Improved Outcome for Aggressive B-Cell Lymphomas with Early High-Dose Chemotherapy


Synopsis: In a randomized study, freedom from progression and event-free survival were significantly higher with high-dose sequential therapy and autologous stem cell or bone marrow rescue, as compared with MACOP-B chemotherapy.

Source: AM Gianni, et al. N Engl J Med 1997; 336:1290-1297.

Almost two decades ago, goldie and coldman theorized that resistance to chemotherapy would be minimized if tumor cells were exposed to multiple active drugs early in the treatment course (generally translated to practice, albeit imperfectly, as an alternating schedule of non-cross resistant regimens). This principle, taken to the maximum, led to the development of combination chemotherapy regimens such as MACOP-B, in which a dose-intense alternating schedule of drugs would be administered over 12 weeks. Initial single-arm studies suggested that this approach (using a variety of similarly constructed regimens) was superior to older therapies (principally CHOP) in aggressive histology lymphoma. However, a randomized comparison conducted by the National Cancer Institute-supported Cooperative Groups failed to demonstrate an improvement in response rate, overall- or disease-free survival for the more dose-intense regimens compared to the older CHOP regimen.1 Some of us did not accept the results of this study as a trump that invalidated the large body of data (including other randomized trials) contradicting its conclusions. But most embraced the notion that all of the regimens cured only one-third of the patients and that CHOP was as good as any of them.

More recently, another cytokinetic model has gained ascendancy: the Norton-Simon model, of which the Bonadonna sequential doxorubicin followed by cyclophosphamide, methotrexate, and 5-fluorouracil (A®CMF) and the Memorial Sloan-Kettering Cancer Center sequential high-dose doxorubicin plus cyclophosphamide (AC) followed by paclitaxel (T) regimens are two examples.2,3 The Norton-Simon hypothesis actually preceded the Goldie-Coldman hypothesis, but the thorough testing of its predictions is more recent. In this model, the strategy is to eliminate rapidly growing cells quickly by using a series of dose-intense regimens (or single agents) as the first therapy given at maximal tolerated dose. In this scheme, an intensive treatment phase is followed by another, given sequentially rather than in an alternating fashion.

Recently, Gianni et al from Milan published the results of a dose-intense, sequential regimen in the treatment of high-grade B-cell lymphomas. Ninety-eight patients (aged 16-70 years) were randomized to either MACOP-B or a sequential high-dose regimen with autologous bone marrow or stem cell support. Patients with bulky disease were eligible, but they all had to be candidates for myeloablative therapy. Exclusion criteria included morphologic bone marrow involvement, T-cell phenotype, or any follicular component. Approximately 70% of patients in both treatment groups had stage III or IV disease. About 90% had Working Formulation group G histology, and 10% had group H histology.

Patients randomized to MACOP-B were reevaluated upon completion of the 12-week regimen. If there was a good partial remission (> 80% tumor reduction), patients received consolidative radiotherapy (3060-3420 cGy) to bulky regions. If there was less successful response, patients were crossed over to high-dose sequential therapy. Similarly, those who were not successfully debulked with sequential therapy were crossed over to MACOP-B. Those who responded with greater than 80% tumor reduction received consolidative high-dose chemotherapy with radiotherapy to residual or bulky lesions.

The high-dose sequential therapy arm consisted of five phases. The first phase was conventional-dose doxorubicin, prednisone, and vincristine, designed to debulk tumor and increase performance status. The second phase consisted of high-dose cyclophosphamide with G-CSF or GM-CSF support. At the end of this phase, peripheral blood stem cell leukapheresis and/or bone marrow harvesting was performed. The third phase consisted of vincristine and high-dose methotrexate with leucovorin rescue. The fourth phase consisted of high-dose etoposide with G-CSF or GM-CSF support. Those who achieved greater than 80% reduction in tumor volume went on to phase five, consisting of myeloablative conditioning (with high-dose melphalan and either total body irradiation or high-dose mitoxantrone) and reinfusion of autologous stem cells and/or bone marrow. Myeloid growth factors were again used to support recovery. The last 13 patients received peripheral blood stem cells alone, whereas the others were rescued with stem cells from both bone marrow and peripheral blood. The median duration of MACOP-B therapy was just one week longer than the prescribed 12 weeks. Excluding the Phase 1 of sequential therapy (which lasted 21 days), the median time from Phase 2 to final hospital discharge was also 12 weeks.

The response rate was substantially higher with high-dose therapy, with 96% complete and 4% partial remissions, as opposed to 70% complete and 13% partial remissions in the MACOP-B arm. The seven-year freedom-from-progression rate was 84% following high-dose therapy compared to only 49% in the MACOP-B arm. The seven-year event-free survival was 76% for high-dose sequential therapy vs. only 49% for MACOP-B. Each of these differences was statistically significant. Overall survival was 81% for high-dose therapy and 55% for MACOP-B—a difference that did not quite achieve statistical significance (P = 0.09). Relapse-free survival was 88% for high-dose therapy and 70% for MACOP-B (P = 0.055).

Not surprising, toxicity was greater in the sequential high-dose therapy arm. However, toxic death rates were similar on both arms: high-dose therapy (8%) compared with MACOP-B (6%). With a median follow-up of 55 months, there were a total of nine deaths in the group randomized to high-dose sequential therapy and 18 deaths in patients randomized to receive MACOP-B.

Following crossover, all five patients who failed high-dose sequential therapy or whose disease relapsed, also failed MACOP-B and died of progressive disease. Of the 23 patients who did not respond or who relapsed after MACOP-B, only 14 received the full course of high-dose therapy. The others refused this treatment, were judged ineligible, or could not tolerate it. Of these 14, four remain in complete remission for 2.5 to six years.


High dose therapy with autologous bone marrow and/or stem cell transplantation is still a "work in progress," in that, for most diseases, it is still not known how and when high-dose chemotherapy is best implemented in relation to conventional-dose chemotherapy. Many consider high-dose therapy as a technology best reserved to treat refractory or relapsed disease. Indeed, this paradigm has been successful in the treatment of aggressive lymphomas where randomized trials have demonstrated a significant benefit for high-dose therapy compared to conventional salvage therapy for relapsed disease. In other malignancies, the efficacy of high-dose therapy in metastatic or refractory disease may not be significantly different from non-myeloablative therapies. However, in clinical trials in testicular carcinoma, metastatic breast carcinoma, and multiple myeloma, where there appears to be a benefit from high-dose therapy, the benefit appears greatest when high-dose therapy is applied earlier in the course of the disease.

The current study examines the use of high-dose sequential therapy with autologous bone marrow or peripheral blood stem cell support as initial therapy for aggressive B-cell lymphomas, and the study compares it to another aggressive, non-myeloablative, combination chemotherapy regimen which takes about the same length of time to complete. Note that the sequence of administration matters: Although this was a crossover study, MACOP-B was ineffective when given secondarily, whereas high-dose therapy was able to salvage a few MACOP-B-refractory cases.

The early use of high-dose therapy distinguishes this study from that of Verdonck et al, in which ABMT was reserved for those who exhibited a partial response to several cycles of conventional CHOP chemotherapy.4 In that study, Verdonck et al were unable to demonstrate a survival advantage with high-dose chemotherapy over the standard-dose regimen. However, the Verdonck et al study appeared to be designed with a bias against high-dose therapy: patients underwent autologous BMT only after having been defined as having disease that was relatively refractory to CHOP. Haioun et al compared intensive sequential chemotherapy (ifosfamide, etoposide, asparaginase and cytarabine) to high-dose therapy with autotransplantation to consolidate patients who had achieved complete remission after initial induction chemotherapy on LNH-84.5 The study design differs from the current study in that it compares two approaches to treating patients who have achieved a complete response to conventional dose therapy, a setting where the value of additional adjuvant therapy is unclear. The study failed to detect a difference between the two arms in terms of disease-free and overall survival. However, an almost-significant 20% increase in five-year survival rate was observed in the subgroup of patients with higher-risk features who underwent high-dose consolidation. The survival benefit of early high-dose chemotherapy for patients with poor-prognosis lymphoma in first remission is also suggested by two other non-randomized studies.6,7

It is important to notice a few details about the patient selection on this study. Patients with T-cell phenotype lymphomas, older patients, and those with intercurrent illnesses are excluded from the study. These criteria may lead to some bias in favor of high-dose therapy. However, there are many more patients with poor prognostic factors (International Index high-intermediate and high-risk) on this study than on the Intergroup Study. Furthermore, there was an imbalance in the distribution of prognostic features on this study: significantly more low-intermediate risk patients were randomized to the MACOP-B arm of this study (24% of those on MACOP-B vs 6% of those on high-dose therapy). This would lead to a bias against high-dose therapy.

Another interesting feature is how much better the results are with MACOP-B on this study than the Intergroup Study. Overall survival was 55% in this study, roughly twice that seen with MACOP-B in the Intergroup Study (as updated by Dr. Fisher in Lugano, June 1996). Gianni et al appear to have gotten results from their control arm that are significantly better than those obtained by the Intergroup investigators and much more similar to those originally reported by Connors and Klimo (Connors JM, Klimo P. Semin Hemotol 1988;25:41-46). Is this improvement entirely related to the use of radiation therapy to sites of bulk disease in patients achieving a complete response or could it be that more experience with MACOP-B regimen leads to better results?

Should the data from this randomized study compel us to abandon conventional chemotherapeutic approaches in favor of initial high-dose therapy? When we agree with the results of randomized studies, the usual response is that the study is definitive and should dictate practice. When we do not agree with the results or when implementing the results is difficult, the usual response is to wait for a confirmatory study before making any change. Certainly, there is increased cost and complexity associated with high-dose therapy and stem cell transplantation, and not all facilities and practitioners are equipped or trained to provide this therapy. However, in our opinion, especially for patients with high-intermediate and high-risk prognostic factors, high-dose therapy has produced a level of success that cannot be matched by CHOP. That level of success is not even achieved by a regimen that we believe is superior to CHOP (MACOP-B). Indeed, the results in these poor prognosis patients are comparable to CHOP plus radiation therapy in localized disease. Not every oncologist chooses to treat acute leukemia because the demands of the treatment intensity are too great. We may be on the verge of a similar situation for a subset of patients with aggressive lymphoma.


1. Fisher RI, et al. N Engl J Med 1993;328:1002-1006.

2. Bonnadonna G, et al. JAMA 1995;273:542-547.

3. Demetri GD, et al. Proc ASCO 1997;16:143a.

4. Verdonck LF, et al. N Engl J Med 1995;332: 1045-1051.

5. Haioun C, et al. J Clin Oncol 1997;15:1131-1137.

6 Santini G, et al. Proc ASCO 1993; 12:A1284.

7. Baro J, et al. Bone Marrow Transplant 1991; 8:283-289.