Prenatal exclusion testing for Huntington's disease: a problem of too much information.

At eight weeks of pregnancy a couple were informed that the prospective father's mother had died of Huntington's disease (HD). There were no living affected members in the immediate family to confirm the diagnosis. By inspection of the local genetic register, it was established that it was indeed HD segregating in the extended family. Genotyping of the prospective mother and father, the father's unaffected father, and his unaffected maternal grandmother was carried out using a battery of polymorphic DNA markers, including a new probe which has a very low recombination rate with the HD locus. Analysis of DNA from a chorionic villus sample taken at 10 weeks of pregnancy showed that the fetus must have inherited a chromosome from its father's affected mother. Its risk of developing HD was 47%. If the genotype of the unaffected maternal grandmother was taken into account, the risk was reduced to 42%. Neither risk was considered acceptable by the prospective parents and the pregnancy was terminated at 12 weeks' gestation. Prospects for future pregnancies are good, with a 50% chance of having a child whose risk of inheriting the HD gene is less than 1.5%. In retrospect it was noted that although genotyping of the maternal grandmother had refined the fetal risk, it had also nearly contributed to an inadvertent and unwanted predictive test for HD on the father. This case makes the point that in prenatal exclusion testing, linkage information must be generated with considerable care.