The fatal outcome in one of two multiple sclerosis (MS) patients with idiopathic thrombocytopenic purpura (ITP) prompted our interest in platelet activity and function in the context of MS. Although Putnam investigated a possible role of venule thrombosis as a factor in central nervous system (CNS) demyelination in 1935 1, a role for platelets in CNS demyelination was not further considered until a series of papers in the 1960s, such as that of Wright et al. 2 For example, Nathanson and Savitsky 3 employed a measure of platelet adhesiveness in 132 subjects, 60 of whom had MS. The investigators reported increased platelet adhesiveness in both MS and Guillain-Barre correlating with disease activity. Although other investigations confirmed their findings, they contributed little additional information. More recently, a central role for platelets in inflammation has emerged, as reviewed 45.

Our observation of platelet abnormalities in MS 6 and subsequent observation of thrombosis in cutaneous venules and capillaries adjacent to subcutaneous ulcers complicating subcutaneous injections of interferon-beta1b 7 heightened our interest in a possible role of platelet dysfunction in MS. To investigate the basis of these observations, we have applied the flow cytometric analysis of platelet-derived microparticles (PMP) and CD62p expression, as well as other more conventional assays. For this study, we employed consecutively recruited patients and measured, in addition to routine tests such as platelet counts, the expression of platelet activation marker P-selectin (CD62p), platelet microparticles (PMP) in plasma, platelet micro-aggregates (PAg), protein S activity, and platelet-associated immunoglobulins IgG and IgM, as described following.

By all three measures (PMP, PAg, CD62p), we have documented the presence of platelet activation in MS patients. Increased PMP and increased CD62p expression are accepted markers of platelet activation 1521. Our observations strongly support the conclusion that platelets are chronically activated in MS, and support results in the earlier literature 2.

Platelet activation may be an epiphenomenon consequent to the disease process in MS, possibly secondary to endothelial injury. Indeed, we have reported elevated plasma endothelial microparticles (EMP) in MS, positive for platelet endothelial cell adhesion molecule-1 (CD31/PECAM-1) 22 and suggested that this increase in EMP may reflect the interaction of activated T-cells with endothelium. It is also possible that interaction of released PMP with underlying endothelium contributes to the endothelial abnormalities involved in the pathophysiology of MS. Binding of P-selectin on PMPs to P-selectin glycoprotein ligand-1 (PSGL-1) and PECAM-1 on lymphocytes is known to induce or increase the expression of integrins such as integrin α4β1 (VLA-4), and thus facilitate lymphocyte binding to endothelium 23. However; the in vivo role of CD62p+ PMP and CD31+ EMP and their interactions in MS remains unknown. It is known that PMP can transport a variety of bioactive agents including platelet activating factor (PAF), amyloid precursor protein, complement factors, and other molecules 2425, all of which could contribute to the disease process in some manner.

Although study of the role of cell-derived microparticles in disease processes is still in its infancy, it is clear that PMP are involved in angiogenesis 2627 and in certain malignant processes 2829. In light of these developments, it is reasonable to postulate that PMP may play an active role in the pathogenesis of MS. Several chemokines have been detected on platelets 30 and some are carried by PMP 3132. The chemokine receptor CXCR4 can be transferred by PMP from cell to cell 33. Leukocyte activation by PMP has been also reported 34, illustrating another potential link between PMP and the inflammatory disease process.

It remains unclear how platelet activation and the release of PMP pertain to our clinical observation of 2 cases of ITP associated with MS at our center. This finding, in accord with the earlier literature 2, suggests the presence of a procoagulant state in MS, and elements of the coagulation system such as fibrin and tissue factor (TF) are found in MS lesions 35. However, review of the available literature has identified additional MS patients with ITP complicating their illness. Granier et al in 2001 reported a 40 year old French patient with well-documented MS who had a transient thrombocytopenia 36. Four additional cases have been identified recently in Spain 37. However, two earlier cases [1997] with Hodgkin's and thrombocytopenia cannot be accepted as ITP 38. Recent epidemiological studies in Maryland 39 and in Nova Scotia, Canada 40 have brought to light additional cases. Segal et al. identified 4 such cases in children 39 and Kirby et al. have identified 22 MS patients with "immune thrombocytopenia" 40. Segal's report reflects a 25-fold higher incidence of ITP in MS in the Maryland study and, although details from the Canadian study are not available, an even higher coincidence may occur there. Although PMPs were thought to possess solely procoagulant properties, it appears that some PMP phenotypes may exert predominantly anticoagulant functions by modulating the protein C/S system 41.

Platelet associated IgM was increased in the MS group (p < 0.01) but platelet associated IgG was not. Our finding of elevated platelet-associated IgM and platelet activation may reflect the presence low levels of serum anti-phospholipid antibodies in MS. This finding is consistent with reports of predominately IgM anti-phospholipid antibodies in MS. Sugiyama et al found 14 of 32 patients positive for IgM anti-phospholipid antibody but only 3 of 32 positive for IgG 42. Using another method, we have found anti-phospholipid antibodies exclusively of IgM class in MS 43. Patients with the anti-phospholipid syndrome are prone to recurrent stroke, thrombocytopenia, and neurological manifestations resembling those of MS are possible 44.

As the figures show, not all MS patients show elevated platelet activation or platelet-associated IgM. Since this study was not longitudinal, it is not known if the patients with elevated values represent a subset of MS patients, or if most MS patients have occasional or episodic elevations, possibly preceding exacerbations.

The classic neuropathology of MS consists of the perivenular plaque surrounding venular endothelial cells with infiltrating lymphocytes and macrophages 45. The interactions of these cellular elements and resulting myelin and axonal damage are complex and incompletely understood. Our findings collectively suggest a complex interplay of platelets with endothelial and other cells in the pathogenesis of the MS plaque. The interaction of CD62p, PMP, macrophages, anti-phospholipid antibodies, and the CNS venular endothelium requires further study.

Findings of this study demonstrate that platelets are significantly activated in MS patients. However, their role in pathogenesis of MS remains unknown. Further longitudinal studies of larger cohorts of MS patients are required to validate the significance of the present study and demonstrate the role of platelet activation in neuro-pathogenesis of MS.

MS: Multiple sclerosis; ITP: idiopathic thrombocytopenic purpura; PMP: plasma platelet microparticles; CD62p: P-selectin expression; CNS: central nervous system; PAg: platelet micro-aggregates; FITC: fluorescein isothiocyanate; CD62p:

P-selectin;

The authors declare that they have no competing interests.

WAS, WH, YSA and AM designed the clinical research protocol, provided expertise for laboratory measurement of platelet activity markers.

WJ and LLH contributed to the manuscript by collecting laboratory data, providing expertise for analysis of the obtained laboratory data, and doing biostatistical analysis and generating figures.

JSA contributed by preparing the manuscript and providing expertise in interpreting the obtained data on platelet adhesion molecules.