- © 2005 by American Society of Clinical Oncology
Efficacy and Safety of Tositumomab and Iodine-131 Tositumomab (Bexxar) in B-Cell Lymphoma, Progressive After Rituximab
- Sandra J. Horning,
- Anas Younes,
- Vinay Jain,
- Stewart Kroll,
- Jennifer Lucas,
- Donald Podoloff and
- Michael Goris
- From Stanford University, Stanford, CA; The University of Texas M.D. Anderson Cancer Center, Houston, TX; US Oncology, Dallas, TX; Corixa Corp, South San Francisco, CA
- Address reprint requests to Sandra J. Horning, MD, Stanford University, 875 Blake Wilbur Dr, Suite CC-2338, Stanford, CA 94305-5821; e-mail: sandra.horning{at}stanford.edu
Abstract
Purpose To determine overall response (OR) and complete response (CR) rates, response duration, progression-free (PFS) and overall survival and safety with the tositumomab and iodine-131 tositumomab (131I tositumomab) therapeutic regimen in patients with indolent, follicular large-cell, or transformed B-cell lymphoma, progressive after rituximab.
Patients and Methods From July 1998 to November 1999, 40 patients (24 rituximab nonresponders: 11 with response < 6 months, and five with response ≥ 6 months) received a therapeutic dose (0.65 to 0.75 Gy per platelet count) of 131I tositumomab based on total-body dosimetry in this prospective phase II study. The median number of prior treatments was four; 59% of patients were chemotherapy-resistant.
Results Confirmed OR (65%) and CR (38%) rates were not significantly associated with prior rituximab response. With a median follow-up of 3.3 years, the median PFS was 10.4 months, 24.5 months for responders, and not reached for CR patients. Among follicular grade 1 or 2 patients with tumors ≤ 7 cm (n = 21), the OR and CR rates were 86% and 57%. Estimated 3-year PFS in this subgroup was 48%, compared with 11% for all others (P = .002). Transient grade 3 to 4 marrow toxicity was seen in 50% of patients. Two patients, one of whom received two subsequent chemotherapy regimens, developed secondary myelodysplasia.
Conclusion 131I tositumomab is effective in CD20-positive lymphoma progressive after rituximab, with a 65% OR rate and median PFS of 24.5 months for responders. Patients with follicular grade 1 or 2 histology and tumors ≤ 7 cm achieved very high OR and CR rates, with 48% PFS at 3 years.
INTRODUCTION
Indolent lymphomas are characterized by advanced disease, initial response to chemotherapy, and relatively long survival. However, a pattern of multiple relapses ensues in the majority of patients throughout their disease course, during which time these lymphomas become more resistant to treatment and prone to transformation to an aggressive subtype.1,2
Monoclonal antibodies reactive with lymphoid-associated antigens are attractive as less toxic but potentially effective agents. The CD20 antigen is an ideal target due to expression on more than 90% of B-cell non-Hodgkin's lymphomas3,4; lack of shedding into the circulation or internalization after antibody binding3,5; and lack of expression on early B-cell progenitors or plasma cells. Rituximab is a chimeric immunoglobulin (Ig) G1 kappa anti-CD20 antibody that mediates complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity in vitro,6 with some evidence of a direct antiproliferative effect and induction of apoptosis in CD20-positive cell lines.7 A fundamental role for complement was suggested in nonimmunodeficient mice,8 whereas Fc receptor–dependent mechanisms were found to contribute substantially to the action of rituximab in another murine system.9 In the pivotal trial in relapsed or refractory low-grade B-cell lymphoma, an overall response (OR) rate of 50% was observed.10 However, because half of the patients did not respond in the pivotal trial and the overall progression-free survival (PFS) was 9 months, additional treatments are required after rituximab.
Tositumomab is an IgG2a murine anti-CD20 monoclonal antibody.3 In vitro, the binding of tositumomab to CD20-positive cells invoked antibody-dependent cellular cytotoxicity,11 complement-dependent cytotoxicity,12 and apoptosis.13 Conjugation of tositumomab to a radionuclide such as iodine 131 (131I) affords irradiation of lymphoma cells, both by direct cellular binding and through a “cross-fire effect” to neighboring lymphoma cells within the path-length of the radionuclide. The dosing methodology and maximum tolerated total-body dose (TBD) of 0.75 Gy were established in a phase I/II study of tositumomab and 131I tositumomab (Bexxar; Corixa Corp, South San Francisco, CA, and GlaxoSmithKline, Philadelphia, PA).15 Despite extensive prior treatment, an OR rate of 83% was reported in this study population. Infusional side effects were minimal, and hematologic toxicity with nadirs at 6 to 10 weeks, followed by recovery, was observed.14-16 The safety and efficacy of 131I tositumomab, as well as the reproducibility of the dosimetry method for calculation of a patient-specific prescribed TBD of 0.75 Gy, were confirmed in a phase II multicenter study of 47 patients in which OR and complete response (CR) rates of 57% and 32% were reported.17 In the pivotal trial of 131I tositumomab, conducted in 60 low-grade and transformed lymphoma patients who had received two or more prior regimens and failed to respond to or progressed within 6 months of last chemotherapy, the OR was 65%—a significantly greater response rate and with longer duration than last chemotherapy.18 The most common adverse event was transient marrow suppression similar to that observed in the earlier trials. Based on these encouraging outcomes and the increased use of rituximab in indolent lymphoma, we sought to establish the safety and efficacy of 131I tositumomab in patients who progressed after rituximab.
PATIENTS AND METHODS
Study Design and Eligibility
This prospective, phase II study evaluated 131I tositumomab in patients with indolent, follicular large-cell, or transformed B-cell lymphoma who had progressive disease after rituximab. The objectives were to determine the OR and CR rate, duration of response, PFS, survival, and safety profile. Efficacy end points were assessed by investigators and were subsequently reviewed by the Masked Independent Randomized Radiology and Oncology Review (MIRROR) panel. The study population included patients 18 years and older with CD20-positive follicular small cleaved or mixed (indolent, grade 1 or 2), de novo follicular large-cell (grade 3), or transformed (from low to intermediate-/high-grade according to the International Working Formulation classification) lymphoma.19 Patients had progressed after at least one course of standard rituximab, defined as failure to achieve a partial response (PR) or CR or disease progression after response. Patients had bidimensionally measurable disease with at least one lesion measuring ≥ 2 × 2 cm. Bone marrow biopsy had to demonstrate 0% to 25% lymphoma infiltration. Other eligibility criteria included Karnofsky performance status ≥ 60%, anticipated survival ≥ 3 months, baseline absolute neutrophil count (ANC) ≥ 1,500 cells/mm3 and platelet count ≥ 100,000/mm3, and adequate renal and hepatic function. In addition, no evidence of circulating human antimurine antibody (HAMA) at baseline and within 48 hours before the therapeutic dose was allowed. Patients had to provide written informed consent according to each individual institutional review board.
Treatment and Monitoring
Tositumomab was produced by Lonza Biologics PLC (Slough, England)/CYTOGEN Corp (Princeton, NJ), and Boehringer Ingelheim Pharma KG (Biberach, Germany). Radiolabeled 131I tositumomab was manufactured in a central facility (MDS Nordion, Kanata, Canada) using the Iodogen method.20 A dosimetric dose on day 0 was administered to calculate the day-7 to -14 therapeutic TBD of 0.75 Gy or attenuated to 0.65 Gy if the platelet count was less than 150,000/mm3 as previously described.21-23 The dose was adjusted to 137% lean body weight for obese patients.18 To prevent thyroid uptake of 131I, oral iodine was started ≥ 24 hours before the dosimetric dose and continued for 14 days after the therapeutic dose (potassium iodide four drops orally three times a day, or potassium iodide tablets 130 mg one tablet daily).
Patients were examined at weeks 7, 13, and 25, with CBCs obtained weekly after week 3. After the therapeutic dose, computed tomography monitoring was done at weeks 7, 13, and 25 and, beginning at 1 year, every 6 months for up to 2 years, and then annually or until lymphoma progression or death. Thyroid function was assessed at weeks 25 and 36. A centralized assay for HAMA was performed at day 5 and at weeks 7, 13, and 25. After 1 year, chemistry and thyroid tests were checked every 6 months until disease progression, follow-up for 2 years, or death. After August 2001, HAMA evaluations were performed on all patients until 2 years, independent of disease status, and thyroid-stimulating hormone (TSH) was to be assessed until death. Patients with progressive disease or followed up for 2 years entered long-term follow-up, where additional lymphoma therapy, thyroid function, and secondary malignancies were monitored annually.
Assessment of Efficacy
CR was defined as disappearance of all disease-related radiographic abnormalities, signs, and symptoms. A complete clinical response was defined as a CR, but allowed a stable residual radiographic abnormality ≤ 2 cm in diameter or ≤ 1 cm by physical examination. PR was defined as a ≥ 50% decrease in the sum of the products of the longest perpendicular diameters (SPPD) of measurable lesions and no new lesions. PD was defined as more than 25% in the SPPD of measurable lesions or the presence of new lesions. All responses were confirmed ≥ 4 weeks after initial response. Herein we report investigator responses but also discuss concordance with an independent, blinded review panel composed of two radiologists and two oncologists. Investigator assessments were used to supplement the durations for patients in continued response at their last independent assessment.
Statistical Methods
This study was designed to enroll 40 assessable patients. Analyses were performed with a data cutoff of May 2003. Duration of response and PFS were measured according to Kaplan-Meier analysis. PFS was measured from the dosimetric dose to the time of progression. Two-sided 95% CIs were calculated for response rates. All subgroup comparisons had two-sided test statistics. P values and CIs are presented without adjustment for multiple comparisons, outcomes, or looks. Exploratory univariate analyses comparing the OR and CR rates used the χ2 test with Yate's correction for categories with two subgroups. The log-rank test and Cox proportional hazards model were used for PFS in exploratory univariate and multivariate analyses.
RESULTS
Patient Characteristics
Forty-three patients were enrolled, but three withdrew before receiving the dosimetric dose due to ineligibility (two patients) or decision to initiate other therapy (one patient). Characteristics of the 40 patients treated at the three institutions from July 1998 to November 1999 are detailed in Table 1. The median age was 57 years (range, 35 to 78 years). Histology determined at each institution was: follicular small cleaved (grade 1, n = 12); follicular mixed (grade 2, n = 14); and one each of small lymphocytic and marginal zone. Two patients had de novo follicular large-cell (grade 3) lymphoma. Transformations were reported to the following: small cleaved with a diffuse pattern (n = 2), mixed with a diffuse pattern (n = 2), follicular large cell (n = 1), diffuse large cell (n = 3), and other (n = 2). Bone marrow was involved in 30%. Adverse characteristics included elevated lactic dehydrogenase in 31% and high or high-intermediate International Prognostic Index score in 21%. Thirteen patients (32%) had tumor masses larger than 7 cm, with 50% > 5 cm. The median number of prior chemotherapy regimens was four (range, one to 11), and the investigators considered 58% refractory to last chemotherapy. Twenty-four patients (60%) had no response to prior rituximab, 11 had a response duration less than 6 months, and five had a response duration ≥ 6 months.
Selected Characteristics in 40 Patients
Treatment and Efficacy
The median therapeutic dose of 131I was 88.7 mCi (range, 52.6 to 168.6); 80% received 0.75 Gy, and 20% received 0.65 Gy TBD. With a median follow-up from the dosimetric dose of 39 months, the confirmed OR rate was 65% (panel assessed = 68%). A confirmed CR was documented in 38% (panel assessed = 32%). There was 98% concordance between the panel and the investigator assessments of confirmed OR, and 85% concordance for confirmed CR (34 of 40 patients). The latter discordance consisted of four investigator CRs interpreted as partial by the panel, and two investigator PRs interpreted as complete by the panel. At the first (7-week) evaluation, the unconfirmed response rate was 66.7%, and the unconfirmed CR rate was 16.7%. The median time to CR was 13 weeks (range, 7 to 57 weeks).
The median PFS was 10.4 months (95% CI, 5.7 to 18.6) for all patients and 24.5 months for confirmed responders (95% CI, 16.8 to not reached [NR]; Fig 1). As illustrated in Figure 2, the PFS for 15 confirmed CR patients was NR with an estimated 3-year PFS of 73%. To date, 12 patients continue in response (11 CR, one PR) from 3.0 to 4.6 years after therapy. Figure 2 also illustrates the PFS for 13 confirmed CR patients according to panel assessment. The investigator-assessed group included four patients scored as PR by the panel; three of these patients continue in remission with no further therapeutic intervention at 3.4+, 3.7+, and 4.0+ years; and one progressed at 2.1 years.
Progression-free survival for all patients (N = 40) and confirmed responders (n = 26).
Progression-free survival for investigator-assessed complete responders (n = 15) and panel-assessed complete responders (n = 13). CRs, complete responses; MIRROR, Masked Independent Randomized Radiology and Oncology Review.
Table 2 indicates that prior response to rituximab did not significantly affect the confirmed OR rate, duration of response, or median PFS. No significant differences in confirmed CR rate or duration were observed. As demonstrated in Figure 2, patients who achieved CR after 131I tositumomab enjoyed prolonged PFS despite multiple prior therapies. In univariate analyses, age ≤ 60 years, tumor bulk ≤ 7 cm, and follicular grade 1 or 2 histology were associated with significantly longer PFS. In multivariate analysis, only follicular grade 1 or 2 histology was associated with significantly longer PFS.
Efficacy of 131I Tositumomab According to Prior Rituximab Response
Most CR patients (80%) had both follicular grade 1 or 2 histology and tumor bulk ≤ 7 cm. As indicated in Table 3, 18 (86%) of 21 patients with these characteristics responded to 131I tositumomab, and 12 (57%) of 21 had CRs. In contrast, patients with greater tumor bulk or other histologies (n = 19) had a 42% OR rate (P = .004) and a 16% CR rate (P = .007). The median durations of response and CR were not reached in patients with both follicular grade 1 or 2 histology and tumor bulk ≤ 7 cm. PFS at 3 years was 48% in this group compared with 11% for all other patients (P = .002; Fig 3).
Progression-free survival for patients with follicular grade 1/2 non-Hodgkin's lymphoma and tumor size ≤ 7 cm (n = 21) and patients with other histologies, or tumor size greater than 7 cm (n = 19).
Efficacy of 131I Tositumomab According to Histology and Tumor Size
Adverse Events
Two patients required adjustments in tositumomab infusions. A minority (16%) of nonhematologic adverse events was grade 3 or 4; these included pneumonia (n = 4), pain (n = 2), and nausea (n = 2). Twenty patients (50%) experienced grade 3 or 4 hematologic toxicity, including ANC less than 1,000/mm3 (n = 17), platelets less than 50,000/mm3 (n = 10), and hemoglobin less than 8.0 g/dL (n = 4). Seven patients developed grade 4 neutropenia, and four had grade 4 thrombocytopenia (including one patient with platelets < 10,000/mm3). Median nadirs were ANC 1,200/mm3, platelets 85,000/mm3, and hemoglobin 11.0 g/dL, with respective median times to nadir of 42, 34, and 42 days. Nine patients received hematologic supportive care, including erythropoietin (n = 6), red cell transfusions (n = 6), platelet transfusions (n = 4), and granulocyte colony-stimulating factor (n = 3). Twenty-two patients (55%) developed infections within 12 weeks of the therapeutic dose, of which 15 (68%) were grade 1 or grade 2 upper respiratory or viral infections. Serious infections were recorded in two patients with pneumonia.
Among 31 assessable patients, three developed an elevation of TSH including one patient who initiated medication following 131I tositumomab therapy, translating to a 3-year cumulative risk of 9.7%. Of the nine nonassessable patients, seven had abnormal thyroid function documented before treatment, and two were missing baseline TSH data but continued to have normal TSH after 4 years. Two patients (5%) developed HAMA 5 and 12 days after the dosimetric dose, and before the therapeutic dose. At week 7, both patients tested HAMA-negative, and no further time points were assessed.
Two patients were diagnosed with myelodysplasia (MDS) with evolution to acute myelogenous leukemia 2.1 and 3.3 years after receiving 131I tositumomab. The first patient was previously treated with fludarabine and cyclophosphamide and para-aortic and hemipelvic radiation therapy (RT). The second patient had received three chemotherapy regimens and RT before 131I tositumomab therapy and two additional courses of chemotherapy after radioimmunotherapy. Two patients developed solid tumors after treatment: prostate cancer at 3.9 years and squamous cell skin cancer at 1.5 years.
Overall Survival
The estimated overall survival was 60% at 3+ years (Fig 4). Sixteen (40%) of 40 patients were dead at the data cutoff time. Ten patients died from progressive disease, two from complications of subsequent transplantation, two from infectious complications of further lymphoma therapy, and one each from complications after gall bladder surgery and secondary leukemia. The median duration of survival from the dosimetric dose was not reached (95% CI, 24.8 months to NR).
Overall survival for all patients (N = 40). Dotted lines indicate 95% CI.
DISCUSSION
These results demonstrate the efficacy and safety of 131I tositumomab therapy in patients with indolent and transformed B-cell lymphoma who developed progressive disease after rituximab. The majority (88%) would be considered rituximab-refractory based on failure to respond or sustain a response for ≥ 6 months. The efficacy of 131I tositumomab might be attributable to the cytotoxic effects of tositumomab, irradiation, or both. The mechanisms by which lymphomas acquire or harbor endogenous resistance to single-agent rituximab are incompletely understood. Clinical samples from patients treated with rituximab have not confirmed preclinical data implicating inhibitors of complement as a resistance mechanism.24,25 In two separate reports, polymorphisms in the Fcγ receptor associated with IgG1 antibodies correlated with the rate and duration of response to rituximab in follicular lymphoma.26,27 Such polymorphisms are not clearly relevant for the murine antibody tositumomab. Although both antibodies target the CD20 epitope, distinct mechanisms of action were proposed based on observations in B-cell lines where F(ab′)2 fragments of tositumomab but not rituximab resulted in apoptosis in the absence of crosslinking.28 As a murine unconjugated antibody, tositumomab is less well equipped than rituximab to mediate cytotoxicity through host effector mechanisms, and this may explain the excellent tolerance for both dosimetric and therapeutic doses. In lymphoma xenograft models, complement depletion, elimination of NK cells, and binding to CD20 by tositumomab or F(ab′)2 resulted in cytotoxicity, implying that non–Fc dependent mechanisms were operative with tositumomab.29 In these studies, much higher levels of apoptosis were seen with tositumomab compared with rituximab. The independent efficacy of unconjugated tositumomab in recurrent or refractory B-cell NHL was demonstrated to be 19% in a randomized trial in which two patients had CRs of long duration.30
Although unconjugated tositumomab can contribute to tumor responses, results of a randomized trial indicate that the major efficacy is the consequence of low-dose-rate irradiation from 131I tositumomab.30 Similarly, in a randomized trial, significantly higher response rates were seen with the murine anti-CD20 parent antibody of rituximab (ibritumomab) conjugated with Yttrium-90 compared with rituximab in recurrent indolent and transformed lymphoma.31 Indolent lymphomas are exquisitely sensitive to radiation as evidenced by prolonged remissions after external beam radiation for limited-stage disease, and high OR and CR rates with low-dose radiation in patients with recurrent indolent lymphoma.32-35 In a preclinical model, the administration of radiation at a low dose rate (LDR) throughout a long period, simulating radioimmunotherapy, was associated with greater antitumor efficacy.36 The radiobiology of LDR RT is incompletely understood, but lymphoma cells may be more susceptible to LDR-induced apoptosis.37 The efficiency of DNA repair systems and the existence of soluble mediators may also be operative.
We observed an OR rate of 65%, with a median duration of response of 24.5 months in patients with a median of four prior chemotherapy courses (more than half were considered chemotherapy refractory); such patients have few therapeutic options. The clinical benefit from 131I tositumomab was remarkable for patients achieving CR in our study: 11 of 15 continue progression free between 3.0 and 4.6 years. The 73% PFS (95% CI, 54% to 100%) at 3 years in CR patients defines a highly durable and clinically important therapeutic result. The lack of ongoing progression or requirement for additional treatment for 3+ years in three of four patients assessed by investigators as continuing in CR, in contrast to the independent panel assessment of progressive disease, likely reflects the true absence of active lymphoma in these patients and the perils of measuring very small changes in lymph nodes.
Witzig et al reported the effects of Yttrium-90 ibritumomab tiuxetan radioimmunotherapy in a group of primarily follicular grade 1 or 2 lymphoma patients with progressive disease after rituximab.38 In this study, in which all patients had failed to respond to rituximab or relapsed within 6 months, the OR rate was 74%, and the CR rate was 15%. With a maximum follow-up of approximately 18 months, the median time to progression was 8.7 months for responders. In the current 131I tositumomab study, all patients have been followed for 36 to 55 months, and the median PFS was 25.4 months. Although patients in both studies had received a median of four prior treatments, dissimilarities in other patient characteristics may have influenced response to the different radioimmunotherapy treatments.
While it is expected that patients with nontransformed histology and tumors less than 7 cm in the current study had better outcomes, the high rate of CR (57%) and the duration of response (48% PFS estimated at 3 years) is remarkable in this multiply treated subset. The excellent long-term results from a pilot study in which 131I tositumomab was administered as primary therapy for follicular lymphoma and the absence of late toxicity in that patient population are notable.39,40 Together, these data suggest that 131I tositumomab should not be reserved until late in a follicular lymphoma patient's disease course when histologic transformation, bulky disease and possible resistance to LDR-induced apoptosis are more likely. Rather, our results provide rationale for use of 131I tositumomab earlier in the course of disease, and potentially following other treatment that lessens the tumor burden.
The development of MDS/leukemia in two patients may be an unfortunate consequence of the cumulative effects of cytotoxic therapy for lymphoma, recognized complications of alkylating agents and total-body irradiation.41 In an analysis of 995 patients treated with 131I tositumomab, Bennett et al estimated the annualized incidence to be 1.1% per year.42 To date, no MDS or leukemia has been seen after primary therapy with 131I tositumomab alone or after CHOP chemotherapy (cyclophosphamide, doxorubicin, vincristine, prednisolone).40,43 Although a relatively high rate of infection was reported in our study population, the majority were grade 1 or 2, of presumed viral etiology, and the incidence of serious infections (n = 2; 5%) was consistent with the overall incidence of serious infection, 4.3%, in the 131I tositumomab safety database.
In conclusion, 131I tositumomab is effective and well-tolerated therapy for patients with indolent and transformed B-cell lymphoma progressive after rituximab therapy. Patients with follicular grade 1 or 2 histology and tumors ≤ 7 cm had very high OR and CR rates, with a PFS of 48% at 3 years. Secondary MDS/leukemia resulting from the cumulative effects of all cytotoxic therapy was observed in two patients and remains a potential concern. Taken together, these data argue that 131I tositumomab therapy should be considered early in the disease course of follicular lymphoma. Future efforts directed toward understanding the in vivo biology of LDR radiation and continued support for clinical trials that seek to determine the optimal timing and combination of therapeutics for B-cell lymphoma are indicated.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Employment: Stewart Kroll, Corixa Corp. Consultant/Advisory Role: Sandra J. Horning, GlaxoSmithKline; Donald Podoloff, Corixa Corp; Michael Goris, Corixa Corp. Stock Ownership: Stewart Kroll, Corixa Corp. Honoraria: Donald Podoloff, Corixa Corp. Research Funding: Sandra J. Horning, Corixa Corp; Anas Younes, Corixa Corp; Vinay Jain, Corixa Corp; Donald Podoloff, Corixa Corp. Expert Testimony: Sandra J. Horning, Corixa Corp. Other Remuneration: Vinay Jain, Corixa Corp, GlaxoSmithKline. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration form and the Disclosures of Potential Conflicts of Interest section of Information for Contributors found in the front of every issue.
Footnotes
-
Supported by a grant from Corixa Corporation, South San Francisco, CA.
Presented in part at the Annual Meeting of the American Society of Hematology, San Francisco, CA, December 5-8, 2000.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
- Received July 9, 2004.
- Accepted October 20, 2004.
