Electronic Letters to:
|
|
E-letters: Electronic letters published:
-
![[Read eLetter]](/icons/misc/sectionDOWN.gif)
Reply: Poor results and patient selection bias
- Daniel C Betticher, Giovanni Martinelli, Thomas Cerny, David C Linch (13 October 2006)
-
Poor results and patient selection bias
- Alessandro M Gianni, Corrado Tarella (10 October 2006)
|
|
|||
|
Daniel C Betticher, Médecin-Chef d'Oncologie Hôpital Cantonal, 1708 Fribourg, Giovanni Martinelli, Thomas Cerny, David C Linch
Send e-letter to journal:
|
We thank Gianni and Tarella for their letter criticising certain aspects of the Mistral trial, which failed to reproduce the improved survival with high dose sequential chemotherapy and autologous stem cell transplantation (HDS) as compared to standard CHOP. 1. We agree with Gianni and Tarella that the patients included in the Mistral trial had very aggressive lymphomas. This might explain our low response rate, but does not negate our results: HDS is not better than CHOP in our selected patients. We included patients with bone marrow involvement at presentation as we proceeded to stem cell harvest after an initial cycle of CHOP, one cycle of high dose cyclophosphamide, one cycle of high dose MTX and one cycle of VP16 (4 cycles of therapy) at which time the marrow was morphologically clear in the harvested patients. We agree that more exacting determination of lymphomatous contamination of the stem graft would now be possible but the Mistral trial was designed over 10 years ago. 2. We agree also with the authors that anaplastic large cell lymphomas are a specific lymphoma entity, but there was no ‘a priori’ reason to believe that these types of lymphoma would not have the potential to benefit from HDS. Furthermore, we do not believe that the inclusion of these patients influenced our results as exploratory subgroup analysis has not revealed any heterogeneity. 3. As mentioned in our publication, the trial was closed in 2003 after a planned interim analysis based on a review of preliminary efficacy and safety results. The independent advisory board, consisting of two clinicians and one biostatistician, not involved in the trial, recommended stopping the trial early, primarily due to the absence of a likelihood of detecting a relevant difference in OS between the treatment arms, with the inevitably increased toxicity (and expense) in the HDS arm. 4. The low rate of enrolment was not the reason for closing the trial, although continuation of the trial would have required the addition of rituximab to both arms as CHOP alone had ceased to be the ‘gold standard’. 5. Finally, we agree with the authors that patient selection has the potential to markedly influence outcome, but this merely emphasizes the importance of performing phase III randomized trials. We are thankful to the editor of Annals of Oncology for having accepted for publication the long awaited results of the Mistral trial even though the results were negative and did not support the earlier trial by Gianni et al. Conflict of Interest:None declared |
|||
|
|
|||
|
Alessandro M Gianni, Director, Department of Leukemia and Lymphoma Research Istituto Nazionale Tumori, 20133 Milano, Italy, Corrado Tarella
Send e-letter to journal:
|
The paper entitled “Sequential high-dose chemotherapy as initial treatment for aggressive sub-types of non-Hodgkin Lymphoma: results of the international randomized phase III trial” by Betticher DC et al., published in the October issue of Annals of Oncology, reports the results of a randomized study known as Mistral trial, comparing CHOP and the high-dose sequential chemotherapy (HDS) regimen as first-line therapy for younger patients with aggressive lymphoma [1]. An overall poor response and a startling dismal survival are reported in both treatment arms, without any significant difference between CHOP and HDS. Indeed, the conclusive point made by the Authors is that HDS is not superior to CHOP, thus questioning the results of our previous study published in 1997, indicating a better outcome following HDS vs. MACOP-B [2]. We believe that some features of the study design may justify the very poor results of the MISTRAL trial, and that several weak points make the comparison between CHOP and HDS unreliable. We wish to highlight the following biases: 1. a high proportion of patients (23%) had Bone Marrow (BM) involvement. The use of an autograft-based program for BM+ aggressive lymphoma, without any assessment of the level of tumor contamination in the graft, exposes these patients to a high-risk of receiving residual neoplastic cells with the graft. This risk is particularly harmful due to their aggressive histology. Indeed, in our previous reports, where HDS was still given without Rituximab, DLCL patients presenting with BM involvement were rigorously excluded from the trial. In order to reduce the risk of tumor contamination, the HDS program was modified in the MISTRAL study by postponing stem cell harvest following hd-VP16 in BM+ patients. It is quite doubtful that the only addition of hd-VP16 could eradicate marrow disease in patients with aggressive lymphoma. In the pre-Rituximab era, we have employed the HDS approach in BM+ low grade lymphoma: however, in those studies, harvesting was scheduled after several chemotherapy courses (2 APO + 2 DHAP + hd-VP16 + hd-Dex + hd-CY), along with a careful molecular monitoring of tumor cell contamination [3,4]. Last, but not least, the program was designed for indolent, not for aggressive lymphoma. The original HDS was not designed to be used in BM+ve DLCL patients, since the risk of tumor cell reinfusion, in the absence of a reliable and sensitive molecular assay, was deemed ethically unacceptable in a treatment-naïve population. This major departure from the original design weakens any comparison between the two studies; 2. Furthermore, a subgroup of patients as high as 11% (15% CHOP and 7% HDS arms) had a histology of anaplastic large cell lymphoma (ALCL), which in published series accounts for approximately 2% of NHL's; unfortunately, the ALK status (positive or negative) of this subgroup is not detailed. The efficacy of the autograft approach in aggressive non-B cell lymphoma is questionable, particularly in ALK-negative ALCL [5]. Again, the unproven efficacy of autograft in non-B-cell subtypes led us to exclude those subtypes in our original trial comparing HDS vs MACOP-B. In the Mistral study, the inclusion of ALCL of undetermined ALK status, with unbalanced distribution of patients between the two arms, makes any comparison futile; 3. patient enrollment was closed after 6.5 years, before completing the accrual of the planned sample size; occurrence of severe toxicities in the HDS program was among the reasons reported by the Authors for the early stopping of the study protocol. The most relevant HDS toxicities compared to the CHOP regimen were stomatitis and myelotoxicity, as pointed out in page 4 (“toxicity”). Indeed, stomatitis and myelotoxicity are two universal side effects in patients undergoing intensive chemotherapy with autograft. These toxicities had to be taken into account when the study was designed. Thus, it is somehow bizarre that stomatitis and myelotoxicity contributed to the decision to stop the trial; 4. a major reason for protocol discontinuation seems the very low rate of enrollment. As mentioned in page 2 (“trial design”): “The trial was initially activated in Switzerland and at the EIO (European Institute of Oncology, Milano, Italy). Through the continuous effort to improve accrual rate, several centers joined the trial later (notably, the German High-Grade Non-Hodgkin’s Lymphoma Study Group, and the UKCCCR lymphoma subcommittee)”. In spite of the several Centers involved in the protocol, a total of 129 patients were enrolled and found eligible to the trial over 6.5 years. It is quite unlikely that such a group of well known specialists was unable to enroll less than 20 DLCL per year. One is tempted to think that the only explanation for such a remarkably slow accrual rate might be patient selection. In other words, it is quite doubtful that all DLCL patients eligible for the trial have been enrolled by the several Swiss-German-UK Centers participating to the study; 5. a wild patient selection seems the most likely explanation also for the very poor results observed. In fact, besides the 34% CR rate for patients receiving HDS, the 37% CR rate reported with CHOP is among the lowest CR rates ever reported in the literature for DLCL. For instance, the CR rates reported in the original age-adjusted IPI study were 57% and 46% for high-intermediate and high prognostic scores, respectively. Furthermore, the study included 21% of patients with a low-intermediate score. In the IPI study this good prognostic group had a CR rate as high as 71%, thus rendering the 37% figure even more baffling. In conclusion, the over-representation of patients with bone marrow involvement, the large proportion of non-B cell hystotypes, and the disappointing response reported in the Mistral study by well recognized specialists in the field of lymphoma, all point to a peculiar selection of patients with unusually poor prognostic presentation, and with characteristics that had been excluded from our original study for both scientific and ethical reasons. In conclusion, any comparison between the present MISTRAL trial and our previously reported study appears inappropriate and misleading. The characteristics of the patients studied are unique in terms of histology and clinical presentation, as well as in terms of responsiveness to standard, conventional chemotherapy regimens. It is surprising that the reviewers of this distinguished journal accepted the manuscript in its present form, without requiring a thorough discussion of the critical points raised above, dealing with an exceedingly slow accrual rate (in particular for such a large cooperative study), with the very poor results in the control arm, and the unstratified inclusion of disparate disease entities lumped together as ‘aggressive lymphoma’ The latter procedure, hardly acceptable when the study was designed, makes the paper of very limited value today. With regards, Alessandro M Gianni and Corrado Tarella References 1. Betticher D, Martinelli G, Radford J, Kaufmann M, Dyer M, Kaiser U, Aulitzky W, Beck J, von Rohr A, Kovascovics T, Cogliatti S, Cina S, Maibach R, Cerny T, Linch D. Sequential high dose chemotherapy as initial treatment for aggressive sub-types of Non-Hodgkin Lymphoma: results of the international randomized phase III trial (MISTRAL). Ann Oncol. 2006 Oct;17(10):1546-52 2. Gianni AM, Bregni M, Siena S, Brambilla, Di Nicola M, Lombardi F, Gandola F, Tarella C, Pileri A, Ravagnani F, Valagussa P, Bonadonna G. High- dose chemotherapy and autologous bone marrow transplantation compared with MACOP-B in aggressive B-cell lymphoma. New Engl J Med 1997, 336: 1290-1297 3. Tarella C, Zallio F, Caracciolo D, Cherasco C, Bondesan P, Gavarotti P, Corradini P, Tassi V, Pileri A. Hemopoietic progenitor cell mobilization and harvest following an intensive chemotherapy debulkying in indolent lymphoma patients. Stem Cells 1999, 17: 55-61 4. Corradini P, Astolfi M, Cherasco C, Caracciolo A, Pileri A, Tarella C. Molecular monitoring of minimal residual disease in follicular and mantle cell non-Hodgkin's lymphomas treated with high-dose chemotherapy and peripheral blood progenitor cell autografting. Blood 1997, 89: 724-731 5. Gallamini A, Stelitano C, Calvi R, Bellei M, Mattei D, Vitolo U, Morabito F, Martelli M, Brusamolino E, Iannitto E, Zaja F, Cortelazzo S, Rigacci L, Devizzi L, Todeschini G, Santini G, Brugiatelli M, Federico M; Intergruppo Italiano Linfomi. Peripheral T-cell lymphoma unspecified (PTCL-U): a new prognostic model from a retrospective multicentric clinical study. Blood. 2004 Apr 1;103 (7): 2474-9 Conflict of Interest:None declared |
|||