Feasibility study of community control programmes for cystic fibrosis: memorandum from a WHO/ICF(M)A meeting.

MemorandalMemorandums Feasibility study of community control programmes for cystic fibrosis: Memorandum from a WHO/ICF(M)A meeting* As a result of improved medical services and diagnosis of cystic fibrosis (CF), it has in recent years become clear that this is a common genetic disorder with a worldwide distribution. The average life expectancy of CF patients is very low in the developing world, but in developed countries this fatal childhood disease is becoming a chronic disorder persisting into adult life. In western Europe and North America the average life expectancy is now about 25 years, with a fatal outcome, so that CF represents one of the most serious of inherited life-threatening conditions. It remains to be shown whether very early diagnosis and treatment can further improve the prognosis. Introduction The first joint meeting between the International Cystic Fibrosis (Mucoviscidosis) Association (ICF(M)A) and WHO took place in Vienna in 1983.a The report of that meeting covered the clinical fea- tures and management of cystic fibrosis (CF), and emphasized the limited nature of current knowledge about its distribution, frequency, pathophysiology, and the underlying molecular defect. Subsequent progress in these areas has been reported at regular joint WHO/ICF(M)A meetings.b',, d, 6 The fourth *This Memorandum was drafted by the following participants at a joint WHO/ICF(M)A meeting in London on 12-14 June 1989: G. Adam, Athens, Greece; V. Baranov, Leningrad, USSR; A. Boue, Paris, France; J.A. Dodge, Belfast, Northern Ireland (Chairman) N. Fost, Madison, WI, USA; M. Gotz, Vienna, Austria (Rapporteur); J. Mangos, San Antonio, TX, USA; G. Mastella, Verona, Italy; B. Modell, London, England; A. Mohsin Mohammed, Manama, Bah- rain; O.H. Pivetta, Buenos Aires, Argentina; P. Scambler, London, England; B. Strandvik, Huddinge, Sweden (Co-Rapporteur); E. Tempany, Dublin, Ireland; B. Wilcken, North Ryde, N.S.W., Austra- lia; Y. Yamashiro, Tokyo, Japan. ICF(M)A Secretariat: M. Weibel, Uetendorf, Switzerland, and L. Heidet, Geneva, Switzerland. WHO Secretariat: V. Bulyzhenkov, Geneva, Switzerland. Requests for reprints of this article should be addressed to the Hereditary Diseases Programme, Division of Noncommunicable Diseases and Health Technology, World Health Organization, 1211 Geneva 27, Switzerland. A French translation will appear in a later issue of the Bulletin. a Report of a Joint WHO/ICF(M)A Meeting on Cystic Fibrosis, Vienna, 6-8 October 1983. Unpublished WHO document HMG/ICF(M)A/83.8, 1983. bReport of a WHO/ICF(M)A Meeting on the Distribution of Cystic Fibrosis, Nicosia, 30 April-2 May 1985. Unpublished WHO docu- ment HIM`G/ICF(M)A/85.2, 1985. c Report of a Joint WHO/ICF(M)A Meeting on Prevention and Control of Cystic Fibrosis, Oslo, 19 June 1987. Unpublished WHO document HDP/ICF(M)A/WG/87.3, 1987. Reprint No. 5125 meeting, which took place in London on 12-14 June 1989, evaluated progress in knowledge of epidemiol- ogy, diagnosis, treatment and prevention of CF, considered the steps that should now be taken in the developing and developed countries, and examined the implications of the impending identification of the CF gene for the future control of cystic fibrosis.' The identification of the gene for the cystic fibrosis transmembrane conductance regulator (CFTR), in which mutation can cause CF, and the nature of the commonest CF mutation were announced in Septem- ber 1989 (see Annex). It was concluded that the identification of the gene will transform the situation by facilitating the diagnosis of patients and prospec- tive carriers. It may also have an impact on treat- ment. One of the most important steps to prepare the population for these changes is to develop good quality educational and training materials for both developing and developed countries. Present state of knowledge Epidemiology Knowledge of the incidence of cystic fibrosis at present depends on correct diagnosis of affected d Report of a Joint WHO/ICF(M)A Meeting on Prevention and Control of Cystic Fibrosis, Sydney, 8 March 1988. Unpublished document WHO/HDP/ICF(M)AIWG/88.1, 1988. ' Report of a WHO Consultation "Proposal for a Joint WHO/ICF(M)A Programme for Neonatal Screening for Cystic Fi- brosis': Geneva, 22-23 August 1988. Unpublished document WHO/HDP/CONS/88.2, 1988. 'A control programme for an inherited disease requires a com- prehensive strategy that combines the best possible patient treatment and prevention based on community information, car- rier screening, counselling, and the availability of prenatal diag- nosis. Bulletin of the World Health Organization, 68 (6): 709-715 (1990) © World Health Organization 1990 709 Memorandum children (homozygotes), and calculation of the carrier (heterozygote) frequency from their birth incidence. Diagnosis of homozygotes is not simple. The sweat test is still the most reliable diagnostic method. In some developed countries there are neonatal pro- grammes based either on the meconium albumin test or, increasingly, on the measurement of trypsinogen in dried (Guthrie) blood spots. Infants with a positive result are followed up by sweat testing. The com- plexity of testing largely accounts for continuing limited knowledge of the distribution of cystic fibro- sis. Progress in molecular biology leading to the identification of the CF gene will greatly simplify diagnosis, both for clinical and for epidemiological purposes. Present information indicates that there are con- siderable differences between various parts of the world and among different ethnic groups. Neonatal screening of 400 000 babies for assessment of the incidence in various ethnic groups gave an apparent incidence among Caucasians from northern Europe including the United Kingdom, USA, South America, South Africa and New Zealand of 1:2200. Mothers from southern Europe, the Mediterranean area, Greece, Yugoslavia, Italy and Malta, appear to have CF children with a frequency of 1:3500. Mothers from the Middle East, Viet Nam and Oceania seem to have CF children with a frequency of 1:12 000. It can be assumed that the frequency in the European part of the USSR is the same as in northern Europe, and that 2-3000 newborn infants with CF can be expected in the USSR every year. The majority (95%) of developed countries are members of the ICF(M)A. Grouped together, they have a total of about 548 million people with 38 000 reported CF patients, i.e., a prevalence of 7:100 000; 20% of these patients are over 18 years old. By contrast, there is little reliable information on the incidence in countries in Asia, Latin America and elsewhere. Case-finding efforts have established the occurrence of cystic fibrosis in Egypt, India, Pakis- tan, Jordan, Kuwait and Cyprus and there is good evidence that CF also occurs in China. However, the prevalence of CF in developing countries collectively is only 0.19/100 000, and only 9% of known patients are more than 18 years old. (The figure increases to 0.4:100000 if Pakistan and India are excluded, as they have very few known cases). Even leaving out China with an approximate population of 1000 mil- lion and an uncertain, probably low, incidence of CF, there still remains a population of 3200 million people in 132 countries where the diagnosis of CF is grossly underestimated. The contribution of low awareness and of different mutations to the difference in incidence of the disease among different ethnic groups is not yet known. The development of DNA diagnostic methods based on identification of specific mutations is likely to radically change the confused epidemiological picture in the next few years. Neonatal screening There are two functions for neonatal screening of homozygotes with cystic fibrosis: (i) research (to ob- tain epidemiological information), and (ii) service (to prevent loss of children, through genetic counselling, and to minimize false alarms). There is some evidence that babies identified early, in the first one or two months of life, may have the best prognosis. Reliable diagnostic methods are needed for both purposes. At present the following two methods may be used: Meconium assay for Increased albumin and lactase levels. These tests have been used extensively. They are easy to perform and can be done locally, but interpretation is unreliable and results have been disappointing. Even in experienced hands and under ideal conditions, false negative rates of up to 40% have been reported and there is an appreciable false positive rate. The test can no longer be recommended for case-finding in developed countries or for epi- demiological studies. However, where equipment for other methods is not available and CF clinics are just beginning, it may be useful as a cheap alternative. Immuno-reactive trypsin (trypsinogen) (IRT) a"ay. This method has been widely used since it was first de- scribed in 1979. At present, at least six commercial kits are available, three of which have been widely used. There is also a noncommercially available radioimmunoassay. Monoclonal-antibody-based- sandwich analysis may also be used on dried blood spots, but experience so far is limited. Three strategies for screening using dried blood spots have been used. In 14 of 17 laboratories carry- ing out large-scale screening programmes, a two-tier system has been used in which a first elevated IRT level leads to a repeat IRT testing. The mean recall rate was 0.73% (range 0.29 to 4.66%). A second elevated IRT test makes sweat-testing mandatory, usually by the age of four to five weeks. Positives on the first test were on average 4.7%, false positives at the second test 0.2%. The false negative rate was low, at 6.4% of all CF cases (range 1.4 to 24), comparable to figures for congenital hypothyroidism screening (5%). Alternatively, babies with an elevated IRT level in the neonatal blood sample can be sweat-tested. This approach has been used in a large-scale pro- gramme in Wisconsin, USA. The Verona (Italy) lab- WHO Bulletin OMS. Vol, 68, 1990.710 Community control programmes for cystic fibrosis oratory uses meconium lactase assays in babies where elevated blood spot IRT levels are found. In this way a diagnosis can be made by two to three weeks of life, and this early diagnosis may be an advantage. At this stage, DNA analysis to confirm suspec- ted cases was considered too time-consuming and expensive, as it would depend on analysis of RFLP (restriction fragment length polymorphisms) haplo- type according to ethnic background. However, in the future, DNA diagnosis may become the method of choice. Difficulties regarding specificity and sensitivity have still not been overcome. False negative results pose a special problem, and most programme data are not adequate to assess completeness of ascertain- ment. The three strategies employed showed little difference in results; however, the efficiency of case- finding was quite variable. In New South Wales (Australia) newborn screening of 642 000 infants up to February 1989 revealed 239 with cystic fibrosis. Only 6.3% were missed by screening and diagnosed later, giving an incidence of 1:2680. An ELISA (enzyme-linked im- munosorbent assay) method seemed to miss more patients than radioimmunoassay: the reasons for this difference are not known. Conceivably, different mutations could have im- plications in neonatal screening, since screening de- pending on pancreatic abnormality may miss pa- tients with mutations associated with pancreatic sufficiency. However, it appears that 37% of all children of one to two years of age who were found to have a high neonatal IRT in Australia had ad- equate pancreatic function. In 14 programmes, well over 2 million newborns have now been screened using the IRT method, and a total of more than 730 cases have been found. In New South Wales, short-term cost-effectiveness for the first two years, due to a reduced number of hospital admissions, has been demonstrated. The benefit of establishing the diagnosis for the family is unquestionable. However it is agreed that a ran- domized controlled study is the only way to establish the clinical benefit of early diagnosis. There are as yet no results from such studies, and it was felt that until there are, screening might have to remain a research tool. In a randomized study in Wisconsin, USA, all newborns are screened but only half the cases are decoded. Access to data is always possible in case CF is suspected in a "control" child. Since it was felt that no harm arose for control children, this study is planned to continue for at least four years, and follow-up may possibly continue for 10 years. Ethical lessons learned from neonatal screening for phenylketonuria (PKU) are worth remembering. Once a screening programme has become mandatory it is difficult to stop it since zeal could obscure critical thinking. With neonatal screening for PKU it took about 10 years for minor and major problems asso- ciated with a high rate of false-positives and difficul- ties in getting the diet right. Screening is certainly not for everyone and requires parental consent and care- ful weighing of benefits and risks. Factors in favour of screening procedures are: the lack of carrier stigma associated with CF, the existence of a treatment network, and absence of ethnic overtones. Combination of neonatal screening for CF with screening for PKU may become feasible when a DNA-based test is available, and could reduce costs. It would also identify carriers, allowing appropriate genetic counselling to be given to their parents. Treatment New directions emphasized in treatment are region- alization of CF care in special centres, nutritional support, advanced new anti-microbial agents, inflam- mation modifiers (steroids), enhancement of the pulmonary defences, clarification of the role of other infective agents, psychosocial support and rehabili- tation, and heart/lung transplantation. Several important issues concern the use of antibiotics (continuous versus intermittent, and pro- phylactic treatment), immunizations, parental smok- ing, and new respiratory pathogens (Pseudomonas cepacia). Recent aspects of nebulized medication in- cluding bronchodilators, physiotherapy, heart/lung transplantation, gastrointestinal problems, diabetes and sclerotherapy for liver disease in CF were re- viewed. Virtually no studies on the long-term effects of the different kinds of treatment are available. Very few investigations have been documented even on long-term effects and the metabolic implications of physiotherapy. Treatment must be individualized and there is no evidence that any one kind is clearly superior to the others. Prolongation of life, improved quality of life, and support by those around the CF patient are of overriding importance. The apparent achievement of a survival "plateau" may imply that present methods are now reaching their maximum in terms of improved survival. The possibility of specific therapy directed against the underlying basic defect in cystic fibrosis would become clearer once the actual defect was characterized. DNA diagnosis will probably be useful in prog- nosis and possibly in treatment. At the time of writing, the CF locus had been established within an area of 200 KB between the D9 and G2 locus. There WHO Bulletin OMS. Vol. 68.1990. 711 Memorandum was some evidence for a second mutation indepen- dent of the major one leading to more cases with pancreatic sufficiency. Differences in clinical ex- pression related to different mutations might be specially pertinent for liver disease. The current cost of treatment for patient care in a developed country varies with the treatment given, but is estimated to be about US$ 7500 to US$ 15 000 per year per patient. Prevention At present, prevention of cystic fibrosis is "retrospec- tive". That means, it is based on diagnosis of affected infants and genetic counselling for their families. Since 1970, this has included the possibility of pre- natal diagnosis, which may allow parents with an affected child to continue to build up their family, including normal children. There has been steady progress in the reliability and acceptability of methods for prenatal diagnosis. Prenatal diagnosis of cystic fibrosis was first achieved by amniotic fluid microvillar intestinal en- zyme assay (MIE) in the mid-trimester of pregnancy. Subsequently, localization of the CF gene and identi- fication of restriction fragment length polymorphisms (RFLPs) closely linked to the CF locus has allowed DNA diagnosis to be made in the first trimester of pregnancy after chorionic villus sampling (CVS). By 1989, 90% of families with a 25% risk were fully informed, and a diagnosis could be made in 50% by using the KM19 probe only. In semi-informed fami- lies, 50% of fetuses could be confidently classified as healthy, but the rest must proceed to amniotic fluid MIE analysis. Among 321 families tested, 421 diag- noses had been made at the time of writing. In couples with a deceased affected child, it is now possible to combine the result of amniotic fluid MIE and the probabilities calculated from RFLP linkage data, especially when the parents are hetero- zygous for the KM19 probe. This allows the use of PCR (polymerase chain reaction) amplification tech- niques on amniotic fluid or chorionic villus material. The calculations are improved when there is DNA from a living normal child. Stored dried Guthrie blood spots can be used for DNA analysis of deceased children. The figures for the possibility of DNA diagnosis using linked RFLPs differ with ethnic background, as the haplotypes of the chromo- somes carrying the CF gene differ with ethnic group. RFLP linkage data can also be used in counselling couples with less than a 25% risk of a CF child, e.g., a CF-affected pregnant woman, the remarried parent of a CF child, and brothers and sisters of a patient. One study of 75 couples with a 25% risk referred for counselling for a second prenatal diagnosis showed limited uptake of further prenatal diagnosis. Of 44 couples who had a first affected fetus, 44% wanted further prenatal diagnosis; of 31 who had a normal fetus, only 17% wanted a further prenatal diagnosis. For families with an affected child, prenatal diagnosis based on analysis of the segregation of DNA markers linked to the CF gene is a reliable technique. The risk of chorionic villus sampling (CVS) is around 1-2%, which is comparable to the risk of amniocentesis, and lower than the risk of spontaneous abortion in the first trimester of preg- nancy. However, if even all known retrospectively-diag- nosed couples requested prenatal diagnosis, the effect on the birth rate of infants with CF would be very small, as only 10-15% of cases are born into families with an existing CF child. Most are born to couples not previously known to be at risk. Only "prospec- tive" detection of couples at risk, by population screening for carriers using a reliable test, can make a major impact on the incidence of the disease. Such tests are likely to become available when the exact locus of the gene has been determined, especially as cheap and non-invasive methods for sample collec- tion, such as a buccal mouth wash, may prove applicable. CF control programmes should begin to be planned now, for both developed and developing countries. Some implications of the availability of such techniques might be foreseen by considering existing experience in the haemoglobin disorders, where pros- pective carrier diagnosis and prenatal diagnosis have been possible for more than 10 years. The control of thalassaemia in the Mediterranean area may be seen as a working model. These programmes have shown that availability of technology for carrier screening and prenatal diagnosis is not enough. There must also be a national programme for community infor- mation, and professional education. An infrastruc- ture for carrier testing has to be developed, and programme monitoring is essential. Support associa- tions have a vital role to play in developing, and in obtaining government support for such programmes. The design of a control programme is highly depen- dent on local social and economic factors. Monitor- ing is best achieved by a patient register that is annually updated. Counselling of carriers may be done by trained health care professionals such as nurses, midwives and district nurses. Establishing an effective programme is more difficult in larger coun- tries, because of the need to incorporate it into the entire health care system. Cost-benefit calculations for prevention of cystic fibrosis must be viewed in financial, social and physi- cal terms, but financial aspects are usually the only WHO Bulletin OMS. Vol. 68. 1990.712 Community control programmes for cystic fibrosis ones measured, and can be very persuasive to admin- istrators. The life-expectancy of an average CF patient in a developed country today is about 25 years, so prevention by prenatal diagnosis and termi- nation of pregnancy when there is an affected fetus would save 25 times US$ 500-15 000, i.e., a minimum of US$ 187 500 (undiscounted). Assuming the cost of carrier screening and counselling to be about US$ 25/person, with a CF incidence of 1: 3000, and assuming 50% uptake of abortion, the saving per case would be at last US$ 65 000. Promotion of professional education, equitable carrier testing with quality control and counselling for all couples at risk must be stressed. The final aim would be incorporation of screening and counselling into primary health care. Models for the future must be developed carefully in each country in accordance with the local national health structure. Research needs At the time of the meeting, it was foreseen that progress in molecular research would soon lead to understanding of the basic defect in cystic fibrosis. This might lead to improved control of the disease by: -establishing a cure; -carrier detection with prenatal diagnosis; and improved genetic counselling. Progress in all these areas is to be expected. Molecular methods will also probably make it pos- sible to further research by creating transgenic ani- mal models of the disease, though the suggestion may not meet with universal approval. Despite tremendous progress in symptomatic treatment, further clinical research efforts are needed on a variety of topics, such as antibacterial treatment, surgical interventions (transplantation), overcoming colonization and infection, and evaluating the effec- tiveness of early clinical management of patients diagnosed by neonatal screening. Support for CF research is essential. Further trends may concentrate on joint programmes, exchange of materials, and recruitment of experts. Organization of control programmes Strategies in developed countries with a known CF burden Every child has the right to as good a quality of life as possible; this right applies to CF patients as well, and prevention and control efforts are needed to decrease human suffering, to satisfy human rights, and to eliminate the high cost of care. Two categories of CF prevention programmes are feasible for devel- oped countries: * Programmes without effective heterozygote detec- tion (retrospective prevention, which is now possible). * Programmes based on effective CF heterozygote detection in whole populations (prospective preven- tion, which is likely in the future). Since programmes without heterozygote detec- tion are based on diagnosis of affected children, the education of physicians and other health profes- sionals is of prime importance. The availability of well-trained health professionals who can diagnose CF is currently the most powerful means of controll- ing CF in developed countries where prenatal diag- nosis should be provided to all high-risk families. Education of the public on CF is essential and should be aimed at adults and could be delivered through schools to children. National CF associations are critically import- ant in involving the public and influencing the gov- ernment to contribute facilities for diagnosis, delivery of care, and improved quality of life. The associations should also make funds available for improvement of care and for research. Neonatal diagnostic programmes using various tests have been introduced in developed countries on the assumption that early diagnosis leads to better prognosis and more effective prevention and control, but, based on their evaluation, the conclusions for or against the implementation and methods for neonatal screening in developed countries have not been reached. The results of the controlled trial at present under way will be useful. Neonatal screening also contributes to genetic counselling. Future programmes with prospective CF hetero- zygote detection will involve screening the whole population before the reproductive age and will re- quire a high level of organization and funding (see Annex). Appropriate government agencies must be involved, and governments must be convinced in order to allocate special funds, separate from other health care needs of the country. It is important at an early stage to develop a plan for evaluating the effectiveness of the programme, using methods that will guarantee objectivity. Predicting the effectiveness of a prevention and control programme in any given country is difficult and may not be possible at this stage. Certainly, the goal of eliminating CF does not appear realistic. On a warning note, problems may arise from the stigma of gene detection. However, the Cyprus thalassaemia control programme has shown that information on carrier status is used not for marital but mainly for reproductive decision-making. Not enough educational material on CF is available for WHO Bulletin OMS. Vol. 68. 1990. 713 Memorandum worldwide use and these should be developed by individual countries. Recommendations on tests and on the qualifications of screening centres also need to be developed. Strategies for developing countries with an unknown cystic fibrosis burden It may be difficult to interest the local health authori- ties in many developing countries to consider a search for children affected with CF because of the relative lack of epidemiological information and their preoccupation with other major causes of infant mortality such as infectious diseases. Only when basic health care is improved does CF come to the surface as a major problem. In addition, if there is nothing to offer after the diagnosis is made, the effort of finding patients may be questionable. In developing countries the burden of inherited disease falls even more heavily on the family than in developed countries, because of the limited amount of social and medical support available. Therefore prevention is even more important in developing than in developed countries, and new possibilities for prevention are more likely to gain support even if prevention and treatment are inseparable. At present, the question of the right approach to cystic fibrosis in developing countries is also a matter for research. Centres for diagnosis and research should be established in developing countries, the cost of pro- viding these resources for diagnosis and treatment being small compared with the funds spent by major CF foundations throughout the world. Local spon- sors should be found, so that local centres could start to become independent. Guidelines for people starting to learn about cystic fibrosis in each country are of great import- ance. ICF(M)A should recognize these problems and, together, with WHO, should offer support to start regional CF centres. At present, Africa (excluding the north) and China should not be included in the list of countries where the establishment of CF diagnosis and care facilities is important. In contrast, in India the frequency of CF could be similar to that in Europe and the supported programmes should be concerned with prevention and control as well as education about cystic fibrosis. The question of whether unearthing the problem may be a Pandora's box cannot be answered, but it is likely that the search would lead to an eventual reduction in the numbers of affected patients through genetic coun- selling. Ignoring the problem cannot be the answer. It was concluded that one centre per country might be a sensible starting point for research, diag- nosis and treatment. This would apply especially to South America. Once the epidemiology of the disease was established in a country, other centres could follow. The setting up of a clinic, even with limited care of patients, must be considered as an education- al exercise. Comprehensive treatment as well as pre- vention activities should spread from such centres. Conclusions and recommendations Since the joint WHO/ICF(M)A meetings in Vienna (1983) and Oslo (1987), considerable progress has been made in prenatal diagnosis. The goals to be tackled are the following: reaching a decision on neonatal screening programmes, establishing the gene frequency in different populations, implemen- ting pilot programmes for prenatal diagnosis and selective abortion of affected pregnancies, assessing the costs, and introducing educational programmes. The latter were considered to be of paramount im- portance on the basis of recommendations of the ICF(M)A to WHO. Providing expertise through WHO and ICF(M)A, and inclusion of cystic fibrosis in the curriculum for medical students will be an inexpensive and effective way of spreading knowledge on this disease. There was agreement on the follow- ing major proposals: * The programme of neonatal screening (outlined in WHO document WHO/HDP/CONS/88.2) should be pursued and should also include Bahrain (the pro- gramme in Costa Rica has now started). The methods to be employed are open at this time, and although radioimmunoassay appears the most re- liable, the ease of performance remains questionable. Quality control must be introduced in view of a lack of reference standards. A control committee should be established under the auspices of WHO. * Educational programmes must be established in conjunction with experts in haemoglobin disorders and medical genetics. Educational material devel- oped at an international level should go to national organizations. At the same time, the available nation- al materials should be sent to health education ex- perts including medical specialists, health educators, teachers and others. * Preparation of a manual for the comprehensive management of cystic fibrosis might be worth under- taking under the auspices of WHO/ICF(M)A. This could be made available to the public and centres worldwide. Collection of information on DNA analy- sis should be undertaken and could be processed by the ICF(M)A Scientific Medical Advisory Committee. WHO Bulletin OMS. Vol. 68. 1990.714 Community control programmes for cystic fibrosis Annex Identification of the CF gene The isolation, sequence and major mutation of the gene at the cystic fibrosis locus was described early in September 1989.8, b The gene itself is large and has the structure of a trans-membrane conductance regu- lator (CFTR). It is similar to a previously known gene family encoding the P-glycoproteins, alterna- tively known as multiple drug resistance (MDR) proteins. It is not clear whether the protein itself transports chloride ions or whether it regulates the activity of a protein which does. The major mutation discovered consists of a deletion of a 3-base pair sequence encoding a phenylalanine residue, which is absent from the protein in CF cells. Some 68% of chromosomes carrying CF had this deletion; work is in progress to determine the nature of the other mutations. It is of interest that CF patients with pancreatic sufficiency (PS) tend to have one chromo- some with the 3-base pair deletion and one chromo- some with another, as yet unknown, mutation, i.e., they are compound heterozygotes. Patients with pan- creatic insufficiency (PI), on the other hand, are usually homozygous for the 3-base pair deletion, suggesting that either this mutation alters protein function more severely than the others or that the ' Rommens, J.M. et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science, 245: 1059-1065 (1989). bRiordan, J.R. et al. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science, 245: 1066-1073 (1989). phenylalanine residue is particularly important for functional regulation in the pancreas. At the clinical level the immediate benefits of this advance, after all the mutations have been char- acterized, are antenatal diagnosis with a theoretical accuracy of 100% and carrier detection for random members of the population. In the same way it will be possible to confirm a diagnosis of cystic fibrosis in doubtful cases. However, diagnostic difficulty is more likely to occur in milder forms of the disease for which the mutation has not yet been determined. With the availability of the polymerase chain reac- tion (PCR) it is now possible to detect the major CF mutation using minute quantities of blood. Should this procedure become automated, a method of neonatal screening/heterozygote detection would be feasible using the blood on a Guthrie card. With the necessary political will and financial resources, it may be possible as well as desirable to offer total popula- tion screening for the CF mutation(s). Such action would perhaps not only improve family planning services but potentially ameliorate the course of the disease in affected individuals. Current estimates are that there may be more than 30 mutations, with differing frequencies in different populations. It is likely to be some time before these have all been characterized. WHO Bulletin OMS. Vol. 68. 1990. 715