Recent studies have suggested that one mechanism of immune escape could be catabolism of the essential amino acid tryptophan, via indoleamine 2,3-dioxygenase (IDO) along the kynurenine pathway. Some metabolites derived from tryptophan by IDO, such as L-kynurenine, block antigen-driven specific T-cell proliferation and induce T-cell death. IDO activity thus plays an important role in regulating immune responses exerted by antigen-presenting cells (APCs) and also provides transformed cells with a potent tool to help escape from assault by the immune system. We have recently been able to clarify the utility of either serum L-kynurenine or tissue IDO expression as prognostic factors for DLBCL patients treated using R-CHOP. The present study therefore examined the serum level of L-kynurenine and tissue IDO expression simultaneously in DLBCL patients in an attempt to identify patients showing truly poor prognosis.
The study protocol used a retrospective case series design that was approved by our institutional review board. We investigated DLBCL patients who had been histologically diagnosed according to the WHO classification of hematopoietic tumors between December 2003 and November 2006. All serum samples were obtained on admission before starting any treatments. Informed consent was obtained from all patients for the use of samples in this study. Tumor tissue was obtained from every patient by lymph node biopsy before initiation of therapy. Patients were assigned to receive six to eight cycles of R-CHOP or R-THP-COP therapy. Each regimen consisted of rituximab (375 mg/m2 on day 1), cyclophosphamide (750 mg/m2 on day 3), doxorubicin (DOX) or tetrahydropyranyl-adriamycin (THP) (50 mg/m2, given as a 30-min infusion on day 3), vincristine (1.4 mg/m2 on day 3, with a maximal dose of 2.0 mg), and prednisolone (100 mg/day on days 3–7). We measured serum concentrations of L-kynurenine by high-performance liquid chromatography (HPLC), and serum levels of sTNF-Rs (p55, TNF-R1; p75, TNF-R2) by ELISA. And we examined IDO expression in DLBCL tissues by immunohistochemical staining in a total of 119 patients.
Median serum levels of L-kynurenine were significantly higher in DLBCL patients (1.598 mM; range, 0.489–5.066 mM) than in healthy volunteers (1.133 mM; range, 0.752–1.388 mM). And we found that lymphoma cells expressed IDO in 32% cases. We therefore classified cases as IDO-positive DLBCL or IDO-negative DLBCL based on whether lymphoma cells themselves expressed IDO. No significant correlation was identified between serum L-kynurenine level and IDO expression. Serum L-kynurenine level showed a significant positive correlation with concentrations of sTNF-R1 and sTNF-R2 in IDO-negative DLBCL patients. CR rates with IDO-positive and IDO-negative DLBCL were 42% and 75%, respectively (p=0.012). Three-year OS rates were 55% and 78% for patients with L-kynurenine ≥1.5 and <1.5 mM, respectively (p=0.027), and 44% and 75% for patients with IDO-positive and IDO-negative DLBCL, respectively (p=0.0017). Univariate Cox analysis identified the following prognostic factors: IDO expression, serum kynurenine, PS, and multiple site extranodal involvement. Multivariate analysis of these four parameters showed that positive IDO expression and serum kynurenine ≥1.5 mM were significant prognostic factors. We analyzed patients' outcome according to serum concentration of L-kynurenine and IDO expression of lymphoma cells at diagnosis. Patients with low kynurenine and IDO-negative showed the best outcomes, while patients with high kynurenine and IDO-positive showed the poorest outcome. This combination distinguishes 3 separate prognostic groups of “very good”, “good”, and “poor”, with 3-year OS ranging from 84.1% to 31.1% (p<0.001, Figure).
IDO activity might play an important role in the disease activity of DLBCL. The combination of tissue IDO expression and serum level of L-kynurenine is a useful tool to predict the prognosis of patients with DLBCL.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.