by Alyssa M. Voss, M.P.H.
In September 2014, the Journal of the National Cancer Institute (JNCI) published the first-ever comprehensive evaluation of risk factors for non-Hodgkin lymphoma (NHL) subtypes. The compendium of 13 papers, reporting pooled analyses from the International Lymphoma Epidemiology Consortium, or InterLymph, was led in part by DCEG’s Lindsay M. Morton, Ph.D. Reflecting five years of work by investigators from 20 studies, and including 17,471 NHL cases and 23,096 controls, the monograph is the largest evaluation of NHL to date.
NHL is the seventh most commonly diagnosed cancer in the U.S., with a five-year relative survival rate of 60 percent. It is also the most common hematologic malignancy worldwide, affecting more than 500,000 individuals each year. There are more than 30 biologically and clinically distinct subtypes, and while scientists have long suspected etiologic similarities and differences among the various lymphoid diseases, previous studies had been limited by small sample sizes and consideration of limited risk factors.
To address these concerns, investigators from nine studies from the U.S., Europe, and Australia joined together in 2001 to launch InterLymph. Members of the consortia pooled their case-control study data and collaborated on studies of mutual interest to expand knowledge on lymphoma etiology and survival. In 2005, Dr. Morton and colleagues published results from the first two analyses on NHL risks associated with smoking and alcohol. Since that time, 11 additional studies have joined the consortium and 19 more analyses have been published from collaborative efforts evaluating occupational, environmental, molecular, and genetic risk factors.
As the number of studies in InterLymph grew, Dr. Morton and colleagues envisioned an ambitious effort to conduct a coordinated, comprehensive evaluation of risk factors for NHL subtypes across the consortium. In 2009, they initiated the InterLymph Non-Hodgkin Lymphoma Subtypes Project. Combining the data from 20 InterLymph studies provided sufficient power to investigate 11 different subtypes. The goals were two-fold: to comprehensively evaluate risk factor profiles, including medical history, family history, lifestyle, and occupational factors and to quantitatively compare risk factor profiles across subtypes.
The project was truly interdisciplinary, involving nearly 100 investigators from the participating studies, and overseen by the Project Coordinating Committee led by epidemiologists Dr. Morton and Martha S. Linet, M.D., M.P.H., from DCEG and Dr. James Cerhan from Mayo Clinic; biostatisticians Joshua N. Sampson, Ph.D. of DCEG and Dr. Susan Slager from the Mayo Clinic; and pathologist Dr. Dennis Weisenburger from City of Hope. Experts at the Mayo Clinic, the Project’s data coordinating center, harmonized the data from all 20 studies and implemented the quality control measures needed to maintain data integrity across the separate analyses. Statistical analyses were conducted centrally at NCI, led by Dr. Sampson, who designed approaches to assess risk factor associations by subtype, evaluate risk factor heterogeneity among subtypes, and cluster the subtypes into groups with similar risk factor profiles.
Martha S. Linet, M.D., M.P.H.
Sam M. Mbulaiteye, M.D.
Lindsay M. Morton, Ph.D.
Joshua N. Sampson, Ph.D.
City of Hope
Sophia Wang, Ph.D.
Dennis D. Weisenburger, M.D.
Comprehensive Cancer Center of Bordeaux and South West, France
Alain Monnereau, M.D., Ph.D.
INSERM, CESP, University Paris South, Villejuif, France
Jacqueline Clavel, M.D., Ph.D.
Institut Català d'Oncologia-Catalan Institute of Oncology, Spain
Silvia de Sanjosé, M.D.
Karin E. Smedby, Ph.D.
James R. Cerhan, M.D., Ph.D.
Susan L. Slager, Ph.D.
University of California, Berkeley
Christine F. Skibola, Ph.D.
University of California, San Francisco
Paige M. Bracci, Ph.D., M.P.H., M.S.
University of Chicago
Briseis Aschebrook-Kilfoy, Ph.D.
Brian C.H. Chiu, Ph.D.
University of New South Wales, Australia
Claire M. Vajdic, Ph.D.
Yale School of Public Health
Yawei Zhang, M.D., Ph.D., M.P.H.
Overall, the findings confirmed many of the differences previously reported for risk factors across subtypes, but also revealed patterns that had not been observed in smaller studies. Notably, a distinctive pattern was seen between T-cell and B-cell lymphomas. Autoimmune diseases that activated B-cells, such as Sjögren’s syndrome, were more likely to increase risk of B-cell lymphomas, whereas diseases that activated T-cells were more likely to increase the risk for T-cell lymphomas. Immune dysfunction was strongly associated with certain subtypes of lymphomas (peripheral T-cell lymphoma, marginal zone, Burkitt, diffuse large B cell, and lymphomplasmacytic lymphoma), but not others (mycosis fungiodes/Sézary syndrome, chronic lymphocytic leukemia/small lymphocytic lymphoma, follicular lymphoma, mantle cell lymphoma). Because of the large number of cases, the team was able to identify novel associations for rare subtypes such as hairy cell leukemia and acute lymphocytic leukemia that had previously not been possible. Still, some risk factors, particularly lifestyle and occupational factors, only showed relatively modest associations.
The Subtypes Project highlights the need for further investigation into the specific types of immune perturbations that are linked to certain NHL subtypes. Dr. Morton also indicated that the search for new risk factors continues. “The factors we report in the monograph still only explain a modest number of NHL cases. There are some subtypes for which very few risk factors have been identified. Additional infectious agents, chemical exposures, and hormonal pathways are a few examples of areas that should be examined.”
In addition, Dr. Morton is in the midst of planning a study to combine information on risk factors with genomic data from another large-scale effort—the NHL Genome-Wide Association Study (GWAS), led in part by DCEG’s Nathaniel Rothman, M.D., M.P.H., M.H.S, and Sonja I. Berndt, Pharm.D., Ph.D.—to unlock the relationship between a patient’s genetic make-up, risk factor information, and risk for these malignancies.
“By effectively collaborating and combining resources on such a large scale, we were able to achieve some of the founding goals of the InterLymph Consortium,” stated Dr. Morton. “We have advanced our knowledge and understanding of NHL much more rapidly by simultaneously considering many different risk factors among numerous NHL subtypes. For me, this has been most satisfying outcome of this project.”