Natural course of HTLV-1 carriers with monoclonal proliferation of T lymphocytes ("pre-ATL") in a 20-year follow-up study.

Previously we reported that asymptomatic human T-cell lymphotropic virus type 1 (HTLV-1) carriers with a monoclonal proliferation of HTLV-1–infected T cells (“pre–adult T-cell leukemia [ATL]”) could account for about 1.7% among a total of HTLV-1 carriers in Japan, and that the pre-ATL subjects had a high potential to develop overt adult T-cell leukemia/lymphoma (ATL/L) and a poor prognosis.1-3 Here we report a further analysis of 50 pre-ATL subjects3 who were enrolled and followed during October 1976 to December 2003 to evaluate their outcomes in Nagasaki prefecture, an endemic area of ATL/L in Japan. Detection assays of the integration band of the HTLV-1 provirus genome were described previously.4 There were 3 end points evaluated: the development of overt ATL/L, death from ATL/L, and overall survival. The diagnostic criteria for overt ATL/L were based on the Lymphoma Study Group (LSG) classification.5 Time-to-event probability was estimated by the Kaplan-Meier methods. Entry age, sex, white blood cell (WBC) count, relative lymphocyte (rLy) counts, and relative abnormal lymphocyte (rAbLy) count were considered as potential prognostic factors. Effects of these factors were evaluated by the Cox proportional hazard regression. Among 50 subjects, 21 (42%) progressed to overt ATL/L (the incidence rate: 48.0 per 1000 person-years), and 31 (62%) died, of whom 21 died from ATL/L itself. The remaining 10 subjects died of opportunistic infections such as Carinii pneumonia or other malignancies (skin carcinoma, lung cancer, etc). In the univariate analyses, subjects with a WBC count more than 9000/ L showed a higher risk for the development of overt ATL/L compared with those with a WBC count less than 9000/ L (hazard ratio [HR], 3.93; 95% confidence interval [CI], 1.50-10.0). The effect did not change even in an adjusted analysis for other factors. A higher rAbLy count (more than 50%) was also associated with the development of overt ATL/L (HR, 3.04; 95% CI, 1.16-7.95). A higher rAbLy count and older entry age (older than 56 years) were marginally associated with overall survival (data not shown). The comparison of the cumulative probability between the 2 groups of WBC counts showed a significant difference in the development of overt ATL/L (85.7% vs 33.0%) and in death from ATL/L (87.0% vs 72.3%), but no difference in the overall survival rate (Figure 1). Several Japanese cross-sectional surveys reported a crude annual incidence of ATL/L ranging from 50 to 150 per 100 000 HTLV-1 carriers.6,7 Although our results may overestimate the risk because we did not take a competing risk of death into consideration, the results indicate that the pre-ATL subjects have a high potential to develop overt ATL/L among HTLV-1 carriers. Notably, most of the subjects with a WBC count above 9000/ L developed overt ATL/L, suggesting that those should be considered as the “early stage of ATL/L.” On the other hand, about two thirds of the subjects with a WBC count below 9000/ L never developed overt ATL/L, suggesting that those would be literally in the “premalignant state.” The screening of the pre-ATL subjects from the