We thank Dr Spencer for bringing to our attention the published first-trimester maternal α-fetoprotein (AFP) results for additional open spina bifida cases. Of the 9 cases tested at 11-13 weeks’ gestation, the AFP levels (7 derived from Figure 4 in the report by Aitken et al ) were 0.2, 0.6, 0.9, 1.0, 1.1, 1.1, 1.2, 1.4, and 1.6 multiple of the median (MoM). Combined with our own series and that of Bredaki et al, the weighted geometric mean becomes 1.28 MoM.

Modeling with this value and the SDs in our series, the predicted proportion of spina bifida pregnancies with levels above the normal 97th percentile would be 14%. This is consistent with the UK Collaborative AFP Study, in which the observed total spina bifida detection rate at 10-12 weeks’ gestation using a 97th percentile cutoff level was 21% (6 of 29).

First-trimester AFP is indeed a relatively weak marker of open spina bifida. However, if it were to be measured at 11-13 weeks for other reasons, such as enhanced Down syndrome screening, modeling predicts a worthwhile increase in spina bifida detection compared with screening by biparietal diameter alone.

It remains questionable whether maternal free β-human chorionic gonadotropin (hCG), which is already part of standard Down syndrome screening protocols, would lead to a further increase in spina bifida detection. We thank Dr Spencer for pointing out that Aitken et al reported normal first- and second-trimester intact hCG levels in open spina bifida, but this is not inconsistent with our finding of reduced first trimester free β-hCG levels. In Down syndrome, the distribution of these 2 isoforms of hCG is markedly different in the first trimester and they could occur in spina bifida.