Diffuse large B-cell non-Hodgkin lymphoma is typically a chemotherapy-sensitive malignancy, justifying dose-intense therapy with hematopoietic stem cell transplantation for patients unlikely to achieve cure with standard-dose regimens.
Diffuse large B-cell non-Hodgkin lymphoma (DLBCL) is typically a chemotherapy-sensitive malignancy, justifying dose-intense therapy with hematopoietic stem cell transplantation (HSCT) for patients unlikely to achieve cure with standard-dose regimens. The value of autologous HSCT was demonstrated in the PARMA study, which randomized patients with relapsed but chemotherapy-sensitive large-cell lymphomas to salvage therapy consisting of several cycles of cisplatin/cytarabine/dexamethasone (DHAP) with or without dose-intense chemotherapy with autologous bone marrow transplantation.1 The actuarial probability of 5-year event-free survival (EFS) was 46% for patients randomized to transplantation compared with 12% for patients who received DHAP without dose-intense therapy. Patient eligibility criteria included age <60 years, no history of bone marrow or central nervous system involvement, and chemotherapy sensitivity; thus, transplantation was greatly restricted to a small subset of patients otherwise eligible for this treatment. Subsequently, most phase II studies demonstrated efficacy of dose-intense therapy for patients with chemorefractory disease or other adverse-risk features.2
The primary cause of failure of dose-intense therapy with autologous transplantation is posttransplant relapse, which raises the questions of patient eligibility and treatment strategies. The recently published CORAL study examined the outcome of transplantation for patients classified by the inclusion of rituximab in initial treatment and by duration of initial remission.3 EFS probabilities ranged from about 20% to 50% (duplicating the results of the PARMA study), with the highest probability enjoyed by patients who received initial chemotherapy regimens without rituximab and with an initial remission >12 months’ duration. Virtually all patients now referred for dose-intense therapy for recurrent or refractory DLBCL will have received rituximab, requiring patient-specific evaluation and treatment plans.
Chemotherapy—refractory disease continues to be a primary predictor for posttransplant relapse. Functional staging with PET scanning defines a group of patients with a higher probability of progression-free survival (PFS). A dramatic difference in survival was found in 1 study that reported a 81% probability of PFS for patients in complete response defined by PET scanning after pretransplant salvage therapy undergoing dose-intense therapy compared to 35% for patients not achieving a complete response (P = .003).4 Most clinical studies of autologous transplantation allowed only limited salvage therapy before the dose-intense cycle. What is not known is the effect of administering multiple salvage regimens in order to achieve a complete remission by PET criteria on transplant outcome, although in this study of PET staging a greater number of salvage regimens was correlated with lower PFS.4 PET negativity could be a surrogate for disease sensitivity to the initial salvage regimen, but otherwise be of no predictive value for transplant outcome. Conceivably, patients may lose the option of dose-intense therapy because of treatment-related toxicities if multiple cycles of salvage therapy are administered in an unsuccessful attempt to achieve PET negativity, and PET response should not be used as an exclusion criterion for dose-intense therapy. PET scanning should be a component of pretransplant evaluation, and, if serial salvage regimens are intended in order to achieve ultimate PET negativity, HSC should be collected earlier in the treatment strategy before severe hematological toxicity is incurred. Patients should be allowed to proceed to autologous (or allogeneic) HCT even if PGT positive, before nonhematological toxicities preclude these treatment options.
It is unlikely that the dose-intense transplant conditioning regimen can be escalated, especially for older patients and patients with comorbid conditions. There are no data to indicate that any particular pretransplant conditioning regimen achieves a better transplant out-come, and regimens used are likely to be transplant center—specific, with differences in reported results between centers likely affected by other factors, including patient selection criteria. However, pre- and posttransplant treatments may increase the probability of achieving durable remissions. Higher intensity pre-collection debulking/mobilization regimens may be important in reducing the risk of relapse. Cells can be collected with filgrastim with or without plerixafor mobilization, or with mild cyclophosphamide-based chemotherapy mobilization regimens. More aggressive mobilization regimens that will also achieve tumor debulking appear to improve the outcome of autologous transplantation, with some regimens approaching what would be considered dose-intense levels equivalent to tandem autologous transplantation.5 The CORAL study found no differences in transplant outcome between 2 commonly used, aggressive salvage regimens (R-ICE or R-DHAP).3 Administration of radioimmunoconjugates after HSC collection but before dose-intense therapy does not appear to increase the regimen-related toxicities of transplantation and also may reduce the risk of subsequent relapse.6 The use of these agents in the treatment of patients with relapsed DLBCL is the subject of a recently concluded phase III study that randomized patients to either rituximab or tositumomab (Bexxar) before autologous peripheral blood progenitor cell (PBPC) transplantation. The results of this study are currently being analyzed and have not yet been reported.
Maintenance therapy with posttransplant rituximab has been proposed as a strategy to reduce the risk of relapse after autologous transplantation, but raises the risk of delayed neutropenia. Preliminary publication of the CORAL study, which randomized patients to posttransplant rituximab maintenance or observation for patients with DLBCL, found no benefit from maintenance therapy.7 Posttransplant radiotherapy will benefit a subset of patients with localized bulky disease, resulting in a reported 10% improvement in posttransplant relapse of disease.8 The theoretical risk is a possibly increased incidence of secondary malignancies for patients receiving both chemotherapy and radiotherapy.
Autologous HSCT is the preferred treatment for patients with recurrent DLBCL because of the higher risks of transplant-related complications and mortality and the lower PFS after allogeneic transplantation. A review of registry data from the Center for International Blood and Marrow Transplant Research reported close to a 5-fold higher risk of transplant-related mortality (relative risk during the first 12 months after transplantation: 5.11; P <.0001) and lower survival (22% vs 49% at 5 yr) for patients undergoing sibling donor transplantation using myeloablative conditioning regimens compared to patients treated using autologous transplantation.9 Although the selection of patients for allogeneic transplantation could induce a selection bias in this treatment, matched-pair analysis in that study continued to show higher treatment-related mortality and lower PFS for the allograft recipient.
The development of reduced-intensity conditioning regimens greatly reduced the risks of immediate posttransplant toxicities to levels equivalent to the more dose-intense autologous transplant conditioning regimens.10 Graft-versus-host disease and relapse are the primary causes of treatment failure after transplantation using reduced-intensity conditioning, justifying novel approaches to separate the immunological benefits and disadvantages of allogeneic HSCT. The Stanford University transplant group reported the use of total-lymphoid irradiation with antithymocyte globulin in the treatment of 111 patients with lymphoid and myeloid malignancies achieving a nonrelapse mortality of less than 4% at 1 year after transplantation and a 3-year EFS probability of 40%.11 The intent of this approach was to develop a skewing of residual host T-cell subsets to favor regulatory natural killer T cells. Investigators from the National Institutes of Health are exploring immunodepletion with cycles of EPOCH with fludarabine and rituximab to deplete host CD4 cells before transplantation using a nonmyeloablative transplant- conditioning regimen of cyclophosphamide and fludarabine.12 In preliminary analysis of 143 patients with high-risk lymphoid malignancies who underwent transplantation, patients who achieved a complete response or partial response after EPOCH-FR (42%) enjoyed a median EFS of 77.4 months and overall survival of 98.5 months. This compared to 4.8 months and 16.2 months, respectively, for patients not responding to this salvage regimen; chemotherapy sensitivity remains a predictor for transplant outcomes for allogeneic transplantation.
Patients with relapsed or refractory DLBCL should be referred to a transplant program to determine eligibility for HSCT, regardless of age, initial remission-induction therapy, or duration of initial remission. Older patients will be excluded from allografting by Medicare rules. For patients considered eligible for autologous or allogeneic HSCT, a comprehensive individual treatment strategy should be developed, including appropriate salvage chemotherapy regimen(s), use of PET scanning to define response, collection of HSC before serious marrow toxicity is incurred, and initiation of dose-intense chemotherapy in a timely manner. Pre- or posttransplant strategies such as administration of a radioimmunoconjugate or involved field radiation/surgical extirpation of residual disease should be components of the treatment plan.
Address correspondence to:
Scott D. Rowley, MD, Adult Blood and Marrow Stem Cell & Transplantation Program, John Theurer Cancer Center at Hackensack University Medical Center, 92 Second St, Hackensack, NJ 07601, or e-mail [email protected].