Drug resistance in the Swedish/Norwegian BCG strain

Bull. Org. mond. Santc 1972, 46, 503-507 Bull. Wid Hlth Org. Drug resistance in the Swedish/Norwegian BCG strain I. HESSELBERG 1 When a BCG strain is maintained by serial subcultures for several years, a change in its genetic properties may occur spontaneously. Such a change occurred in a strain maintained in the Norwegian State BCG Laboratory for 11 years. When received from Sweden in 1953, the strain was sensitive to the main tuberculostatics. However, in 1964, by mere chance, it was found to have become resistant to isoniazid. An examination of extant samples from the batches offreeze-dried vaccine produced during the period 1953-64 enabled the time and speed of the change to be determined. Although an alteration in drug sensitivity may not be harmful, it could cause a loss in the protective potency ofa vaccine in man. It was to safeguard against this danger that the system of serial subcultures was discontinued and the seed lot system adopted, in accordance with WHO requirements. Between 1953 and 1965, the Norwegian State BCG Laboratory, Bergen, employed a BCG strain obtained from the BCG Laboratory in Goteborg, Sweden, in August 1953. Shortly after its receipt, this strain was tested at the Gades Institute, Bergen, and found to be sensitive to streptomycin, isoniazid, and PAS. During a vaccination experiment in Stavanger in September 1964, one of the vaccinators accidentally pricked a finger and the lesion became infected with vaccine from a newly opened ampoule (No. 535) from reference batch No. 110, produced at the laboratory on 13 February 1958 from the above- mentioned Swedish strain. Following this accident, the vaccine remaining in the ampoule was examined by the State Bacterio- logical Laboratory, Stavanger. Marginal resistance to isoniazid was found in the culture grown from this BCG. MATERIALS AND METHODS In order to verify this finding, cultures from ampoule No. 535, from another ampoule of vaccine selected randomly from the same batch, and from an ampoule of vaccine produced from a new culture at the BCG Laboratory in Goteborg (No. 171) were simultaneously examined at three Norwegian 1 Chief, Statens BCG Laboratorium, Bergen, Norway. laboratories' for resistance to streptomycin, iso- niazid, and PAS. At all three laboratories, the cultures were found to be sensitive to streptomycin. The results for isoniazid and PAS are shown in Table 1. All the Table 1. Tests of resistance to isoniazid and PAS in three BCG vaccines: results from three Norwegian laboratories Results of resistance tests Laboratory No. 110 No. 110 (ampoule (randomly select- No. 171 No. 535) ed ampoule) ISONIAZID Bergen resistant resistant sensitive Stavanger marginal marginal sensitive Oslo marginal marginal sensitive PAS Bergen sensitive a resistant b sensitive Stavanger sensitive sensitive sensitive Oslo sensitive marginal sensitive a Resistant with the micromethod. b Marginal with the micromethod. I At Bergen, Stavanger, and Oslo, in 1965. 2825 503- 504 I. HESSELBERG Table 2. Tests of resistance to isoniazid and PAS in three BCG vaccines and Swedish/ Norwegian Primary Seed Lot I (PSL 1): results from one Norwegian and three foreign laboratories Results of resistance tests Laboratory Pasteur No. 8No. 110 No. 183 PSL I (control) ISONIAZID Bergen marginal sensitive/marginal sensitive resistant G6teborg sensitive/marginal sensitive/marginal sensitive resistant Copenhagen sensitive sensitive sensitive resistant Paris resistant sensitive sensitive resistant PAS Bergen sensitive/marginal sensitive sensitive sensitive/marginal G6teborg resistant sensitive/marginal sensitive resistant/marginal Copenhagen sensitive sensitive sensitive sensitive Paris resistant sensitive sensitive resistant laboratories demonstrated resistance to isoniazid in the cultures from ampoule No. 535 and the other ampoule from batch No. 110. The findings con- cerning resistance to PAS varied for the randomly selected ampoule from batch No. 110, whereas vaccine No. 171 was found to be sensitive to all three drugs. There is thus close agreement between the results obtained by the three laboratories with regard to resistance to isoniazid. In 1968-69 a further examination was made in co- operation with the Gades Institute, Bergen; the Gote- borg State Bacteriological Laboratory; the State Serum Institute, Copenhagen; and the Pasteur Insti- tute, Paris. Two coded ampoules of each of the following freeze-dried vaccines were sent to each of these laboratories: (1) vaccine from batch No. 110; (2) vaccine No. 183, produced in Bergen on 4 Feb- ruary 1965 from a Swedish strain sent from Goteborg in September 1964; 1 (3) the joint Swedish/Norwegian Primary Seed Lot I, produced in Goteborg on 5 April 1965; and (4) Pasteur No. 8 vaccine produced in Paris from an isoniazid-resistant BCG strain, as a control. 1 This strain was sent as a direct result of the discovery that resistance to isoniazid had been found in vaccine from a previous batch examined in Bergen (see page 503). The vaccines were examined for sensitivity to streptomycin, isoniazid, and PAS. On the whole, the results from Bergen, Goteborg, and Paris tally satisfactorily. All the vaccines were shown to be sensitive to streptomycin. The results for isoniazid and PAS are given in Table 2. Primary Seed Lot I was characterized as sensitive in all cases, whereas, for the control vaccine, Pasteur No. 8, resistance to isoniazid was confirmed and a certain degree of resistance to PAS was indicated. Three of the four laboratories found batch No. 110 to be resistant or to show marginal resistance to both isoniazid and PAS. The divergence in the results from Copenhagen may perhaps be due to the method employed, as the vaccines were examined only in dilutions 10-5 and 10-6. It was thought to be of interest to ascertain at what time and how rapidly this resistance to iso- niazid occurred in the BCG culture maintained since 1953. The BCG laboratory in Bergen stores ampoules from all batches of freeze-dried vaccines that it pro- duces. Therefore it was possible to send ampoules of the vaccines produced at intervals of about 1 year during the period 1953-60 to the Gades Institute, Bergen, for drug resistance testing. Ampoules from Primary Seed Lot I were sent at the same time as a control. The vaccines were tested for sensitivity to strepto- DRUG RESISTANCE IN SWEDISH/NORWEGIAN BCG STRAIN mycin, isoniazid, and PAS. No sign that resistance to streptomycin had developed during the period 1953-65 could be demonstrated. The results for isoniazid and PAS are given in Table 3. This table also shows the results for one fresh vaccine (No. A-303), produced in 1965 from the strain under study. The strain was shown to have been sensitive to PAS up to 1956. It was designated as having been " sensitive? " in 1957 and " marginal " from 1958 on. Sensitivity to isoniazid up to 1956 was demonstrated. The designation for 1956 and 1957 was " sensitive ", for 1958 " marginal ", and, from 1959 on, " resis- tant ". The control vaccine-Primary Seed Lot I- was sensitive to all three drugs. Thus it seems that the strain under study developed resistance to isoniazid between 1956 and 1959, while it was being continually subcultured in Bergen, and that it was still resistant in 1965. Resistance to PAS developed simultaneously, although to a lower degree. During the same period, the original strain, which is still being maintained at the BCG Labora- tory in Goteborg, has shown no modification in its resistance characteristics. Other changes in the strain under study also appear to have occurred during the first years of its Table 3. Development of resistance to isoniazid and PAS in the Swedish/Norwegian BCG strain during 1953-65 Year Vaccine Results (drug concentrations in Mg/ml of Lowenstein-Jensen medium) a Conclusion No. 0.5 1 3 10 50 ISONIAZID 1953 23 + - - - - sensitive 1954 37 + - + - - - sensitive 1955 61 + - + - - - sensitive 1956 78 + - + - + - + - marginal 1957 94 + _ + - + _ - marginal 1958 105 + + + + marginal 1959 123 ++ ++ ++ ++ resistant 1960 138 ++ ++ ++ ++ resistant 1965 A-303 ++ ++ ++ + + resistant 1965 b PSL I b + - + - _ - sensitive PAS 1953 23 ++ + - + - - sensitive 1954 37 + - - - sensitive 1955 61 ++ + - + - - sensitive 1956 78 ++ + - - - sensitive 1957 94 ++ + + - + - sensitive? 1958 105 + + + + + + marginal 1959 123 + + + + + + + marginal 1960 138 + + + + + - - marginal 1965 A-303 + + + + + marginal 1965 b PSL I b + _ - - - sensitive a +++ = full growth as in the control tube; +- = a few scattered colonies. b The Swedish/Norwegian Primary Seed Lot was used as a control. 6 505 506 I. HESSELBERG subculture at Bergen. The bacilli changed in appear- ance from short and blunt to relatively long and slender. The number of culturable particles per mg of semi-dried culture gradually decreased from an average of 40-50 million to 20-30 million. At the same time the oxygen consumption per mg of cul- ture mass dropped by approximately 50%. Despite the decrease in the number of culturable particles, however, no noticeable reduction in postvaccinal allergy was observed, and experiments on guinea- pigs revealed that the protective potency of the vaccine had remained unaltered. DISCUSSION Until 1965, the strain under study was maintained at the BCG Laboratory in Bergen by re-seeding every 14 days on bile-potato and Sauton potato media al- ternately, whereas bile-potato medium exclusively had been used at the BCG Laboratory in Goteborg. It has not been possible to pick out any other diffe- rence in working methods between the two labora- tories, but it is conceivable that the Norwegian meth- od of culture may have provided growth conditions especially favourable for rapid mutation in the origi- nal Swedish strain. It has been demonstrated earlier (Jespersen & Weis Bentzon, 1964; Bunch-Christensen et al., 1968, 1970; Ladefoged et al., 1970) that the characteristics ofBCG daughter strains may alter when they are grown in various places and by different methods. Such chan- ges have been observed not only in morphology and metabolism but also in virulence and protective potency. The strains may also develop varying resistance patterns to tuberculostatics (Rist et al., 1967). According to G. Canetti (personal communication, 1967) a BCG culture normally contains a certain number of isoniazid-resistant mutants-rather more than are usually found in a wild strain of Mycobac- terium tuberculosis. In the case of BCG, the number of mutants found is of the magnitude of 1 isoniazid- resistant mutant per 104 particles capable of germi- nating, whereas the corresponding figure for a wild strain is 1-10 mutants per 106 particles.' Canetti had not previously observed a spontaneous trend towards full resistance to isoniazid in BCG cultures- i.e., to a ratio greater than 1 mutant per 102 particles -except on passage through a medium containing isoniazid. Rist et al. (1967) assert that a description of the sensitivity of the BCG strain to PAS and isoniazid presents difficulties. Sensitivity appears to vary not only with the geographical origin of the strains but also with the examination methods employed. How- ever, it has been demonstrated that BCG strains con- tain a certain number of mutants that are resistant to up to 0.25-0.50 fig of PAS and 0.1 ,ug of isoniazid. This tendency towards a certain resistance to PAS and a reduced sensitivity to isoniazid is a well- known phenomenon, also where strainsofM. bovisare concerned. Exceptionally, however, the mutants in a BCG strainmay be so numerous that the strain must be described as resistant to PAS and/or isoniazid. In the past, many production laboratories have experienced spontaneous changes in their BCG strains, in a way similar to that described above. Some changes-e.g., those in morphology or the growth rate-are readily detected, whereas others (such as the one reported here) may go unnoticed or be detected only by chance. This is particularly true in the conceivable case of loss of protective potency in man, since nowhere in the world is this property systematically monitored by direct experi- mentation. Changes in the properties of strains ap- pear to be inherent in the classical procedure of maintenance by serial subcultures, and the only way to avoid them and to keep the genetic properties of a strain constant once these have been determined would seem to be to use the Seed Lot System (WHO Expert Committee on Biological Standardization, 1966). This was adopted for routine vaccine pro- duction by the BCG laboratories in Bergen and G6teborg in 1965, when a Primary Seed Lot of the Swedish strain was introduced. 1 This test was performed at the Pasteur Institute, Paris, on Lowenstein-Jensen medium with an isoniazid concen- tration of 0.2,ug/ml. The reading was made after 40 days. RESUME RESISTANCE AUX MIDICAMENTS DE LA SOUCHE SUEDO-NORVEGIENNE DE BCG En 1953, le Laboratoire norvegien du BCG a requ une culture de la souche su6doise de BCG provenant du Laboratoire du BCG a Goteborg. Jusqu'en 1965, cette culture a ete entretenue continuellement et utilisee pour la preparation usuelle de vaccin frais et lyophilise. Le Laboratoire est ensuite passe au systeme e lot de semence * DRUG RESISTANCE IN SWEDISH/NORWEGIAN BCG STRAIN 507 avec un lot primaire commun su6do-norv6gien. DNs la r6ception de la culture su6doise en 1953, son schema de resistance avait 6te controle, et on avait constat6 qu'elle 6tait sensible A la streptomycine, a l'isoniazide et au PAS. On s'est apercu par hasard, en 1964, que le vaccin de reference norvegien lyophilise, produit en 1958, pr6sen- tait une r6sistance a l'isoniazide et au PAS. La determination des caract6ristiques de r6sistance de diff6rentes souches de BCG peut presenter certaines difficultes, 6tant donne que les r6sultats d6pendent des methodes de travail des differents laboratoires. C'est pourquoi quatre laboratoires bact6riologiques, un norve- gien et trois etrangers, pratiquerent en collaboration une etude de contr6le de la r6sistance. Cette 6tude permit d'6tablir qu'une certaine resistance A l'isoniazide et au PAS s'6tait d6veloppee dans la culture entretenue au Laboratoire norvegien du BCG. La meme culture, entre- tenue au Laboratoire du BCG de Goteborg entre 1953 et 1965, presentait un schema de resistance inchange, et continuait d'etre sensible A l'isoniazide et au PAS. Pour etablir a quel moment et de quelle faron cette resistance etait apparue, on a etudi6 les caract6ristiques de r6sistance d'une serie de vaccins lyophilises pr6par6s a un an d'intervalle entre 1953 et 1960. Cette 6tude a rev6le qu'entre 1956 et 1959 la culture 6tait progressive- ment devenue resistante a l'isoniazide, et en meme temps, mais a un degr6 moindre, au PAS. On n'a pas trouv6 d'explication certaine de ce deve- loppement de la r6sistance dans la culture entretenue au Laboratoire norvegien, alors que la culture faite en meme temps au Laboratoire de Goteborg est rest6e inchang6e. La raison peut en etre une difference dans les methodes de culture des deux laboratoires. Dans le Laboratoire norv6gien, la culture a ete entretenue par des repiquages altematifs bimensuels d'un milieu pomme de terre-bile a un milieu pomme de terre de Sauton, tandis que le Laboratoire de G6teborg utilisait exclusive- ment un milieu pomme de terre-bile. On peut donc supposer que la m6thode norv6gienne de culture offre des conditions favorables au d6veloppement d'un mutant a croissance rapide dans la culture su6doise initiale. REFERENCES Bunch-Christensen, K. et al. (1968) Bull. Wld Hlth Org., 39, 821-828 Bunch-Christensen, K. et al. (1970) Bull. Wld Hlth Org., 43, 65-70 Jespersen, A. & Weis Bentzon, M. (1964) Acta tuberc. scand., 44, 222-249 Ladefoged, A. et al. (1970) Bull. Wld Hlth Org., 43, 71-90 Rist, N. et al. (1967) Rev. Tuberc. (Paris), 31, 1060- 1065 WHO Expert Committee on Biological Standardization (1966) Wld Hlth Org. techn. Rep. Ser., No. 329