It had been 10 years since the National Institutes of Health (NIH) last hosted a “State of the …” meeting on chronic fatigue syndrome (CFS)(1), and interest in the meeting was high in weeks leading up to the workshop and during the event. Use of the term ME/CFS in the workshop title (a first for a U.S. government agency), anticipation created by the late posting of speakers and topics and the sustained tensions over XMRV’s association with CFS generated both heat and light during the planning process. By the time of the workshop, held Apr. 7-8, 2011, more than 150 people attended the meeting and more than 900 viewed the meeting via the live webcast.
This workshop also marked the first time an NIH director addressed a meeting specifically dedicated to CFS. Dr. Francis Collins spoke at the conclusion
of the second day of the meeting, in spite of the looming threat of a federal government shut-down. Dr. Collins said that Secretary of Health and Human Services Kathleen Sebelius had directed the NIH and other agencies to give CFS special attention. He emphasized the need for further definition of the illness, identification of subsets, and the difficulties caused by lumping many potential causes together.
Another member of the NIH leadership, Dr. James Anderson (director of the Division of Program Coordination, Planning and Strategic Initiatives), opened the meeting and shared a letter from Secretary Sebelius with the Workshop attendees. Secretary Sebelius acknowledged the devastating effects of ME/CFS, and said, “I want to assure you that several agencies within the Department of Health and Human Services (HHS) are working together with the Chronic Fatigue Syndrome Advisory Committee and advocacy groups to develop interdisciplinary initiatives that will address important aspects of this illness, including improved diagnosis and treatment.” In his opening remarks, Dr. Dennis Mangan (senior advisor, Office of Research on Women’s Health, NIH) identified three goals for the Workshop: 1) document what we currently know about this illness, 2) look for gaps in our knowledge needing more research and 3) identify outstanding opportunities in science and technology that might advance biomedical research on ME/CFS.
Dr. Anthony Komaroff (Harvard Medical School) offered an overview of the clinical presentation of CFS. Drawing on his experience as a physician treating CFS patients, Dr. Komaroff described the illness symptoms, common lab abnormalities and the severe disability endured by some patients. Dr. Komaroff identified three possible causes of CFS: primary or secondary central nervous system disturbance; dysregulation in the production of cytokines; and an infectious trigger or perpetuation of the illness.
Dr. Leonard Jason (DePaul University) introduced the attendees to the importance and pitfalls of the diagnostic criteria and case definitions for CFS. The selection and application of a case definition has profound impacts on research cohorts and comparability of data. Dr. Jason pointed out that CFS can be distinguished from major depression by the presence/absence of self-reproach and post-exertional malaise, but that many research studies do not account for these differences in patient selection. He called upon the participants to address the definitional issues head on, and stated that it is time to create “gold standard” criteria.
Dr. Ronald Glaser (The Ohio State University) presented his work on the pathophysiology of Epstein-Barr virus (EBV). EBV can be reactivated in the body, possibly by excess cortisol, and there may be unique antibodies to EBV proteins.
Reviewing his work on enteroviruses and CFS, Dr. John Chia (EVMED Research) stated that he has found enteroviruses in muscle biopsies, gut biopsies and serum from CFS patients. The overall rate of positivity in CFS cohorts ranges from 20 percent to 87 percent. In his experience, the sickest patients tend to have higher incidence of these viruses, and he hypothesized that chronic infections could stimulate ongoing immune responses leading to CFS symptoms.
Dr. Judy Mikovits (Whittemore Peterson Institute) presented her latest data on evidence of XMRV infection in CFS patients. Using multiple detection methods, she reported that samples at the Whittemore Peterson Institute are positive for XMRV and polytropic MLVs, and that they are finding variation in the sequences. Dr. Mikovits stated that there is no evidence for contamination in WPI’s cell lines, the Lombardi et al. study or the Lo et al. study. Serology is the best method for detecting XMRV, and the next step in research is more full-length sequencing of clinical XMRV isolates.
Taking a different view of XMRV, Dr. John Coffin (National Cancer Institute, NIH and Tufts University) presented data on the possible origins of XMRV. More than 70 mouse strains have been surveyed for unique XMRV sequences and none have been positive. However, a provirus called pre-XMRV2 that is highly similar to XMRV was found in half of the lab mouse strains examined. Work by Dr. Vinay Pathak (National Cancer Institute, NIH) has shown that another provirus, pre-XMRV1, has been found in different lab mouse strains. Pre-XMRV1 and 2 together are almost identical to XMRV, and this data strongly suggests that XMRV is a lab recombinant virus. Dr. Coffin also highlighted the ease with which contamination can spread through cell lines, reagents and equipment. (This research has since been published in Science on May 31, 2011.)
These contrasting viewpoints sparked charged debate during the discussion period; Dr. Harvey Alter (Clinical Center, NIH) called the presentations “provocative” and reiterated his belief that whichever way the XMRV story ultimately unfolds, he is convinced there is something real that needs to be found. He suggested that it might be one or more infectious agents, a unique host response or a combination of factors. He called the data for XMRV as a recombinant “very convincing” but in a later summary said that the two large studies being funded by NIH(2) will likely determine the outcome of the debate.
In summary, there is evidence that numerous pathogens have been detected in CFS patients, and are known to trigger and/or exacerbate symptoms. Infectious pathogens could contribute to the etiology of CFS in several ways: a single infectious agent initiates a chronic host response; a variety of agents trigger an immune-mediated post-viral syndrome; and/or susceptibility factors influence response to infectious agents. At this time the evidence is insufficient to fulfill Koch’s postulates or prove cause and effect.
Dr. Gordon Broderick (University of Alberta) presented his network modeling of 30 neuroendocrine immune biomarkers, including those related to immune function, immune cell signaling and immune metabolism. The networks of pathways in these systems clearly distinguish CFS patients from patients with Gulf War Illness and healthy controls. The CFS networks have fewer connections and actually “shed” connections, while the Gulf War Illness networks have tighter networks than healthy controls.
Examining CFS samples collected by Centers for Disease Control & Prevention (CDC), Dr. Mangalathu Rajeevan (CDC) has identified 38 single-nucleotide polymorphisms (SNPs) associated with CFS. These genetic variations affect components of the immune system, including the complement cascade, chemokines, cytokines and toll-like receptors. Dr. Rajeevan suggested that these results should be validated in larger studies examining pathway-genomics.
Dr. Keith Kelley (University of Illinois at Urbana-Champaign) presented his data from mouse models showing that systemic infection leads to brain inflammation. This systemic periphery inflammation affects behavior including fatigue, sleep impairment, pain sensitivity and depressive behavior. Dr. Kelley proposed that an immunomodulatory enzyme called IDO could be the master switch wherein an infection can lead to prolonged inflammation and symptoms. It is possible that therapeutic interventions could be tested in such an animal model prior to trials in humans.
Discussion highlighted the possibility that systems modeling can identify differences in patients with clinically identical presentations. Much more data is needed to overcome the possible heterogeneity of CFS sample cohorts. Systems modeling can break down the research “silos” of specialization. A systems approach could help researchers combine disparate information and disciplines, leading to new theories and testable hypotheses. There is also promise in using systems modeling to identify regulatory networks in the body that might benefit from “rebooting” to correct imbalances.
Dr. Mary Ann Fletcher (Miller School of Medicine, University of Miami) discussed immune abnormalities found in CFS patients, including low/poor functioning natural killer cells and abnormal levels of cytokines. Dr. Fletcher’s group has incorporated exercise testing and multiple time point sampling into some studies to identify differences in immune function correlated to severity of symptoms. Treatment possibilities include targeting cytokine networks and testing drugs already approved for multi-system illnesses like rheumatoid arthritis and multiple sclerosis.
Dr. Benjamin Natelson (Albert Einstein College of Medicine) reviewed data on the immunology of CFS and fibromyalgia. He pointed out that CFS symptoms could be reproduced by injection of pro-inflammatory cytokines. While there are very few immune markers that are consistently abnormal in CFS patients, the recent study by Schutzer et al. demonstrated a unique protein signature in the cerebrospinal fluid of CFS patients. Dr. Natelson stated that his own data and experiences suggest that CFS is a central nervous system disorder and not an immune disorder.
In summary, immune parameters have been studied in CFS, but the findings are inconsistent. However, those studies that find immune disturbances support a hypothesis that CFS, or a subset of CFS, is an immune dysfunction state (e.g., activated and inflamed).
Dr. Kathleen Light (University of Utah) presented her data showing differences in gene expression between CFS and other groups after an exercise challenge. CFS patients, with and without co-morbid fibromyalgia, showed excessive increases in gene expression on adrenergic and sensory receptors and cytokines as compared to healthy controls, patients with multiple sclerosis and patients with fibromyalgia without CFS. The exercise challenge was an essential component of measuring the group differences, supporting the unique role of post-exertional malaise in CFS. Dr. Light stated that several of the receptors could serve as a biomarker for CFS. (This data has since been published in the Journal of Internal Medicine, May 26, 2011.)
Dr. Roy Freeman (Beth Israel Deaconess Medical Center, Harvard Medical School) reviewed the types of orthostatic intolerance that affect people with CFS. Orthostatic hypotension, delayed orthostatic hypotension, neurally mediated syncope, and postural tachycardia syndrome have all been found in CFS patients. However, Dr. Freeman observed that these conditions are more severe in patients with other (primary) autonomic nervous system disorders.
Dr. James Baraniuk (Georgetown University) presented his data on proteomic studies of cerebrospinal fluid in CFS patients. There is evidence for central nervous and immune system involvement in CFS, and Dr. Baraniuk offered several theories related to neuroimmunology in CFS.
Taken together, there is substantial evidence for cross-talk between the nervous and immune systems. Orthostatic intolerance points to sympathetic nervous system over-activation. There is likely central sensitization and cross-regulating of the central nervous system and immune system in CFS patients.
Exercise Physiology and Energy Metabolism
Dr. Jane Kent-Braun (University of Massachusetts, Amherst) indicated that the scientific literature on CFS does not suggest a material difference in muscle strength, muscle fatigue or muscle oxidative capacity between patients and controls. However, CFS patients and controls do differ in their ratings of perceived exertion for the same task. Dr. Kent-Braun suggested that altered cortical response could explain the different experiences of exertion and proposed this as an area for future studies.
Dr. Christopher Snell (University of the Pacific) presented data on the results of exercise testing in patients with CFS versus sedentary controls. He argued that cardiopulmonary exercise testing (CPET) with gas exchange is the best way to measure the effects of exercise and that subgroups of CFS patients can be identified based on the level of disability identified through the testing. CPET can show evidence of abnormality across multiple body systems including the immune and autonomic nervous system. Dr. Snell suggested that CPET can be an objective measure of post-exertional fatigue in CFS and could serve as a functional endpoint for treatment.
Returning to orthostatic intolerance, Dr. Peter Rowe (Johns Hopkins University School of Medicine) showed that orthostatic disturbances are strongly associated with CFS. Upright posture aggravates CFS symptoms, often before any change in heart rate or blood pressure is detected. Treatment of orthostatic intolerance is associated with improvement in a subset of CFS patients. However, Dr. Rowe said that very large treatment studies are needed to compensate for the many other factors involved in CFS.
In summary, CFS patients do not appear to have abnormal muscle structure, function or physiology. There is no overt mitochondrial dysfunction, although mitochondria may play a role in CFS symptoms. The evidence points to disturbances in signaling and coordination in the neuroendocrine immune system that affect metabolism.
Diagnosis and Biomarkers
Dr. Nancy Klimas (Miller School of Medicine, University of Miami) focused on candidate immune biomarkers, including reduced natural killer cell function, elevated pro-inflammatory cytokines, reduced CD26 and elevated neuropeptide Y. Gene expression for inflammation and immune activation increases in CFS patients with an exercise challenge. Dr. Klimas noted that methods and standardization for measurement of cytokines has been a recent development and this should support more consistent results going forward.
Dr. Dane Cook (University of Wisconsin-Madison) reviewed the promise of neuroimaging to differentiate CFS patients from controls. Possible measures include gray matter volume, brain blood flow, brain lactate levels and functional neuroimaging to identify active regions during a cognitive task. Dr. Cook said that while neuroimaging does not provide a biomarker at present, he is optimistic for its potential.
Dr. Michael Dean (National Cancer Institute, NIH) addressed the challenges of identifying genes associated with complex diseases. He pointed out that very large cohorts of 5,000 to 10,000 patients are needed to properly scale studies for fine mapping characterization. Three factors that are essential for these studies are an accurate shared phenotype, cell lines and a shared data repository.
The immune system is a reasonable place to discover and validate biomarkers, but the reproducibility and reliability of candidate markers is unknown. The evidence suggests that it is unlikely that CFS will have a single biomarker. Rather, multiple biomarkers from disparate systems may function together to identify – and potentially subtype – CFS patients. Neuroimaging shows promise, but cannot yet be considered a biomarker. Genetic predisposition and genomics may describe pathophysiology as well as possible subtypes of CFS. In order to successfully validate any biomarker, studies must be carefully designed with standardized sample collection and researchers must share those samples for cross-validation.
Dr. Fred Friedberg (State University of New York, Stony Brook) presented his treatment protocol for CFS patients, including pacing, coping skills, scheduled pleasurable activities, stress reduction and customized low-level exercise. He challenged the notion that cognitive behavioral therapy and graded exercise therapy are successful treatments, pointing out that neither therapy has been shown to increase patients’ overall activity level.
Dr. Italo Biaggioni (Vanderbilt University) described the results of his studies looking at postural orthostatic tachycardia syndrome in CFS. Treatment with salt, intravenous saline and short-acting beta-blockers has been successful in some patients.
Dr. Theoharis Theoharides (Tufts University School of Medicine) presented his work on mast cells, a cell of the immune system. When mast cells are activated, the cells release pro-inflammatory mediators. Neuro-immune interactions can augment mast cells activation, and this has implications for many diseases. Dr. Theoharides stated that flavonoids can block mast cells activity and this has implications for treatment. As an area of future inquiry, he suggested looking for evidence of mast cell activation in serum and cerebrospinal fluid from CFS patients. (This research has since been published in the Journal of Clinical Psychopharmacology, June 2011.)
Based on her extensive experience treating CFS patients, Dr. Lucinda Bateman (Fatigue Consultation Clinic) presented her approach to treatment. Dr. Bateman takes a stepwise approach, beginning with an accurate diagnosis of CFS and continuing to pacing, treatment of symptoms that affect function and illness severity (pain, disordered sleep and physical conditioning) and treatment based on the presumed mechanisms and pathophysiology of the illness (autonomic and immune function, hormonal manipulation, and vitamin supplements). She noted that cognitive behavioral therapy is inaccessible to most patients and she focuses on teaching patients self-management techniques.
Discussion focused on the best combination of objective and subjective measures for treatment trials. No single outcome measure was agreed upon, but suggestions included activity levels, improved activity tolerance, fatigue, pain and sleep. Self-report measures are problematic in part because patients may rate their improvement relative to times of worse symptoms, rather than normative function levels. In addition to identifying consensus outcome measures, treatment trials must be large enough to capture the actual effect of treatment amid the “noise” of normal illness variation.
In summary, a systematic approach to management should take into account the variations between case definitions and should include supportive care and self-management. Possible therapeutic targets include rebooting regulatory networks and targeting specific phenotypes such as orthostatic intolerance.
Five presentations centered on the topic of communications. Kim McCleary (CFIDS Association of America) discussed best practices for communicating with research stakeholders; Dr. Kenneth Friedman (Castleton State College) recounted the challenges to his academic career which he attributed to being identified as a CFS advocate; Patricia Fero (Wisconsin ME/CFS Association) challenged the level of funding allocated per CFS patients and the tragedy of there being now three generations of CFS sufferers; Dr. Mary Schweitzer (University of Pennsylvania) described how establishing community with other CFS patients through the Internet had been her lifeline; and John Burklow (Office of Communications, NIH) provided an overview of how NIH provides information about its activities to the various communities it serves.
Gaps and Opportunities
Following summaries by each group of session co-chairs, Dr. Suzanne Vernon (CFIDS Association of America) presented a comprehensive summary of the gaps and opportunities in CFS research identified during the workshop. Gaps in current knowledge and research studies include:
- Different terminology and case definitions blur the lines between research and clinical medicine and create problems in comparing study results.
- CFS research falls into discipline silos.
- There is a lack of standard operating procedures for studies, a lack of common data elements and a lack of coordination and collaboration.
- Study design must account for subtypes, demographics, disease and healthy controls, biomarkers, time course data, clinical trials, early detection, pediatric vs. adult presentations and outcome measures.
- Many types of studies are needed, including longitudinal, natural history, early detection, pediatric vs. adult CFS, genetics and clinical trials.
- Biomarkers must be reproducible, replicated and validated.
- Animal models are needed.
- There are not enough doctors to treat the millions of CFS patients.
- There are not enough researchers.
- There is not enough money.
However, these difficulties can be addressed and progress can be achieved. In lean times, collaboration and communication is needed to leverage existing infrastructure, resources and knowledge. Opportunities for advancing the science of CFS include:
- XMRV has given the field unprecedented awareness, attention and opportunities.
- Leadership is needed to agree on and embrace opportunities.
- Platforms for collaboration should be silo-transparent and include patient derived/reported outcomes and a clinical network to share information.
- Information standardization and aggregation should help identify and prioritize studies, including biomarker discovery, biomarker replication and validation, preclinical and clinical studies and leveraging existing infrastructure and resources.
- Learning from other research areas and diseases must be incorporated.
The NIH will produce its own summary of the workshop that will shape future funding opportunity announcements. The webcast recording can be viewed at http://videocast.nih.gov/PastEvents.asp and other resources are available at http://www.cfids.org/research/sok-resources.asp.
(1) Two “State of the Science” meetings were held in 2000. The first was held Feb. 6-7, 2000. A closed meeting was planned by NIH staff that included just three CFS experts among the speakers, all of whom were psychiatrists. Following vocal objections from members of the Department of Health and Human Services CFS Coordinating Committee (CFSCC, the predecessor to the CFS Advisory Committee) and other advocates, the panel was expanded and members of the public were permitted to attend. At that time, DHHS also announced that a “true” State of the Science meeting would be held later that year. The second meeting took place Oct. 23-24, 2000, and the planning committee included members of the CFSCC and other patient advocates. Six topic areas were addressed: neuroendocrinology, neurocognition, pain, immunology, fatigue and orthostatic intolerance.
(2) The National Institute of Allergy and Infectious Diseases is funding a multi-center study of XMRV being coordinated by Ian Lipkin, M.D., of Columbia University. The National Heart, Lung and Blood Institute is conducting a four-phase blood safety study employing various testing methods and laboratories to detect XMRV and other murine leukemia virus-related viruses. For more information about XMRV, please visit http://www.cfids.org/xmrv/default.asp.
Author Jennifer M. Spotila, J.D., is a member of the CFIDS Association of America’s Board of Directors. Ms. Spotila participated in the conference by webcast.