Virological aspects of tropical spastic paraparesis/HTLV-I associated myelopathy and HTLV-I infection.

TSP/HAM, a chronic spastic paraparesis or paraplegia with sphincter disturbances and minimal sensory loss, is characterized pathologically by a meningo-myelitis of the lower thoracic cord with axonal degeneration and demyelinisation of the lateral and anterior spinal tracts. High titer specific anti HTLV-I antibody is present in the serum and in the cerebrospinal fluid with specific intrathecal IgG synthesis, elevated IgG index and intra blood-brain barrier IgG synthesis. Most TSP/HAM patients also exhibit IgG oligoclonal bands in the CSF (also sometimes in the serum), some of those being HTLV-I specific, directed against p24 or against various antigens of the disrupted virus. HTLV-I specific cytotoxic T lymphocytes (CTL), mainly CD8+HLA class I restricted, and recognizing several HTLV-I epitopes especially of the tax, the rex and env proteins are present at high levels in the blood and CSF of TSP/HAM patients. These findings of high CTL activity have led some authors to suggest that these specific CD8 cells may play a major role in TSP/HAM pathogenesis by destruction of HTLV-I infected cells within the central nervous system. Such an hypothesis remains a matter of controversy since some groups have shown that such CTL (both CD8+ or CD4+) are also present in asymptomatic HTLV-I carriers. Recent data have confirmed the high proviral load in PBMC of TSP/ HAM patients (as compared to asymptomatic HTLV-I carriers); however viral expression is very low. There is also recent evidence that more than one copy of HTLV-I proviral DNA may be present in an individual PBMC. One major unanswered question is whether HTLV-I can infect central nervous system cells (neurons, or astrocytes) or if the only HTLV-I infected cells present in the CNS are the infiltrating CD4+ cells. There are published reports supporting both hypotheses. Futhermore, no specific HTLV-I sequences have been related to a given disease outcome, whereas several molecular studies have clearly demonstrated that nucleotide changes in some parts of the HTLV-I genome are correlated with the geographical origin of the patients.