When smoke gets in your sperm

Males are conceived at a higher frequency than females (ratio 1.53), but are more frequently lost through spontaneous abortions and stillbirths. Thus, the usual sex ratio at birth is reported to be about 1.06. However, in the past thirty years or so, there has been a notable reduction in the ratio of male to female births in several developed countries. While the reasons for this are not clear, there is speculation that exposures that affect males or the male reproductive system could lead to a decrease in this ratio. In a study by Fukuda et al [Lancet 2002; 359:1407-1408] of 11,815 liveborn singleton births (from 5372 mothers), maternal and paternal smoking habits were assessed by asking the mother about her and her spouse’s smoking habits during the periconceptional period (from 3 months before the last menstrual period to when the conception was confirmed). Smoking was categorized as non-smoking, < 19 cigarettes per day, and ? 20 cigarettes per day. The authors found that the sex ratio (M:F) of births declined with increasing cigarette smoking by both the mother and the father. The ratio for infants whose parents were non-smokers was 1.21, whereas the lowest ratio (0.68) was seen in the group in which both mother and father smoked ? 20 cigarettes per day. While intermediate ratios were found for the other groups, there were very few children (n=98) whose mothers smoked but father did not. Thus, the statistically significant findings were confined to the groups where the father smoked. There were no differences in either maternal or paternal age among the different smoking categories. The authors suggest that parental periconceptional cigarette smoking reduces the frequency of conceiving male children.

COMMENT: This is a fascinating use of basic data that explores an important question—whether environmental factors can disrupt the usual male:female birth ratio that nature intended. The results of this study, however, raise additional questions: Why are males so vulnerable to these environmental insults? and during what time period are they affected? Results of this study may also be pertinent to our identification of risk factors for infant leukemia. While parental cigarette smoking has not been identified as an important risk factor for infant leukemia, it is interesting to note that the sex ratio for infant leukemia is approximately 1.5:1 (F:M). Thus, either males are not surviving to manifest their leukemia, or females are somehow at an increased risk. Further research in the area of reproductive toxicology is needed. 

Antisocial Dads--  A possible benefit

There is intense interest in the identification of a possible infectious etiology in the risk of childhood leukemia. One hypothesis, described by Professor Leo Kinlen of the United Kingdom, involves population mixing, since a significant excess of childhood leukemia has been noted in some rural areas affected by a major influx of ‘outside’ individuals. Kinlen et al [Br J Cancer 2002; 86:732-7] recently examined whether children of fathers who have a high level of work contacts have a higher risk of leukemia than children of fathers who have low levels of work contacts, particularly in rural areas that are generally isolated. Using a national case-control study from Sweden, 1935 cases diagnosed with leukemia during the period 1958-1998 (representing 84% of all leukemias) together with 7736 age-matched controls were linked with paternal occupations as recorded in the Census. Paternal occupations were classified using a two-digit Nordic Classification of Occupations and included very high contact (fathers who are teachers, as they come into contact with children who experience the highest level of new infections), high contact (service and professional workers such as dentists, doctors, nurses, journalists, students, sales workers, etc who come into contact with many different people), and low/medium/indeterminate contact (all occupations not specified above). Unemployed individuals were excluded from analyses. Rural/urban status was determined by assembling counties of Sweden into three groups of decreasing rurality as follows: most rural (density 4 per square mile), intermediate (32/square mile), and most dense (189/square mile). For the most rural counties, there was an increased risk of childhood leukemia diagnosed < 5 years of age (OR=3.47, p for trend=0.02) for fathers with the highest level of work contacts compared to the intermediate (OR=1.18), and medium/low level (OR=1.00, p trend=0.02). No such association was observed for children diagnosed at older ages, nor was there as apparent an association with parental contact levels in urban or intermediate counties. These data suggest that population mixing (via the father’s level of work contacts) in a rural area may be a risk factor for childhood leukemia.

COMMENT: This is an intriguing concept, and Dr. Kinlen is to be commended for identifying ways in which to test his population mixing hypothesis using epidemiological data. As we have noted previously [see C3 Vol 12, No 3; Vol 11, No 2; Vol 10, No 6; Vol 8, No 3 ], it is extremely difficult to pinpoint a single infectious agent in the etiology of childhood leukemia. We intend to examine the question of paternal occupation and rural status by using epidemiological data collected from past studies of childhood leukemia in the Children’s Cancer Group. We would also encourage other groups who have similar data to examine this question. 

Down Syndrome- where are the other malignancies?

As has been noted in previous studies [Satge et al, Am J Hum Genetics 1998; 78:207-216; Hasle et al, Lancet 2000; 355:165-69], there appears to be a conspicuous lack of solid tumors diagnosed in individuals with Down syndrome (DS), although there is a substantial excess of leukemia, particularly in the first 10 years of life. In a new study by Yang et al [Lancet 2002; 359:1019-25], 17,897 United States death certificates of people reported to have DS were evaluated for the period of 1983-1997. Median age at death was calculated by sex and racial group. Standardized mortality odds ratios (SMORs) were calculated for 17 different conditions using ICD-9 codes and included malignancies, congenital heart defects, diabetes, infectious illnesses and the odds of having DS was compared among cases and controls (a random sample of 25% of the deaths in the United States).  For all individuals with DS, the median age at death increased by 24 years (25 years in 1983, 49 years in 1997); there were no differences between men and women. However, there were highly significant differences among whites, blacks, and people of other races, with white individuals having a substantially longer life span. Death certificates for people with DS were significantly more likely to list congenital heart defects, leukemia, dementia, hypothyroidism, seizure disorder, aspiration, pneumonia, influenza, and viral hepatitis than people without DS. These associations varied with age, with congenital heart disease deaths more often occurring in the third decade of life.  Children with DS younger than 10 years of age, were over three times more likely to have leukemia mentioned on the death certificate than children without DS. This association decreased with age and disappeared after the age of 40 years. In contrast, other malignancies were much less often listed on the death certificates of people with DS (OR=0.07, 95% CI=0.06-0.08). This was true for all ages, races, and both sexes. The only exception was cancer of the testis, where there was an elevated risk noted for individuals with DS less than 60 years of age. The authors emphasize that a) reduced exposure to environmental factors; b) tumor suppressor genes on chromosome 21; or c) a slower replication rate/higher apoptosis rate could contribute to the striking lack of malignancy noted in individuals with DS.

COMMENT: While several members of the Children’s Oncology Group are working to identify why children with Down syndrome are at such an increased risk of leukemia, it seems equally important for investigators of adult malignancies to identify why there is such a deficit of most malignancies (in particular, breast). 

Heel stick spots: back to the future

We have previously reported studies showing the presence of leukemic clones in neonatal heel stick cards [See C3 vol 9, No 1]. Additional data come from Taub et al, [Blood, 2002; 99, 2992] who studied 17 randomly selected patients with B-precursor ALL. Cases had a median age of 46 months (range 18 months to 13 years). A clonal rearrangement of the immunoglobulin heavy chain gene was identified in diagnostic lymphoblasts and sequenced. Patient specific primers were used to amplify DNA from blood samples on the patient’s newborn screening cards. Twelve of the 17 (71%) analyzed newborn cards had detectable IgH rearrangements amplified by semi-nested PCR. DNA sequencing confirmed that the IgH rearrangements matched the IgH sequences identified from diagnostic leukemia cells. This study nicely complements the earlier studies that used leukemia specific translocations eg t(4:11) to identify leukemia cells and expands the observation to B-precursor ALL without specific translocations.

COMMENT: These data augment the evidence that many cases of childhood ALL are initiated in utero. Many fascinating questions remain unanswered in this area. Most pressingly, what is the frequency of translocations in children who do not develop leukemia? An abstract from Mel Greaves’ group [Blood 96, 2000, 88a] examined 450 normal cord blood sample and found translocations in 5 (4 TEL-AML1 and 1 AML-ETO), suggesting that this is a frequent event. It is then intriguing to ask how the children who do not develop leukemia differ from those who do. Additionally, how do the children with in utero initiated leukemia differ from those with later initiating events. 

C3 Quarterly Newsletter
Children's Cancer Research Fund
Epidemiology Research Unit
Division of Pediatric Epidemiology
Clinical Research
University of Minnesota
420 Delaware St. SE, Box 422
Minneapolis, MN 55455
pedsepi@umn.edu

Editors: 
Stella M. Davies, MD, PhD, and Julie A. Ross, PhD