Early diagnosis of acute myocardial infarction

SIR,-Further to the case report by Dr Felix H Sennhauser and colleagues' we have screened 15 partners of people heterozygous for cystic fibrosis from Aberdeen. We found one couple in which the partner also carried a cystic fibrosis mutation (AF508). Prenatal diagnosis was not undertaken for this couple because the mutation in the known carrier (the mother, who had had an affected child by her previous partner) could not be identified and because gestation was too far advanced at presentation. The mother's cystic fibrosis-bearing chromosome was subsequently identified by using a linked marker (pJ3. 11), permitting future prenatal diagnosis. In six other couples we originally screened the partner for only AF508, and in the other eight nine mutations were excluded. The residual risk of having an affected child despite a negative test result in the partner with no family history should be calculated by using Bayes's theorem, and not by the simple method used by Dr Sennhauser and colleagues. They calculated the risk that the man is a carrier and that the woman is A508 negative, given that she is a carrier (the joint probability), whereas we wish to know the risk that the woman is a carrier, given that she is AF508 negative (the posterior probability). The Bayes's risk (with Sennhauser and colleagues' notation) is 1/4X ((I -a) x q)/((l -a) x q)+ p) or 1 in 257-a small but important increase over their result (1 in 267). It should also be noted that the probability of AF508 not occurring in an unrelated partner is 1-aq, or 96-5%, and not 98 4% as stated by Sennhauser and colleagues. The mutation AF508 accounts for 73% of cystic fibrosis mutations in Scotland, and a further 16% are accounted for by the less common mutations G551D, G542X, R117H, R553X, 1717G--A, R560T, 621+1G--*T, and A1507. For couples screened only for AF508 the residual risk calculated with a population carrier frequency of 1 in 25 is 1 in 360. If all the above mutations are excluded then the risk falls to 1 in 877, an extremely low value. Couples at increased risk commonly request screening when they know that it is available, although pregnancy is often the stimulus to presentation. If the obligate carrier does not carry an identifiable mutation (expected in about 11% of cases) a family study with linked markers is required. This takes much longer to organise than direct mutation screening of the presenting couple. There is a need for increased awareness among the public (and medical profession) of the possibilities and limitations of cystic fibrosis genetic testing.