Increased Risk Of Brain Tumors Among First-Degree Relatives Of Patients With Therapy-Related Myelodysplasia and Acute Myeloid Leukemia (t-MDS/AML)

t-MDS/AML is a leading cause of non-relapse mortality among patients undergoing autologous hematopoietic cell transplantation (aHCT) for Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL). A combination of therapeutic exposures prior to aHCT, aHCT-related conditioning regimens, peripheral blood stem cell mobilization and post-aHCT hematopoietic regeneration contribute to the development of t-MDS/AML. However, significant inter-individual variability is observed despite similar therapeutic exposures. This inter-individual variability could possibly be explained by low-penetrance, high-prevalence polymorphisms in genes involved in DNA damage and repair. On the other hand, a small but quantifiable fraction of the inter-individual variability observed in the risk of t-MDS/AML (in the context of comparable genotoxic exposures) could be attributable to the familial predisposition to cancer that is associated with rare polymorphisms in high-penetrance genes – an area that has not been explored thus far.

Methods

This study aimed to describe the excess risk of specific cancers among first-degree relatives of HL/NHL patients with t-MDS/AML after aHCT compared to first-degree relatives of HL/NHL patients without t-MDS/AML. We have constructed a prospective, longitudinal cohort of patients undergoing aHCT for HL and NHL, where patients are followed from pre-aHCT to 10 yrs post-aHCT (period of risk for t-MDS/AML) allowing for complete ascertainment of t-MDS/AML cases. Study participants were invited to complete a family history questionnaire. Probands with at least one sibling, parent or offspring with cancer were defined as having a positive family history of cancer. Person-yrs at risk for the cohort of relatives were determined from birth to the development of cancer, death or date of questionnaire completion (whichever occurred first). Person-yrs at risk were stratified by age, sex and calendar yr and applied to SEER registry data to yield expected numbers of each type of cancer to compute standardized incidence ratios (SIR: observed/expected) with 95% confidence intervals (CI) for cancer incidence in first-degree relatives of probands with and without t-MDS/AML.

Results

The 446 HL/NHL patients in the cohort reported on 2,664 first-degree relatives, yielding 130,578 person-yrs of follow up. A positive family history of cancer was reported by 103 patients, with 370 incidences of cancer. When stratified by t-MDS/AML status, a positive family history of cancer was identified in 15 of 64 patients with t-MDS/AML (23%) vs. 88 of 382 (23%) patients without t-MDS/AML, yielding an overall SIR of 1.09 (0.82-1.42) for relatives of HL/NHL patients with t-MDS/AML vs. 1.01 (0.89-1.13) for relatives of HL/NHL patients without t-MDS/AML (p=0.60) (Table). However, examination of risk by family history of specific cancer type demonstrated an excess risk of brain tumors among relatives of HL/NHL patients with t-MDS/AML (SIR=5.27, 95% CI, 1.89-11.31) as opposed to relatives of lymphoma patients without t-MDS/AML (SIR=1.55, 95% CI, 0.71-2.89, p=0.03) (Table). This excess risk was contributed to largely by brain tumors among fathers of patients with t-MDS/AML (SIR=5.84, 95% CI, 0.97-18.05) when compared with fathers of patients without t-MDS/AML (SIR=1.05, 95% CI, 0.18-3.27, p=0.09).

Conclusions

This study demonstrates an excess risk of brain tumors among first-degree relatives (and in particular fathers) of HL/NHL patients with t-MDS/AML after aHCT, as compared to HL/NHL patients without t-MDS/AML. This information may serve as a basis for the discovery of underlying genetic predisposition syndromes as well as specific genes responsible for their development.

No relevant conflicts of interest to declare.