Human T-lymphotropic virus type 1 (HTLV-1)-associated myelopathy/tropical spastic paraparesis (HAM/TSP) is a rare chronic neuroinflammatory disease. While the disease usually progresses slowly without remission, there is a subgroup of patients with rapid progression and another subgroup with very slow progression. However, there have been no reports to date that have successfully determined the criteria to differentiate these subgroups. Therefore, we initially conducted a statistical modeling analysis to explore representative patterns of disease progression using data from our nationwide HAM/TSP patient registration system ("HAM-net"). The latent class mixed model analysis on the retrospective data (n = 205) of disease progression measured by the change in Osame Motor Disability Score from the onset of the disease to diagnosis demonstrated three representative progression patterns of HAM/TSP. Next, to test the effect of the progression rate at the initial phase of the disease on long-term prognosis, we divided 312 "HAM-net" registered patients into three groups (rapid, slow, and very slow progressors) based on the progression rate, then analyzed long-term functional prognosis of each group using the Kaplan-Meier method. Our data clearly demonstrated that the rapid progression at the early phase of the disease is an important poor prognostic factor. Moreover, to determine the biomarkers capable of discriminating the difference in disease activity, we compared the value of potential biomarkers of HAM/TSP among rapid (n = 15), slow (n = 74), very slow (n = 7), and controls (non-HAM/TSP patients, n = 18). The cerebrospinal fluid (CSF) levels of neopterin and C-X-C motif chemokine 10 (CXCL10) were the most valuable markers to discriminate among rapid, slow, and very slow progressors. To differentiate between rapid and slow progressors, the cut-off values of neopterin and CXCL10 were determined to be 44 pmol/mL and 4400 pg/mL, respectively. Furthermore, to differentiate between slow and very slow progressors, these values were determined to be 5.5 pmol/mL and 320 pg/mL, respectively. Notably, we found that CSF levels of these markers in very slow progressors were within the reference range. Thus, we propose a new classification criteria for disease activity of HAM/TSP that may contribute to improving the treatment algorithm for HAM/TSP.
Keywords: CXCL10; HAM/TSP; HTLV-1; biomarker; classification criteria; disease activity; neopterin.