A phase 1 dose escalation study of bortezomib combined with rituximab, cyclophosphamide, doxorubicin, modified vincristine, and prednisone for untreated follicular lymphoma and other low-grade B-cell lymphomas†
Rajni Sinha MD, MRCP,
Department of Hematology and Medical Oncology, Winship Cancer Institute, Emory University-School of Medicine, Atlanta, Georgia
S. Lonial, R. Sinha, J. L. Kaufman, H. J. Khoury, Jr., M. J. Lechowicz, L. T. Heffner, M. Tighiouart, and C. R. Flowers conceived and designed the study; R. Sinha, P. J. Shenoy, C. Lewis, N. King, A. Hutchison-Rzepka, K. Bumpers, M. Tighiouart, and C. R. Flowers were responsible for data collection and assembly; R. Sinha, P. J. Shenoy, A. Hutchinson-Rzepka, K. Bumpers, J. L. Kaufman, H. J. Khoury, Jr., S. Lonial, M. J. Lechowicz, L. T. Heffner, M. Tighiouart, and C. R. Flowers were responsible for data analysis and interpretation; and R. Sinha, P. J. Shenoy, J. L. Kaufman, H. J. Khoury, Jr., M. J. Lechowicz, M. Tighiouart, and C. R. Flowers wrote the article.
Bortezomib has demonstrated efficacy in patients with relapsed B-cell non-Hodgkin lymphoma (NHL) both alone and in combination with other agents; however, limited data exist regarding its toxicity in combination with common frontline therapies for indolent NHL. A phase 1 study of bortezomib combined with rituximab, cyclophosphamide, doxorubicin, modified vincristine, and prednisone (R-CHOP) was conducted in patients with untreated follicular lymphoma (FL) and other indolent NHLs.
Nineteen patients, including 10 patients with FL, were enrolled. The median patient age was 59 years (range, 29-71 years). Seven patients had a FL International Prognostic Index score ≥3. R-CHOP with the vincristine dose capped at 1.5 mg was administered on a 21-day cycle for 6 to 8 cycles, and 1 of 3 dose levels of bortezomib (1.0 mg/m2 [n = 1], 1.3 mg/m2 [n = 6], or 1.6 mg/m2 [n = 12]) was administered on days 1 and 8 of each cycle using a Bayesian algorithm for dose escalation.
The maximum tolerated dose (MTD) of bortezomib with modified R-CHOP was reached at 1.6 mg/m2. Dose-limiting toxicity was observed in 5 patients (1 patient at a bortezomib dose of 1.0 mg/m2, 1 patient at a bortezomib dose of 1.3 mg/m2, and 3 patients at a bortezomib dose of 1.6 mg/m2). Neuropathy occurred in 16 patients (84%), including 2 patients (11%) who experienced grade 3 sensory neuropathy. Grade 4 hematologic toxicity occurred in 4 patients. Of 19 evaluable patients, 100% responded, and the complete response rate was 68%. At a median follow-up of 32 months, the 3-year progression-free survival rate was 89.5%.
Patients with indolent non-Hodgkin lymphomas (NHL) comprise approximately 25% of all new NHL cases, and follicular lymphoma (FL) represents 75% of indolent NHLs. The majority of patients with indolent NHL present with advanced stage disease and require treatment. Although current treatment strategies with chemoimmunotherapy are associated with higher response rates and longer remission durations,1-5 historically, the clinical course of these lymphomas with standard treatments is characterized by multiple relapses and a median survival of 6.2 years.6, 7 Common initial management options include radiation therapy, single agents, and combination chemoimmunotherapy, but no clear standard of care has been established for first-line therapy.8
Although the addition of rituximab (R) to chemotherapy improves survival for patients with B-cell lymphomas,2, 3, 6, 9, 10 patients still relapse after receiving R chemotherapy regimens. Some studies suggest that achieving completed remission (CR) and eradication of FL and possibly other forms of indolent NHL may be a key to the development of strategies aimed at cure.9, 11, 12 An initial step toward this end involves developing first-line treatment strategies with improved complete and molecular response rates. Our meta-analysis of frontline therapies for FL demonstrated that R plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) has a very high CR rate,5 and other data have indicated that it is the most commonly used first-line regimen for FL in the United States,8 making R-CHOP suitable as a backbone for a novel treatment approach.
One mechanism by which lymphoma cells become resistant to R and chemotherapy is through hyperactivation of the nuclear factor-κB (NF-κB) pathway, leading to overexpression of the long, alternatively spliced variant of bcl-x, B-cell lymphoma-extra large (bcl-XL).13, 14 Bortezomib is a proteasome inhibitor with antitumor activity in B-cell malignancies that produces responses in patients with relapsed FL either alone15 or combined with R16 and can overcome Bcl-XL-mediated R and chemotherapy resistance in vitro.13 These data suggest that adding bortezomib to R-CHOP could produce a therapy that may maximize response rates in the frontline setting and provide a platform for establishing long-term disease-free survival with maintenance therapy.17-19
Bortezomib has been combined safely with R in the treatment of relapsed FL, demonstrating equal efficacy and less toxicity when given on a weekly basis compared with a twice-weekly schedule.16 Neurotoxicity has been of particular concern when combining bortezomib with vincristine. Peripheral neuropathy rates have been as high as 65% in some reported studies when combining bortezomib with R-CHOP for aggressive lymphomas.17, 19 Furman and colleagues reported a 65% rate of peripheral neuropathy in a phase 1 trial using full-dose R-CHOP with 1 of 3 dose levels of bortezomib (0.7 mg/m2, 1.0 mg/m2, and 1.3 mg/m2) administered on days 1 and 4 of each cycle.17 Grade 3 neuropathy was observed at the 1.0 mg/m2 dose level. Ruan and colleagues confirmed those findings, describing neuropathy in 64% of patients with diffuse large B-cell lymphoma (DLBCL) and mantle cell lymphoma with additional follow-up from an extension of this trial.19 Given the concerns for higher rates of neuropathy when combining these 2 drugs with overlapping toxicities, additional monitoring of neurologic events is needed for such trials. Neurotoxicity-directed questionnaires can be useful tools for determining the presence and intensity of neuropathic pain and/or peripheral neuropathy from the patient's perspective. For this reason, we incorporated a neurotoxicity questionnaire into our study to capture signs and symptoms of neuropathy and performed electrophysiologic testing to help identify markers of chemotherapy-induced nerve damage. To our knowledge, this is the first study to use an escalating schema for both bortezomib and vincristine to treat patients with indolent NHL and to evaluate the incidence and severity of neuropathy in this manner.
MATERIALS AND METHODS
Patients who satisfied all the following criteria were enrolled into the study: 1) histologically confirmed diagnosis of a previously untreated FL (grade 1, 2, 3a, or 3b), small lymphocytic lymphoma (SLL), marginal zone lymphoma (MZL), or lymphoplasmacytic lymphoma; 2) an indication for treatment based on modified French Follicular Lymphoma Study Group criteria20 or a Follicular Lymphoma International Prognostic Index score ≥321; 3) an Eastern Cooperative Oncology Group (ECOG) performance status ≤2; 4) aged ≥18 years at the time of registration; 5) at least 1 measurable tumor mass (>1.5 cm in greatest dimension and >1.0 cm in the short axis) on a computed tomography (CT) scan or physical examination; 6) adequate renal function demonstrated by a serum creatinine level <1.5 mg/dL; 7) alanine and aspartate aminotransferase levels ≤3.5 the upper limit of normal and a total bilirubin level ≤2.0 mg/dL unless secondary to lymphoma; 8) cardiac left ventricular ejection fraction ≥50%; and 10) no previous treatment of the disease other than external-beam radiation (≤2 sites of therapy) or steroids (<2 months of therapy). Patients with any other malignancies within the past 5 years except basal carcinoma of the skin or carcinoma in situ of the cervix were excluded from the study. Patients were excluded for pregnancy, central nervous system lymphoma, hypersensitivity to boron or mannitol, human immunodeficiency virus, or active hepatitis B or C. All patients provided voluntary, written, informed consent and authorization according to the Health Insurance Portability and Accountability Act before the performance of any study-related procedures. The protocol was approved by the Institutional Review Board before the study was initiated.
Pretreatment Study Evaluation
Pretreatment evaluations included a physical examination; concomitant medication assessment; high-resolution CT scans of the neck, chest, abdomen, and pelvis (alone or in combination with positron emission tomography [PET] studies); and bone marrow aspiration and biopsy. In addition, blood was drawn for serum chemistry analysis, a complete blood count, measurement of serum β2-microglobulin, human immunodeficiency virus antibody, and hepatitis panel.
Chemotherapy Regimen and Supportive Care
The primary objective was to determine the maximum tolerated dose (MTD) and toxicity profile of bortezomib (Velcade; Millennium Pharmaceuticals, Cambridge, Mass) administered in combinations with R-CHOP (VR-CHOP) to patients with untreated FL and indolent NHL. Secondary objectives included assessment of the overall response (OR) and complete Response (CR) rates, progression-free survival (PFS), and overall survival (OS). Because this study also incorporated maintenance strategies, another objective was to estimate the conversion to CR rate associated with maintenance R plus bortezomib after VR-CHOP and to assess the tolerance of maintenance R in combination with bortezomib, which will be evaluated in the phase 2 portion of this study.
Six to 8 cycles of therapy administered every 21 days were planned, and the total number of cycles was determined by the treating physician. Because our inclusion criteria allowed for the treatment of grade 3b FL and previous trials in low-grade lymphoma reported on the feasibility of delivering up to 8 cycles of R with chemotherapy,22 patients were able to receive a total of 8 cycles of chemotherapy. We wanted to examine the safety of bortezomib in combination given up to 8 cycles.
Standard premedications consisted of intravenous ondansetron 8 mg plus prednisone 100 mg orally (on day 1 of 5 for prednisone) for nausea and diphenhydramine 50 mg orally plus acetaminophen 650 mg orally for the prevention of hypersensitivity reaction to R. Bortezomib was administered by intravenous push on day 1 and was repeated on day 8 of each cycle. On day 1, bortezomib administration was followed by R. R infusion was followed by doxorubicin 50 mg/m2 administered intravenously over 3 to 5 minutes followed by intravenous cyclophosphamide 750 mg/m2 and vincristine capped at 1.5 mg. Table 1 provides the dose-finding schema with titration of bortezomib and vincristine. No other drugs were titrated. White blood cell growth factor support was not allowed for any patient during cycle 1 to permit the assessment of hematologic toxicity in establishing the MTD. Maintenance therapy was explored for all patients. Maintenance therapy was started 3 months after the completion of induction therapy (Fig. 1). For patients who achieved CR, R 375 mg/m2 was administered every 12 weeks until progressive disease (PD) or 2 years of therapy. For patients who achieved stable disease or PR, R 375 mg/m2 along with bortezomib 1.6 mg/m2 was administered weekly for 4 doses every 6 months until PD or 2 years of completed maintenance. The data reported in this article focus on the results from patients who were treated during the induction phase. The results from the maintenance strategy will be reported in an analysis of the phase 2 portion of this study.
Table 1. Phase 1 Dosing Schema
Bortezomib Day 1, Day 8 mg/m2
Vincristine Day 1, Day 8 mg/m2
Number of Patients
No other drugs were titrated. Day 1 infusion following bortezomib included rituximab 375 mg/m2, doxorubicin 50 mg/m2, cyclophosphamide 750 mg/m2, and vincristine with prednisone given at 100mg oral days 1-5.
Abbreviations: NTE, not to exceed.
1.4 (NTE 1.5 mg)
1.4 (NTE 1.5 mg)
1.4 (NTE 1.5 mg)
1.4 (NTE 2.0 mg)
Safety Data Collection
Toxicities were captured and graded according to the National Cancer Institute Common Toxicity Criteria for Adverse Events (NCI CTCAE), version 3.0. Patients had a complete blood count with differential and platelet count performed on days 1 and 8 of each cycle. A cycle of chemotherapy could be held for up to 2 weeks until the patient had a platelet count >100 × 109/L and an absolute neutrophil count of >1.0 × 109/L. Before each cycle, the patient had a complete history and physical examination, including a detailed neurologic evaluation using the Functional Assessment of Cancer Therapy/Gynecologic Oncology Group (FACT/GOG)-Neurotoxicity questionnaire.
The FACT/GOG patient self-assessment questionnaire includes 11 questions that assess the severity and impact of neuropathy symptoms.23 It is comprised of questions regarding sensory symptoms (numbness and discomfort in hands and feet, difficulty feeling the shape of objects), motor symptoms (general weakness, trouble walking, trouble buttoning buttons, joint pain/muscle cramps), and auditory problems (difficulty hearing, buzzing/ringing in ears). We also used nerve-conduction studies, which involved recording responses from both sensory nerves and motor nerves. Both the amplitude (size) of the response and the speed with which an impulse traveled along the nerve (conduction velocity) were measured. These parameters were recorded along distal nerve segments. Proximal segments of the nerves were evaluated by recording the latency of late responses (F-waves and H-reflexes). We also recorded responses from the sural nerves (bilateral), the radial nerve on 1 side, the peroneal and tibial nerves (with F-responses) on 1 side, and the H-reflexes (bilaterally). These electromyography/nerve conduction velocity (EMG/NCV) studies were performed at baseline and after cycle 4.24 The FACT/GOG questionnaire and EMG/NCV testing were additional assessments of neuropathy but were not incorporated formally into the dose-limiting toxicity (DLT) assessment. DLT assessment was based solely on NCI CTCAE coding.
CT scans were performed after every 2 cycles during the induction phase and then every 3 months during maintenance. Additional PET images were obtained in selected patients at follow-up at the discretion of the treating physician, but insufficient data are available to evaluate the usefulness of PET studies with this regimen in these lymphoma subtypes. Revised response criteria for lymphomas were used for the categorization of responses after completion of induction therapy.25
The objective of this phase 1 trial was to determine tolerable dose levels for the combination vincristine and bortezomib (with fixed dosing of R, cyclophosphamide, doxorubicin, and prednisone). The MTD was defined as the regimen at which <30% grade ≥3 nonhematologic toxicity or grade ≥4 hematologic toxicity (lasting >14 days) occurred. The dose-escalation scheme followed the Escalation with Overdose Control (EWOC) protocol, a statistical method that directly incorporates formal safety constraints into the design of phase 1 cancer trials.26-28 EWOC controls the frequency of over dosage by selecting dose levels for use in the trial and stipulates that the predicted proportion of patients administered a dose exceeding the MTD equals a specified upper bound (θ = 0.3). This allows more patients to be treated with potentially therapeutic doses of promising new agents and exposes fewer patients to the deleterious effects of a toxic dose. The dose escalation was conducted in 2 stages. In the first stage, the dose of vincristine was fixed and capped at 1.5 mg, whereas the dose of bortezomib was escalated from 1.0 mg/m2 to its MTD. After the MTD for bortezomib was reached, we calculated the 25th percentile of the posterior distribution of the MTD (with the prior distribution for MTD based on the known probabilities of neuropathy and hematologic toxicity associated with bortezomib and R-CHOP from this study and from other available data; Fig. 2a).17-19 In the second stage, vincristine escalation was planned according to the EWOC algorithm, and bortezomib was to be held constant at the previously determined dose from stage 1. Dose escalation of vincristine was never performed during the execution of the trial after data were presented indicating that the combination of 1.6 mg/m2 of bortezomib with vincristine capped at 2 mg produced excessive neurologic toxicity,18 and we updated the EWOC algorithm and probabilities with these data and stage 1 data. EWOC predicted a posterior median MTD of 1.34 mg/m2 of bortezomib with vincristine capped at 2 mg (Fig. 2b), which was less than the dose established in stage 1; thus, vincristine was not escalated.
Between September 2007 and September 2009, 19 patients were enrolled in phase 1 of the study. All 19 patients completed induction therapy and were assessable for response and toxicity. The clinical characteristics of these patients at presentation are listed in Table 2. The median patient age was 59 years, 58% of patients were men, and most patients presented with stage IV disease (63%). FL was the diagnosis in 53% patients; 4 patients had SLL, 3 patients had MZL, and 2 patients had lymphoplasmacytic lymphoma/Waldenstrom macroglobulinemia (WM). Patients received an average of 7 cycles (range, 6-8 cycles) of therapy. One patient received dose level 0, 6 patients received dose level 1, and 12 patients received dose level 2.
Table 2. Baseline Characteristics of Patients in the Trial
No. of Patients (%), n = 19
Abbreviations: ECOG, Eastern Cooperative Oncology Group; FLIPI, Follicular Lymphoma International Prognostic Index; LDH, lactate dehydrogenase; ULN, upper limit of normal.
Small lymphocytic lymphoma
Marginal zone lymphoma
ECOG performance status
Hematologic and nonhematologic toxicities are listed in Table 3. Grade ≥3 neutropenia was observed in 58% of patients in the study, whereas grade ≥3 thrombocytopenia was observed in 1 patient. Five patients experienced grade 4 neutropenia; however, it did not persist for >14 days in any patient. The most common nonhematologic toxicities were sensory neuropathy (84%; 13 patients, 1 patient, and 2 patients with grade 1, grade 2, and grade 3 sensory neuropathy, respectively), nausea (13% of patients), pain (63% of patients), and fatigue (58% of patients).
Table 3. Adverse Events in Patients in the Trial, n = 19
No. of Patients (%)
Dose-limiting grade 4 neutropenia was observed in 4 patients.
Dose-limiting grade 3 sensory neuropathy was observed in 2 patients.
Across all dose levels, DLT was observed in 5 patients (1 patient at dose level 0, 1 patient at dose level 1, and 3 patients at dose level 2). The only patient at dose level 0 developed grade 4 neutropenia, whereas 1 of 6 patients at dose level 1 developed grade 4 neutropenia. Of the 12 patients at dose level 2, 1 patient each developed grade 4 neutropenia and grade 3 sensory neuropathy, and 1 patient developed both grade 4 neutropenia and grade 3 sensory neuropathy.
To examine neurotoxicity, we used 3 modalities: the FACT/GOG questionnaire for patient self-assessments, NCI CTCAE neurotoxicity scores assigned by investigators, and EMG/NCV studies to detect subclinical neurologic deficits. Only the NCI CTCAE neurotoxicity assessment was used in determining DLT. FACT neurotoxicity questionnaires were completed by 18 patients at baseline and after cycle 2, by 16 patients after cycle 4, and by 15 patients after the completion of therapy. After cycle 4, ≥88% of patients reported little or no neuropathy for each FACT item. By the end of the treatment, >65% of patients reported little or no neuropathy for each FACT item (Fig. 3). The most common neurologic issues reported on the FACT questionnaire at the end of treatment were numbness in the hands or feet (17%), discomfort in the hands (17%), and discomfort in the feet (17%). On the basis of EMG/NCV studies, 3 of 18 patients had subclinical sensory or motor deficits at baseline (Table 4). Two patients who had normal baseline EMG/NCV studies experienced painful sensory neuropathy (grade 3) after the onset of therapy. Two different patients who had abnormal baseline NCV studies had no symptoms according to the FACT/GOG questionnaire, whereas all patients who reported baseline symptoms on the FACT/GOG had normal NCV studies. Repeat EMG/NCV studies were conducted in 16 patients after cycle 4 and demonstrated sensory deficit in 2 patients at dose level 2, motor deficit in 3 patients (2 patients at dose level 1, 1 patient at dose level 2), and both sensory and motor deficit in 1 patient at dose level 1. However, none of these deficits were associated with neuropathy according to FACT self-reporting or NCI CTCAE neurotoxicity scores.
Table 4. Neurotoxicity Based on Nerve Conduction Studies
Bortezomib was held for a single cycle in 2 patients with sensory neuropathy, and 1 patient with grade 4 neutropenia had a cycle delayed for 2 weeks because of poor recovery of neutrophil counts. All other patients completed 6 or 8 cycles as planned.
Among 19 patients across all dose levels, the OR rate at the end of induction treatment was 100% (CR, 68%; PR, 32%). The diagnosis-specific CR rates were 80% for FL, 25% for SLL, 100% for MZL, and 50% for WM. The median OS and PFS were not reached. One patient with stage IV FL who achieved CR after 6 cycles of chemotherapy relapsed within 1 month of completing induction therapy, had transformation to Burkitt lymphoma with central nervous system involvement within 7 months, and subsequently died with PD. Of the remaining 12 patients who achieved CR, 11 patients remained in CR at last follow-up, and 1 patient with MZL relapsed after 3 years. Of the 6 patients who achieved PR, 1 patient with FL relapsed after 1 year. The 3-year PFS and OS rates were 89.5% (Fig. 4) and 94.7%, respectively.
In this prospective, phase 1, single-institutional study, the addition of bortezomib to R-CHOP with modified vincristine dosing did not increase or add significant toxicity to the regimen. The current regimen produced an OR rate of 100% and a CR rate of 68%, and responses were observed across indolent NHL subtypes and were most prominent in FL. One patient at dose level 0 achieved CR, 5 of 6 patients at dose level 1 achieved CR, and 7 patients at dose level 2 achieved CR. The data suggest that bortezomib at doses up to 1.6 mg/m2 given on days 1 and 8 with vincristine capped at 1.5 mg can be safely combined with acceptable toxicity in patients who have untreated, low-grade lymphomas. The combination of bortezomib with R-CHOP was well tolerated, and adverse events were manageable. In particular, neurotoxicity with this regimen was not increased substantially compared with R-CHOP1, 29 based on objective assessments with EMG/NCV studies, subjective investigator evaluations, and patient self-report. After cycle 4, >88% of patients reported no or very little neuropathy; and, at the end of therapy, >65% of patients reported little or no neuropathy for each item on the FACT neurotoxicity questionnaire. The FACT questionnaire also revealed that >80% of patients were able to perform tasks of daily living and reported no problems with buttoning, feeling small objects, and walking. To our knowledge, this is the first study of neurotoxicity assessments in patients with lymphoma to incorporate not only CTCAE grading of neurotoxicity but also patient questionnaires and nerve-conduction studies to broadly characterize regimen-related neurotoxicity in patients with NHL. It is noteworthy that baseline EMG/NCV studies did not correlate with patient self-assessments. The nerve-conduction studies revealed that both sensory nerve and motor nerve conduction velocities were affected but that these did not correlate well with clinical assessments. Thus, combined modalities of assessment are needed to fully characterize the neurologic toxicity associated with this regimen and probably with other cancer therapies. Furthermore, the EWOC methodology used for dosing illustrates the advantages of a Bayesian method for dose escalation that can make use of available data to update probabilities and reduce the likelihood of exposing patients to regimens identified with toxicity in other settings.
Novel approaches with combination chemoimmunotherapy appear to have improved OS for patients with FL over time.6, 10 Although R-CHOP has been the most commonly used first-line regimen in the United States, other approaches have emerged that may rival or supplant R-CHOP in the future. Rummel and colleagues presented data on the use of bendamustine and R in patients with FL and mantle cell lymphoma and demonstrated that, whereas the OR rate for patients who received bendamustine-R (93.8%) was similar to the rate for those who received R-CHOP (93.5%), the CR rate was significantly higher for the bendamustine-R group (40.1%) compared with the R-CHOP group (30.8%), and the median PFS was significantly longer for the bendamustine-R group (54.8 months) compared with the R-CHOP group (34.8 months).30 Moreover, R-CHOP was associated more frequently with serious adverse events than bendamustine-R. These data suggest that bendamustine-R may serve as another useful backbone for the exploration of novel combination therapies. The combination of bortezomib-bendamustine-R already has been examined in patients with relapsed, indolent lymphomas31 and is being examined in a first-line ECOG clinical trial.32
Other studies have investigated the combination of bortezomib and R-CHOP predominantly in aggressive lymphoma subtypes. Like our trial, 1 concern in those trials was the neurotoxicity that may develop when using bortezomib and vincristine in combination. In a phase 1 trial, patients with DLBCL (n = 40) and mantle cell lymphoma (n = 36) received standard-dose R-CHOP on a 21-day schedule for 6 cycles with 1 of 3 dose levels of bortezomib (0.7 mg/m2, 1.0 mg/m2 or 1.3 mg/m2) administered on days 1 and 4 of each cycle.17, 33 The MTD of bortezomib with R-CHOP was not reached, and the 1.3 mg/m2 dose level had acceptable toxicity. Peripheral neuropathy was observed in 64% of patients. In 69% of patients, the neuropathy returned to baseline within 6 months after the completion of therapy.33
In another trial of patients with previously untreated B-cell lymphomas, 49 patients received standard R-CHOP and were randomized to biweekly or weekly schedules of bortezomib, with both schedules dose-escalated (biweekly up to 1.3 mg.m2 on days 1, 4, 8, and 11 or weekly up to 1.6 mg/m2 on days 1 and 8).18 The CR/unconfirmed CR rate was 84% for all subtypes combined. However, those investigators noted that grade 2 and 3 neurotoxicity was more common with higher bortezomib dosing (1.3 mg/m2 biweekly and 1.6 mg/m2 weekly) than with lower dosing (1.0 mg/m2 biweekly and 1.3 mg/m2 weekly). For example, grade 3 neurologic toxicity increased from 7% (1 of 14 patients) to 27% (4 of 15 patients) in those who received bortezomib on days 1 and 8 of each cycle when the dose was increased from 1.3 mg/m2 to 1.6 mg/m2.
Bortezomib also has been combined with other vincristine-containing and anthracycline-containing regimens. Romaguera colleagues reported on the results from bortezomib in combination with R plus hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (R-HyperCVAD) alternating with R plus methotrexate/cytarabine (R-M/A) in patients with untreated mantle cell lymphoma. Bortezomib was added to the R-HyperCVAD combination as a fixed dose of 1.3 mg.m2 on days 2 and 5 and was added to the R-M/A regimen after R in increasing doses of 0.7 mg/m2, 1.0 mg/m2, and 1.3 mg/m2 in cohorts of 3 patients. Those authors reported that no significant sensory neuropathy was observed with this combination, and the low incidence of neuropathy may have been related to scheduling the combination of bortezomib and vincristine every other cycle (every 6 weeks).34 In contrast, Kahl and colleagues presented data on a combination of bortezomib with the R-HyperCVAD regimen in which patients received bortezomib on days 1 and 4 and vincristine on day 3. This regimen was repeated every 3 weeks. The dose of bortezomib had to be decreased to 1.3 mg/m2, and vincristine had to be capped at 1 mg to avoid sensory neuropathy.35
Although our study included only 19 patients, R-CHOP and bortezomib demonstrated encouraging efficacy in patients with indolent lymphomas. Of 10 patients with FL who received this regimen, 8 patients achieved CR. In addition, all 3 patients with MZL had CR after 6 cycles of treatment. The neurotoxicity, hematologic toxicities, and other toxicities were limited when the vincristine dose was capped at 1.5 mg and bortezomib was given at the MTD of 1.6 mg/m2. However, data from other trials suggest that substantial neurotoxicity can occur when bortezomib 1.6 mg/m2 is given with standard-dose vincristine. On the basis of our Bayesian EWOC algorithm and the published trials mentioned above, bortezomib can be given safely at 1.6 mg/m2 on days 1 and 8 of R-CHOP when vincristine is capped at 1.5 mg, but it should be reduced to 1.3 mg/m2 if vincristine is capped at 2 mg. Given the in vitro effects of bortezomib in overcoming R-resistant B-cells and the single-agent activity of bortezomib in relapsed lymphomas, we propose that maximizing the bortezomib dosing is of greater importance than maximizing vincristine.
At the time that the current trial was designed, extended schedules of maintenance R had been assessed in patients with FL after initial treatment with single-agent R and after chemotherapy alone, and the results demonstrated a substantial impact of R on PFS and suggested an OS benefit.36-38 However, our trial was designed during the period before results from the Primary Rituximab and Maintenance (PRIMA) trial were known and used a slightly different maintenance strategy than PRIMA38 based on pharmacokinetic studies indicating that, with maintenance R given every 2 to 3 months, nearly all patients maintained active R levels.39-42 Data from separate trials by the Swiss Group for Clinical Cancer Research (SAKK)40 and by Gordan and colleagues41 suggested that repeat dosing was needed approximately every 2 to 3 months to maintain a therapeutic level of 25 μg/mL (the median serum level at 3 months after treatment of responders in the pivotal trial that led to R approval).39 We chose the more convenient schedule for patients in CR of every 3 months in which less R was given with similar expected results. In addition, based on the promising results from a phase 2 trial in patients with relapsed FL,16 a maintenance schedule was designed to test whether giving bortezomib with R could produce CRs for patients who achieved PR or SD after bortezomib plus modified R-CHOP.16, 43 The evaluation of this approach is ongoing and will be assessed as a secondary endpoint with the completion of the phase 2 trial. On the basis of the safety profile, high response rate, and early data on PFS, we continue to enroll patients with untreated FL and MZL in a phase 2 study of R-CHOP and bortezomib given at the established MTD.
This work was supported by a Georgia Cancer Coalition Distinguished Scientist Award to Dr. Flowers, by an American Society of Hematology Amos Medical Faculty Development Award, and by funding from Millennium Pharmaceuticals to Emory University to support the conduct of the trial.
CONFLICT OF INTEREST DISCLOSURES
Dr. Kaufman serves as a consultant to Millennium Pharmaceuticals, Onyx Pharmaceuticals, Keryx Biopharmaceuticals, and Novartis Pharmaceutical. Dr. Lonial serves as a consultant to Millennium Pharmaceuticals, Bristol-Myers Squibb, Celgene Corporation, Novartis Pharmaceutical, Merck and Company, and Onyx Pharmaceuticals. Dr. Lechowicz serves as a consultant to Celgene Corporation and Allos Therapeutics and previously served as a consultant to Eisai Pharmaceuticals. Dr. Heffner serves as a consultant to Millennium Pharmaceuticals. Dr. Flowers serves as a consultant to Celgene Corporation and as an unpaid consultant to Millennium Pharmaceuticals and Genentech. Dr. Sinha receives research funding from Celgene Corporation. Dr. Kaufman receives research funding from Millennium Pharmaceuticals, Merck and Company, and Celgene Corporation. Dr. Lechowicz receives research funding from Allos Therapeutics and previously received research funding from Eisai Pharmaceuticals. Dr. Flowers receives research funding from Millennium Pharmaceuticals, Calistoga Pharmaceuticals, Pfizer, Novartis, and Spectrum Pharmaceuticals.