DNA probes or microvillar enzymes or both for prenatal diagnosis of cystic fibrosis.

SIR, Now that at least two DNA probes tightly linked to the cystic fibrosis (CF) gene are available,12 it makes sense to offer first trimester prenatal diagnosis to high risk couples when this is possible. The question that has to be considered is whether this should be combined with second trimester diaignosis based on measurement of amniotic fluid microvillar enzymes.3 In the most favoured situation a pregnancy will be fully informative for DNA analysis. This means the availability of DNA samples from both parents and an affected index child, a phase relationship that is unambiguous, and enough time to work up the family before chorionic villus sampling. If recombination frequencies between pJ3. 11, met, and the CF gene remain as low as reported,'2 DNA based prenatal diagnosis will be more accurate than that using microvillar enzymes. Posterior odds on either an affected or normal fetus will be sufficiently high to enable clinical decisions to be made immediately. The advantages of a first trimester termination of pregnancy are so great that they probably outweigh the possible benefit of being able to confirm the diagnosis on an abortus after a second trimester termination.34 If a fetus is deemed to be normal after DNA diagnosis, microvillar enzyme analysis two or three months later will either confirm already substantial odds or throw the whole situation into hopeless confusion. I believe that the temptation to 'make assurance doubly sure' should be resisted. At the opposite extreme are those cases where DNA based methods cannot be used. There may be