In this chapter we review what is known about the effects of anti-cancer treatments and of certain other therapeutic and environmental agents that could conceivably have an injurious effect on chromosomal distribution at gametogenesis or that might cause chromosomal breakage or rearrangement in the cells of the gonad. In other words, the focus is on factors that might disturb the course of meiosis or have clastogenic effects upon the chromosomes of gametocytes. We do not consider other categories of genetic damage.
Given the inherent vulnerability of gametogenesis, a logical starting position might be that any potential damaging agent should be presumed guilty until proven innocent. As discussed in Chapter 21, large fractions of sperm and eggs, in the vicinities of 10%–20%, are chromosomally abnormal, due to aneuploidies or structural change acquired for the most part during meiosis. If this is what happens naturally, if gametogenesis is so susceptible normally, then surely, would not agents known to compromise the integrity of the DNA and of the spindle apparatus (not to mention various artificial dietary and environmental exposures) compound the effect dramatically? Perhaps surprisingly, this seems not to be the case. Gametogenesis, provided the damage is not irreversible, often proceeds normally, or at any rate recovers, even in the setting of some heavy exposures, and no discernible increase in chromosomal abnormality is recorded in the subsequently born children. Nevertheless, if only on the pure grounds of what seems biologically reasonable and plausible, the question should not be regarded as being closed. The fact that sperm chromosomes may, with certain agents, show an increased rate of cytogenetic abnormality is a more practical grounds for maintaining a cautious view.
BIOLOGY AND EPIDEMIOLOGY
A majority of children and young adults who receive modern cancer treatment survive. Some treatments cause sterility, but in quite a number fertility is unscathed or, at any rate, subsequently recovers. For those who are potentially capable of having children, the question arises: could there be an increased risk to have a child with a chromosomal abnormality? For most, in fact, the short answer may be apparently not. Longer answers follow.
The chemotherapeutic agents and radiation used to rid the body of cancer are essentially cellular toxins, some of which specifically target DNA or the mitotic apparatus. Thus the starting hypothesis is that the chromosomes in exposed bystander tissues, in particular of the gonad, could be vulnerable. The fact that these treatments can damage chromosomes is well known, and this is actually the basis of one of the in vitro laboratory tests for ataxia-telangiectasia.1 Rapidly dividing cells are the most vulnerable to anti-cancer treatments (this being, of course, the rationale for their use), and this would suggest, in theory, a susceptibility for spermatogenesis in the postpubertal male (millions of cell divisions daily), and a relative resistance in the prepubertal male (division yet to commence) and in oogenesis (cell division in suspension).
In practice, it is to the experience of the “therapeutic experiments” of oncological medicine that we mostly appeal: the in vivo observations of those who have survived their cancer, recovered from their treatment, and who have gone on to have or to attempt to have children. Have the children shown any excess of cytogenetic abnormality? The best data addressing the question of the effects on offspring are those from the Five Center2 Study (Byrne et al., 1998). A total of 1062 survivors of several types of childhood cancer (Hodgkin's disease, soft tissue sarcoma, and thyroid and central nervous system tumors accounting for the majority) who were born in the period 1945–74 were interviewed in 1980–83. They had had a total of 2198 children, and the findings have been reasonably reassuring. The rate of birth defects did not differ significantly from those of their sibling controls (3.4% vs. 3.1%). On the specific issue of cytogenetic defects, the fractions for the children of survivors and sibling controls were 4/2198 (0.2%) and 6/4544 (0.1%), respectively, the difference not being statistically significant. With particular reference to Down syndrome, the rates were 3/2198 (0.1%) and 4/4544 (0.1%), respectively.
The direct assessment of gametic chromosomes offers insight. For obvious reasons, the male gamete is more accessible for study. Sperm chromosome analysis can be done in men who have survived cancer treatment. Table 26-1 provides a review of 12 such studies, and shows that several therapeutic regimens can cause sperm karyotypic defects.3 Why, then, is no increase in chromosomally abnormal offspring to be seen? One suggestion is that these abnormal sperm may be selected against at fertilization, or in the conceptus during early embryogenesis (Arnon et al., 2001). Gonosomal (X and Y) abnormalities might be less subject to this sort of selection, as Foresta et al. (2000) propose. They assessed 10 men with severe oligozoospermia who had been treated for cancer at least 5 years previously with combinations of high-dose radiotherapy and chemotherapy. The frequency of 24,XY, 24,XX, and 24,YY disomic sperm was significantly higher in patients treated for cancer (and also in a group of men with idiopathic oligozoospermia) than that in normal controls. These authors suggest that damaged testicular structure, whether due to cancer treatment or other factors, might of itself compromise sex chromosome distribution at meiosis. They offer the cautious advice that, if assisted conception with intracytoplasmic sperm injection (ICSI) is offered, this (presumably rather small) sex chromosome risk should be discussed. For women who had been treated as children, there is the practical question that cancer treatments may have lessened the ovarian reserve, with the prospect of an earlier menopause, and thus decisions about childbearing should not be too long delayed (Larsen et al., 2003).
Levy and Stillman (1991) and Arnon et al. (2001) review in detail the effects of various chemotherapeutic regimens on fertility; for some, data are also available concerning mutagenicity. The six classes of chemotherapeutic agents are the following: alkylating agents, cisplatin and its analogs, vinca alkaloids, antimetabolites, topoisomerase inhibitors, and “newer agents.” In broad terms, the relationship between type of drug and risk for gonadal damage is outlined in Table 26-2. Alkylating agents (cyclophosphamide and chlorambucil being major representatives), which have their damaging effect by adding an alkyl group to DNA, can cause testicular hypotrophy in the male, with oligospermia or azoospermia. With cyclophosphamide, the effect is dose related; pubertal boys may be more susceptible than prepubertal boys, although this point is arguable (Ben Arush et al., 2000). In follow-up into adulthood, the reproductive potential for males having been treated with an alkylating agent in childhood is considerably reduced, with a relative fertility of 0.4 (Byrne et al., 1987). Levy and Stillman review a number of papers, which offer a generally optimistic picture for girls in terms of pubertal development, but, as they point out, longer-term studies relating to the specific question of fertility are not so numerous. In one follow-up study women actually had a relative fertility of 1.0 (Byrne et al., 1987). In adult women who have had chemotherapy with cyclophosphamide for Hodg-kin's disease or breast cancer, oocyte depletion and ovarian failure are documented (Familiari et al., 1993; Goodwin et al., 1999).
Table 26.1. Findings from 12 Sperm Chromosome Studies on Patients Treated According to Certain Anti-Cancer Regimens
Table 26.2. Three Categories of Risk for Germ Cell Depletion, Based on the Drugs Used
Antimetabolites, alkaloids, and antibiotics (including methotrexate, vincristine, and actinomycin D) seem not to cause compromise of ovarian function when given alone or as combination therapy, but in conjunction with radiotherapy some agents will cause ovarian failure.
Topoisomerase inhibitors affect the integrity of the mechanical apparatus of the meiotic chromosome, including the centromere and the microtubules of the spindle, and also act directly upon the DNA. Mouse studies with etoposide show an actual increase in sperm and zygote aneuploidies (De Mas et al., 2001; Marchetti et al., 2001).
Multiagent chemotherapy is, as would be expected, more damaging, based on the observations of ovarian histology postmortem in a group of girls who had succumbed to their cancer (Nicosia et al., 1985).
Fertility is diminished in females who have had radiation therapy to the abdomen, and there is an increased risk of obstetric complication. But their children appear to have no increased incidence of birth defects (Nicholson and Byrne, 1993). Martin et al. (1986c) studied 13 male cancer patients (mostly seminoma) at intervals up to 36 months after radiotherapy, in whom the doses of testicular radiation were estimated to be in the range 0.4–5.0 Gray. While most were azoospermic in the first year following treatment, in those in whom spermatogenesis recovered, variable increases in sperm chromosome abnormalities were seen, averaging twofold overall compared with controls, but with wide ranges. The frequencies correlated with the estimated “bystander” testicular radiation.
Radioiodine is used in thyroid cancer, and in a review of 408 offspring of survivors in Ehrenheim et al. (1997), no increase in congenital malformations was noted, although in one case a 7/14 translocation, not further described (and not identified as de novo or familial), was detected at prenatal diagnosis. One case does not make a case, but it can be noted.
Treatment for Hodgkin's disease typically involves radiation to the chest and abdomen, and multidrug chemotherapy (e.g., MOPP, NOVP; see explanation of abbreviations in footnote of Table 26-1). Spermatogenesis is compromised and may or may not recover, although ovarian function typically is resistant or, if affected, more readily returns to normal (Marmor and Duyck, 1995; Papadakis et al., 1999). Brandriff et al. (1994) studied six male patients who had had from two to six cycles of MOPP treatment, with or without radiotherapy, from 3 to 20 years previously. Using the sperm human-hamster (humster) model, 9.8% of sperm had structural aberrations and 1.6% were hyperhaploid, compared with figures of 6.9% and 0.8%, respectively, from a control group. Robbins et al. (1997a) used multicolor FISH to study eight patients who were treated with NOVP, and observed a fivefold increase in sperm with abnormal karyotypes involving the chromosomes tested (chromosomes X, Y, and 8). Some subjects had increased aneuploidy rates before starting treatment, pointing to an effect upon spermatogenesis of the disease state itself.
Damage to sperm may not necessarily be permanent, and in some men the chromosomal defects decline to pretreatment levels after a matter of only months or even weeks (Robbins et al., 1997a; De Palma et al., 2000). Monteil et al. (1997) studied a single case with FISH sperm analysis, in a man treated with radiotherapy and chemotherapy. This man had multiple structural abnormalities in most spermatozoa immediately following the radiotherapy, but none 5 weeks later, whereas multicolor FISH showed disomy for chromosomes 1, 6, 11, X, and Y on both occasions. The structural defects may have been due to the effects of radiotherapy, and the aneuploidies, due to chemotherapy.
Actual observed reproductive outcomes from previously treated patients are the proof positive. Aisner et al. (1993) interviewed 35 women and 25 men who had had 68 living children. There was no increase in spontaneous abortions or congenital abnormalities. Six women in the study of Papadakis et al. (1999) had had eight normal children. Similar findings were obtained by Swerdlow et al. (1996), from 11 men and 16 women who had had a total of 49 children. Chromosome analyses were done in 45 of these children, and all were normal, except for one child with trisomy 21, the additional chromosome having been transmitted from the other parent. These authors commented that “offspring of patients treated in adulthood for Hodgkin's disease are not at greatly raised risk of genotoxic or other adverse outcomes as a consequence of their parent's treatment.” This provides a quite reassuring counterpoint to the concern raised by the in vitro sperm studies noted above. Some patients might nevertheless choose to have prenatal diagnosis.
Childhood Acute Lymphoblastic Leukemia
About half of males with acute lymphoblastic leukemia (ALL) will suffer sterility, following the typical regimen of alkylating agents. López Andreu et al. (2000)undertook sperm analyses in a group of 22 childhood leukemia survivors when they were teenagers or young adults. Five were azoospermic or severely oligoasthenozoospermic. The ovary, on the other hand, is more resistant to a permanent effect of alkylating agents. In those in whom fertility recovers, there has been no indication of any increase in the rates of congenital malformation in the offspring (Levy and Stillman, 1991). In a study based on 140 children born to ALL survivors, Kenney et al. (1996) recorded one case of trisomy 13, but overall there was no significant difference in malformation rates between the children of survivors (3.6%) and those of their sibling controls (3.5%).
In a review of Wilms tumor survivors, Byrne et al. (1998) documented a particular risk for the female to have children with adverse outcomes, but this effect is apparently due to a damaging influence of radiation on the uterus with a secondary deforming effect on the fetus, or possibly reflecting a transmitted genetic defect that had been responsible for the mother's original tumor as a child. No malformations were recorded in the offspring of 19 children of male survivors, this being, admittedly, a small number.
This is typically a cancer of younger men. The treatment is usually surgical removal of the affected testis, and chemotherapy with such agents as cisplatin, etoposide, and bleomycin. Fertility is maintained in some, although persisting oligospermia is a frequent observation, especially in those having had high-dose cisplatin (Stephenson et al., 1995). Martin and colleagues (1997, 1998, 1999) studied four men variously before, during, and some 2–13 years after treatment with the three drugs noted. On sperm chromosome analysis, using certain chromosomes as surrogates (nos. 1, 12, X, and Y), there was an increase in 24,XY disomy from pretreatment (0.14%) to immediately posttreatment (0.34%). These men had had a total of two children previously, and, after successful treatment, six more children, all normal. Another eight normal children are recorded in the series of Stephenson et al. (1995), three of the fathers being oligospermic. A risk to produce aneuploid sperm may diminish with the passage of time after the chemotherapy. De Mas et al. (2001) document sperm disomies for chromosomes 16, 18, and XY, using five-color FISH, in the period of 6–18 months after treatment; while Martin et al.'s (1999) work shows that, some years thereafter, the sperm karyotypic indices return to normal. These observations are consonant with what is seen in etoposide exposure in a mouse model (Marchetti et al., 2001). While it appears that the risk for chromosome abnormality in a child is small, it may nevertheless be reasonable to offer prenatal diagnosis for fathers-to-be.
No cytogenetic abnormalities were seen in the sperm of a 26-year-old man treated with cisplatin, vinblastine, and bleomycin, analyzed some 9 months after treatment (Jenderny et al., 1992).
This chromosomal condition is discussed in detail in Chapter 21. There is no increased risk for other abnormal outcome in a subsequent pregnancy, and in particular the incidence of congenital malformations is no greater (Berkowitz et al., 1994). Chemotherapy, either at the time of evacuation of complete mole or for “persistent gestational trophoblastic tumor” following the index pregnancy, seems also to be without untoward effect in a subsequent pregnancy.
Infertility Associated with Cancer Therapy, and Prior Gamete Banking
Banking of gametes prior to treatment for cancer is a logical management, and successful accounts have been reported (Horne et al., 2001). Hallak et al. (2000b) speak of sperm banking as “fertility insurance” for men with cancer. Patients with carcinoma, but less so in sarcoma, may have poor semen indices ahead of any cancer treatment, and so they might in any event have required ICSI (Hallak et al., 2000b). Blackhall et al. (2002) report 33 couples over a study period 1978–1990, the male being a survivor of Hodgkin's disease, using cryopreserved sperm for artificial insemination or IVF, with 9 of the 33 couples having 10 pregnancies. One pregnancy was terminated because of fetal hydrocephalus (46,XX karyotype), the others all producing normal infants. For prepubertal boys, surgical removal of testicular tissue and the separation of spermatogonia is a theoretical route yet to be developed (Hovatta, 2001; Shinohara et al., 2002). Cryopreservation of ovarian tissue is more advanced as a procedure, but still in a research stage (Gook et al., 2001; Oktay, 2001; Boiso et al., 2002; Poirot et al., 2002). Any risk for chromosomal abnormalities in offspring due to these procedures is unknown, and data will need to be collected (it would be rather hypothetical to think that the risk in women might actually be reduced, if they were to reanimate their oocytes from a younger age).
Azathiaprine is a major immunosuppresive agent, and its inimical effect on the ovary is known (McDermott and Powell, 1996). Reports of any possible effect on offspring are few. One example to be noted is of a dysmorphic child with two de novo abnormalities, an interstitial deletion and an apparently balanced translocation, 46,XY,t(6;14)(q21;q12), del(7)(q21), whose mother had been treated with azathioprine and prednisone for systemic lupus erythematosis, although not much weight can be put on a single case report (Ostrer et al., 1984). Jenderny et al. (1992) studied a 36-year-old man who had been treated for 4 years with these same two drugs for chronic active hepatitis, and showed no significant differences in sperm chromosome distributions from controls. Alkylating agents are used in the treatment of nephrotic syndrome. In women who were treated in childhood with cyclophosphamide, fertility may be little compromised, with 17 out of 18 girls in one series going on to have normal menstruation and fertility (Watson et al., 1986). Pubertal boys, and those receiving higher doses of cyclophosphamide, are likely to become azoospermic, whereas prepubertal boys appear to have greater gonadal resistance. The use of chlorambucil for this condition is also associated with a high likelihood of azoospermia (Levy and Stillman, 1991).
Other Pharmaceutical Agents
Griseofulvin is a fungicide quite commonly used for the treatment of tinea. It is known to interfere with formation of microtubules, and can damage the mitotic spindle. Mouse studies (although at very high dosages) suggest that meiosis may also be vulnerable, more so in oogenesis than in spermatogenesis (Shi et al., 1999). Single studies of mouse model sperm analyses suggest an increase in aneuploidies with exposure to pyrimethamine, an antimalarial agent, and miconazole, used in the treatment of candidiasis (Aydemir and Bilaloglu, 1996; Hassan, 1997). Whether the doses of the above agents typically used in humans might have practical reproductive implications is quite unknown.
Diazepam is used in psychiatry, and Baumgartner et al. (2001) studied patients having been on chronic high-dose treatment. Sperm disomy rates, for the chromosomes analyzed (X, Y, 13), were approximately twofold those of controls (and sperm counts were reduced).
Most studies point to no effect of previous use of the oral contraceptive pill, in terms of a subsequent risk of having a child with Down syndrome (DS) (Källén, 1989). It is speculative whether there might be an additive effect from use of the pill together with cigarette smoking (Yang et al., 1999).
Views differ whether there exists an increased risk for women with diabetes mellitus to have a child with DS. Narchi and Kulaylat (1997) proposed a risk approximately twofold the age-related figure, but such an effect was not identified in the study of Martínez-Frías et al. (2002).
A small effect may possibly exist for DS with respect to previous X-rays to the abdomen and pelvic area; that is to say, for X-rays in which the gonads may have been within or not far off the field of the film. In a study of 156 mothers and 149 fathers in whose DS children the “nondisjunctional parent” could be identified (using Q-banding polymorphisms), a history of X-ray exposure was more often recorded in older fathers and in younger mothers (Strigini et al., 1990). The odds ratio for the whole group was 1.85, although the lower limit of the 95% confidence interval was 1.0. If such an effect truly exists in the younger mothers (and the statistics were borderline), it would seem that this slight influence becomes diluted as they get older, and the age effect comes to be predominant.
Humans may be relatively resistant to the mutagenic effects of radiation, compared with some animal models (Neel et al., 1990). The atomic bomb blasts at Hiroshima and Nagasaki in 1945 were not followed by a statistically significant difference in the rate of chromosome abnormalities in children subsequently conceived, in a study commenced in 1967 (Awa et al., 1987; Neel and Schull, 1991). The study population comprised 8322 individuals born from 1946–1972, age range at the time of study 12–38 years, one or both of whose parents were within 2000 meters of the hypocenter “ATB” (at the time of the bomb), alongside a contemporaneous local control group of 7976, who were either more than 2500 meters from the hypocenter or not present in the city. Sex chromosomal abnormalities were seen in 2.28 per 1000 in the exposed group and in 3.01 per 1000 of the controls. The only instance of autosomal trisomy was a 15-year-old with standard trisomy DS, whose father had been exposed at Hiroshima. (Given the structure of this study, deceased younger children and infants with autosomal trisomy were not included, although it is also to be noted that in separate analyses in Neel and Schull no significant correlation existed between parental exposure ATB and the frequency of stillbirth or congenital malformations.) More children of exposed parents had a small supernumerary abnormal chromosome than in the controls (5 cf. 2, a difference not specifically commented upon in Awa et al.). Of the balanced structural rearrangements, only two were confirmed as having arisen de novo (one each in the exposed and control groups).4 An earlier study with specific reference to clinically diagnosed DS in 9-month-old infants, undertaken during 1948–1954 (before the chromosomal basis of DS was known), had shown no increase among offspring of 5582 exposed cf. 9452 unexposed mothers, and indeed the figures were in the other direction (0.54 cf. 1.27 per 1000), and in spite of the exposedmothers being on average slightly older (Schull and Neel, 1962).
The Chernobyl nuclear plant explosion occurred in 1986, and a cloud of radioactivity was dispersed over Europe. With respect to DS, no subsequent rise in incidence was identified in a number of European countries, apart from a small cluster in Berlin (Little, 1993). In contrast, Bound et al. (1995) suggest a possible link between events in 1957 (a fire in a nuclear reactor) and the early 1960s (increased levels of fallout from nuclear testing) and peaks of DS prevalence in 1958 and 1963–64 in the Fylde district of Lancashire, England. But by no means is a firm case made: post hoc does not necessarily mean propter hoc,5 and some fluctuation is normal. The same 1957 nuclear reactor accident had been proposed as the possible reason for a cluster of six cases of DS among the children of women who had attended the same high school in Dundalk, Ireland, during 1956–57, when they would have been 12 to 19 years of age. Of the 387 births to the former pupils from this period, the expectation would have been 0.69 children with DS. However, a stringent review of the evidence, and including molecular analyses that showed one case to have arisen post-fertilization, led to the conclusion that, in fact, chance was the probable basis for the cluster (Dean et al., 2000).
Paternal occupation provides a surrogate marker for a variety of potential industrial agents. Olshan et al. (1989) assessed the father's occupation for a thousand DS children born in British Columbia in 1952 through 1973. Seven employment categories out of 59 showed odds ratios in the range 1.4–3.3, the lower confidence limit being not less than 0.9, in some of which exposure to various industrial agents could plausibly have occurred (including mechanics, janitors, metal workers, sawmillers). But the increases in risk were small, and with 59 items there was of course the possibility of chance fluctuation. One category that might have seemed risky, namely “other chemical workers,” in fact had the lowest odds ratio of all (0.2). Exposure to industrial air pollution has been proposed as a cause of the Prader-Willi deletion and as contributing to sperm aneuploidy (Ducatman, 1988; Perreault et al., 2000). Regardless of whether or not hydrocarbons really could damage testes and cause deletions, it is implausible that this would be a universal explanation for the Prader-Willi deletion. More likely, proximal 15q is simply a region of inherent instability, that will, from time to time, be the basis of deletions and other rearrangements (Nicholls, 1993). A study of professional pesticide sprayers in the Red River Valley area of Minnesota showed, in the preliminary analysis, no increase in sperm chromosome abnormalities in these men compared with controls, and the ongoing analysis is confirming this conclusion (Smith et al., 2000b; J. L. Smith, pers. comm., 2001). Likewise, a studyof Danish farmers exposed to fungicides in the course of their work showed no significant differences in sperm aneuploidy before and after exposure (Härkönen et al., 1999). These are questions for the reproductive toxicologist to address (Wyrobek and Adler, 1996).
Tobacco, Alcohol, Caffeine
Tobacco smoking in mothers had no influence on the incidence of DS in the study of Chen et al. (1999a), based on data of a population case–control analysis in Washington state from 1984 to 1994, and in which the authors had been at pains to account for a confounding effect of maternal age. The odds ratio was exactly 1.0—that is, no effect either way—for smokers versus nonsmokers. Similar findings are also reported from Sweden, California and England (Källén, 1997; Torfs and Christianson, 2000; Rudnicka et al., 2002). Nevertheless, a tentative role has been proposed for one particular mechanism: trisomy 21 due to nondisjunction in maternal meiosis II (MMII). In a case–control study in Atlanta, cigarette smoking around the time of conception gave an odds ratio of 7.6 in mothers of MMII trisomic offspring, compared with controls, in the <35-year age group (Yang et al., 1999). Very speculatively, smoking might diminish blood flow in the microvasculature of the perifollicular bed, and the resultant hypoxia could compromise some aspect of the oocyte's functioning as meiosis II gets underway. Alcohol and coffee intake by the mother prior to conceiving might actually reduce the DS risk. In the study of Torfs and Christianson, the odds ratios for “high” alcohol and coffee consumption (≥4 drinks per week, ≥4 cups per day) were 0.54 and 0.63, respectively. If these figures reflect biological reality, a possible mechanism would be a selective reduction in viability of a trisomic 21, as compared to a normal, conceptus.
Concerning spermatogenesis, in reviewing the literature, Shi and Martin (2000c) concluded that personal habits with respect to smoking, alcohol, and caffeine ingestion appear not to have any consistent effect on disomy rates in sperm. But since there were somewhat varying findings in the different studies, it had to be acknowledged that a definitive answer was not at hand. Shi et al. (2001a) proceeded to a study of cigarette smoking and aneuploidy using FISH analysis of sperm, with reference to chromosomes 13, 21, X, and Y. They divided their subjects into nonsmokers, light smokers (<20 cigarettes per day), and heavy smokers (≥20/day). The smokers showed an increase in disomic 13 sperm (0.2% of sperm 24, + 13, vs. 0.07% in controls), which was statistically significant. The rates of disomies 21, X, and Y were within the control ranges. Chromosome 13 and, from other studies, chromosome 1 may be more susceptible, as they go through meiotic disjunction, to an untoward influence of toxic substances in cigarette smoke. Since most trisomy 13 is due to a maternal meiotic error, and given the fact that the excess is, in absolute terms, very small, it seems safe to suppose that fathers who smoke contribute scarcely, if at all, to the totality of this particular aneuploidy. As for alcohol, the observation of a negative association between sperm disomy frequencies and alcohol consumption in one study6 (Härkönen et al., 1999), and noting also the figures above on maternal consumption, should not lead one to advise that couples drink more heavily prior to a planned impregnation!
As Wyrobek and Adler (1996) comment, “it has been more than half a century  since Muller demonstrated that X-rays can induce germinal mutations in Drosophila, yet questions as basic as the existence of even a single human germinal mutagen remain unresolved.” McFadden and Friedman, writing in 1997, noted that no environmental agent has been identified in which it could be stated, beyond reasonable doubt, that this agent would cause chromosome abnormalities in the offspring of exposed parents. While some studies have shown increased rates of aneuploidy in sperm, the practical fact remains that there is no excess of children born with chromosomal syndromes. Only in 2001 could Marchetti et al. claim, with respect to their work on etoposide exposure with a mouse model (noted above), that “we know of no other report of an agent for which paternal exposure leads to an increased incidence of aneuploidy in the offspring.”
Encouragement can be drawn from this largely negative information, and the counselor will usually be justified in offering substantially reassuring advice from the particular focus of chromosome abnormality. Reference to the commentaries above may provide useful supporting information for the individual agent of specific interest. Prenatal diagnosis would be a discretionary option, as would be preimplantation diagnosis for those whose treatment-related infertility requires IVF. In terms of the specific question of previous treatment for cancer, Byrne (1999) does remain cautious, considering that on the rather few data available there are only “limited grounds for reassurance,” and sees us as being “in the infancy of studies of germ cell mutagenesis in cancer survivors.” She emphasizes that the opportunity to assess newer cancer treatments has not yet arisen, and that the time frame for these assessments is to be measured in decades. Certainly, and more particularly for those agents in which a biological link could plausibly be proposed, it is right that research in this area continue.
1. Radiation and bleomycin, both having potent DNA-breaking properties, cause lymphocyte chromosome rearrangements in normal cells, and considerably more so in ataxia-telangiectasia cells (p. 305).
2. The five centers are the Universities of Iowa, Kansas, and Texas, California Department of Health Services, and Connecticut Tumor Registry.
3. Male cancer patients may show abnormal sperm genetic studies ahead of having received any treatment, suggesting that there is a harmful effect of the malignant disease per se (Hallak et al., 2000a; Kobayashi et al,. 2001).
4. Nevertheless, it cannot be overlooked that radiation-induced chromosome rearrangements can remain more or less permanently in the individual, at least in marrow, as shown in long-term follow up of nuclear-weapons workers (Hande et al., 2003); and thus a possibility that the same thing might happen in germinal stem tissue cannot be completely discounted.
5. Post hoc, ergo propter hoc (Latin) = Something happened after the event, and therefore it must have been due to the event.
6. But another study showed a positive association with alcohol, as well as with caffeine (Robbins et al., 1997b).