- © 2008 by American Society of Clinical Oncology
Phase II Study of Rituximab Plus Three Cycles of CHOP and Involved-Field Radiotherapy for Patients With Limited-Stage Aggressive B-Cell Lymphoma: Southwest Oncology Group Study 0014
- Daniel O. Persky,
- Joseph M. Unger,
- Catherine M. Spier,
- Baldassarre Stea,
- Michael LeBlanc,
- Matthew J. McCarty,
- Lisa M. Rimsza,
- Richard I. Fisher and
- Thomas P. Miller
- From the Arizona Cancer Center, University of Arizona, Tucson, AZ; Southwest Oncology Group Statistical Center, Seattle, WA; Thompson Cancer Survival Center, Knoxville, TN; and the James P. Wilmot Cancer Center, University of Rochester Medical Center, Rochester, NY
- Corresponding author: Southwest Oncology Group (SWOG-S0014) Operations Office, 14980 Omicron Drive, San Antonio, TX 78245-3217; e-mail: pubs{at}swog.org
Abstract
Purpose To evaluate the effect of rituximab in limited-stage diffuse large B-cell lymphoma (DLBCL), we conducted a multicenter phase II trial combining rituximab with three cycles of CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone; R-CHOP) followed by involved-field radiation therapy (IFRT).
Patients and Methods Southwest Oncology Group (SWOG) study S0014 enrolled patients with newly diagnosed, aggressive, CD20-expressing non-Hodgkin's lymphoma (NHL). Patients had limited-stage disease and at least one adverse risk factor as defined by the stage-modified International Prognostic Index (nonbulky stage II disease, age > 60 years, WHO performance status of 2, or elevated serum lactate dehydrogenase). Four doses of rituximab were infused on days −7, 1, 22, and 43, and CHOP was administered on days 3, 24, and 45, followed 3 weeks later by 40 to 46 Gy of IFRT.
Results Sixty patients with aggressive NHL were eligible. With the median follow-up of 5.3 years, treatment resulted in a progression-free survival (PFS) of 93% at 2 years and 88% at 4 years. Overall survival (OS) was 95% at 2 years and 92% at 4 years. These results were compared with those from a historic group of patients treated without rituximab on S8736, demonstrating PFS of 78% and OS of 88% at 4 years.
Conclusion In limited-stage DLBCL, the addition of rituximab to three cycles of CHOP plus IFRT met prespecified study criteria of efficacy, with 2-year PFS of at least 84%, meriting further investigation. There is a pattern of continuing relapse with modest survival gains. We hypothesize that such a pattern may be the result of biologic differences between limited- and advanced-stage lymphoma.
INTRODUCTION
Successful outcome for patients with limited stage aggressive histologies of non-Hodgkin's lymphoma (NHL) requires initial treatment with doxorubicin-containing chemotherapy.1 For the last two decades, the most commonly used strategy has been three cycles of CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) chemotherapy followed by involved-field radiotherapy [CHOP(3) plus IFRT] resulting in 5-year overall survival (OS) estimates that vary from 68% to 97%.2-4 However, the proportion of patients surviving is markedly affected by patient selection. Commonly used clinical variables such as stage, age, serum lactate dehydrogenase (LDH), and performance status are strong prognostic predictors of outcome.
The International Prognostic Index (IPI) has gained wide acceptance for use in advanced disease.5 However, the IPI has limited utility in patients with limited stages of NHL because two of the five defined risk factors are specific for advanced stages of NHL (stages I/II v III/IV, and extranodal sites of disease of < 2 or v 2 or more). Consequently, we have developed a stage-modified version of the IPI for use in patients with limited-stage disease. This schema successfully separates outcome among patients with limited-stage disease. Using stage (I v nonbulky II), age (< 60 v ≥ 60), serum LDH (normal v elevated) and performance status (ambulatory v less than fully ambulatory), 5-year OS varies from 95% to 42%.4 Furthermore, the stage-modified IPI is reproducible.3
The current phase II therapeutic trial used the stage-modified IPI to select patients. Limited-disease patients with no adverse risk factors (stage I, younger age, normal serum LDH, and good performance status) were excluded from the trial because their overall survival is 95% to 97% at 5 years.2,3 Patients with bulky stage II disease were excluded because their course was similar to patients with advanced disease. Therefore, inclusion required stage I disease with an adverse risk factor or nonbulky stage II disease. Historically, OS for these patients was 71% at 5 years, and the risk of progression or death justified inclusion in a therapeutic phase II trial.
The trial was designed to test the effect of adding rituximab to standard treatment in the phase II setting. The design was based on historical data for patients treated with CHOP(3) plus IFRT.2 Rituximab was chosen on the basis of proven efficacy in advanced disease.6-9 Rituximab was added to CHOP (R-CHOP) in a manner adapted from the initial study of rituximab and CHOP by Czuczman et al.10
PATIENTS AND METHODS
Patient Selection
Southwest Oncology Group (SWOG) study 0014 was a phase II open-label trial for patients with aggressive, CD20-expressing NHL, including DLBCL, mantle-cell lymphoma, Burkitt's or Burkitt-like lymphoma, and what was formerly known as B-cell phenotype of anaplastic large-cell lymphoma. The study had to be approved by a local institutional review board and be in accord with state and federal guidelines. The study required central hematopathology review of specimens to confirm the diagnoses. Fine-needle aspirations were not accepted. The patients had to have stage I, IE, or nonbulky II or IIE disease by Ann Arbor classification. Bulky disease was defined as any mass exceeding 10 cm in maximal diameter, or a mediastinal mass with a maximal diameter exceeding one third of maximal chest diameter. Patients also had to have at least one adverse risk factor as defined by the stage-modified IPI (nonbulky stage II disease, age > 60 years, WHO performance status of 2, or elevated serum LDH).2
All patients had WHO performance status of 0 to 2, no prior lymphoma therapy, no prior malignancy (except for adequately treated basal cell or squamous cell skin carcinoma, in situ cervical carcinoma, or other cancer from which the patient was disease-free for at least 5 years), and no hiv infection. Pretreatment evaluation included a physical examination with performance status evaluation; computed tomography (CT) scan of chest, abdomen, and pelvis; bone marrow biopsy; and serum LDH. Patients were determined to be medically fit to receive CHOP chemotherapy. Radiation oncology consultation at a SWOG-approved RT facility occurred within 21 days before registration, confirming that treatment per protocol was medically and technically feasible.
Treatment
Rituximab was infused at 375 mg/m2 on days −7, 1, 22, and 43, and CHOP was administered at standard doses (cyclophosphamide 750 mg/m2, doxorubicin 50 mg/m2, vincristine 1.4 mg/m2 (capped at 2 mg) and prednisone 100 mg for 5 days on days 3, 24, and 45. Involved-field radiation therapy (IFRT) began 3 weeks after the last cycle of R-CHOP (day 66) as previously described.2 Briefly, IFRT consisted of 40 to 46 Gy for patients achieving a complete response (CR), with a small boost volume to a maximum of 50 to 55 Gy for patients failing to achieve a CR, in daily fractions of 1.8 to 2 Gy (5 days a week), in predetermined standard ports involving only the lymph node region(s) or organ affected by overt disease and calculated at midplane of the target volumes or 3-cm depth for supraclavicular regions. Use of growth factors was at the discretion of the treating physician.
Statistical Analysis
The primary objective of the study was to estimate 2-year progression-free survival (PFS). We assumed a 2-year PFS of at least 75% on the basis of prior results without rituximab, and a 2-year PFS of at least 84% in this trial would, therefore, merit further investigation in the phase III setting. Sixty patients were sufficient to estimate the 2-year PFS to within 11% and were sufficient to determine the probability of any toxicity to within 13%. Any adverse event occurring with at least 5% probability was likely to be seen at least once (95% probability).
The analysis was performed with all eligible patients. Estimates of PFS and OS were performed using the Kaplan-Meier method.11 PFS was measured from the time of registration until disease progression or relapse, or death resulting from any cause. OS was measured from the time of registration until death resulting from any cause. The adjustment for prognostic factors (ie, IPI) for PFS and OS was performed using Cox regression analysis.12 Toxic effects and therapy responses were coded according to the National Cancer Institute Common Toxicity Criteria (CTC), version 2.
To better assess the benefit of adding rituximab to CHOP(3) plus IFRT, we identified all patients (n = 68) from SWOG study 8736 treated with CHOP(3) plus IFRT who met all eligibility criteria for the current study. Specifically, the SWOG 8736 patients must have had one or more adverse risk factors and must have had histology and immunophenotype demonstrating diffuse large, Burkitt-like, or high-grade B-cell lymphoma.
RESULTS
Patient Characteristics
Seventy-one patients were registered on the study between April 1, 2000, and January 15, 2002. Of these, 11 patients were ineligible: six patients had incorrect stage of disease, four patients had follicular NHL according to central pathology review (specimens with any trace of follicular structure were excluded from the study), and one patient had insufficient prestudy information. All 60 eligible patients have finished treatment, with 55 of the 60 eligible patients (92%) completing all therapy as planned. Treatment was not completed in five patients because of adverse events (n = 3), therapy refusal (n = 1), and early death unlikely related to treatment (n = 1). The clinical characteristics of 60 eligible patients are summarized in Table 1. The risk factors were predominantly older age (73%) and stage II disease (43%).
Sixty Eligible Patients Receiving R-CHOP(3) Plus IFRT (from S0014) Compared With 68 Eligible Patients Receiving CHOP(3) plus IFRT (from S8736), Matched for Limited Disease, Aggressive B-Cell Histologies, and Presence of at Least One Stage-Modified IPI Risk Factor
Outcomes
There were no major chemotherapy violations, defined as more than 10% deviation from expected dose rate. One patient with bilateral neck lymphadenopathy did not have radiation therapy administered to the side where lymph nodes were resected and was considered a major violation. Fifty-eight patients (97%) received three cycles of R-CHOP as planned and 57 patients (95%) received radiation therapy. Fifty-one of 54 patients (95%) on whom dose data are available completed radiation therapy at planned dosage (40 to 55 Gy). Most of the patients (91%) received doses in the range of 40 to 46 Gy. The median dose delivered was 41.4 Gy. Radiation therapy was initiated at a median of 24 days after administration of intravenous CHOP chemotherapy. All but two patients began radiation therapy before day 35.
Median follow-up among patients still alive is 5.3 years (range, 4.1 to 6.3 years). Fourteen of the 60 eligible patients have relapsed or died. Eight of these 14 patients had documented lymphoma relapse; five of these eight patients have died. One relapse occurred in the radiation field, five occurred outside the radiation field, and the location of the other two relapses could not be confirmed. The 2-year PFS estimate is 93% (95% CI, 87% to 100%) and the 4-year PFS estimate is 88% (95% CI, 80% to 96%; Fig 1A). Eleven of the 60 eligible patients have died. The 2-year OS estimate is 95% (95% CI, 89% to 100%) and the 4-year OS estimate is 92% (95% CI, 85% to 99%; Fig 1B). Of six patients who died without documented recurrence, one died as a result of treatment toxicity, one died as a result of hepatocellular carcinoma, one died as a result of interstitial pneumonitis, one died as a result of pulmonary embolism, one died as a result of acute myocardial infarction, and in one patient the cause of death was not known.
(A) Progression-free and (B) overall survival of 60 eligible patients enrolled in a Southwest Oncology Group (SWOG) trial of three cycles of R-CHOP followed by involved-field radiation therapy. R-CHOP, rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone.
Comparison With the Prior SWOG Trial
To assess the promise of adding rituximab to CHOP(3) plus IFRT (and therefore, whether the addition of rituximab would be worth testing in a comparative trial), we identified 68 patients from SWOG study 8736 treated with CHOP(3) plus IFRT using the same eligibility criteria as for S0014 (limited-stage disease with at least one stage-modified IPI adverse risk factor and aggressive diffuse B-cell histologies). Of those 68 patients, 60 had DLBCL, seven had Burkitt-like lymphoma, and one had high-grade B lineage lymphoma. The characteristics of patients on S8736 (historical control group) and S0014 are comparable (Table 1). The survival curves for S8736 were truncated to match the follow-up for S0014 (Figs 2 and 3). PFS at 2 years on S8736 (historical control group) was 85% (95% CI, 77% to 94%) compared with 93% on S0014, whereas PFS at 4 years was 78% (95% CI, 68% to 88%) as opposed to 88%, respectively. For OS, the 2-year estimate was 93% (95% CI, 86% to 99%) for S8736 (historical control group) versus 95% for S0014, and the 4-year estimate was 88% (95% CI, 81% to 96%) compared with 92%, respectively.
Progression-free survival of 60 eligible patients receiving R-CHOP(3) plus IFRT (from S0014) is compared with that of 68 eligible patients receiving CHOP(3) plus IFRT (from S8736), matched for limited disease, aggressive B-cell histologies, and presence of at least one stage-modified IPI risk factor. The curve for S8736 was truncated to match the follow-up for S0014. R-CHOP(3), three cycles of rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone; IFRT, involved-field radiation therapy.
Overall survival of 60 eligible patients receiving R-CHOP(3) plus IFRT (from S0014) is compared with that of 68 eligible patients receiving CHOP(3) plus IFRT (from S8736), matched for limited disease, aggressive B-cell histologies, and presence of at least one stage-modified International Prognostic Index (IPI) risk factor. The curve for S8736 was truncated to match the follow-up for S0014. R-CHOP(3), three cycles of rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone; IFRT, involved-field radiation therapy.
Survival according to study and risk group was similar between the two studies, with 2-year PFS in patients with a single stage-modified IPI risk factor of 90% in the historical control group S8736 versus 100% for S0014, and 2-year PFS in patients with more than 1 risk factor of 79% in S8736 versus 78% in S0014.
Toxicity
All 60 eligible patients have been evaluated for toxicity (Table 2). One patient (2%) died as a result of treatment-related anorexia. Most toxicities were hematologic, with 59 patients (98%) having some hematologic toxicity and 29 patients (48%) having grade 4 hematologic toxicity. The hematologic toxicity was primarily neutropenia (90% any grade, 20% grade 3, 45% grade 4). flu-like symptoms consisted mostly of fatigue/malaise/lethargy (46 patients [77%] any grade and none with grade 4); and these symptoms were likely related to infusion reactions. One patient (2%) developed grade 3 pharyngeal constriction with first dose of rituximab and had to discontinue the trial. One patient developed acute myelogenous leukemia 2.8 years after completion of all therapy, with t(8;21) and subsequently died of CNS lymphoma relapse; one other patient developed myelodysplastic syndrome 2.6 years after completion of therapy and remains alive as of last contact. In addition, seven patients have developed solid tumors, including breast (n = 3), hepatic (n = 1), rectal (n = 1), sarcoma (n = 1), and GIST (n = 1). These cancers developed outside the radiation fields. One patient subsequently developed follicular lymphoma.
Most Common Toxicities for 60 Eligible Patients on S0014 Summarized by Grade
DISCUSSION
Building on the results of trials incorporating rituximab with CHOP in advanced-stage DLBCL,6,8 S0014 incorporated four doses of rituximab into standard treatment with CHOP(3) plus IFRT. Patients with high-risk, limited-stage, aggressive lymphoma were selected for the trial. With the median follow-up of 5.3 years, we found that addition of rituximab resulted in PFS of 93% at 2 years and 88% at 4 years. Overall survival was 95% at 2 years and 92% at 4 years.
The toxicity profile was similar to that seen with CHOP(3) plus IFRT, with the notable exception of higher rates of neutropenia.2 Grade 4 neutropenia was present in 45% of the patients in the current trial, compared with 27% in S8736. The increased toxicity could be attributed to incorporation of rituximab, or to an older cohort of patients (median age, 69 years in the current trial v 59 years in S8736). One patient (2%) died as a result of treatment-related complications, in line with prior experience. Development of AML with t(8;21) 2.8 years after completion of therapy in another patient would be unusual with CHOP and radiotherapy. The proportion of patients developing solid tumors (12%) probably reflects an older population.
We compared the results of the current trial with all patients from S8736 who would have met eligibility criteria for S0014 and received protocol treatment, which did not include rituximab. Although this represents a nonrandomized comparison, the results of this study are nonetheless promising in the context of historical cohort. PFS was 88% and OS was 92% at 4 years in the current study, as opposed to PFS of 78% and OS of 88% in the historical control group. The results of this study satisfy our prespecified study criteria for efficacy in having PFS at 2 years of at least 84% (93% in the study).
The experience from British Columbia Cancer Agency (BCCA) using CHOP(3) plus IFRT closely mirrors S8736 patient population and outcomes, strengthening the possibility that adding rituximab improved survival in the current trial.3 Three prospective trials in limited-stage DLBCL (Eastern Cooperative Oncology Group [ECOG] 1484 and two Groupe d'Etude des Lymphomes de l'Adulte [GELA] trials, LNH 93-1 and LNH 93-4) have different patient populations and are more difficult to compare.
In GELA trial LNH 93-1 comparing CHOP(3) plus IFRT to aggressive dose-intense chemotherapy (doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone [ACVBP]) with sequential consolidation, median age was 47 years, 67% of patients had stage I disease, and 95% of patients had a 0 age-adjusted IPI (aaIPI) risk score, ensuring that the majority of patients had no risk factors using the stage-modified IPI. Nevertheless, in that study (LNH 93-1) with patients selected for lower risk, the results were comparable to our study, even for the arm that received ACVBP and consolidation (90% 5-year OS).13
In another GELA trial (LNH 93-4) patients older than 60 years (median, 68 years), were randomly assigned to four cycles of CHOP [CHOP(4)] or CHOP(4) plus IFRT. In that study, 95% had a 0 aaIPI risk score, and 65% had stage I disease. Using the stage-modified IPI, 65% had one risk factor (older age and stage I disease) and 35% had two risk factors (older age and stage II disease), comparable with the current study (S0014). The results were inferior to S0014, and slightly inferior to S8736, with 5-year OS of 72% in CHOP(4) arm and 68% in CHOP(4) plus IFRT arm.14 By comparison, 5-year OS for S8736 is 79%, and for the current study is 82%, though follow-up through 5 years is not yet complete. The analysis of our patients older than 60 years (44 patients, 73%) shows 4-year estimates of 86% for PFS and 89% for OS (data not shown). ECOG 1484 did not record LDH and thus cannot be compared.15 Mabthera International Trial (MInT)9 had 18% of patients with stage I and 55% with stage II disease but was not intended to study limited-stage disease but rather low-risk disease in young patients (and reported it as such), making extrapolation to early disease in older patients difficult. Furthermore, 49% of the patients on that trial received radiation therapy, based on varying definitions of bulky disease.
Intriguingly, the impact of rituximab in a limited-stage setting appears to be smaller than in advanced-disease setting. There is 4% absolute difference in 4-year OS as opposed to 10% to 15% absolute difference in 5-year OS in advanced disease, although the relative difference is 33%, more in line with reduction seen in advanced disease. There seems to be a continuing pattern of relapse without a plateau in either PFS or OS curve. One explanation for this may be that our phase II trial has smaller patient numbers and may not predict 4-year OS as accurately as larger studies. Further, our trial was not randomized and was also compared with a historic cohort. Another possibility, however, is that limited-stage DLBCL may be biologically different.
We have previously shown preliminary evidence that limited-stage disease has a unique molecular fingerprint relative to advanced disease.16 In that study of 35 stage I and 131 stage III/IV DLBCL patients, limited-stage disease was largely (77%) of germinal center origin (GCB), whereas in the advanced-stage disease only 49% presented with GCB subtype (P = .01), as defined by a previously published gene expression profile.17
Interestingly, two recent studies suggest that adding rituximab to CHOP preferentially benefits the non-GCB phenotype and thus eliminates the adverse impact associated with non-GCB subtypes.17 Analysis of an Intergroup trial by Winter et al demonstrated that bcl-6–negative but not bcl-6–positive cases showed a significant improvement with addition of rituximab to CHOP (bcl-6 is mostly, but not exclusively, associated with the GCB subtype).18 A recent study by Nyman et al demonstrated that outcomes of GCB and non-GCB phenotypes were similar after rituximab-containing therapy, but only GCB phenotype retained favorable prognosis after CHOP-like chemotherapy without rituximab.19 This is further supported by studies reported in abstract form from the British Columbia Cancer Agency (BCCA),20 Japan,21 and the National Institutes of Health.22 Thus, if limited-stage DLBCL is primarily of GCB origin, we hypothesize that treatment with rituximab might result in a smaller incremental benefit in limited-stage DLBCL compared with advanced-stage disease.
In summary, we found that the addition of rituximab to CHOP(3) plus IFRT in patients with high-risk, limited-stage DLBCL resulted in PFS of 93% at 2 years and 88% at 4 years, and OS of 95% at 2 years and 92% at 4 years. When the results of the current trial are compared with a historical cohort treated with CHOP(3) plus IFRT, the improvement is clinically significant but nonetheless shows a pattern of continuous relapse. We hypothesize that the modest improvement observed in this trial may be explained by a unique biology associated with limited-stage disease. Thus, the paradigm of looking to improvements in advanced disease to incorporate into the treatment of limited-stage disease may not be entirely appropriate.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: None Consultant or Advisory Role: Richard I. Fisher, Genentech (C), GlaxoSmithKline (C), Millenium (C); Thomas P. Miller, Genentech (C) Stock Ownership: None Honoraria: Richard I. Fisher, Genentech, GlaxoSmithKline, Millenium; Thomas P. Miller, Genentech, Biogen Idec Research Funding: Thomas P. Miller, Biogen Idec, Genentech Expert Testimony: None Other Remuneration: None
AUTHOR CONTRIBUTIONS
Conception and design: Joseph M. Unger, Baldassarre Stea, Michael LeBlanc, Richard I. Fisher, Thomas P. Miller
Administrative support: Richard I. Fisher, Thomas P. Miller
Provision of study materials or patients: Baldassarre Stea, Matthew J. McCarty, Lisa M. Rimsza, Thomas P. Miller
Collection and assembly of data: Daniel O. Persky, Joseph M. Unger, Catherine M. Spier, Michael LeBlanc, Thomas P. Miller
Data analysis and interpretation: Daniel O. Persky, Joseph M. Unger, Catherine M. Spier, Michael LeBlanc, Richard I. Fisher, Thomas P. Miller
Manuscript writing: Daniel O. Persky, Joseph M. Unger, Lisa M. Rimsza, Richard I. Fisher, Thomas P. Miller
Final approval of manuscript: Daniel O. Persky, Joseph M. Unger, Catherine M. Spier, Baldassarre Stea, Richard I. Fisher, Thomas P. Miller
Footnotes
-
published online ahead of print at www.jco.org on April 14, 2008
Supported in part by Public Health Service Cooperative Agreement Grants No. CA13612, CA46282, CA42777, CA45450, CA04919, CA35090, CA20319, CA45808, CA35431, CA45807, CA35176, CA45560, CA35192, CA45377, CA35261, CA35128, CA35119, CA46441, CA38926, and CA32102 awarded by the National Cancer Institute.
Presented in part at the 46th Annual Meeting of the American Society of Hematology and Exposition, December 4-7, 2004, San Diego, CA.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received August 1, 2007.
- Accepted January 25, 2008.
