Risky Business

By Suzanne D. Vernon, Ph.D.
Scientific Director

A paper titled “CFS and subsequent risk of cancer among elderly US adults” was published online on May 30, 2012 in the journal Cancer. (1) The authors of the paper, Cindy M. Chang, Joan L. Warren and Eric A. Engels, are from the division of cancer epidemiology and genetics at the National Cancer Institute (NCI), part of the National Institutes of Health in Rockville, Maryland. These investigators state that they conducted this important study because of the parallels they saw between CFS and cancer. That is, both are associated with immune abnormalities and infection. Over the past 20 years, other studies have hinted at a possible link between CFS and cancer. However, these studies involved relatively few patients and their findings were inconsistent. In contrast, for this study the NCI team linked nearly 1.2 million records from the NCI’s Surveillance, Epidemiology and End Results (SEER) registry with 100,000 Medicare records.

The investigators identified 1,176,950 people with first-time cancer between the ages of 66-99 from 1992-2005 in the SEER database; these were the “cases” in this study. “Controls” were 100,000 living individuals who were cancer-free Medicare beneficiaries. The cancer cases and cancer-free controls were similar as far as sex (53% male) and age. There were more whites in the cancer cases (85.4 percent vs 83.5 percent). Overall, the cancer cases and controls were comparable.

CFS was identified through Medicare claims using the ICD-9 codes for “chronic fatigue syndrome” (780.71) and “neurasthenia” (300.5). They used both codes because some of the cases predated the 780.71 CFS code, which was introduced in 1998. The neurasthenia code, 300.5, was available in claims going back to 1991. CFS was further verified with documentation of a hospital claim or at least two physician or outpatient claims in the individual’s record. Importantly, the CFS or neurasthenia diagnosis had to precede the cancer diagnosis by at least one year. This reduced the possibility that the symptoms were a consequence of a known cancer.

CFS occurred at a rate of 0.5% in both the cancer (case) and cancer-free (control) populations. This is slightly higher than the prevalence reported from two other region-specific community based studies of smaller populations. The SEER program covers cancers occurring in about 26 percent of the U.S. population, while the Medicare program covers approximately 97 percent of the U.S. population aged greater than 65 years.

The investigators assessed the rates of 40 different cancers (and 1 miscellaneous cancer category) in the cases identified with a prior CFS diagnosis. They found a significant association with non-Hodgkin lymphoma (NHL) in the cancer cases. This significant association with NHL occurred with CFS defined with both ICD codes combined, or just the CFS code (and not the neurasthenia code). It also held up if they changed the criteria to require that the CFS diagnosis occurred at least two years before the cancer diagnosis.

In the initial analysis, there were four other cancers that showed association with CFS. When they adjusted for multiple comparisons, these other cancers were no longer significantly associated. They found some other medical conditions were also associated with CFS in the control (cancer-free) population: transfusion, hepatitis C infection, Sjögren’s syndrome and rheumatoid arthritis. The paper does not state how these conditions were temporally associated with CFS – whether they followed CFS or preceded it. When these medical conditions were excluded from the cancer cases, the association of CFS with NHL remained. The NHL subtypes most strongly associated with CFS were diffuse large B cell lymphoma, marginal zone lymphoma and B cell NHL (not otherwise specified). They qualify the findings with this statement: “We could not estimate the absolute risk of NHL associated with CFS, but the risk is likely too small to affect the clinical management of patients with CFS.” They also recognize that it may not be appropriate to generalize these results to a younger population.

It’s important to note that there were some cancers that were found to be less prevalent than expected among individuals who had CFS diagnoses before their cancer diagnosis: breast cancer and cancers of the oral cavity and pharynx. They did not find an association between CFS and prostate cancer, arguing against XMRV as a potential cause of both conditions, and they did not find an association between CFS and risk of brain tumors, as has been reported in earlier studies of a northern Nevada cohort by Paul Levine, Dan Peterson, et al., referenced in this paper. CFS was not associated with overall cancer in this study.

The authors provide a brief but elegant review of immunological findings and infectious agents associated with CFS documented in the literature. They draw links between similar findings in NHL:

“In concert with these immune abnormalities, viruses and bacteria such as Epstein-Barr virus, cytomegalovirus, parvovirus B19, hepatitis C virus and chlamydia pneumoniae, have been implicated in CFS. Although the link between CFS and these infections is disputed, it is noteworthy that some of these infections and other immune-related medical conditions are associated with diffuse large B cell lymphoma and marginal zone lymphoma in other contexts.”

They conclude, “Chronic immune activation or an infection may play a role in explaining the observed association between CFS and NHL. Our study results support continued efforts to understand the biology of CFS.” This is an important statement from an expert group of investigators at a top cancer institution, derived from a (extremely) large population data set that supports earlier observations of increased rates of certain cancers in CFS. This association between CFS and NHL may also provide an explanation for the demonstrated effectiveness of Rituxan in a majority of the CFS patients who have been studied in small clinical trials. (2) Notably, Rituxan was first approved to treat NHL.

Look-back studies like this one, using existing datasets, are likely to become more common as cost-effective means of identifying tantalizing associations in large populations. It is our hope that this important study lends further credibility to the infectious and immunologic underpinnings of CFS and that it spurs research into related treatments for CFS. It’s also possible that CFS might be recognized as an early “warning signal” for people at risk for NHL, but that possibility will require more study too.


1) Chang CM, Warren JL, Engels EA. Chronic fatigue syndrome and subsequent risk of cancer among elderly US adults. Cancer. Published online ahead of print, May 30, 2012. Link to abstract: http://onlinelibrary.wiley.com/doi/10.1002/cncr.27612/abstract

2) Fluge Ø, Bruland O, Risa K, Storstein A, Kristoffersen EK, Sapkota D, Næss H, Dahl O, Nyland H, Mella O. Benefit from B-lymphocyte depletion using the anti-CD20 antibody rituximab in chronic fatigue syndrome. A double-blind and placebo-controlled study. PLoS One. 2011;6(10):e26358. Link to study: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0026358

What are the risk factors for non-Hodgkin lymphoma? American Cancer Society. Accessed June 3, 2012: http://www.cancer.org/Cancer/Non-HodgkinLymphoma/DetailedGuide/non-hodgkin-lymphoma-risk-factors

Information about odds ratios: Wikipedia. Accessed June 3, 2012: http://en.wikipedia.org/wiki/Odds_ratios

Suzanne D. Vernon, PhD, is the Solve ME/CFS Initiative’s scientific director. She has nearly two decades of experience as a microbiologist.

June 3, 2012