Program Updates: 2002-2003
Chronic fatigue syndrome (CFS) is a complex, debilitating illness characterized by at least 6 months of severe persistent or relapsing fatigue and a group of characteristic but nonspecific symptoms. Despite more than a decade of extensive research, the cause and pathophysiology of CFS remain unknown and no diagnostic tests exist. Thus, effective prevention and control remain elusive. VEHB conducts a multifaceted, integrated program to study CFS in the United States and internationally. The program's objective is to develop control and prevention measures for CFS. Five specific aims address this objective:
- Estimate the magnitude of CFS as a public health problem.
- Determine if CFS represents a single disease or a common illness response to a variety of insults.
- Define the natural history and clinical parameters of CFS.
- Identify etiologic agents, risk factors and diagnostic markers associated with CFS.
- Provide current and appropriate technical information on CFS to various audiences.
These aims are achieved through the integration of epidemiologic and laboratory investigation. This collaborative strategy includes five components:
- Refinement of the CFS case definition
- Clinical studies
- Molecular Epidemiology
From 1989 to 2001, the CFS program focused on surveillance. Initially, we used sentinel physicians representing Atlanta, Grand Rapids, Reno, and Wichita. This allowed us to determine the burden CFS imposed on the health care system, refine the research case definition, evaluate the clinical course of CFS, and it provided rigorously defined patients who we enrolled in epidemiologic studies of demographic, behavioral, environmental, infectious, and immunologic risk factors for CFS. Beginning in 1995, we shifted surveillance to defined populations and terminated physician surveillance. We initially surveyed the San Francisco population (to test methods). We next conducted a 4-year longitudinal surveillance of CFS and other fatiguing illnesses in the general population of Wichita. Most recently, we conducted a pilot national survey for CFS and other fatiguing illnesses. These large population-based surveillance activities have yielded rigorous estimates of the magnitude of CFS as a public health problem, enhanced our efforts to more rigorously define CFS, and provided specimens (from defined populations of fatigued and non-fatigued subjects followed over time) to test in genomic, proteomics, immunologic, and neuroendocrine assays. Rigorous clinical studies are necessary to define the pathophysiology of CFS, define clinical parameters, and identify diagnostic markers but are difficult to accomplish in the context of population-based surveillance. Currently, the CFS program focuses on detailed clinical studies of persons with CFS and other fatiguing illnesses identified from defined populations and on clinical modeling studies of CFS-like illness following various stresses.
Objectives of the surveillance program are to estimate the magnitude of CFS as a public health problem in the United States, describe the clinical course of CFS, revise the case definition, collect specimens to identify markers of disease activity, and identify subjects for enrollment in clinical studies.
Prevalence and incidence of CFS in adults
We concluded surveillance of CFS in Wichita in 2001, after 3 years of follow-up. Between February and September 1997, we queried 33,997 randomly selected Wichita households representing 90,316 residents and identified those with prolonged fatigue (>1 month). At baseline, 7,162 adults aged 18 to 69 years (all fatigued respondents and a random selection of those not reporting fatigue) were studied in detail. Approximately 4,000 subjects were followed annually. The weighted point prevalence of CFS at baseline, adjusted for non-response, was 235 per 100,000 persons. Only 16% of the subjects we classified as CFS had been diagnosed or treated for CFS by a physician. CFS prevalence was significantly higher among women, 373 per 100,000 persons than among men, 83 per 100,000 persons. Among subjects non-fatigued and fatigued for less than 6 months, the 1-year incidence of CFS was 180 per 100,000 persons.
CFS in adolescents
CFS is an important public health problem in the adult population because of the high prevalence and its long-term debilitating nature. CFS appears to be considerably less common in adolescents and children, but may be even more devastating to adolescents because of their unique physical, emotional, and social attributes. CFS surveillance in Wichita also evaluated fatiguing illness in adolescents. Of the 90,316 residents initially queried, 8,586 were adolescents (12-17 years) and 138 had fatigue for greater than 1 month. Most (107 or 78%) had chronic fatigue (= 6 mo) and 28 of them had exclusionary diagnoses. Thirty-one were considered as CFS-like and eligible for clinic evaluation, but only 11 agreed to participate, and none met the 1994 CFS research case definition. The weighted prevalence of CFS-like illness was 338 per 100,000. Based on the ratio between CFS-like and CFS prevalence in adults, the weighted prevalence of CFS among adolescents was estimated as 50 per 100,000.
Economic impact of CFS
The ability of CFS patients to carry out productive lives can be severely limited and the overall effect of CFS on society is potentially significant. Economic cost of illness provides an important element for understanding the overall impact and in deciding the allocation of health recourses. However, scant empirical scientific work exists to quantify the impact of CFS. Persons with CFS incur direct costs associated with healthcare services and products used for the diagnosis, assessment, and management of an illness. Perhaps more important, persons with CFS suffer indirect costs irrespective of their health care. Indirect costs are affiliated with the loss in productivity attributed to a particular illness - that is, forgone income due to a decrease in hours worked or required job change. Medical and nonmedical costs are usually described as resources expended, and productivity losses are described as resources foregone. We are using data from our surveillance program to estimate lost productivity associated with CFS. These are costs to society as a whole (i.e., society's costs related to reduced levels of output, time spent to obtain health care, and lost productivity resulting from change in employment status as a result of morbidity or mortality).
Clinical course of CFS
Using data from the longitudinal population-based study of fatiguing illnesses in Wichita, we characterized the clinical course of CFS. We identified and followed 65 subjects with CFS for up to 3 years. We evaluated changes in clinical status (partial or total remission, alternative medical or psychiatric diagnoses), CFS case-defining criteria, wellness scores, hours of activities and sleep, and treatments used to reduce fatigue. Associations between risk factors and outcomes were determined by use of logistic regression and generalized estimating equations models. Only 20 to 33% of the subjects were classified as having CFS at follow-up, 57% experienced partial or total remission, 10% sustained total remission, and 23% received alternative diagnoses. Higher fatigue severity scores and total numbers of symptoms were negatively associated with ever remitting. Duration of illness = 2 years was positively associated with sustained remission. Unrefreshing sleep persisted in more than 80% of the subjects across all periods but, as with most of the CFS symptoms, tended to be less frequent over time. The number of activities affected by fatigue decreased over time, while wellness scores increased. Findings confirm the intermittent pattern of CFS relapse and remission. Remission estimates were similar to those reported in clinical studies. The persistence of sleep complaints and identification of sleep disorders suggest that sleep studies might be essential in determining differential diagnoses in CFS subjects.
Sleep assessment in a population-based study of Chronic Fatigue Syndrome
Sleep physiology may be central to understanding CFS. Unrefreshing sleep is the most prevalent of the 8 CFS case-defining symptoms, being endorsed by between 70 to 90% of cases. Most of the postulated etiologies of CFS affect sleep (e.g. infection, cytokines, stress, and hormones) and sleep deprivation produces many of the features of CFS (e.g., fatigue, impaired cognition, joint pain and stiffness). Accurately identifying the relationship of sleep abnormalities to CFS requires as a start, characterizing sleep in persons with CFS. Formal sleep studies require all-night polysomnography and a multiple sleep latency assessment under standardized conditions. Because of the complexity and expense of formal sleep testing, few studies characterizing sleep in persons with CFS are available, and none have been performed in cases identified from population-based studies.
We have screened and profiled sleep abnormalities of persons with CFS identified in the general population of Wichita. To do this, we used two self-administered questionnaires, the Epworth Sleepiness Scale and the Toronto Centre for Sleep and Chronobiology Sleep Assessment Questionnaire© (SAQ©). We administered the questionnaires to 339 subjects from the Wichita surveillance study during their clinical evaluation. Eighty-one percent of subjects had an abnormality in at least one SAQ© sleep factor. Women were more likely than men to have non-restorative disorder and insomnia, and men were more likely to have sleep apnea. Persons classified as CFS had significantly increased risk of abnormal scores in the non-restorative and restlessness SAQ© factors compared to non-fatigued, but not for factors of sleep apnea or excessive daytime somnolence. This is consistent with studies finding that, while fatigued, CFS subjects are not sleepy.
National Survey for CFS
In order to resolve questions concerning associations between CFS, geographic area, type of environment (metropolitan, urban, rural), age, socioeconomic status, and race/ethnicity, we planned a National Survey of CFS. Between May 2001and January 2002, we conducted a pilot national survey to test questionnaires and methods, verify response rates, field-test physical examination protocols, and address issues of response bias. The September 11 tragedies and ensuing events disrupted the pilot survey and likely would continue to affect conduct, and results, of the National Survey, which was scheduled to begin in March 2002. So, we did not begin the larger survey. However, sufficient numbers of subjects were evaluated in the pilot study to permit analyses in several areas.
In the pilot national study, we screened 2,728 randomly selected households (97% response rate) representing 7,317 residents. Sampling included 8 strata (Northeast urban/Buffalo-Niagara Falls, NY, Midwest urban/Chicago, IL, South urban/Baton Rouge, LA, West urban/Oakland, CA, Northeast rural/Franklin County, PA, Midwest rural/Ripley County, IN, South rural/Monroe County, GA, West rural/Chaves County, NM). We identified 744 individuals with fatigue lasting for at least 1 month and 440 between 18 and 69 years of age met eligibility criteria and completed a detailed telephone interview. For comparison, a sample of 444 non-fatigued individuals between the ages of 18 and 69 also completed the detailed interview. The 884 respondents completing the detailed interview represent a weighted population of 180 million individuals in the eight strata. We are currently analyzing the data.
A second major aim of the CFS research program is to determine if CFS represents a single illness or a common response to a variety of insults. CFS is identified by symptoms and has no confirmatory physical signs or laboratory abnormalities. The etiology and pathophysiology remain unknown and there is an amazing lack of consensus in the findings of many well-conducted studies. This may reflect failure of the case definition to accurately identify CFS and confounding of results due to misclassification of persons with other fatiguing illnesses as CFS. Since CFS first became apparent as a public health problem, CDC has had a lead role in developing case definitions for CFS. CDC was responsible for the first such definition in 1988, an international group convened by CDC published the current international CFS Research Case Definition in 1994, and in 2000 CDC convened an International CFS Study Group whose initial goal was to standardize instruments and derive an empiric case definition.
Empiric Definition of CFS in Wichita
The 1994 international CFS Research Case Definition is based on clinical consensus and was not derived empirically from data. We have used baseline data from the Wichita surveillance project to compare symptom complexes specified in the 1994 definition with symptom groups empirically identified among subjects with unexplained chronic fatigue. Dichotomous factor analyses revealed that unexplained chronic fatigue was multi-dimensional and that CFS symptoms were not included in a single dimension. CFS symptoms from the 1994 definition (except severe headaches) were distributed across 3 factors: musculoskeletal, infection, and cognition-mood-sleep symptoms. Thus, CFS is empirically multi-dimensional and it overlaps with other dimensions of unexplained chronic fatigue. Scores measuring the degree of chronic unwellness in each factor indicated that CFS subjects had significantly higher scores than those fatigued for at least 6 months but without the defining symptom complex (CF). Musculoskeletal symptoms seemed to discriminate CFS and CF subjects better than the other 2 factors. Cluster analysis of factor scores also showed overlap between CFS and non-CFS subjects.
International CFS Study Group
In 2000, CDC organized an International CFS Study Group comprised of experts in CFS, epidemiology, infectious diseases, endocrinology, immunology, neurology, psychology, psychiatry, biostatistics, and patient advocacy to consider difficulties in applying the current CFS case definition. At the Group's first workshop in May 2000 participants agreed that the 1994 International CFS Research Case Definition should be revised and that future revisions should be empirically derived based on data from defined populations. We considered how a case definition should be used for research, what population groups should be studied to revise the case definition, and standardization of classification instruments. The second meeting in June 2001 focused on ambiguities of the 1994 case definition and what specific instruments would provide the best objective measures of the major dimensions of CFS. We met for a third time in May 2002 to discuss how standard instruments measuring the major symptom domains of CFS could be used internationally in clinical research settings. The Group gave final approval to a review article critiquing the CFS case definition, which has been submitted for publication. At the meeting we agreed upon specific standard instruments that could be uniformly used in international clinical settings to measure fatigue severity, disability, cognition, sleep, psychiatric comorbidity, and other symptoms. We also discussed study designs and analytic procedures that would identify symptom dimensions among persons with CFS and other unexplained chronically fatiguing illnesses in international clinical research settings and would define the variability of symptom dimensions manifest by different patient populations.
To explore whether there is a common phenotype (symptom complex) characteristic of CFS across cultures, the International CFS Study Group recommended an analysis of existing data from investigators around the world. This would provide additional evidence for the existence (or lack of existence) of a core symptom complex and would also provide information concerning numbers of subjects available for a larger study. We have begun work to locate data sets from CFS patients seen by investigators who care for CFS patients from different health care settings in the English-speaking world (U.S., Australia, U.K.), continental Europe (Germany, France, Scandinavia), and Far East (Hong Kong, Japan). The datasets we selected for analysis include at minimum:
- Demographic details and information as to the setting in which the data was collected.
- Measures of fatigue duration and severity using a defined instrument (e.g., Krupp, Chalder).
- Symptoms specified in the 1994 CFS case definition.
- Other symptoms.
- Identify exclusionary and non-exclusionary medical and psychiatric illnesses.
We will use results from this systematic review in final development of a larger international study. This exercise will also identify information related to our CFS Patient Registry effort.
To address clinical aspects of CFS, we have developed collaborations with a multidisciplinary team from Emory University's Division of Endocrinology, Department of Psychiatry, and Department of Neurology and support their staff through Interagency Personnel Agreements. We also sponsored sabbaticals for Dr. Jim Jones from the National Jewish Medical Center in Denver and Dr. Nancy Klimas from Miami University. The clinical collaborative group was formed to enhance and complement the CFS Research Program's laboratory-based epidemiology core with studies of the neuroendocrine, immune, and central nervous system aspects of CFS.
Clinical Research Unit Studies of CFS in Wichita
We have initiated a study, developed with our clinical collaborators, to characterize neuroendocrine, immune function, cognition, sleep characteristics, and psychologic function of persons with CFS. The study enrolls subjects classified during the Wichita surveillance study as:
- Meeting the CFS research case definition (n=70).
- Meeting the CFS research case definition but accompanied by a major depressive disorder with melancholic features (CFS+MDD/M)(n=41).
- Chronically fatigued but not meeting the CFS research case definition because of insufficient number of symptoms or fatigue severity (ISF)(n=70).
- ISF accompanied by MDD/M (n=39).
- Non-fatigued controls matched to CFS subjects on age, race/ethnicity, sex, and body mass index (n=70).
- Clinical evaluation of medical and psychiatric status.
- Sleep characteristics
- Cognitive functioning
- Measures of stress and coping mechanisms.
- Laboratory testing for neuroendocrine status, autonomic nervous system function, and cytokine profiles.
Finally, we will measure lymphocyte gene expression patterns, proteomics, and polymorphisms in neurotransmission/immune regulation genes. The first study subjects were enrolled in early January 2003 and the study should be completed by June.
Investigations into the pathophysiology of CFS have focused on immune system activation secondary to infection, autoimmune reactivity, and dysfunction of regulatory neuroendocrine pathways. However, these studies have been complicated by patient heterogeneity with respect to chronicity and co-morbid illnesses. "Model" systems in which symptom-free subjects develop CFS-like illness following exposure to a known stimulus (or stress) obviate these problems and allow controlled studies of the pathophysiology of fatigue and associated symptoms as they relate to immune and endocrine activation and central nervous system response.
Emory Collaboration - Department of Psychiatry
Interferon-α (IFN-α), a cytokine widely used in the treatment of hepatitis C (HCV) and malignant melanoma, represents a model system of great promise, because IFN-α activates the immune system and produces a high rate of symptoms commonly observed in CFS (fatigue, cognitive complaints, pain, sleep disturbance and depression). Thus, a CDC/Emory University Collaborative Group has undertaken a series of integrated studies using IFN-α-associated fatigue to model CFS.
Interferon-α-Induced Sickness Syndrome as a Model for the Pathophysiology of Chronic Idiopathic Fatigue. We measure immune and neuroendocrine parameters, sleep, metabolism, brain activity, mood changes, and mRNA expression patterns in IFN-α recipients with HCV infection. Subjects are studied prior to, and after 12 weeks, of receiving IFN-α. HCV subjects randomized to postpone treatment until study completion serve as HCV+ controls. All study subjects undergo sleep polysomnography, body composition analysis, resting metabolic rate assessment and physical activity assessment at home via actigraphy. Subjects receive hourly blood draws for assessment of plasma cortisol, ACTH, norepinephrine, epinephrine, and cytokines (interleukin (IL)-1, IL-6, tumor necrosis factor (TNF)- α, and soluble receptors for TNF-α and IL-6). Subjects also receive a lumbar puncture following 12 weeks of treatment with IFN-α. The lumbar puncture assesses cerebrospinal fluid concentrations of corticotropin-releasing hormone (CRH), ACTH, cortisol, norepinephrine, epinephrine, and proinflammatory cytokines (i.e. IL-1, IL-6, TNF-α).
Eleven patients have been enrolled with a target enrollment of 50 subjects. Findings from this study will be integrated with results of a similar analysis of these functions in CFS patients from the Wichita clinical study and with data from a study of post-infectious fatigue in Australia. We are measuring gene expression profiles in these three studies so the data can be compared and integrated to define relevant common pathways requisite to the expression of fatigue and CFS.
Emory Collaboration - Division of Endocrinology, Department of Medicine
Proinflammatory cytokines, such as interleukin-6 (IL-6), have been implicated in the pathophysiology of CFS. Administration of IL-6 to human subjects causes fatigue and malaise, fever, headaches, and somnolence. IL-6 participates in the pathogenesis of adrenal insufficiency, a condition that closely resembles CFS. Adipose tissue is a major source of IL-6 and other circulating cytokines. Due to their illness, many CFS patients lead a sedentary life and are over weight, yet the metabolic status of the CFS patients has not been formally addressed. To investigate the role of adipose tissue in the pathophysiology of CFS, CDC is collaborating with and supporting laboratory studies by Emory University's Department of Medicine Division of Endocrinology.
Corticotropin-Releasing Hormone and Adipose Tissue Cytokine Production as a Model for CFS.
This study's objective is to investigate the role of IL-6, produced by adipocytes, in the pathogenesis of CFS. There are three specific aims:
- Identify the optimal dose and duration of CRH infusion.
- Investigate the role of endogenous IL-6 in the pathophysiology of CFS.
- Develop a diagnostic test for CFS using CRH administration.
The study is enrolling normal Caucasian female volunteers between 30 and 50 years of age, with a body mass index between 27 - 30, who are menstruating, and taking no medications. The study is randomized, double-blinded, and placebo controlled. Groups of 5 subjects receive 24-hour infusions of placebo or ovine CRH. Study parameters include: circadian rhythms of plasma cortisol, adrenocorticotropin hormone (ACTH), IL-6, TNF-α, IL-6 soluble receptor (IL-6sR), TNF-α soluble receptor 1 (TNFsR1), CFS symptoms, cognitive function, sleep quality, and ex vivo cytokine production by adipose cells. Because plasma concentrations of ACTH, cortisol, and cytokines exhibit a circadian rhythm, hourly blood samples for measurement of the above hormones and cytokines are obtained before and during CRH administration for a total of 48 h. Subjects also complete a Symptom Severity Scale hourly for 48 h (except while sleeping), before and during CRH administration. Effects of CRH on cognitive function are assessed using the Cambridge Neuropsychological Test Automated Battery (CANTAB) administered twice (before and during CRH administration). Sleep quality is assessed using a self-administered Visual Analog Scale for Sleep, completed upon awakening, on two consecutive days, before and after CRH administration. To assess the effects of CRH on local production of IL-6 by fat, we obtain baseline and post-CRH infusion fat samples from the subcutaneous paraumbilical region. Samples (500-1000 mg) are cultured for 24 h followed by measurement of TNF-α, IL-6, and glucose uptake. Eighteen patients have been enrolled thus far.
Post-infection fatigue - Collaboration with the University of New South Wales, Australia
Fatigue, cognitive disability, and myoarthralgia (i.e., CFS symptoms) are common during the acute phase of many infectious diseases, and appear to reflect the normal immune response. Cohort studies suggest that at least 10% of individuals experience disabling fatigue, accompanied by constitutional symptoms, for six months or more after acute infections due to Coxiella burnetii (Q fever), Epstein-Barr virus (EBV; infectious mononucleosis) or Ross River virus (RRV; epidemic polyarthritis). Such post-infective fatigue states represent a subset of the more heterogeneous group of disorders labeled as CFS. A range of host, microbial, and environmental variables has been identified as possible risk factors for post-infectious fatigue. It is unclear how they might interact to produce post-infectious fatigue or whether different factors contribute to the onset, persistence, and level of disability associated with post-infectious fatigue. The objective of this study is to define the risk factors and pathophysiological processes that result in post-infectious fatigue.
We support the project through a Cooperative Agreement with the University of New South Wales (UNSW), Sydney Australia. The UNSW uses hospital-based surveillance to identify, recruit, and follow three post-infection cohorts (EBV, RRV, and Q fever) where the enrolled subjects are in the acute phase of the illness as evidenced by IgM positive antibody response results to the infectious agent. They use standardized instruments to measure fatigue, psychological distress and psychiatric morbidity and are conducting laboratory studies of immune response. CDC has responsibility for gene expression and proteomics studies on the cohort. The post-infective fatigue cohorts will serve as model systems to identify measurable characteristics of gene expression and proteomics and correlate this with immune, endocrine and neurocognitive characteristics that differentiate and stratify the various categories of chronic fatigue from non-fatigued individuals. We will also be able to evaluate the similarities and differences of host gene expression between the three post-infection cohorts and compare to the gene expression patterns defined for CFS in other studies.
To date, 144 subjects have been enrolled (44 with Q fever, 33 with EBV, and 62 with RRV). Their average age is 33 years, and 60% are men (particularly reflecting the occurrence of Q fever in male meat-workers and farmers). All have remained under follow-up and 11 (8%) have remained fatigued =6 months, and also meet diagnostic criteria for CFS. This year we hosted a pre-doctoral scientist from UNSW who conducted gene expression analysis of 3,800 genes from 4 individuals who had not yet recovered 12 months post acute EBV infection and 8 HLA-matched controls who had recovered within 3 months of acute EBV infection. Each of these individuals has serial samples and the longitudinal gene expression changes and the point prevalent changes are being analyzed. Preliminary results indicate that individuals who fail to recover from acute EBV and are considered post-infectious fatigue cases have different longitudinal gene expression patterns than recovered individuals. Gene expression analysis phase of the EBV cohort samples is ongoing. We are also completing the laboratory component of the gene expression analysis on individuals with acute Q fever compared to those with QCFS.
Allergy and Exercise Challenge of CFS and normal subjects - National Jewish Medical Center (Denver) Collabroation
Patients with CFS have exacerbated symptoms after physical activity or exposure to allergens. Both allergens and exercise also induce a variety of changes in physiologic function mimicking an infectious or inflammatory state. This study serves as a model system of CFS where patients and healthy subjects were challenged with an exercise or allergen stress. Blood samples were drawn before, 6 hours and 24 hours after challenge. Microarrays were used to simultaneously assess the expression of 3,800 genes. We anticipate that by evaluating response to a physiological challenge we can better understand some of the pathways underlying the symptomatology observed in individuals with CFS. Since we know that age, sex and race can influence gene expression, we are limiting this study to a pre- and post-challenge sample from similarly aged, Caucasian women in the following categories. The laboratory component for expression profiling of the pre-challenge and 24 hour post-challenge samples has been completed. These results are currently being analyzed to determine if there is differential expression between cases and controls associated with response to challenge.
The molecular epidemiology component of the CFS program allows us to apply rapidly evolving cutting-edge genomics, proteomics, and bioinformatics technology to epidemiologic studies whose objective is CFS prevention and control. All components of the CFS program (surveillance, case definition, and clinical studies) include a molecular epidemiology effort. For presentation, we will discuss genomics, proteomics, and bioinformatics separately below. Genomics studies of post-infectious fatigue and exercise induced CFS were discussed above.
Gene Expression Profiling with Microarrays
Gene expression profiling examines the activity (transcription) of genes by analyzing cell's messenger RNA (mRNA). Gene expression is altered by many factors including cell differentiation, metabolic state, and disease status. By comparing expression of many genes simultaneously, gene expression profiles can be compared between persons (and populations) with and without disease and characteristic differences (or associations) can be identified. These differences, known as differential gene expression, can point to markers for diagnosis or altered physiologic pathways to target for prevention or therapy.
As summarized in the discussion of clinical studies, the pathophysiology of CFS is thought to involve dysregulation of the HPA axis. Gene expression analysis is particularly suited to studying CFS because it examines RNA transcripts from thousands of genes at the same time, covering all from immune system proteins, neuroendocrine pathways, stress response proteins, metabolism, apoptosis, cell adhesion receptors, cell cycles and many more. Microarrays simultaneously measure thousands of expressed genes, and such data is essential for understanding biological pathways and processes. Most gene expression studies have focused on samples derived from cells or tissues with a known lesion, but CFS has no known to sample. We have hypothesized that peripheral blood mononuclear cells (PBMCs) reflect the systemic state, thus allowing for evaluation of multiple pathophysiologic pathways. Some of these pathways should be affected by CFS.
Using a case control design, we have shown that microarray profiling of PBMC gene expression can distinguish CFS from non-fatigued subjects and have published a manuscript. We are currently using this technology to screen for biological correlates of CFS in population-based studies and model systems. Some were discussed above and the remainder are reviewed below.
Analysis of symptom patterns in CFS using gene expression methodologies
Although gene expression patterns can distinguish cases from controls this does not address possible causal associations. CFS is difficult to study using classic case control designs because incident disease is uncommon and hard to identify. Prevalent cases in most studies have been ill an average of five years so are enrolled into studies many years after causal associations occurred. We are focusing our current gene expression profiling effort to detect markers of disease activity, elucidate physiologic pathways, and clarify factors associated with sustaining illness. This will permit precise identification of cases for case control studies and will point to specific hypotheses.
Persons with CFS can be subgrouped according to whether illness begins acutely or gradually. In addition to different onsets, persons with CFS in Wichita appear to fall into two distinct groups based on symptomatology. We have examined PBMC gene expression patterns of 3,800 genes on 26 women classified as CFS in the Wichita surveillance program. Seven, all with gradual onset illness, fell into one symptom group. The remaining 19, 13 with a gradual onset and 6 with a sudden onset illness, were in the second group. Testing is complete and we are analyzing the data using the CDC MADB bioinformatics program (described below). We are comparing expression levels between different definition groups on a gene-by-gene basis to identify biologically relevant patterns. The analysis groups include:
- symptom groups
- onset type (e.g., gradual versus sudden)
- duration of CFS (e.g., less than 7 years or greater than 10 years
- number and type of symptoms
- blinded unsupervised cluster analysis to see if gene expression patterns identify "gene" groups independent of symptoms.
Longitudinal Analysis of Gene Expression in Subjects with CFS from the Wichita Surveillance Study
The objective of this study is to follow a well-characterized cohort of people with CFS over a four-year period and determine if unique gene expression profiles are associated with symptom occurrence or persistence.
Testing will include subjects identified in the Wichita surveillance study classified as:
- Meeting the CFS research case definition at each of the time points.
- Meeting the CFS definition and having the fatigue resolve at some point in the 4 year period.
- Having fluctuations between CFS and being fatigued but not meeting the case definition because of insufficient number of symptoms or fatigue severity.
- Non-fatigued controls.
To extend our efforts in gene expression analysis, we have initiated a program to study protein expression profiles of CFS patients. While mRNA levels (transcription) represent the genetic potential of a cell, most phenotypic attributes of biological systems stem from protein expression (translation). Therefore, a natural progression for analysis of CFS samples is to measure both protein patterns and gene expression profiles. Our major effort utilizes mass Spectrometry (MS) to profile the total protein composition of blood to define protein patterns that discriminate CFS cases from controls and ultimately to define specific protein biomarkers.
We utilize surface-enhanced laser desorption ionization (SELDI) spectrometry, which measures the mass/charge (m/z) of each protein in a mixture by the actual time it takes for the ionized protein to travel (fly) across a detection tube and strike a detector. SELDI-TOF utilizes a retention chromatography chip to capture specific families of proteins from a crude sample. This surface enhancement facilitates fractionation of complex samples to allow detection of specific classes of proteins.
SELDI-TOF has three major components: a ProteinChip Array, a reader, and software. The ProteinChip array is a stainless steel or aluminum-based solid supports with defined bait regions containing different surface binding chemistries including hydrophobic, normal-phase, metal-affinity and cationic or anionic. Samples (cell lysates or body fluids) are directly applied to the ProteinChip array at volumes as low as 0.5ul. ProteinChips are washed and then analyzed. Molecules are recorded as mass signature - peak- as they are ionized and are displayed as a standard chromatograph. The output data includes streams of data comprised of tens of thousands of data points over a wide dynamic range (1000 to 300000 Daltons), however, most observed differences of differential value have been observed between 200 and 20,000 Daltons.
In addition, we pursue more target studies directed at cytokine profiling studies of serum samples. These efforts use a Luminex bead-based Multiple-Analyte Profiling (MAP) assay to detect cytokines involved in pathways of immune function that may play a role in fatiguing illnesses. The "cytokine hypothesis" of the pathophysiology of CFS has not been systematically evaluated in well-defined cohorts or in prospective studies. The preliminary studies in this project are examining the role of pro-inflammatory cytokines in the development of post-infective fatigue. However, the unique design of this cohort study, and the established protocols for cytokine detection will allow examination of other relevant hypotheses, such as the relative production of type 1 (IL-2, IFN-α , IL-12) and type 2 (IL-4, IL-5, 1L-10) cytokine profiles, during acute and convalescent Q fever, EBV or RRV infections.
The objective of our proteomic effort in the Molecular Epidemiology Activity is to use MAP and SELDI-TOF to generate and define serum protein profiles of CFS. The goal is the discovery of a serum biomarker(s) of this illness. Combined with immunological and gene expression information, these analyses may establish a novel relational database for pursuing treatment and prevention of CFS in the future. Our hypothesis is that different molecular patterns can be identified in serum samples from subjects with CFS compared versus controls. Ours specific aims for the proteomic effort are:
- To explore the use of modern protein analysis technologies to identify biomarkers of CFS or CFS like patient populations based on empiric experimentation.
- To identify and to report biomarkers for CFS discovered using MAP and SELDI-TOF mass spectrometry and to evaluate existing as well as to provide unique methods of data analysis through exploration of available informatics technologies and the development of novel tools for such analysis.
- To provide validation of analysis and technology methods regarding CFS biomarker identification.
Microarray and proteomics data presents unique challenges for data processing and analysis. Interpretation of results in the context of epidemiologic and clinical studies adds another level of complexity. The rapidly evolving discipline of bioinformatics addresses these issues. The CFS bioinformatics effort has developed a microarray database (MADB), for which we have received a provisional patent. CDC's MADB allows analysis of microarray gene expression data in the context of epidemiology. CDC MADB is a Sybase relational database, (SYBASE Advanced Server Enterprise 12) one of the most powerful relational database systems that allows web interfacing.
Currently, CDC MADB uses data from microarray gene expression and SELDI-TOF experiments and provides several analytical and statistical tools for analysis. CDC MADB has dynamic interaction with supporting data from public repositories such as GenBank, LocusLink, GeneCards, and UniGene. Finally, CDC MADB includes statistical and mathematical analysis capability. These include scatter plotting for determining correlation and outliers, multidimensional scaling and clustering. The clustering programs, used for pattern discovery, perform standard agglomerative hierarchical clustering, k-means clustering, and self-organizing maps. CBEL also provides the ability to export results to external programs (Excel, S-plus, MATLAB) for further analysis.
Educational activities comprise the fifth component of CDC's CFS program. Our major priority is to inform health care professionals, state and local public health officials, health maintenance organizations, and insurance professionals concerning CFS. Education of patients and the public, although important, is a lower priority. Activities involve supporting formal education of health care professionals, providing technical liaison, publishing research results in peer-reviewed medical journals, helping to organize and support scientific/medical conferences and workshops on CFS, preparing and distributing printed materials for patients and health care professionals, and the CDC CFS world wide web home page and telephone information lines. In fiscal years 2001-2002, CDC cosponsored an NIH CFS State of the Science Meeting, cosponsored the American Association for CFS international conference, sponsored two workshops of the International CFS Study Group, and cosponsored three CFS Assessment Symposia Series workshops with the CIFDS Association of America. Symposia addressed the role of the autonomic nervous system in the pathophysiology of CFS, neuroendocrine aspects of CFS, and immunological aspects of CFS. This year CDC cosponsored a Banbury Center/Cold Spring Harbor Laboratory Conference (Towards Understanding of Cellular and Molecular Mechanisms of Unexplained Fatigue) and the American Association for CFS International Conference.
Health Care Professional Education
In FY 2000 CDC provided funding to the Health Resources and Services Administration (HRSA) to support training of primary health care professionals in the diagnosis and management of CFS. Throughout 2001, these funds supported a cooperative effort (Chronic Fatigue Syndrome: A Diagnostic and Management Challenge) with HRSA, the Illinois Area Health Education Centers (AHEC) Program, and the CFIDS Association of America. The project's objective was to increase health professionals' awareness and understanding of CFS. A group of CFS experts (including representatives from CDC) collaborated with the CFIDS Association of America to develop educational materials designed to improve the ability of primary care professionals to detect, diagnose, and manage CFS. HRSA was unable to continue this effort and in 2002 CDC awarded a 5-year competitive contract to the CFIDS Association of America to continue direct training, develop self-study modules, and evaluate the effectiveness of these methods to improve health professional knowledge about CFS.
In 2001, 3 Train-the-Trainer sessions were held and enrolled 52 trainers, who were taught methods of peer-to-peer continuing education. The 52 "core trainers" agreed to return home and educate peers or students in health care disciplines about CFS. They have given 28 presentations to date, and more training sessions are scheduled for 2003. A monthly newsletter helps maintain regular contact with core trainers. Supporting services, such as help with program planning, securing continuing education credits and preparing handouts are provided to core trainers to facilitate successful seminars. Refresher courses for veteran core trainers and additional training sessions for new recruits are planned for 2003.
Three self-study formats have been developed to reach a greater number of professionals, particularly rural professionals and those working with underserved populations who may have fewer opportunities to attend continuing education programs:
- A print-based self-study course was implemented in January 2002.
- Extensive editing and additions to a videotaped course (produced in 2001) were completed in November 2002. Distribution of the two-tape set and accompanying written resources will begin in January 2003.
- A Web-based self-study went "live" on December 20, 2002 and links to this learning module are on the CDC Web site's chronic fatigue syndrome home page. CDC's Division of Professional Development and Evaluation has awarded continuing education credits for each format, an important incentive for health care professionals to complete the course.
The modular display developed with pilot funds has attracted hundreds of medical professionals at 6 national professional conferences. Visitors to the display demonstrated interest in helping their CFS patients by learning more about symptom management. Nearly all expressed frustration in successfully managing the illness. Participation in these conferences also generates interest in the self-study modules and train-the-trainer workshops and presents opportunities to network with others who interact with medical professionals, including pharmaceutical representatives, continuing education accreditation staff and public health educators.
Four major conferences accepted proposals to present the CFS program in 2002:
- Nancy Klimas, M.D., University of Miami, presented at the American Medical Women's Association's Annual Meeting in San Antonio in February.
- Pilot project co-directors, Kim Kenney and Richard Wansley, presented to the National Area Health Education Centers (AHEC) Organization in Little Rock, Arkansas in April.
- Constance Van der Eb, PhD, DePaul University and Kathleen Gale, RN, presented to over 300 Chicago school nurses annual back-to-school conference in August.
- Patricia Major, M.D., University of Miami, lectured at the Nurse Practitioner Associates for Continuing Education (NPACE) in Boston in November.