LAUSR

ated with age (≥ or <45 years), serum Alb (< or ≥4.0 g/dL), Ann Arbor stage (IV or I–III), sex, or WBC (≥ or <15 000 /mm). OS was significantly shorter in patients with a high serum Kyn/Trp ratio compared to those with a low ratio (OS rate at 5 years, 60.0 vs 92.2%, P = .018; Figure). OS was also shorter in patients with stage IV compared to those with stage I–III lymphoma (OS rate at 5 y, 78.8% vs 91.9%, P = .015), and in those with lymphocytopenia compared to those without (OS rate at 5 y, 71.4% vs 92.0%, P = .010). There were no significant differences in OS according to age, sex, Hb or Alb level, or leukocytosis. Accordingly, multivariate analysis for OS in the 52 HL patients was performed using the following 3 variables: Ann Arbor stage (I–III or IV), lymphocytopenia (presence or absence), and serum Kyn/Trp ratio (≦ or >42.57). Of these, only the serum Kyn/Trp ratio significantly affected OS (Hazard ratio 7.577; 95% confidence interval, 1.362‐42.160). In immunostaining analyses, HL tumor cells (Hodgkin or Reed–Sternberg cells) were negative, but macrophages and dendritic cells in the microenvironment were positive for IDO. A significant positive correlation existed between the serum Kyn/ Trp ratio and the degree of histologically IDO positive cells in the tumor microenvironment. Conclusions: Quantification of serum Kyn and Trp is useful for predicting the prognosis of an individual HL patient. Furthermore, HL, especially in those patients with a high serum Kyn/Trp ratio, is an appropriate disease for testing novel cancer immunotherapies targeting IDO.