Congenital Abnormalities: A Legacy of Cancer Treatment?

Abstract

Both childhood and adult cancer survivors are keenly interested in understanding the potential battery of long-term health effects related to their disease and treatment. Unfortunately, the potential hazards are many. For example, large numbers of cancer survivors are left permanently infertile by therapy-induced gonadal failure (1), heart disease is among the serious sequelae of anthracycline treatment and mediastinal radiotherapy (2), and subsequent malignancies related to chemotherapy and radiotherapy can arise in several organ systems (3). In this issue of the Journal, Stahl et al. (4) tackle another pressing health question for cancer survivors—might they somehow pass along genetic damage capable of causing adverse outcomes in their children? Among the outcomes studied in this investigation, including preterm birth, low birth weight, and size for gestational age, only major congenital abnormalities emerged as a potential outcome associated with a paternal cancer diagnosis. With a prevalence of 3.7% among the children of male cancer survivors and 3.2% for men without a history of cancer, the children of cancer survivors had a modest 17% relative increase in major congenital abnormalities (adjusted relative risk 1.17, 95% confidence interval 1.05 to 1.31). In this article, we see excellent use of the nationwide registries from Denmark and Sweden. The medical histories of the over 1.75 million children included in the analysis could be linked back not only to the cancer status of their father but also to their mothers for whom data such as maternal age, parity, and even smoking behavior during pregnancy were available. Nevertheless, there are inherent shortcomings to the use of these registries to address the study hypotheses. An important one in this case is the limited treatment data available for the cancer survivors, which prevented more than educated guesses about the types and doses of exposures that were applied (a daunting task given the nearly 50-year span since the start of the cancer registries, a time period during which cancer treatment practices evolved substantially). Thus, some of the investigator-defined groupings of cancers are unlikely to reflect homogeneous treatment exposures, and dose –response cannot be evaluated. But aside from these shortcomings, the analyses do not tell a cogent story. For example, why would a paternal history of skin cancer have among the strongest associations with future risk of major congenital abnormalities, where a paternal history of testicular cancer, a prime suspect, shows no association? Also, chromosomal abnormalities are among the only outcomes showing an inverse association with paternal cancer history, against expectation when looking for germline damage. Further telling, for the male cancer survivors who used assisted reproductive technologies, an equal proportion (4.4%) of children conceived with cryopreserved pretreatment sperm and with post treatment sperm had a major congenital abnormality. This sub analysis is an important one, and the investigators deserve kudos