ably from the area cerebrovasculosa which is open floor of the fetal cranium.
CHROMOSOME 1 IN CERVICAL CARCINOMA
SIR,—Little is known about specific chromosome changes in forms of cancer. Chromosome no. 1 is the one altered in carcinoma of the ovary,’ bladder2 and breast.3 We have now found abnormal or extra no. 1 chromosomes in each of 14 consecutive near-diploid carcinomas of the uterine cervix (see table). 5 tumours had trisomy for no. 1 but banded preparations were too poor to reveal structural
Pædiatric Research Unit,
London SE1 9RT
ABO BLOOD-GROUP DISTRIBUTION AND HEART-DISEASE
SIR,—Mitchell1 has reported a strong positive association between ischsmic heart-disease (I.H.D.) mortality in England, Wales, and Scotland and the prevalence of blood-group 0. In a partial correlation analysis, he showed that when allowance is made for the effect of water hardness the association between i.H.D. mortality and % group 0 is little diminished and that allowance for frequency of blood-group 0 had little effect on the association between water hardness and I.H.D.
CHROMOSOMES IN CERVICAL CARCINOMA
mortality. We have studied2 the relationship between I.H.D. mortality and water calcium, twelve other water trace-elements, mean annual rainfall, mean annual temperature, and a socioeconomic index derived by West and Lowe.3 There was a strong negative association between I.H.D. mortality and water calcium which could not be explained in terms of the other variables. As the 61 towns in our study were among the 81 used by Mitchell, we have reanalysed our data with % bloodgroup 0 as an additional variable. Regression analysis shows that % group 0 partially explains the variation in I.H.D. mortality but does not remove the effect of calcium. Furthermore the contribution of blood-group to the regression was less than that of calcium, and was more influenced by other independent variables. This suggests that while % blood-group 0 may make an independent contribution to the differences in I.H.D. mortality, it is confounded with the other variables, and to a greater degree than is calcium. Our results and conclusions are therefore broadly in accord with those of Mitchell. However, the interrelationships between the independent variables make studies of this kind difficult to interpret. Whilst acknowledging that Mitchell has discovered an interesting association, we must agree with Allan4that further studies on other European populations are desirable. P. C. ELWOOD M.R.C. Epidemiology Unit, A. S. ST. LEGER Cardiff CF2 3AS Tenovus Research Institute, Cardiff
BLOOD-GROUP A AND GIARDIASIS
SIR,—Barnes and KayS found an association between Giardia lamblia infestation and blood-group A, and suggested that similarity between parasite and blood-group antigens prevented immunological recognition of the parasite. However, the association might also be explained by a weakening of the normal gut defence mechanisms in these patients. Acids in gastric secretions are important in this defence and I have found achlorhydria in a large proportion of the patients with Salmonella, Shigella, and Giardia infections, even in the convalescent stage. Achlorhydria has a higher than normal frequency in people with blood-group A. Such people are thus likely to be at risk from enteric infections such as G. lamblia. Clinic of Infectious Diseases. Boden Central Hospital, 96100 Boden, Sweden
1. Mitchell, J. R. A. Lancet, 1977, i, 295. 2. Elwood, P. C., St. Leger, A. S., Morton, M. Br. J. prev. 178. 3. West, R. R., Lowe, C. R. Int. J. Epidem. 1976, 5, 195. 4. Allan, T. M. Lancet, 1977, i, 541. 5. Barnes, G. L., Kay, R. ibid. p. 808.
trisomy for chromosome
1 but gave
preparations. abnormalities. Of the remainder (see table), isochromosomes were found in three tumours, 2 for the long arm and 1 for the short arm. In tumours with high chromosome numbers, structurally changed, and/or a relative excess of normal no. 1 chromosomes, were frequently present. In an anaplastic malignant small round-cell tumour involving the vault of the vagina and base of the bladder in a patient aged 68, the modal chromosome number was 47 and the only abnormality in G-banded preparations was an extra no.l-derived chromosome in which the short arm was replaced by most of a long arm of the same chromosome. It thus had an almost completely duplicated long arm and resembled an isochromosome i(lq) except that the near-centromeric heterochromatic band (q 12) was not duplicated. Abnormalities of chromosome no. 1 have been found in several tumours but not in chronic leukaemia and myeloproliferative disorders, suggesting that they are associated with the later stages of malignant transformation.4 We have, however, found a lp- chromosome in a cervical carcinoma. There may be a relationship between heteromorphism for the heterochromatic region of chromosome no. 1,4 pericentric inversions of this regionsand an increased risk of cancer.1.s Department of Cancer Research, Mount Vernon Hospital, Northwood, Middlesex HA6 2RN
N. B. ATKIN MARION C. BAKER
RESEARCH ON ALIMENTARY CANCER accrues from detailed studies of rarities than from the commonplace. You note (Sept. 24, p. 645) the expense and doubt the cost effectiveness of epidemiological studies in alimentary cancer. More then may be learnt from detailed studies of the few individuals who acquire alimentary
1. Atkin, N. B., Pickthall, V. J. Hum. Genet. 1977, 38, 25. 2. Atkin, N. B., Baker, M. C. Cytobios (in the press). 3. Cruciger, Q. V. J., Pathak, S., Cailleau, R. Cytogenet. Cell Genet. 1976, 17, 231. 4. Atkin, N. B. Br. med. J. 1977, i, 358. 5. Atkin, N. B., Baker, M. C. Cytogenet. Cell Genet. (in the press).