Reply to Duncan: Fragile X leads to X chromosome loss: Comments from a cytogeneticist

The comments of Alessandra Duncan are based on pre-1991 cytogenetic literature that she mistakenly interprets as implying that the fragile X full mutation is not associated with mosaic Turner syndrome. It is not surprising that older cytogenetic literature is not ‘‘congruent’’ with our observations. The data in Rosenberg et al. [1991] that Alessandra Duncan refers to were from analyses in more than a dozen publications that focused on fragile site expression in females who were fragile X ‘‘obligate carriers.’’ These cytogenetic studies predate both the identification of the FMR1 gene in 1991 and, more importantly, they predate the distinction between premutation and full mutation females that analysis of the fragile X triplet repeat mutation made possible. Examination of the articles summarized in Rosenberg et al. reveals that approximately 30% of the ‘‘obligate carriers’’ were listed as mentally impaired. The other ‘‘obligate carriers’’ among the 119 females in Rosenberg et al. were diagnosed by cytogenetic expression of the fragile site. Diagnosis by fragile site expression has been shown to have substantially inflated fragile X prevalence estimations [Turner et al., 1996; Morton et al., 1997] and, although at the time it was the accepted diagnostic, comparisons with DNA analyses have since shown that it was not reliable, especially in females. It is very likely that a significant proportion of the 119 ‘‘obligate carriers’’ in Rosenberg et al. did not carry the full mutation but were premutation carriers or carried normal FMR1 alleles. Contrary to the assumption made by Dr. Duncan, cytogenetic abnormalities, such as cells with only one X chromosome, may not have been reported in the publications summarized in Rosenberg et al. These cytogenetic analyses were focused on expression of the fragile site—observation of other abnormalities was not a goal of the studies. In addition the cytogenetic analyses of the 119 females were done in more than a dozen different laboratories, each laboratory with its own criteria for reporting abnormal karyotypes. Of the 13 publications summarized in Rosenberg et al., only Jacobs et al. [1980] and Carpenter et al. [1982] mentioned the presence of abnormal karyotypes and only the former tabulated them. Another of the summarized publications, Howell and McDermott [1982] explicitly excluded metaphases that did not contain two recognizable X chromosomes. No mention of abnormal cytogenetic findings was made in the other 10 articles from which the data in Rosenberg et al. were drawn. Dr. Duncan also mentions the large number of cells analyzed in these studies but it is not clear how this is relevant. Clinical cytogenetic analysis usually involves 20–150 cells at most, and that is how mosaic Turner syndrome is detected. The association of the fragile X full mutation and mosaic Turner syndrome are based on clinical prenatal studies with a fragile X diagnostic reliability substantially greater than 99% and on completely independent clinical cytogenetic analyses that note all abnormalities. In our published data set there were five cases of mosaic Turner syndrome in 105 full mutation fetuses and none in the 318 normal, intermediate and premutation female fetuses. The probability that this occurred by chance was conservatively estimated (Fishers exact test) to be less than 1 in 1,000. Currently our data for females show six reports of mosaic Turner syndrome in 111 full mutation fetuses and none in 392 normal, intermediate and premutation fetuses analyzed. The odds that this association occurs merely by chance are approximately 1 in 10,000. In conclusion it seems likely that absence of mosaic Turner syndrome in fragile X ‘‘obligate carriers’’ that Dr. Duncan noted is due to the limitations of the cytogenetic data she examined.