Childhood Cancer and Neural Tube Defects

Abstract

To the Editor: In the March 1997 issue of the Journal,Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar concluded that cases of spina bifida and of abnormalities of the eye, ribs, and spine were more common in children with cancer than among population-based controls. In an invited editorial in the same issue, Friedman, 1997Friedman JM Genetics and epidemiology, congenital anomalies and cancer.Am J Hum Genet. 1997; 60: 469-473PubMed Google Scholar called for additional work to define more clearly the associations between congenital anomalies and cancer and to differentiate those associations that are spurious from those that are biologically important. I wish to comment on the reported association between spina bifida and cancer, which was based on comparison of the relative frequency of specific types of anomalies in children with cancer who were diagnosed in Great Britain during the period 1971–86 with the corresponding relative frequency in liveborn children born in British Columbia during the period 1969–88. In addition to the differences in methods of ascertainment discussed by Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar, this comparison appears to be inappropriate, since the prevalence of neural tube defects at birth was markedly higher in the British Isles than in Canada, from the 1950s to the early 1980s (Little and Elwood, 1992Little J Elwood JM Geographical variation.in: Elwood JM Little J Elwood JH Epidemiology and control of neural tube defects. Oxford University Press, Oxford1992: 96-145Google Scholar). This difference was apparent despite the decline in the prevalence of these defects at birth in the British Isles since the early 1970s. In Canada, there has been a persistent east-west gradient, with the prevalence at birth declining from the east toward the west. In the study by Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar, when the 275 cases with an established genetic cause were excluded, the prevalence of neural tube defects at birth among children diagnosed with childhood cancer was 1.2 per 1,000 births. As noted by Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar, anencephalus is associated with a high rate of infant mortality, and most children with neural tube defects are likely to have had spina bifida. In studies of births in Great Britain prior to 1979, the prevalence of spina bifida at birth was within the range of 1.5–4.1 per 1,000 births, and, in studies in the early 1980s, the prevalence was within the range of 0.7–1.9 per 1,000 births (Little and Elwood, 1992Little J Elwood JM Geographical variation.in: Elwood JM Little J Elwood JH Epidemiology and control of neural tube defects. Oxford University Press, Oxford1992: 96-145Google Scholar). Some of these infants would have been stillborn. For example, in Northern Ireland during the period 1974–79, 15.5% of 569 cases of spina bifida or encephalacoele were stillborn (Little and Nevin, 1989Little J Nevin NC Congenital anomalies in twins in Northern Ireland. II. Neural tube defects, 1974–1979.Acta Genet Med Gemellol (Roma). 1989; 38: 17-25PubMed Google Scholar). In Glasgow and Liverpool during the period 1980–92, when fetuses from terminated pregnancies were excluded, 16% of 262 cases of spina bifida were recorded to have resulted in fetal deaths (EUROCAT Working Group, 1995EUROCAT Working Group Surveillance of congenital anomalies of Europe 1980–1992. Institute of Hygiene and Epidemiology, Brussels1995Google Scholar). Therefore, it appears that the proportion of cases of childhood cancer with neural tube defects is similar to what would be expected on the basis of data on the prevalence of these defects at birth, in Great Britain. In the study by Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar, eight of the children with tumors of the brain or of the spinal cord were recorded as having spina bifida, compared with the 5.6 expected on the basis of the frequency of spina bifida among children with other types of cancer in Great Britain and with the 2.4 expected on the basis of the data for British Columbia. Again, the proportion of children with tumors of the brain or of the spinal cord who were recorded as having spina bifida (1.7 per 1,000 births) would appear to be within the range of prevalences at birth reported for Great Britain during the period in which the children included in the study by Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar would have been born. Thus, the study by Narod et al., 1997Narod SA Hawkins MM Robertson CM Stiller CA Congenital anomalies and childhood cancer in Great Britain.Am J Hum Genet. 1997; 60: 474-485PubMed Google Scholar does not appear to support the hypothesis of a common maternal factor for brain tumors and spina bifida.