Outcomes of Salvage Chemotherapy and Autologous Stem Cell Transplantation for Relapsed or Refractory Primary Mediastinal Large B Cell Lymphoma (PMLCL) Are Inferior to Diffuse Large B Cell Lymphoma (DLBCL).

INTRODUCTION: PMLCL is an aggressive subtype of non-Hodgkin’s lymphoma with a distinct clinical and gene expression profile. While the results of primary therapy for PMLCL have been shown to be superior to those of DLBCL, the outcomes of salvage chemotherapy and autologous stem cell transplantation (ASCT) for relapsed or refractory disease are not well characterized. We studied the overall response rate (ORR) of this group of patients (pts) to second-line chemotherapy and compared their progression-free (PFS) and overall survival (OS) to a group of pts with DLBCL.

METHODS: We retrospectively reviewed our computerized database and charts between Jan 1^st 1995 until Dec 31^st 2004 and identified 37 pts with PMLCL who underwent salvage chemotherapy. 143 consecutive pts with DLBCL that received salvage chemotherapy between Jan 1^st 1999 and Dec 31^st 2004 were also identified to serve as a cohort for comparison. All pts had disease that did not respond to or relapsed after initial anthracycline-based chemotherapy. Responses have been retrospectively assessed using International Workshop Criteria. Pts typically received 2–3 cycles of platinum-based salvage therapy to assess chemotherapy sensitivity; responding pts proceeded to PBSC mobilization and subsequent ASCT. The conditioning regimen consisted of high-dose VP16 60 mg/kg day −4 and melphalan 160 mg/m2 day −3 with PBSC infusion day 0. Pts with bulk disease at relapse (> 5cm) received involved field radiation post-ASCT.

RESULTS: The median age at the time of salvage chemotherapy was 34 (range 21–64) in the PMLCL group vs. 50 (18–64) in the DLBCL group (P < 0.0001, Wilcoxon test). Advanced stage disease at salvage was found in 46% of PMLCL pts. vs. 49% of DLBCL pts (p=0.74 chi-square test). The distribution of primary refractory (57% PMLCL, 44% DLBCL) and relapsed pts (43% PMLCL, 56% DLBCL) was similar (p=0.17). The ORR to first line salvage chemotherapy was 25% in PMLCL pts vs. 48% in DLBCL pts (p=0.01, chi-square test). 22% of PMLCL and 50% of DLBCL pts were able to proceed to ASCT after all salvage chemotherapy and PBSC mobilization (p=0.002, chi-square). At two years post diagnosis of relapsed or refractory disease, the OS of the PMLCL pts was 31% vs. 50% in DLBCL pts (p=0.03). For pts undergoing ASCT, the 2-year post-ASCT OS (67% PMLCL vs. 59% DLBCL, p=0.99) and PFS (57% PMLCL vs. 36% DLBCL, p=0.64) were similar. Multivariate analysis to determine predictors of survival after progression found that ECOG performance status (2–3 vs. 0, HR=2.9,p=0.01) and primary refractory disease (HR=2, p=0.003) were predictive of OS. Survival was not significantly different between PMLCL and DLBCL types (p=0.18). Multivariate analysis to determine predictors of ORR to salvage chemotherapy showed that ECOG (1 vs. 0, OR=2.6, 2–3 vs. 0, OR=2.6, p=0.05), primary refractory lymphoma (OR= 6.1, p<0.0001) and PMLCL type (OR =2.5, p=0.05) were predictive of lack of response to salvage chemotherapy.

CONCLUSIONS: In this “intent to transplant” analysis, relapsed or refractory PMLCL had an inferior ORR and survival from time of progression when compared to DLBCL. PMLCL pts who have chemosensitive disease have an OS and PFS post-ASCT that is similar to pts with DLBCL. Strategies for PMLCL should focus on improved salvage therapies for refractory and relapsed disease.