High-dose chemotherapy in lymphomas, and mainly non-Hodgkin’s lymphomas, has been advancing since the 1970s. This therapeutic strategy is based on the supposed existence of a dose-response curve for cytotoxic agents. However, the available data are contradictory, so high-dose chemotherapy cannot be guaranteed as consolidation treatment for first-remission follicular lymphoma or diffuse large cell lymphoma. The objective of this paper is to review the current knowledge about high-dose chemotherapy followed by hematopoietic stem cell transplantation in follicular non-Hodgkin’s lymphoma. The published studies on follicular lymphoma after first remission, recurrent follicular lymphoma, and transformed follicular lymphoma were assessed together with the data available on diffuse large cell lymphoma. During analysis of the studies, difficulties were encountered in comparing studies due to the heterogeneous nature of the data. High-dose chemotherapy as consolidation treatment after first remission or in recurrent or refractory disease was also analyzed.
Tumori 2015; 101(1): 2 - 7
Article Type: ORIGINAL RESEARCH ARTICLE
AuthorsIgnacio García Escobar, Blanca Cantos Sánchez de Ibarg?en, Virginia Calvo de Juan, C. Maximiano Alonso, Miriam Méndez García, Antonio Carlos Sánchez Ruíz, Mariano Provencio Pulla
- • Accepted on 09/09/2014
- • Available online on 29/01/2015
This article is available as full text PDF.
High-dose chemotherapy in lymphomas has been advancing since the 1970s, mainly in non-Hodgkin’s lymphomas (NHL) and especially follicular lymphoma (FL) and diffuse large B cell lymphoma (DLBCL). The basic principle underlying high-dose chemotherapy is the supposed existence of a dose-response curve for cytotoxic agents with a short half-life and fundamentally medullar toxicity that permit subsequent hematological salvage.
Considering the 2 forms of hematopoietic stem cell transplant (HSCT), autologous HSCT (auto-HSCT) is not usually curative but may result in prolonged remissions, while allogeneic HSCT (allo-HSCT) may cure but is associated with high treatment-related mortality (TRM).
FL is the second most common type of NHL and accounts for between 15% and 30% of all lymphomas. Individual prognoses are difficult to make, although there have been some advances, especially after publication of the Follicular Lymphoma International Prognostic Index (FLIPI) (1).
The initial treatment of FL depends on the stage of the disease at presentation. Patients with early-stage disease may be cured with radiation therapy while patients with advanced stages are initially managed with an immunotherapy-based regimen (2). Defining the role of HSCT in the therapeutic algorithm is one of the major challenges in the management of this disease – especially in the rituximab era – and is the subject of ongoing clinical trials.
However, when we analyzed and compared the studies carried out to date, we encountered various difficulties due to differing selection criteria and conditioning or follow-up regimens.
Considering that the ideal timing of HSCT in FL is unknown, this review focuses on the current situation regarding the indications for autologous and allogeneic HSCT.
Role of high doses of chemotherapy to consolidate the first remission
Several phase II trials have been published, with most data deriving from pre-rituximab studies in which high-dose chemotherapy followed by hematopoietic stem cell transplantation improved relapse free survival compared with historic controls (3-4-5-6-7-8).
On the basis of these data, 4 multicenter randomized trials were initiated, 3 in the pre-rituximab era and 1 in the rituximab era. The German Low-Grade Lymphoma Study Group (GLSG), reported on by Lenz et al (9), studied the value of high-dose versus maintenance interferon in patients under 60 years of age who were in partial or complete first remission after 2 cycles of cyclophosphamide-doxorubicin-vincristine-prednisone (CHOP) or mitoxantrone-chlorambucil-prednisone (MCF). A total of 240 patients were included. With a mean follow-up of 4.2 years, 31 relapses (27.2%) were observed in the high-dose arm versus 76 (60.3%) in the maintenance interferon arm, with 5-year progression-free survival (PFS) rates of 64.7% and 33.3%, respectively (p<0.0001), but without sufficient follow-up to detect differences in overall survival (OS).
Similar results were published by Gyan et al (10) from the French Groupe Ouest-Est des Leucémies et Autres Maladies du Sang (GOELAMS). Out of 172 patients randomized to receive high-dose chemotherapy followed by autologous purged stem cell transplantation or immunochemotherapy with cyclophosphamide, doxorubicin, teniposide, prednisone and interferon, higher response rates (69% vs 81%, p = 0.045) and longer 9-year progression-free survival (64% vs 39%; p = 0.004) were reported for patients receiving high-dose therapy than those receiving conventional treatment; the patients with worse FLIPI benefited most from high doses. A PFS plateau was observed in the high-dose-therapy group after 7 years, suggesting that in a subgroup of patients FL may be cured by auto-HSCT. However, no differences were observed in OS, due in part to excess mortality owing to the development of second tumors in patients given high doses.
This aspect was also indicated in the monitoring of GLSG patients, especially secondary to myelodysplastic syndromes (sMDS) and acute myeloid leukemia (AML), with an estimated risk at 5 years of 3.8% versus 0% in the interferon arm (11).
Moreover, the Groupe D’Etude des Lymphomes de l’Adulte (GELA) trial (12), which included 401 patients, did not find any advantage for those undergoing intensive treatment. Patients received 12 cycles of cyclophosphamide, doxorubicin, teniposide, and prednisone (CHVP) plus interferon alpha for 18 months, or 4 courses of CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone), followed by high-dose therapy with total-body irradiation and autologous stem cell transplantation. An intent-to-treat analysis after a median follow-up of 7.5 years showed that there were no differences in event-free survival (EFS) (p = 0.11) or OS (p = 0.53). This study showed that there was no statistically significant benefit in favor of first-line high-dose therapy in patients with FL.
To address the role of auto-HSCT in upfront consolidation of FL in the rituximab era, Ladetto et al of the Italian Gruppo Italiano Trapianto di Midollo Osseo (GITMO/IIL) published a phase III trial comparing a rituximab-supplement version of high-dose sequential chemotherapy (R-HDT) with 6 cycles of CHOP supplemented by an identical number of rituximab courses (CHOP-R) in 136 high-risk patients with FL (13). Intention to treat was analyzed with EFS as the primary endpoint. At a median follow-up of 51 months, the 4-year EFS rates were 28% and 61%, respectively (p<0.001), with no statistically significant difference in OS between the 2 arms (p = 0.96). Molecular remission was achieved in 44% of the patients treated with CHOP-R and 80% in those treated with R-HDT (p<0.001), and was the strongest independent predictor of outcome.
In conclusion, a significant increase in PFS was demonstrated in favor of auto-HSCT, but to date no benefit in OS has been reported for autografting FL patients in first remission, with similar PFS results in patients receiving first-line rituximab chemoimmunotherapy versus those in the auto-HSCT arm. Therefore, taking into consideration strategies such as consolidation with radio-immunotherapy (16), rituximab maintenance (17), and/or rituximab retreatment (18), and in accordance with international guidelines (19, 20), routine use of auto-HSCT in first remission for FL patients should not be considered the first option. However, this is a field for clinical research seeking to reduce relapses, define better conditioning, select patients who can benefit from high doses, and reduce the risk of secondary neoplasms.
Role of allogeneic HSCT in first remission
Allo-HSCT is a potentially curative treatment modality for FL patients who would be considered incurable with conventional chemoimmunotherapy or auto-HSCT treatments in light of the lymphoma-free constitution of the graft and the graft-versus-lymphoma (GVL) immunological effect. Unfortunately, there are no randomized controlled trials to support its use for chemosensitive FL patients in first remission, and its indication is only supported by data from single institutions in small groups of high-risk FL patients with primary refractory disease despite multiple lines of treatment (21, 22). Due to its high toxicity, TRM is one of the major limitations of this procedure. However, with the introduction of reduced-intensity conditioning (RIC) regimens, this procedure will be considered more as a future treatment option. So that, according to existing evidence, currently allo-HSCT should only be offered in the context of clinical trials for this group of refractory FL patients in first remission.
Role of high doses in recurrent FL
Auto-HSCT has a better established role in patients with relapsed chemosensitive FL. The 5-year PFS ranges from 40% to 50% in most prospective and retrospectives studies (8, 23-24-25-26-27-28-28). As a result, several prognostic factors are being studied to determine their impact on the outcome in these patients. In this context, the number of chemotherapy regimens (especially in patients treated with more than 3 regimens), chemosensitivity, FLIPI at the time of transplant, and histological grade have to be considered (8, 23-24-25-26-27-28-29-30).
In the pre-rituximab era, Montoto et al (8) reported on a retrospective series from the European Group for Blood and Marrow Transplantation (EBMT) including 693 FL patients who underwent auto-HSCT. The 10-year PFS and OS were 31% and 52%, respectively, with a non-relapse mortality rate of 9%. In the multivariate analysis, advanced age, chemoresistant disease and conditioning regimens with total-body irradiation were correlated with inferior survival. Similarly, in a series of 241 FL patients, Kornacker et al (27) reported 10-year OS and PFS rates of 75% and 49%, respectively. Finally, Schouten et al (25) published the results of the CUP trial, the only randomized prospective study that compared auto-HSCT with standard therapy in relapsed FL patients. One hundred and forty patients were randomized to receive conventional chemotherapy, auto-HSCT with a purged graft, or auto-HSCT with an unpurged graft. The 4-year OS rates were 46%, 71% and 77%, respectively, with the limitation that the sample size was too small to extract definitive conclusions. The trial was closed prematurely due to low recruitment.
In the rituximab era, the GELA/GOELMANS group compared the impact of auto-HSCT with conventional treatment in 175 FL patients at first relapse (28), of whom 40% had received rituximab. The 3-year OS rate was significantly higher for the HSCT group (92% vs 63%, p = 0.0003). The study was limited by its retrospective nature and the fact that it was probably affected by a selection bias because only patients who were chemosensitive to salvage therapy underwent HSCT. In addition, Sebban et al (12) reported an analysis of the results obtained in groups of FL patients treated with various salvage therapies (GELF) including auto-HSCT. In patients treated with rituximab, no statistical differences were observed in PFS or OS when comparing groups with and without auto-HSCT. Similar to these authors, Rohatiner et al (24) published the results of a retrospective analysis which included 121 patients who received myeloablative therapy supported by autologous HSCT as consolidation of second or subsequent remissions. The survival of patients treated in second remission was significantly longer than the survival of patients treated later in the course of the lymphoma.
Therefore, in conclusion, it is difficult to assess which patients are better candidates for auto-HSCT. It seems clear that in chemoresistant patients, those treated with several lines of chemotherapy, patients with a poor performance status or those older than 70 years, auto-HSCT should be considered with caution. On the other hand, taking the consensus of international experts into consideration (31), it is probably a good alternative in first relapse with chemosensitive disease in patients with a short response duration to conventional immunochemotherapy (less than 3 years), patients with high FLIPI at relapse, or in order to maintain chemosensitivity in further relapses.
Several studies have explored the role of ex vivo purging (monoclonal antibodies, CD34+ cell selection, etc.) (32, 33) and
Rituximab maintenance after auto-HSCT
There is evidence that relapsed FL patients could benefit from rituximab maintenance after autologous transplant. Phase II trials demonstrated the effectiveness and safety of maintenance therapy, and showed that it could eradicate minimal residual disease that persists after auto-HSCT (37, 40). Furthermore, a large randomized study from the EBMT showed significantly higher PFS rates in patients treated with maintenance rituximab (41). Therefore, there are data supporting this strategy, although it is not clear whether OS is improved and this continues to be an area of active investigation. Rituximab maintenance cannot yet be considered a standard option.
Role of myeloablative allo-HSCT in relapsed FL
Allo-HSCT is a curative modality for patients with FL. Indolent lymphomas are among the histological types most sensitive to the effect of GVL (42). This aspect can be of special relevance in disease regression after relapse, for decreasing doses of immunosuppression treatment, or after leukocyte infusions (43-44-45). When comparing auto and allo procedures, relapse rates are clearly favorable for allo- HSCT (25% vs 55% at 5 years), and the available data show a relapse-risk plateau 2-3 years after allo-HSCT, therefore suggesting the probability of cure in a proportion of LF patients (46, 47). Along these lines, 3 retrospective studies reported durable remissions with 5-year EFS rates ranging from 45% to 75% (48-49-50). However, no differences in OS were found, probably due to the high rate of TRM (35%-40%). In 2008 Kuruvilla et al (49) reported a 15% reduction in TRM, which was probably due to the chemosensitivity of the series and the use of syngeneic HSCT in 10% of the treated patients.
Role of RIC allo-HSCT for relapsed FL
While the effect of conventional allo-HSCT is based to a large extent on cytoreduction of high-dose chemotherapy, RIC-HSCT uses a donor-mediated GVL effect. The objective is to reduce TRM and extend the transplant age to more than 70 years.
Khouri et al (51) published the prospective trial with the longest follow-up of 107 months. The 47 patients with relapsed FL received fludarabine, cyclophosphamide and rituximab (FCR); high-dose rituximab was added at a maximum dose of 1000 mg/m2 in 75% of cycles to prolong serum concentrations. The 11-year EFS and OS rates were 72% and 78%, respectively. The effects of this high rituximab dose should be considered for its probable regulation of GVL or acute graft-versus-host disease (GVHD) (56).
The Blood and Marrow Transplant Clinical Network (BMT CTN) (Protocol 0701) is currently conducting a multicenter phase II trial of RIC for patients with relapsed FL after first complete response, with PFS at 2 years from time of transplant as the primary endpoint. Final data collection and primary outcome measurement are programmed for November 2014.
With regard to the origin of donors, some authors have proposed strategies to improve the high rates of GVHD associated with unrelated donors. Thomson et al (53) employed
Taking these findings into consideration, the customary and appropriate question is now, Which is the correct transplant to offer relapsed FL patients? Unfortunately, there are no prospective trials comparing these options. The BMT CTN 0202 trial was closed early due to poor accrual. The 3-year PFS and OS were 63% vs 86% and 73% vs 100% for auto-HCST vs allo-HSCT, respectively. Cohen et al (57) recently reported a 96% 3-year PFS and OS with auto-HSCT followed by tandem RIC allo-HSCT. The superiority or not of this strategy is currently unknown and is included as part of an investigational strategy. Nevertheless, current evidence shows a low rate of TRM, a low risk of relapse, and no risk of sMDS/AML for RIC allo-HSCT (53, 54, 58) compared to auto-HSCT, so it could be considered a good curative-intent treatment option for clinically fit FL patients with a suitable donor and chemosensitive disease after several lines of treatment. Nonetheless, auto-HSCT could be reserved for patients unfit for allograft or without a suitable donor, with the disease in these cases being considered incurable.
Role of high doses in transformed FL
Over time a certain proportion of FL patients suffer a transformation to more aggressive histological forms. It is difficult to measure the exact percentage of patients in whom this occurs, but the most important studies calculate that around 26%-60% of patients will develop histological transformation (2). These patients have a very poor short-term prognosis, with a mean survival of 11 to 28 months. Transformed FL is usually treated as if it was an aggressive de novo lymphoma. Studies evaluating the effect of high-dose chemotherapy on relapses are few and had low numbers of patients (59). According to data from the European Bone Marrow Transplant Registry (60), disease-free survival at 5 years in 50 patients was 30%. When only patients with chemosensitive disease were considered, the figure was 69%, with no long-term survivors among those with chemoresistant disease. Similarly, no benefit was found for patients over 60 years of age (61). In any case, the poor prognosis of these patients suggests that high-dose chemotherapy could be a reasonable option for a selected subgroup of patients.
The optimal timing of auto-HSCT in patients with FL remains controversial. The results of 3 large randomized studies show no benefit of auto-HSCT as consolidation therapy in first complete remission. The same conclusion can be drawn when considering allo-HSCT outside clinical trials.
For patients with relapsed disease, it is difficult to define a group of better candidates for auto-HSCT. It is probably a good alternative in first relapse for patients with chemosensitive disease and with a short response duration to conventional immunochemotherapy (less than 3 years), especially those with high FLIPI at relapse. In that sense, strategies such as ex
- Escobar, Ignacio García [PubMed] [Google Scholar] , * Corresponding Author (firstname.lastname@example.org)
- de Ibarg?en, Blanca Cantos Sánchez [PubMed] [Google Scholar]
- de Juan, Virginia Calvo [PubMed] [Google Scholar]
- Alonso, C. Maximiano [PubMed] [Google Scholar]
- García, Miriam Méndez [PubMed] [Google Scholar]
- Ruíz, Antonio Carlos Sánchez [PubMed] [Google Scholar]
- Pulla, Mariano Provencio [PubMed] [Google Scholar]
Medical Oncology Department, Puerta de Hierro-Majadahonda Universitary Hospital, Madrid - Spain