Date: Wed, 16 Feb 2000 22:35:29 -0700

~~~ Thanks Kathy Nye ~~~

Authors: R Hugh Dunstan

Authors: Neil R McGregor

Authors: Henry L Butt

Authors: Timothy K Roberts

A diagnosis of chronic fatigue syndrome (CFS) requires the exclusion of other known fatigue-related diseases and requires compliance with a clinical definition. The patient set derived by this process is heterogeneous in its polysymptomatic presentation and has proved very difficult to study clinically and scientifically.

Copyright Carfax Publishing Company Jun 1999

A diagnosis of chronic fatigue syndrome (CFS) requires the exclusion of other known fatigue-related diseases and requires compliance with a clinical definition. The patient set derived by this process is heterogeneous in its polysymptomatic presentation and has proved very difficult to study clinically and scientifically. Our laboratory has been involved in research to evaluate changes in biochemical and physiological homeostasis which might occur in CFS patients. The studies have included investigating urinary excretion, blood lipids, immunology, organochlorine pesticides and microbiology.

All these research projects indicated that the CFS patients have multiple anomalies in their homeostasis. The multivariate cluster analyses of these data sets indicated that different types of CFS patient could be characterized on the basis of similarities in metabolite profiles. The measured changes in metabolite composition were strongly associated with symptom incidence and severity, suggesting that these subtle patterns of metabolism have a strong influence on CFS symptom presentation.

Keywords: chronic fatigue syndrome, chronic pain, myalgia, pathology, pesticides, organochlorines, homeostasis, essential fatty acids, proteolysis, staphylococci.

Outbreaks of a mysterious disease with an apparently infective onset, currently called chronic fatigue syndrome (CFS) or myalgic encephalomyelitis (ME), have occurred on numerous occasions since the 1934 outbreak in Los Angeles

[1] . CFS is currently diagnosed by

(1) compliance with a clinical definition, and

(2) exclusion of other known fatigue-related diseases [2-3].

The American Centers for Disease Control (CDC Atlanta) have devised diagnostic criteria to allow a standardized selection of patients for research studies [2]. These criteria have since been modified to be more applicable to general practice [3] and are currently accepted as the basic clinical tool for the diagnosis of CFS.

No aetiological agent, consistent cellular or biochemical alteration has been found which could be used to differentiate the condition from similar fatigue-related diseases. The lack of objective biochemical and/or cellular diagnostic markers has resulted in considerable confusion in diagnosing the syndrome and hampered scientific investigation of its aetiology and pathophysiology.

The primary symptomatic features of CFS, including fatigue and pain, represent common host responses and can be initiated by many different mechanisms. The CFS patient groups defined by the CDC criteria may therefore actually represent a number of different polysymptomatic illnesses with separate aetiologies, which are artificially clustered under the umbrella of CFS. This phenomenon would result in confounding scientific investigations of CFS, and would explain why no single causal factor could be identified for all CFS patients. Thus CFS, as it is currently defined, has been suggested as resulting from a common host response, particularly involving immune and neural function, to a heterogeneous group of aetiological agents.

Effective design of management protocols can only be conceived after patient sets have been sorted into relatively homogeneous subsets, and the pathophysiologies of these subsets have been properly defined. The differential anomaliesmight include combinations of underlying genetic, dietary, immunological, environmental and/or pathogen-altered responses. Classification of CFS patients into their respective subgroups would enable meaningful trials to be conducted to evaluate management/treatment schemes derived from empirical data.

Chronic fatigue syndrome patients have a polysymptomatic symptom presentation which can often be difficult to characterize and manage. The use of (questionnaires to evaluate symptom incidence and severity can enable the construction of symptom indices to assist in the evaluation of patients [4]. There are several types of symptom profile which can bc observed in the CFS patients. Three simplistic groups can be defined on the basis of those having primarily (1) muscle pain, (2) fatigue and (3) pain and fatigue. A review of data compiled from a number of studies in our laboratory indicates that these groups of patients can be separated from each other and control subjects by discriminant function analysis of their corresponding symptom profiles (Fig. 1). These analyses provide strong evidence that those patients with both muscle pain and fatigue have a different symptom profile compared with those patients presenting with only fatigue. These separable groups may represent different disease entities.

Analyses of the CFS biochemical data by cluster analysis and multivariate techniques indicated that different types of CFS patient could be characterized on the basis of similarities in the metabolite profiles [5]. The urine profiles could be used to classify patients with similar excretion patterns to give relatively homogeneous sets of patients for investigation. Individual patients can be aligned with the metabolic profiles from our CFS research to assign their altered homeostasis to a particular CFS type.

The processes used to classify the patients involved assessing multiple aspects of the urine excretion pattern. The groups so far delineated include urinary marker (UM) patterns with either CFSUMI, fl-alanine or UMISb as the major feature of the profile category.

Discriminant function analysis is a technique which can be used to assess the multivariate profile data and compare group differences in their excretion patterns. The net result is the formation of a discriminant function model which can be used to generate scatter plots to graphically represent the differences between patient and control sets as shown in Fig. 2. This same model can then be used to align individual excretion profiles with those of the study reference data and thereby provide a classification of the patient on the basis of an empirical data set.

Three symptom clusters were characterized by multivariate clustering techniques, and these were coded according to the predominant features of pain in each cluster: (Type 1 ) localized pain, (Type 2) pain with gut dysfunction and (Type 3) generalized pain. The urine excretion patterns for members of each cluster were then assessed to determine whether urine excretion patterns could differentiate between the different clusters. Fig. 3 shows the remarkably clear separation of symptom groups and controls by discriminant function analysis of the urine excretion profiles. These data suggest that an important relationship exists between changes in homeostasis and symptom presentation.

In a similar manner, the plasma fatty acids profile can be used to Differentiate CFS patients from control subjects [6, 7]. Analyses of these lipid data by cluster analysis and multivariate techniques also indicated that

(1) the controls represented a homogeneous set of subjects,

(2) the sudden onset CFS subjects could be differentiated from the radual onset CFS

[IMAGE GRAPH] Captioned as: FIG. 1.

[IMAGE GRAPH] Captioned as: FIG. 2 patients [7],

(3) CFS subjects with Epstein Barr virus (EBV) early antigen could be differentiated from the other CFS patients [8],

(4) CFS subjects could be divided into five well-resolved subgroups.

These results indicated that the lipid metabolic profile represents a potentially sensitive and powerful tool for assessing host responses and altered homeostasis in chronic fatigue disorders. The resolving power of the discriminant functionapproach is shown in Fig. 4 and Fig. 5, where the canonical plots show well-resolved scatters of members of the various CFS subgroups compared with the control subjects.

Our laboratory has been involved in the development of gas chromatography mass spectrometry techniques lo measure a wide range of metabolites in body fluid samples, including urine and plasma. The technique for urine analysis assesses the levels of amino acids and organic acids. In addition, the technique can identify unusual metabolites which cannot be identified from the mass spectral data, and allow investigation of the association of these products with symptom presentation.

[IMAGE GRAPH] Captioned as: FIG. 3.

[IMAGE GRAPH] Captioned as: FIG 4 Captioned as: FIG 5

Analyses of the urinary excretion of metabolites in CDC-defined CFS patients and control subjects have indicated that CFS patients have a significantly altered pattern of excretion, suggestive of an altered homeostasis [9]. An increase in tyrosine output was observed which is an indication of an active non-fibrillar proteolytic response in the CFS patients [10-12]. Increases in an unusual marker, coded CFSUMI, and fi-alanine as well as decreases in serine and alanine were also observed.

Elevations in the excretion of CFSUM1 and l-alanine were positively associated with increased symptom incidence (see Fig. 6) and severity. A range of symptom indices was constructed to further evaluate the relationships between the urine anomalies and the symptom presentation. The symptom indices were evaluated for associations with changes in the urine profiles by multiple regression analysis. The results indicated that each symptom index was strongly associated with changes in the urine excretion profile and the primary correlates are summarized for each index in Table 1. Elevations in CFSUMI and beta-alanine were positively associated with the reporting of CFS core symptoms and the total numbers of symptom respectively. Conversely, reduced output of serine was associated with increasing severity of neurological and pain symptoms.

Psychological symptoms were also assessed in this study, and somatization was the most important factor differentiating CFS and control subjects [13] Depression and anxiety were not important inter-group determinants. The somatization score was strongly associated with the excretion of CFSUMI, which is again consistent with the physical symptom reporting summarized in Table 1. Depression was associated with the excretion of UM15, and was independent of the excretion of CFSUM1. These results indicated that the "psychological" phenomena were associated with alterations in biochemical homeostasis in the host.

These data therefore provide evidence of a relationship between alterations in homeostasis and symptom expression. Although these do not necessarily indicate a cause and effect situation, they do establish associations with biochemical changes that require further investigation, and currently support an organic basis to CFS.

The alterations in amino and organic acids in CFS patients may be the result of dysregulation of protein turnover where there is a persistent increase in proteolysis over protein synthesis. This enhanced proteolysis is a mechanism to increase the availability of amino acids to the body following trauma, infection or highly stressful situations and can result from interferon-associated increased protein degradation. This mechanism has also been described in patients with certain genetic and acquired disease states, and has been observed in the muscle wasting associated with late-stage cancer.

In muscle, enhanced proteolysis can be related to non-fibrillar (short-term protein) and fibrillar (long-term protein-actin, myosin) mechanisms. Non-fibrillar proteolysis is associated with increased release of tyrosine from the cytoplasmic protein pool, whereas fibrillar proteolysis is associated with increased release of the unique actin component 3-methylhistidine from muscle fibers. Although the detection of proteolysis in patients is nonspecific to the aetiology, it can provide important data for the clinician to consider in designing patient management programs. The amino acid leucine controls non-fibrillar proteolysis in muscle and adipose tissue and has a distinct control over RNA degradation [14]. Alterations in leucine homeostasis may lead to dysregulated and sustained "shortterm" proteolysis in CFS patients.

Solvents and organophosphate pesticides are rapidly excreted by the body and corresponding serum levels can therefore only be validly estimated immediately after exposure [15]. It is possible, however, to measure persistent organochlorine molecules such as hexachlorobenzene (HCB) and dichlorodiphenyl-trichloroethane (DDT, or its metabolic product l,1-dichloro-2,2-bis (p-chlorophenyl) ethene, DDE) in serum or biopsy samples. DDT and HCB are chlorinated hydrocarbon pesticides which are an extremely stable group of lipophilic compounds and have been utilized extensively around the world for pest control since the early 1940s [16]. Recent findings in our laboratory [17] have indicated that CFS patients can have significantly higher levels of serum DDT compared with control subjects. In addition, DDE and HCB levels correlated with changes in blood cell parameters such as red cell distribution width (an index of the variation of red cell morphology, or anisocytosis: a standard pathology item in the measurement of blood cell parameters) and haemoglobin content. Exploration of the early literature revealed interesting associations of pesticide usage and viral activities, as well as an association of DDT with substantial alterations in metabolic homeostasis (Table 2).

Symptoms developed from prolonged low-level exposure to organochlorine pesticides may be gradual in onset and may not necessarily be associated with chemical insults. However, insecticide workers have been reported to have an elevated relative risk of mental disorders including neurotic, depressive and sleep disorders and an acute reaction to stress [18]. Other reported effects of organochlorine exposures include an elevated relative risk of breast cancer [19]; impairment of immune function [20]; development of endometriosis [21 ]; significant increases in chromosome aberrations [22]; decreases in male fertility; decreases in the frequency of live births; and increases in still births, neonatal deaths and congenital defects in the offspring of pesticide-exposed males [23].

The results of our studies suggested that low-level exposures to organochlorines could be associated with subtle, yet significant, changes in blood cell parameters which fall within the "normal" clinical laboratory range. The associations between HCB and the blood cell parameters were different between the fatigued group and the control group, suggesting substantial alterations in the functional capacity of the blood cells to respond to external stimuli. HCB was positively associated with the CD4:CD8 ratio and neutrophil count in the CFS patients. Although the subjects in this study had had no clinical evidence of overt infection, these immune responses tend toward those seen for bacterial infections rather than viral infections. It has been previously reported that sub-lethal levels of DDT and HCB caused a significant decrease in the capacity for bovine macrophages and neutrophils to phagocytose the standard Staphylococcus aureus strain 305 [24]. This finding has further significance in regard to our recent microbiology research where 89% of patients with chronic muscle pain had multiple carriage of strains of coagulase negative staphylococci (CNS) which produced a membrane damaging toxin [25], and the controls had none.

The extensive range of fatigue and pain symptoms reported by CFS patients, combined with changes observed in monoamine metabolism and in the urinary excretion of metabolites [26], suggest substantial systemic alterations in metabolism and homeostasis. The symptom profiles inclusive of low-grade fever, lymphodynia and myalgia, are suggestive of a pathogenic challenge. The results of a recent study with chronic muscle pain patients [25] (as well as a separate study with CFS patients-unpublished data) have implicated the role of toxic coagulase negative staphylococci (T-CNS) in the aetiology of chronic pain and may further constitute an important component in the development of CFS.

[IMAGE TABLE] Captioned as: TABLE 1

[IMAGE TABLE] Captioned as: TABLE 2.

The subjects in this study had had no clinical evidence of overt infection, but the chronic muscle pain patients had 23 versus 9 isolates/10 subjects compared with the controls. There was also a higher incidence of carriage of two or more CNS strains in the muscle pain patients as shown in Fig. 7. The muscle pain subjects carried strains of CNS which produced delta-like membrane damaging toxins, whereas control subjects did not carry toxin-producing strains as shown in Fig. 8. The carriage of membrane-damaging toxins was strongly correlated with increases in symptom severity, including palpitations, urinary frequency, fatigue, pain and weakness. These chronic muscle pain patients also had changes in urinary excretion patterns which suggested an altered homeostasis. The major urine excretion features included an increase in tyrosine output suggestive of an active proteolysis event and a reduction in leucine suggestive of possible disruption of the proteolytic control.

The urine and plasma lipid profiles can be used to classify patients with similar metabolic profiles to give relatively homogeneous sets of patients for investigation. As well as having more closely aligned lipid and/or urine anomalies, these sets also have less diverse symptom presentation profiles. Investigation of the pathophysiology of these types of patient set is more likely to yield results that lead to an understanding of the underlying genetic, dietary, immunological, environmental and/or pathogen altered responses. A better understanding of the aetiologies of these subgroups of CFS would then allow the development and trial of therapeutic interventions based on empirical data.

Individual patients can be aligned with the metabolic profiles from our CFS research to assign their altered homeostasis patterns to a particular CFS type, and the list of associated symptoms can be checked by the clinician. Patient management can then be initiated based on the most up-to-date understanding of the various pathophysiologies available for each type of chronic fatigue disorder.

[IMAGE GRAPH] Captioned as: FIG 7

Captioned as: FIG. 8

* It is important to appreciate that the pesticide formulations used for various applications usually contain less than 1% of the active ingredient. Most of the pesticides are lipophilic to facilitate entry into target organisms and they require delivery in appropriate solvent mixtures to ensure that they are dissolved and they usually also require the presence of dispersal agents. The formulations can therefore contain a complex cocktail of solvents and additives. These solvents can have substantial effects on exposed organisms, either by causing oxidative damage or by membrane disruption. They are rapidly excreted leaving no trace, but sometimes leaving considerable cellular damage.

[1] Wilson JC, Walker PJ. Acute anterior poliomyelitis: orthopaedic aspects of the California outbreak of 1934. Arch Intern Med 1936; 57: 477-91.

[2] Holmes GP, Kaplan JE, Gantz NM, et al. Chronic fatigue syndrome: a working case definition. Ann Intern Med 1988; 108: 387-9.

[3] Fukuda K, Straus SE, Hickie I, et aL The chronic fatigue syndrome: a comprehensive approach to its definition and study. Ann Intern Med 1994; 121: 953-9.

[4] McGregor NR, Dunstan RH, Zerbes M, Butt ItL, Roberts TK, Klineberg IJ. Preliminary determination of the association between symptom expression and urinary metabolites in subjects with chronic fatigue syndrome. Biochem Molec Med 1996; 58: 85-92.

[5] Dunstan RH, McGregor NR, Butt HL, Roberts TK, Zerbes M, Klineberg IJ. The development of laboratory tests for investigating chronic fatigue and pain disorders. The Clinical and Scientific Basis of Chronic Fatigue Syndrome: from Myth Towards Management. International Meeting for Clinicians and Scientists, TK Roberts ed. The University of Newcastle and combined CFS Consumer Groups, 1998.

[6] McGregor RN, Dunstan RH, Roberts TK, Donohoe M, Watkins JA, Butt HL, Murdoch RN. Classification of chronic fatigue syndrome patients by assessing plasma lipid homeostasis. The

Clinical and Scientific Basis of Chronic Fatigue Syndrome: from Myth Towards Management. International Meeting for Clinicians and Scientists, ed. TK Roberts. The University of Newcastle and Combined CFS Consumer Groups, 1998.

[7] McGregor RN, Dunstan RIl, Roberts TK, Donohoe M, Watkins JA, Butt HL, Murdoch RN. Assessment of lipid homeostasis in sudden and gradual onset chronic fatigue syndrome patients. The Clinical and Scientific Basis of Chronic Fatigue Syndrome: from Myth Towards Management. International Meeting for Clinicians and Scientists, ed. TK Roberts. The University of Newcastle and Combined CFS Consumer Groups, 1998.

[8] Dunstan RII, McGregor NR, Butt IlL, Roberts TK, Zerbes M, Klineberg IJ. Assessment of plasma lipid homeostasis in relationship to epstein barr virus antibody titres in patients reporting sudden onset chronic fatigue. The Clinical and Scientific Basis of Chronic Fatigue Syndrome: from Myth Towards Management. International Meeting for Clinicians and Scientists, ed. TK Roberts. The University of Newcastle and Combined CFS Consumer Groups,1998.

[9] McGregor NR, Dunstan RH, Zerbes M, Butt HL, Roberts TK, Klineberg IJ. Preliminary determination of a molecular basis to chronic fatigue syndrome. Biochem Molecul Med 1996; 57: 73-80.

[10] Mortimore GE, Poso AR. Intracellular protein catabolism and its control during nutrient deprivation and supply. Annu Rev Nutr 1987; 7: 539-64.

[11] Goodman MN. Interleukin-6 induces skeletal muscle protein breakdown in rats. Proc Soc Exp Biol Med 1994; 205: 182-5.

[12] McGregor NR, Dunstan RH, Butt HL, Roberts TK, Zerbes M, Klineberg IJ. Chronic Facial Muscle Pain and Dysregulated Cellular Proteolysis. The Clinical and Scientific Basis of Chronic Fatigue

Syndrome: from Myth Towards Management. International Meeting for Clinicians and Scientists, ed. TK Roberts. The University of Newcastle and Combined CFS Consumer Groups, 1998.

[13] McGregor NR, Dunstan RH, Butt HL, Zerbes M, Roberts TK, Klineberg IJ. A preliminary assessment of the association of SCL-90-R psychological inventory responses with changes in urinary metabolites in patients with chronic fatigue syndrome. J Chronic Fatigue Syndrome 1997; 3: 17-37.

[14] Balavoine S, Feldmann G, Lardeux B. Regulation of RNA degradation in cultured rat hepatocytes: effects of specific amino acids and insulin. J Cell Physiol 1993; 156: 56-62.

[15] Joy RM Chlorinated hydrocarbon pesticides, In: Pesticides and Neurological Diseases, 2nd edition, eds. DJ Ecobichon and RM Joy. CRC Press: London, 1993.

[16] Dunstan Rtl, Donohoe M, Taylor W, et al. A preliminary investigation of chlorinated hydrocarbons and chronic fatigue syndrome. Med J Aus 1995; 163: 294-7.

[171 Dunstan Rll, Roberts TK, Donohoe M, McGregor NR, Hope D, Taylor WG, Watkins JA, Murdoch RN, Butt 11. Bioaccumulated chlorinated hydrocarbons and red/white blood cell parameters. Biochem Molec Med 1996; 58: 77-84.

[ 18] de Jong G. Long term health effects of aldrin and dieldrin. Toxicol Lett, Supplement. Elsevier Science: Amsterdam, 1991.

[19] Wolfe MS, Toniolo PG, Lee EW, Rivera M, Dubin N. Blood levels of organochlorine residues and risk of breast cancer. J Nat Cancer Inst 1993; 85: 648-52.

[20] Neubert R, Jacob-Miller U, Helge lI, Stahlmann R, Neubert D. Polyhalogenated dibenzo-p-dioxins and dibenzofurans and the immune system. 2. In vitro effects of 2,3,7, 8-tetrachlorodibenzo-p-dioxin (TCDD) on lymphocytes of venous blood from man and a non-human primate (Callithrix jacchus). Arch Toxicol 1991; 65: 213-9.

211 Rier SE, Martin DC, Bowman RE, Dmowski WP, Becker JL. Endometriosis in Rhesus monkeys (Macaca mulatta) following exposure to 2,3,7, 8-tetrachlorodibenzo-p-dioxin. Fundam Appl Toxicol 1993; 21: 433 41.

[22] Kourakis A, Mouratidou M, Kokkinos G, Barbouti A, Kotsis A, Mourclatos D, Dozi-Vassiliades J. Frequencies of chromosomal aberrations in pesticide sprayers working in plastic green houses. Mutat Res 1992; 279: 145-8.

[23] Rupa DS, Reddy PP, Reddi OS. Reproductive performance in population exposed to pesticides in cotton fields in India. Environ Res 1991; 55: 1238.

[24] Sitarska E, Winnicka A, Klucinski W. Effect of organochlorine pesticides on the phagocytic activity of bovine milk cells. Zentralbl Veterinarmed A 1990; 37: 471-6.

[25] Butt IlL, Dunstan RES, McGregor NR, Roberts TK, Zerbes M, Klineberg IJ. An association of membrane damaging toxins from coagulase negative staphylococcus and chronic orofacial muscle pain. J Med Microbiol 1998; 47: 577-84.

[26] Demitrack MA, Gold PW, Dale JK, Krahn DD, Kling MA, Straus SE. Plasma and cerebrospinal fluid monoamine metabolism in patients with chronic fatigue syndrome: preliminary findings. Biol Psychiatry 1992; 32: 1065-77.

R. HUGH DUNSTAN DPHIL, NEIL R. McGREGOR MDSc, HENRY L. BUTT PhD AND TIMOTHY K. ROBERTS PhD

Collaborative Pain Research Unit, Department of Biological Sciences, The University of Newcastle, Callaghan, NSW 2308 Australia

Reproduced with permission of the copyright owner. Further reproduction or distribution is prohibited without permission.

End of Document