Reply to J.J. Castillo et al

Abstract

We thank Castillo and Beltran for their comments regarding our work and are pleased to have this opportunity to address their interest in the evaluation of the adult-T cell leukemia/lymphoma (ATL) prognostic index (PI) separately for patients with acute-type versus lymphoma-type ATL. Intheircorrespondence,Castilloetal describedtheclinical, genetic, and pathophysiologic differences between acuteand lymphoma-type ATL. We completely agree with their descriptions. Combination therapy with interferon alfa and zidovudine, which is one of the standard treatments outside of Japan, has been reported to be more effective against acute-type ATL than conventional chemotherapy, but ineffective against lymphoma-type ATL. Moreover, a recent study reported that the genetic differences between acuteand lymphoma-type ATL are acquired by clonal evolution in the lymph nodes. It remains to be determined whether the genetic alterations that have been observed in ATL cells have significant influence on the clinical outcomes of patients with ATL, or whether these alterations should be taken into consideration when tailoring therapy to individual patients. In our study, we attempted to establish an ATL-PI that is applicable to both acuteand lymphoma-type ATL, given that the subtypes have not been considered separately when determining therapeutic options in Japan, and also to make a simple PI that can be applied easily in clinics. Consistent with previous reports, in our study, patients with acute-type ATL (n 564) showed worse median overall survival time than patients with lymphoma-type ATL (n 243): 6.9 months (95% CI, 6.2 to 7.8 months) and 9.9 months (95% CI, 8.1 to 11.3 months), respectively (P .001). Application of the ATL-PI to each subtype revealed median overall survival times for acute-type ATL of 4.3 (95% CI, 3.3 to 5.1), 7.3 (95% CI, 6.5 to 8.4), and 15.7 (95% CI, 13.9 to 23.2) months for patients at high, intermediate, and low risk, respectively, whereas those for lymphoma-type ATL were 3.6 (95% CI, 1.5 to 4.7), 9.0 (95% CI, 7.4 to 11.1), and 14.0 (95% CI, 12.4 to 17.2) months, respectively. The three risk groups according to the ATL-PI were thus effectively prognostic both in acute-type (P .001 by log-rank test with 2 df) and lymphoma-type (P .001 by log-rank test with 2 df) ATL. In conclusion, ATL-PI is a PI for patients with aggressive ATL, and it is applicable for both acute and lymphoma types separately. We believe ATL-PI is a promising tool for risk stratification. We hope that it will be validated internationally in studies that include patients treated with a combination of interferon alfa and zidovudine.