Characteristics of Environmental Exposure-Associated Versus Unexposed High-Grade B-Cell Lymphomas in Veterans: A Single Center, Matched Case-Control Study

Introduction: Environmental exposures, such as Agent Orange and other chemicals, are associated with the development of non-Hodgkin lymphomas (NHL). Military personnel and civilians in theaters of conflict are especially at high risk. The mechanism is still not well understood. We have initiated studies to examine the association of dioxin exposure, changes in DNA methylation, and the development of high-grade B-cell NHL.

Methods: Patients diagnosed with high grade B-cell NHL were identified from the VA Long Beach medical record system for database abstraction and tissue analysis. Subjects with sufficient clinical data including diagnosis, date of diagnosis, age of diagnosis, exposure to toxins, treatments, date of last follow up, date of death, cause of death were included in the study. Patients were matched by exposure vs no exposure and international prognostic index (IPI), as well as duration in military and tour of duty when possible. Formalin-fixed paraffin-embedded specimens from a subset of these patients (where sufficient tissue was available for analysis) with and without known exposures were analyzed for differences in DNA methylation by enzymatic methyl sequencing. Descriptive statistics were used to summarize patient characteristics. Two sample t- and chi-square tests were used to compare the exposed and unexposed lymphoma groups. Overall survival was calculated from time of diagnosis to death from any cause and plotted on a Kaplan-Meier curve and compared using the Log rank test. The Fisher's exact test was used to compare the causes of death between the two groups using SAS 9.

Results: Fifty-two patients were included in the study. Patient characteristics were well-balanced between the two groups. The duration of active duty, interval from military enlistment to diagnosis of NHL, and first-line therapy (generally anthracycline-based) were similar. There was a trend toward higher proportion of patients with advanced (stage III or IV) disease and younger age of diagnosis in patients with exposure-associated NHL (p=0.07). The median duration of response to first-line therapy (DOR1) trended towards being shorter (median, 0.8 years; range 0.03-25 years) in patients with exposure related NHL compared to median DOR1 of 2.7 (range, 0.02-25 years) in those without exposure, p=0.056. Median overall survival was 11.5 years in the exposure group and 10 years in the unexposed group, p=0.31. A higher portion of patients in the exposed group died from NHL compared to unexposed group (38% vs 15%), p=0.057. Other competing causes of death, which were more likely in the unexposed group, included other cancers, cardiac disease, infection, end stage kidney disease, dementia and unknown. Six exposed and six unexposed samples were sequenced with over 10,000 statistically significant differentially methylated regions (DMRs) out of 40,605 DMRs tested. Details of differences in methylation of specific genes and their potential pathophysiological and therapeutic implications will be discussed in the presentation.

Conclusions: Patients with exposure-related high-grade B cell NHL who were matched by IPI with unexposed lymphomas did not differ significantly in baseline characteristics including time from enlistment to diagnosis and overall survival. In the exposure-related NHL group, certain findings trended toward significance, including younger age at diagnosis, higher proportion diagnosed at an advanced stage, shorter DOR1, and higher proportion died of NHL. This single center cohort, however, was small and heterogeneous. We are therefore currently expanding this study to examine a larger, multi-center cohort to confirm and extend our clinical and translational findings.