Abstract

JCOG consecutively conducted 6 multicenter clinical trials for advanced aggressive lymphoma in 1990s, including 4 CHOP-based chemotherapies (JCOG9505, 9506, 9508 and 9809) and 2 second or third generation, multidrug combination chemotherapies (JCOG9002 and 9203). Among the whole 1,141 enrolled patients (pts), histopathological specimens from 1,084 pts (95%) were reviewed by 6 hematopathologists and diagnosed according to WHO classification. Factors affecting the overall survival (OS) of pts with each B- or T-cell lymphoma were analyzed. In all the 6 trials, major eligibility criteria were as follows: age between 15 and 69 except one trial for elderly pts, pathological subtypes of intermediate- or high-grade lymphoma by Working Formulation excluding adult T-cell leukemia/lymphoma, lymphoblastic lymphoma, and cutaneous T-cell lymphoma. ECOG performance status 4 was excluded. Numbers of the cases of major B-cell lymphoma subtypes were 642 diffuse large B-cell lymphoma(DLBCL), 104 follicular lymphoma(FL), 30 mantle cell lymphoma(MCL) and 24 marginal zone B-cell lymphoma(MZL). One hundred and thirty six cases of PTCL were analyzed, including 18 anaplastic large cell lymphoma (ALCL), 46 angioimmunoblastic T-cell lymphoma (AILT), 53 PTCL unspecified (PTCL-U), 17 extranodal NK/T-cell lymphoma, nasal type (NK/T). Five year OS rates of major subtypes were; 83% for MZL, 76% for FL, 59% for DLBCL, 59% for MCL, 61% for ALCL, 55% for AILT, 41% for NK/T and 38% for PTCL-U. In the whole B-cell lymphomas, 5 year OS rates were discriminated well by International Prognostic Index (IPI); i.e. low (L): 75.0%, low-int (LI): 60.3%, high-int (HI): 45.3%, high(H): 25.7%. On the contrary, those of PTCL could not be discriminated well according to IPI, i.e. L: 63.6% LI: 45.1%, HI: 30.8%, and H: 47.4%. Other factors influencing the OS of pts with PTCL were clarified by the univariate or multivariate analysis. In the univariate analysis of 136 PTCL cases, 4 factors related to unfavorable prognosis were identified as follows: low serum total protein level (TP<6.3g/dl) [hazard ratio (HR) 2.39, (95%CI: 1.50–3.80, p=0.0003)], involvement of gastrointestinal tract [HR 1.81 (95%CI: 1.00–3.28, p=0.049)], low serum albumin level (<3.7g/dl) [HR 1.68(95%CI: 1.08–2.61, p=0.02)], and pathological subtype (ALCL+AILT vs. PTCL-U+NK/T) [HR 0.59 (95%CI:0.38–0.90, p=0.02)]. In multivariate analysis, low TP(<6.3g/dl) [HR 2.20(95%CI:1.27–3.80, p=0.0048)] and pathological subtype (ALCL,AILT vs. PTCL-U,NK/T) [HR 0.59(95%CI:0.36–0.95, p=0.028) ] were significant. These results indicated that pts with PTCL showed worse OS than those with B-cell lymphoma. In B-cell lymphoma, IPI was confirmed as the important prognostic model, however, in PTCL, the significance of IPI is limited by other significant factors, i.e. low TP (< 6.3 g/dl) and unfavorable pathological subtype (PTCL-U and NK/T). In conclusion, in prospective multicenter trials settings, IPI model for aggressive B-cell lymphoma is not suitable for PTCL, and further studies on establishing novel prognostic models for PTCL are warranted.

Author notes

Disclosure: No relevant conflicts of interest to declare.