Response to Rituximab Induction Is a Predictive Biomarker in Post-Transplant Lymphoproliferative Disorder (PTLD) and Allows Successful Treatment Stratification in an International Phase II Trial Including 152 Patients

Background: The PTLD-1 trial has demonstrated the efficacy and safety of 4 cycles of weekly rituximab followed by 4 cycles of CHOP-21 + G-CSF in CD20-positive PTLD after solid organ transplantation. Median overall survival (OS) was 6.6 years, a clear improvement over the preceding rituximab monotherapy trials (2.4 years). However, response to rituximab induction predicted OS after completion of therapy. Based on the hypothesis that rituximab consolidation might be sufficient treatment for patients already in a complete response (CR) after rituximab induction, trial treatment was changed in 2007 through a protocol amendment introducing risk-stratified sequential treatment (RSST): rituximab consolidation for patients in CR after rituximab induction and R-CHOP-21 consolidation for all others.

Methods: In this international, multicenter phase II trial (PTLD-1, 3^rd amendment; NCT00590447), treatment-naïve adult solid organ transplant recipients diagnosed with CD20-positive PTLD were treated with rituximab (375 mg/m² IV) on days 1, 8, 15 and 22. After restaging, patients in CR continued with four three-weekly courses of rituximab monotherapy while all others received 4 cycles of R-CHOP-21 + G-CSF. In case of disease progression during rituximab monotherapy R-CHOP was commenced immediately. The primary endpoint was treatment efficacy measured as response rates and response duration. Analysis was by intention to treat. This is the final analysis of 152 patients treated with RSST from 2007 to 2014 at centers in Germany (72), Belgium (36), France (24), Australia (7), Poland (7) and Italy (6) with a median follow-up of 4.5 years. The 70 patients treated with rituximab followed by CHOP-21 in the original PTLD-1 trial (median follow-up 5.1 years) served as a control population. Inclusion criteria and follow-up schedule were identical; there were no significant differences in the transplant- and lymphoma-related baseline factors listed below.

Results: 115/152 patients were male. 69/152 were kidney, 40 liver, 18 lung, 15 heart, 5 heart/kidney, 3 kidney/pancreas and 2 heart/lung transplant recipients. Median age at diagnosis was 56 years. PTLD was late (> 1 year after transplantation) in 120/152 (79%) of patients. 67/145 (46%) PTLD were EBV-associated. 130/152 patients had monomorphic, 20 polymorphic and 2 early lesion PTLD. The overall response rate (ORR) was 111/126 (88%, CR: 88/126 [70%]). Median duration of remission (DR) was not reached; the 3-year Kaplan-Meier estimate was 82% (compared to 71% in PTLD-1). In the intention-to-treat population (152 patients), the median time to progression (TTP) was not reached either. The 3-year Kaplan-Meier estimate was 78% (69% in PTLD-1). Median OS by intention-to-treat was 6.6 years (95% CI 5.5 - 7.6) with a 3-year estimate of 70% in comparison to 61% in PTLD-1. There was no significant difference in ORR, DR, TTP or OS between EBV-positive and EBV-negative PTLD. On the other hand, response to 4 applications of rituximab was a highly significant predictor of OS, TTP and progression-free survival (PFS) despite treatment stratification (all p<0.001).

37/148 patients (25%) achieved CR with 4 cycles of rituximab and were allocated to rituximab consolidation. In this group, TTP in the intention-to-treat population was significantly longer than in the corresponding group in the PTLD-1 trial (37 patients versus 14 patients, p<0.05). In the 111 patients allocated to R-CHOP consolidation, ORR was 78/92 (85%) with 55/92 (60%) complete remissions (89% and 60%, respectively, in PTLD-1). Median TTP was not reached, the 3-year estimate was 73% (69% in PTLD-1). In patients refractory to rituximab induction, the CR rate was 22/38 (58%) with R-CHOP compared to 3/11 (27%) in PTLD-1 with CHOP (p=0.07); median PFS was 1.4 years versus 0.3 years in PTLD-1, p<0.05. The frequency of grade 3/4 leukopenia and infections was 63% and 34%, respectively. Treatment-related mortality occurred in 7%.

Conclusions: This largest trial cohort in PTLD to date demonstrates for the first time that treatment stratification by response to rituximab induction is feasible, safe and effective. Rituximab consolidation in early rituximab responders results in significantly better disease control compared to CHOP consolidation. The addition of rituximab to CHOP chemotherapy improves outcome in patients refractory to rituximab monotherapy.