CFS

IMMUNOLOGICAL ABNORMALITIES IN PATIENTS WITH CHRONIC FATIGUE SYNDROME

Umberto Tirelli M.D.°, Giuseppe Marotta M.D.*, Salvatore Improta M.D.*, Antonio Pinto M.D.*

° CFS Unit, Division of Medical Oncology and AIDS, Centro di Riferimento
Oncologico (CRO), Aviano - Italy
* The Leukemia Unit, Division of Medical Oncology, C.R.O., Aviano - Italy

Running head:	CFS: clinical and immunological study
Corresponding author:	Prof. Umberto Tirelli Division of Medical Oncology and AIDS Centro di Riferimento Oncologico, IRCCS 33081 Aviano - Italy phone: Italy - 434 - 659284 fax: Italy - 434 - 652997
key words:	chronic fatigue syndrome, lymphocyte subsets, NK cells, flow cytometry

Between January 1991 and January 1993, 265 patients who fulfilled the CDC criteria of the working case definition of Chronic Fatigue Syndrome (CFS) have been observed at our Institution and submitted to clinical and laboratory evaluation. One hundred and sixty-three patients were females and 102 males, the median age was 35 years (range 4-55 years); all patients reported profound and prolonged fatigue, lasting for a median of 3 years (range 6 months - 10 years), preceded or accompanied at appearance by fever in 185 cases, and neuropsychologic problems including inability to concentrate, difficulty in thinking, confusion, irritability, forgetfulness, and depression. The fatigue was so severe to induce 102 patients to stop their working activities for a period of time ranging from 3 months to 2 years (range 7 months). In 40 consecutive patients a comprehensive immunologic testing by single and two-color flow cytometry was performed and results compared with a group of 35 healthy, age- and sex-matched controls. Whilst no significant differences were found in the absolute numbers of circulating total T cells (CD3+) and of total helper/inducer (CD4+) or suppressor/cytotoxic (CD8+) T cells, an evident reduction in CD3-/CD16+ and CD57+/CD56+ NK lymphocytes along with an expansion of the CD8+/CD56+ and CD16-/CD56+ NK subsets, were found in the CFS group. In addition, CD56+ NK cells from CFS subjects were found to express an increased amount of cell adhesion molecules (CD11b, CD11c, CD54) and activation antigens (CD38). Both the percentage and absolute numbers of CD4+ T cells bearing the CD45RA antigen appeared significantly reduced in CFS patients, and CD4+ T lymphocytes from CFS subjects displayed an increased expression of the intercellular adhesion molecule-1 (ICAM-1/CD54). Finally, the total numbers of circulating (CD19+) B lymphocytes, were significantly higher in CFS cases than in controls, and in 11 out of 30 CFS patients the increase in circulating B cells was sustained by the expansion of the CD5+/CD19+ subset of B lymphocytes. We conclude that CFS is a syndrome not previously described in Italy, with the already known clinical characteristics and appears associated with several immunologic abnormalities, including those reported previously in cohort of patients from different countries. We also show for the first time that CD56+ NK cell subsets from CFS patients display an abnormally increased expression of cell adhesion molecules and activation markers.

INTRODUCTION
Chronic Fatigue Syndrome (CFS) is a recently defined illness of unknown etiology, characterized by unexplained, disabling fatigue lasting more than six months, chronic and recurrent low-grade fever, adenopathy, pharyngitis, and neuropsychological symptoms such as difficulties with concentration and depression [1]. The Centers for Disease Controls (CDC) have produced a working case definition for such heterogeneous disease which relies on clinical and laboratory criteria [2]. A viral etiology for CFS has been suspected owing to the overlapping clinical features with post-viral fatigue, and a number of different viruses including Human herpes virus-6 (HHV-6) [3, 4], enteroviruses [5] and HTLV II-related retroviruses [6], have been implicated in the development of CFS.
Results on immunological abnormalities associated with CFS have been far more conflicting and include a decreased number and function of NK cells [7-9], the presence of chronically activated circulating T cells [10, 11], monocyte alterations [9], changes in B cells subsets [8, 12], and abnormalities in cytokine serum levels or in vitro response of lymphocytes to mitogenic stimulation [13-16]. More recently, an altered distribution of CD4+ T cells subsets bearing CD45RA (naive T cells) and CD45RO (memory T cells) antigens, along with the increased expression of cell adhesion molecules on CD4+/CD45R0+ T lymphocytes have been described in CFS patients by Straus and coworkers [17].
CFS has been mainly described in the U.S., Australia, and Great Britain, but also in Canada, New Zealand, Israel, Spain, and France [1] while it has never been described, to our knowledge, in Italy with the exception of a preliminary report from our group [18].
We describe here the results of a prospective clinical study on 265 patients with CFS observed in Italy, and of a detailed immunophenotypic study of circulating lymphocytes from 40 consecutive CFS subjects and 35 matched healthy controls.

METHODS
Patients selection. Between January 1991 and January 1993, 265 patients who fulfilled the CDC criteria of the working case definition of CFS [2] have been observed at our Institution and submitted to clinical and laboratory evaluation [18]. This large number of patients is explained by the fact that our group was the first in Italy to report cases of CFS and at least during 1991 we were the only referral center. Moreover, large coverage by the media on this apparently new syndrome in Italy was made.
All subjects had a longer than 6 months history of a new onset of debilitating fatigue and fullfilled 6 or more minor criteria (symptoms and physical findings), along with the exclusion of any alternative diagnosis for their symptoms. In particular, chronic psychiatric disease, such as endogenous depression, hysterical personality disorder, anxiety neurosis, chronic use of antidepressive medications were excluded as required by the case definition also by the use of the Minnesota Multiphasic Personality Inventory (MMPI) [19] and structured interview. In the same period of time, 1750 individuals were referred at our Institution with suspicion of CFS, but they were actually diagnosed as affected by other diseases. The diagnosis of depression was made in several patients by psychiatrists and neurologists at other institutions, while patients of this series were submitted to the MMPI [19]. From September 1991 to March 1992, 40 consecutive patients underwent extensive immunophenotypic analysis of peripheral blood lymphocytes. The reason of the selection of such limited group for extensive immunological analysis was mainly economic. However these 40 patients reported a debilitating fatigue of more than 6 months duration along with the other symptoms fulfilling the CDC criteria of the working case definition of CFS [2], therefore with characteristics superimposable to the overall group of 265 patients. As a control group, 35 age- and sex- matched healthy individuals were selected among blood donors and hospital staff. All control subjects underwent physical examination and a routine laboratory workup to assess their healthy condition. None of the control subjects had experienced a period of fatigue or unexplained malaise lasting more than one week, in the six months preceding immunological assessment. Both patients and control subjects were free from medications (with a special regard to steroids and non-steroideal antiflammatory drugs) for a minimum period of two months before blood sampling, as assessed by the attending and referring physicians.

Cell separation and immunophenotyping. Mononuclear cells were isolated by standard Ficoll-Hypaque centrifugation of citrate-dextrose anticoagulated peripheral blood. Specimens were processed within two hours from sampling. In general, two patient samples and at least one control specimen were tested toghether on the same day. For direct immunofluorescence analysis, cells (5.0 x 105) were first preincubated (30 min at 4°C) with Hank's balanced saline solution containing 10% rabbit serum and 0.01% sodium azide to prevent Fc receptor binding, and subsequently incubated (30 min at 4°C) with saturating concentrations of fluorescein- (FITC) and phycoerythrin- (PE) labeled monoclonal antibodies as described [20, 21]. Two-color direct imunofluorescence was performed as described [20, 21]. Monoclonal antibodies used along with their recognized cluster of differentiation (CD) and source were as follows: Leu 4/CD3-PE, Leu 3/CD4-PE, Leu 2/CD8-PE and -FITC, Leu 11a/CD16-FITC, Leu 12/CD19-FITC, anti-HLA-DR-FITC, Leu 19/CD56-PE, Leu 7/CD57-FITC (Becton-Dickinson, Mountain View, CA); MO1/CD11b-FITC, 4B4/CD29-FITC, 2H4/CD45RA-FITC (Coulter Immunology, Hialeah, FL); 84H10/CD54-FITC, UCHL.1/CD45RO-FITC, T16/CD38-FITC (Immunotech, Marseille, France); KB90/CD11c-FITC (Dakopatts A/S, Glostrup, Denmark). In 30 CFS patients and 23 healthy controls CD5+ B cells were also evaluated with a preconstituted CD5-PE/CD19-FITC antibody combination (Immunotech). The percentage of CD5+ B cells was expressed as a percentage of the total lymphocyte population in the analysis gate. Non specific binding of monoclonal antibodies was assessed by labeling cells with phycoerythrinated and fluoresceinated isotype-matched control mouse Igs (Becton-Dickinson). Viable, antibody-labeled lymphocytes were identified according to their forward and right angle scattering, electronically gated and analyzed for surface fluorescence on a FACScan flow cytometer (Becton Dickinson). Cells identified by the lymphocyte gating were checked with a anti-CD45-FITC+anti-CD14-PE antibody combination (LeucoGATE, Becton-Dickinson). Less than 1% of CD14+ monocytes were identified in the lymphocyte gate. Fluorescence data were collected and analyzed by a Lysis II software (Becton-Dickinson). Laser output was mantained constant and the fluorescence gain settings were adjusted daily to obtain the same fluorescence signals from calibrating microspheres (Polyscience Inc., Warrington, PA) to allow fluorescence intensity comparisons among different experiments. In 10 patient samples and 9 control specimens immunophenotypic characterization was performed on the same day with a whole blood lysis method [22], obtaining superimposable results, as also supported by previous studies [17].
The statistical significance of mean antibody reactivities and of absolute numbers of antigen positive lymphocytes comparisons among CFS and control groups were analyzed by the Mann-Whitney U test for non-gaussian distribution [23].

RESULTS
Clinical findings. One hundred and sixty-three patients were females and 102 males, the median age was 35 years (range 4-55 years); all patients reported profound and prolonged fatigue, lasting for a median of 3 years (range 6 months - 10 years), preceded or accompanied at appearance by fever in 185 cases, and neuropsychologic problems including inability to concentrate, difficulty in thinking, confusion, irritability, forgetfulness, and depression (see Table 1). The other symptoms detected were, among the others, in agreement with the CDC definition, muscle weakness in all patients, low grade fewer in 185 patients, sore throat in 190, painful laterocervical or axillary lymphnodes in 140, myalgia in 260, sleep disturbances, either insomnia or hypersomnia in 261, headaches in 260, migratory arthralgia in 245, photophobia and transient visual scotomata in 235 patients. Depression, diagnosed mainly at other institutions, was reported to appear several months after the onset of fatigue in 95 patients, being therefore rather considered reactive to a debilitating and unpredicatable disease. The fatigue was so severe to induce 102 patients to stop their working activities for a period of time ranging from 3 months to 2 years (range 7 months). There were three pairs of brothers, one mother and daughter and three children (11, 11 and 4 years old) in this case series. Laboratory examinations, in particular complete blood count, erythrocyte sedimentation rates, standard serum chemistry test, testing for collagen vascular diseases and thyroid function test, were normal. Antibodies to HIV were negative in the patients tested. On clinical examination, non essudative pharyngitis was observed in 155 patients, while in 65 patients lymph nodes were increased in size without any suspicious characteristics for malignancies.
During the 2 year period of the study, 8 patients experienced the complete disappearance of symptoms, and 22 patients referred a substantial decrease of symptoms, while in the other 235 cases symptoms persisted although with spontaneous remissions and relapses. No patient died for any reason.

Immunophenotyping of peripheral lymphocytes. Immunological studies were performed in 40 consecutive patients and results compared with those obtained from 35 age- and sex- matched healthy controls. The absolute numbers of circulating lymphocytes were comparable in the two groups (CFS, 1804 ± 354 x 106/l; normal controls 1927± 401 x 106/l). No statistically significant differences were found in the absolute numbers of circulating total T cells (CD3+) and of total helper/inducer (CD4+) or suppressor/cytotoxic (CD8+) T cells (Table 2). CD4/CD8 ratios in CFS patients were also superimposable to values obtained in the control group (data not shown). Circulating B cells, as evaluated by an anti-CD19 antibody appeared to be significantly higher in the CSF patients as compared to healthy subjects (210 ± 71 vs 116 ± 36 x 106/l; p<0.05) (Table 2). As shown in table 3, two-color immunofluorescence disclosed a significant decrease in the numbers of circulating CD3-/CD16+ NK lymphocytes in the CFS group (80 ± 14 vs 136 ± 27 x 106/l; p<0.05) (Table 3). Subset analysis of NK cells showed that whilst CD56+/CD57+ cells appeared to be decreased in the CFS group (123 ± 35 x 106/l) as opposed to normal controls (186 ± 50 x 106/l) (p < 0.05), the CD8+/CD56+ subpopulation of NK cells appeared significantly increased in CFS cases (165 ± 55 vs 126 ± 37 x 106/l) (p < 0.05). In addition, the CD16-/CD56+ subset of NK lymphocytes appeared also significantly expanded in CFS cases (138 ± 48 x 106/l) as compared to control subjects (86 ± 23 x 106/l) (p<0.05). Interestingly, most of CD56+ cells in CFS patients displayed an increased expression of adhesion molecules CD11b, CD11c, CD54 and of the activation antigen CD38 as shown by two-color flow cytometry (Fig. 1). The percentages of CD56+ NK cells expressing such membrane molecules appeared higher than in healthy controls, even though a statistical significance was reached only when differences in CD56+/CD38+ cells were analyzed (Fig. 1). Subset analysis of CD4+ lymphocytes with antibodies recognizing different isoforms of the CD45 antigen, evidenced a significant reduction of both absolute numbers and relative percentages of CD4+/CD45RA+ T cells in CFS patients as opposed to healthy subjects (Table 3 and Fig 2). No significant changes were conversely detected when subpopulations of CD4+ cells bearing CD45RO and CD29 molecules were analyzed in both patients and controls (Table 3 and Fig 1).
Circulating T cells of CFS patients were found to display an "activated" phenotype. The total number of CD3+ T lymphocytes expressing HLA-DR antigens (Table 3) was significantly increased (138.6 ± 61 x 106/l) as compared to normal controls (81.4 ± 39 x 106/l) (p < 0.05). In addition, the percentage of CD3+ T cells expressing the activation antigens CD11b, CD11c and CD54 appeared also increased in the CSF group (Fig 3A), but such difference did not reach a statistical significance. Conversely, the subset analysis of the intercellular adhesion molecule-1 (ICAM-1/CD54) expression revealed a significant increase of both relative percentages (Fig 3B) and absolute numbers (Table 3) of CD4+/CD54+ T cells in CFS patients (228.8 ± 142 x 106/l) as compared to controls (71.4 ± 31 x 106/l) (p < 0.05). No differences were conversely found in the levels of CD8+/CD54+ T cells among CFS cases and normal subjects (Fig 3B and Table 3). Representative dot plots showing the phenotypic profile (CD45RA, CD29 and CD54) of circulating CD4+ lymphocytes from a CFS patient and from a normal control subject are presented in Fig. 4.
In 11 out of 30 CSF patients analyzed for B lymphocyte subpopulations, CD19+ B cells were found to coexpress the CD5 antigen. The percentage of circulating CD19+/CD5+ B cells in CSF patients appeared variable but usually superior to 15% (Fig 5). CD19+/CD5+ B cells ranged from 0.9 to 5.3% of total lymphocytes in 25 subjects of control group analyzed for such subpopulation of B cells (data not shown). A representative flow cytometry dot plot showing the coexpression of CD5 antigen on CD19+ B cells from CFS patients is shown in Fig. 6.

DISCUSSION
We have observed a significant number of cases of CFS at a single institution, suggesting that CFS, so far never reported in Italy in a case series, is present in this country. This high number of cases of CFS at a single institution can be explained by the fact that our center was the only referral center for CFS in Italy in 1991.
The clinical characteristics of our patients are similar to those reported in other countries [1-2]. We were able to detect 3 groups of patients with distinct natural history. While in 8 patients symptoms of CFS disappeared during the 2 year follow-up of this study, and in 22 other patients there was a significant decrease of symptoms, in other 235 patients symptoms of CFS persisted although with spontaneous remissions and relapses. It must be stressed that the fatigue was so severe to induce 102 subjects to stop their working activities for a relatively long period of time, therefore affecting significantly the social life of these patients. A debate exists in the literature as whether the increased prevalence of psychiatric diseases seen in the CFS patients is the primary cause of the syndrome or secondary to the debilitation of having a chronic disease. Buchwald et al [4] reported abnormal magnetic resonance imaging of the brain in 78% of patients with CFS tested and this was significantly increased in comparison with healthy controls.
A detailed immunophenotypic study of circulating lymphocytes was performed in 40 consecutive, strictly CDC-defined CFS patients and 35 healthy matched controls. In our cohort of patients with CFS a number of immunologic abnormalities involving specific subpopulations of NK cells, T lymphocytes and B cells were detected. First, the absolute number of CD3-/CD16+ and CD57+/CD56+ NK cells were reduced along with a concomitant significant increase of CD56+/CD8+ and CD16-/CD56+ subpopulations. CD56+ NK cells also expressed an increased amount of CD11b, CD11c, and CD54 antigens. Second, CD4+ T cells from CFS patients were found to show an increased expression of the ICAM-1 (CD54) antigen as opposed to CD8+ T cells. Third, the absolute numbers and relative percentage of CD4+/CD45RA+ "naive" T cells were reduced in CFS patients, whilst CD4+/CD45RO+ and CD4+/CD29+ T cell subsets did not show appreciable variations in patients as compared to healthy controls. Fourth, CD19+ B cells were overall increased in patients and in 11 out of 30 patients the B cell expansion was sustained by an increase of the CD5+/CD19+ subset of B lymphocytes.
The immunological results observed in this study, although in agreement with previously data showing a decrease in circulating NK cells [7-9], did not confirm the decreased proportions of CD4+ and CD8+ T cells reported by Lloyd et al. [10] and by Behan et al. [15] in CFS. In addition, we have shown that the NK cell decrease in CFS appears to be subset-specific in that the numbers of CD8+/CD56+ cells in our patients were increased. Interestingly, the CD16-/CD56+ subset of NK cells was also expanded in our patients. This latter appears the most immature subpopulation of NK cells and is thought to represent an "early precursor" subset giving rise to mature CD16+/CD56+ NK lymphocytes [24, 25]. In addition, CD16-/CD56+ cells were found to express the highest levels of CD11c, but unable to secrete appropriate amounts of g-interferon (IFN) and turned to be less efficient effectors of non-MHC restricted cytotoxicity than CD16+ NK cells (24-26). Accordingly, we have found an increased proportion of CD56+/CD11c+ and CD56+/CD11b+ cells in our patients and previous studies have shown an impaired release of g-IFN by mitogen stimulated cultures [14, 16], and a reduced cytotoxic activity against K562 target cells of CD56+ lymphocytes from CFS subjects [8]. Whether the expansion of CD16-/CD56+ NK subsets in our CFS patients is a sort of "rebound" response to diminished levels of CD16+ and CD57+ mature NK cells, or is directly related to the action of a still unknown pathogen, remains to be established. The umbalanced distribution of NK subsets evidenced in our cohort of CFS patients, represents an intriguing finding which can explain at least in part the reduced levels of NK activity detected in CFS by different investigators [7-9]. Abnormalities in NK cells although reported previously in CFS patients have not been confirmed by more recent studies [17]. The reason for such discrepancies is not clear at the moment even though patient selection, disease status, immunophenotyping techniques and different antibody combinations employed to evaluate NK subsets, might explain divergent results on NK cells enumeration in studies of CFS patients. As suggested by Strauss et al. the use of fresh samples or cryopreserved cells may represent one of such biases [17]. The issue of NK function in CFS remains therefore an open matter of debate warranting further careful investigations. Our results suggest that the reduced NK functional activity in CFS may be in part related to an absolute or relative increase of the less efficient CD16-/CD56+ subset of NK cells. The increased expression of adhesion molecules and of activation markers detected by us on CD56+ cells, further supports the hypothesis of a preferential recruitment of such NK subset in CFS subjects.
We have also detected a reduction in the relative proportion of CD4+/CD45RA+ "naive T cells", along with normal numbers of CD4+/CD45RO+ in italian patients with CFS. These results are in agreement with a recent report by Straus et al [17], even though the absolute number of CD4+ T cells and the CD4/CD8 ratios appeared normal in our patients. Accordingly, the rate of expression of the adhesion molecule ICAM-1 (CD54) on CD4+ T cells appeared increased in our patients, as also shown by a previous report by Gupta et al. [9]; even though we did not perform CD45RO/CD54 double-labeling experiments, it is possible to speculate that CD4+/CD54+ T cells in our patients were of CD45RO type as clearly indicated by Straus et al. [17]. If so, it could be proposed as suggested by Strauss et al. that CD4+/CD45RO+ lymphocytes owing to a high expression of surface molecule mediating cell adhesion and tissue trafficking, may be responsible, following migration into specific tissues, for at last some of the typical symptoms of CFS, i.e. myalgies, arthralgia and lymph node tenderness. Cytokine produced focally by such cells within specific tissues, might in fact mediate mild inflammatory response [27]. Further experiments are needed to verify such an hypothesis.
The increase in CD5+/CD19+ B cells detected in some of our patients also deserves some speculations. Relative expansions of such subset of B lymphocytes have been previously reported in other series of CFS patients (8, 12) and in infectious mononucleosis [28]. None of our patients displaying increased CD5+/CD19+ cells had a positive serology for EBV but all presented with adenopathy, which appeared sustained in two biopsied cases by a lymphoid follicular hyperplasia. Since CD4+/CD45RA+ T cells are implicated in the control of B cell proliferation and differentiation [29, 30], it could be speculated that the significant reduction of such T cell subset as detected in different studies of CFS patients [8, 17, 31] and by us, might in turn result in a dysregulation of the B lymphocyte compartment. As a matter of fact our CFS subjects showing a relative increase of CD5+ B cells, presented the lowest numbers of CD4+/CD45RA+ T cells (data not shown). In addition, decreased levels of CD4+/CD45RA+ T cells have been reported to correlate with the extent of disease and the presence of autoimmune phenomena in patients with lymphoproliferative disorders sustained by CD5+/CD19+ B cells [32]. Finally, selective loss of CD45RA positive cells has been shown to play a role in the pathogenesis of autoimmune disorders [33-36], including those accompanied by the expansion of the CD5+/CD19+ subsets of B cells [37-38]. It could be therefore proposed that the increased level of CD5+ B cells detected by us and other investigators in CFS patients, is somehow related to the loss of regulating CD45RA+ T cells, and that the expansion of such specific B cell pool might in part contribute to CFS symptoms, which often overlap with those of autoimmune diseases characterized by abnormalities of CD5+ B lymphocytes [37-41].
We conclude that CFS is a new disease present in Italy, with clinical characteristics similar to those reported in patients from other countries and associated to a number of heterogeneous disfunctions of the immune system. Further investigations are needed in order to better define immunological abnormalities detected in patients with CFS with a special regard to the derangement in NK cell subpopulatons reported in our study and to the defect in CD45RA T cells described by different groups of investigators.

REFERENCES

1. Shafran SD. The chronic fatigue syndrome. Am J Med 1991;90:730-739.

2. Holmes GP, Kaplan JE, Gantz NM et al. Chronic fatigue syndrome: a working case definition. Ann Intern Med 1988;108:387-389.

3. Salahuddin SZ, Ablashi DV, Markham P et al. Isolation of a new HBLV in patients with lymphoproliferative disorders. Science 1986;234:596-601.

4. Buchwald D, Cheney PR, Peterson DL et al. A chronic illnes characterized by fatigue, neurologic and immunologic disorders, and active human herpesvirus type 6 infection. Ann Intern Med 1992;116:103-113.

5. Archard LC, Behan PO, Bowles NE, Bell EJ, Doyle D. Postiviral fatigue syndrome: persistance of enterovirus RNA in muscle and elevated creatine kinase. J R Soc Med 1988;81:326-329.

6. DeFreitas E, Hilliard B, Cheney PR et al. Retroviral sequences related to human T-lymphotropic virus type II in patients with chronic fatigue immune dysfunction syndrome. Proc Natl Acad Sci USA 1991;88:2922-2926.

7. Caligiuri M, Murray C, Buchwald D et al. Phenotypic and functional deficiency of natural killer cells in patients with chronic fatigue syndrome. J Immunol 1987;139:3306-3313.

8. Klimas NG, Salvato FR, Morgan R, Fletcher MA. Immunologic abnormalities in chronic fatigue syndrome. J Clin Microbiol 1990;28:1403-1410.

9. Gupta S, Vayuvegula B. A comprehensive immunological analysis in chronic fatigue syndrome. Scand J Immunol 1991;33:319-327.

10. Lloyd AR, Wakefield D, Broughton CR, Dwyer JM. Immunological abnormalities in the chronic fatigue syndrome. Med J Austral 1989;151:122-124.

11. Landay AL, Jessop C, Lennette ET, Levy JA. Chronic fatigue syndrome: clinical condition associated with immune activation. Lancet 1991;338:707-712.

12. Buchwald D, Komaroff AL. Review of the laboratory findings for patients with chronic fatigue syndrome. Rev Infect Dis 1991;13, (suppl. 1):S12-S18.

13. Straus SE, Dale JK, Peter JB, Dinarello CA. Circulating lymphokine levels in the chronic fatigue syndrome. J Infect Dis 1989;160:1085-1086.

14. Chao CC, Janoff EN, Hu S et al. Altered cytokine release in peripheral blood mononuclear cell culture from patients with the chronic fatigue syndrome. Cytokine 1991;3:292-298.

15. Behan PO, Behan WM, Bell EJ. The postviral fatigue syndrome. An analysis of the findings in 50 cases. J Infect 1985;10:211-222.

16. Morte S, Castilla A, Civeira MP, Serrano, Prieto J. Gamma-interferon and chronic fatigue syndrome. Lancet 1988;ii:623-624.

17. Straus SE, Fritz S, Dale JK, Gould B, Strober W. Lymphocyte phenotype and function in the chronic fatigue syndrome. J Clin Immunol 1993;13:30-40.

18. Tirelli A, Pinto A, Marotta G et al. Clinical and immunological study of 205 patients with chronic fatigue syndrome: a case series from Italy. Arch Intern Med 1993;153:116-120.

19. Hathaway SR, McKinley JC. The minnesota Multiphasic Personality Inventory Manual. New York, Psychological Corporation 1951, 1967.

20. Pinto A, Zagonel V, Carbone A et al. CD13 expression in B-cell chronic lymphocytic leukemia is associated with the pattern of bone marrow infiltration. Leukemia and lymphoma 1992;6:209-218.

21. Carbone A, Gloghini A, Volpe R et al. Follicular lymphoma of compartmentalized small cleaved center cells and mantle zone lymphocytes: evidence for a common derivation. Am J Clin Pathol 1992;98:437-448.

22. Fletcher MA, Baron GC, Ashman M, Fischl MA, Klimas NG. Use of whole blood methods in assessment of immune parameters in immunodeficiency states. Diagn Clin Immunol 1987;5:69-81.

23. Siegel S. Nonparametric statistic for the behavioral sciences. New York: Mc Graw-Hill book Co, 1956.

24. Lanier LL, My Le A, Civin CI, Loken MR, Phillips JH. The relationship of CD16 (Leu-11) and Leu-19 (NKH-1) antigen expression on human peripheral blood NK cells and cytotoxic T lymphocytes. J Immunol 1986;136:4480-4486.

25. Nagler A, Lanier LL, Cwirla S, Phillips JH. Comparative studies of human FcRIII-positive and negative natural killer cells. J Immunol 1989;143:3183-3191.

26. Robertson MJ, Ritz J. Biology and clinical relevance of human natural killer cells. Blood 1990;76:2421-2438.

27. Sanders ME, Makgoba MW, Sharrow SO et al. Human memory T lymphocytes express increased levels of three cell adhesion molecules (LFA-3, CD2, and LFA-1) and three other molecules (UCHL1, CDw29, and Pgp-1) and have enhanced IFN-g production. J Immunol 1988;140:1401-1407.

28. Hassan J, Feighery C, Bresnihan B, Whelan A. Increased CD5+ B cells in patients with infectious mononucleosis. Br J Haematol 1990;74:375-376.

29. Janeway CA, Carding S, Jones B, Murray J et al. CD4+ T cells: specificity and function. Immunol Rev 1988;101:39-80.

30. Sanders ME, Makgoba MW, Shaw S. Human naive and memory T cells: reinterpretation of helper-inducer and suppressor-inducer subsets. Immunol Today 1988;9:195-198.

31. Salvato F, Fletcher MA, Ashman M, Klimas NG. Immune dysfunction among chronic fatigue syndrome patients with clear evidence of Epstein-Barr virus reactivation. J Exp Clin Can Res 1988;7(suppl.):89.

32. Peller S, Kaufman S. Decreased CD45RA T cells in B-cell chronic lymphatic leukemia patients: correlation with disease stage. Blood 1991;78:1569-1573.

33. Morimoto C, Steinberg AD, Letvin NL et al. A defect of immuno regulatory T cell subsets in systemic lupus erythematosus patients demonstated with anti-2H4 antibody. J Clin Invest 1987;79:762-766.

34. Morimoto C, Hafler DA, Weiner HL et al. Selective loss of the suppressor-inducer T cell subset in progressive multiple sclerosis. Analysis with anti-2H4 monoclonal antibody. N Engl J Med 1987;316:67-72.

35. Takeuchi T, Tanaka S, Steinberg AD et al. Defective expression of 2H4 molecule after autologous mixed lymphocyte reaction activation in systemic lupus erythematosus patients. J Clin Invest 1988;82:1288-1294.

36. Morimoto C, Romain PL, Fox DA et al. Abnormalities in CD4+ T-lymphocyte subsets in inflammatory rheumatic diseases. Am J Med 1988;84:817-825.

37. Casali P, Burastero SE, Nakamura M et al. Human lymphocytes making rheumatoid factor and antibody to ssDNA belong to Leu-1+ B-cell subsets. Science 1987;236:77-81.

38. Plater-Zyberk C, Maini RN, Lam K, Kennedy TD, Janossy G. A rheumatoid arthritis B cell subset expresses a phenotype similar to that in chronic lymphocytic leukemia. Arthritis Rheum 1985;28:971-976.

39. Burastero SE, Casali P, Wilder RL, Notkins AL. Monoreactive high affinity and polyreactive low affinity rheumatoid factors are produced by CD5+ B cells from patients with rheumatoid arthritis. J Exp Med 1988;168:1979-1992.

40. Dauphinee M, Tovar Z, Talal N. B cells expressing CD5 are increased in Sjogren's syndrome. Arthritis Rheum 1988;31:642-647.

41. Iwatani Y, Amino N, Kaneda T et al. Marked increase of CD5+ B cells in hyperthyroid Graves' disease. Clin Exp Immunol 1989;78:196-200.

LEGENDS TO THE FIGURES

Figure 1. Relative expression of cell adhesion (CD11b, CD11c, CD54) and activation (CD38) molecules on CD56-positive NK subpopulations. Flow cytometry data are expressed as a percent of double positive cells. Closed bars (healthy subjects); shaded bars (CFS patients). * p < 0.05.

Figure 2. Relative expression of CD29, CD45RA and CD45RO antigens on CD4-positive lymphocytes. Flow cytometry data are expressed as a percent of double positive cells. Closed bars (healthy subjects); shaded bars (CFS patients). * p < 0.05.

Figure 3. (A) Relative expression of adhesion molecules (CD11b, CD11c, CD54) on CD3-positive lymphocytes. (B) Relative expression of intercellular adhesion molecule-1 (ICAM-1/CD54) on CD4- and CD8-positive lymphocyte subpopulations. Flow cytometry data are expressed as a percent of double positive cells. Closed bars (healthy subjects); shaded bars (CFS patients). * p < 0.05.

Figure 4. Representative dot-plots showing two-color immunofluorescence analysis of peripheral CD4-positive lymphocytes from an healthy control subjects (CNT; left panels) and a patient with active CFS (CFS; right panels). Four regions are identified by setting quadrant markers at the highest levels of non specific fluorescence. X-axes represents log-green fluorescence; Y-axes represents log-red fluorescence. Double-positive cells are identified in the upper-right region.

Figure 5. The percentage of CD5+/CD19+ B cells in peripheral blood of eleven CFS patients. Closed bars, CD19+ cells; shaded bars, CD19+/CD5+ cells.

Figure 6. Representative dot-plot showing two-color immunofluorescence analysis of circulating CD19-positive B cells from a CFS patient. Four regions are identified by setting quadrant markers at the highest levels of non specific fluorescence. X-axes represents log-green fluorescence (CD19); Y-axes represents log-red fluorescence (CD5). Double-positive cells (CD5+/CD19+) are identified in the upper-right region.