Allogeneic transplant for peripheral T-cell lymphoma: a sparkle of hope and many questions

Peripheral T-cell lymphoma (PTCL) comprises a heterogeneous group of diseases with overall aggressive clinical course and dismal prognosis on average [1,2]. For subtypes other than ALK-positive anaplastic large cell lymphoma (ALCL), the application of standard lymphoma regimens (e.g. cyclophosphamide, doxorubicin, vincristine, and prednisone [CHOP]) results in low remission rates and unacceptably high rates of recurrence of progression. Other combinations are under investigation, including those incorporating etoposide and/or alemtuzumab [3,4], but progress is hampered by the relative rarity of PTCL.

Given these findings, autologous stem cell transplant (auto-SCT) has been employed as a consolidative therapy, either up-front or at relapse. Several small prospective trials have been completed [5–9 ] in the front-line setting. Outcomes (progression-free survival [PFS], overall survival [OS]) in some were not substantially different compared to chemotherapy alone. A significant selection bias complicates interpretation in those trials with superior outcomes. Paralleling its application in B-cell non-Hodgkin lymphoma (B-NHL), auto-SCT has been utilized in the relapsed/refractory setting, but outcomes were generally unsatisfactory in retrospective trials [10,11]; no prospective data exist.

Allogeneic stem cell transplant (allo-SCT) has also been proposed as a potentially curative therapy for relapsed PTCL, bolstered by optimism for a ‘graft-versus-lymphoma’ (GVL) effect. The largest study to date, a retrospective analysis of 77 patients, demonstrated a 5-year OS of 57%, with significant differences in outcomes between T-NHL subtypes: patients with angioimmunoblastic T cell lymphoma (AITL) had the best outcomes (5-year OS, 80%), followed by those with PTCL and ALCL (5-year OS, 58% and 48%, respectively) [12]. Patients achieving a complete or partial response prior to transplant had substantially better OS compared to those who did not (69% vs. 29%, p = 0.04). These survival numbers are offset by a 33% transplant-related mortality (TRM). Acute graft-versus-host disease (GVHD) was an adverse prognostic factor, but the authors posited a GVL effect based on two patients who achieved a second complete response after donor lymphocyte infusion (DLI). A second retrospective study incorporating patients with PTCL reported three patients with a response to DLI [13], but similarly indicated an adverse effect of GVHD as well as a TRM of 42%.

This combination of high TRM and purported GVL effect has stimulated interest in reduced-intensity conditioning (RIC) allo-SCT. The only prospective trial reported thus far using allo-SCT in PTCL analyzed this approach. Corradini et al. reported a 3-year OS of 81% with a corresponding TRM of only 6% in 17 patients with relapsed PTCL [14]. Additionally, two patients had a response to DLI or withdrawal of immunosuppression. These outcomes are encouraging; however, all patients but one had demonstrated chemosensitivity, challenging the overall applicability of these results in a disease for which attaining a remission remains a major challenge. Moreover, a recent retrospective analysis of 126 patients undergoing allo-SCT found no differences in PFS, OS, or TRM comparing patients receiving myeloablative versus reduced-intensity conditioning [15].

Against this confusing backdrop, the article by Zain et al. [16], in this issue of Leukemia and Lymphoma, provides useful additional information. They report on a single-center study of 24 patients undergoing allo-SCT for PTCL (they also included 13 patients with cutaneous T-cell lymphomas, either mycosis fungoides or Sezary syndrome). The study population is notable in that 17 patients, or 46%, had primary induction failure indicating chemoresistant disease. Both myeloablative (13 patients) and RIC (24 patients) regimens were used. The 5-year PFS, OS, and TRM are 52.2%, 46.8%, and 27%, respectively. For those patients with active disease at transplant, the 5-year OS was 42%, compared to 73% for those in complete or partial remission (CR/PR). There was no difference in outcomes between myeloablative and non-myeloablative conditioning regimens.

Although this study does not report on DLI or withdrawal of immunosuppression, the authors probed the potential role of a GVL effect, using GVHD as a surrogate measure. By utilizing a time-varying covariate in their analysis, they obviated a potential source of bias which may have compromised previous investigations: namely, patients have to survive some length of time in order to develop GVHD, which may lend the illusion that those who develop GVHD live longer compared to those who do not. They found no impact of chronic or acute GVHD on PFS/OS outcomes. Indeed, although the documented cases of responses to DLI constitute proof of GVL in lymphoma, the overall contribution of GVL effects to the outcome of allogeneic SCT in patients with lymphoma remains in doubt. A comparison of syngeneic SCT with allogeneic SCT in non-Hodgkin lymphoma indicates similar rates of relapse [17], suggesting that the major benefit of allo-SCT may be the avoidance of occult tumor contamination in the graft.

Regardless of the mechanism governing allogeneic transplant effects, this study injects significant information into the controversy surrounding optimal treatment of patients with PTCL. As the authors indicate in their discussion, a 43% 5-year OS for chemoresistant patients compares favorably to previous reports, and implies that there is hope for disease control even in patients with very high-risk disease. It is unclear how the addition of patients with cutaneous T-cell lymphoma (CTCL) affected this study; these patients may have improved outcomes after allo-SCT [18]. The debate is not yet settled, and further prospective data are needed. Additionally, the role of a host of novel agents will need to be settled; if initial response rates may be augmented, accumulated data indicate that outcomes after consolidative transplant (auto or allo) may be improved as well.

Potential conflict of interest:

Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.