To determine the mortality rate and its associated variables from an open cohort started in July 1997 at the HTLV Clinic, Emilio Ribas Institute (IIER), a major infectious diseasehospital in São Paulo, Brazil.
METHODS:
Since inception up to September 2018, we admitted 727 HTLV-1-infected individuals, with a rate of 30-50 new admissions per year. All patient data, including clinical and laboratory data, were regularly updated throughout the 21-year period, using a dedicated REDCap database. The Ethical Board of IIER approved the protocol.
RESULTS:
During 21 years of clinical care to people living with HTLV-1 in the São Paulo region, we recruited 479 asymptomatic HTLV-1-infected individuals and 248 HAM/TSP patients, of which 632 remained under active follow-up. During a total of 3800 person-years of follow-up (maximum follow-up 21.5 years, mean follow-up 6.0 years), 27 individuals died (median age of 51.5 years), of which 12 were asymptomatic, one ATLLpatient and 14 HAM/TSP patients. HAM/TSP diagnosis (but neither age nor gender) was a significant predictor of increased mortality by univariate and multivariate (hazard ratio (HR) 5.03, 95% CI [1.96-12.91], p = 0.001) Cox regression models. Coinfection with HIV/HCV was an independent predictor of increased mortality (HR 15.08; 95% CI [5.50-41.32]; p < 0.001), with AIDS-related infections as a more frequent cause of death in asymptomatics (6/13; p = 0.033). HIV/HCV-negative fatal HAM/TSP cases were all female, with urinary tract infection and decubitus ulcer-associated sepsis as the main cause of death (8/14, p = 0.002).