Critical proportion – is the proportion of resistant organisms within a particular cultured isolate that is used to determine resistance to a particular drug. A 1% critical proportion is used to differentiate susceptible and resistant isolates. Any culture that shows less than 1% growth on a medium containing a critical concentration of the agent being tested when compared with the growth on a control without the agent is considered to be susceptible; a culture that has 1% or more growth on the medium containing the critical concentration of the agent is considered to be resistant, and the patient whose sample is being tested may not respond to the agent. The critical concentration and proportion criteria are used for testing most first-line and second-line anti-TB agents. Cross-resistance is resistance to multiple anti-tuberculosis agents caused by a single genetic change (or multiple changes, in case the given resistance mechanisms requires several genetic alterations), although in practice, such mutations may not be known. Epidemiological cut-off value (ECOFF), phenotypically wild type (pWT) and non-wild type (pNWT) isolates • Typically, when MICs that are tested using a standardised method are aggregated for one species, a single Gaussian-shaped MIC distribution is formed, which corresponds to the pWT distribution for that species (i.e. the distribution for organisms that lack phenotypically detectable resistance mechanisms). Additional distributions with higher overall MICs are sometimes identified, even prior to the clinical use of the particular drug in question (or prior to the clinical use of another, related drug that shares the same resistance mechanism), that xii correspond to intrinsically or naturally resistant organisms. In this case, the distribution with the lowest MICs corresponds to the pWT distribution and the other distributions correspond to one or more pNWT distributions. • The ECOFF corresponds to the upper end of the pWT distribution (i.e. it typically encompasses 99% of pWT isolates). • Excluding the scenario where it is difficult to distinguish pWT and pNWT isolates because of methodological variation in MIC testing (i.e. where both distributions overlap), pWT isolates are, by definition, genotypically WT (gWT). However, this does not mean that gWT isolates are identical genotypically since they may harbour mutations in genes associated with resistance that do not change the MIC (e.g. the gyrA S95T mutation does not affect the MICs of fluoroquinolones). • Conversely, organisms with MICs above the ECOFF are by definition pNWT. Again, excluding the possibility of methodological testing variation close to the ECOFF, there should be a genetic basis for this phenotype (i.e. the isolates should be genotypically NWT (gNWT)). Yet in practice, these gNWT isolates may appear to be gWT if: o The gene conferring the phenotype was not interrogated. o The gene was interrogated, but the genetic change conferring the phenotype was not detected, as it occurred at a frequency below the level of detection of the molecular test (i.e. heteroresistance). o The genetic change was detected but could not be interpreted because of an incomplete understanding of the genotype-phenotype relationship. Indirect susceptibility test – a procedure based on inoculation of drug-containing media using organisms grown in culture. Minimum inhibitory concentration (MIC) – the lowest concentration of an antimicrobial agent that prevents growth of more than 99% a microorganism in a solid medium or broth dilution susceptibility test. Potency – All antimicrobial agents are assayed for standard units of activity or potency. The assay units may differ widely from the actual weight of the powder and often may differ between drug production lots. Thus, a laboratory must standardise its antimicrobial solutions based on assays of the antimicrobial powder lots that are being used. The value for potency supplied by the manufacturer should include consideration for: • Purity measures (usually by high-performance liquid chromatography assay) • Water content (e.g. by Karl Fischer analysis or by weight loss on drying) • Salt/counter-ion fraction (if the compound is supplied as a salt instead of free acid or base) The potency may be expressed as a percentage, or in units of micrograms per milligrams (w/w). Proportion method: The proportion method was originally proposed by Canetti and colleagues, and modified later; it is the most common method used for testing the susceptibility of M. tuberculosis complex isolates. In this method, the inoculum used is monitored by testing two dilutions of a culture suspension, and the growth (that is, the number of colonies) on a control medium without an anti-TB agent is compared with the growth (the number of colonies) present on a medium containing the critical concentration of the anti-TB agent being tested; the ratio of the number of colonies on the medium containing the anti-TB agent to the number of colonies on the medium without the anti-TB xiii agent is calculated, and the proportion is expressed as a percentage. A 1% critical proportion is used to differentiate the proportion of resistant organisms within a particular sample that is used to determine resistance to a particular drug. xiv Executive summary The effective management of multidrug-resistant tuberculosis (MDR-TB) relies upon the rapid diagnosis and treatment of resistant infections. Growth-based phenotypic drug susceptibility testing (pDST) methods are currently the gold standard for drug resistance detection, but these methods are time-consuming, need strict quality control and are not always reproducible. Traditionally, pDST for the Mycobacterium tuberculosis complex (MTBC) has relied on the testing of anti-TB agents at a single, critical concentration (CC) particular to each drug.1 Traditionally, CCs were set based on expert opinion, taking into account the culture medium, reading time, and minimum inhibitory concentrations (MICs) of the phenotypically wild type (pWT) populations. In 2017, the World Health Organization (WHO) Global TB Programme commissioned FIND to perform a systematic review to inform a Technical Expert Group (TEG) tasked with evaluating the evidence base to establish or revise the CCs for 13 second-line anti-TB drugs, including bedaquiline and delamanid. The following media were considered: Löwenstein-Jensen (LJ), Middlebrook 7H10 (7H10), Middlebrook 7H11 (7H11) and BACTEC™ Mycobacterial Growth Indicator Tube™ 960 (MGIT). Based upon the TEG’s review of the systematic review data, revised CCs were established for many of those compounds.2 In 2020, the WHO Global TB Program (GTB) convened a second TEG meeting to review the results of an equivalent systematic review of the published literature for the first-line anti-TB drugs isoniazid (INH) and the rifamycins (rifampicin (RIF), rifabutin (RFB) and rifapentine (RPT)). The quality and quantity of MIC data were limited for most drugs and media. Therefore, the TEG adopted a pragmatic approach by lowering CCs that were clearly above the epidemiological cut-off value (i.e. the 7H10 and MGIT CCs for RIF) but maintaining the remaining CCs (Table 1).3 However, the validity of the 7H11 CC for RIF was questioned and the need for more data to evaluate the RIF CC on LJ was apparent. These findings underscored the need for greater standardization and validation of pDST.4,5 The TEG recommended that seven “borderline resistance” rpoB mutations, which have been referred to as “discordant”, “disputed”, “occult” or “(sub-breakpoint) low-level resistance” mutations in the literature (e.g. L430P (L511P) and I491F (I572F)) need to be treated with an MDR-TB regimen according to the latest WHO guidelines. 6,7 In addition, the interpretation of the remaining mutations in the RIF resistance-determining region (RRDR) was clarified (Section 3.3). 1 Ängeby K, Juréen P, Kahlmeter G, Hoffner S, Schön T. Challenging a dogma: antimicrobial susceptibility testing breakpoints for Mycobacterium tuberculosis. Bull World Health Organ. 2012. doi:10.2471/blt.11.096644. 2 World Health Organization. Technical report on critical concentrations for TB drug susceptibility testing of medicines used in the treatment of drug-resistant TB; 2018 (https://apps.who.int/iris/bitstream/handle/10665/260470/WHO-CDS-TB- 2018.5-eng.pdf, accessed 12 December 2018). 3 Köser CU, Maurer FP, Kranzer K. 'Those who cannot remember the past are condemned to repeat it': Drug-susceptibility testing for bedaquiline and delamanid. Int J Infect Dis. 2019;80S:S32-S35. doi:10.1016/j.ijid.2019.02.027. 4 Schön T, Köser CU, Werngren J, et al. What is the role of the EUCAST reference method for MIC testing of the Mycobacterium tuberculosis complex? Clin Microbiol Infect. 2020;26(11):1453-1455. doi:10.1016/j.cmi.2020.07.037. 5 Schön T, Matuschek E, Mohamed S, et al. Standards for MIC testing that apply to the majority of bacterial pathogens should also be enforced for Mycobacterium tuberculosis complex. Clin Microbiol Infect. 2019;25(4):403-405. doi:10.1016/j.cmi.2019.01.019. 6 Please refer to Section 3.2 for more details regarding the different rpoB numbering systems. 7 World Health Organization. WHO consolidated guidelines on tuberculosis. Module 4: treatment – drug-resistant tuberculosis treatment (https://apps.who.int/iris/rest/bitstreams/1280998/retrieve, accessed 15 June 2020). xv Changes to the CCs for INH were not warranted based on the available evidence. A detailed review of pharmacokinetic/pharmacodynamic and clinical outcome data for different INH resistance mechanisms is planned for 2021 to potentially set a clinical breakpoint for INH to stratify the level of resistance. Table 1. Critical concentrations for INH and the rifamycins. Drug LJ 7H10 7H11 MGIT Isoniazid Rifampicina Rifabutinb Rifapentinec 0.2 40 – – 0.2 0.5 – – 0.2 1.0 – – 0.1 0.5 – – All concentrations are in mg/L and apply to the proportion method with 1% as the critical proportion. Changes to the previous version of the table are highlighted in red.8 a Additional data are needed to clarify whether the RIF CC for LJ is set correctly. The RIF CC for 7H11 was based on limited data and might be too high in light of the fact that the RIF CC for 7H10 had to be lowered to 0.5 mg/L. b No CCs were adopted as RFB is not currently recommended for TB treatment by WHO, but the validity of the current CCs of 0.5 mg/L for 7H10, 7H11 and MGIT set by the Clinical and Laboratory Standards Institute could not be confirmed in this review.9 As a conservative approach, gDST and, where applicable, pDST for RIF should serve as surrogate for RFB DST (Section 3.3). c gDST and, where applicable, pDST for RIF should serve as surrogate for RPT DST (see Section 3.3). 8 World Health Organization. Updated interim critical concentrations for first-line and second-line DST (as of May 2012) (http://www.stoptb.org/wg/gli/assets/documents/Updated critical concentration table_1st and 2nd line drugs.pdf, accessed 7 June 2019). 9 Clinical & Laboratory Standards Institute. Clinical and Laboratory Standards Institute. Susceptibility testing of mycobacteria, nocardiae, and other aerobic actinomycetes, 3rd edition approved standard. CLSI Document M24; 2018. 1 SECTION 1: Introduction 2 1.0 Background Tuberculosis (TB) causes 10 million cases and 1.4 million deaths annually and it is estimated that 2.9 million cases go undiagnosed by public health services each year.10 The global declines in TB incidence, as well as the global reduction in the total number of TB deaths observed in recent years, fall far short of the End TB Strategy milestones for 2020. Ending the global TB epidemic will only be achievable given intensive action by all countries, including a commitment to enhanced, multisectoral actions that have been demonstrated to drive down the epidemic at a rapid pace. TB drug resistance is a major global public health problem that has threatened global progress made in TB care and prevention in recent decades. Drug resistance in the Mycobacterium tuberculosis complex (MTBC) is caused by selection of naturally occurring mutants. There are two ways that people get drug-resistant TB (DR-TB). Firstly, acquired DR-TB occurs when TB treatment is suboptimal due to inadequate policies and failures of health systems and care provision, poor quality of TB drugs, poor prescription practices, patient non-adherence, or a combination of the above. Secondly, primary DR- TB results from the direct transmission of DR-TB from one person to another, which is responsible for most multidrug-resistant TB (MDR-TB) in high-burden settings. In 2019, approximately half a million people worldwide developed TB that was resistant to rifampicin (RIF), the most effective first-line drug.10 78% of these patients were infected with TB that was additionally resistant to isoniazid (INH) and, consequently, had MDR-TB.10 The End TB Strategy calls for early diagnosis and prompt treatment of all persons of all ages with any form of drug-susceptible or -resistant TB. This requires ensuring access to rapid diagnostics recommended by the World Health Organization (WHO) and universal access to drug susceptibility testing (DST) for all patients with signs and symptoms of TB and no longer only prioritizing persons at greater risk of MDR-TB and-or HIV-associated TB. WHO defines universal access to DST as rapid DST for at least RIF, and further DST for at least fluoroquinolones among all TB patients with RIF resistance.11 The effective management of both drug-susceptible and -resistant TB relies upon the rapid diagnosis and treatment of infections. Culture-based phenotypic DST (pDST) methods are currently the gold standard for drug resistance detection, but these methods are time-consuming and require sophisticated laboratory infrastructure, qualified staff and strict quality control. Traditionally, DST for MTBC has relied on the testing of a single, critical concentration (CC), which is used to differentiate resistant from susceptible isolates of MTBC, and is specific for each anti-TB agent and test method. However, the definition of the CC for MTBC DST has evolved over time and now considers more explicitly phenotypically wild type (pWT) vs. phenotypically non-wild type (pNWT) isolates, as defined by the European Committee on Antimicrobial Susceptibility Testing (EUCAST).12 Laboratory tests of the susceptibility of tubercle bacilli to anti-TB agents serve three main purposes. Firstly, they can be used as guidance in the choice of chemotherapy to be given to a patient. Secondly, 10 World Health Organization. Global tuberculosis report 2020 (https://apps.who.int/iris/bitstream/handle/10665/336069/9789240013131-eng.pdf, accessed 14 October 2020). 11 World Health Organization. WHO operational handbook on tuberculosis. Module 3: diagnosis – rapid diagnostics for tuberculosis detection (https://apps.who.int/iris/rest/bitstreams/1284635/retrieve, accessed 30 June 2020). 12 Kahlmeter G. The 2014 Garrod Lecture: EUCAST - are we heading towards international agreement? J Antimicrob Chemother. 2015;70(9):2427-2439. doi:10.1093/jac/dkv145. 3 they are of value in confirming that drug resistance has emerged when a patient has failed to show a satisfactory response to treatment and, thirdly, they can be used for the surveillance of emerging drug resistance. 1.1 Scope of the Technical Expert Consultation Meeting The WHO Global TB Program (GTB) initiated and provided oversight to the process of evidence retrieval and analysis, was responsible for selection of members for the Technical Expert Group (TEG), for management of declarations of interest, and, finally, conducting the face-to-face TEG meeting. As a part of evidence retrieval and analysis, the WHO GTB commissioned the systematic review, which was performed by FIND in 2018-2019. The aim of the review was to collect the available data on minimum inhibitory concentrations (MICs) of pWT and pNWT isolates, including associated sequencing data for relevant gene regions, for the following anti-TB drugs: • INH • Rifamycins (RIF, rifabutin (RFB), and rifapentine (RPT)) The following media were considered: • Löwenstein-Jensen (LJ) • Middlebrook 7H10 (7H10) • Middlebrook 7H11 (7H11) • BACTEC™ Mycobacterial Growth Indicator Tube™ 960 (MGIT) The objectives of the TEG were to revise and update the CCs for pDST for INH and the rifamycins. The TEG meeting was convened by the Global TB Programme, WHO on 24 February 2020 in Geneva, Switzerland. During that meeting, the group assessed the MIC and sequencing data for each drug- medium combination, with a particular focus on potential sources of bias. Depending on the quality and quantity of the data, CCs were either established, maintained or revised. The decisions on the breakpoints for all anti-TB drugs in the review were based on majority view of TEG members. The outcome of the TEG was an updated table of the CCs for INH and rifamycin DST (Table 1). 1.2 Systematic review 1.2.1 Search methodology A MEDLINE/PubMed search without date restrictions was conducted of all publications reporting quantitative DST results for the selected antibiotics. The search terms for each drug or group of drugs, which can be found in the supplement of this report, were intentionally broad since the titles or abstracts of papers do not necessarily mention MIC data. Moreover, MIC data were also solicited from the WHO Supranational Reference Laboratory Network and directly from key researchers, as identified through the literature search and laboratory network. 4 Studies in the following languages were reviewed independently by one or more people: 1. English: Sophia Georghiou and Mikashmi Kohli or Vaidehi Nafade 2. Chinese: Hairong Huang, Suan Wen or Tingting Zhang 3. Croatian and Serbian: Ivan Barilar 4. Danish and Swedish: Thomas Schön or Erik Svensson 5. Dutch: Jakko van Ingen 6. French: Alexandra Aubry or Nicolas Veziris 7. German: Claudio Köser 8. Italian: Paolo Miotto 9. Japanese: Kiyohiko Izumi, Satoshi Mitarai or Akiko Takaki 10. Norwegian: Céline Cunen 11. Portuguese: Miguel Viveiros 12. Polish: Tomas Jagielski 13. Russian: Alexei Korobitsyn or Natalia Shubladze 14. Spanish: Victoria Furio 15. Turkish: Ferda Yılmaz 16. Ukrainian: Vlad Nikolayevskyy 1.2.2 Inclusion criteria Studies identified as containing any MIC data through the full-text screening were further reviewed in detail by Sophia Georghiou and/or Claudio Köser. Studies that met the following criteria were included in the review: 1. The MICs for at least one of the anti-TB compounds of interest (with at least three concentrations tested per drug) were determined using the proportion method with a critical proportion of 1%, using LJ, 7H10, 7H11 or MGIT. 2. The drug concentrations tested were clearly defined (i.e. to assess potential truncations of the MIC results). 3. The number of isolates tested at each concentration was given (i.e. to evaluate the shape of the MIC distributions and determine the mode of the distributions). 4. The MIC data were available for at least 10 isolates per drug. For studies that reported only MIC ranges (i.e. did not meet the third criterion), raw study data were solicited directly from the corresponding authors and/or their co-authors. These studies were excluded if detailed MIC data could not be obtained. In exceptional circumstances, studies that did not meet all of these criteria were still included if they presented data that were particularly valuable, such as studies with sequencing data for anti-TB drugs. 1.2.3 Studies identified through the systematic review For the INH review, a total of 1,408 records were identified for possible inclusion, along with 79 additional datasets from other sources. As shown in Figure 1, 70 of these studies were included in the review, which were stratified further by medium (NB: the sum of the studies for individual media does not correspond to 70 as some studies featured MICs for multiple media). 5 For the rifamycins review, a total of 7,359 records were identified for possible inclusion, along with 23 additional datasets from other sources. As shown in Figure 2, 72 of these studies were included in the review, which were stratified further by medium (NB: the sum of the studies for individual media does not correspond to 72 as some studies featured MICs for multiple media). Figure 1. PRISMA diagram for isoniazid search results and exclusion criteria 6 Figure 2. PRISMA diagram for rifamycin search results and exclusion criteria 1.3 Data presentation 1.3.1 Format of this report Each chapter in the report covers a single drug or group of drugs that share at least one resistance mechanism (i.e. the rifamycins). The results in each chapter are grouped by different media (LJ, 7H10, 7H11 and MGIT). For each medium, data are organised into three sections: (1) MICs for pWT isolates, including laboratory control isolates (e.g. H37Rv), (2) MICs for isolates with mutations in relevant resistance genes (i.e. MICs from in vitro, animal or clinical isolates as well as allelic exchange experiments, where available), (3) conclusion for CC for each combination of drug and medium, including the rationale for the revised or existing breakpoints. 1.3.2 Format of MIC tables This report contains abridged versions of the full Excel MIC data files, which are included in the supplement. Details for the information provided in each column of these files can be found below. However, only essential columns were included in this report. For example, the column with the “total [number of] MICs” performed was included only if these numbers differed from the numbers of unique isolates tested (i.e. when isolates were tested repeatedly, as was the often the case for H37Rv). 7 The following points are relevant for the interpretation of the data: • If a cell is empty, no information regarding the particular category were available (i.e. in the case of the “genotypic results” column, blank cells are not equivalent to gWT (where sequencing or another genotypic method was carried out but no relevant genetic changes were found)). • MICs from different studies cannot be compared unless the concentrations and ranges of concentrations tested are considered. Shaded cells therefore designate the concentrations tested for each group of isolates (NB: some studies tested a wide range of concentrations. Table 2 provides an overview of how MIC data are displayed. Table 2. Overview of MIC data presentation. Shaded cells correspond to the concentrations tested in a particular study (e.g. concentrations of 0.5, 1 and 2 mg/L were tested for study A, whereas 0.5 and 2 mg/L were not tested in study B, which means that MICs of 1 mg/L in both studies are not equivalent). Truncated MIC values were highlighted in red. If red was used in a shaded cell, the MIC was either ≤ or ≥ the concentration in question. For example, the lowest MIC value for study B was ≤0.25 mg/L, whereas the highest MICs were 8 mg/L. If red was used in an unshaded cell, the MIC was > the last concentration tested (for study A, the highest MICs were >2 mg/L, as opposed to 4 mg/L). The mode of the putative pWT MIC distribution was indicated by highlighting the corresponding number of MICs in bolded text (e.g. 1 mg/L for study A). In the case of study B, the truncation of the MIC values meant that a mode could not be identified (e.g. it was possible that the MICs of all 20 isolates with MICs ≤0.25 mg/L were actually 0.25 mg/L, in which case 0.25 mg/L would be the mode of the MIC distribution). The following information are provided in each data column. “Studies” column: • The names of the studies with notable limitations were highlighted in red (e.g. if the same laboratory participated in multiple studies that used the same medium or a method other than sequencing was used for genotypic DST (gDST)). The corresponding limitations were detailed below the tables in the footnotes in this report and in the ‘comment’ column in red in the supplementary MIC file. “Lab” column: • The laboratories that participated in multiple studies using the same medium were highlighted in red. “Unique isolates” & “total MICs” columns: • Red entries correspond to isolates that were tested multiple times. Studies 0.25 0.5 1 2 4 8 study A 15 2 2 study B 20 15 2 RIF MIC [m g/L] 8 “Comment” column: • Additional remarks regarding the study in question were included in this column. Important limitations were highlighted in red. 9 SECTION 2: Isoniazid 10 2.0 INH resistance mechanisms Despite the simple structure of INH, a large number of genes have been implicated in resistance to this drug, but the clinical relevance of many of these mechanisms remains poorly understood.13 The two WHO-endorsed line probe assays (LPAs) by Hain (Hain Lifescience, Nehren, Germany) and Nipro (NIPRO Corporation, Osaka, Japan) rely upon the detection of well-characterized INH resistance mutations in katG and the promoter of inhA (Table 3). The S315T mutation in katG reduces but does not fully abolish the function of this catalase-peroxidase which is needed for the activation of the pro- drug INH.14,15 By contrast, inhA encodes an enoyl acyl carrier protein reductase that represents the shared target of INH as well as the second-line compounds ethionamide and prothionamide.14 The promoter-up mutations upstream of the fabG1-inhA operon interrogated by both LPAs, therefore, confer cross-resistance to all three compounds. The same effect is achieved by a mutation in fabG1, which creates an alternative promoter for inhA.16 Finally, some gDST assays interrogate mutations upstream of ahpC, which compensate for loss-of-function mutations in katG and, thus, serve as indirect markers for INH resistance.17 Numerous studies have found that mutations in inhA confer lower MIC increases to INH than katG mutations.18,19 However, no study has been conducted to date to compare all four WHO-endorsed media systematically. Such a review would guide the interpretation of molecular tests by confirming the level of INH resistance conferred by the aforementioned mechanisms and could inform the setting of a clinical breakpoint (CB). 13 Merker M, Kohl TA, Barilar I, et al. Phylogenetically informative mutations in genes implicated in antibiotic resistance in Mycobacterium tuberculosis complex. Genome Med. 2020;12(1):27. doi:10.1186/s13073-020-00726-5. 14 Vilchèze C, Jacobs JR. WR. Resistance to isoniazid and ethionamide in Mycobacterium tuberculosis: Genes, mutations, and causalities. Microbiol Spectr. 2014;2(4):MGM2-0014-02913. doi:10.1128/microbiolspec.mgm2-0014-2013. 15 Seifert M, Catanzaro D, Catanzaro A, Rodwell TC. Genetic mutations associated with isoniazid resistance in Mycobacterium tuberculosis: A systematic review. PLoS One. 2015;10(3):e0119628. doi:10.1371/journal.pone.0119628. 16 Ando H, Miyoshi-Akiyama T, Watanabe S, Kirikae T. A silent mutation in mabA confers isoniazid resistance on Mycobacterium tuberculosis. Mol Microbiol. 2014;91(3):538-547. doi:10.1111/mmi.12476. 17 Sherman DR, Mdluli K, Hickey MJ, et al. Compensatory ahpC gene expression in isoniazid-resistant Mycobacterium tuberculosis. Science. 1996;272(5268):1641-3. doi:10.1126/science.272.5268.1641. 18 Lempens P, Meehan CJ, Vandelannoote K, et al. Isoniazid resistance levels of Mycobacterium tuberculosis can largely be predicted by high-confidence resistance-conferring mutations. Sci Rep. 2018;8(1):3246. doi:10.1038/s41598-018-21378-x. 19 Ghodousi A, Tagliani E, Karunaratne E, et al. Isoniazid resistance in Mycobacterium tuberculosis is a heterogeneous phenotype composed of overlapping MIC distributions with different underlying resistance mechanisms. Antimicrob Agents Chemother. 2019;63(7):e00092-19. doi:10.1128/AAC.00092-19. 11 Table 3. Probe-binding regions and katG and inhA promoter mutation coverage of WHO-endorsed LPAs. Assay Probe(s) Codons/nucleotides analysed or specific mutation detecteda Mutations covered as per package insert Hain GenoType MTBDRplus V220,21 katG WT 315 region S315T katG MUT1 agc/acc S315Tb katG MUT2 agc/aca S315Tb katG WT1, MUT1 and MUT2 all negative katG deletion inhA WT1 -15 region c-15t, a-16g inhA WT2 -8 region t-8c, t-8a inhA MUT1 c-15tb inhA MUT2 a-16g inhA MUT3A t-8c inhA MUT3B t-8a Nipro NTM+MDRTB II22 katG S7 294-299 katG S8 313-317 katG S9 323-327 katG S10 325-330 katG R8a S315Tb katG R8b S315Nb inhA S6 -17 to -3 inhA R6a a-16g inhA R6b c-15t inhA R6c t-8c inhA R6d t-8a a Hain has not disclosed which precise katG or inhA codons are covered by the corresponding “WT” or “MUT” probes. The same applies to the “R” probes of the Nipro LPA. The regions covered by the Hain “WT” or Nipro “S” probes may not be covered completely and not all mutations in these regions, particularly if they occur at the edges of the probes, prevent binding.21 b Mutations associated with INH resistance according to Miotto et al. (see Section 2.1).23 2.1 Current statements and policies regarding genotypic markers of INH resistance The following statements, policy recommendations and practices regarding genotypic INH resistance are relevant to this review: 1. Miotto et al. conducted a comprehensive analysis of the association between INH resistance mutations and pDST results, which yielded only katG S315N and S315T and the c-15t inhA promoter mutations as markers for INH resistance.23 No or insufficient evidence was found for the remaining inhA promoter mutations interrogated by the WHO-endorsed LPAs (Table 3). 20 Hain Lifescience. GenoType MTBDRplus VER 2.0. Instructions for use. IFU-304A-09. (https://www.hain- lifescience.de/include_datei/kundenmodule/packungsbeilage/download.php?id=2877, accessed 1 Nov 2020). 21 World Health Organization. Regional Office for Europe. Interpretation guide for GenoType MTBDRplus VER 2.0 and GenoType MTBDRsl VER 2.0. A technical guidance document developed by the European Laboratory Initiative. Version 1.0. (https://openwho.org/courses/multi-drug-resistant-tb, accessed 9 May 2020). 22 FIND. Report for WHO: Non-inferiority evaluation of Nipro NTM+MDRTB and Hain GenoType MTBDRplus V2 line probe assays. Version 4.1. Geneva, Switzerland; 2015 (https://www.finddx.org/wp-content/uploads/2016/04/LPA- report_noninferiority-study_oct2015.pdf, accessed 22 May 2020). 23 Miotto P, Tessema B, Tagliani E, et al. A standardised method for interpreting the association between mutations and phenotypic drug resistance in Mycobacterium tuberculosis. Eur Respir J. 2017;50(6):1701354. doi:10.1183/13993003.01354- 2017. 12 2. For the purposes of surveillance, WHO has adopted a composite reference standard for INH resistance.24 Specifically, the three aforementioned mutations from Miotto et al. were considered to be true markers for INH resistance (i.e. the presence of any of these mutations was deemed to be necessary and sufficient to confirm INH resistance). Any susceptible pDST result for an isolate with one of these mutations was thereby corrected to resistant.24 Without this correction, 3.8% (exact binomial 95% confidence interval (CI) 2.9-4.8%) of INH-R isolates would have been characterized as INH-S in a recent multi-country surveillance study conducted by WHO. 3. The Global Laboratory Initiative (GLI) has recommended that the INH results for the Hain LPA are stratified into low-level resistance (LLR) and high-level resistance (HLR). 25 , 26 This has been implemented in a recently published guide by the European Laboratory Initiative (ELI), where isolates with a katG mutation or deletion are reported as HLR, whereas those with only an inhA promoter mutation are reported as “at least LLR” (“at least” was used to signal that higher MICs are possible due to mechanisms that are not interrogated by the LPA).21 2.2 INH MIC data stratification and current breakpoints All mutations within the INH resistance-associated gene regions katG, fabG1-inhA, inhA, oxyR-ahpC, ahpC, kasA, ndh, nat, mshA and fabG1 were noted for this report and included whenever reported. oxyR-ahpC mutations were numbered relative to the start of the ahpC gene (e.g. g-48a corresponds to g-6a relative to the transcriptional start site). Synonymous mutations were considered to be gWT for this report with the exception of the INH resistance-associated mutation fabG1 L203L. 27 Additionally, polymorphisms that are likely not associated with phenotypic INH resistance (inhA g- 102a, t-80g, g-47c, and T4I, and katG A110V, R463L and L499M, mshA A187V and N111S) were considered to be gWT and not reported herein.28 For the purposes of this report, MIC data were stratified based upon the most common mutations reported in katG, the inhA promoter, the inhA coding regions, and the oxyR-ahpC intergenic region as well as mutations in more than one of these gene regions. For the correlation between resistance mutations and phenotypic results, binomial 95% CIs were calculated, where applicable (i.e. including all isolates where sufficient concentrations were tested to determine whether the isolates were resistant or susceptible at a particular concentration). Table 4 provides an overview of the current WHO and CLSI CCs and CBs for INH. 24 Zignol M, Cabibbe AM, Dean AS, et al. Genetic sequencing for surveillance of drug resistance in tuberculosis in highly endemic countries: a multi-country population-based surveillance study. Lancet Infect Dis. 2018;18(6):675-683. doi:10.1016/S1473-3099(18)30073-2. doi:10.1016/S1473-3099(18)30073-2. 25 Global Laboratory Initiative. Line probe assays for drug-resistant tuberculosis detection: interpretation and reporting guide for laboratory staff and clinicians; 2019 (http://www.stoptb.org/wg/gli/assets/documents/LPA_test_web_ready.pdf, accessed 22 May 2020). 26 World Health Organization. WHO operational handbook on tuberculosis. Module 3: diagnosis – rapid diagnostics for tuberculosis detection (https://apps.who.int/iris/rest/bitstreams/1284635/retrieve, accessed 30 June 2020). 27 Ando H, Miyoshi-Akiyama T, Watanabe S, Kirikae T. A silent mutation in mabA confers isoniazid resistance on Mycobacterium tuberculosis. Mol Microbiol. 2014;91(3):538-547. doi:10.1111/mmi.12476. 28 Seifert M, Catanzaro D, Catanzaro A, Rodwell TC. Genetic mutations associated with isoniazid resistance in Mycobacterium tuberculosis: A systematic review. PLoS One. 2015;10(3):e0119628. doi:10.1371/journal.pone.0119628. 13 Table 4. Overview of current INH CCs. LJ 7H10 7H11 MGIT Drug WHO CLSI WHO CLSI WHO CLSI WHO CLSI INH 0.2 0.25/1.0 0.2/1.0 0.2/1.0 0.1/0.4 Green CCs were set by both the WHO and CLSI; red CCs were set by WHO; blue CCs and CBs were set by CLSI.29,30 All concentrations are in mg/L. 29 Clinical & Laboratory Standards Institute. Clinical and Laboratory Standards Institute. Susceptibility testing of mycobacteria, nocardiae, and other aerobic actinomycetes, 3rd edition approved standard. CLSI Document M24; 2018. 30 World Health Organization. Technical manual for drug susceptibility testing of medicines used in the treatment of tuberculosis. (http://apps.who.int/iris/bitstream/handle/10665/275469/9789241514842-eng.pdf, accessed 17 November 2018). 14 2.A.1 INH MIC data on LJ 2.A.1.1 INH MICs for pWT isolates on LJ Eleven studies were identified that reported INH MIC data for the pWT population on LJ (Table 5). Most of these studies were enriched for INH-R isolates and all identified pWT MIC distributions were severely truncated at the lower end, precluding an assessment of the shape of the distributions. Table 5. INH MICs for pWT isolates on LJ. The red line denotes the current WHO CC for INH DST on LJ (0.2 mg/L), whereas the blue lines denote the CLSI CC and CB (0.25 and 1 mg/L). Notable limitations: studies I2, I4, I11, and studies I9 and I10 were conducted in the same laboratory; study I7 presented data for isolates only characterized by LPA; and studies I12 and I10 presented data mostly for INH-R isolates. 2.A.1.2 INH MICs for mutated isolates on LJ katG S315 mutants Seven studies reported MIC data for 352 clinical isolates with katG S315 mutations and no co-occurring inhA promoter or coding mutations (Table 6). Based on the current INH CCs, all but one (99.7%, 95% CI 98-100%) of these mutated isolates were phenotypically INH-R. The modes of most MIC distributions were 2.5-10 mg/L on this medium; above the current CCs and CLSI CB. Studies Lab Isolate origin Unique isolates Type of isolates Genotypic summary 0.05 0.1 0.2 0.25 0.5 0.8 1 1.6 2 2.5 3.2 4 5 8 10 2) Vincent 2012 2 clinical 98 Mix of first-line resistance profiles gWT 46 44 4 1 1 2 4) Lempens 2018 2 clinical 52 Mix of first-line resistance profiles gWT 22 29 1 2 282 99 56 56 26 45 2 80 5 2 22 12 2 37 2 59 9 5 21 24 2 27 6 4 17 2 124 9 17 20 1 11 18 48 2 84 9 14 12 5 9 1 34 12) Farhat 2019 2 clinical 46 Mostly MDR gWT 27 13 4 2 5 1 H37Rv 1 5 clinical 17 15 2 9) Brossier 2009 & 9 clinical 20 Mix of first-line resistance profiles gWT 17 3 9 1 H37Rv 1 9 clinical 4 Mostly INH-R gWT 3 1 7) Alame-Emane 2015 6, 7 clinical 50 Mix of first-line resistance profiles gWT 49 1 1 1 H37Rv 1 1 clinical 66 Mix of MDR, INH mono-resistant and pan-S gWT 48 12 4 2 3) Lee 2000 3 1 H37Rv 1 8) Thai 2018 8 clinical 214 Mix of first-line resistance profiles 26 81 94 8 2 3 INH MIC [mg/L] 1) Jagielski 2015 & Jagielski 2013 & Jagielski 11) Rigouts 2013 & Van Deun (unpublished) clinical Mix of first-line resistance profiles 6) Beckers 1985 10) Brossier 2016 15 Table 6. INH MICs for clinical katG S315 mutants on LJ. The red line denotes the current WHO CC for INH DST on LJ (0.2 mg/L), whereas the blue lines denote the CLSI CC and CB (0.25 and 1 mg/L). Notable limitations: studies I9 and I10, and studies I2, I4 and I12 were conducted in the same laboratory and study I7 presented data for isolates only characterized by LPA and real-time PCR melting curve analysis. inhA promoter mutants Seven studies reported MIC data for 83 clinical isolates with inhA promoter mutations and no co- occurring katG S315 or inhA coding mutations (Table 7). Based on the current INH CC, all (100%, 95% CI 95-100%) of these mutated isolates were phenotypically INH-R. The modes of most MIC distributions were 0.8-1 mg/L on this medium (i.e. equivalent to the CLSI CB), though these INH MICs may be higher than the true MICs, given that these were usually the first concentrations tested above the CC. Table 7. INH MICs for clinical inhA promoter mutants on LJ. The red line denotes the current WHO CC for INH DST on LJ (0.2 mg/L), whereas the blue lines denote the CLSI CC and CB (0.25 and 1 mg/L). Notable limitations: studies I2, I4 and I12 were conducted in the same laboratory and study I7 presented data for isolates only characterized by LPA and real-time PCR melting curve analysis. Studies Lab Isolate origin Unique isolates Genotypic summary 0.05 0.1 0.2 0.25 0.5 0.8 1 1.6 2.5 3.2 5 6.4 10 12.8 19.2 2 clinical 1 katG S315R 1 2 clinical 6 katG S315N 2 3 1 2 clinical 1 katG S315I 1 2 clinical 79 katG S315T 3 35 41 2 clinical 1 katG S315I 1 2 clinical 1 katG S315R 1 2 clinical 2 katG S315N 1 1 2 clinical 44 katG S315T 1 2 15 26 2 clinical 63 katG S315T 9 50 4 2 clinical 3 katG S315T + kasA G269S 2 1 2 clinical 1 katG S315T + oxyR-ahpC c-52t 1 2 clinical 1 katG S315T + T275A 1 2 clinical 1 katG S315T + G192A 1 2 clinical 1 katG S315T + T625A 1 2 clinical 1 katG S315T + nat A210R 1 2 clinical 1 katG S315T + M84T + nat A210R 1 2 clinical 1 katG S315N 1 2 clinical 1 katG S315N + kasA G269S 1 2 clinical 1 katG S315N + S140N 1 2 clinical 1 katG S315N 1 2 clinical 1 katG S315T + nat deletion 1 2 clinical 1 katG S315T + ndh frameshift 1 2 clinical 1 katG S315T + T275A 1 6, 7 clinical 2 katG S315N 2 6, 7 clinical 27 katG S315T 27 9 clinical 51 katG S315T 3 44 4 9 clinical 2 katG S315N 2 10) Brossier 2016 9 clinical 2 katG S315T 2 1 clinical 30 katG S315T 1 25 2 2 1 clinical 1 katG S315T + mabA G48V + V75A 1 1 clinical 1 katG S315T + ahpC I38M 1 1 clinical 12 katG S315T + kasA G269S 2 9 1 1 clinical 1 katG S315T + ndh L50V 1 1 clinical 2 katG S315T + ndh 18A + nat G207R 1 1 1 clinical 1 katG S315T + oxyR-ahpC c-54t 1 1 clinical 1 katG S315T + ndh V18A + G207R + mshA frameshift 1 1 clinical 1 katG S315N + nat A103G 1 1 clinical 1 katG S315T + kasA G269S + mshA A362V 1 1 clinical 2 katG S315T + kasA G269S 1 1 1 clinical 1 katG S315T + kasA A321T 1 2) Vincent 2012 INH MIC [mg/L] 1) Jagielski 2015 & Jagielski 2013 & Jagielski 2014 9) Brossier 2009 & Brossier 2006 7) Alame-Emane 2015 4) Lempens 2018 12) Farhat 2019 Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.2 0.25 0.5 0.8 1 1.6 2.5 3.2 5 6.4 10 12.8 7) Alame-Emane 2015 6, 7 clinical 1 mabA-inhA t-8c 1 2) Vincent 2012 2 clinical 1 mabA-inhA g-9a 1 2) Vincent 2012 2 clinical 18 mabA-inhA c-15t 11 3 1 3 4) Lempens 2018 2 clinical 1 mabA-inhA c-15t + kasA G312S 1 4) Lempens 2018 2 clinical 1 katG T251K + mabA-inhA c-15t 1 4) Lempens 2018 2 clinical 1 katG G127P + mabA-inhA c-15t 1 4) Lempens 2018 2 clinical 6 mabA-inhA c-15t 6 12) Farhat 2019 2 clinical 6 mabA-inhA c-15t 5 1 1) Jagielski 2015 & 1 clinical 1 mabA-inhA c-15t 1 1) Jagielski 2015 & 1 clinical 1 katG W91R + mabA-inhA c-15t 1 1) Jagielski 2015 & 1 clinical 1 katG P131Q + mabA-inhA c-15t 1 1) Jagielski 2015 & 1 clinical 1 katG M126I + mabA-inhA c-15t 1 3) Lee 2000 3 clinical 14 mabA-inhA c-15t 11 3 7) Alame-Emane 2015 6, 7 clinical 15 mabA-inhA c-15t 14 1 9) Brossier 2009 & 9 clinical 14 mabA-inhA c-15t 11 3 4) Lempens 2018 2 clinical 1 katG L458R + mabA-inhA g-17t 1 INH MIC [mg/L] 16 inhA coding mutants Four studies reported MIC data for 10 clinical isolates with inhA coding mutations and no co-occurring katG S315 or inhA promoter mutations (Table 8). Where it could be assessed, 4 (57%, 95% CI 18-90%) of these mutated isolates were phenotypically INH-R based on the current INH CC. The three isolates with an inhA S94A mutation that tested phenotypically INH-S had an MIC at the CC. These mutations have been previously shown to result only in modest MIC increases.31 By contrast, the inhA I194T mutation correlated with MICs above the CLSI CB. Table 8. INH MICs for clinical inhA coding mutants on LJ. The red line denotes the current WHO CC for INH DST on LJ (0.2 mg/L), whereas the blue lines denote the CLSI CC and CB (0.25 and 1 mg/L). Notable limitations: studies I2 and I12 were conducted in the same laboratory. oxyR-ahpC intergenic region mutants Only 2 studies were identified that reported MIC data for 8 clinical isolates with oxyR-ahpC intergenic region mutations and no co-occurring katG S315, inhA coding or inhA promoter mutations (Table 9). Where it could be assessed, 6 (86%, 95% CI 42-100%) of these mutated isolates were phenotypically INH-R based on the current INH CC. Table 9. INH MICs for clinical oxyR-ahpC intergenic region mutants on LJ. The red line denotes the current WHO CC for INH DST on LJ (0.2 mg/L), whereas the blue lines denote the CLSI CC and CB (0.25 and 1 mg/L). Double mutants katG S315 and inhA promoter double mutants Six studies reported MIC data for 43 clinical isolates with katG S315 and inhA promoter double mutations (Table 10). Based on the current INH CC, all (100%, 95% CI 92-100%) of these mutated isolates were phenotypically INH-R, with MICs from 1->25 mg/L, slightly higher than those MIC distributions reported for the single mutants (Table 6 and Table 7). katG S315 and inhA coding double mutants Only one study reported MIC data for 1 clinical isolate with a katG S315 and inhA coding mutation (Table 10). The isolate was INH-R, with an MIC of 5 mg/L. 31 Vilchèze C, Wang F, Arai M, et al. Transfer of a point mutation in Mycobacterium tuberculosis inhA resolves the target of isoniazid. Nat Med. 2006;12(9):1027-9. doi:10.1038/nm1466. Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.2 0.25 0.5 0.8 1 1.6 2.5 3.2 5 4) Lempens 2018 2 clinical 1 inhA S94A 1 4) Lempens 2018 2 clinical 1 katG M126I + inhA S94A 1 12) Farhat 2019 2 clinical 2 inhA S94A 2 9) Brossier 2009 & 9 clinical 3 inhA S94A 3 4) Lempens 2018 2 clinical 1 inhA I194T 1 1) Jagielski 2015 & 1 clinical 1 inhA I194T 1 1) Jagielski 2015 & 1 clinical 1 katG K537E + inhA I194T 1 INH MIC [mg/L] Studies Lab Isolate origin Unique isolates Genotypic summary 0.2 0.25 0.5 0.8 1 1.6 2.5 3.2 5 6.4 10 12.8 20 25 40 60 64 1) Jagielski 2015 & 1 clinical 1 katG T394P + oxyR-ahpC g-48a 1 1) Jagielski 2015 & 1 clinical 1 katG R128Q + oxyR-ahpC g-48a + nat STOP127 1 1) Jagielski 2015 & 1 clinical 1 katG R128Q + oxyR-ahpC g-48a + ndh A300P + nat STOP127 1 4) Lempens 2018 2 clinical 1 oxyR-ahpC g-48a 1 4) Lempens 2018 2 clinical 1 katG D612G + 970_971delACinsCT + T324L + oxyR-ahpC c-52t 1 1) Jagielski 2015 & 1 clinical 1 katG STOP46 + A235G + oxyR-ahpC c-57t 1 4) Lempens 2018 2 clinical 1 katG Y229C + oxyR-ahpC c-81t 1 1) Jagielski 2015 & 1 clinical 1 oxyR-ahpC tca-82/-80atc + ndh S281S 1 INH MIC [mg/L] 17 inhA promoter and inhA coding double mutants Four studies reported MIC data for 12 clinical isolates with inhA promoter and inhA coding mutations (Table 10). Based on the current INH CC, all (100%, 95% CI 63-100%) of these mutated isolates were phenotypically INH-R. The reported MICs were in line with the MICs reported for the single mutants tested on LJ (Table 7 and Table 8). Table 10. INH MICs for clinical katG S315, inhA promoter and inhA coding double mutants on LJ. The red line denotes the current WHO CC for INH DST on LJ (0.2 mg/L), whereas the blue lines denote the CLSI CC and CB (0.25 and 1 mg/L). Notable limitations: studies I2, I4 and I12 were conducted in the same laboratory and study I7 presented data for isolates only characterized by LPA and real-time PCR melting curve analysis. 2.A.1.3 Conclusion for INH CC for LJ Owing to the limited quantity and quality of MICs for pWT isolates, it was not possible to assess whether the current CC of 0.2 mg/L corresponds to the epidemiological cut-off (ECOFF) for INH. However, given that the CC appeared to adequality detect known resistance mechanisms, with the exception of inhA S94A, 0.2 mg/L was reaffirmed. Some experts noted that 0.25 mg/L should have been adopted in accordance with the standards of the International Organization for Standardization (ISO) and the CLSI CC. Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.2 0.25 0.5 0.8 1 1.6 2.5 3.2 5 6.4 10 12.8 19.2 25 30 9) Brossier 2009 & 9 clinical 1 katG S315T + mabA-inhA t-8c 1 1) Jagielski 2015 & 1 clinical 1 katG S315T + mabA-inhA t-8c 1 2) Vincent 2012 2 clinical 7 katG S315T + mabA-inhA c-15t 7 2) Vincent 2012 2 clinical 1 katG S315T + mabA-inhA c-15t 1 4) Lempens 2018 2 clinical 12 katG S315T + mabA-inhA c-15t 1 11 7) Alame-Emane 2015 6, 7 clinical 5 katG S315T + mabA-inhA c-15t 1 4 9) Brossier 2009 & 9 clinical 10 katG S315T + mabA-inhA c-15t 3 7 1) Jagielski 2015 & 1 clinical 1 katG S315T + mabA-inhA c-15t 1 12) Farhat 2019 2 clinical 2 katG S315T + mabA-inhA c-15t 2 4) Lempens 2018 2 clinical 3 katG S315T + mabA-inhA g-47a + nat G78D 3 1) Jagielski 2015 & 1 clinical 1 katG S315T + inhA K57R + mshA G446S 1 4) Lempens 2018 2 clinical 1 katG P367L + mabA-inhA c-15t + inhA I16T + kasA G269S 1 4) Lempens 2018 2 clinical 1 mabA-inhA c-15t + inhA S94A 1 4) Lempens 2018 2 clinical 1 mabA-inhA c-15t + inhA S94A + kasA G312S 1 4) Lempens 2018 2 clinical 1 mabA-inhA c-15t + inhA S94A + ndh V317A + fabG1 E102D 1 12) Farhat 2019 2 clinical 2 mabA-inhA c-15t + inhA S94A 1 1 9) Brossier 2009 & 9 clinical 4 mabA-inhA c-15t + inhA S94A 4 1) Jagielski 2015 & 1 clinical 1 mabA-inhA c-15t + inhA S94A 1 1) Jagielski 2015 & 1 clinical 1 mabA-inhA c-15t + inhA I194T 1 INH MIC [mg/L] 18 2.A.2 INH MIC data on 7H10 2.A.2.1 INH MICs for pWT isolates on 7H10 Eighteen studies were identified that reported INH MIC data for the pWT population on 7H10 (Table 11). Five of these studies were conducted in the same laboratories, and 14 of these studies featured MIC data for 10 or more pWT isolates. Given that not all studies sequenced all INH resistance genes and many studies only tested a few concentrations, it was difficult to assess the shapes of many of the MIC distributions. Considering that the modes of most gWT distributions were between 0.05 and 0.125 mg/L, and that the MICs reported for the H37Rv control strain were 0.06-0.2 mg/L, 0.2 mg/L or the corresponding ISO concentration of 0.25 mg/L likely represents the tentative ECOFF for 7H10. Table 11. INH MICs for pWT isolates on 7H10. The green line denotes the current WHO and CLSI CC for INH DST on 7H10 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). Notable limitations: studies I21, I22 and I23, and studies I33 and I27 were conducted in the same laboratory; study I28 presented data for isolates only characterized by LPA; and studies I17 and I33 presented data for mostly INH-R isolates. 2.A.2.2 INH MICs for mutated isolates on 7H10 katG S315 mutants Twelve studies reported 1113 MICs for isolates with katG S315 mutations and no co-occurring inhA promoter or coding mutations (Table 12). Based on the current INH CC, 1085 (97%, 95% CI 96-98%) of these mutated isolates were phenotypically INH-R. Most isolates that tested INH-S at the current CC were from one of two studies by Farhat et al. or Karunaratne et al., which reported lower MICs for mutants than reported by other studies. Although many studies only tested a few concentrations in a non-standard dilution series to establish MICs, the modes of the distributions appeared to be 5-10 mg/L, when they could be defined (i.e. above the CLSI CB). Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic summary 0.0125 0.016 0.025 0.03 0.05 0.06 0.1 0.12 0.15 0.2 0.25 0.4 0.5 0.613 0.8 1 2 3 4 5 8 10 16 20 32 64 128 19 2 2 H37Rv 2 19 clinical 13 26 Mix of first-line resistance profiles gWT 4 2 12 4 4 24 1 4 H37Rv ATCC27294 1 3 24 1 1 BTB 08-049 1 24 clinical 106 106 1 43 35 1 1 1 2 9 8 3 1 1 29 1 2 H37Rv ATCC27294 2 29 1 2 Erdman 1 1 29 clinical 55 57 Mix of first-line resistance profiles gWT 15 17 18 7 11 1 2 H37Rv ATCC27294 2 11 clinical 10 20 1 19 26 1 1 H37Rv 1 26 clinical 11 11 Mix of first-line resistance profiles gWT 3 7 1 10 1 1 H37Rv ATCC25618 1 10 1 1 ATCC35822 1 10 36 36 gWT 30 6 10 12 12 INH-S 12 26) Abe 2008 21 clinical 15 15 Mostly INH-S gWT 2 4 9 27 clinical 42 42 gWT 36 1 5 22 17 17 gWT 1 1 3 1 2 9 28 203 203 gWT 24 100 15 11 7 4 3 2 1 36 23 clinical 1 13 H37Rv ATCC27294 11 2 23 46 46 MTBDRplus WT gWT 7 29 5 3 2 27) van Klingeren 2007 22 clinical 7956 7956 Mostly INH-S 6985 374 159 48 32 358 25) Gali 2006 20 clinical 7 7 Mix of first-line resistance profiles gWT 3 2 1 1 19) Pretorius 1995 16 1 1 H37Rv 1 15) Campbell 2011 & 12 clinical 122 122 gWT 102 5 5 10 17) Cavusoglu 2006 14 clinical 9 9 Mostly RIF-R gWT 6 1 2 21) DeCoster 2005 18 clinical 26 26 Mix of first-line resistance profiles 7 9 6 4 22) Moore 1999 18 clinical 17 17 Mix of first-line resistance profiles 6 7 2 2 23) Kirk 1998 18 clinical 35 35 Mix of first-line resistance profiles 17 9 4 5 31) Karunaratne 2018 25 1 1 H37Rv 1 28) Wedajo 2014 Mostly MDR INH MIC [mg/L] 24) Rancoita 2018 29) Schön 2009 34) Gygli 2019 14) de Steenwinkel 2012, de Steenwinkel 2012 & 32) Pholwat 2011 13) Alonso 2013 33) Farhat 2019 19 Table 12. INH MICs for clinical katG S315 mutants on 7H10. The green line denotes the current WHO and CLSI CC for INH DST on 7H10 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). Notable limitations: study I28 presented data for isolates only characterized by LPA; studies I33 and I31 reported lower MICs for mutants than reported for other studies. inhA promoter mutants Nine studies reported 257 MICs for clinical isolates with inhA promoter mutations and no co-occurring katG S315 or inhA coding mutations (Table 13). Where it could be assessed, 238 (95%, 95% CI 92-98%) of these mutated isolates were phenotypically INH-R based on the current INH CC. Most isolates that tested INH-S at the current CC were from one of two studies by Farhat et al. or Karunaratne et al., which reported lower MICs for mutants than reported by other studies. Excluding these studies and the study by Pholwat et al., for which MICs were higher than reported for other studies, the modes of the distributions appeared to be 0.25-1 mg/L, when they could be defined (i.e. below the CLSI CB). Table 13. INH MICs for clinical inhA promoter mutants on 7H10. The green line denotes the current WHO and CLSI CC for INH DST on 7H10 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). Notable limitations: study I28 presented data for isolates only characterized by LPA; studies I33 and I31 reported lower MICs for mutants than reported for other studies. Studies Isolate origin Unique isolates Total MICs Genotypic summary 0.05 0.1 0.2 0.4 0.5 0.61 0.8 1 1.6 2 3 3.2 4 5 6.4 8 10 15 16 20 32 35 24) Rancoita 2018 clinical 7 14 katG S315T 6 8 4 4 katG S315N 4 1 1 katG S315G 1 2 2 katG S315R 2 531 531 katG S315T 2 11 2 6 4 3 10 32 27 434 1 1 katG S315N 1 83 83 katG S315T 7 5 47 19 2 3 2 2 katG S315N 2 46 46 katG S315T 1 1 2 42 28) Wedajo 2014 clinical 16 16 katG MUT inhA WT 16 32) Pholwat 2011 clinical 22 22 katG S315T 2 3 2 1 14 34) Gygli 2019 clinical 43 49 katG S315T 1 2 15 18 2 5 6 4 4 katG S315T 3 1 1 1 katG S315N 1 16 16 katG S315T 1 1 5 4 5 11 11 katG S315T 1 9 1 1 1 katG S315R 1 1 1 katG S315N 1 1 1 katG S315T + A234G 1 34 34 katG S315T 8 14 12 46 46 katG S315T 3 18 25 3 3 katG S315T 3 2 2 katG S315N 1 1 1 1 katG S315I 1 151 151 katG S315T 1 48 102 2 2 katG S315T + I335V 2 7 7 katG S315T 7 1 1 katG S315R 1 17) Cavusoglu 2006 clinical 27 27 katG S315T 27 15 15 katG S315T 1 4 4 1 1 4 1 1 katG S315N 1 1 1 katG S315T + S275A 1 1 1 katG S315N + G212D + ahpC C37T 1 1 1 katG S315T + oxyR-ahpC c-52t 1 1 1 katG S315T + oxyR-ahpC g-88a 1 1 1 katG S315T + ahpC T39S 1 8 8 katG S315T + ahpC C37T 5 2 1 3 3 katG S315N + ahpC C37T 3 1 1 katG S315R + ahpC C37T 1 31) Karunaratne 2018 clinical clinical clinical clinical 30) Otto-Knapp 2016 16) Abbadi 2009 INH MIC [mg/L] 18) van Doorn 2003 clinical 25) Gali 2006 15) Campbell 2011 & Ramirez 2010 clinical clinical33) Farhat 2019 Studies Lab Isolate origin Unique isolates Total MICs Genotypic summary 0.05 0.1 0.12 0.2 0.25 0.4 0.5 0.61 0.8 1 2 4 5 8 10 16 32 35 33) Farhat 2019 28 clinical 1 1 mabA-inhA t-8a 1 33) Farhat 2019 27 clinical 1 1 mabA-inhA t-8a 1 33) Farhat 2019 28 clinical 4 4 mabA-inhA t-8c 1 1 1 1 25) Gali 2006 20 clinical 2 2 mabA-inhA t-8c 2 32) Pholwat 2011 26 clinical 1 1 mabA-inhA t-8c 1 33) Farhat 2019 28 clinical 1 1 mabA-inhA t-8g 1 33) Farhat 2019 27 clinical 1 1 mabA-inhA t-8g 1 24) Rancoita 2018 19 clinical 1 2 mabA-inhA c-15t 2 26) Abe 2008 21 clinical 12 12 mabA-inhA c-15t 5 7 26) Abe 2008 21 clinical 1 1 katG W204R + mabA-inhA c-15t 1 33) Farhat 2019 28 clinical 138 138 mabA-inhA c-15t 1 4 1 12 11 9 20 8 2 70 33) Farhat 2019 22 clinical 9 9 mabA-inhA c-15t 3 1 3 1 1 34) Gygli 2019 29 clinical 14 18 mabA-inhA c-15t 2 13 3 25) Gali 2006 20 clinical 10 10 mabA-inhA c-15t 3 5 2 25) Gali 2006 20 clinical 1 1 katG Y678C + mabA-inhA c-15t 1 25) Gali 2006 20 clinical 1 1 katG D189H + mabA-inhA c-15t 1 32) Pholwat 2011 26 clinical 4 4 mabA-inhA c-15t 1 3 34) Gygli 2019 29 clinical 1 1 katG Y155S + mabA-inhA c-15t 1 15) Campbell 12 clinical 11 11 mabA-inhA c-15t 7 4 31) Karunaratne 201825 clinical 1 1 mabA-inhA c-15t 1 31) Karunaratne 201825 clinical 2 2 mabA-inhA c-15t + ahpC C37T 2 33) Farhat 2019 27 clinical 5 5 mabA-inhA c-15t 1 3 1 33) Farhat 2019 28 clinical 23 23 mabA-inhA g-17t 1 2 4 5 1 3 2 5 33) Farhat 2019 22 clinical 1 1 mabA-inhA g-17t 1 25) Gali 2006 20 clinical 1 1 mabA-inhA g-17t 1 34) Gygli 2019 29 clinical 1 1 mabA-inhA g-17t 1 33) Farhat 2019 27 clinical 3 3 mabA-inhA g-17t 1 1 1 28) Wedajo 2014 23 clinical 1 1 katG WT inhA MUT 1 INH MIC [mg/L] 20 inhA coding mutants Two studies were identified that reported 17 MICs for clinical isolates with inhA coding mutations and no co-occurring katG S315 or inhA promoter mutations (Table 14). The data were insufficient to confirm whether these mutations conferred resistance, as the data were few and conflicting even for the S94A mutations, for which a greater number of MICs were identified. Table 14. INH MICs for clinical inhA coding mutants on 7H10. The green line denotes the current WHO and CLSI CC for INH DST on 7H10 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). Notable limitations: study I33 reported lower MICs for mutants than reported for other studies. oxyR-ahpC intergenic region mutants No studies presenting MIC data for oxyR-ahpC intergenic region mutants on 7H10 were identified. Double mutants katG S315 and inhA promoter double mutants Six studies reported 92 MICs for clinical isolates with katG S315 and inhA promoter double mutations tested on 7H10 (Table 15). Based on the current INH CC, all (100%, 95% CI 96-100%) of these mutated isolates were phenotypically INH-R. Most studies tested insufficiently high concentrations to enable a comparison of the MICs for these double mutations with those of katG S315 single mutants (Table 12). katG S315 and inhA coding double mutants Three studies reported MIC data for clinical isolates with both a katG S315 and inhA coding mutation (Table 15). The 13 MICs reported for the 11 clinical isolates were all >5 mg/L. inhA promoter and inhA coding double mutants Two studies presenting MIC data for inhA promoter and inhA coding double mutants on 7H10 were identified (Table 15). Based on the current INH CC, all 50 (100%, 95% CI 93-100%) of these mutated isolates were phenotypically INH-R, with MICs ranging from 0.6 to >32 mg/L. Studies Lab Isolate origin Unique isolates Total MICs Genotypic summary 0.1 0.12 0.2 0.25 0.4 0.5 0.613 1 2 4 5 6.4 34) Gygli 2019 29 clinical 1 2 inhA I21V 1 1 33) Farhat 2019 28 clinical 13 13 inhA S94A 1 1 2 1 8 33) Farhat 2019 27 clinical 1 1 inhA S94A 1 34) Gygli 2019 29 clinical 1 1 inhA S94A 1 33) Farhat 2019 28 clinical 1 1 inhA I194T 1 INH MIC [mg/L] 21 Table 15. INH MICs for clinical katG S315, inhA promoter and inhA coding double mutants on 7H10. The green line denotes the current WHO and CLSI CC for INH DST on 7H10 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). Notable limitation: study I28 presented data for isolates only characterized by LPA; studies I33 and I31 reported lower MICs for mutants than reported for other studies. 2.A.2.3 Conclusion for INH CC for 7H10 The current CC of 0.2 mg/L likely corresponds to the tentative ECOFF for INH and was consequently reaffirmed, although some experts noted that 0.25 mg/L should have been adopted in accordance with ISO standards. Studies Isolate origin Unique isolates Total MICs Genotypic summary 0.2 0.4 0.5 0.613 0.8 1 1.6 2 3 4 5 6.4 8 10 15 16 20 32 35 34) Gygli 2019 clinical 4 4 katG S315T + mabA-inhA t-8a 1 3 33) Farhat 2019 clinical 2 2 katG S315T + mabA-inhA t-8c 2 33) Farhat 2019 clinical 7 7 katG S315T + mabA-inhA t-8c 7 34) Gygli 2019 clinical 8 8 katG S315T + mabA-inhA t-8c 1 5 2 34) Gygli 2019 clinical 1 1 katG S315T + K557N + mabA-inhA t-8c 1 33) Farhat 2019 clinical 2 2 katG S315T + mabA-inhA t-8c 2 15) Campbell 2011 & clinical 3 3 katG S315T + mabA-inhA t-8c 3 33) Farhat 2019 clinical 1 1 katG S315T + mabA-inhA t-8g 1 33) Farhat 2019 clinical 7 7 katG S315T + mabA-inhA c-15t 1 1 2 3 34) Gygli 2019 clinical 6 8 katG S315T + mabA-inhA c-15t 1 1 6 15) Campbell 2011 & clinical 20 20 katG S315T + mabA-inhA c-15t 20 15) Campbell 2011 & clinical 1 1 katG S315T + I335V + mabA-inhA c-15t 1 31) Karunaratne 2018 clinical 1 1 katG S315T + mabA-inhA c-15t + ahpC C37T 1 33) Farhat 2019 clinical 7 7 katG S315T + mabA-inhA c-15t 7 33) Farhat 2019 clinical 1 1 katG S315T + mabA-inhA g-17t 1 24) Rancoita 2018 clinical 2 4 katG S315T + mabA-inhA -34c deletion 4 31) Karunaratne 2018 clinical 7 7 katG S315T + mabA-inhA -34c deletion 7 33) Farhat 2019 clinical 1 1 katG S315T + mabA-inhA c-34t 1 34) Gygli 2019 clinical 1 1 katG S315T + mabA-inhA c-34t 1 31) Karunaratne 2018 clinical 1 1 katG S315T + mabA-inhA c-34t + oxyR-ahpC g-88a 1 24) Rancoita 2018 clinical 2 4 katG S315T + mabA-inhA c-60t 4 28) Wedajo 2014 clinical 1 1 katG MUT inhA MUT 1 33) Farhat 2019 clinical 1 1 katG S315T + inhA I21V 1 33) Farhat 2019 clinical 1 1 katG S315T + inhA I21V 1 33) Farhat 2019 clinical 1 1 katG S315T + inhA S94A 1 33) Farhat 2019 clinical 1 1 katG S315G + inhA S94A 1 33) Farhat 2019 clinical 1 1 katG S315N + inhA S94A 1 31) Karunaratne 2018 clinical 3 3 katG S315T + inhA S94A 3 33) Farhat 2019 clinical 1 1 katG S315T + inhA I194T 1 24) Rancoita 2018 clinical 2 4 katG S315T + inhA N231D 4 34) Gygli 2019 clinical 2 2 mabA-inhA c-15t + inhA I21T 2 33) Farhat 2019 clinical 32 32 mabA-inhA c-15t + inhA I21T 1 2 2 8 4 15 33) Farhat 2019 clinical 5 5 mabA-inhA c-15t + inhA I21T 1 4 34) Gygli 2019 clinical 1 1 mabA-inhA c-15t + inhA I21V 1 33) Farhat 2019 clinical 1 1 mabA-inhA c-15t + inhA I21V 1 34) Gygli 2019 clinical 1 1 mabA-inhA c-15t + inhA S94A 1 33) Farhat 2019 clinical 1 1 mabA-inhA c-15t + inhA S94A + A26T 1 34) Gygli 2019 clinical 1 1 mabA-inhA g-17t + inhA S94A 1 33) Farhat 2019 clinical 2 2 mabA-inhA g-17t + inhA S94A 2 33) Farhat 2019 clinical 2 2 mabA-inhA g-17t + inhA S94A 2 33) Farhat 2019 clinical 1 1 mabA-inhA c-15t + inhA A239V 1 33) Farhat 2019 clinical 1 1 mabA-inhA c-15t + inhA I194T 1 INH MIC [mg/L] 22 2.A.3 INH MIC data on 7H11 2.A.3.1 INH MICs for pWT isolates on 7H11 Eight studies were identified that reported INH MIC data for the pWT population on 7H11 (Table 16). Four of these studies reported MICs for at least 10 pWT isolates. The modes of most pWT MIC distributions appeared to be between 0.05 and 0.125 mg/L, suggesting that 0.2 mg/L or the corresponding ISO concentration of 0.25 mg/L likely represents the tentative ECOFF for 7H11. Table 16. INH MICs for pWT isolates on 7H11. The green line denotes the current WHO and CLSI CC for INH DST on 7H11 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). 2.A.3.2 INH MICs for mutated isolates on 7H11 katG S315 mutants Only two studies were identified that reported INH MICs for 8 clinical isolates with katG S315 mutations and no co-occurring inhA promoter or coding mutations (Table 17). Based on the current INH CC, all 8 (100%, 95% CI 63-100%) of these mutated isolates were phenotypically INH-R. Table 17. INH MICs for clinical katG S315 mutants on 7H11. The green line denotes the current WHO and CLSI CC for INH DST on 7H11 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). inhA promoter mutants Only one study reported MICs for 13 clinical isolates with inhA promoter mutations and no co- occurring katG S315 or inhA coding mutations on 7H11 (Table 18). Given that only one study was identified, no comments could be made as to whether the current CC for 7H11 adequately differentiated these mutations. Table 18. INH MICs for clinical inhA promoter mutants on 7H11. The green line denotes the current WHO and CLSI CC for INH DST on 7H11 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). Studies Isolate origin Unique isolates Total MICs Type of isolates Genotypic summary 0.025 0.05 0.06 0.1 0.12 0.2 0.25 0.4 0.5 1 2 4 8 16 32 64 102.4 1 1 H37Rv 1 clinical 10 10 INH-S 4 6 43) Lee 1987 clinical 17 17 Pre-treatment 12 5 1 1 H37Rv 1 1 2 ATCC35822 1 1 1 ATCC35838 1 1 1 ATCC35820 1 1 1 ATCC35837 1 clinical 40 40 Mostly INH-S 15 10 1 2 2 1 3 4 2 41) Rodriguez Diaz 2003 clinical 16 16 6 4 6 1 3 BCG Tokyo seed lot 3 1 14 BCG Tokyo 14 1 1 H37Rv 1 clinical 1 1 mix of first-line resistance profiles gWT 1 1 1 H37Rv ATCC27294 1 1 1 ATCC35822 1 36) Guo 2006 1 1 H37Rv 1 40) Fattorini 1999 INH MIC [mg/L] 37) Cockerill 1995 39) Rey-Jurado 2013 & Rey-Jurado 2012 38) Coban 2013 42) Shishido 2007 Studies Isolate origin Unique isolates Genotypic summary 0.2 0.25 0.4 0.5 1 2 4 5 8 16 1 katG S315T + S302R 1 1 katG S315T + M609I 1 5 katG S315T 5 1 katG S315N + A379T 1 36) Guo 2006 37) Cockerill 1995 INH MIC [mg/L] clinical clinical Studies Isolate origin Unique isolates Genotypic summary 0.2 0.4 1 5 36) Guo 2006 clinical 13 mabA-inhA c-15t 12 1 INH MIC [mg/L] 23 inhA coding mutants No studies presenting MIC data for inhA coding mutants on 7H11 were identified. oxyR-ahpC intergenic region mutants No studies presenting MIC data for oxyR-ahpC intergenic region mutants on 7H11 were identified. Double mutants katG S315 and inhA promoter double mutants Only one study was identified that reported INH MICs for 3 clinical isolates with katG S315 and inhA promoter double mutations on 7H11. Based on the current INH CC, all of these mutated isolates were phenotypically INH-R, with MICs of 5 mg/L. katG S315 and inhA coding double mutants No studies presenting MIC data for katG S315 and inhA coding double mutants on 7H11 were identified. inhA promoter and inhA coding double mutants No studies presenting MIC data for inhA promoter and inhA coding double mutants on 7H11 were identified. Table 19. INH MICs for clinical katG S315, inhA promoter and inhA coding double mutants on 7H11. The green line denotes the current WHO and CLSI CC for INH DST on 7H11 (0.2 mg/L), whereas the blue line denotes the CLSI CB (1 mg/L). 2.A.3.3 Conclusion for INH CC for 7H11 The data for 7H11 were more limited than 7H10. Nevertheless, 0.2 mg/L was reaffirmed as the current CC given that it likely corresponds to the tentative ECOFF for INH on 7H11. As for 7H10, some experts noted that the ISO concentration of 0.25 mg/L should have been adopted. Studies Isolate origin Unique isolates Genotypic summary 0.2 0.4 1 5 clinical 1 katG S315T + mabA-inhA t-8a 1 clinical 1 katG S315T + mabA-inhA g-22c 1 clinical 1 katG S315T + mabA-inhA t-8c 1 36) Guo 2006 INH MIC [mg/L] 24 2.A.4 INH MIC data in MGIT 2.A.4.1 INH MICs for pWT isolates in MGIT Seventeen studies were identified that reported INH MIC data for the pWT population by MGIT (Table 20). Over half of these studies were conducted in overlapping laboratories. Ten of these studies reported MICs for at least 10 pWT isolates. The modes of most of the reported distributions appeared to be between 0.03 and 0.06 mg/L, supporting 0.1 mg/L or the corresponding ISO concentration of 0.125 mg/L as the tentative ECOFF for MGIT. Table 20. INH MICs for pWT isolates in MGIT. The green line denotes the current WHO and CLSI CC for INH DST in MGIT (0.1 mg/L), whereas the blue line denotes the CLSI CB (0.4 mg/L). Notable limitations: studies I53, I54 and I55, studies I57 and I58, studies I46 and I59, studies I44 and I45, and studies I52 and I56 were conducted in the same laboratory; study I47 presented data for isolates characterized by a mix of LPA and pyrosequencing and studies I48 and I53 presented data mostly for INH-R isolates. 2.A.4.2 INH MICs for mutated isolates in MGIT katG S315 mutants Eleven studies reported INH MIC data for 603 clinical isolates with katG S315 mutations tested by MGIT (Table 21). Based on the current INH CCs, all (100%, 95% CI 99-100%) of these mutated isolates were phenotypically INH-R at the current CC. The modes of the various distributions ranged from 3- 10 mg/L (i.e. two dilutions above the CLSI CB). Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic summary 0.016 0.02 0.025 0.03 0.04 0.05 0.06 0.1 0.25 0.4 0.5 1 3 4 8 10 16 60) Colangeli 2018 46 clinical 25 25 Validation cohort 6 5 3 3 6 1 1 36 clinical 1 1 H37Rv ATCC27294 1 36 1 3 H37Rv ATCC27294 3 36 14 14 INH-S gWT 7 7 24 1 3 H37Rv ATCC27294 3 39 1 2 H37Rv ATCC27294 2 24 clinical 26 26 non-MDR gWT 17 8 1 39 1 2 H37Rv 2 39 clinical 22 22 gWT 14 8 19 1 1 H37Rv ATCC27294 1 19 clinical 4 4 gWT 3 1 53) Tessema 2017 36 clinical 29 29 Mostly MDR gWT 11 11 1 1 3 2 41 1 1 H37Ra 1 41 30 30 INH-S gWT 5 25 16 94 94 5 60 21 3 3 2 16 23 23 pre-treatment (paired) 1 14 6 2 16 24 24 post-treatment (paired) 1 15 6 2 52) Chigutsa 2015 45 clinical 54 54 1 39 12 2 56) Rockwood 2017 45 clinical 100 100 Xpert RIF-S 5 63 24 6 1 1 48) Torres 2015 42 clinical 2 2 Mostly INH-R gWT 1 1 35 1 1 H37Rv ATCC27294 1 35 clinical 6 6 INH-S gWT 6 45) Machado 2013 35 1 1 H37Rv ATCC27294 1 49) Rueda 2015 43 clinical 4 4 INH-S gWT 4 44 1 1 H37Rv 1 44 1 1 BCG 1 44 1 1 AN5 1 44 seals 7 7 3 4 55) Andres 2014 & 36 clinical 12 12 gWT 4 5 1 2 29 1 1 H37Rv ATCC27294 1 29 1 1 Erdman 1 29 clinical 8 8 mix of first-line resistance profiles gWT 8 57) Niward 2018 51) Bernardelli 2004 INH MIC [mg/L] 44) Machado 2018 and Machado 54) Heyckendorf 2018 & Sturegård 2015 47) Kambli 2015 50) Rockwood 2017 58) Groenheit (unpublished) 61) Gygli 2019 59) Ghodousi 2019 clinical clinical 25 Table 21. INH MICs for clinical katG S315 mutants in MGIT. The green line denotes the current WHO and CLSI CC for INH DST in MGIT (0.1 mg/L), whereas the blue line denotes the CLSI CB (0.4 mg/L). Notable limitations: studies I46, I53, I54, and I55, studies I46, I44 and I45, studies I46 and I58, and studies I46 and I59 were conducted in the same laboratory and study I47 presented data for isolates characterized by a mix of LPA and pyrosequencing. inhA promoter mutants Nine studies reported INH MIC data for 73 clinical isolates with inhA promoter mutations (Table 22). Of these, only two (3% (95% CI, 0-10%)) were INH-S at the current CC. Based on the current INH CCs, 71 (97%, 95% CI 90-100%) of these mutated isolates were phenotypically INH-R at the current CC. The current CC appeared to be sufficient to differentiate the vast majority of these mutants. Assuming equivalence between 0.4 and 0.5 mg/L, the current CLSI CB appears to divide the MIC distribution for inhA promoter mutants, which was not the case for LJ (Table 7) and 7H10 (Table 13), although more data are needed to support this observation. Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.4 0.5 1 2 3 4 5 8 10 16 20 36 2 katG S315T 1 1 36 16 katG S315T 9 6 1 35 7 katG S315T 3 3 1 35 1 katG S315N 1 45) Machado 2013 and 35 1 katG S315T 1 9, 19, 22, 35-40 42 katG S315T 14 26 2 9, 19, 22, 35-40 57 katG S315T 1 25 27 4 9, 19, 22, 35-40 2 katG S315N 1 1 47) Kambli 2015 41 50 katG S315T 10 35 5 43 4 katG S315T 3 1 43 7 katG S315T + I248M + mshA I460R 6 1 43 1 katG S315T + ndh V18A 1 43 1 katG S315T + ndh N316K 1 36 22 katG S315T 11 7 4 36 1 katG S315G 1 36 1 katG S315N 1 39 70 katG S315T 14 1 40 15 39 1 katG S315T + I317V 1 39 1 katG S315T + T677I 1 39 1 katG S315T + F657L 1 61) Gygli 2019 29 clinical 23 katG S315T 8 13 2 36 2 katG S315T + mshA N69S 1 1 36 1 katG S315T + mshA G106V + ndh A209V 1 36 3 katG S315T + N493S 3 36 1 katG S315G + ndh deletion 1 36 1 katG S315T + ndh L221R 1 36 4 katG S315T + ndh V18A 3 1 36 1 katG S315T + ahpC D73H 1 36 2 katG S315T + oxyR-ahpC g-48a 1 1 36 1 katG S315T + oxyR-ahpC c-52a 1 36 4 katG S315T + oxyR-ahpC c-52t 3 1 36 1 katG S315T + oxyR-ahpC c-57t 1 36 1 katG S315N + oxyR-ahpC c-72t 1 36 1 katG S315T + oxyR-ahpC g-105a 1 36 1 katG T275A + S315T + oxyR-ahpC g-142a 1 36 3 katG S315T + oxyR-ahpC g-142a 3 36 2 katG S315T + oxyR-ahpC t-519c 1 1 36 2 katG S315T + oxyR-ahpC c-581g 2 36 2 katG S315N + I317V 1 1 36 1 katG S315T + I335V 1 36 1 katG D448A + S315T 1 36 1 katG S315T + T677P 1 36 2 katG S315T + G712A 2 36 1 katG S315N 1 36 226 katG S315T 2 185 38 1 19 26 katG S315T 8 16 2 19 1 katG S315T + oxyR-ahpC g-47gt insertion 1 59) Ghodousi 2019 clinical 53) Tessema 2017 clinical 46) Cambau 2015 49) Rueda 2015 55) Andres 2014 & unpublished data clinical clinical clinical 58) Groenheit (unpublished) INH MIC [mg/L] clinical 44) Machado 2018 and Machado (unpublished) 54) Heyckendorf 2018 & Sturegård 2015 clinical clinical 26 Table 22. INH MICs for clinical inhA promoter mutants in MGIT. The green line denotes the current WHO and CLSI CC for INH DST in MGIT (0.1 mg/L), whereas the blue line denotes the CLSI CB (0.4 mg/L). Notable limitations: studies I46 and I58, studies I46, I53 and I54, studies I44 and I46, and studies I46 and I59 were conducted in the same laboratory and study I47 presented data for isolates characterized by a mix of LPA and pyrosequencing. inhA coding mutants Only two studies were identified that presented MIC data for inhA coding mutants without co- occurring katG S315 or inhA promoter mutations by MGIT (Table 23). MIC data were only available for 2 isolates, including 1 inhA S94A mutant with an MIC of 0.4 mg/L. The other inhA coding mutant, with an I21V mutation, had an MICs <0.1 mg/L, below the current CC. Table 23. INH MICs for clinical inhA coding mutants in MGIT. The green line denotes the current WHO and CLSI CC for INH DST in MGIT (0.1 mg/L), whereas the blue line denotes the CLSI CB (0.4 mg/L). oxyR-ahpC intergenic region mutants Two studies were identified that presented MIC data for oxyR-ahpC intergenic region mutants without co-occurring katG S315, inhA promoter, or inhA coding mutations by MGIT (Table 24). These studies were conducted in the same laboratory and presented MIC data for 7 clinical mutants. Six of the 7 mutants (86%, 95% CI 42-100%) had MICs above the current CC. Table 24. INH MICs for clinical oxyR-ahpC intergenic region mutants in MGIT. The green line denotes the current WHO and CLSI CC for INH DST in MGIT (0.1 mg/L), whereas the blue line denotes the CLSI CB (0.4 mg/L). Notable limitations: both studies were conducted in the same laboratory. Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.12 0.25 0.4 0.5 1 2 3 4 8 10 16 64 102.4 58) Groenheit 39 clinical 1 katG G285D + mabA-inhA t-8c 1 47) Kambli 2015 41 16 mabA-inhA c-15t 12 4 53) Tessema 2017 36 clinical 1 mabA-inhA c-15t + fabG1 M98V 1 53) Tessema 2017 36 clinical 1 mabA-inhA c-15t + ndh K78N 1 53) Tessema 2017 36 clinical 2 oxyR-ahpC g-142a + mabA-inhA c-15t 1 1 53) Tessema 2017 36 clinical 11 mabA-inhA c-15t 9 1 1 54) Heyckendorf 36 clinical 1 mabA-inhA c-15t 1 59) Ghodousi 2019 19 clinical 6 mabA-inhA c-15t 2 2 1 1 44) Machado 2018 35 clinical 1 mabA-inhA c-15t 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 3 mabA-inhA c-15t 2 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 3 mabA-inhA c-15t 3 49) Rueda 2015 43 clinical 1 katG V442G + mabA-inhA c-15t 1 58) Groenheit 39 clinical 1 mabA-inhA c-15t 1 61) Gygli 2019 29 clinical 11 mabA-inhA c-15t 11 61) Gygli 2019 29 clinical 1 katG Y155S + mabA-inhA c-15t 1 53) Tessema 2017 36 clinical 1 katG A109T + mabA-inhA c-15t + ahpC P44R 1 53) Tessema 2017 36 clinical 1 katG A130E + mabA-inhA c-15t 1 53) Tessema 2017 36 clinical 1 katG R249H + mabA-inhA c-15t 1 53) Tessema 2017 36 clinical 1 katG insertion + mabA-inhA c-15t 1 59) Ghodousi 2019 19 clinical 5 katG deletion + mabA-inhA c-15t 5 59) Ghodousi 2019 19 clinical 1 katG T271A + mabA-inhA c-15t 1 59) Ghodousi 2019 19 clinical 1 katG insertion + mabA-inhA c-15t 1 59) Ghodousi 2019 19 clinical 1 katG A172V + mabA-inhA c-15t + oxyR-ahpC g-51a 1 53) Tessema 2017 36 clinical 1 mabA-inhA g-17t 1 INH MIC [mg/L] Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.4 1 3 10 61) Gygli 2019 29 clinical 1 inhA I21V 1 53) Tessema 2017 36 clinical 1 inhA S94A 1 INH MIC [mg/L] Studies Lab Isolate origin Unique isolates Genotypic summary 0.05 0.1 0.4 1 3 10 12.8 55) Andres 2014 & 36 clinical 1 oxyR-ahpC c-15t 1 53) Tessema 2017 36 clinical 1 katG I87W + oxyR-ahpC c-52t 1 53) Tessema 2017 36 clinical 1 katG W191R + oxyR-ahpC c-72t 1 55) Andres 2014 & 36 clinical 1 oxyR-ahpC g-115a 1 55) Andres 2014 & 36 clinical 1 oxyR-ahpC c-121t 1 53) Tessema 2017 36 clinical 1 oxyR-ahpC g-142a 1 53) Tessema 2017 36 clinical 1 oxyR-ahpC g-552a 1 INH MIC [mg/L] 27 Double mutants katG S315 and inhA promoter double mutants Nine studies were identified that reported INH MICs for 200 isolates with katG S315 and inhA promoter double mutations by MGIT (Table 25). Based on the current INH CC, all (100%, 95% CI 98- 100%) of these mutated isolates were phenotypically INH-R, with MICs >3 mg/L, far above the current CC. Moreover, the majority of isolates appeared to have MICs >10 mg/L, above those MICs reported for single katG S315 mutants (Table 21). katG S315 and inhA coding double mutants Five studies were identified that reported INH MICs for six clinical isolates with katG S315 and inhA coding mutations by MGIT (Table 25). Based on the current INH CC, all of these mutated isolates were phenotypically INH-R, with MICs >3 mg/L. inhA promoter and inhA coding double mutants Six studies were identified that reported INH MICs for 67 clinical isolates with inhA promoter and inhA coding mutations by MGIT (Table 25). Based on the current INH CC, all of these mutated isolates were phenotypically INH-R, with MICs >0.5 mg/L. Table 25. INH MICs for clinical katG S315, inhA promoter and inhA coding double mutants in MGIT. The green line denotes the current WHO and CLSI CC for INH DST in MGIT (0.1 mg/L), whereas the blue line denotes the CLSI CB (0.4 mg/L). Notable limitations: studies I46, I53, I54 and I55, studies I46 and I59, studies I44, I45 and I46, and studies I46 and I58 were conducted in the same laboratory and study I47 presented data for isolates characterized by a mix of LPA and pyrosequencing. Studies Lab Isolate origin Unique isolates Genotypic summary 0.1 0.4 1 1.25 2 2.5 3 4 5 7.5 8 10 15 16 24 25 32 35 48 64 55) Andres 2014 & 36 clinical 1 katG S315T + mabA-inhA t-8a 1 53) Tessema 2017 36 clinical 4 katG S315T + mabA-inhA t-8a 4 59) Ghodousi 2019 19 clinical 1 katG S315T + mabA-inhA t-8a 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 1 katG S315T + mabA-inhA t-8c 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 1 katG S315T + mabA-inhA t-8c 1 47) Kambli 2015 41 4 katG S315T + mabA-inhA t-8c 3 1 49) Rueda 2015 43 clinical 1 katG S315T + mabA-inhA t-8c 1 55) Andres 2014 & 36 clinical 1 katG S315T + mabA-inhA t-8c 1 53) Tessema 2017 36 clinical 3 katG S315T + mabA-inhA t-8c 1 2 59) Ghodousi 2019 19 clinical 1 katG S315T + mabA-inhA t-8c 1 49) Rueda 2015 43 clinical 2 katG S315T + mabA-inhA t-8g 1 1 44) Machado 2018 and 35 clinical 2 katG S315T + mabA-inhA c-15t 2 46) Cambau 2015 9, 19, 22, 35-40 clinical 2 katG S315T + mabA-inhA c-15t 2 46) Cambau 2015 9, 19, 22, 35-40 clinical 2 katG S315T + mabA-inhA g-17t 2 46) Cambau 2015 9, 19, 22, 35-40 clinical 4 katG S315T + mabA-inhA c-15t 4 47) Kambli 2015 41 20 katG S315T + mabA-inhA c-15t 20 49) Rueda 2015 43 clinical 1 katG S315G + mabA-inhA c-15t 1 49) Rueda 2015 43 clinical 2 katG S315T + mabA-inhA c-15t 2 55) Andres 2014 & 36 clinical 11 katG S315T + mabA-inhA c-15t 1 10 58) Groenheit 39 clinical 6 katG S315T + mabA-inhA c-15t 6 61) Gygli 2019 29 clinical 5 katG S315T + mabA-inhA c-15t 1 4 61) Gygli 2019 29 clinical 1 katG S315T + K155N+ mabA-inhA c-15t 1 53) Tessema 2017 36 clinical 1 katG S315T + mabA-inhA g-17t + oxyR-ahpC g-142a 1 53) Tessema 2017 36 clinical 107 katG S315T + mabA-inhA c-15t 3 104 59) Ghodousi 2019 19 clinical 14 katG S315T + mabA-inhA c-15t 1 2 5 6 59) Ghodousi 2019 19 clinical 1 katG S315T + mabA-inhA -34c deletion 1 59) Ghodousi 2019 19 clinical 1 katG S315N + mabA-inhA c-5t 1 53) Tessema 2017 36 clinical 1 katG S315T + inhA I21V + oxyR-ahpC a-432c 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 1 katG S315T + inhA G40W 1 49) Rueda 2015 43 clinical 2 katG S315T + inhA S94A 2 58) Groenheit 39 clinical 1 katG S140N + S315N + inhA S94A 1 59) Ghodousi 2019 19 clinical 1 katG S315T + inhA S94A 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 1 mabA-inhA t-8c + inhA S94A 1 58) Groenheit 39 clinical 1 mabA-inhA c-15t + inhA I21T 1 59) Ghodousi 2019 19 clinical 2 mabA-inhA c-15t + inhA I21T 2 53) Tessema 2017 36 clinical 1 mabA-inhA c-15t + ndh V18A + inhA I21T 1 53) Tessema 2017 36 clinical 1 mabA-inhA c-15t + inhA I21T 1 58) Groenheit 39 clinical 1 mabA-inhA c-15t + inhA I21V 1 59) Ghodousi 2019 19 clinical 1 mabA-inhA c-15t + inhA I21V 1 58) Groenheit 39 clinical 1 mabA-inhA c-15t + inhA G40W 1 44) Machado 2018 and 35 clinical 7 mabA-inhA c-15t + inhA S94A 6 1 45) Machado 2013 and 35 10 mabA-inhA c-15t + inhA S94A 9 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 7 mabA-inhA c-15t + inhA S94A 6 1 54) Heyckendorf 2018 36 clinical 2 mabA-inhA c-15t + inhA S94A 2 58) Groenheit 39 clinical 3 mabA-inhA c-15t + inhA S94A 2 1 61) Gygli 2019 29 clinical 1 mabA-inhA c-15t + inhA S94A 1 59) Ghodousi 2019 19 clinical 1 mabA-inhA c-15t + inhA S94A 1 44) Machado 2018 and 35 clinical 7 mabA-inhA c-15t + inhA I194T 7 45) Machado 2013 and 35 6 mabA-inhA c-15t + inhA I194T 6 46) Cambau 2015 9, 19, 22, 35-40 clinical 5 mabA-inhA c-15t + inhA I194T 5 53) Tessema 2017 36 clinical 2 mabA-inhA c-15t + inhA I194T 2 58) Groenheit 39 clinical 2 mabA-inhA c-15t + inhA I194T 2 59) Ghodousi 2019 19 clinical 2 mabA-inhA c-15t + inhA I194T 1 1 46) Cambau 2015 9, 19, 22, 35-40 clinical 1 mabA-inhA c-15t + inhA I200T 1 59) Ghodousi 2019 19 clinical 1 mabA-inhA c-15t + inhA A239V 1 61) Gygli 2019 29 clinical 1 mabA-inhA g-17t + inhA S94A 1 INH MIC [mg/L] 28 2.A.4.3 Conclusion for INH CC in MGIT The current CC of 0.1 mg/L was reaffirmed as it likely corresponds to the tentative ECOFF based on the available data. ISO dilutions should be adopted for testing and future breakpoints. 29 2.3 INH conclusions and comments The quality and quantity of MIC data for all media did not meet EUCAST standards to define ECOFFs.32 Nevertheless, the current CCs were reaffirmed as they are likely close to the ECOFFs for INH and adequately identified most known resistance mechanisms. However, ISO concentrations should be used in the future for MIC testing and DST breakpoints. INH resistance is a heterogenous phenotype, but the level of resistance conferred can be predicted, to some extent, based on the combination of inhA promoter and katG S315 mutations. 33 inhA promoter mutations confer only modest MIC increases (0.25-2 mg/L in MGIT) and, consequently, are at risk of being misclassified as susceptible because of the inherent technical variation in pDST.34 By contrast, katG S315 mutations correlate with marked MIC increases (4-16 mg/L in MGIT) and test reliably resistant at the current CC.34 The effects of the two mechanisms are likely additive, which means that double mutants typically have higher MICs than single mutants (8-64 mg/L in MGIT). 33,34 This is reflected in the time to sputum culture conversion for these three groups (i.e. isolates with only inhA promoter mutations have faster conversion times than katG S315 mutants, which, in turn, respond better than double mutants).35 Finally, katG loss-of-function mutations, which can sometimes be inferred with the Hain LPA, typically correlate with very high INH MICs.34,36,37 The TEG considered clinical trial data from the AIDS Clinical Trials Group A5312 that provided some of the strongest evidence to date that isolates with only inhA promoter mutations, and corresponding modest MIC increases, may benefit from high-dose INH therapy.38 Nevertheless, a CB was not set. As a result, INH resistance is now stratified genotypically into LLR and HLR by the Hain LPA but not by pDST.37,39 This inconsistency was highlighted by the TEG as the topic for a future meeting that will consider detailed pharmacokinetic/pharmacodynamic (PK/PD) modelling and additional clinical outcome data (e.g. from the clinical trials NCT01589497, NCT01936831 and NCT02236078, registered at ClinicalTrials.gov).35,38,40 32 European Committee on Antimicrobial Susceptibility Testing. Standard Operating Procedure. MIC distributions and the setting of epidemiological cut-off (ECOFF) values. SOP 10.1. 30 November 2019 (https://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/EUCAST_SOPs/EUCAST_SOP_10.1_MIC_distributions_a nd_epidemiological_cut-off_value__ECOFF__setting_20191130.pdf, accessed 26 July 2020). 33 Lempens P, Meehan CJ, Vandelannoote K, et al. Isoniazid resistance levels of Mycobacterium tuberculosis can largely be predicted by high-confidence resistance-conferring mutations. Sci Rep. 2018;8(1):3246. doi:10.1038/s41598-018-21378-x. 34 Ghodousi A, Tagliani E, Karunaratne E, et al. Isoniazid resistance in Mycobacterium tuberculosis is a heterogeneous phenotype composed of overlapping MIC distributions with different underlying resistance mechanisms. Antimicrob Agents Chemother. 2019;63(7):e00092-19. doi:10.1128/AAC.00092-19. 35 Click ES, Kurbatova EV, Alexander H, et al. Isoniazid and rifampin-resistance mutations associated with resistance to second-line drugs and with sputum culture conversion. J Infect Dis. 2020;221(12):2072-2082. doi:10.1093/infdis/jiaa042. 36 Ando H, Kitao T, Miyoshi-Akiyama T, Kato S, et al. Downregulation of katG expression is associated with isoniazid resistance in Mycobacterium tuberculosis. Mol Microbiol. 2011;79(6):1615-28. doi:10.1111/j.1365-2958.2011.07547.x. 37 World Health Organization. Regional Office for Europe. Interpretation guide for GenoType MTBDRplus VER 2.0 and GenoType MTBDRsl VER 2.0. A technical guidance document developed by the European Laboratory Initiative. Version 1.0. (https://openwho.org/courses/multi-drug-resistant-tb, accessed 9 May 2020). 38 Dooley KE, Miyahara S, von Groote-Bidlingmaier F, et al. Early bactericidal activity of different isoniazid doses for drug resistant TB (INHindsight): A randomized open-label clinical trial. Am J Respir Care Med. 2020;201(11):1416-1424. doi:10.1164/rccm.201910-1960OC. 39 World Health Organization. WHO operational handbook on tuberculosis. Module 3: diagnosis – rapid diagnostics for tuberculosis detection (https://apps.who.int/iris/rest/bitstreams/1284635/retrieve, accessed 30 June 2020). 40 Diacon A, Miyahara S, Dawson R, et al. Assessing whether isoniazid is essential during the first 14 days of tuberculosis therapy: a phase 2a, open-label, randomised controlled trial. Lancet Microbe. 2020;1(2):E84-E92. doi:10.1016/S2666- 5247(20)30011-2. 30 2.4 References for INH MIC studies 1. (a) Jagielski T, Bakuła Z, Roeske K, et al. Mutation profiling for detection of isoniazid resistance in Mycobacterium tuberculosis clinical isolates. J Antimicrob Chemother. 2015;70(12):3214-3221. doi:10.1093/jac/dkv253. (b) Jagielski T, Grzeszczuk M, Kamiński M, et al. Identification and analysis of mutations in the katG gene in multidrug-resistant Mycobacterium tuberculosis clinical isolates. Pneumonol Alergol Pol. 2013;81(4):298-307. (c) Jagielski T, BakuŁa Z, Roeske K, et al. Detection of mutations associated with isoniazid resistance in multidrug-resistant Mycobacterium tuberculosis clinical isolates. J Antimicrob Chemother. 2014;69(9):2369-2375. doi:10.1093/jac/dku161. 2. Vincent V, Rigouts L, Nduwamahoro E, et al. The TDR Tuberculosis Strain Bank: a resource for basic science, tool development and diagnostic services. Int J Tuberc Lung Dis. 2012;16(1):24-31. doi:10.5588/ijtld.11.0223. 3. Lee H, Cho SN, Bang HE, et al. Exclusive mutations related to isoniazid and ethionamide resistance among Mycobacterium tuberculosis isolates from Korea. Int J Tuberc Lung Dis. 2000;4(5):441-447. 4. Lempens P, Meehan CJ, Vandelannoote K, et al. Isoniazid resistance levels of Mycobacterium tuberculosis can largely be predicted by high-confidence resistance-conferring mutations. Sci Rep. 2018;8(1):3246. doi:10.1038/s41598-018-21378-x. 5. Schönfeld N, Vesenbeckh S, Rüssmann H, et al. Minimal inhibitory concentrations of first-line drugs of multidrug-resistant tuberculosis isolates. Lung India. 2012;29(4):309. doi:10.4103/0970-2113.102794. 6. Beckers B, Lang HRM, Schimke D, Lammers A. Evaluation of a bioluminescence assay for rapid antimicrobial susceptibility testing of mycobacteria. Eur J Clin Microbiol. 1985;4(6):556-561. doi:10.1007/BF02013394. 7. Alame-Emane AK, Xu P, Pierre-Audigier C, et al. Pyrazinamide resistance in Mycobacterium tuberculosis arises after rifampicin and fluoroquinolone resistance. Int J Tuberc Lung Dis. 2015;19(6):679-684. doi:10.5588/ijtld.14.0768. 8. Thai PVK, Ha DTM, Hanh NT, et al. Bacterial risk factors for treatment failure and relapse among patients with isoniazid resistant tuberculosis. BMC Infect Dis. 2018;18(1):112. doi:10.1186/s12879-018-3033-9. 9. (a) Brossier F, Veziris N, Jarlier V, Sougakoff W. Performance of MTBDRplus for detecting high/low levels of Mycobacterium tuberculosis resistance to isoniazid. Int J Tuberc Lung Dis. 2009;13(2):260-265. (b) Brossier F, Veziris N, Truffot-Pernot C, Jarlier V, Sougakoff W. Performance of the genotype MTBDR line probe assay for detection of resistance to rifampin and isoniazid in strains of Mycobacterium tuberculosis with low- and high-level resistance. J Clin Microbiol. 2006;44(10):3659-3664. doi:10.1128/JCM.01054-06. 10. Brossier F, Boudinet M, Jarlier V, Petrella S, Sougakoff W. Comparative study of enzymatic activities of new KatG mutants from low- and high-level isoniazid-resistant clinical isolates of Mycobacterium tuberculosis. Tuberculosis. 2016;100:15-24. doi:10.1016/j.tube.2016.06.002. 11. (a) Rigouts L, Gumusboga M, de Rijk WB, et al. Rifampin resistance missed in automated liquid culture system for Mycobacterium tuberculosis isolates with specific rpoB mutations. J Clin Microbiol. 2013;51(8):2641-2645. doi:10.1128/JCM.02741-12. (b) Van Deun A. Personal communication. 2018. 12. Farhat MR, Freschi L, Calderon R, et al. GWAS for quantitative resistance phenotypes in Mycobacterium tuberculosis reveals resistance genes and regulatory regions. Nat Commun. 2019;10(1):2128. doi:10.1038/s41467-019-10110-6. 13. Alonso M, Martínez-Lirola M, Palacios JJ, et al. Evaluation of the potential role of a new mutation in mabA in modifying the response of Mycobacterium tuberculosis to isoniazid. Tuberculosis. 2013;93(6):664-667. doi:10.1016/j.tube.2013.07.003. 14. (a) de Steenwinkel JEM, ten Kate MT, de Knegt GJ, et al. Drug susceptibility of Mycobacterium tuberculosis Beijing genotype and association with MDR TB. Emerg Infect Dis. 2012;18(4):660-663. doi:10.3201/eid1804.110912. (b) de Steenwinkel JEM, ten Kate MT, de Knegt GJ, et al. Consequences of noncompliance for therapy efficacy and emergence of resistance in murine tuberculosis caused by the Beijing genotype of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2012;56(9):4937-4944. doi:10.1128/AAC.00124-12. 31 (c) Bax HI, Bakker-Woudenberg IAJM, de Vogel CP, et al. The role of the time-kill kinetics assay as part of a preclinical modeling framework for assessing the activity of anti-tuberculosis drugs. Tuberculosis. 2017;105:80-85. doi:10.1016/j.tube.2017.04.010. 15. (a) Campbell PJ, Morlock GP, Sikes RD, et al. Molecular detection of mutations associated with first- and second-line drug resistance compared with conventional drug susceptibility testing of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2011;55(5):2032-2041. doi:10.1128/AAC.01550-10. (b) Ramirez M V., Cowart KC, Campbell PJ, et al. Rapid detection of multidrug-resistant Mycobacterium tuberculosis by use of real-time PCR and high-resolution melt analysis. J Clin Microbiol. 2010;48(11):4003-4009. doi:10.1128/JCM.00812-10. 16. Abbadi SH, Sameaa GA, Morlock G, Cooksey RC. Molecular identification of mutations associated with anti-tuberculosis drug resistance among strains of Mycobacterium tuberculosis. Int J Infect Dis. 2009;13(6):673-678. doi:10.1016/j.ijid.2008.10.006. 17. Cavusoglu C, Turhan A, Akinci P, Soyler I. Evaluation of the Genotype MTBDR assay for rapid detection of rifampin and isoniazid resistance in Mycobacterium tuberculosis isolates. J Clin Microbiol. 2006;44(7):2338-2342. doi:10.1128/JCM.00425-06. 18. Van Doorn HR, Claas ECJ, Templeton KE, et al. Detection of a point mutation associated with high-level isoniazid resistance in Mycobacterium tuberculosis by using real-time PCR technology with 3′-minor groove binder-DNA probes. J Clin Microbiol. 2003;41(10):4630-4635. doi:10.1128/JCM.41.10.4630- 4635.2003. 19. Pretorius GS, Van Helden PD, Sirgel F, Eisenach KD, Victor TC. Mutations in katG gene sequences in isoniazid-resistant clinical isolates of Mycobacterium tuberculosis are rare. Antimicrob Agents Chemother. 1995;39(10):2276-2281. doi:10.1128/AAC.39.10.2276. 20. Schechter MC, Bizune D, Kagei M, et al. Time to sputum culture conversion and treatment outcomes among patients with isoniazid-resistant tuberculosis in Atlanta, Georgia. Clin Infect Dis. 2017;65(11):1862-1871. doi:10.1093/cid/cix686. 21. DeCoster DJ, Vena RM, Callister SM, Schell RF. Susceptibility testing of Mycobacterium tuberculosis: Comparison of the BACTEC TB-460 method and flow cytometric assay with the proportion method. Clin Microbiol Infect. 2005;11(5):372-378. doi:10.1111/j.1469-0691.2005.01127.x. 22. Moore A V., Kirk SM, Callister SM, Mazurek GH, Schell RF. Safe determination of susceptibility of Mycobacterium tuberculosis to antimycobacterial agents by flow cytometry. J Clin Microbiol. 1999;37(3):479-483. doi:10.1128/jcm.37.3.479-483.1999. 23. Kirk SM, Schell RF, Moore A V., Callister SM, Mazurek GH. Flow cytometric testing of susceptibilities of Mycobacterium tuberculosis isolates to ethambutol, isoniazid, and rifampin in 24 hours. J Clin Microbiol. 1998;36(6):1568-1573. doi:10.1128/jcm.36.6.1568-1573.1998. 24. Rancoita PM V., Cugnata F, Gibertoni Cruz AL, et al. Validating a 14-drug microtiter plate containing bedaquiline and delamanid for large-scale research susceptibility testing of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2018;62(9):e00344-18. doi:10.1128/AAC.00344-18. 25. Galí N, Domínguez J, Blanco S, et al. Use of a mycobacteriophage-based assay for rapid assessment of susceptibilities of Mycobacterium tuberculosis isolates to isoniazid and influence of resistance level on assay performance. J Clin Microbiol. 2006;44(1):201-205. doi:10.1128/JCM.44.1.201-205.2006. 26. Abe C, Kobayashi I, Mitarai S, et al. Biological and molecular characteristics of Mycobacterium tuberculosis clinical isolates with low-level resistance to isoniazid in Japan. J Clin Microbiol. 2008;46(7):2263-2268. doi:10.1128/JCM.00561-08. 27. van Klingeren B, Dessens-Kroon M, van der Laan T, Kremer K, van Soolingen D. Drug susceptibility testing of Mycobacterium tuberculosis complex by use of a high-throughput, reproducible, absolute concentration method. J Clin Microbiol. 2007;45(8):2662-2668. doi:10.1128/JCM.00244-07. 28. Wedajo W, Schön T, Bedru A, et al. A 24-well plate assay for simultaneous testing of first and second line drugs against Mycobacterium tuberculosis in a high endemic setting. BMC Res Notes. 2014;7:512. doi:10.1186/1756-0500-7-512. 29. Schön T, Juréen P, Giske CG, et al. Evaluation of wild-type MIC distributions as a tool for determination of clinical breakpoints for Mycobacterium tuberculosis. J Antimicrob Chemother. 2009;64(4):786-793. doi:10.1093/jac/dkp262. 30. Otto-Knapp R, Vesenbeckh S, Schönfeld N, et al. Isoniazid minimal inhibitory concentrations of tuberculosis strains with katG mutation. Int J Tuberc Lung Dis. 2016;20(9):1275-1276. doi:10.5588/ijtld.16.0148. 32 31. Karunaratne GHRE, Wijesundera SS, Vidanagama D, Adikaram CP, Perera J. Significance of coexisting mutations on determination of the degree of isoniazid resistance in Mycobacterium tuberculosis strains. Microb Drug Resist. 2018;24(6):844-851. doi:10.1089/mdr.2017.0330. 32. Pholwat S, Heysell S, Stroup S, Foongladda S, Houpt E. Rapid first- and second-line drug susceptibility assay for Mycobacterium tuberculosis isolates by use of quantitative PCR. J Clin Microbiol. 2011;49(1):69-75. doi:10.1128/JCM.01500-10. 33. Farhat MR, Freschi L, Calderon R, et al. GWAS for quantitative resistance phenotypes in Mycobacterium tuberculosis reveals resistance genes and regulatory regions. Nat Commun. 2019;10(1):2128. doi:10.1038/s41467-019-10110-6. 34. Gygli SM, Keller PM, Ballif M, et al. Whole-genome sequencing for drug resistance profile prediction in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2019;63(4):e02175-18. doi:10.1128/AAC.02175-18. 35. Park S-W, Tasneen R, Converse PJ, Nuermberger EL. Immunodeficiency and intermittent dosing promote acquired rifamycin monoresistance in murine tuberculosis. Antimicrob Agents Chemother. 2017;61(11):e01502-17. doi:10.1128/AAC.01502-17. 36. Guo H, Seet Q, Denkin S, Parsons L, Zhang Y. Molecular characterization of isoniazid-resistant clinical isolates of Mycobacterium tuberculosis from the USA. J Med Microbiol. 2006;55(11):1527-1531. doi:10.1099/jmm.0.46718-0. 37. Cockerill FR, Uhl JR, Temesgen Z, et al. Rapid identification of a point mutation of the Mycobacterium tuberculosis catalase-peroxidase (katG) gene associated with isoniazid resistance. J Infect Dis. 1995;171(1):240-245. doi:10.1093/infdis/171.1.240. 38. Coban AY. Blood agar validation for susceptibility testing of isoniazid, rifampicin, ethambutol, and streptomycin to Mycobacterium tuberculosis isolates. PLoS One. 2013;8(2):e55370. doi:10.1371/journal.pone.0055370. 39. (a) Rey-Jurado E, Tudó G, de la Bellacasa JP, Espasa M, González-Martín J. In vitro effect of three-drug combinations of antituberculous agents against multidrug-resistant Mycobacterium tuberculosis isolates. Int J Antimicrob Agents. 2013;41(3):278-280. doi:10.1016/j.ijantimicag.2012.11.011. (b) Rey-Jurado E, Tudó G, Martínez JA, González-Martín J. Synergistic effect of two combinations of antituberculous drugs against Mycobacterium tuberculosis. Tuberculosis. 2012;92(3):260-263. doi:10.1016/j.tube.2012.01.005. 40. Fattorini L, Iona E, Ricc ML, et al. Activity of 16 Antimicrobial agents against drug-resistant strains of Mycobacterium tuberculosis. Microb Drug Resist. 1999;5(4):265-270. doi:10.1089/mdr.1999.5.265. 41. Rodriguez Díaz JC, Ruiz M, López M, Royo G. Synergic activity of fluoroquinolones and linezolid against Mycobacterium tuberculosis. Int J Antimicrob Agents. 2003;21(4):354-356. doi:10.1016/s0924- 8579(02)00387-4. 42. Shishido Y, Mitarai S, Otomo K, et al. Anti-tuberculosis drug susceptibility testing of Mycobacterium bovis BCG Tokyo strain. Int J Tuberc Lung Dis. 2007;11(12):1334-1338. 43. Lee C-N, Heifets LB. Determination of minimal inhibitory concentrations of antituberculosis drugs by radiometric and conventional methods. Am Rev Respir Dis. 1987;136(2):349-352. doi:10.1164/ajrccm/136.2.349. 44. Machado D, Perdigão J, Portugal I, et al. Efflux activity differentially modulates the levels of isoniazid and rifampicin resistance among multidrug resistant and monoresistant Mycobacterium tuberculosis strains. Antibiotics. 2018;7(1):18. doi:10.3390/antibiotics7010018. 45. Machado D, Perdigão J, Ramos J, et al. High-level resistance to isoniazid and ethionamide in multidrug- resistant Mycobacterium tuberculosis of the Lisboa family is associated with inhA double mutations. J Antimicrob Chemother. 2013;68(8):1728-1732. doi:10.1093/jac/dkt090. 46. Cambau E, Viveiros M, Machado D, et al. Revisiting susceptibility testing in MDR-TB by a standardized quantitative phenotypic assessment in a European multicentre study. J Antimicrob Chemother. 2015;70(3):686-696. doi:10.1093/jac/dku438. 47. Kambli P, Ajbani K, Sadani M, et al. Defining multidrug-resistant tuberculosis: Correlating GenoType MTBDRplus assay results with minimum inhibitory concentrations. Diagn Microbiol Infect Dis. 2015;82(1):49-53. doi:10.1016/j.diagmicrobio.2015.01.009. 48. Torres JN, Paul L V., Rodwell TC, et al. Novel katG mutations causing isoniazid resistance in clinical M. tuberculosis isolates. Emerg Microbes Infect. 2015;4(7):e42. doi:10.1038/emi.2015.42. 49. Rueda J, Realpe T, Mejia GI, et al. Genotypic analysis of genes associated with independent resistance and cross-resistance to isoniazid and ethionamide in Mycobacterium tuberculosis clinical isolates. Antimicrob Agents Chemother. 2015;59(12):7805-7810. doi:10.1128/AAC.01028-15. 33 50. Rockwood N, Sirgel F, Streicher E, et al. Low frequency of acquired isoniazid and rifampicin resistance in rifampicin-susceptible pulmonary tuberculosis in a setting of high HIV-1 infection and tuberculosis coprevalence. J Infect Dis. 2017;216(6):632-640. doi:10.1093/infdis/jix337. 51. Bernardelli A, Morcillo N, Loureiro J, Quse V, Davenport S. In vitro susceptibility testing of Mycobacterium tuberculosis complex strains isolated from seals to antituberculosis drugs. Biomedica. 2004;24 Supp 1:85-91. 52. Chigutsa E, Pasipanodya JG, Visser ME, et al. Impact of nonlinear interactions of pharmacokinetics and MICs on sputum bacillary kill rates as a marker of sterilizing effect in tuberculosis. Antimicrob Agents Chemother. 2015;59(1):38-45. doi:10.1128/AAC.03931-14. 53. Tessema B, Nabeta P, Valli E, et al. FIND Tuberculosis Strain Bank: A resource for researchers and developers working on tests to detect Mycobacterium tuberculosis and related drug resistance. Land GA, ed. J Clin Microbiol. 2017;55(4):1066-1073. doi:10.1128/JCM.01662-16. 54. (a) Heyckendorf J, Andres S, Köser CU, et al. What is resistance? Impact of phenotypic versus molecular drug resistance testing on therapy for multi- and extensively drug-resistant Tuberculosis. Antimicrob Agents Chemother. 2017;62(2):e01550-17. doi:10.1128/AAC.01550-17. (b) Sturegård E, Ängeby KA, Werngren J, et al. Little difference between minimum inhibitory concentrations of Mycobacterium tuberculosis wild-type organisms determined with BACTEC MGIT 960 and Middlebrook 7H10. Clin Microbiol Infect. 2015;21(2):148.e5-7. doi:10.1016/j.cmi.2014.08.021. 55. (a) Andres S, Hillemann D, Rüsch-Gerdes S, Richter E. Occurrence of rpoB mutations in isoniazid- resistant but rifampin-susceptible Mycobacterium tuberculosis isolates from Germany. Antimicrob Agents Chemother. 2014;58(1):590-592. doi:10.1128/AAC.01752-13. (b) Andres S. Personal communication. 2018. 56. Rockwood N, Pasipanodya JG, Denti P, et al. Concentration-dependent antagonism and culture conversion in pulmonary tuberculosis. Clin Infect Dis. 2017;64(10):1350-1359. doi:10.1093/cid/cix158. 57. Niward K, Davies Forsman L, Bruchfeld J, et al. Distribution of plasma concentrations of first-line anti- TB drugs and individual MICs: a prospective cohort study in a low endemic setting. J Antimicrob Chemother. 2018;73(10):2838-2845. doi:10.1093/jac/dky268. 58. Groenheit R. Personal communication. 2018. 59. Ghodousi A, Tagliani E, Karunaratne E, et al. Isoniazid resistance in Mycobacterium tuberculosis is a heterogeneous phenotype composed of overlapping MIC distributions with different underlying resistance mechanisms. Antimicrob Agents Chemother. 2019;63(7):e00092-19. doi:10.1128/AAC.00092-19. 60. Colangeli R, Jedrey H, Kim S, et al. Bacterial factors that predict relapse after tuberculosis therapy. N Engl J Med. 2018;379(9):823-833. doi:10.1056/NEJMoa1715849. 61. Gygli SM, Keller PM, Ballif M, et al. Whole-genome sequencing for drug resistance profile prediction in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2019;63(4):e02175-18. doi:10.1128/AAC.02175-18. 34 35 SECTION 3: Rifamycins 36 3.0 Rifamcyin resistance mechanisms RIF, RFB and RPT belong to the rifamycin group, and inhibit bacterial DNA-dependent RNA synthesis.41,42,43 The vast majority (~96%) of rifamycin resistance is caused by genetic changes in the RIF resistance-determining region (RRDR), a well-defined, 81-base-pair central region of the rpoB gene stretching from M. tuberculosis codons 426 to 452 (Escherichia coli numbering 507-533).44,45,46,47,48,49 This RRDR forms the RIF drug-binding pocket, and, as a result, non-synonymous changes and, more rarely, in-frame deletions or insertions in this region confer resistance to rifamycins. rpoB mutations well outside of the RRDR have also been associated with RIF resistance. These mutations have been most commonly reported at rpoB codons 170 (146 by E. coli numbering) and 491 (572 by E. coli numbering).44 The V170F mutation has been shown to confer RIF resistance in E. coli through site-directed mutagenesis studies.50 The same has been demonstrated for the I491F mutation in laboratory mutagenesis studies in E. coli.51 Importantly, transformation experiments of both of these mutations confirmed that they also result in elevated RIF MICs in M. tuberculosis.52,53 Each of these mutations was shown to be responsible for 0.5% (95% CI 0.2-1.2%) of RIF resistance respectively based on a recent WHO multi-country population-based surveillance study.54 However, the I491F (I572F) mutation accounted for more than half of RIF resistance in Eswatini even though the frequency in neighboring South Africa was <1%.55,56 41 Calvori C, Frontali L, Leoni L, Tecce G. Effect of rifamycin on protein synthesis. Nature. 1965;207(4995):417-418. doi:10.1038/207417a0. 42 Wehrli W. Rifampin: Mechanisms of action and resistance. Rev Infect Dis. 1983;5:S407-11. doi:10.2307/4453139. 43 Yang B, Koga H, Ohno H, et al. Relationship between antimycobacterial activities of rifampicin, rifabutin and KRM-1648 and rpoB mutations of Mycobacterium tuberculosis. J Antimicrob Chemother. 1998;42(5):621-628. doi:10.1093/jac/42.5.621. 44 Please refer to Section 3.2 for more details regarding the different rpoB numbering systems. 45 Telenti A, Imboden P, Marchesi F, et al. Detection of rifampicin-resistance mutations in Mycobacterium tuberculosis. Lancet. 1993;341(8846):647-650. doi:10.1016/0140-6736(93)90417-f. 46 Van Rie A, Warren R, Mshanga I, et al. Analysis for a limited number of gene codons can predict drug resistance of Mycobacterium tuberculosis in a high-incidence community. J Clin Microbiol. 2001;39(2):636-641. doi:10.1128/JCM.39.2.636-641.2001. 47 Ramaswamy S, Musser JM. Molecular genetic basis of antimicrobial agent resistance in Mycobacterium tuberculosis: 1998 update. Tuber Lung Dis. 1998;79(1):3-29. doi:10.1054/tuld.1998.0002 48 Zhang Y, Telenti A. Genetics of drug resistance in Mycobacterium tuberculosis. In: Hatfull GF, Jacobs WR, eds. Molecular Genetics of Mycobacteria. ASM Press; 2000:363 (https://library.villanova.edu/Find/Record/538689/TOC, accessed 7 June 2019). 49 World Health Organization. Technical manual for drug susceptibility testing of medicines used in the treatment of tuberculosis; 2018. (https://apps.who.int/iris/bitstream/handle/10665/275469/9789241514842-eng.pdf, accessed 7 June 2019). 50 Severinov K, Soushko M, Goldfarb A, Nikiforov V. RifR mutations in the beginning of the Escherichia coli rpoB gene. MGG Mol Gen Genet. 1994;244:120-126. doi:10.1007/BF00283512. 51 Rodríguez-Verdugo A, Gaut BS, Tenaillon O. Evolution of Escherichia coli rifampicin resistance in an antibiotic-free environment during thermal stress. BMC Evol Biol. 2013;13:50. doi:10.1186/1471-2148-13-50. 52 Siu GK, Zhang Y, Lau TC, et al. Mutations outside the rifampicin resistance-determining region associated with rifampicin resistance in Mycobacterium tuberculosis. J Antimicrob Chemother. 2011;66(4):730-733. doi:10.1093/jac/dkq519. 53 Lai LY, Hsu LY, Weng SH, et al. A glutamine insertion at codon 432 of RpoB confers rifampicin resistance in Mycobacterium tuberculosis. Front Microbiol. 2020;11:583194. doi:10.3389/fmicb.2020.583194. 54 Zignol M, Cabibbe AM, Dean AS, et al. Genetic sequencing for surveillance of drug resistance in tuberculosis in highly endemic countries: a multi-country population-based surveillance study. Lancet Infect Dis. 2018;18(6):675-683. doi:10.1016/S1473-3099(18)30073-2. doi:10.1016/S1473-3099(18)30073-2. 55 Sikhondze W, Dlamini T, Joloba Moses, et al. Xpert MTB/RIF miss es more than 50% of rifampicin resistant TB cases in Eswatini: results of the 2nd national anti-TB drug resistance survey (2017/2018). Int J Tuberc Lung Dis. 2019;23(10 Suppl 1):S585. 56 Ismail NA, Omar SV, Mvusi L, Madhi SA. Prevalence of drug-resistant tuberculosis in South Africa - Authors' reply. Lancet Infect Dis. 2018;18(8):836-837. doi:10.1016/S1473-3099(18)30422-5. 37 A recent report by Huseby et al. demonstrated that a frameshift in the RRDR of E. coli is possible, despite rpoB being an essential gene, and also confers RIF resistance through spontaneous frameshift suppression. 57 Similar mutations could, in principle, occur throughout the entire length of rpoB. Frameshifts and nonsense mutations (i.e. premature stop codons) have been described in MTBC in the literature, but it is unclear what proportion of these may be due to experimental errors and, consequently, which percentage of RIF resistance in clinical isolates is caused by these mechanisms. 3.1 Current statements and policies regarding genotypic markers of rifamycin resistance Currently, there is a lack of consensus on a list of genetic alterations that confer rifamycin resistance, and the degree of cross-resistance between the three rifamycins is unclear.58 In particular, there is a need for alignment regarding the following statements, policy recommendations and practices regarding RIF resistance: 1. Miotto et al. conducted a comprehensive analysis of the association between rpoB mutations and pDST results, which yielded 24 high-, moderate-, or minimal-confidence mutations that were associated with RIF resistance after performing a p-value correction (Table 26).59 A further 13 rpoB mutations (or combinations thereof) were associated with RIF resistance based on their nominal p-values, but were statistically indeterminate following a p-value correction. 2. For the purposes of surveillance, WHO has adopted a composite reference standard for RIF resistance.54 Specifically, the aforementioned list of 24 p-value corrected rpoB mutations from Miotto et al. were considered to be true markers for RIF resistance (i.e. the presence of any of these mutations is necessary and sufficient to confirm RIF resistance). Any susceptible pDST result for an isolate with one of these mutations was thereby corrected to resistant. 3. In the recently published WHO technical manual for TB DST, DNA sequencing of the entire rpoB gene was proposed as a reference standard for RIF resistance. Furthermore, an expert rule was established, that “[a]ny mutation (excluding silent mutations) observed in the 81bp RRDR [as defined above] hotspot region of the rpoB gene are known or assumed to be associated with rifampicin resistance” (i.e. even mutations that were statistically indeterminate in Miotto et al. or have never been reported to date within the RRDR were considered RIF-R).60 This expert rule was introduced to address the confusion surrounding the “borderline resistance” rpoB mutations, which are also referred to as “disputed”, “discordant”, “occult” or “(sub-breakpoint) low-level resistance” mutations. These account for 12% (95% CI 10-15%) of RIF resistance based on WHO surveillance data from seven countries but can be considerably more frequent in some settings 57 Huseby DL, Brandis G, Praski Alzrigat L, Hughes D. Antibiotic resistance by high-level intrinsic suppression of a frameshift mutation in an essential gene. Proc Natl Acad Sci U S A. 2020;117(6):3185-3191. doi:10.1073/pnas.1919390117. 58 Schön T, Juréen P, Chryssanthou E, et al. Rifampicin-resistant and rifabutin-susceptible Mycobacterium tuberculosis strains: a breakpoint artefact? J Antimicrob Chemother. 2013;68(9):2074-2077. doi:10.1093/jac/dkt150 59 Miotto P, Tessema B, Tagliani E, et al. A standardised method for interpreting the association between mutations and phenotypic drug resistance in Mycobacterium tuberculosis. Eur Respir J. 2017;50(6):1701354. doi:10.1183/13993003.01354- 2017. 60 World Health Organization. Technical manual for drug susceptibility testing of medicines used in the treatment of tuberculosis. (http://apps.who.int/iris/bitstream/handle/10665/275469/9789241514842-eng.pdf, accessed 17 November 2018). 38 (e.g. in Eswatini or São Paulo state, Brazil).54,55,61 These were originally defined as “disputed” given that they are more likely to test RIF-S by MGIT compared to LJ.62 However, studies differ on which exact mutations are associated with borderline RIF resistance. Based on the published literature and a review of MIC data collected as part of this systematic review, six RRDR mutations as well as I491F (I572F) were classified as borderline resistance mutations for the purposes of this report, which are highlighted in Table 26.62,63,64 In order to justify the expert rule that any RRDR mutation, apart from synonymous mutations, should be considered to be a valid marker for RIF-R, it is necessary to determine why these borderline resistance mutations are more likely to test RIF-S (see Section 3.3). 4. In the recently published WHO technical guide for the use of next-generation sequencing (NGS), an abridged review of MIC evidence for rpoB mutations from Miotto et al. was performed, though there were a few differences in the confidence grading of rare mutations.65 For example, the N437 (N518) deletion in RRDR was listed as a high-confidence resistance mutation in the NGS guide even though it actually was statistically indeterminate in Miotto et al. following p-value correction.59 5. The WHO-endorsed LPAs and Xpert MTB/RIF Ultra (Cepheid, Sunnyvale, CA, USA) cover codons immediately downstream and/or upstream of RRDR (Table 27). Because the gDST results for these assays are not always confirmed in many settings, the aforementioned expert rule for RRDR is de facto extended to these adjacent codons (e.g. 424-425 (505-506) in the case of the Hain LPA).60,66 6. Even when confirmatory genotypic testing is conducted, the interpretation of sequencing results has not been well aligned. For example, the GLI recommends that sequencing results are interpreted according to the NGS technical guide but does not mention the RRDR expert rule from the DST manual. 60,66 61 Brandao AP, Pinhata JMW, Simonsen V, et al. Transmission of Mycobacterium tuberculosis presenting unusually high discordance between genotypic and phenotypic resistance to rifampicin in an endemic tuberculosis setting. Tuberculosis (Edinb). 2020;125:102004. doi:10.1016/j.tube.2020.102004. 62 Torrea G, Ng K, Van Deun A. Variable ability of rapid tests to detect Mycobacterium tuberculosis rpoB mutations conferring phenotypically occult rifampicin resistance. Sci Rep. 2019;9(1):11826. doi:10.1038/s41598-019-48401-z. 63 Rigouts L, Gumusboga M, de Rijk WB, et al. Rifampin resistance missed in automated liquid culture system for Mycobacterium tuberculosis isolates with specific rpoB mutations. J Clin Microbiol. 2013;51(8):2641-2645. doi:10.1128/JCM.02741-12. 64 Miotto P, Cabibbe AM, Borroni E, Degano M, Cirillo DM. Role of disputed mutations in the rpoB gene in interpretation of automated liquid MGIT culture results for rifampin susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol. 2018;56(5):e01599-17. doi:10.1128/JCM.01599-17. 65 World Health Organization. The use of next-generation sequencing technologies for the detection of mutations associated with drug resistance in Mycobacterium tuberculosis complex: Technical guide; 2018 (https://apps.who.int/iris/bitstream/handle/10665/274443/WHO-CDS-TB-2018.19-eng.pdf, accessed 7 June 2019). 66 Global Laboratory Initiative. Line probe assays for drug-resistant tuberculosis detection: interpretation and reporting guide for laboratory staff and clinicians; 2019 (http://www.stoptb.org/wg/gli/assets/documents/LPA_test_web_ready.pdf, accessed 22 May 2020). 39 Table 26. Overview of confidence-graded rpoB RIF resistance mutations, including the seven borderline resistance mutations. High confidence M. tuberculosis resistance mutation (E. coli numbering) Moderate confidence M. tuberculosis resistance mutation (E. coli numbering) Minimal confidence M. tuberculosis resistance mutation (E. coli numbering) Q432K (Q513K) D435Y (D516Y) L430P (L511P) Q432L (Q513L) S441L (S522L) H445N (H526N) Q432P (Q513P) L452P (L533P) I491F (I572F) F433dupl (F514dupl) D435A (D516A) D435F (D516F) D435G (D516G) D435V (D516V) S441Q (S522Q) H445C (H526C) H445D (H526D) H445G (H526G) H445L (H526L) H445R (H526R) H445Y (H526Y) S450F (S531F) S450L (S531L) S450W (S531W) Cells with borderline resistance mutations are shaded in grey. Mutations that were statistically indeterminate after p-value correction are not listed, which is the case for H445S (H526S), the seventh borderline resistance mutation.59 40 Table 27. Probe-binding regions and rpoB mutation coverage of WHO-endorsed gDST assays. Assay rpoB probe Codons analysed or specific mutation detecteda Mutations covered as per package insert Hain GenoType MTBDRplus V267,68 WT1 424-428 (505-509)b F424L (F505L), T427A (T508A), S428T (S509T) WT2 429-432 (510-513) Q429H (Q510H), L430P (L511P) WT3 432-436 (513-517) Q432L (Q513L), Q432P (Q513P), del433-435 (del514-516) WT4 435-438 (516-519) D435V (D516V), D435Y (D516Y), del434 (del515) WT5 437-441 (518-522) del437 (del518), N437I (N518I) WT6 441-444 (522-525) S441L (S522L), S441Q (S522Q) WT7 445-448 (526-529) H445C (H526C), H445D (H526D), H445L (H526L), H445N (H526N), H445P (H526P), H445Q (H526Q), H445R (H526R), H445S (H526S), H445Y (H526Y) WT8 449-452 (530-533) S450L (S531L), S450Q (S531Q), S450W (S531W), L452P (L533P) MUT1 gac/gtc D435V (D516V) MUT2A cac/tac H445Y (H526Y) MUT2B cac/gac H445D (H526D) MUT3 tcg/ttg S450L (S531L) Nipro NTM+MDRTB II69 S1 428-433 (509-514) S2 434-439 (515-520) S3 439-444 (520-525) S4 444-449 (525-530) S5 449-454 (530-535)b R2 D435V (D516V) R4a H445Y (H526Y) R4b H445D (H526D) R5 S450L (S531L) Cepheid Xpert MTB/RIF G470 A 426-431 (507-512) Bc 430-438 (511-519) C 436-442 (517-523) D 441-447 (522-528) E 447-453 (528-534)b Cepheid Xpert Ultra70 1 426-435 (507-516) 2 433-443 (514-524) 3 441-450 (522-531) 4 446-455 (527-536)b Molbio Truenat MTB-RIF Dx Not published to date a Hain has not published which precise codons are covered by “MUT” probes. The same applies to the “R” probes of the Nipro LPA. The codons covered by the Hain “WT” or Nipro “S” probes may not be covered completely and not all mutations 67 Hain Lifescience. GenoType MTBDRplus VER 2.0. Instructions for use. IFU-304A-09. (https://www.hain- lifescience.de/include_datei/kundenmodule/packungsbeilage/download.php?id=2877, accessed 1 Nov 2020). 68 World Health Organization. Regional Office for Europe. Interpretation guide for GenoType MTBDRplus VER 2.0 and GenoType MTBDRsl VER 2.0. A technical guidance document developed by the European Laboratory Initiative. Version 1.0. (https://openwho.org/courses/multi-drug-resistant-tb, accessed 9 May 2020). 69 FIND. Report for WHO: Non-inferiority evaluation of Nipro NTM+MDRTB and Hain GenoType MTBDRplus V2 line probe assays. Version 4.1. Geneva, Switzerland; 2015 (https://www.finddx.org/wp-content/uploads/2016/04/LPA- report_noninferiority-study_oct2015.pdf, accessed 22 May 2020). 70 Chakravorty S. Personal communication. 2020. 41 in these codons, particularly if they occur at the edges of the probes, prevent binding.68 Similarly, the codons listed for Xpert and Ultra probes may only be covered partially.70 b Probes include nucleotides or codons outside of the RRDR: 426-452 (507-533).71 c Xpert G3 and earlier versions used a different probe B.72,73,74 3.2 Rifamycin MIC data stratification and current breakpoints As the RRDR was originally defined in E. coli, it is standard practice to use the numbering system of this species even when describing specific features of rpoB from other bacteria.75 For the purposes of this report, both the M. tuberculosis and E. coli numbering systems were used for reporting of rpoB mutations, with the E. coli sequence annotation noted in parentheses.76 Specifically, the annotation adopted for the genome of the M. tuberculosis H37Rv reference genome is used throughout this report (GenBank accession AL123456.3).77 The start of this annotation is six codons downstream of the one first proposed in 1994 (GenBank accession AAA21416.1), which is occasionally used in the literature.78,79 All rpoB mutations that were identified in codons 424-456 (505-537), as covered by WHO-endorsed molecular assays (Table 27), as well as codons 170 (146) and 491 (572), were recorded and reported in this analysis. MIC data were only excluded if rpoB mutations occurred in codons other than those of interest (M. tuberculosis codons 170, 424-456, and 491), or if multiple rpoB single nucleotide polymorphisms (SNPs) were reported in the same codon, as this might signify either mixtures or translation errors, or if the exact concentrations used for MIC testing of isolates or final MICs for any isolates could not be clarified after contacting the lead study authors. Synonymous rpoB mutations were considered to be gWT. MIC data were stratified based on mutation location: RRDR mutations were grouped as S450 (S531) codon mutations (because these confer large MIC increases, they can be used accordingly to assess the quality of MIC testing of a given study80); RRDR borderline resistance mutations (Table 26); other RRDR mutations; and rpoB mutations outside the RRDR. Only single mutants are included in this report, though MIC data for isolates with multiple mutations can be found in the corresponding Excel file. Similarly, rpoB indels were only recorded in the Excel files. For the 71 World Health Organization. Technical manual for drug susceptibility testing of medicines used in the treatment of tuberculosis. (http://apps.who.int/iris/bitstream/handle/10665/275469/9789241514842-eng.pdf, accessed 17 November 2018). 72 FIND. Performance of Xpert MTB/RIF version G4 assay. Version and date: 1.0/30 Nov 2011. (http://www.stoptb.org/wg/gli/assets/documents/map/findg4cartridge.pdf, accessed 15 August 2020). 73 Helb D, Jones M, Story E, et al. Rapid detection of Mycobacterium tuberculosis and rifampin resistance by use of on- demand, near-patient technology. J Clin Microbiol. 2010;48(1):229-237. doi:10.1128/JCM.01463-09. 74 Chakravorty S, Simmons AM, Rowneki M, et al. The new Xpert MTB/RIF Ultra: Improving detection of Mycobacterium tuberculosis and resistance to rifampin in an assay suitable for point-of-care testing. mBio. 2017;8(4) e00812-17. 75 Calvori C, Frontali L, Leoni L, Tecce G. Effect of rifamycin on protein synthesis. Nature. 1965;207(4995):417-418. doi:10.1038/207417a0. 76 Andre E, Goeminne L, Cabibbe A, et al. Consensus numbering system for the rifampicin resistance-associated rpoB gene mutations in pathogenic mycobacteria. Clin Microbiol Infect. 2017;23(3):167-172. doi:10.1016/j.cmi.2016.09.006. 77 Camus JC, Pryor MJ, Médigue C, Cole ST. Re-annotation of the genome sequence of Mycobacterium tuberculosis H37Rv. Microbiology (Reading). 2002;148(Pt 10):2967-2973. doi:10.1099/00221287-148-10-2967. 78 Miller LP, Crawford JT, Shinnick TM. The rpoB gene of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1994;38(4):805-811. doi:10.1128/AAC.38.4.805. 79 Heep M, Brandstätter B, Rieger U, et al. Frequency of rpoB mutations inside and outside the cluster I region in rifampin- resistant clinical Mycobacterium tuberculosis isolates. J Clin Microbiol. 2001;39(1):107-110. doi:10.1128/JCM.39.1.107- 110.2001. 80 Nebenzahl-Guimaraes H, Jacobson KR, Farhat MR, Murray MB. Systematic review of allelic exchange experiments aimed at identifying mutations that confer drug resistance in Mycobacterium tuberculosis. J Antimicrob Chemother. 2014;69(2):331- 342. doi:10.1093/jac/dkt358. 42 correlation between resistance mutations and phenotypic results, binomial 95% CIs were calculated, where applicable. Table 28 provides an overview of the current WHO and CLSI CCs for RIF, RFB and RPT.81,82 Table 28. Overview of current rifamycin CCs. LJ 7H10 7H11 MGIT Drug WHO CLSI WHO CLSI WHO CLSI WHO CLSI RIF 40.0 1.0 1.0 1.0 RFB – – – 0.5 – 0.5 – 0.5 RPT – – – – – – – – Green CCs were set by both the WHO and CLSI; blue CCs were set by CLSI. All concentrations are in mg/L. 81 Clinical & Laboratory Standards Institute. Clinical and Laboratory Standards Institute. Susceptibility testing of mycobacteria, nocardiae, and other aerobic actinomycetes, 3rd edition approved standard. CLSI Document M24; 2018. 82 World Health Organization. Technical manual for drug susceptibility testing of medicines used in the treatment of tuberculosis. (http://apps.who.int/iris/bitstream/handle/10665/275469/9789241514842-eng.pdf, accessed 17 November 2018). 43 3.A.1 RIF MIC data on LJ 3.A.1.1 RIF MICs for pWT isolates on LJ Four studies reported RIF MIC data for the pWT population on LJ, of which two were from the same laboratory (Table 29). All of these studies reported MICs for more than 10 pWT isolates, though the majority of the MIC distributions were truncated at the lower end, meaning that little insight could be gained about the shape of the pWT MIC distribution. It was, therefore, not clear whether the current CC of 40 mg/L or 32 mg/L (i.e. the equivalent ISO dilution for 30 mg/L) corresponds to the ECOFF. Table 29. RIF MICs for pWT isolates on LJ. The green line denotes the current WHO and CLSI CC for RIF on LJ (40 mg/L). Notable limitations: studies RF1 and RF5 were conducted in the same laboratory. 3.A.1.2 RIF MICs for mutated isolates on LJ rpoB 450 (531) mutants Three studies reported MICs for 160 rpoB S450 (S531) codon mutants tested on LJ (Table 30). Based on the current RIF CCs, only 2 (1%, 95% CI 0-4%) of these mutated isolates tested phenotypically RIF- S. An area of technical uncertainty (ATU)83 at 40 mg/L would lower this percentage to 1% (95% CI 0- 3%). Table 30. RIF MICs for rpoB S450 (S531) mutants on LJ. The green line denotes the current WHO and CLSI CC for RIF on LJ (40 mg/L). Notable limitations: studies RF1 and RF5 were conducted in the same laboratory. 83 European Committee on Antimicrobial Susceptibility Testing. Area of Technical Uncertainty (ATU) in antimicrobial susceptibility testing (http://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/Breakpoint_tables/Area_of_Technical_Uncertainty_- _guidance_2019-1.pdf, accessed 15 February 2020). Studies Lab Isolate origin Unique isolates Type of isolates Genotypic results - Mtb2 8 10 20 30 32 40 60 80 100 120 128 160 200 320 400 800 4) Fabry 1995 4 clinical 20 2 17 1 1) Vincent 2012 1 clinical 115 Mixture of first-line resistance profiles gWT 74 32 5 3 1 5 5 gWT 5 6 5 gWT 4 1 3 5 gWT 5 1 5 gWT 1 3 1 1 clinical 237 gWT 222 12 2 1 1 clinical 81 gWT 55 16 3 1 2 1 1 1 1 7 1 H37Rv 1 7 clinical 53 mixture of first-line resistance profiles gWT 52 1 RIF MIC [mg/L] clinical 6) Jagielski 2018 5) Rigouts 2013 & Van Deun 2009 & Van Deun (unpublished) Studies Lab Dataset Isolate origin Unique isolates Genotypic results 10 20 30 40 48 60 64 80 100 120 128 160 200 320 400 640 800 1 1 clinical 55 rpoB S450L (S531L) 4 2 49 1 1 clinical 3 rpoB S450W (S531W) 3 1 1 clinical 1 rpoB S450F (S531F) 1 5 5 clinical 1 rpoB S450L (S531L) 1 6 5 clinical 1 rpoB S450L (S531L) 1 3 5 clinical 1 rpoB S450L (S531L) 1 1 5 clinical 1 rpoB S450L (S531L) 1 1 6 clinical 22 rpoB S450L (S531L) 1 6 15 1 8 clinical 27 rpoB S450L (S531L) 1 1 1 1 13 2 1 7 1 8 clinical 1 rpoB S450W (S531W) 1 1 7 clinical 11 rpoB S450L (S531L) 1 10 1 7 clinical 4 rpoB S450W (S531W) 4 7 9 clinical 31 rpoB S450L (S531L) 1 6 24 7 9 clinical 1 rpoB S450W (S531W) 1 RIF MIC [mg/L] 6) Jagielski 2018 5) Rigouts 2013 & Van Deun 2009 & Van Deun (unpublished) 1) Vincent 2012 44 rpoB borderline RRDR mutants Four studies reported MICs for 115 rpoB borderline RRDR mutants tested on LJ (Table 31). At the current RIF CC, 32 (28%, 95% CI 20-37%) of these 115 mutated isolates tested phenotypically RIF-S, which could be lowered to 22% (95% CI 15-30%) with an ATU at 40 mg/L. Table 31. RIF MICs for rpoB borderline RRDR mutants on LJ. The green line denotes the current WHO and CLSI CC for RIF on LJ (40 mg/L). Notable limitations: studies RF1 and RF5 and studies RF3 and RF5 were conducted in the same laboratory. Other rpoB RRDR mutants One hundred thirty isolates from 3 studies harboured rpoB RRDR mutations other than S450 (S531) and the borderline mutations (Table 32). Based on the current RIF CC on LJ, only 3 (2%, 95% CI 0-7%) of these 130 mutated isolates tested phenotypically RIF-S. This would be only slightly lowered to 2% (95% CI 0-5%) with an ATU at 40 mg/L. Studies Lab Isolate origin Unique isolates Genotypic results 5 10 20 30 40 48 60 80 96 100 120 128 160 200 320 400 640 800 1) Vincent 2012 1 clinical 1 rpoB L430P (L511P) 1 5) Rigouts 2013 & Van 5 clinical 2 rpoB L430P (L511P) 1 1 5) Rigouts 2013 & Van 6 clinical 3 rpoB L430P (L511P) 1 2 5) Rigouts 2013 & Van 3 clinical 3 rpoB L430P (L511P) 2 1 5) Rigouts 2013 & Van 1 clinical 3 rpoB L430P (L511P) 1 2 5) Rigouts 2013 & Van 1 clinical 1 rpoB L430P (L511P) 1 5) Rigouts 2013 & Van 1 clinical 5 rpoB L430P (L511P) 1 1 1 2 5) Rigouts 2013 & Van 1 clinical 6 rpoB L430P (L511P) 3 1 1 1 1) Vincent 2012 1 clinical 2 rpoB D435Y (D516Y) 2 5) Rigouts 2013 & Van 5 clinical 4 rpoB D435Y (D516Y) 2 1 1 5) Rigouts 2013 & Van 6 clinical 4 rpoB D435Y (D516Y) 1 1 2 5) Rigouts 2013 & Van 3 clinical 4 rpoB D435Y (D516Y) 2 2 5) Rigouts 2013 & Van 1 clinical 2 rpoB D435Y (D516Y) 1 1 5) Rigouts 2013 & Van 1 clinical 6 rpoB D435Y (D516Y) 3 2 1 5) Rigouts 2013 & Van 1 clinical 6 rpoB D435Y (D516Y) 2 3 1 6) Jagielski 2018 7 clinical 2 rpoB D435Y (D516Y) 2 5) Rigouts 2013 & Van 5 clinical 1 rpoB H445L (H526L) 1 5) Rigouts 2013 & Van 6 clinical 1 rpoB H445L (H526L) 1 5) Rigouts 2013 & Van 3 clinical 1 rpoB H445L (H526L) 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445L (H526L) 1 5) Rigouts 2013 & Van 1 clinical 10 rpoB H445L (H526L) 5 5 6) Jagielski 2018 7 clinical 1 rpoB H445L (H526L) 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445N (H526N) 1 5) Rigouts 2013 & Van 1 clinical 5 rpoB H445N (H526N) 1 1 1 1 1 1) Vincent 2012 1 clinical 1 rpoB H445S (H526S) 1 5) Rigouts 2013 & Van 5 clinical 1 rpoB H445S (H526S) 1 5) Rigouts 2013 & Van 6 clinical 1 rpoB H445S (H526S) 1 5) Rigouts 2013 & Van 3 clinical 1 rpoB H445S (H526S) 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445S (H526S) 1 3) Andres 2014 3 clinical 2 rpoB L452P (L533P) 1 1 1) Vincent 2012 1 clinical 6 rpoB L452P (L533P) 1 1 1 3 5) Rigouts 2013 & Van 5 clinical 2 rpoB L452P (L533P) 2 5) Rigouts 2013 & Van 6 clinical 2 rpoB L452P (L533P) 1 1 5) Rigouts 2013 & Van 3 clinical 2 rpoB L452P (L533P) 2 5) Rigouts 2013 & Van 1 clinical 1 rpoB L452P (L533P) 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB L452P (L533P) 1 5) Rigouts 2013 & Van 1 clinical 3 rpoB L452P (L533P) 2 1 5) Rigouts 2013 & Van 1 clinical 15 rpoB L452P (L533P) 4 5 5 1 6) Jagielski 2018 7 clinical 1 rpoB L452P (L533P) 1 RIF MIC [mg/L] 45 Table 32. RIF MICs for other rpoB RRDR mutants on LJ. The green line denotes the current WHO and CLSI CC for RIF on LJ (40 mg/L). Notable limitations: studies RF1 and RF5 were conducted in the same laboratory. rpoB mutants outside the RRDR Fourteen isolates from 2 studies harbored rpoB I491F (I572F) mutations. Based on the current RIF CC on LJ, 5 (36%, 95% CI 13-65%) of these 14 mutated isolates were classified as phenotypically RIF-S. This percentage would be lowered to 21% (95% CI 5-51%) given an ATU at 40 mg/L. Table 33. RIF MICs for rpoB mutants outside the RRDR on LJ. The green line denotes the current WHO and CLSI CC for RIF on LJ (40 mg/L). Notable limitations: studies RF1 and RF5 were conducted in the same laboratory. 3.A.1.3 Conclusion for RIF CC for LJ Given that the MIC data for RIF on LJ were limited (Table 29), it was not clear whether 40 mg/L corresponds to the ECOFF. To avoid potentially misclassifying susceptible isolates as resistant, 40 mg/L was reaffirmed as the CC. However, the TEG noted that this decision would be revisited should data Studies Lab Isolate origin Unique isolates Genotypic results 10 20 30 40 48 60 80 100 120 128 160 200 320 400 640 800 5) Rigouts 2013 & Van 1 clinical 1 rpoB Q432K (Q513K) 1 5) Rigouts 2013 & Van 1 clinical 2 rpoB Q432K (Q513K) 2 5) Rigouts 2013 & Van 1 clinical 1 rpoB Q432P (Q513P) 1 5) Rigouts 2013 & Van 1 clinical 4 rpoB Q432P (Q513P) 4 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445C (H526C) 1 6) Jagielski 2018 7 clinical 2 rpoB H445C (H526C) 1 1 1) Vincent 2012 1 clinical 5 rpoB H445D (H526D) 5 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445D (H526D) 1 5) Rigouts 2013 & Van 1 clinical 4 rpoB H445D (H526D) 1 2 1 5) Rigouts 2013 & Van 1 clinical 10 rpoB H445D (H526D) 1 9 6) Jagielski 2018 7 clinical 2 rpoB H445D (H526D) 1 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB D435F (D516F) 1 5) Rigouts 2013 & Van 1 clinical 4 rpoB D435F (D516F) 1 3 1) Vincent 2012 1 clinical 7 rpoB D435V (D516V) 1 6 5) Rigouts 2013 & Van 1 clinical 2 rpoB D435V (D516V) 2 5) Rigouts 2013 & Van 1 clinical 1 rpoB D435V (D516V) 1 5) Rigouts 2013 & Van 1 clinical 15 rpoB D435V (D516V) 1 1 13 6) Jagielski 2018 7 clinical 2 rpoB D435V (D516V) 2 5) Rigouts 2013 & Van 1 clinical 1 rpoB S441L (S522L) 1 5) Rigouts 2013 & Van 1 clinical 2 rpoB S441L (S522L) 2 5) Rigouts 2013 & Van 1 clinical 1 rpoB S441Q (S522Q) 1 5) Rigouts 2013 & Van 1 clinical 11 rpoB S441Q (S522Q) 1 1 1 8 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445P (H526P) 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445P (H526P) 1 1) Vincent 2012 1 clinical 5 rpoB H445R (H526R) 1 4 5) Rigouts 2013 & Van 1 clinical 1 rpoB H445R (H526R) 1 5) Rigouts 2013 & Van 1 clinical 5 rpoB H445R (H526R) 5 1) Vincent 2012 1 clinical 9 rpoB H445Y (H526Y) 9 5) Rigouts 2013 & Van 1 clinical 3 rpoB H445Y (H526Y) 3 5) Rigouts 2013 & Van 1 clinical 6 rpoB H445Y (H526Y) 5 1 5) Rigouts 2013 & Van 1 clinical 14 rpoB H445Y (H526Y) 3 11 6) Jagielski 2018 7 clinical 3 rpoB H445Y (H526Y) 1 1 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB K446Q (K527Q) 1 5) Rigouts 2013 & Van 1 clinical 1 rpoB R447L (R528L) 1 RIF MIC [mg/L] Studies Lab Isolate origin Unique isolates Genotypic results 10 20 30 40 60 80 120 160 320 640 800 1) Vincent 2012 1 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 5 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 6 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 3 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 1 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 1 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 1 clinical 1 rpoB I491F (I572F) 1 5) Rigouts 2013 & Van Deun 1 clinical 7 rpoB I491F (I572F) 1 1 1 2 2 RIF MIC [mg/L] 46 from appropriately designed studies become available.84 The effect of introducing an ATU at this concentration was considered, though no ATU was endorsed (see Section 3.3 for more details). Table 34. Effect of introducing an ATU for RIF DST on LJ. Mutation type Percentage of mutants classified RIF-S at CC 40 mg/L CC+ATU 40 mg/L S450 (S531) 1% (95% CI 0-4%) 1% (95% CI 0-3%) borderline RRDR 28% (95% CI 20-37%) 22% (95% CI 15-30%) other RRDR 2% (95% CI 0-7%) 2% (95% CI 0-5%) borderline I491F (I572F) 36% (95% CI 13-65%) 21% (95% CI 5-51%) The current CC is underlined. 84 Schön T, Köser CU, Werngren J, et al. What is the role of the EUCAST reference method for MIC testing of the Mycobacterium tuberculosis complex? Clin Microbiol Infect. 2020;26(11):1453-1455. doi:10.1016/j.cmi.2020.07.037. 47 3.A.2 RIF MIC data on 7H10 3.A.2.1 RIF MICs for pWT isolates on 7H10 Fifteen studies were identified that reported RIF MIC data for the pWT population on 7H10 (Table 35). Nine of these studies featured MIC data for more than 10 pWT isolates. Seven of these studies reported MIC data from the same laboratories. Most distributions were not truncated and therefore provided a good understanding of the shape of the pWT MIC distributions. García et al. clearly had a systematic error with testing given that no on-scale MIC results were obtained for all presumably pWT isolates despite 0.016 mg/L being tested as the lowest concentration (this was also the case for RFB on 7H10 for the same study, as shown in Table 50). The modes of the pWT distributions for most of the remaining studies were all within plus or minus one dilution (i.e. 0.06-0.25 mg/L). Only 1 dataset (de Steenwinkel et al.) had an elevated mode of 0.5 mg/L, for which the current CC of 1 mg/L would be optimal. Therefore, the tentative ECOFF for 7H10 is likely 0.5 mg/L. Lowering the CC to 0.5 mg/L would misclassify at most 37 of 8537 isolates (0.4%, 95% CI 0.3-0.6%) that are currently regarded as RIF-S. However, this is likely an overestimate given that 32 of those 37 isolates were from a single laboratory from Farhat et al., and the quality of the results is questionable. This study stood out because of the unusually high number of isolates that either tested phenotypically INH-S despite harbouring classical INH resistance mutations (Table 12 and Table 13) or that were phenotypically RIF- S isolates despite having rpoB S450 (S531) mutations (Table 36). In addition, this study was enriched for MDR isolates and had an unusually long tail of isolates that were phenotypically RIF-R despite being characterized as gWT by whole genome sequencing. The pWT MIC distribution from this laboratory was strongly truncated and no MICs for H37Rv were available. Excluding this study would reduce the number of potentially misclassified isolates to just 5 of 8362 (0.1%, 95% 0.0-0.1%). Table 35. RIF MICs for pWT isolates on 7H10. The green line denotes the current WHO and CLSI CC for RIF DST on 7H10 (1 mg/L). Notable limitations: studies RF19 and RF20, RF25 and RF11, RF25 and RF10, and RF13 and RF15 were conducted in the same laboratory, and RF21 had a systematic error with testing. 3.A.2.2 RIF MICs for mutated isolates on 7H10 rpoB 450 (531) mutants Eleven studies reported 731 MICs for isolates with rpoB S450 (S531) codon mutations tested on 7H10 (Table 36). To err on the side of caution, Mvelase et al. was excluded from all calculations in this report as the S450 MICs reported in the study were systematically lower than those in all other studies for Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic results - Mtb0.02 0.03 0.04 0.06 0.08 0.1 0.12 0.2 0.25 0.3 0.5 0.7 1 1.25 2 3 4 5 8 10 16 25 32 50 64 128 16 1 2 H37Rv ATCC25618 1 1 16 1 1 BTB 08-049 1 16 22 22 Mostly RIF-R gWT 1 7 8 2 1 1 1 1 16 1 4 H37Rv ATCC27294 4 16 1 1 BTB 08-049 1 16 clinical 92 92 8 11 24 33 14 2 12 1 2 H37Rv 2 12 17 34 gWT 10 4 4 10 4 2 17 reference 16 48 30 18 17 clinical 63 189 138 51 21 1 3 H37Rv ATCC27294 1 1 1 21 1 3 Erdman 1 1 1 21 clinical 76 81 mixture of first-line resistance profiles gWT 5 14 36 15 9 2 9 6 6 gWT 1 2 3 10 222 222 gWT 100 43 32 1 14 2 2 3 25 11) Heifets 1999 10 1 10 H37Rv 9 1 15 1 2 H37Rv ATCC27294 2 15 clinical 11 22 1 4 16 1 19 1 1 H37Rv 1 19 clinical 13 13 Mostly MDR gWT 3 4 4 1 1 14) Kusunoki 1995 13 clinical 65 65 24 14 4 1 2 2 2 16 10) van Klingeren 2007 9 clinical 7956 7956 Mostly RIF-S 1234 5766 844 3 7 7 95 22) Marubini 2016 18 clinical 9 9 gWT 7 2 12) Moghazeh 1996 11 clinical 1 1 Pan-S gWT 1 12 1 1 H37Rv 1 12 clinical 7 7 gWT 7 18) Cavusoglu 2004 8 1 1 H37Rv ATCC27294 1 23) Pholwat 2011 17) de Steenwinkel 2012, de Steenwinkel 2012 & Bax 2017 RIF MIC [mg/L] 25) Farhat 2019 clinical Mostly MDR 15) Miotto 2018 19) Schön 2013 20) Schön 2009 13) Rancoita 2018 21) García 2010 26) Gygli 2019 48 both 7H10 and MGIT (Table 45). Excluding Mvelase et al. left 727 acceptable MICs. Farhat et al. also showed a wide MIC range for S450 mutants, suggesting a higher random error rate for testing or that the phenotypic effect of this mutation may depend on the genetic background. However, because the majority of MICs in this study were high, Farhat et al. was not excluded. Based on the current RIF CC, 20 (3%, 95% CI 2-4%) of the 727 mutated isolates tested phenotypically RIF-S. If the current CC was lowered one dilution to 0.5 mg/L, only 15 (2%, 95% CI 1-3%) MICs would be classified as RIF-S. The corresponding percentage given an ATU at 0.5 mg/L would be 1% (95% CI 1-3%). Table 36. RIF MICs for rpoB S450 (S531) mutants on 7H10. The green line denotes the current WHO and CLSI CC for RIF DST on 7H10 (1 mg/L). Notable limitations: studies RF13 and RF15 were conducted in the same laboratory and study RF24 had systematically lower MIC distributions for all mutants tested and was, therefore, excluded from calculations in this report. rpoB borderline RRDR mutants Eleven studies reported 112 MICs for rpoB borderline RRDR mutants tested on 7H10 (Table 37). Excluding the study by Mvelase et al., 21 (24%, 95% CI 16-35%) of 86 total mutated isolates tested phenotypically RIF-S at the current CC. If the CC was lowered one dilution to 0.5 mg/L, 13 (15%, 95% CI 8-24%) of the 86 mutants would still test phenotypically RIF-S on 7H10. This could be reduced to 6% (95% CI 2-13%) with an ATU at 0.5 mg/L. Studies Lab Isolate origin Unique isolates Total MICs Genotypic results 0.12 0.25 0.5 1 1.25 2 4 5 8 10 16 20 32 50 64 128 256 320 13) Rancoita 2018 12 10 20 rpoB S450L (S531L) 4 6 2 8 16 46 46 rpoB S450L (S531L) 1 3 6 19 11 6 16 1 1 rpoB S450W (S531W) 1 18 2 2 rpoB S450L (S531L) 2 18 1 1 rpoB S450W (S531W) 1 18 1 1 rpoB S450Q (S531Q) 1 18 1 1 rpoB S450P (S531P) 1 23) Pholwat 2011 19 clinical 18 18 rpoB S450L (S531L) 18 20 3 3 rpoB S450L (S531L) 1 1 1 20 1 1 rpoB S450P (S531P) 1 26) Gygli 2019 21 clinical 43 45 rpoB S450L (S531L) 1 1 2 15 18 8 11 4 4 rpoB S450L (S531L) 4 11 1 1 rpoB S450W (S531W) 1 14 5 5 rpoB S450L (S531L) 5 14 1 1 rpoB S450W (S531W) 1 10 12 12 rpoB S450F (S531F) 12 10 442 442 rpoB S450L (S531L) 8 4 1 3 4 1 6 15 400 10 4 4 rpoB S450M (S450M) 1 3 10 5 5 rpoB S450Q (S531Q) 5 10 7 7 rpoB S450W (S531W) 1 6 9 80 80 rpoB S450L (S531L) 2 2 76 9 5 5 rpoB S450W (S531W) 5 12 2 2 rpoB S450F (S531F) 2 12 1 1 rpoB S450L (S531L) 1 8 19 19 rpoB S450L (S531L) 2 1 1 15 8 4 4 rpoB S450W (S531W) 4 RIF MIC [mg/L] 15) Miotto 2018 19) Schön 2013 22) Marubini 2016 clinical 24) Mvelase 2019 clinical 12) Moghazeh 1996 clinical 16) Park 2017 mouse clinical 25) Farhat 2019 clinical clinical clinical18) Cavusoglu 2004 49 Table 37. RIF MICs for rpoB borderline RRDR mutants on 7H10. The green line denotes the current WHO and CLSI CC for RIF DST on 7H10 (1 mg/L). Notable limitations: studies RF9 and RF25 were conducted in the same laboratory, and study RF24 had systematically lower MIC distributions for all mutants tested and was, therefore, excluded from calculations in this report. Other rpoB RRDR mutants Eleven studies reported 339 MICs for isolates that harbored rpoB RRDR mutations other than S450 (S531) and the borderline mutations tested on 7H10 (Table 38). Excluding Mvelase et al., only 3 (1%, 95% CI 0-3%) of 300 isolates would test phenotypically RIF-S at the current CC or if the CC were lowered one dilution to 0.5 mg/L. An ATU at 0.5 mg/L would eliminate this misclassification altogether. Studies Lab Isolate origin Unique isolates Total MICs Genotypic results 0.12 0.25 0.5 1 2 3 4 5 8 10 16 32 50 64 128 19) Schön 2013 16 1 1 rpoB L430P (L511P) 1 22) Marubini 2016 18 clinical 1 1 rpoB L430P (L511P) 1 23) Pholwat 2011 19 clinical 1 1 rpoB L430P (L511P) 1 24) Mvelase 2019 20 clinical 7 7 rpoB L430P (L511P) 2 4 1 25) Farhat 2019 10 clinical 3 3 rpoB L430P (L511P) 1 1 1 19) Schön 2013 16 3 3 rpoB D435Y (D516Y) 1 2 22) Marubini 2016 18 clinical 2 2 rpoB D435Y (D516Y) 1 1 24) Mvelase 2019 20 clinical 11 11 rpoB D435Y (D516Y) 1 3 4 3 25) Farhat 2019 10 clinical 11 11 rpoB D435Y (D516Y) 1 2 4 2 2 9) van Ingen 2011 9 clinical 4 7 rpoB D435Y (D516Y) 1 1 5 15) Miotto 2018 12 clinical 3 3 rpoB D435Y (D516Y) 3 25) Farhat 2019 9 clinical 3 3 rpoB D435Y (D516Y) 1 2 18) Cavusoglu 2004 8 clinical 2 2 rpoB D435Y (D516Y) 2 19) Schön 2013 16 1 1 rpoB H445L (H526L) 1 24) Mvelase 2019 20 clinical 6 6 rpoB H445L (H526L) 2 1 3 26) Gygli 2019 21 clinical 1 1 rpoB H445L (H526L) 1 25) Farhat 2019 10 clinical 13 13 rpoB H445L (H526L) 1 1 2 1 5 1 2 12) Moghazeh 1996 11 clinical 2 2 rpoB H445L (H526L) 2 25) Farhat 2019 9 clinical 1 1 rpoB H445L (H526L) 1 22) Marubini 2016 18 clinical 1 1 rpoB H445N (H526N) 1 25) Farhat 2019 10 clinical 3 3 rpoB H445N (H526N) 1 1 1 12) Moghazeh 1996 11 clinical 1 1 rpoB H445N (H526N) 1 15) Miotto 2018 12 clinical 1 1 rpoB H445N (H526N) 1 24) Mvelase 2019 20 clinical 2 2 rpoB H445S (H526S) 1 1 25) Farhat 2019 10 clinical 1 1 rpoB H445S (H526S) 1 15) Miotto 2018 12 clinical 5 5 rpoB H445S (H526S) 2 1 2 19) Schön 2013 16 1 1 rpoB L452P (L533P) 1 22) Marubini 2016 18 clinical 1 1 rpoB L452P (L533P) 1 26) Gygli 2019 21 clinical 2 3 rpoB L452P (L533P) 1 1 1 25) Farhat 2019 10 clinical 10 10 rpoB L452P (L533P) 1 1 2 1 1 1 1 2 12) Moghazeh 1996 11 clinical 1 1 rpoB L452P (L533P) 1 16) Park 2017 14 mouse 1 1 rpoB L452P (L533P) 1 15) Miotto 2018 12 clinical 1 1 rpoB L452P (L533P) 1 18) Cavusoglu 2004 8 clinical 1 1 rpoB L452P (L533P) 1 RIF MIC [mg/L] 50 Table 38. RIF MICs for other rpoB RRDR mutants on 7H10. The green line denotes the current WHO and CLSI CC for RIF DST on 7H10 (1 mg/L). Notable limitations: studies RF7 and RF18 were conducted in the same laboratory and study RF24 had systematically lower MIC distributions for all mutants tested and was, therefore, excluded from calculations in this report. rpoB mutants outside the RRDR Only 2 studies were identified that presented MIC data for isolates with rpoB mutations outside of the RRDR on 7H10 (Table 39). All 14 mutants had MICs above the RIF CC on 7H10, except for the one P454L (P535L) mutation, which would potentially be identified as false-resistant by the Nipro NTM MDR-TB2 and Xpert Ultra assays, assuming that this mutation is neutral. An ATU at 0.5 mg/L would not change these results. Table 39. RIF MICs for rpoB mutants outside the RRDR on 7H10. The green line denotes the current WHO and CLSI CC for RIF DST on 7H10 (1 mg/L). Studies Lab Isolate origin Unique isolates Total MICs Genotypic results 0.12 0.25 0.5 1 1.25 2 3 4 5 8 10 16 20 32 50 64 128 256 320 24) Mvelase 2019 20 clinical 1 1 rpoB S428T (S509T) 1 23) Pholwat 2011 19 clinical 1 1 rpoB S431N (S512N) 1 15) Miotto 2018 12 clinical 1 1 rpoB S431R (S512R) 1 25) Farhat 2019 10 clinical 2 2 rpoB Q432E (Q513E) 2 7) Cavusoglu 2006 8 clinical 1 1 rpoB Q432E (Q513E) 1 25) Farhat 2019 10 clinical 3 3 rpoB Q432K (Q513K) 3 12) Moghazeh 1996 11 clinical 1 1 rpoB Q432K (Q513K) 1 12) Moghazeh 1996 11 clinical 1 1 rpoB Q432L (Q513L) 1 15) Miotto 2018 12 clinical 3 3 rpoB Q432L (Q513L) 3 24) Mvelase 2019 20 clinical 21 21 rpoB Q432P (Q513P) 3 11 3 4 25) Farhat 2019 10 clinical 2 2 rpoB Q432P (Q513P) 2 25) Farhat 2019 9 clinical 1 1 rpoB Q432P (Q513P) 1 18) Cavusoglu 2004 8 clinical 1 1 rpoB Q432P (Q513P) 1 25) Farhat 2019 10 clinical 1 1 rpoB D435F (D516F) 1 19) Schön 2013 16 9 9 rpoB D435V (D516V) 2 2 1 3 1 23) Pholwat 2011 19 clinical 2 2 rpoB D435V (D516V) 1 1 24) Mvelase 2019 20 clinical 16 16 rpoB D435V (D516V) 1 1 3 7 3 1 26) Gygli 2019 21 clinical 18 20 rpoB D435V (D516V) 1 3 10 4 1 1 25) Farhat 2019 10 clinical 130 130 rpoB D435V (D516V) 3 1 2 3 2 4 4 38 73 25) Farhat 2019 9 clinical 9 9 rpoB D435V (D516V) 3 6 18) Cavusoglu 2004 8 clinical 1 1 rpoB D435V (D516V) 1 25) Farhat 2019 10 clinical 3 3 rpoB S441L (S522L) 3 16) Park 2017 14 mouse 1 1 rpoB S441L (S522L) 1 18) Cavusoglu 2004 8 clinical 2 2 rpoB S441W (S522W) 1 1 25) Farhat 2019 10 clinical 7 7 rpoB H445C (H526C) 1 1 4 1 15) Miotto 2018 12 clinical 2 2 rpoB H445C (H526C) 2 25) Farhat 2019 9 clinical 2 2 rpoB H445C (H526C) 2 18) Cavusoglu 2004 8 clinical 1 1 rpoB H445C (H526C) 1 19) Schön 2013 16 3 3 rpoB H445D (H526D) 1 2 26) Gygli 2019 21 clinical 7 9 rpoB H445D (H526D) 5 4 12) Moghazeh 1996 11 clinical 2 2 rpoB H445D (H526D) 2 25) Farhat 2019 10 clinical 7 7 rpoB H445D (H526D) 1 1 5 25) Farhat 2019 9 clinical 4 4 rpoB H445D (H526D) 4 18) Cavusoglu 2004 8 clinical 1 1 rpoB H445D (H526D) 1 25) Farhat 2019 10 clinical 1 1 rpoB H445G (H526G) 1 25) Farhat 2019 9 clinical 1 1 rpoB H445G (H526G) 1 25) Farhat 2019 10 clinical 1 1 rpoB H445P (H526P) 1 15) Miotto 2018 12 clinical 1 1 rpoB H445Q (H526Q) 1 22) Marubini 2016 18 clinical 1 1 rpoB H445R (H526R) 1 24) Mvelase 2019 20 clinical 1 1 rpoB H445R (H526R) 1 25) Farhat 2019 10 clinical 10 10 rpoB H445R (H526R) 1 1 8 16) Park 2017 14 mouse 2 2 rpoB H445R (H526R) 2 25) Farhat 2019 9 clinical 2 2 rpoB H445R (H526R) 2 18) Cavusoglu 2004 8 clinical 4 4 rpoB H445R (H526R) 4 19) Schön 2013 16 3 3 rpoB H445Y (H526Y) 1 2 23) Pholwat 2011 19 clinical 2 2 rpoB H445Y (H526Y) 2 26) Gygli 2019 21 clinical 5 6 rpoB H445Y (H526Y) 3 3 12) Moghazeh 1996 11 clinical 4 4 rpoB H445Y (H526Y) 4 25) Farhat 2019 10 clinical 16 16 rpoB H445Y (H526Y) 1 15 16) Park 2017 14 mouse 4 4 rpoB H445Y (H526Y) 4 25) Farhat 2019 9 clinical 5 5 rpoB H445Y (H526Y) 5 18) Cavusoglu 2004 8 clinical 2 2 rpoB H445Y (H526Y) 2 25) Farhat 2019 10 clinical 2 2 rpoB R448K (R529K) 2 RIF MIC [mg/L] Studies Lab Isolate origin Unique isolates Genotypic results 0.25 0.5 1 1.25 2 3 4 5 8 16 32 50 64 128 256 26) Gygli 2019 21 clinical 1 rpoB V170F (V146F) 1 9 clinical 1 rpoB V170F (V146F) 1 10 clinical 5 rpoB V170F (V146F) 5 10 clinical 1 rpoB P454L (P535L) 1 10 clinical 6 rpoB I491F (I572F) 2 1 1 1 1 25) Farhat 2019 RIF MIC [mg/L] 51 3.A.2.3 Conclusion for RIF CC for 7H10 The experts were in agreement that the ECOFF for 7H10 corresponds to 0.5 mg/L (Table 35). Only 0.4% (95% CI 0.3-0.6%) of isolates currently considered to be RIF-S would be potentially misclassified as RIF- R at this concentration. However, this estimate was almost entirely driven by MIC data from Farhat et al., for which many results were not in keeping with other well-established evidence. Excluding this one study reduces the potential rate of misclassification of RIF-S isolates to just 0.1% (95% 0.0-0.1%), whilst reducing the rate of misclassification of borderline RRDR mutations from 24% (95% CI 16-35%) to 15% (95% CI 8-24%), as shown in Table 40. An ATU at 0.5 mg/L would further reduce the rate of misclassification to 6% (95% CI 2-13%). Consequently, the CC was lowered to 0.5 mg/L but an ATU was not endorsed (see Section 3.3 for more details). Table 40. Effect of changing the CC for RIF DST on 7H10 for detection of rpoB mutants. Mutation type Percentage of mutants classified RIF-S at CC 1 mg/L CC+ATU 1 mg/L or CC 0.5 mg/L CC+ATU 0.5 mg/L S450 (S531) 3% (95% CI 2-4%) 2% (95% CI 1-3%) 1% (95% CI 1-3%) borderline RRDR 24% (95% CI 16-35%) 15% (95% CI 8-24%) 6% (95% CI 2-13%) other RRDR 1% (95% CI 0-3%) 1% (95% CI 0-3%) 0% (95% CI 0-1%) V170F (V146F) 0% (95% CI 0-41%) borderline I491F (I572F) 0% (95% CI 0-46%) The newly adopted CC is underlined. 52 3.A.3 RIF MIC data on 7H11 3.A.3.1 RIF MICs for pWT isolates on 7H11 Ten studies were identified that reported RIF MIC data for the pWT population on 7H11 (Table 41). Ten of these studies reported MICs for at least 10 pWT isolates, and the MIC distributions were largely untruncated. The mode of the distributions ranged from 0.25 up to 1 mg/L, the current CC. Modes that are identical to the ECOFF are unusual but based on the data from Heifets et al., this did not appear to result in an unusually high false-resistance rate. Table 41. RIF MICs for pWT isolates on 7H11. The green line denotes the current WHO and CLSI CC for RIF DST on 7H11 (1 mg/L). Notable limitation: studies RF35 and RF36 were conducted in the same laboratory. 3.A.3.2 RIF MICs for mutated isolates on 7H11 rpoB 450 (531) mutants Only 1 study reported RIF MICs for rpoB S450 (S531) codon mutants tested on 7H11 (Table 42). Based on the current RIF CC, all 5 mutants reported in this study (100%, 95% CI 48-100%) tested phenotypically RIF-R. Table 42. RIF MICs for rpoB S450 (S531) mutants on 7H11. The green line denotes the current WHO and CLSI CC for RIF DST on 7H11 (1 mg/L). rpoB borderline RRDR mutants No studies presenting MIC data for rpoB borderline RRDR mutants on 7H11 were identified. Other rpoB RRDR mutants Only 1 study reported RIF MICs for rpoB RRDR mutants harboring mutations other than S450 (S531) and the borderline mutations on 7H11 (Table 43). Based on the current RIF CC, all 16 mutants reported in this study (100%, 95% CI 79-100%) tested phenotypically RIF-R. Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic results - Mtb0.03 0.06 0.12 0.25 0.5 1 2 4 5 8 10 16 32 40 22 1 1 H37Rv 1 22 1 1 M. africanum ATCC 25420 1 22 1 1 M. bovis ATCC 19210 1 22 5 5 M. bovis BCG 2 1 1 1 22 6 6 Pan-S 4 2 22 4 4 M. africanum 2 2 22 4 4 M. bovis 3 1 24 1 1 H37Rv 1 24 1 1 M. tuberculosis ATCC 35822 1 24 1 1 M. tuberculosis ATCC 35820 1 24 1 1 M. tuberculosis ATCC 35837 1 24 clinical 40 40 1 7 14 3 3 1 1 10 31) Shishido 2007 26 1 15 BCG (Tokyo) 9 6 35) Lee 1987 10 clinical 17 17 Pre-treatment 4 9 4 37) Truffot-Pernot 1988 29 clinical 10 10 RIF-S 1 5 4 33) Rodriguez 2003 28 clinical 16 16 3 7 5 1 23 1 1 H37Rv ATCC27294 1 23 1 1 M. tuberculosis ATCC 35838 1 23 clinical 46 46 Mostly MDR 5 5 2 1 2 3 28 25 1 2 H37Rv 1 1 25 10 20 INH-S 1 3 6 10 25 12 12 INH-R 6 6 36) Heifets 1985 10 clinical 180 180 untreated 180 32) Anthony 2005 & 27 lab 2 2 gWT 2 RIF MIC [mg/L] 27) Rastogi 2000 clinical clinical 28) Fattorini 1999 29) Coban 2013 30) Rey-Jurado 2013 & Rey-Jurado 2012 Studies Isolate origin Unique isolates Genotypic results 1 2 4 8 16 32 64 128 256 lab 4 rpoB S450L (S531L) 4 lab 1 rpoB S450W (S531W) 1 RIF MIC [mg/L] 32) Anthony 2005 & Bergval 2007 53 Table 43. RIF MICs for other rpoB RRDR mutants on 7H11. The green line denotes the current WHO and CLSI CC for RIF DST on 7H11 (1 mg/L). rpoB mutants outside the RRDR No studies presenting MIC data for rpoB isolates with mutations outside the RRDR on 7H11 were identified. 3.A.3.3 Conclusion for RIF CC for 7H11 The quality and quantity of MIC data for pWT isolates for 7H11 (Table 41) were considerably worse than for 7H10 (Table 35). A change of the current CC of 1 mg/L was not warranted based on these limited data. However, the fact that the CC for 7H10 had to be lowered from 1 mg/L to the ECOFF of 0.5 mg/L raises the possibility that 1 mg/L may be too high for 7H11 as well. Additional data are needed to clarify this question. The value of adopting an ATU could not be assessed given that no MICs for borderline resistance mutations were identified. Studies Isolate origin Unique isolates Genotypic results 1 2 4 8 16 32 64 128 256 512 lab 1 rpoB Q432L (Q513L) 1 lab 1 rpoB S441L (S522L) 1 lab 4 rpoB H445D (H526D) 1 3 lab 5 rpoB H445R (H526R) 1 3 1 lab 1 rpoB H445P (H526P) 1 lab 4 rpoB H445Y (H526Y) 1 3 lab 5 rpoB H445R (H526R) 1 3 1 RIF MIC [mg/L] 32) Anthony 2005 & Bergval 2007 54 3.A.4 RIF MIC data in MGIT 3.A.4.1 RIF MICs for pWT isolates in MGIT Twenty-one studies were identified that reported RIF MIC data for the pWT population by MGIT (Table 44). Thirteen of these studies reported MICs for at least 10 pWT isolates, although many of the MIC distributions were truncated, precluding an assessment of the shape of the distributions. Because genotypic information was not available for key datasets (e.g. Chigutsa et al. and Rockwood et al.), it was not clear whether 0.5 mg/L was the upper end of their respective pWT MIC distributions. It is therefore not clear whether 0.25 or 0.5 mg/L represents the tentative ECOFF. Lowering the CC to 0.5 mg/L would result in only one of 523 isolates (0.2%, 95% 0.0-1.1%) being misclassified as RIF-R. Notably, this isolate from Bernardelli et al. was isolated from a seal and might represent an animal variant of MTBC. Table 44. RIF MICs for pWT isolates in MGIT. The green line denotes the current WHO and CLSI CC for RIF DST in MGIT (1 mg/L). Notable limitations: studies RF55 and RF58; studies RF44, RF41, RF61 and RF54; studies RF57 and RF63; and studies RF50 and RF60 were conducted in the same laboratory. 3.A.4.2 RIF MICs for mutated isolates in MGIT rpoB 450 (531) mutants Sixteen studies reported MICs for 796 isolates with rpoB S450 (S531) codon mutations tested by MGIT (Table 45). Excluding Mvelase et al., which also had unusual MICs for 7H10 (Table 36), and based on the current RIF CC, only 2 (0%, 95% CI 0-1%) of the 792 mutated isolates tested phenotypically RIF-S. If the current CC were lowered by one dilution to 0.5 mg/L, at least 1 (0%, 95% CI 0-1%) mutant would still test RIF-S by MGIT. An ATU in this case would have minimal effect (0% (95% CI 0-1%)). Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic results - Mtb0.016 0.022 0.03 0.04 0.06 0.08 0.1 0.12 0.25 0.5 1 2 4 5 8 10 16 20 50 100 120 16 1 3 H37Rv ATCC27294 1 2 33 1 2 H37Rv ATCC27294 2 16 clinical 26 26 Pre-treatment gWT 1 8 14 3 33 1 2 H37Rv 2 33 clinical 11 11 gWT 2 7 2 48) Chigutsa 2015 41 clinical 54 54 Pre-treatment 1 5 12 26 7 3 46 107 107 Development cohort gWT 3 7 24 20 28 19 6 46 24 24 Validation cohort gWT 1 12 3 3 5 3 1 1 H37Rv ATCC27294 1 3 1 4 H37Rv ATCC27294 2 2 3 clinical 16 16 gWT 1 4 11 39 93 93 gWT by Xpert 12 35 33 10 3 39 21 21 pre-treatment (paired) 3 11 5 1 1 39 21 21 post-treatment (paired) 3 7 8 3 41) Feuerrigel 2012 3 1 1 H37Rv ATCC27294 1 45 1 10 H37Rv 7 2 1 45 2 2 gWT 2 45 1 1 H37Rv ATCC27294 1 45 clinical 6 6 gWT 6 44 1 1 H37Rv ATCC27294 1 44 1 1 M. bovis BCG 1 44 1 1 M. bovis AN5 1 44 seals 7 7 3 3 1 49) Jamieson 2014 42 clinical 1 1 mix of first-ine drug resistance gWT 1 59) Mvelase 2019 20 1 9 H37Rv 1 7 1 1 5 5 H37Rv 3 2 1 clinical 13 13 gWT 11 2 61) Tessema 2017 3 clinical 18 18 gWT 12 1 5 30 1 1 H37Ra 1 30 clinical 36 36 gWT 36 50) Van Deun (unpublished) 1 clinical 5 5 gWT 5 51) Sirgel 2013 & 43 clinical 26 26 gWT 26 5 1 1 H37Rv ATCC27294 1 5 clinical 4 4 Pan-S and mono-INH-R gWT 4 54) Andres 2014 & 3 clinical 71 71 gWT 70 1 56) Whitfield 2018 31 25 25 gWT 20 2 1 2 21 1 1 H37Rv ATCC27294 1 21 1 1 Erdman 1 21 clinical 22 22 gWT 22 RIF MIC [mg/L] 55) Niward 2018 clinical 58) Groenheit - unpublished clinical 57) Lin - unpublished 63) Berrada 2016 52) Bernardelli 2004 62) Colangeli 2018 44) Heyckendorf 2018 46) Rockwood 2017 & Rockwood 2017 60) Torrea 2019 53) Machado 2018 & Machado (unpublished) 64) Gygli 2019 38) Kambli 2015 55 Table 45. RIF MICs for rpoB S450 (S531) mutants in MGIT. The green line denotes the current WHO and CLSI CC for RIF DST in MGIT (1 mg/L). Notable limitations: studies RF44, RF39, RF61 and RF54; studies RF57 and RF63; studies RF39 and RF56; studies RF53 and RF39; and studies RF50 and RF60 were conducted in the same laboratory and study RF59 had systematically lower MIC distributions for all mutants tested and was, therefore, excluded from calculations in this report. rpoB borderline RRDR mutants Seventeen studies were identified that reported RIF MICs for 181 isolates that harbored rpoB RRDR borderline mutations and were tested by MGIT (Table 46). Excluding the study by Mvelase et al. and based on the current RIF CC on MGIT, the majority, or 116 (74%, 95% CI 67-81%) of 156 mutated isolates were phenotypically RIF-S. If the CC was lowered one dilution to 0.5 mg/L, at least 83 (53%, 95% CI 45-61%) of the 156 mutants would still test phenotypically RIF-S by MGIT. If the CC was lowered two dilutions to 0.25 mg/L, at least 55 (35%, 95% CI 28-43%) mutants would still test phenotypically RIF-S by MGIT. An ATU at 0.25 mg/L would reduce this percentage to 11% (95% CI 6-17%). Studies Lab Isolate origin Unique isolates Genotypic results 0.12 0.25 0.5 1 2 4 5 8 10 20 32 40 50 64 80 100 128 160 320 640 43) Abanda 2018 37 clinical 1 rpoB S450W (S531W) 1 3 clinical 1 rpoB S450F (S531F) 1 3 clinical 12 rpoB S450L (S531L) 12 3 clinical 1 rpoB S450L (S531L) 1 42 clinical 19 rpoB S450L (S531L) 6 11 2 42 clinical 3 rpoB S450W (S531W) 3 21 3 rpoB S450L (S531L) 1 1 1 21 1 rpoB S450P (S531P) 1 60) Torrea 2019 1 clinical 5 rpoB S450L (S531L) 5 3 clinical 324 rpoB S450L (S531L) 324 3 clinical 6 rpoB S450W (S531W) 6 45 clinical 1 rpoB S450C (S531C) 1 45 clinical 1 rpoB S450F (S531F) 1 45 clinical 3 rpoB S450W (S531W) 3 45 clinical 6 rpoB S450L (S531L) 6 38) Kambli 2015 30 clinical 57 rpoB S450L (S531L) 57 50) Van Deun 1 clinical 1 rpoB S450L (S531L) 1 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB S450W (S531W) 1 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB S450P (S531P) 1 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB S450M (S531M) 1 3, 5, 9, 12, 29, 31-34 clinical 101 rpoB S450L (S531L) 1 2 98 40 clinical 14 rpoB S450L (S531L) 14 40 clinical 5 rpoB S450W (S531W) 5 53) Machado 2018 5 clinical 31 rpoB S450L (S531L) 22 1 1 5 2 3 clinical 40 rpoB S450L (S531L) 40 3 clinical 1 rpoB S450W (S531W) 1 31 135 rpoB S450L (S531L) 1 2 14 26 92 31 4 rpoB S450W (S531W) 4 57) Lin - unpublished 45 1 rpoB S450Q (S531Q) 1 64) Gygli 2019 21 clinical 16 rpoB S450L 16 RIF MIC [mg/L] 39) Cambau 2015 54) Andres 2014 & Andres (unpublished) 56) Whitfield 2018 47) El Maraachii 2015 44) Heyckendorf 2018 49) Jamieson 2014 59) Mvelase 2019 61) Tessema 2017 62) Berrada 2016 56 Table 46. RIF MICs for rpoB borderline RRDR mutants in MGIT. The green line denotes the current WHO and CLSI CC for RIF DST in MGIT (1 mg/L). Notable limitations: studies RF44, RF39, RF41, RF61 and RF54; studies RF39 and RF56; studies RF61 and RF39; and studies RF50 and RF60 were conducted in the same laboratory and study RF59 had systematically lower MIC distributions for all mutants tested and was, therefore, excluded from calculations in this report. Other rpoB RRDR mutants Fifteen studies were identified that reported RIF MICs for 402 isolates with rpoB RRDR mutations other than S450 (S531) and the borderline mutations (Table 47). Excluding Mvelase et al. and based on the current RIF CC on MGIT, 11 (3%, 95% CI 2-5%) of 364 isolates had MICs below the current CC. If the CC were set at 0.5 mg/L, 5 (1%, 95% CI 0-3%) of these isolates would be classified as RIF-S. If the CC were set at 0.25 mg/L, only 2 (1%, 95% CI 0-2%) isolates would be classified as RIF-S, which would be the same with an ATU at 0.25 mg/L. Studies Lab Isolate origin Unique isolates Genotypic results 0.12 0.25 0.5 1 2 2.5 4 5 8 10 20 50 100 120 40) Ho 2013 35 clinical 3 rpoB L430P (L511P) 1 1 1 41) Feuerrigel 2012 3 clinical 1 rpoB L430P (L511P) 1 43) Abanda 2018 37 clinical 2 rpoB L430P (L511P) 2 49) Jamieson 2014 42 clinical 2 rpoB L430P (L511P) 2 59) Mvelase 2019 21 7 rpoB L430P (L511P) 3 3 1 60) Torrea 2019 1 clinical 10 rpoB L430P (L511P) 3 4 1 2 61) Tessema 2017 3 clinical 5 rpoB L430P (L511P) 4 1 63) Berrada 2016 45 clinical 3 rpoB L430P (L511P) 3 50) Van Deun (unpublished) 1 clinical 3 rpoB L430P (L511P) 3 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB L430P (L511P) 1 54) Andres 2014 & Andres (unpublished) 3 clinical 1 rpoB L430P (L511P) 1 56) Whitfield 2018 31 2 rpoB L430P (L511P) 2 41) Feuerrigel 2012 3 clinical 3 rpoB D435Y (D516Y) 1 2 42) Williamson 2012 36 clinical 1 rpoB D435Y (D516Y) 1 43) Abanda 2018 37 clinical 4 rpoB D435Y (D516Y) 4 44) Heyckendorf 2018 3 clinical 1 rpoB D435Y (D516Y) 1 59) Mvelase 2019 20 11 rpoB D435Y (D516Y) 2 4 2 3 60) Torrea 2019 1 clinical 9 rpoB D435Y (D516Y) 1 3 4 1 61) Tessema 2017 3 clinical 2 rpoB D435Y (D516Y) 2 63) Berrada 2016 45 clinical 4 rpoB D435Y (D516Y) 2 2 38) Kambli 2015 30 clinical 3 rpoB D435Y (D516Y) 1 1 1 50) Van Deun (unpublished) 1 clinical 4 rpoB D435Y (D516Y) 3 1 51) Sirgel 2013 & Sirgel 2015 43 clinical 1 rpoB D435Y (D516Y) 1 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB D435Y (D516Y) 1 47) El Maraachii 2015 40 clinical 2 rpoB D435Y (D516Y) 2 56) Whitfield 2018 31 12 rpoB D435Y (D516Y) 10 2 42) Williamson 2012 36 clinical 1 rpoB H445L (H526L) 1 43) Abanda 2018 37 clinical 2 rpoB H445L (H526L) 1 1 49) Jamieson 2014 42 clinical 2 rpoB H445L (H526L) 1 1 59) Mvelase 2019 21 5 rpoB H445L (H526L) 1 2 2 61) Tessema 2017 3 clinical 4 rpoB H445L (H526L) 2 2 63) Berrada 2016 45 clinical 4 rpoB H445L (H526L) 2 1 1 50) Van Deun (unpublished) 1 clinical 1 rpoB H445L (H526L) 1 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 3 rpoB H445L (H526L) 2 1 47) El Maraachii 2015 40 clinical 2 rpoB H445L (H526L) 1 1 64) Gygli 2019 21 clinical 1 rpoB H445L (H526L) 1 56) Whitfield 2018 31 3 rpoB H445L (H526L) 2 1 43) Abanda 2018 37 clinical 1 rpoB H445N (H526N) 1 49) Jamieson 2014 42 clinical 1 rpoB H445N (H526N) 1 61) Tessema 2017 3 clinical 6 rpoB H445N (H526N) 1 2 3 63) Berrada 2016 45 clinical 3 rpoB H445N (H526N) 2 1 38) Kambli 2015 30 clinical 3 rpoB H445N (H526N) 2 1 56) Whitfield 2018 31 4 rpoB H445N (H526N) 3 1 59) Mvelase 2019 21 2 rpoB H445S (H526S) 1 1 63) Berrada 2016 45 clinical 3 rpoB H445S (H526S) 1 1 1 50) Van Deun (unpublished) 1 clinical 1 rpoB H445S (H526S) 1 40) Ho 2013 35 clinical 1 rpoB L452P (L533P) 1 41) Feuerrigel 2012 3 clinical 1 rpoB L452P (L533P) 1 43) Abanda 2018 37 clinical 1 rpoB L452P (L533P) 1 44) Heyckendorf 2018 3 clinical 1 rpoB L452P (L533P) 1 60) Torrea 2019 1 clinical 8 rpoB L452P (L533P) 2 5 1 61) Tessema 2017 3 clinical 6 rpoB L452P (L533P) 6 63) Berrada 2016 45 clinical 2 rpoB L452P (L533P) 2 38) Kambli 2015 30 clinical 4 rpoB L452P (L533P) 1 3 50) Van Deun (unpublished) 1 clinical 2 rpoB L452P (L533P) 1 1 64) Gygli 2019 21 clinical 1 rpoB L452P (L533P) 1 56) Whitfield 2018 31 4 rpoB L452P (L533P) 3 1 RIF MIC [mg/L] 57 Table 47. RIF MICs for other rpoB RRDR mutants in MGIT. The green line denotes the current WHO and CLSI CC for RIF DST in MGIT (1 mg/L). Notable limitations: studies RF44, RF39, RF61 and RF54; studies RF39 and RF56; studies RF53 and RF39; and studies RF57 and RF63 were conducted in the same laboratory and study RF58 had systematically lower RIF MICs for mutants than reported for other studies and study RF59 had systematically lower MIC distributions for all mutants tested and was, therefore, excluded from calculations in this report. rpoB mutants outside the RRDR Fourteen isolates from 6 studies harbored rpoB mutations outside the RRDR (Table 48). The most frequently reported mutation, rpoB I491F (I572F), had an MIC distribution with an upper end that Studies Lab Isolate origin Unique isolates Genotypic results 0.12 0.25 0.5 1 2 4 5 8 10 15 20 24 50 80 100 160 320 47) El Maraachii 2015 40 clinical 1 rpoB S428R (S509R) 1 59) Mvelase 2019 20 1 rpoB S428T (S509T) 1 63) Berrada 2016 45 clinical 2 rpoB Q432E (Q513E) 2 61) Tessema 2017 3 clinical 2 rpoB Q432K (Q513K) 2 63) Berrada 2016 45 clinical 3 rpoB Q432K (Q513K) 3 47) El Maraachii 2015 40 clinical 4 rpoB Q432K (Q513K) 4 54) Andres 2014 & Andres (unpublished) 3 clinical 1 rpoB Q432K (Q513K) 1 56) Whitfield 2018 31 3 rpoB Q432K (Q513K) 1 2 63) Berrada 2016 45 clinical 1 rpoB Q432L (Q513L) 1 56) Whitfield 2018 31 2 rpoB Q432L (Q513L) 2 43) Abanda 2018 37 clinical 2 rpoB Q432P (Q513P) 2 59) Mvelase 2019 20 21 rpoB Q432P (Q513P) 5 10 5 1 61) Tessema 2017 3 clinical 2 rpoB Q432P (Q513P) 2 63) Berrada 2016 45 clinical 2 rpoB Q432P (Q513P) 2 56) Whitfield 2018 31 2 rpoB Q432P (Q513P) 1 1 38) Kambli 2015 30 clinical 1 rpoB D435A (D516A) 1 56) Whitfield 2018 31 1 rpoB D435A (D516A) 1 63) Berrada 2016 45 clinical 1 rpoB D435F (D516F) 1 38) Kambli 2015 30 clinical 1 rpoB D435F (D516F) 1 51) Sirgel 2013 & Sirgel 2015 43 clinical 4 rpoB D435S (D516S) 2 2 44) Heyckendorf 2018 3 clinical 1 rpoB D435V (D516V) 1 49) Jamieson 2014 42 clinical 3 rpoB D435V (D516V) 1 2 61) Tessema 2017 3 clinical 26 rpoB D435V (D516V) 2 12 12 59) Mvelase 2019 21 15 rpoB D435V (D516V) 1 2 3 2 6 1 63) Berrada 2016 45 clinical 18 rpoB D435V (D516V) 3 15 38) Kambli 2015 30 clinical 3 rpoB D435V (D516V) 3 51) Sirgel 2013 & Sirgel 2015 43 clinical 29 rpoB D435V (D516V) 4 15 10 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 12 rpoB D435V (D516V) 2 9 1 47) El Maraachii 2015 40 clinical 15 rpoB D435V (D516V) 15 53) Machado 2018 and Machado (unpublished) 5 clinical 2 rpoB D435V (D516V) 2 54) An res 2014 & Andres (unpublished) 3 clinical 1 rpoB D435V (D516V) 1 64) Gygli 2019 21 clinical 2 rpoB D435V (D516V) 2 56) Whitfield 2018 31 79 rpoB D435V (D516V) 4 14 32 18 9 1 1 57) Lin - unpublished 45 1 rpoB N438K (N519K) 1 49) Jamieson 2014 42 clinical 3 rpoB S441L (S522L) 3 61) Tessema 2017 3 clinical 1 rpoB S441L (S522L) 1 63) Berrada 2016 45 clinical 1 rpoB S441L (S522L) 1 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB S441L (S522L) 1 47) El Maraachii 2015 40 clinical 1 rpoB S441L (S522L) 1 56) Whitfield 2018 31 1 rpoB S441L (S522L) 1 57) Lin - unpublished 45 1 rpoB S441T (S522T) 1 63) Berrada 2016 45 clinical 1 rpoB H445A (H526A) 1 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB H445A (H526A) 1 57) Lin - unpublished 45 1 rpoB H445A (H526A) 1 63) Berrada 2016 45 clinical 2 rpoB H445C (H526C) 1 1 38) Kambli 2015 30 clinical 1 rpoB H445C (H526C) 1 49) Jamieson 2014 42 clinical 1 rpoB H445D (H526D) 1 61) Tessema 2017 3 clinical 10 rpoB H445D (H526D) 10 63) Berrada 2016 45 clinical 3 rpoB H445D (H526D) 3 38) Kambli 2015 30 clinical 7 rpoB H445D (H526D) 7 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 6 rpoB H445D (H526D) 6 47) El Maraachii 2015 40 clinical 2 rpoB H445D (H526D) 2 53) Machado 2018 and Machado (unpublished) 5 clinical 1 rpoB H445D (H526D) 1 54) An res 2014 & Andres (unpublished) 3 clinical 2 rpoB H445D (H526D) 2 64) Gygli 2019 21 clinical 5 rpoB H445D (H526D) 5 56) Whitfield 2018 31 18 rpoB H445D (H526D) 1 17 63) Berrada 2016 45 clinical 1 rpoB H445G (H526G) 1 61) Tessema 2017 3 clinical 1 rpoB H445P (H526P) 1 57) Lin - unpublished 45 1 rpoB H445P (H526P) 1 57) Lin - unpublished 45 1 rpoB H445Q (H526Q) 1 43) Abanda 2018 37 clinical 3 rpoB H445R (H526R) 3 59) Mvelase 2019 20 1 rpoB H445R (H526R) 1 61) Tessema 2017 3 clinical 5 rpoB H445R (H526R) 5 63) Berrada 2016 45 clinical 2 rpoB H445R (H526R) 2 38) Kambli 2015 30 clinical 1 rpoB H445R (H526R) 1 47) El Maraachii 2015 40 clinical 2 rpoB H445R (H526R) 2 56) Whitfield 2018 31 3 rpoB H445R (H526R) 1 2 49) Jamieson 2014 42 clinical 2 rpoB H445Y (H526Y) 1 1 61) Tessema 2017 3 clinical 15 rpoB H445Y (H526Y) 15 63) Berrada 2016 45 clinical 1 rpoB H445Y (H526Y) 1 38) Kambli 2015 30 clinical 3 rpoB H445Y (H526Y) 3 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 4 rpoB H445Y (H526Y) 1 3 47) El Maraachii 2015 40 clinical 3 rpoB H445Y (H526Y) 3 53) Machado 2018 and Machado (unpublished) 5 clinical 3 rpoB H445Y (H526Y) 1 1 1 64) Gygli 2019 21 clinical 2 rpoB H445Y (H526Y) 2 56) Whitfield 2018 31 13 rpoB H445Y (H526Y) 1 12 54) Andres 2014 & Andres (unpublished) 3 clinical 1 rpoB R448K (R529K) 1 RIF MIC [mg/L] 58 overlapped the CC for MGIT, whereas another mutation that was reported in multiple studies, rpoB V170F (V146F), had MICs at least two dilutions higher than the current RIF CC for MGIT. At the current CC (i.e. 1 mg/L), all 10 (100%, 95% CI 69-100%) I491F (I572F) mutants would be classified as RIF-S. If the RIF CC were set at 0.5 mg/L, at least 8 (80%, 95% CI 44-97%) I491F (I572F) mutants would be classified as RIF-S. If the RIF CC were set at 0.25 mg/L, at least 7 (70%, 95% CI 35-93%) I491F (I572F) mutants would be classified as RIF-S. With an ATU at 0.25 mg/L, it would be at least 7% (95% CI 0- 34%). Table 48. RIF MICs for rpoB mutants outside the RRDR in MGIT. The green line denotes the current WHO and CLSI CC for RIF DST in MGIT (1 mg/L). Notable limitations: studies RF50 and RF60; studies RF39 and RF54 were conducted in the same laboratory. 3.A.4.3 Conclusion for RIF CC for MGIT The TEG agreed that 1 mg/L is higher than the ECOFF for MGIT (Table 44). The CC was consequently lowered to 0.5 mg/L given that only a single isolate (i.e. 0.2%, 95% CI 0.0-1.1%) would be misclassified as RIF-R at this concentration. Notably, this change can be adopted easily in routine clinical practice by reconstituting the lyophilised RIF from the BD SIRE kit in 8 mL, instead of 4 mL, of sterile distilled/deionised water. 85 This off-label concentration for MGIT would reduce the rate of misclassification of borderline RRDR mutations from 74% (95% CI 67-81%) to 53% (95% CI 45-61%), which could be minimized further by adopting an ATU and/or lowering the CC by one more dilution (Table 49). However, there was a consensus that more MIC data for pWT isolates, ideally from studies meeting EUCAST standards, are needed urgently to investigate whether the CC can be lowered to 0.25 mg/L without resulting in high rate of misclassification of susceptible isolates. 86 An ATU was not adopted, in line with the decision for other media (see Section 3.3 for more details). 85 BD. BBL™ MGIT™ AST SIRE System for the antimycobacterial susceptibility testing of Mycobacterium tuberculosis. 88-2041- 1JAA(04). 2019-09. (https://www.bd.com/resource.aspx?IDX=18298, accessed 13 September 2020). 86 Schön T, Köser CU, Werngren J, et al. What is the role of the EUCAST reference method for MIC testing of the Mycobacterium tuberculosis complex? Clin Microbiol Infect. 2020;26(11):1453-1455. doi:10.1016/j.cmi.2020.07.037. Studies Lab Isolate origin Unique isolates Genotypic results 0.12 0.25 0.5 1 2 4 8 20 50 49) Jamieson 2014 42 clinical 1 rpoB V170F (V146F) 1 39) Cambau 2015 3, 5, 9, 12, 29, 31-34 clinical 1 rpoB V170F (V146F) 1 54) Andres 2014 & Andres (unpublished) 3 clinical 1 rpoB V170F (V146F) 1 57) Lin - unpublished 45 1 rpoB V170F (V146F) 1 60) Torrea 2019 1 clinical 8 rpoB I491F (I572F) 1 5 1 1 50) Van Deun (unpublished) 1 clinical 1 rpoB I491F (I572F) 1 57) Lin - unpublished 45 1 rpoB I491F (I572F) 1 RIF MIC [mg/L] 59 Table 49. Effect of changing the CC for RIF DST on MGIT for detection of rpoB mutants. Mutation type Percentage of mutants classified RIF-S at CC 1 mg/L CC+ATU 1 mg/L or CC 0.5 mg/L CC+ATU 0.5 mg/L or CC 0.25 mg/L CC+ATU 0.25 mg/L S450 (S531) 0% (95% CI 0-1%) 0% (95% CI 0-1%) 0% (95% CI 0-1%) 0% (95% CI 0-1%) borderline RRDR 74% (95% CI 67-81%) 53% (95% CI 45-61%) 35% (95% CI 28-43%) 11% (95% CI 6-17%) other RRDR 3% (95% CI 2-5%) 1% (95% CI 0-3%) 1% (95% CI 0-2%) 1% (95% CI 0-2%) V170F (V146F) 0% (95% CI 0-60%) borderline I491F (I572F) 100% (95% CI 69-100%) 80% (95% CI 44-97%) 70% (95% CI 35-93%) 7% (95% CI 0-34%) The newly adopted CC is underlined. 60 3.A.5 References for RIF MIC studies 1. Vincent V, Rigouts L, Nduwamahoro E, et al. The TDR Tuberculosis Strain Bank: a resource for basic science, tool development and diagnostic services. Int J Tuberc Lung Dis. 2012;16(1):24-31. doi:10.5588/ijtld.11.0223. 2. Schönfeld N, Bergmann T, Vesenbeckh S, et al. Minimal inhibitory concentrations of first-line drugs of multidrug-resistant tuberculosis isolates. Lung India. 2012;29(4):309-12. doi:10.4103/0970- 2113.102794. 3. Andres S, Hillemann D, Rüsch-Gerdes S, Richter E. Occurrence of rpoB mutations in isoniazid-resistant but rifampin-susceptible Mycobacterium tuberculosis isolates from Germany. Antimicrob Agents Chemother. 2014;58(1):590-592. doi:10.1128/AAC.01752-13. 4. Fabry W, Schmid EN, Ansorg R. Comparison of the E test and a proportion dilution method for susceptibility testing of Mycobacterium tuberculosis. Zentralbl Bakteriol. 1995;282(4):394-401. doi:10.1016/s0934-8840(11)80710-3. 5. (a) Rigouts L, Gumusboga M, de Rijk WB, et al. Rifampin resistance missed in automated liquid culture system for Mycobacterium tuberculosis isolates with specific rpoB mutations. J Clin Microbiol. 2013;51(8):2641-2645. doi:10.1128/JCM.02741-12. (b) Van Deun A, Barrera L, Bastian I, et al. Mycobacterium tuberculosis strains with highly discordant rifampin susceptibility test results. J Clin Microbiol. 2009;47(11):3501-3506. doi:10.1128/JCM.01209-09 (c) Van Deun A. Personal communication. 2018. 6. Jagielski T, Bakuła Z, Brzostek A, et al. Characterization of mutations conferring resistance to rifampin in Mycobacterium tuberculosis clinical strains. Antimicrob Agents Chemother. 2018;62(10):e01093-18. doi:10.1128/AAC.01093-18. 7. Cavusoglu C, Turhan A, Akinci P, Soyler I. Evaluation of the Genotype MTBDR assay for rapid detection of rifampin and isoniazid resistance in Mycobacterium tuberculosis isolates. J Clin Microbiol. 2006;44(7):2338-2342. doi:10.1128/JCM.00425-06. 8. Schönfeld N, Bergmann T, Vesenbeckh S, et al. Minimal inhibitory concentrations of first-line drugs of multidrug-resistant tuberculosis isolates. Lung India. 2012;29(4):309-12. doi:10.4103/0970- 2113.102794. 9. van Ingen J, Simons S, de Zwaan R, et al. Comparative study on genotypic and phenotypic second-line drug resistance testing of Mycobacterium tuberculosis complex isolates. J Clin Microbiol. 2010;48(8):2749-2753. doi:10.1128/JCM.00652-10. 10. van Klingeren B, Dessens-Kroon M, van der Laan T, Kremer K, van Soolingen D. Drug susceptibility testing of Mycobacterium tuberculosis complex by use of a high-throughput, reproducible, absolute concentration method. J Clin Microbiol. 2007;45(8):2662-2668. doi:10.1128/JCM.00244-07. 11. Heifets L, Sanchez T, Vanderkolk J, Pham V. Development of rifapentine susceptibility tests for Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1999;43(1):25-28. doi:10.1128/AAC.43.1.25. 12. Moghazeh SL, Pan X, Arain T, et al. Comparative antimycobacterial activities of rifampin, rifapentine, and KRM-1648 against a collection of rifampin-resistant Mycobacterium tuberculosis isolates with known rpoB mutations. Antimicrob Agents Chemother. 1996;40(11):2655-2657. doi:10.1128/AAC.40.11.2655. 13. Rancoita PM V., Cugnata F, Gibertoni Cruz AL, et al. Validating a 14-drug microtiter plate containing bedaquiline and delamanid for large-scale research susceptibility testing of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2018;62(9):e00344-18. doi:10.1128/AAC.00344-18. 14. Kusunoki S, Takemura K, Takagi K, et al. A new genetic assay for rifampicin susceptibility of Mycobacterium tuberculosis. Kansenshogaku Zasshi. 1995;69(12):1342-1347. doi:10.11150/kansenshogakuzasshi1970.69.1342. 15. Miotto P, Cabibbe AM, Borroni E, Degano M, Cirillo DM. Role of disputed mutations in the rpoB gene in interpretation of automated liquid MGIT culture results for rifampin susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol. 2018;56(5):e01599-17. doi:10.1128/JCM.01599-17. 16. Park S-W, Tasneen R, Converse PJ, Nuermberger EL. Immunodeficiency and intermittent dosing promote acquired rifamycin monoresistance in murine tuberculosis. Antimicrob Agents Chemother. 2017;61(11):e01502-17. doi:10.1128/AAC.01502-17. 17. (a) de Steenwinkel JEM, ten Kate MT, de Knegt GJ, et al. Consequences of noncompliance for therapy efficacy and emergence of resistance in murine tuberculosis caused by the Beijing genotype of 61 Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2012;56(9):4937-4944. doi:10.1128/AAC.00124-12. (b) de Steenwinkel JEM, ten Kate MT, de Knegt GJ, et al. Drug susceptibility of Mycobacterium tuberculosis Beijing genotype and association with MDR TB. Emerg Infect Dis. 2012;18(4):660-663. doi:10.3201/eid1804.110912. (c) Bax HI, Bakker-Woudenberg IAJM, de Vogel CP, et al. The role of the time-kill kinetics assay as part of a preclinical modeling framework for assessing the activity of anti-tuberculosis drugs. Tuberculosis. 2017;105:80-85. doi:10.1016/j.tube.2017.04.010. 18. Cavusoglu C, Karaca-Derici Y, Bilgic A. In-vitro activity of rifabutin against rifampicin-resistant Mycobacterium tuberculosis isolates with known rpoB mutations. Clin Microbiol Infect. 2004;10(7):662- 665. doi:10.1111/j.1469-0691.2004.00917.x. 19. Schön T, Juréen P, Chryssanthou E, et al. Rifampicin-resistant and rifabutin-susceptible Mycobacterium tuberculosis strains: a breakpoint artefact? J Antimicrob Chemother. 2013;68(9):2074-2077. doi:10.1093/jac/dkt150. 20. Schön T, Juréen P, Giske CG, et al. Evaluation of wild-type MIC distributions as a tool for determination of clinical breakpoints for Mycobacterium tuberculosis. J Antimicrob Chemother. 2009;64(4):786-793. doi:10.1093/jac/dkp262. 21. Garcia A-B, Palacios JJ, Ruiz M-J, et al. Strong in vitro activities of two new rifabutin analogs against multidrug-resistant Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2010;54(12):5363- 5365. doi:10.1128/AAC.00149-10. 22. Marubini E. Evaluation of Mycobacterium tuberculosis clinical isolates with discordant rifampicin genotypic and phenotypic susceptibility testing results from Pretoria, South Africa. 2016. https://repository.up.ac.za/bitstream/handle/2263/56939/Marubini_Evaluation_2016.pdf. 23. Pholwat S, Heysell S, Stroup S, Foongladda S, Houpt E. Rapid first- and second-line drug susceptibility assay for Mycobacterium tuberculosis isolates by use of quantitative PCR. J Clin Microbiol. 2011;49(1):69-75. doi:10.1128/JCM.01500-10. 24. Mvelase NR, Pillay M, Sibanda W, et al. rpoB mutations causing discordant rifampicin susceptibility in Mycobacterium tuberculosis: Retrospective analysis of prevalence, phenotypic, genotypic, and treatment outcomes. Open Forum Infect Dis. 2019;6(4):ofz065. doi:10.1093/ofid/ofz065. 25. Farhat MR, Freschi L, Calderon R, et al. GWAS for quantitative resistance phenotypes in Mycobacterium tuberculosis reveals resistance genes and regulatory regions. Nat Commun. 2019;10(1):2128. doi:10.1038/s41467-019-10110-6. 26. Gygli SM, Keller PM, Ballif M, et al. Whole-genome sequencing for drug resistance profile prediction in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2019;63(4):e02175-18. doi:10.1128/AAC.02175-18. 27. Rastogi N, Goh KS, Berchel M, Bryskier A. Activity of rifapentine and its metabolite 25-O- desacetylrifapentine compared with rifampicin and rifabutin against Mycobacterium tuberculosis, Mycobacterium africanum, Mycobacterium bovis and M. bovis BCG. J Antimicrob Chemother. 2000;46(4):565-570. doi:10.1093/jac/46.4.565. 28. Fattorini L, Iona E, Ricc ML, et al. Activity of 16 antimicrobial agents against drug-resistant strains of Mycobacterium tuberculosis. Microb Drug Resist. 1999;5(4):265-270. doi:10.1089/mdr.1999.5.265. 29. Coban AY. Blood agar validation for susceptibility testing of isoniazid, rifampicin, ethambutol, and streptomycin to Mycobacterium tuberculosis isolates. PLoS One. 2013;8(2):e55370. doi:10.1371/journal.pone.0055370. 30. (a) Rey-Jurado E, Tudó G, de la Bellacasa JP, Espasa M, González-Martín J. In vitro effect of three-drug combinations of antituberculous agents against multidrug-resistant Mycobacterium tuberculosis isolates. Int J Antimicrob Agents. 2013;41(3):278-280. doi:10.1016/j.ijantimicag.2012.11.011. (b) Rey-Jurado E, Tudó G, Martínez JA, González-Martín J. Synergistic effect of two combinations of antituberculous drugs against Mycobacterium tuberculosis. Tuberculosis. 2012;92(3):260-263. doi:10.1016/j.tube.2012.01.005. 31. Shishido Y, Mitarai S, Otomo K, et al. Anti-tuberculosis drug susceptibility testing of Mycobacterium bovis BCG Tokyo strain. Int J Tuberc Lung Dis. 2007;11(12):1334-1338. 32. (a) Anthony RM, Schuitema AR, Bergval IL, et al. Acquisition of rifabutin resistance by a rifampicin resistant mutant of Mycobacterium tuberculosis involves an unusual spectrum of mutations and elevated frequency. Ann Clin Microbiol Antimicrob. 2005;4(1):9. doi:10.1186/1476-0711-4-9. 62 (b) Bergval IL, Klatser PR, Schuitema ARJ, Oskam L, Anthony RM. Specific mutations in the Mycobacterium tuberculosis rpoB gene are associated with increased dnaE2 expression. FEMS Microbiol Lett. 2007;275(2):338-343. doi:10.1111/j.1574-6968.2007.00905.x. 33. Rodriguez Díaz JC, Ruiz M, López M, Royo G. Synergic activity of fluoroquinolones and linezolid against Mycobacterium tuberculosis. Int J Antimicrob Agents. 2003;21(4):354-356. doi:10.1016/s0924- 8579(02)00387-4. 34. Pretet S, Lebeaut A, Parrot R et al. Combined chemotherapy including rifabutin for rifampicin and isoniazid resistant pulmonary tuberculosis. G.E.T.I.M. (Group for the Study and Treatment of Resistant Mycobacterial Infection). Eur Respir J. 1992;5(6):680-4. 35. Lee C-N, Heifets LB. Determination of minimal inhibitory concentrations of antituberculosis drugs by radiometric and conventional methods. Am Rev Respir Dis. 1987;136(2):349-352. doi:10.1164/ajrccm/136.2.349. 36. Heifets L, Sanchez T, Vanderkolk J, Pham V. Development of rifapentine susceptibility tests for Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1999;43(1):25-28. doi:10.1128/AAC.43.1.25. 37. Truffot-Pernot C, Giroir AM, Grosset J. A study of the minimal inhibitory concentration of rifabutin (ansamycin LM 427) for Mycobacterium tuberculosis, Mycobacterium xenopi and Mycobacterium avium-intracellulaire. Rev Mal Respir. 1988;5(4):401-406. 38. Kambli P, Ajbani K, Sadani M, et al. Defining multidrug-resistant tuberculosis: Correlating GenoType MTBDRplus assay results with minimum inhibitory concentrations. Diagn Microbiol Infect Dis. 2015;82(1):49-53. doi:10.1016/j.diagmicrobio.2015.01.009. 39. Cambau E, Viveiros M, Machado D, et al. Revisiting susceptibility testing in MDR-TB by a standardized quantitative phenotypic assessment in a European multicentre study. J Antimicrob Chemother. 2015;70(3):686-696. doi:10.1093/jac/dku438. 40. Ho J, Jelfs P, Sintchencko V. Phenotypically occult multidrug-resistant Mycobacterium tuberculosis: dilemmas in diagnosis and treatment. J Antimicrob Chemother. 2013;68(12):2915-2920. doi:10.1093/jac/dkt284. 41. Feuerriegel S, Oberhauser B, George AG, et al. Sequence analysis for detection of first-line drug resistance in Mycobacterium tuberculosis strains from a high-incidence setting. BMC Microbiol. 2012;12:90. doi:10.1186/1471-2180-12-90. 42. Williamson DA, Roberts SA, Bower JE, et al. Clinical failures associated with rpoB mutations in phenotypically occult multidrug-resistant Mycobacterium tuberculosis. Int J Tuberc Lung Dis. 2012;16(2):216-220. doi:10.5588/ijtld.11.0178. 43. Abanda NN, Djieugoué JY, Khadka VS, et al. Absence of hybridization with the wild-type and mutant rpoB probes in the Genotype MTBDRplus assay detects ‘disputed’ rifampicin mutations. Clin Microbiol Infect. 2018;24(7):781.e1-781.e3. doi:10.1016/j.cmi.2017.11.021. 44. Heyckendorf J, Andres S, Köser CU, et al. What is resistance? Impact of phenotypic versus molecular drug resistance testing on therapy for multi- and extensively drug-resistant Tuberculosis. Antimicrob Agents Chemother. 2017;62(2):e01550-17. doi:10.1128/AAC.01550-17. 45. Sanker P, Ambika AP, Santhosh VT, et al. Are WHO approved nucleic acid amplification tests causing large-scale "false identification" of rifampicin-resistant tuberculosis?: Programmatic experience from south india. Int J Mycobacteriol. 2017;6(1):21-26. doi:10.4103/2212-5531.201900. 46. (a) Rockwood N, Sirgel F, Streicher E, Warren R, Meintjes G, Wilkinson RJ. Low frequency of acquired isoniazid and rifampicin resistance in rifampicin-susceptible pulmonary tuberculosis in a setting of high HIV-1 infection and tuberculosis coprevalence. J Infect Dis. 2017;216(6):632-640. doi:10.1093/infdis/jix337. (b) Rockwood N, Pasipanodya JG, Denti P, et al. Concentration-dependent antagonism and culture conversion in pulmonary tuberculosis. Clin Infect Dis. 2017;64(10):1350-1359. doi:10.1093/cid/cix158. 47. El Maraachli W, Slater M, Berrada ZL, et al. Predicting differential rifamycin resistance in clinical Mycobacterium tuberculosis isolates by specific rpoB mutations. Int J Tuberc Lung Dis. 2015;19(10):1222-1226. doi:10.5588/ijtld.14.0936. 48. Chigutsa E, Pasipanodya JG, Visser ME, et al. Impact of nonlinear interactions of pharmacokinetics and MICs on sputum bacillary kill rates as a marker of sterilizing effect in tuberculosis. Antimicrob Agents Chemother. 2015;59(1):38-45. doi:10.1128/AAC.03931-14. 49. Jamieson FB, Guthrie JL, Neemuchwala A, et al. Profiling of rpoB mutations and MICs for rifampin and rifabutin in Mycobacterium tuberculosis. J Clin Microbiol. 2014;52(6):2157-2162. doi:10.1128/JCM.00691-14. 63 50. Van Deun A. Personal communication. 2018. 51. (a) Sirgel FA, Warren RM, Böttger EC, et al. The rationale for using rifabutin in the treatment of MDR and XDR tuberculosis outbreaks. PLoS One. 2013;8(3):e59414. doi:10.1371/journal.pone.0059414. (b) Sirgel FA, Warren RM, Böttger EC, et al. Correction: The rationale for using rifabutin in the treatment of MDR and XDR tuberculosis outbreaks. PLoS One. 2015. doi:10.1371/journal.pone.0134190. 52. Bernardelli A, Morcillo N, Loureiro J, Quse V, Davenport S. In vitro susceptibility testing of Mycobacterium tuberculosis complex strains isolated from seals to antituberculosis drugs. Biomedica. 2004;24 Supp 1:85-91. 53. (a) Machado D, Perdigão J, Portugal I, et al. Efflux activity differentially modulates the levels of isoniazid and rifampicin resistance among multidrug resistant and monoresistant Mycobacterium tuberculosis strains. Antibiotics. 2018;7(1):18. doi:10.3390/antibiotics7010018. (b) Machado D, Viveiros M. Personal Communication. 2018. 54. (a) Andres S, Hillemann D, Rüsch-Gerdes S, Richter E. Occurrence of rpoB mutations in isoniazid- resistant but rifampin-susceptible Mycobacterium tuberculosis isolates from Germany. Antimicrob Agents Chemother. 2014;58(1):590-592. doi:10.1128/AAC.01752-13. (b) Andres S. Personal communication. 2018. 55. Niward K, Davies Forsman L, Bruchfeld J, et al. Distribution of plasma concentrations of first-line anti- TB drugs and individual MICs: a prospective cohort study in a low endemic setting. J Antimicrob Chemother. 2018;73(10):2838-2845. doi:10.1093/jac/dky268. 56. Whitfield MG, Warren RM, Mathys V, et al. The potential use of rifabutin for treatment of patients diagnosed with rifampicin-resistant tuberculosis. J Antimicrob Chemother. 2018;73(10):2667-2674. doi:10.1093/jac/dky248. 57. Lin G. Personal communication. 2018. 58. Groenheit R. Personal communication. 2018. 59. Mvelase NR, Pillay M, Sibanda W, et al. rpoB mutations causing discordant rifampicin susceptibility in Mycobacterium tuberculosis: Retrospective analysis of prevalence, phenotypic, genotypic, and treatment outcomes. Open Forum Infect Dis. 2019;6(4):ofz065. doi:10.1093/ofid/ofz065. 60. Torrea G, Ng K, Van Deun A. Variable ability of rapid tests to detect Mycobacterium tuberculosis rpoB mutations conferring phenotypically occult rifampicin resistance. Sci Rep. 2019;9(1):11826. doi:10.1038/s41598-019-48401-z. 61. Tessema B, Nabeta P, Valli E, et al. FIND Tuberculosis Strain Bank: A resource for researchers and developers working on tests to detect Mycobacterium tuberculosis and related drug resistance. J Clin Microbiol. 2017;55(4):1066-1073. doi:10.1128/JCM.01662-16. 62. Colangeli R, Jedrey H, Kim S, et al. Bacterial factors that predict relapse after tuberculosis therapy. N Engl J Med. 2018;379(9):823-833. doi:10.1056/NEJMoa1715849. 63. Berrada ZL, Lin S-YG, Rodwell TC, et al. Rifabutin and rifampin resistance levels and associated rpoB mutations in clinical isolates of Mycobacterium tuberculosis complex. Diagn Microbiol Infect Dis. 2016;85(2):177-181. doi:10.1016/j.diagmicrobio.2016.01.019. 64. Gygli SM, Keller PM, Ballif M, et al. Whole-genome sequencing for drug resistance profile prediction in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2019;63(4):e02175-18. doi:10.1128/AAC.02175-18. 64 3.B.1 RFB MIC data on LJ 3.B.1.1 RFB MICs for pWT isolates on LJ No studies were identified that featured RFB MIC data for pWT isolates on LJ medium. 3.B.1.2 RFB MICs for mutated isolates on LJ No studies were identified that featured RFB MIC data for mutated isolates on LJ medium. 3.B.1.3 Conclusion for RFB CC for LJ Owing to a lack of data not even a tentative ECOFF could be defined for this medium. Even if sufficient MIC evidence had been available, a CC would not have been endorsed as RFB is not currently recommended for TB treatment by WHO. Refer to Section 3.3 for a strategy for RFB DST. 65 3.B.2 RFB MIC data on 7H10 3.B.2.1 RFB MICs for pWT isolates on 7H10 Seven studies were identified that reported RFB MIC data for the pWT population on 7H10 (Table 50). Five of these studies featured MIC data for more than 10 pWT isolates. Two of these studies had data truncations that precluded an assessment of the shape of the MIC distributions, including García et al. with unusually low MICs for pWT isolates, as previously seen for RIF (Table 35). The study was consequently excluded. The modes of the remaining studies, two of which were conducted in the same laboratory, ranged from <0.015 to 0.03 mg/L. 0.125 or 0.06 mg/L represent the tentative ECOFF. Table 50. RFB MICs for pWT isolates on 7H10. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). Notable limitations: studies RB3 and RB5 were conducted in the same laboratory, and study RB6 had unusually low MICs for pWT isolates. 3.B.2.2 RFB MICs for mutated isolates on 7H10 rpoB 450 (531) mutants Five studies were identified that reported 593 RFB MICs for isolates with rpoB S450 (S531) mutations on 7H10 (Table 51). The modes of the three untruncated distributions were 2-4 mg/L. If 0.125 mg/L was selected as the RFB CC for 7H10 only 23 isolates (4%, 95% CI 2-6%) would test RFB-S. At the CLSI CC (i.e. 0.5 mg/L), 41 (7%, 95% CI 5-9%) isolates would be classified as RFB-S. The effect of adopting 0.06 mg/L as a CC with or without an ATU could not be explored, as the lowest concentration tested by Farhat et al. was 0.125 mg/L. It is notable that the MIC distribution for S450 mutants in this study was unusually broad, which may suggest phenotypic differences depending on the genetic background. Table 51. RFB MICs for rpoB S450 (S531) mutants on 7H10. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). rpoB borderline RRDR mutants Four studies were identified that reported RFB MICs for isolates with rpoB borderline mutations tested on 7H10 (Table 52). The vast majority of MICs reported for these isolates were between 0.02-0.5 mg/L. Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic results - Mtb0.004 0.008 0.015 0.03 0.04 0.06 0.1 0.12 0.2 0.25 0.3 0.5 0.7 1 2 2.5 3 4 5 8 10 16 25 50 3 1 2 H37Rv ATCC25618 1 1 3 1 1 BTB 08-049 1 3 22 22 Mostly RIF-R gWT 1 7 9 2 1 1 1 3 1 4 H37Rv ATCC27294 1 2 1 3 1 1 BTB 08-049 1 3 clinical 99 99 8 41 25 23 1 1 1 1 2 H37Rv 2 1 17 17 gWT 7 5 5 7 1 3 H37Rv ATCC27294 1 1 1 7 1 3 Erdman 1 1 1 7 clinical 76 80 mixture of first-line resistance profiles gWT 6 24 28 21 1 5 reference 16 48 30 15 3 5 clinical 63 189 138 45 6 2) Cavusoglu 2004 2 1 1 H37Rv ATCC27294 1 7) van Ingen 2010 4 1 1 H37Rv 1 RFB MIC (mg/L) 3) Schön 2013 6) García 2010 9) Gygli 2019 1) Rancoita 2018 5) Schön 2011 Studies Lab Unique isolates Total MICs Genotypic results 0.06 0.12 0.25 0.5 0.75 0.83 1 1.5 2 2.5 4 5 8 16 1) Rancoita 2018 1 10 10 rpoB S450L (S531L) 2 2 4 2 2 19 19 rpoB S450L (S531L) 1 1 1 16 2 4 4 rpoB S450W (S531W) 4 3 46 46 rpoB S450L (S531L) 1 3 9 17 12 3 1 3 1 1 rpoB S450W (S531W) 1 9) Gygli 2019 7 43 45 rpoB S450L (S531L) 3 28 13 1 6 12 12 rpoB S450F (S531F) 12 6 440 440 rpoB S450L (S531L) 19 3 8 5 10 11 36 75 256 6 4 4 rpoB S450M (S531M) 1 3 6 5 5 rpoB S450Q (S531Q) 1 1 1 1 6 7 7 rpoB S450W (S531W) 1 1 5 RFB MIC [mg/L] 2) Cavusoglu 2004 3) Schön 2013 8) Farhat 2019 66 At the CLSI CC (i.e. 0.5 mg/L), 56 (92%, 95% CI 82-97%) isolates would be classified as RFB-S. If a CC were set at 0.125 mg/L, at least 41 of 61 (67%, 95% CI 54-79%) isolates would be classified as RFB-S. The consequence of a lower CC or an ATU could not be assessed given the severe truncations of most MICs. Table 52. RFB MICs for rpoB borderline RRDR mutants on 7H10. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). Other rpoB RRDR mutants Four studies were identified that reported 246 RFB MICs for isolates with rpoB mutations other than S450 (S531) and the borderline mutations on 7H10 (Table 53). The D435V (D516V) mutants had low MICs, overall, with the majority, or 129 of 159 (81% 95% CI 74-87%) having MICs lower than 0.125 mg/L. At the CLSI CC (i.e. 0.5 mg/L), 147 (60%, 95% CI 53-66%) isolates would be classified as RFB-S. If a CC were set at 0.125 mg/L, 95 (39%, 95% CI 33-45%) of the total isolates would be classified as RFB- S. Given that most MICs at 0.125 mg/L were truncated, the effect of an ATU at this concentration or a CC at 0.06 mg/L (with or without an ATU) could not be calculated. However, a CC and ATU at 0.06 mg/L would eliminate any mutants testing RFB-S for untruncated datasets. Table 53. RFB MICs for other rpoB RRDR mutants on 7H10. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). Studies Isolate origin Unique isolates Total MICs Genotypic results 0.02 0.03 0.06 0.12 0.2 0.25 0.5 0.75 0.83 1 1.5 2 2.5 3 4 3) Schön 2013 1 1 rpoB L430P (L511P) 1 8) Farhat 2019 clinical 3 3 rpoB L430P (L511P) 3 3) Schön 2013 3 3 rpoB D435Y (D516Y) 1 1 1 2) Cavusoglu 2004 2 2 rpoB D435Y (D516Y) 2 8) Farhat 2019 clinical 11 11 rpoB D435Y (D516Y) 11 4) van Ingen 2011 clinical 4 7 rpoB D435Y (D516Y) 4 3 3) Schön 2013 1 1 rpoB H445L (H526L) 1 9) Gygli 2019 clinical 1 1 rpoB H445L (H526L) 1 8) Farhat 2019 clinical 13 13 rpoB H445L (H526L) 12 1 8) Farhat 2019 clinical 3 3 rpoB H445N (H526N) 2 1 8) Farhat 2019 clinical 1 1 rpoB H445S (H526S) 1 3) Schön 2013 1 1 rpoB L452P (L533P) 1 9) Gygli 2019 clinical 2 3 rpoB L452P (L533P) 1 1 1 2) Cavusoglu 2004 1 1 rpoB L452P (L533P) 1 8) Farhat 2019 clinical 10 10 rpoB L452P (L533P) 6 1 2 RFB MIC [mg/L] Studies Unique isolates Total MICs Genotypic results 0.06 0.12 0.25 0.5 0.6 0.7 0.75 0.83 1 1.5 2 2.5 4 5 8 16 32 8) Farhat 2019 2 2 rpoB Q432E (Q513E) 1 1 8) Farhat 2019 1 1 rpoB D435F (Q513F) 1 8) Farhat 2019 3 3 rpoB Q432K (Q513K) 3 2) Cavusoglu 2004 1 1 rpoB Q432P (Q513P) 1 8) Farhat 2019 2 2 rpoB Q432P (Q513P) 1 1 2) Cavusoglu 2004 1 1 rpoB D435V (D516V) 1 3) Schön 2013 9 9 rpoB D435V (D516V) 2 4 3 9) Gygli 2019 18 20 rpoB D435V (D516V) 5 7 3 3 1 1 8) Farhat 2019 129 129 rpoB D435V (D516V) 70 28 6 8 5 1 2 9 8) Farhat 2019 3 3 rpoB S441L (S522L) 2 1 2) Cavusoglu 2004 2 2 rpoB S441W (S522W) 2 2) Cavusoglu 2004 1 1 rpoB H445C (H526C) 1 8) Farhat 2019 7 7 rpoB H445C (H526C) 7 2) Cavusoglu 2004 1 1 rpoB H445D (H526D) 1 3) Schön 2013 3 3 rpoB H445D (H526D) 1 1 1 9) Gygli 2019 7 9 rpoB H445D (H526D) 1 3 4 1 8) Farhat 2019 7 7 rpoB H445D (H526D) 1 6 8) Farhat 2019 1 1 rpoB H445G (H526G) 1 8) Farhat 2019 1 1 rpoB H445P (H526P) 1 2) Cavusoglu 2004 4 4 rpoB H445R (H526R) 4 8) Farhat 2019 10 10 rpoB H445R (H526R) 10 2) Cavusoglu 2004 2 2 rpoB H445Y (H526Y) 2 3) Schön 2013 3 3 rpoB H445Y (H526Y) 1 2 9) Gygli 2019 5 6 rpoB H445Y (H526Y) 2 3 1 8) Farhat 2019 16 16 rpoB H445Y (H526Y) 1 15 8) Farhat 2019 2 2 rpoB R448K (R529K) 2 RFB MIC [mg/L] 67 rpoB mutants outside the RRDR Only 2 studies featured RFB MIC data for isolates with rpoB mutations outside of the RRDR on 7H10 medium (Table 54). 67% (95% CI, 22-86%) of I491F (I572F) mutants would test susceptible at the CLSI CC of 0.5 mg/L. Lowering the CC to 0.125 mg/L would reduce this to 50% (95% CI, 12-88%). Lower concentrations could not be explored. Table 54. RFB MICs for rpoB mutants outside the RRDR on 7H10. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). 3.B.2.3 Conclusion for RFB CC for 7H10 The three well-designed studies by Gygli et al. and Schön et al. (Table 50) provided a good insight into the pWT MIC distribution, but was not clear whether 0.125 mg/L or 0.06 mg/L was the likely tentative ECOFF. Notably, these results were in line with a study by Heifets et al. that did not meet the inclusion criteria for this review because it was not clear precisely which concentrations were tested.87 The effect of adopting an ATU at 0.125 mg/L or a CC at 0.06 mg/L (with or without an ATU) could not be calculated with confidence because insufficiently low concentrations were tested for most mutants. However, neither 0.125 nor 0.06 mg/L were adopted as CCs given that WHO does not currently recommend RFB for TB treatment by WHO. Refer to Section 3.3 for a strategy for RFB DST. Table 55. Effect of adopting a CC for RFB DST on 7H10 for detection of rpoB mutants. Mutation type Percentage of mutants classified RFB-S at CC 0.5 mg/L CC 0.125 mg/L S450 (S531) 7% (95% CI 5-9%) 4% (95% CI 2-6%) borderline RRDR 92% (95% CI 82-97%) 67% (95% CI 54-79%) other RRDR 60% (95% CI 53-66%) 39% (95% CI 33-45%) V170F (V146F) 0% (95% CI 0-46%) borderline I491F (I572F) 67% (95% CI 22-86%) 50% (95% CI 12-88%) The current CLSI CC is underlined. 87 Heifets LB, Lindholm-Levy PJ, Iseman MD. Rifabutine: minimal inhibitory and bactericidal concentrations for Mycobacterium tuberculosis. Am Rev Respir Dis. 1988;137(3):719-721. doi:10.1164/ajrccm/137.3.719. Studies Unique isolates Genotypic results 0.12 0.25 0.5 0.6 0.75 0.83 1 1.5 2 2.5 3 9) Gygli 2019 1 rpoB V170F (V146F) 1 5 rpoB V170F (V146F) 1 2 2 6 rpoB I491F (I572F) 3 1 1 1 RFB MIC [mg/L] 8) Farhat 2019 68 3.B.3 RFB MIC data on 7H11 3.B.3.1 RFB MICs for pWT isolates on 7H11 Three studies were identified that reported RFB MIC data for the pWT population by 7H11 (Table 56). Only one study reported MICs for at least 10 pWT isolates. The RFB MIC distributions were <0.15 to 0.25 mg/L. Table 56. RFB MICs for pWT isolates on 7H11. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). 3.B.3.2 RFB MICs for mutated isolates on 7H11 rpoB 450 (531) mutants Only 1 study was identified that reported RFB MICs for isolates with rpoB S450 (S531) codon mutants on 7H11 (Table 57). The five isolates tested in this study all had RFB MICs >32 mg/L, with a mode of 128 mg/L. Table 57. RFB MICs for rpoB S450 (S531) mutants on 7H11. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). rpoB borderline RRDR mutants No studies were identified that featured RFB MIC data for isolates with rpoB borderline RRDR mutations on 7H11 medium. Other rpoB RRDR mutants Only 1 study was identified that reported RFB MICs for isolates with rpoB RRDR mutations other than the borderline mutations and S450 (S531) codon mutations (Table 58). The 16 MICs reported for these isolates were all >32 mg/L, with a mode of 128 mg/L, except one S522L mutant with an MIC <1 mg/L. Studies Isolate origin Unique isolates Type of isolates Genotypic results - Mtb0.015 0.03 0.06 0.12 0.25 0.5 1 2 4 1 H37Rv 1 1 M. africanum ATCC 25420 1 1 M. bovis ATCC 19210 1 5 M. bovis BCG 2 3 6 Pan-S 2 4 4 M. africanum 1 3 4 M. bovis 4 13) Truffot-Pernot 1988 clinical 10 RIF-S 7 3 11) Anthony 2005 & Bergval 2007 lab 2 gWT 2 10) Rastogi 2000 clinical RFB MIC (mg/L) Studies Isolate origin Unique isolates Genotypic results 0.5 1 2 4 8 16 32 64 128 160 lab 4 rpoB S450L (S531L) 1 2 1 lab 1 rpoB S450W (S531W) 1 RFB MIC [mg/L] 11) Anthony 2005 & Bergval 2007 69 Table 58. RFB MICs for other rpoB RRDR mutants on 7H11. The blue line denotes the current CLSI CC for RFB DST on 7H10 (0.5 mg/L). rpoB mutants outside the RRDR No studies were identified that featured RFB MIC data for isolates with rpoB mutations outside of the RRDR on 7H11 medium. 3.B.3.3 Conclusion for RFB CC for 7H11 Given a lack of genetic data and MIC data truncations, the three studies identified by this review contained insufficient data for pWT isolates to define even a tentative ECOFF for RFB CC for 7H11. Even if sufficient MIC evidence had been available, a CC would not have been endorsed as RFB is not currently recommended for TB treatment by WHO. Refer to Section 3.3 for a strategy for RFB DST. Studies Isolate origin Unique isolates Genotypic results 0.5 1 2 4 8 16 32 64 128 160 lab 1 rpoB Q432L (Q513L) 1 lab 1 rpoB S441L (S522L) 1 lab 4 rpoB H445D (H526D) 3 1 lab 5 rpoB H445R (H526R) 5 lab 1 rpoB H445P (H526P) 1 lab 4 rpoB H445Y (H526Y) 1 2 1 RFB MIC [mg/L] 11) Anthony 2005 & Bergval 2007 70 3.B.4 RFB MIC data in MGIT 3.B.4.1 RFB MICs for pWT isolates in MGIT Nine studies were identified that reported RFB MIC data for the pWT population by MGIT (Table 59). Five of these studies reported MICs for at least 10 pWT isolates, though the majority of MIC distributions were truncated, and so the shape of the distributions could not be determined in most cases. Nevertheless, the results suggested a tentative ECOFF of 0.125 mg/L. Table 59. RFB MICs for pWT isolates in MGIT. The blue line denotes the current CLSI CC for RFB DST in MGIT (0.5 mg/L). Notable limitations: studies RB21 and RB23; and studies RB15 and RB19 were conducted in the same laboratory. 3.B.4.2 RFB MICs for mutated isolates in MGIT rpoB 450 (531) mutants Eight studies were identified that reported RFB MICs for 218 isolates with rpoB S450 (S531) codon mutants by MGIT (Table 60). All MICs reported for these isolates were >1 mg/L except for 2 (1%, 95% CI 0-3%) rpoB S450L (S531L) mutants with MICs <0.25 mg/L from two different studies. If a CC were set at 0.125 mg/L, it is likely that no isolates would be classified as RFB-S. Table 60. RFB MICs for rpoB S450 (S531) mutants in MGIT. The blue line denotes the current CLSI CC for RFB DST in MGIT (0.5 mg/L). Notable limitations: studies RB14 and RB22; studies RB14 and RB20; and studies RB23 and RB21 were conducted in the same laboratory. rpoB borderline RRDR mutants Seven studies were identified that reported RFB MICs for isolates with rpoB borderline mutations tested by MGIT (Table 61). The MICs for these 42 isolates ranged from <0.06 to >2 mg/L. At the CLSI CC (i.e. 0.5 mg/L), 38 (90%, 95% CI 77-97%) isolates would be classified as RFB-S. If a CC were set at Studies Lab Isolate origin Unique isolates Total MICs Type of isolates Genotypic results - Mtb0.015 0.03 0.06 0.1 0.12 0.25 0.4 0.5 1 2 4 15) Heyckendorf 2018 17 1 1 H37Rv ATCC27294 1 16) Jamieson 2014 18 clinical 1 1 mixture of first-line resistance profiles gWT 1 17) Sirgel 2013 & Sirgel 2015 19 clinical 26 26 gWT 3 23 18) Sharma 2011 20 36 36 mixture of first-line resistance profiles 19 2 15 17, 21-22 10 30 Pan-S 30 17, 21-22 21 63 mixture of first-line resistance profiles 8 4 51 20) Whitfield 2018 11 3 3 gWT 2 1 23 1 10 H37Rv 4 5 1 23 2 2 gWT 2 23) Berrada 2016 23 clinical 44 44 gWT 39 5 7 1 1 H37Rv ATCC27294 1 7 1 1 Erdman 1 7 clinical 22 22 gWT 22 19) Rüsch-Gerdes 2006 RFB MIC (mg/L) 24) Gygli 2019 21) Lin - unpublished Studies Lab Isolate origin Unique isolates Genotypic results 0.1 0.12 0.25 0.4 0.5 1 2 4 5 8 10 15 1, 4, 10-16 clinical 1 rpoB S450P (S531P) 1 1, 4, 10-16 clinical 1 rpoB S450M (S531M) 1 1, 4, 10-16 clinical 89 rpoB S450L (S531L) 19 70 22) Machado 12 clinical 27 rpoB S450L (S531L) 9 18 24) Gygli 2019 7 clinical 16 rpoB S450L (S531L) 10 6 17 clinical 1 rpoB S450F (S531F) 1 17 clinical 12 rpoB S450L (S531L) 12 23 clinical 1 rpoB S450C (S531C) 23 clinical 1 rpoB S450F (S531F) 1 23 clinical 4 rpoB S450W (S531W) 1 2 1 23 clinical 34 rpoB S450L (S531L) 7 25 2 20) Whitfield 2018 11 8 rpoB S450L (S531L) 1 4 3 18 clinical 19 rpoB S450L (S531L) 1 18 18 clinical 3 rpoB S450W (S531W) 3 21) Lin - unpublished 23 1 rpoB S450Q (S531Q) 1 RFB MIC [mg/L] 14) Cambau 2015 15) Heyckendorf 2018 16) Jamieson 2014 23) Berrada 2016 71 0.125 mg/L, at least 26 (62%, 95% CI 46-76%) isolates would be classified as RFB-S. An ATU at 0.125 mg/L could reduce this to 33% (95% CI 20-50%). Table 61. RFB MICs for rpoB borderline RRDR mutants in MGIT. The blue line denotes the current CLSI CC for RFB DST in MGIT (0.5 mg/L). Notable limitations: studies RB14 and RB20 were conducted in the same laboratory. Other rpoB RRDR mutants Nine studies were identified that reported RFB MICs for isolates with rpoB RRDR mutations other than the borderline mutations and S450 (S531) codon mutations (Table 62). The MICs reported for these 150 isolates ranged from <0.03 to >8 mg/L, with the modes of some distributions between 0.125 and 0.5 mg/L. At the CLSI CC (i.e. 0.5 mg/L), 93 (62%, 95% CI 54-70%) of these isolates would be classified as RFB-S. If a CC were set at 0.125 mg/L, at least 41 (27%, 95% CI 20-35%) isolates would be classified as RFB-S. An ATU at 0.125 mg/L might lower this substantially to 3% (95% CI 1-8%). Studies Lab Isolate origin Unique isolates Genotypic results 0.06 0.1 0.12 0.25 0.4 0.5 1 2 5 23) Berrada 2016 23 clinical 3 rpoB L430P (L511P) 3 14) Cambau 2015 1, 4, 10-16 clinical 1 rpoB L430P (L511P) 1 20) Whitfield 2018 11 2 rpoB L430P (L511P) 2 16) Jamieson 2014 18 clinical 2 rpoB L430P (L511P) 2 15) Heyckendorf 2018 17 clinical 1 rpoB D435Y (D516Y) 1 17) Sirgel 2013 & Sirgel 2015 19 clinical 1 rpoB D435Y (D516Y) 1 23) Berrada 2016 23 clinical 4 rpoB D435Y (D516Y) 4 14) Cambau 2015 1, 4, 10-16 clinical 1 rpoB D435Y (D516Y) 1 20) Whitfield 2018 11 2 rpoB D435Y (D516Y) 2 23) Berrada 2016 23 clinical 4 rpoB H445L (H526L) 2 1 1 14) Cambau 2015 1, 4, 10-16 clinical 3 rpoB H445L (H526L) 1 1 1 24) Gygli 2019 7 clinical 1 rpoB H445L (H526L) 1 20) Whitfield 2018 11 1 rpoB H445L (H526L) 1 16) Jamieson 2014 18 clinical 2 rpoB H445L (H526L) 2 23) Berrada 2016 23 clinical 3 rpoB H445N (H526N) 2 1 20) Whitfield 2018 11 2 rpoB H445N (H526N) 2 16) Jamieson 2014 18 clinical 1 rpoB H445N (H526N) 1 23) Berrada 2016 23 clinical 3 rpoB H445S (H526S) 2 1 15) Heyckendorf 2018 17 clinical 1 rpoB L452P (L533P) 1 23) Berrada 2016 23 clinical 2 rpoB L452P (L533P) 1 1 24) Gygli 2019 7 clinical 1 rpoB L452P (L533P) 1 20) Whitfield 2018 11 1 rpoB L452P (L533P) 1 RFB MIC [mg/L] 72 Table 62. RFB MICs for other rpoB RRDR mutants in MGIT. The blue line denotes the current CLSI CC for RFB DST in MGIT (0.5 mg/L). Notable limitations: studies RB14 and RB20; studies RB14 and RB22; and studies RB23 and RB21 were conducted in the same laboratory. rpoB mutants outside the RRDR Three studies were identified that reported RFB MICs for isolates with mutations outside the rpoB RRDR by MGIT (Table 63). The MICs for V170F (V146F) mutants were clearly elevated from the wildtype distribution, with MICs >1 mg/L. An ATU at 0.125 mg/L would eliminate the misclassification of the sole I491F (I572F) mutant. Table 63. RFB MICs for rpoB mutants outside of RRDR in MGIT. The blue line denotes the current CLSI CC for RFB DST in MGIT (0.5 mg/L). Studies Lab Isolate origin Unique isolates Genotypic results 0.03 0.06 0.1 0.12 0.25 0.4 0.5 1 2 4 5 8 10 23) Berrada 2016 23 clinical 2 rpoB Q432E (Q513E) 2 23) Berrada 2016 23 clinical 3 rpoB Q432K (Q513K) 3 20) Whitfield 2018 11 2 rpoB Q432K (Q513K) 2 23) Berrada 2016 23 clinical 1 rpoB Q432L (Q513L) 1 20) Whitfield 2018 11 1 rpoB Q432L (Q513L) 1 23) Berrada 2016 23 clinical 2 rpoB Q432P (Q513P) 1 1 20) Whitfield 2018 11 1 rpoB Q432P (Q513P) 1 20) Whitfield 2018 11 1 rpoB D435A (D516A) 1 23) Berrada 2016 23 clinical 1 rpoB D435F (D516F) 1 17) Sirgel 2013 & 19 clinical 4 rpoB D435S (D516S) 4 17) Sirgel 2013 & 19 clinical 29 rpoB D435V (D516V) 20 9 15) Heyckendorf 2018 17 clinical 1 rpoB D435V (D516V) 1 23) Berrada 2016 clinical 18 rpoB D435V (D516V) 2 6 10 24) Gygli 2019 7 clinical 2 rpoB D435V (D516V) 1 1 14) Cambau 2015 1, 4, 10-16 clinical 10 rpoB D435V (D516V) 1 6 2 1 22) Machado 12 clinical 2 rpoB D435V (D516V) 2 20) Whitfield 2018 11 13 rpoB D435V (D516V) 3 9 1 16) Jamieson 2014 18 clinical 3 rpoB D435V (D516V) 1 2 21) Lin - unpublished 23 1 rpoB N438K (N519K) 1 23) Berrada 2016 23 clinical 1 rpoB S441L (S522L) 1 14) Cambau 2015 1, 4, 10-16 clinical 1 rpoB S441L (S522L) 1 20) Whitfield 2018 11 1 rpoB S441L (S522L) 1 16) Jamieson 2014 18 clinical 3 rpoB S441L (S522L) 3 21) Lin - unpublished 23 1 rpoB S441T (S522T) 1 21) Lin - unpublished 23 1 rpoB H445A (H526A) 1 23) Berrada 2016 23 clinical 1 rpoB H445A (H526A) 1 14) Cambau 2015 1, 4, 10-16 clinical 1 rpoB H445A (H526A) 1 23) Berrada 2016 23 clinical 2 rpoB H445C (H526C) 2 23) Berrada 2016 23 clinical 3 rpoB H445D (H526D) 2 1 14) Cambau 2015 1, 4, 10-16 clinical 5 rpoB H445D (H526D) 5 24) Gygli 2019 7 clinical 4 rpoB H445D (H526D) 4 20) Whitfield 2018 11 4 rpoB H445D (H526D) 4 16) Jamieson 2014 18 clinical 1 rpoB H445D (H526D) 1 23) Berrada 2016 23 clinical 1 rpoB H445G (H526G) 1 21) Lin - unpublished 23 1 rpoB H445P (H526P) 1 21) Lin - unpublished 23 1 rpoB H445Q (H526Q) 1 23) Berrada 2016 23 clinical 5 rpoB H445R (H526R) 2 3 20) Whitfield 2018 11 2 rpoB H445R (H526R) 2 23) Berrada 2016 23 clinical 4 rpoB H445Y (H526Y) 2 2 14) Cambau 2015 1, 4, 10-16 clinical 3 rpoB H445Y (H526Y) 3 22) Machado 12 clinical 1 rpoB H445Y (H526Y) 1 24) Gygli 2019 7 clinical 2 rpoB H445Y (H526Y) 2 20) Whitfield 2018 11 2 rpoB H445Y (H526Y) 2 16) Jamieson 2014 18 clinical 2 rpoB H445Y (H526Y) 2 RFB MIC [mg/L] Studies Lab Isolate origin Unique isolates Genotypic results 0.1 0.12 0.25 0.4 0.5 1 2 4 16) Jamieson 2014 17 clinical 1 rpoB V170F (V146F) 1 14) Cambau 2015 1, 4, 9-15 clinical 1 rpoB V170F (V146F) 1 21) Lin - unpublished 22 1 rpoB V170F (V146F) 1 21) Lin - unpublished 22 1 rpoB I491F (I572F) 1 RFB MIC [mg/L] 73 3.B.4.3 Conclusion for RFB CC for MGIT Although 0.125 mg/L appeared to be the tentative ECOFF for RFB for this medium (Table 59), this observation was based on limited data and no CC was adopted given that RFB is not currently recommended for the treatment of TB by WHO. Refer to Section 3.3 for a strategy for RFB DST. Table 64. Effect of adopting a CC for RFB DST on MGIT for detection of rpoB mutants. Mutation type Percentage of mutants classified RFB-S at CC 0.5 mg/L CC 0.125 mg/L CC+ATU 0.125 mg/L S450 (S531) 1% (95% CI 0-4%) 0% (95% CI 0-2%) 0% (95% CI 0-2%) borderline RRDR 90% (95% CI 77-97%) 62% (95% CI 46-76%) 33% (95% CI 20-50%) other RRDR 62% (95% CI 54-70%) 27% (95% CI 20-35%) 3% (95% CI, 1-8%) borderline I491F (I572F) not calculateda The current CLSI CC is underlined. a Too few MICs. 74 3.B.5 References for RFB MIC studies 1. Rancoita PM V., Cugnata F, Gibertoni Cruz AL, et al. Validating a 14-drug microtiter plate containing bedaquiline and delamanid for large-scale research susceptibility testing of Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2018;62(9):e00344-18. doi:10.1128/AAC.00344-18. 2. Cavusoglu C, Karaca-Derici Y, Bilgic A. In-vitro activity of rifabutin against rifampicin-resistant Mycobacterium tuberculosis isolates with known rpoB mutations. Clin Microbiol Infect. 2004;10(7):662- 665. doi:10.1111/j.1469-0691.2004.00917.x. 3. Schön T, Juréen P, Chryssanthou E, et al. Rifampicin-resistant and rifabutin-susceptible Mycobacterium tuberculosis strains: a breakpoint artefact? J Antimicrob Chemother. 2013;68(9):2074-2077. doi:10.1093/jac/dkt150. 4. van Ingen J, Simons S, de Zwaan R, et al. Comparative study on genotypic and phenotypic second-line drug resistance testing of Mycobacterium tuberculosis complex Isolates. J Clin Microbiol. 2010;48(8):2749-2753. doi:10.1128/JCM.00652-10. 5. Schön T, Juréen P, Chryssanthou E, et al. Wild-type distributions of seven oral second-line drugs against Mycobacterium tuberculosis. Int J Tuberc Lung Dis. 2011;15(4):502-9. doi:10.5588/ijtld.10.0238. 6. Garcia A-B, Palacios JJ, Ruiz M-J, et al. Strong in vitro activities of two new rifabutin analogs against multidrug-resistant Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2010;54(12):5363- 5365. doi:10.1128/AAC.00149-10. 7. van Ingen J, Simons S, de Zwaan R, et al. Comparative study on genotypic and phenotypic second-line drug resistance testing of Mycobacterium tuberculosis complex isolates. J Clin Microbiol. 2010;48(8):2749-2753. doi:10.1128/JCM.00652-10. 8. Farhat MR, Sixsmith J, Calderon R, Hicks ND, Fortune SM, Murray M. Rifampicin and rifabutin resistance in 1003 Mycobacterium tuberculosis clinical isolates. J Antimicrob Chemother. 2019;74(6):1477-1483. doi:10.1093/jac/dkz048. 9. Gygli SM, Keller PM, Ballif M, et al. Whole-genome sequencing for drug resistance profile prediction in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2019;63(4):e02175-18. doi:10.1128/AAC.02175-18. 10. Rastogi N, Goh KS, Berchel M, Bryskier A. Activity of rifapentine and its metabolite 25-O- desacetylrifapentine compared with rifampicin and rifabutin against Mycobacterium tuberculosis, Mycobacterium africanum, Mycobacterium bovis and M. bovis BCG. J Antimicrob Chemother. 2000;46(4):565-570. doi:10.1093/jac/46.4.565. 11. (a) Anthony RM, Schuitema AR, Bergval IL, et al. Acquisition of rifabutin resistance by a rifampicin resistant mutant of Mycobacterium tuberculosis involves an unusual spectrum of mutations and elevated frequency. Ann Clin Microbiol Antimicrob. 2005;4(1):9. doi:10.1186/1476-0711-4-9. (b) Bergval IL, Klatser PR, Schuitema ARJ, Oskam L, Anthony RM. Specific mutations in the Mycobacterium tuberculosis rpoB gene are associated with increased dnaE2 expression. FEMS Microbiol Lett. 2007;275(2):338-343. doi:10.1111/j.1574-6968.2007.00905.x. 12. Pretet S, Lebeaut A, Parrot R, et al. Combined chemotherapy including rifabutin for rifampicin and isoniazid resistant pulmonary tuberculosis. G.E.T.I.M. (Group for the Study and Treatment of Resistant Mycobacterial Infection). Eur Respir J. 1992;5(6):680-4. 13. Truffot-Pernot C, Giroir AM, Grosset J. A study of the minimal inhibitory concentration of rifabutin (ansamycin LM 427) for Mycobacterium tuberculosis, Mycobacterium xenopi and Mycobacterium avium-intracellulaire. Rev Mal Respir. 1988;5(4):401-406. 14. Cambau E, Viveiros M, Machado D, et al. Revisiting susceptibility testing in MDR-TB by a standardized quantitative phenotypic assessment in a European multicentre study. J Antimicrob Chemother. 2015;70(3):686-696. doi:10.1093/jac/dku438. 15. Heyckendorf J, Andres S, Köser CU, et al. What is resistance? Impact of phenotypic versus molecular drug resistance testing on therapy for multi- and extensively drug-resistant tuberculosis. Antimicrob Agents Chemother. 2017;62(2):e01550-17. doi:10.1128/AAC.01550-17. 75 16. Jamieson FB, Guthrie JL, Neemuchwala A, et al. Profiling of rpoB mutations and MICs for rifampin and rifabutin in Mycobacterium tuberculosis. J Clin Microbiol. 2014;52(6):2157-2162. doi:10.1128/JCM.00691-14. 17. Sharma M, Thibert L, Chedore P, et al. Canadian multicenter laboratory study for standardized second- line antimicrobial susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol. 2011;49(12):4112-4116. doi:10.1128/JCM.05195-11. 18. Rüsch-Gerdes S, Pfyffer GE, Casal M, Chadwick M, Siddiqi S. Multicenter laboratory validation of the BACTEC MGIT 960 technique for testing susceptibilities of Mycobacterium tuberculosis to classical second-line drugs and newer antimicrobials. J Clin Microbiol. 2006;44(3):688-692. doi:10.1128/JCM.44.3.688-692.2006. 19. Whitfield MG, Warren RM, Mathys V, et al. The potential use of rifabutin for treatment of patients diagnosed with rifampicin-resistant tuberculosis. J Antimicrob Chemother. 2018;73(10):2667-2674. doi:10.1093/jac/dky248 20. Lin G. Personal communication. 2018. 21. Machado D, Perdigão J, Portugal I, et al. Efflux activity differentially modulates the levels of isoniazid and rifampicin resistance among multidrug resistant and monoresistant Mycobacterium tuberculosis strains. Antibiotics. 2018;7(1):18. doi:10.3390/antibiotics7010018. 22. Berrada ZL, Lin S-YG, Rodwell TC, et al. Rifabutin and rifampin resistance levels and associated rpoB mutations in clinical isolates of Mycobacterium tuberculosis complex. Diagn Microbiol Infect Dis. 2016;85(2):177-181. doi:10.1016/j.diagmicrobio.2016.01.019. 23. Gygli SM, Keller PM, Ballif M, et al. Whole-genome sequencing for drug resistance profile prediction in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2019;63(4):e02175-18. doi:10.1128/AAC.02175-18. 76 3.C.1 RPT MIC data on LJ 3.C.1.1 RPT MICs for pWT isolates on LJ No studies were identified that featured RPT MIC data for pWT isolates on LJ medium. 3.C.1.2 RPT MICs for mutated isolates on LJ No studies were identified that featured RPT MIC data for mutated isolates on LJ medium. 3.C.1.3 Conclusion for RPT CC for LJ Given the lack of data, a CC for RPT could not be set for this medium. RIF should, instead, be used as a surrogate (Section 3.3). 77 3.C.2 RPT MIC data on 7H10 3.C.2.1 RPT MICs for pWT isolates on 7H10 Just two studies were identified with RPT MIC data for 7H10 (Table 65). Only one study reported MICs for at least 10 pWT isolates and the MIC distribution was severely truncated at the lower end, precluding an assessment of the shape of the distribution. Table 65. RPT MICs for pWT isolates on 7H10. 3.C.2.2 RPT MICs for mutated isolates on 7H10 rpoB 450 (531) mutants Five clinical isolates from 1 study harbored rpoB S450 (S531) codon mutations (Table 66). All 5 isolates had MICs >32 mg/L. Table 66. RPT MICs for rpoB S450 (S531) mutants on 7H10. rpoB borderline RRDR mutants Only 4 clinical isolates from 1 study harbored rpoB RRDR borderline mutations (Table 67). The isolates were only tested on 7H10, and RPT MICs ranged from 8->32 mg/L. Table 67. RPT MICs for rpoB borderline RRDR mutants on 7H10. Other rpoB RRDR mutants Only 8 clinical isolates from 1 study harbored rpoB RRDR mutations other than S450 (S531) and the borderline mutations (Table 68). The isolates had RPT MICs that ranged from 16->32 mg/L. Studies Isolate origin Unique isolates Total MICs Type of isolates Genotypic results 0.03 0.06 0.12 0.25 0.5 1 2 4 8 16 lab 1 10 H37Rv 7 3 lab 1 1 H37Rv 1 lab 1 1 Erdman 1 lab 1 1 Atencio 1 lab 5 5 Pan-S 3 1 1 clinical 85 128 mixture of first-line resistance profiles 93 2 1 4 28 2) Moghazeh 1996 clinical 1 1 Pan-S gWT 1 1) Heifets 1999 RPT MIC [mg/L] Studies Isolate origin Unique isolates Genotypic results 0.25 0.5 1 2 4 8 16 32 64 clinical 4 rpoB S450L (S531L) 4 clinical 1 rpoB S450W (S531W) 1 2) Moghazeh 1996 RPT MIC [mg/L] Studies Isolate origin Unique isolates Genotypic results 0.25 0.5 1 2 4 8 16 32 64 clinical 1 rpoB H445N (H526N) 1 clinical 2 rpoB H445L (H526L) 2 clinical 1 rpoB L452P (L533P) 1 2) Moghazeh 1996 RPT MIC [mg/L] 78 Table 68. RPT MICs for other rpoB RRDR mutants on 7H10. rpoB mutants outside the RRDR No studies were identified that featured RPT MIC data for isolates with rpoB mutations outside the RRDR. 3.C.2.3 Conclusion for RPT CC for 7H10 The available data (Table 65) were insufficient to set even a tentative CC for this medium. RIF should, instead, be used as a surrogate for RPT resistance (Section 3.3). Studies Isolate origin Unique isolates Genotypic results 0.25 0.5 1 2 4 8 16 32 64 clinical 1 rpoB Q432K (Q513K) 1 clinical 1 rpoB Q432L (Q513L) 1 clinical 2 rpoB H445D (H526D) 2 clinical 4 rpoB H445Y (H526Y) 4 2) Moghazeh 1996 RPT MIC [mg/L] 79 3.C.3 RPT MIC data on 7H11 3.C.3.1 RPT MICs for pWT isolates on 7H11 Just one study was identified that reported RPT MIC data for the pWT population by 7H11 (Table 69). Rastogi et al. (study RP3) reported an untruncated pWT MIC distribution of 0.06-0.5 mg/L for 14 M. tuberculosis complex isolates, with a mode of 0.125 mg/L. The authors also tested the H37Rv control strain, with an MIC of 0.25 mg/L, M. africanum ATCC 25420 with an MIC of 0.25 mg/L, and six M. bovis isolates, with MICs ≤0.015-0.125 mg/L. Table 69. RPT MICs for pWT isolates on 7H11. 3.C.3.2 RPT MICs for mutated isolates on 7H11 No studies were identified that featured RPT MIC data for mutated isolates on 7H11 medium. 3.C.3.3 Conclusion for RPT CC for 7H11 Given that only a single study was identified by this review (Table 69), a RPT CC could not be set for this medium. RIF should, instead, be used as a surrogate for RPT resistance (Section 3.3). Studies Isolate origin Unique isolates Type of isolates 0.02 0.03 0.06 0.12 0.25 0.5 1 1 H37Rv 1 1 M. africanum ATCC 25420 1 1 M. bovis ATCC 19210 1 5 M. bovis BCG 2 1 2 6 Pan-S 1 5 4 M. africanum 1 2 1 4 M. bovis 2 2 3) Rastogi 2000 RPT MIC [mg/L] clinical 80 3.C.4 RPT MIC data in MGIT 3.C.4.1 RPT MICs for pWT isolates in MGIT No studies were identified that featured RPT MIC data for pWT isolates tested by MGIT. 3.C.4.2 RPT MICs for mutated isolates in MGIT No studies were identified that featured RPT MIC data for mutated isolates tested by MGIT. 3.C.4.3 Conclusion for RPT CC for MGIT Given the lack of data, a CC for RPT could not be set for this medium. RIF should, instead, be used as a surrogate for RPT DST (Section 3.3). 81 3.C.5 References for RPT MIC studies 1. Heifets L, Sanchez T, Vanderkolk J, Pham V. Development of rifapentine susceptibility tests for Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1999;43(1):25-28. doi:10.1128/AAC.43.1.25. 2. Moghazeh SL, Pan X, Arain T, et al. Comparative antimycobacterial activities of rifampin, rifapentine, and KRM-1648 against a collection of rifampin-resistant Mycobacterium tuberculosis isolates with known rpoB mutations. Antimicrob Agents Chemother. 1996;40(11):2655-2657. doi:10.1128/AAC.40.11.2655. 3. Rastogi N, Goh KS, Berchel M, Bryskier A. Activity of rifapentine and its metabolite 25-O- desacetylrifapentine compared with rifampicin and rifabutin against Mycobacterium tuberculosis, Mycobacterium africanum, Mycobacterium bovis and M. bovis BCG. J Antimicrob Chemother. 2000;46(4):565-570. doi:10.1093/jac/46.4.565. 82 3.3 Rifamycin conclusions and comments Rifampicin The six RRDR and the I491F (I572F) borderline resistance mutations (Table 26) have been the focus of considerable debate in the literature, particularly in light of the role of L452P (L533P) in KwaZulu-Natal, South Africa, and in the clonal spread of I491F (I572F) in Eswatini.88, 89 Specifically, there is uncertainty as to why these mutations are more likely to test susceptible by pDST and whether they are clinically relevant at the currently recommended RIF dose of 10 mg/kg (8-12 mg/kg) daily (the 600 mg daily cap was removed in 2017 to improve dosing in higher body-weight bands).90,91 This systematic review revealed that one reason for the discordant results for borderline resistance mutations was that the CCs for 7H10 and MGIT had been set too high for the pWT population, which was corrected by lowering these CCs to the ECOFFs of 0.5 mg/L. However, this only reduces rather than eliminates the discordance between genotype and phenotype, as borderline resistance mutations are typically associated with only modest MIC increases. Specifically, their MIC distributions usually overlap with the MIC distributions of pWT isolates, which means that the ECOFF for RIF intersects these MIC distributions, resulting in a poor reproducibility with categorical pDST (i.e. even if the same isolate is tested in the same laboratory, susceptibility results may vary due to the inevitable technical variation in pDST). This issue applies to all media, including the WHO-endorsed microscopic observation of drug susceptibility and the nitrate reductase assay, although the problem is particularly pronounced with MGIT, which may be due to the lower fitness of isolates with these mutations and the shorter incubation time of MGIT.92 Clinical outcome data for the seven borderline resistance mutations, particularly I491F (I572F), are limited and it is unclear whether these associations may be confounded. For example, some isolates with borderline resistance mutations have very high RIF MICs, which could be due to any of the following reasons. A sample might be mixed and also harbour a subpopulation with a HLR mutation, such as rpoB S450L (S531L), below the limit of detection of the gDST method employed.93 Alternatively, secondary mutations in rpoA, rpoB, rpoC, or elsewhere in the genome may raise the MIC of a given 88 Brown TS, Challagundla L, Baugh EH, et al. Pre-detection history of extensively drug-resistant tuberculosis in KwaZulu- Natal, South Africa. Proc Natl Acad Sci U S A. 2019;116(46):23284-23291. doi:10.1073/pnas.1906636116. 89 Sanchez-Padilla E, Merker M, Beckert P, et al. Detection of drug-resistant tuberculosis by Xpert MTB/RIF in Swaziland. N Engl J Med. 2015;372(12):1181-1182. doi:10.1056/NEJMc1413930. 90 World Health Organization. Guidelines for treatment of drug-susceptible tuberculosis and patient care. 2017 update. Annex 6. Essential first-line antituberculosis drugs. (https://www.who.int/tb/publications/2017/tb_guidelines2017_annex6_en_v4.pdf, accessed 18 September 2020) 91 World Health Organization. Technical report on the pharmacokinetics and pharmacodynamics (PK/PD) of medicines used in the treatment of drug-resistant tuberculosis. (http://apps.who.int/iris/bitstream/handle/10665/260440/WHO-CDS-TB- 2018.6-eng.pdf, accessed 15 September 2018). 92 Torrea G, Ng K, Van Deun A. Variable ability of rapid tests to detect Mycobacterium tuberculosis rpoB mutations conferring phenotypically occult rifampicin resistance. Sci Rep. 2019;9(1):11826. doi:10.1038/s41598-019-48401-z. 93 Ng KCS, Supply P, Cobelens FGJ, et al. How well do routine molecular diagnostics detect rifampin heteroresistance in Mycobacterium tuberculosis? J Clin Microbiol. 2019;57(11):e00717-19. doi:10.1128/JCM.00717-19. 83 isolate well above the CC.94,95,96 Nevertheless, the available evidence points towards worse treatment outcomes for at least some of these mutations.97,98,99, 100,101, 102,103,104 In fact, the poor treatment outcomes attributed to INH mono-resistance are likely partly due to missed RIF resistance due to the presence of these borderline resistance mutations.105 In addition, the 10 mg/kg/day dose represents the minimal effective dose for RIF, which was chosen historically because of the limited supply and high cost of the drug and fear for dose-related adverse events.106,107 This dose is likely insufficient for a proportion of patients infected with pWT isolates and, thus, risks the selection of RIF resistance, particularly in light of substantial PK variability.108,109 Further, RIF displays profound concentration-dependent killing, which is not activated by 10 mg/kg/day.110 As a result, several trials are evaluating higher doses.111 Because RIF is likely not optimally dosed even for pWT isolates, any MIC increase caused by rpoB mutations must, by definition, increase the proportion of patients who are not optimally treated. 94 Zaczek A, Brzostek A, Augustynowicz-Kopec E, Zwolska Z, Dziadek J. Genetic evaluation of relationship between mutations in rpoB and resistance of Mycobacterium tuberculosis to rifampin. BMC Microbiol. 2009;9:10. doi:10.1186/1471-2180-9-10. 95 Meftahi N, Namouchi A, Mhenni B, et al. Evidence for the critical role of a secondary site rpoB mutation in the compensatory evolution and successful transmission of an MDR tuberculosis outbreak strain. J Antimicrob Chemother. 2016;71(2):324-332. doi:10.1093/jac/dkv345. 96 Shea J, Halse TA, Kohlerschmidt D, et al. Low-level rifampin resistance and rpoB mutations in Mycobacterium tuberculosis: An analysis of whole-genome sequencing and drug susceptibility test data in New York. J Clin Microbiol. 2020 Online ahead of print. doi:10.1128/JCM.01885-20. 97 Williamson DA, Roberts SA, Bower JE, et al. Clinical failures associated with rpoB mutations in phenotypically occult multidrug-resistant Mycobacterium tuberculosis. Int J Tuberc Lung Dis. 2012;16(2):216-220. doi:10.5588/ijtld.11.0178. 98 Ho J, Jelfs P, Sintchencko V. Phenotypically occult multidrug-resistant Mycobacterium tuberculosis: dilemmas in diagnosis and treatment. J Antimicrob Chemother. 2013;68(12):2915-2920. doi:10.1093/jac/dkt284. 99 Van Deun A, Aung KJ, Bola V, et al. Rifampin drug resistance tests for tuberculosis: challenging the gold standard. J Clin Microbiol. 2013;51(8):2633-2640. doi:10.1128/JCM.00553-13. 100 Pang Y, Ruan YZ, Zhao J, et al. Diagnostic dilemma: treatment outcomes of tuberculosis patients with inconsistent rifampicin susceptibility. Int J Tuberc Lung Dis. 2014;18(3):357-362. doi:10.5588/ijtld.13.0459. 101 Van Deun A, Aung KJ, Hossain A, et al. Disputed rpoB mutations can frequently cause important rifampicin resistance among new tuberculosis patients. Int J Tuberc Lung Dis. 2015;19(2):185-190. doi:10.5588/ijtld.14.0651. 102 Shah N, Lin SYG, Barry PM, Cheng YN, et al. Clinical impact on tuberculosis treatment outcomes of discordance between molecular and growth-based assays for rifampin resistance, California 2003-2013. Open Forum Infect Dis. 2016;3(3):ofw150. doi:10.1093/ofid/ofw150. 103 Hu P, Zhang H, Fleming J, et al. Retrospective analysis of false-positive and disputed rifampin resistance Xpert MTB/RIF assay results in clinical samples from a referral hospital in Hunan, China. J Clin Microbiol. 2019;57(4):e01707-18. doi:10.1128/JCM.01707-18. 104 van Ingen J, Aarnoutse R, de Vries G, Boeree MJ, van Soolingen D. Low-level rifampicin-resistant Mycobacterium tuberculosis strains raise a new therapeutic challenge. Int J Tuberc Lung Dis. 2011;15(7):990-992. doi:10.5588/ijtld.10.0127. 105 Van Deun A, Decroo T, Kya Jai Maug A, et al. The perceived impact of isoniazid resistance on outcome of first-line rifampicin-throughout regimens is largely due to missed rifampicin resistance. PLoS One. 2020;15(5):e0233500. doi:10.1371/journal.pone.0233500. 106 van Ingen J, Aarnoutse RE, Donald PR, et al. Why Do we use 600 mg of rifampicin in tuberculosis treatment? Clin Infect Dis. 2011;52(9):e194-e199. doi:10.1093/cid/cir184. 107 Maug AKJ, Hossain MA, Gumusboga M, et al. First-line tuberculosis treatment with double-dose rifampicin is well tolerated. Int J Tuberc Lung Dis. 2020;24(5):499-505. doi:10.5588/ijtld.19.0063. 108 Gumbo T, Louie A, Deziel MR, et al. Concentration-dependent Mycobacterium tuberculosis killing and prevention of resistance by rifampin. Antimicrob Agents Chemother. 2007;51(11):3781-3788. doi:10.1128/AAC.01533-06. 109 Stott KE, Pertinez H, Sturkenboom MGG, et al. Pharmacokinetics of rifampicin in adult TB patients and healthy volunteers: a systematic review and meta-analysis. J Antimicrob Chemother. 2018;73(9):2305-2313. doi:10.1093/jac/dky152. 110 Peloquin CA, Velásquez GE, Lecca L, et al. Pharmacokinetic evidence from the HIRIF trial to support increased doses of rifampin for tuberculosis. Antimicrob Agents Chemother. 2017;61(8):e00038-17. doi:10.1128/AAC.00038-17. 111 Velásquez GE, Brooks MB, Coit JM, et al. Efficacy and safety of high-dose rifampin in pulmonary tuberculosis. A randomized controlled trial. Am J Respir Crit Care Med. 2018;198(5):657-666. doi:10.1164/rccm.201712-2524OC. 84 Given this direct and indirect evidence and to err on the side of caution, the TEG recommended that all seven mutations should be regarded as clinically relevant for the current dose of RIF (i.e. they should be treated with an MDR-TB regimen according to the least treatment guidelines and are, therefore, best referred to as “borderline resistance” mutations rather than “disputed”, “discordant” or “occult” to underline their clinical importance and that they cannot be confirmed reliably using pDST).112 This decision will be reassessed when more evidence regarding the clinician relevance of these mutations with the current dose becomes available. It was beyond the remit of this review to assess the role of therapeutic drug monitoring or whether a higher of RIF can overcome the MIC increase typically associated with these borderline resistance mutations.98,106,113,114 However, these questions will become a priority should WHO endorse a higher dose of RIF in the future. The TEG considered two approaches to minimize the misclassification of borderline resistance mutations by pDST. First, Torrea et al. found that when the incubation period of MGIT is extended, the MICs of isolates with borderline resistance mutations increased more than those of pWT isolates (i.e. the overlap between the MIC distributions is reduced, resulting in an improved ability to resolve both populations).92 However, any pDST methodology change would require an extensive revalidation of MGIT prior to implementation (e.g. the CC may have to be adjusted), which means that this option would not be feasible in the short term. Second, several experts noted that adopting ATUs by testing two concentrations rather than one (i.e. the CC and the concentration below) would not be currently feasible in most settings.115 In light of these limitations of pDST, the TEG endorsed the following expert rules: 1. gDST is the only reliable way of detecting RIF resistance caused by borderline mutations, which means that countries that do not conduct routine gDST as the initial test using an appropriately validated method risk missing these likely clinically relevant mechanisms of resistance. 2. Because pDST is not a reliable confirmatory method, there is little value in conducting pDST for RIF if a borderline resistance mutation is specifically detected using gDST upfront. Therefore, pDST is not needed in these cases. If pDST is unavoidable, the detection of one of these mutations by sequencing or a mutant probe should overrule a susceptible pDST result at the CC after a review of all results to rule out obvious laboratory or clerical errors (i.e. if there is a concern regarding the positive predictive value of the genotypic result, gDST, ideally using an alternative method, is the 112 World Health Organization. WHO consolidated guidelines on tuberculosis. Module 4: treatment – drug-resistant tuberculosis treatment (https://apps.who.int/iris/rest/bitstreams/1280998/retrieve, accessed 15 June 2020). 113 Jeong DH, Kang YW, Kim JY, et al. Successful treatment with a high-dose rifampin-containing regimen for pulmonary tuberculosis with a disputed rpoB mutation. Intern Med. 2018;57(22):3281-3284. doi:10.2169/internalmedicine.9571-17. 114 van den Elsen SHJ, Akkerman OW, Wessels M, et al. Dose optimisation of first-line tuberculosis drugs using therapeutic drug monitoring in saliva: feasible for rifampicin, not for isoniazid. Eur Respir J. 2020 Oct 22;56(4):2000803. doi:10.1183/13993003.00803-2020. 115 European Committee on Antimicrobial Susceptibility Testing. Area of Technical Uncertainty (ATU) in antimicrobial susceptibility testing (http://www.eucast.org/fileadmin/src/media/PDFs/EUCAST_files/Breakpoint_tables/Area_of_Technical_Uncertainty_- _guidance_2019-1.pdf, accessed 15 February 2020). 85 appropriate confirmatory approach). 116,117 In other words, the genotype is designated as the reference standard for borderline resistance mutations. 3. Because additional rpoB mutations likely exist that are equally difficult to detect using pDST as the seven borderline resistance mutations defined in this review (Table 26) and have the same clinical consequences, the TEG reaffirmed the expert rule that any mutation in RRDR, apart from synonymous mutations, should be assumed to confer RIF resistance, as stated in the current technical manual for DST (i.e. even if these mutations are novel).118 The possibility that some mutations in RRDR are neutral and do not affect the MIC at all and/or have no adverse clinical consequences was acknowledged, but to confirm this possibility conclusively, MIC testing would have to be carried out for multiple replicates of the mutant in question along with an on-scale quality control strain and measures would have to be taken to rule out other confounders (e.g. low-frequency resistance mutations missed by the initial gDST method).94,119,120 This is not possible as part of routine clinical care but should be prioritised by reference laboratories for novel mutations and mutations that increase in frequency over time. Moreover, sufficiently powered studies would have to be conducted to investigate the clinical impact of such mutations. Depending on the outcomes of such studies, WHO will modify this expert rule (e.g. by excluding individual RRDR mutations96). Finally, the TEG agreed that WHO should provide more practical guidance to implement these expert rules, and to explain the possible reasons for discordant DST results (e.g. because not all gDST assays interrogate the same parts of rpoB (Table 27), and to address the possibility of systematic errors due to synonymous mutations).121 In this context, it was emphasized that laboratories ensure that they follow the latest interpretation guides for gDST assays. For instance, Hain recently updated under which conditions “WT” bands are interpreted as negative to minimize false-susceptible results due to the presence of borderline resistance mutations. 122 This change is already included in the ELI interpretation guide but has yet to be recognized by GLI.116,123 116 World Health Organization. Regional Office for Europe. Interpretation guide for GenoType MTBDRplus VER 2.0 and GenoType MTBDRsl VER 2.0. A technical guidance document developed by the European Laboratory Initiative. Version 1.0. (https://openwho.org/courses/multi-drug-resistant-tb, accessed 9 May 2020). 117 Miotto P, Cabibbe AM, Borroni E, Degano M, Cirillo DM. Role of disputed mutations in the rpoB gene in interpretation of automated liquid MGIT culture results for rifampin susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol. 2018;56(5):e01599-17. doi:10.1128/JCM.01599-17. 118 World Health Organization. Technical manual for drug susceptibility testing of medicines used in the treatment of tuberculosis. (http://apps.who.int/iris/bitstream/handle/10665/275469/9789241514842-eng.pdf, accessed 17 November 2018). 119 Jagielski T, Bakuła Z, Brzostek A, et al. Characterization of mutations conferring resistance to rifampin in Mycobacterium tuberculosis clinical strains. Antimicrob Agents Chemother. 2018;62(10):e01093-18. doi:10.1128/AAC.01093-18. 120 Schön T, Matuschek E, Mohamed S, et al. Standards for MIC testing that apply to the majority of bacterial pathogens should also be enforced for Mycobacterium tuberculosis complex. Clin Microbiol Infect. 2019;25(4):403-405. doi:10.1016/j.cmi.2019.01.019. 121 Omar SV, Hillemann D, Pandey S, et al. Systematic rifampicin resistance errors with Xpert® MTB/RIF Ultra: implications for regulation of genotypic assays. Int J Tuberc Lung Dis. 2020;24(12):1307-1311. doi:10.5588/ijtld.20.0396. 122 Hain Lifescience. GenoType MTBDRplus VER 2.0. Instructions for use. IFU-304A-09. (https://www.hain- lifescience.de/include_datei/kundenmodule/packungsbeilage/download.php?id=2877, accessed 1 Nov 2020). 123 Global Laboratory Initiative. Line probe assays for drug-resistant tuberculosis detection: interpretation and reporting guide for laboratory staff and clinicians; 2019 (http://www.stoptb.org/wg/gli/assets/documents/LPA_test_web_ready.pdf, accessed 22 May 2020). 86 Rifabutin RFB is not currently recommended by WHO for the treatment of active or latent TB.112,124 Therefore, the TEG did not set a CC for this rifamycin. Nevertheless, some researchers are of the opinion that RFB is effective for specific rpoB mutations that confer resistance to RIF.125 The most important example is D435V (D516V), as this mutation can be differentiated from other rpoB mutations using the Hain and Nipro LPAs (Table 27).126 This and other rpoB mutations appear to confer lower relative MIC increases for RFB than for RIF. Indeed, the mode of D435V (D516V) for RIF is approximately 20 times higher than the newly endorsed CC of 0.5 mg/L in MGIT (Table 47), whereas the mode of D435V (D516V) for RFB is identical to or only slightly above the tentative ECOFF of 0.125 mg/L (Table 62). Consequently, isolates with this mutation are still clearly pNWT and are only classified as RFB-S (i.e. treatable with RFB) because the CLSI CC of 0.5 mg/L is 4 times higher than the tentative ECOFF in MGIT of 0.125 mg/L.127 The same applies for 7H10, where the CLSI CC of 0.5 mg/L is 4-8 times higher than the tentative RFB ECOFF (i.e. 0.06 or 0.125 mg/L).128,129 The rationale used by CLSI to set the CCs for RFB at 0.5 mg/L is unclear, but the work by Heifets et al. likely influenced this decision.130 Heifets et al. had originally proposed 0.125 mg/L as the RFB CC for 7H10, 7H11 and the radiometric BACTEC™ 460 (BACTEC) method with 7H12 medium given that 17 RIF-S isolates had RFB MICs 0.06 mg/L (0.125 rather than 0.06 mg/L was chosen to buffer for potential methodological differences).131,132 Moreover, the authors raised the possibility that pNWT isolates with only slight MIC increases might be treatable in light of the serum concentrations and concentrations in selected tissues achievable with RFB. However, the MICs that the authors considered to be associated with “moderately susceptible” isolates changed over time (i.e. they narrowed their original range of 0.25-0.5 mg/L for BACTEC, which they found to yield equivalent MICs to 7H10 and 7H11, to just 0.25 mg/L).131,132 A subsequent study by Pfyffer et al., which is also cited by CLSI, only tested 0.5 mg/L as the lowest concentration for 7H10 and did not consider any rpoB sequencing information.133 Therefore, this publication provided no insight into the shape of the pWT MIC distribution and the authors provided few details to justify their choice of 1 mg/L as the RFB CC for 7H10. Similar limitations apply to the subsequent study by Rüsch-Gerdes et al. that informed the 124 World Health Organization. WHO consolidated guidelines on tuberculosis. Module 1: prevention – tuberculosis preventive treatment (https://apps.who.int/iris/rest/bitstreams/1270183/retrieve, accessed 24 May 2020). 125 Domínguez J, Boettger EC, Cirillo D, et al. Clinical implications of molecular drug resistance testing for Mycobacterium tuberculosis: a TBNET/RESIST-TB consensus statement. Int J Tuberc Lung Dis 2016;20(1):24-42. doi:10.5588/ijtld.15.0221. 126 Williams DL, Spring L, Collins L, et al. Contribution of rpoB mutations to development of rifamycin cross-resistance in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1998;42(7):1853-7. doi:10.1128/AAC.42.7.1853. 127 Heyckendorf J, Andres S, Köser CU, et al. What is resistance? Impact of phenotypic versus molecular drug resistance testing on therapy for multi- and extensively drug-resistant tuberculosis. Antimicrob Agents Chemother. 2017;62(2):e01550- 17. doi:10.1128/AAC.01550-17. 128 Ängeby K, Juréen P, Kahlmeter G, Hoffner S, Schön T. Challenging a dogma: antimicrobial susceptibility testing breakpoints for Mycobacterium tuberculosis. Bull World Health Organ. 2012. doi:10.2471/blt.11.096644. 129 Schön T, Juréen P, Chryssanthou E, et al. Rifampicin-resistant and rifabutin-susceptible Mycobacterium tuberculosis strains: a breakpoint artefact? J Antimicrob Chemother. 2013;68(9):2074-2077. doi:10.1093/jac/dkt150. 130 Clinical & Laboratory Standards Institute. Clinical and Laboratory Standards Institute. Susceptibility testing of mycobacteria, nocardiae, and other aerobic actinomycetes, 3rd edition approved standard. CLSI Document M24; 2018. 131 Heifets LB, Lindholm-Levy PJ, Iseman MD. Rifabutine: minimal inhibitory and bactericidal concentrations for Mycobacterium tuberculosis. Am Rev Respir Dis. 1988;137(3):719-721. doi:10.1164/ajrccm/137.3.719. 132 Heifets LB. Drug susceptibility tests in the management of chemotherapy of tuberculosis. In Heifets LB, ed. Drug susceptibility in the chemotherapy of mycobacterial infections. CRC Press; 1991. 133 Pfyffer GE, Bonato DA, Ebrahimzadeh A, et al. Multicenter laboratory validation of susceptibility testing of Mycobacterium tuberculosis against classical second-line and newer antimicrobial drugs by using the radiometric BACTEC 460 technique and the proportion method with solid media. J Clin Microbiol. 1999;37(10):3179-3186. doi:10.1128/JCM.37.10.3179-3186.1999. 87 MGIT CC of 0.5 mg/L (i.e. the lowest concentration tested was only 0.25 mg/L and no rpoB results were considered).134 Thus, there appears to be no strong evidence to support the RFB CCs set by CLSI. It is possible that a minority of rpoB mutations (e.g. some borderline RIF resistance mutations) have no clinically relevant effect for RFB. Yet, until sufficiently powered studies have been conducted to demonstrate this, the cautious approach would be to assume full cross-resistance with RIF. 135 Specifically, gDST and pDST results for RIF instead of RFB would maximize the detection of rpoB mutations (i.e. in the same way that TEG recommends RIF as a surrogate for RPT). Finally, some experts noted that even if some rpoB mutations that were easily identifiable using gDST could be treated with RFB, pDST would still be needed to exclude unusually high RFB MICs (e.g. although the MICs for D435V (D516V) are typically below the current CLSI CC, isolates with considerably higher MICs exist (Table 53)). For this reason, trial NCT02236078, which was designed to investigate the potential utility of RFB and is due to complete at the end of 2020, relies on direct pDST to generate timely results.136 Rifapentine RPT is currently recommended only for the preventative treatment of TB.112,124 Given the lack of available MIC data for this antibiotic, the TEG recommended that complete cross-resistance with RIF should be assumed until sufficient data to the contrary are available (i.e. gDST and pDST results for RIF should be used as the surrogate for RPT).126,137 This point will be reevaluated based on the findings of the phase 3 NCT02410772 trial, which investigates whether the treatment of drug-susceptible TB can be shortened by replacing RIF with high-dose RPT.138 134 Rüsch-Gerdes S, Pfyffer GE, Casal M, Chadwick M, Siddiqi S. Multicenter laboratory validation of the BACTEC MGIT 960 technique for testing susceptibilities of Mycobacterium tuberculosis to classical second-line drugs and newer antimicrobials. J Clin Microbiol. 2006;44(3):688-692. doi:10.1128/JCM.44.3.688-692.2006. 135 Weiner M, Benator D, Burman W, et al. Association between acquired rifamycin resistance and the pharmacokinetics of rifabutin and isoniazid among patients with HIV and tuberculosis. Clin Infect Dis. 2005;40(10):1481-1491. doi:10.1086/429321. 136 Click ES, Kurbatova EV, Alexander H, et al. Isoniazid and rifampin-resistance mutations associated with resistance to second-line drugs and with sputum culture conversion. J Infect Dis. 2020;221(12):2072-2082. doi:10.1093/infdis/jiaa042. 137 Moghazeh SL, Pan X, Arain T, et al. Comparative antimycobacterial activities of rifampin, rifapentine, and KRM-1648 against a collection of rifampin-resistant Mycobacterium tuberculosis isolates with known rpoB mutations. Antimicrob Agents Chemother. 1996;40(11):2655-2657. doi:10.1128/AAC.40.11.2655. 138 Dorman SE, Nahid P, Kurbatova EV, et al. High-dose rifapentine with or without moxifloxacin for shortening treatment of pulmonary tuberculosis: study protocol for TBTC Study 31/ACTG A5349 phase 3 clinical trial. Contemp Clin Trials. 2020;90:105938. doi:10.1016/j.cct.2020.105938. 1 7.D.1