Durable complete responses from therapy with combined epratuzumab and rituximab

Final results from an international multicenter, phase 2 study in recurrent, indolent, non-Hodgkin lymphoma

Authors

  • John P. Leonard MD,

    Corresponding author
    1. Center for Lymphoma and Myeloma, Division of Hematology/Oncology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, New York
    • Weill Cornell Medical College, 520 East 70th Street, New York, NY 10021===

    Search for more papers by this author
    • Fax: (212) 746-3844

  • Stephen J. Schuster MD,

    1. Division of Hematology/Oncology, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania
    Search for more papers by this author
  • Christos Emmanouilides MD,

    1. Division of Hematology/Oncology, University of California-Los Angeles Medical Center, Los Angeles, California
    Search for more papers by this author
  • Felix Couture MD,

    1. Department of Medicine, University Hospital Center of Quebec, Montreal, Quebec, Canada
    Search for more papers by this author
  • Nick Teoh MD,

    1. Immunomedics, Inc., Morris Plains, New Jersey
    Search for more papers by this author
    • William A. Wegener, Nick Teoh, Morton Coleman, and David M. Goldenberg have employment, board membership, and/or stock ownership in Immunomedics, Inc.

  • William A. Wegener MD,

    1. Immunomedics, Inc., Morris Plains, New Jersey
    Search for more papers by this author
    • William A. Wegener, Nick Teoh, Morton Coleman, and David M. Goldenberg have employment, board membership, and/or stock ownership in Immunomedics, Inc.

  • Morton Coleman MD,

    1. Center for Lymphoma and Myeloma, Division of Hematology/Oncology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, New York
    Search for more papers by this author
    • William A. Wegener, Nick Teoh, Morton Coleman, and David M. Goldenberg have employment, board membership, and/or stock ownership in Immunomedics, Inc.

  • David M. Goldenberg MD

    1. Immunomedics, Inc., Morris Plains, New Jersey
    2. Garden State Cancer Center, Belleville, New Jersey
    Search for more papers by this author
    • William A. Wegener, Nick Teoh, Morton Coleman, and David M. Goldenberg have employment, board membership, and/or stock ownership in Immunomedics, Inc.


  • Presented in abstract form at the 49th Annual Meeting of the American Society of Hematology, Atlanta, Georgia, December 8-11, 2007.

  • J.P.L., W.A.W., and D.M.G. were responsible for study design; Immunomedics, Inc. (W.A.W.) provided epratuzumab; J.P.L., S.J.S., C.E., F.C., and M.C. provided patients; W.A.W. and N.T. collected and assembled data; J.P.L., W.A.W., and N.T. provided data analysis; J.P.L., W.A.W., M.C., and D.M.G. interpreted the data; J.P.L., W.A.W., N.T., and D.M.G. wrote the article; and all authors reviewed and approved the article.

Abstract

BACKGROUND.

In this international, multicenter trial, the authors evaluated rituximab (anti-CD20) plus epratuzumab (anti-CD22) in patients with postchemotherapy relapsed/refractory, indolent non-Hodgkin lymphoma (NHL), including long-term efficacy.

METHODS.

Forty-nine patients with follicular NHL (FL) (N = 41) or small lymphocytic lymphoma (SLL) (N = 7) received intravenous epratuzumab 360 mg/m2 and then intravenous rituximab 375 mg/m2 weekly ×4. The regimen was tolerated well.

RESULTS.

Twenty-two of 41 patients with FL (54%) had an objective response (OR), including 10 (24%) complete responses (CR) (CR/unconfirmed CR [CRu]), whereas 4 of 7 patients with SLL (57%) had ORs, including 3 (43%) with CR/CRu. Rituximab-naive patients (N = 34) had an OR rate of 50% (26% CR/CRu rate), whereas patients who previously responded to rituximab (N = 14) had an OR rate of 64% (29% CR/CRu rate). An OR rate of 85% was observed in patients with FL who had Follicular Lymphoma International Prognostic Index (FLIPI) risk scores of 0 or 1 (N = 13), whereas 28 patients with intermediate or high-risk FLIPI scores (≥2) had an OR rate of 39% (18% CR/CRu rate). In patients with FL, the median response duration was 13.4 months, and that duration increased to 29.1 months for 10 patients who had a CR/CRu, including 4 patients who had durable responses with remissions that continued for >4 years. In patients with SLL, the median response duration was 20 months, including 1 patient who had a response that continued for >3 years.

CONCLUSIONS.

The combination of epratuzumab and rituximab induced durable responses in patients with recurrent, indolent NHL. Cancer 2008. © 2008 American Cancer Society.

Although the non-Hodgkin lymphomas (NHLs) are a diverse collection of diseases,1 most express common B-cell antigens. In 1997, the anti-CD20 monoclonal antibody rituximab (Rituxan; Genentech, South San Francisco, Calif; Biogen Idec Pharmaceuticals, San Diego, Calif) was approved for the treatment of patients with relapsed or refractory, low-grade or follicular, CD20-positive, B-cell NHL after a pivotal trial in which 375 mg/m2 rituximab administered once weekly for 4 consecutive weeks resulted in a 48% overall response rate (6% complete responses [CRs]).2 Other B-cell antigens also provide promising targets, including CD22, a 135-kD glycoprotein that is expressed on the cell surface of mature B cells in follicular NHL (FL) and other NHL subtypes, with expression that ceases when B-cells differentiate into plasma cells.3, 4 A recent summary indicted that CD22 may act to regulate B-lymphocyte survival, mediate adhesion and signal transduction, and regulate B-cell receptor (BCR) activation.5 Initially, an anti-CD22 mouse monoclonal antibody (mLL2; formerly EPB-2) was developed and characterized that had promising in vitro properties.6, 7 Thus, epratuzumab, the resulting humanized monoclonal antibody (hLL2), which has 90% to 95% human origin, greatly reduced the potential for immunogenicity.8 In vitro experiments demonstrated that the antitumor activity of epratuzumab was predominantly by antibody-dependent cytotoxicity,9, 10 and in vivo studies in animals demonstrated that combined therapy with epratuzumab and rituximab or with another CD20 antibody, veltuzumab (hA20), was more effective in controlling lymphoma growth and prolonging survival than either agent alone.10, 11 Other in vitro studies have suggested a mechanism of action different from rituximab, possibly also involving a different signaling pathway in the inhibition of proliferation of CD22-positive malignant B cells.12

In initial clinical studies, patients with indolent and aggressive NHL received increasing doses of epratuzumab administered weekly for 4 consecutive weeks.13, 14 Epratuzumab was tolerated well, no dose-limiting toxicity was observed for doses administered up to 1000 mg/m2, and most infusions were administered within 1 hour. Epratuzumab showed activity as a single agent, particularly in FL, with a 43% objective response (OR) rate in 14 patients who received 360 mg/m2 for 4 weeks. Therefore, this dose was selected for further development and has been employed in follow-up trials, including the current study. However, because rituximab and epratuzumab target different antigens, combining epratuzumab with rituximab could result in higher antitumor activity compared with rituximab alone. On the basis of a small, single-center pilot study indicating that the combination of rituximab with epratuzumab had high activity in both indolent and aggressive NHL,15 the current international multicenter trial was designed to determine the safety and efficacy of epratuzumab in combination with rituximab for patients with low-grade, CD20-positive B-cell NHL with disease that progressed after (ie, refractory or relapsed) at least 1 prior regimen of standard chemotherapy for NHL.16, 17 Here, we report the final study results, including long-term follow-up, demonstrating durable responses that, in some patients, continued for >4 years.

MATERIALS AND METHODS

This open-label, single-arm, multicenter phase 2 study was conducted to evaluate the safety and effectiveness of epratuzumab in combination with rituximab for the treatment of patients with refractory or recurrent, indolent NHL. The study endpoints were safety, efficacy (OR and CR rates, duration of response, time to progression), immunogenicity, and pharmacokinetics. At each participating institution, the governing ethics committee approved the study, and written informed consent was obtained from all patients.

Patient Population

Patients aged ≥18 years were eligible if they had indolent, CD20-positive, B-cell NHL based on local diagnosis and had failed at least 1 prior regimen of standard chemotherapy for NHL. Patients who were rituximab-naive and those who had received rituximab as single agent or in combination with chemotherapy as their last treatment and who demonstrated a partial response (PR) or CR with a time to progression ≥12 months were eligible. Included patients had bidimensionally measurable disease demonstrated on a computed tomography (CT) scan with a least 1 lesion ≥1.5 cm in greatest dimension, an Eastern Cooperative Oncology Group (ECOG) performance status ≤1, and a minimal life expectancy of 4 months. Patients who had disease that was refractory to rituximab (ie, who did not meet the criteria described above) or who had received prior antibody therapy for lymphoma other than rituximab were ineligible; also ineligible were patients with primary or secondary central nervous system lymphoma, known or suspected transformed lymphoma, pleural effusions with positive cytology for lymphoma, or prior radiation therapy to the index lesion(s), unless lesions clearly were progressing or were positive on gallium scans. Other exclusion criteria included evidence of rapid disease progression or symptoms that required intervention within 2 weeks of starting therapy; another primary malignancy within 3 years (other than squamous or basal cell skin cancer, cervical cancer in situ, or stage I prostate cancer); known human antichimeric antibody (HACA) or human antihuman antibody (HAHA) positivity or positive for hepatitis B surface antigen, hepatitis C virus, or human immunodeficiency virus (HIV); known type-1 hypersensitivity or anaphylactic reactions to murine proteins; major surgery within 4 weeks (unless fully recovered); other investigational agents within 30 days; pregnant or breastfeeding; refusal to use adequate contraception precautions; unavailable for follow-up assessment; or any serious nonmalignant condition or infection that likely would compromise patient safety.

Treatment

The epratuzumab/rituximab combination was given on a weekly basis for 4 consecutive weeks, with an intravenous infusion of 360 mg/m2 epratuzumab given first within 1 hour followed at least 1 hour later (but within 24 hours) by a 375 mg/m2 dose of rituximab administered intravenously. All patients were premedicated each week with antihistamines and antipyretics, but steroids were not prescribed routinely.

Safety and Laboratory Tests

At study entry, the most recent pathology specimen obtained from each patient was submitted for central NHL histology review. Formalin-fixed, paraffin-embedded tumor sample evaluations by an independent pathologist also included determining positive CD20 and CD22 status using commercially available anti-CD20 and anti-CD22 antibodies. Patients were monitored for adverse reactions during infusions, with vital signs obtained every 15 minutes until both infusions were complete, and then 30 minutes and 60 minutes later. Patients continued to be monitored for safety at 4 weeks after the last infusion and then every 3 months for at least 1 year. Adverse events (AEs) were classified according to the MedDRA system organ class and preferred term, whereas toxicity grading of AEs and laboratory tests used National Cancer Institute Common Toxicity Criteria, version 2.0. Blood samples for routine safety laboratory tests (serum chemistry, hematology), B-cell counts (CD19-positive), and immunoglobulin (Ig) levels were obtained before each infusion, 4 weeks after the last infusion, and then every 3 months as needed at follow-up evaluations. Serum samples to determine epratuzumab and rituximab levels were obtained at the time of these evaluations and also 10 minutes after each infusion, and some patients who participated in an optional full pharmacokinetics study had additional samples taken after the first and last infusions. Serum samples for determining immunogenicity (HAHA/HACA) were obtained at baseline, before the last infusion, 4 weeks later, and then during follow-up until the samples were negative. Epratuzumab levels and HAHA determinations were performed by the sponsor using an enzyme-linked immunosorbent assay.

Response Evaluation

All patients underwent physical examination and had CT scans obtained of the neck, chest, abdomen, pelvis, and other sites of known disease at baseline, at 4 weeks after the last infusion, and every 3 months thereafter either until the occurrence of disease progression or the initiation of additional therapy. A bone marrow biopsy was required at baseline and, in patients who had bone marrow infiltration, only if needed to confirm a CR. Treatment responses were classified by each investigator using International Workshop Criteria,18 with each patient's best response classified as either a CR, an unconfirmed CR (CRu), a PR, stable disease (SD), or progressive disease (PD).

Statistical Analysis

Descriptive statistics were used to summarize demographics, safety and laboratory data, and treatment response rates. Progression-free survival (PFS), defined as the duration from the first day of study drug administration to the day of disease progression (based on physical or radiologic [CT] evidence), death, or last contact, whichever occurred earliest, was analyzed by using the Kaplan-Meier product-limit method. Duration of response (DR), defined as the duration from the first day of onset of an OR, ie, CR, CRu, or PR, to the day of disease progression, death, or last contact, whichever occurred earliest, was summarized using similar methods. Secondary analyses using similar statistical methods also were performed to evaluate the effects of several covariates, including prior treatment with rituximab, bone marrow involvement, and Follicular Lymphoma International Prognostic Index (FLIPI) risk score.19 The Fisher exact test was used to analyze for significant differences in the OR rates according to the FLIPI score.

RESULTS

Patient Characteristics and Demographics

In total, 49 patients (26 men, 23 women; median age, 61 years) were enrolled at 18 institutions in the United States, Australia, and Canada between September 2001 and February 2003. They were a median of 4 years from their initial diagnosis, and most patients had a good performance status (78% ECOG 0) at study entry but had widespread disease (69% stage III/IV). All patients had received at least 1 prior chemotherapy treatment (range, 1-8 prior treatments), and most patients were rituximab-naive: Only 14 patients (29%) had received 1 prior rituximab-containing regimen. Twenty-four patients (49%) had positive bone marrow involvement at study entry, and 7 patients (14%) had elevated lactate dehydrogenase.

On the basis of the World Health Organization classification,20 all patients initially had locally determined diagnostic classifications of indolent NHL: either grade 1 or 2 lymphoma (FL) or small lymphocytic lymphoma (SLL). The central histology review confirmed the original indolent FL or SLL histology in 42 patients and reclassified the histology in 4 patients (2 SLL were reclassified as FL, 1 FL was reclassified as SLL, and 1 FL was reclassified as aggressive NHL with findings of both FL and diffuse large B-cell lymphoma [DLBCL]). The review was inconclusive regarding the other 3 patients (2 FL, 1 SLL), so they remained characterized by their original diagnostic classification. Thus, 48 patients maintained their classification of indolent NHL (41 FL, 7 SLL), whereas 1 patient was reclassified with aggressive lymphoma (FL and DLBCL). For the 41 patients who had FL, FLIPI scores were assigned on a scale from zero to 5 according to published criteria,19 with 13 patients (32%) considered low risk0-1 and 28 patients (68%) considered intermediate to high risk (≥2) for a poor outcome. Demographics and patient characteristics are summarized in Table 1.

Table 1. Demographics and Patient Information
CharacteristicAll Patients, N=49FL Patients, N=41SLL Patients, N=7
  • FL indicates follicular non-Hodgkin lymphoma; SLL, small lymphocytic lymphoma; ECOG, Eastern Cooperative Oncology Group LDH, lactate dehydrogenase; IULN, institutional upper limit of normal; FLIPI, Follicular Lymphoma International Prognostic Index; NA, not available.

  • *

    Two patients with FL had indeterminate bone marrow biopsy results.

  • Two patients with FL and 2 patients with SLL had no LDH measurements.

Sex, men/women26/2323/182/5
Median age (range), y61 (30-88)60 (30-88)63 (38-74)
ECOG performance status   
 038316
 111101
Disease stage at study entry   
 I660
 II981
 III980
 IV25196
Median time after diagnosis (range), y4.1 (0.3-18.4)4.7 (0.3-18.4)4.1 (1.3-11.6)
Prior treatments   
 Median no. of prior chemotherapies (range)1 (1-8)1 (1-8)1 (1-2)
 Rituximab naive35286
 Rituximab-containing   
  114131
  ≥2000
Bone marrow involvement*24186
Elevated LDH. >IULN761
FLIPI   
  0-1NA13NA
  ≥2NA28NA

Study Drug Administration

Forty-nine patients completed all 4 scheduled infusions of epratuzumab, whereas 48 patients received all 4 rituximab infusions. One patient who had infusion reactions during the first rituximab infusion prematurely discontinued that infusion and continued to receive epratuzumab infusions without further rituximab infusions. For epratuzumab, the median infusion time for each infusion was 1.0 to 1.1 hours (range, 0.3-2.3 hours); and, for rituximab, the median infusion time was 4.4 hours for the first infusion and 3.1 to 3.2 hours for subsequent infusions (range, 1.5-6.8 hours).

Efficacy

Of the 48 patients with indolent NHL (41 FL, 7 SLL), 6 patients had disease progression at the first evaluation 4 weeks after the last infusion and underwent no further response evaluation, whereas the other patients continued scheduled evaluations until disease progression. On the basis of the best response (12 CR, 1 CRu, 13 PR, 16 SD, 6 PD), 26 patients had an OR (OR = CR + CRu + PR). For these 26 patients, the median time from first infusion to onset of an OR was 1.7 months (range, 1.4-10.8 months), and the median time to onset of their best response was 4.2 months (range 1.4-22.4, months). Among the 41 patients with FL, 22 patients (54%) had an OR, including 10 (24%) CRs (CR/CRu), and 4 of 7 patients with SLL (57%) had ORs, including 3 (42.9%) CRs (CR/Cru). The 34 rituximab-naive patients with FL and SLL had an OR rate of 50% (26% CR/CRu) compared with 64.3% (28.6%) for 14 patients who received prior rituximab, whereas 22 patients without bone marrow involvement had an OR (CR/CRu) rate of 63.6% (40.9%) compared with 45.8% (16.7%) for 24 patients with bone marrow spread. Thirteen low-risk patients with FL (FLIPI score, 0-1) had an 85% OR rate (43% CR/CRu), whereas 28 patients with intermediate- to high-risk FL (FLIPI ≥2) had a 39% OR rate (18% CR/CRu): This difference in OR rates achieved statistical significance (P = .009). Table 2 summarizes these efficacy results for the 48 patients who had indolent NHL. The single patient who was evaluated by independent histology review and was diagnosed with a more aggressive histology achieved a PR 4 weeks after treatment, and had disease progression at the next evaluation 1 month later.

Table 2. Treatment Response Rates in Patients With Indolent Non-Hodgkin Lymphoma
Patient GroupNo.ORCR/CRu
%No./Total%No./Total
  1. OR indicates objective response; CR, complete response; CRu, unconfirmed complete response; FL, follicular non-Hodgkin lymphoma; FLIPI, Follicular Lymphoma International Prognostic Index; SLL, small lymphocytic lymphoma.

All patients4854.226/4827.113/48
Histology     
 FL4153.722/4124.410/41
  FLIPI 0-11384.611/1338.55/13
  FLIPI ≥22839.311/2817.95/28
 SLL757.14/742.93/7
Prior rituximab     
 No345017/3426.59/34
 Yes1464.39/1428.64/14

The 48 patients with indolent NHL were followed for a median of 11.1 months (range, 1.5-52.3 months). Figure 1 depicts the Kaplan-Meier estimated survival curves for PFS and DR in the FL and SLL patient subgroups. The 41 patients with FL had a median PFS of 10 months, and the 22 patients with FL who had an OR had a median DR of 13.4 months. In addition, the 10 patients with FL who achieved CR/CRu had durable responses with a median DR of 29.1 months, and 4 patients remained in remission at the last evaluation, including patients with responses that currently have continued for >4 years. The 7 patients with SLL had a median PFS of 20 months, and the 4 patients with SLL who had an OR had a median DR of 20.3 months, including 1 patient with a response that currently has continued for >3 years.

Figure 1.

Kaplan-Meier progression-free survival (PFS) and duration of objective response (OR) curves for subgroups of patients with follicular non-Hodgkin lymphoma (FL) or small lymphocytic lymphoma (SLL).

Adverse Events

None of the 49 patients discontinued the study because of AEs. Two patients died approximately 6 months after treatment, both attributed to progression of disease. The other 47 patients continued on study for at least 1 year, as required by protocol for safety monitoring. Three other deaths occurred during subsequent follow-up: 2 were attributed to disease progression and the third was attributed to interstitial pneumonia after high-dose chemotherapy for disease progression.

Without regard to attribution, 43 of 49 patients (88%) had at least 1 AE during the study, and the most frequently reported (≥10% of patients) were rigors (43%), nausea (25%), pyrexia (25%), fatigue (22%), vomiting (22%), headache (20%), cough (16%), dyspnea (16%), flushing (10%), and sinusitis (10%). Most AEs were mild-moderate (grade 1-2) events, and only 7 patients had events of greater severity (grade ≥3), including asymptomatic neutropenia (laboratory finding at 1 evaluation only and without clinical event), cervical lymphadenitis (swollen malignant lymph node during treatment requiring hospitalization for intravenous antibiotics), peripheral/scrotal edema (unrelated worsening of grade 2 baseline condition), febrile neutropenia (6 weeks after treatment requiring hospitalization for intravenous antibiotics and granulocyte-colony-stimulating factor [G-CSF]), tachycardia, and tachypnea (infusion reactions during first rituximab infusion), cholecystitis (unrelated event 10 weeks after treatment), and cardiopulmonary arrest (because of progression of disease 7 months after treatment).

AEs that were considered likely or suspected to be caused by epratuzumab occurred in 24 of 49 patients (49%), and the most frequent (≥10%) were rigors (18%), nausea (14%), fatigue (12%), vomiting (12%), and headache (10%). Most were infusion reactions that occurred predominantly at first infusion, with ≤5 patients having such events at each of the subsequent infusions. In addition, these were predominantly grade 1 and 2 AEs, and only 3 grade ≥3 events were described previously (asymptomatic neutropenia, cervical lymphadenitis, febrile neutropenia), and they were considered possibly study drug related.

Four AEs were reported as serious AEs (SAEs), including the 3 grade ≥3 events described above (cervical lymphadenitis, febrile neutropenia, cardiopulmonary arrest). The remaining SAE occurred in a patient who received fludarabine and cyclophosphamide after disease progression occurred 8 weeks after treatment; 4 weeks later, the patient developed disseminated herpes zoster attributed to neutropenia from the chemotherapy and was hospitalized for intravenous antivirals and GM-CSF.

Seventeen of 49 patients (35%) had 1 or more infections, including 5 upper and 1 lower respiratory tract infections, 5 sinus and 3 nasopharynx infections, 2 dental abscesses, 2 herpes zoster and 2 herpes simplex infections, 1 catheter site infection, and 1 cervical lymphadenitis. Of these, 3 patients had 5 infections that were considered possibly study drug related (upper and lower respiratory tract infections, sinusitis, herpes simplex, cervical lymphadenitis), 2 infections (herpes zoster, cervical lymphadenitis) were SAEs, and 3 infections (dental abscess, herpes zoster, cervical lymphadenitis) required treatment with intravenous antibiotics/antivirals.

Routine Laboratory Tests

Blood samples for standard laboratory safety tests (serum chemistry, hematology), were obtained before each infusion, 4 weeks after the last infusion, and every 3 months thereafter as needed at follow-up evaluations. Postbaseline increases in toxicity grades according to National Cancer Institute Common Toxicity Criteria version 2.0 were infrequent and were limited predominantly to grade 1 and 2 events (Table 3). There were no grade 4 cytopenias, only 4 patients developed grade 3 cytopenias (2 lymphopenias, 1 neutropenia, 1 leukopenia), and no patients developed grade 3 or 4 renal or hepatic abnormalities.

Table 3. Changes in Safety Laboratory Tests: Patients With Increases in Common Toxicity Criteria Toxicity Grades From Baseline (N=49)
TestMaximum Post-Treatment Grade
1234NA
  1. NA indicates not available; WBC, white blood count; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; ALT, alanine aminotransferase; AST, aspartate aminotransferase.

Hemoglobin31000
Platelets40000
WBC56100
ANC12106
ALC22200
Creatinine10001
Total bilirubin21001
ALT30001
AST11002

Pharmacokinetics

Samples for the determination of serum epratuzumab and rituximab concentrations were collected at baseline, just before each infusion, within 10 minutes after the completion of each infusion, 4 weeks after the last treatment, and at 3 months, 6 months, 9 months, and 12 months thereafter. Maximum epratuzumab and rituximab serum concentrations (mean ± standard deviation), respectively, were 194 ± 56 μg/mL and 187 ± 39 μg/mL after completing the first infusion and 322 ± 89 μg/mL and 327 ± 97 μg/mL after the fourth infusion. A subset of patients had optional additional samples collected at 24 hours, 48 hours, 72 hours, and 120 hours after the first and fourth infusions and 336 hours after the fourth infusion. After the fourth infusion, the resulting serum terminal-phase half-lives (mean ± standard deviation) for epratuzumab and rituximab were 22.7 ± 7.2 days and 31.7 ± 7.8 days, respectively. These results, as reported previously in abstract form,21, 22 were comparable to the values determined previously when epratuzumab and rituximab were administered as single agents and, thus, were not affected when the 2 antibodies were administered in combination.

Consistent with data reported in the literature,2 median serum levels of rituximab were higher in responders than nonresponders, and the difference achieved statistical significance in this study before each infusion as well as at 4 weeks and 3 months after the last infusion. For epratuzumab, differences in median serum levels were not apparent between responders and nonresponders at most time points, except at 4 weeks and 3 months after last infusion, when responders had statistically significantly higher values compared with nonresponders.

B-Cell and Immunologic Changes

The independent pathology assessment of submitted pathology samples determined that all patients in this study had tissue samples that were both CD20-negative and CD22-positive. These results were part of a larger assessment in B-cell NHL histologies demonstrating that 99% of all indolent NHL samples tested were CD22-positive.3

Blood samples for determination of B-cell levels (CD19) were collected at baseline, before each infusion, at 4 weeks after the last infusion, and at 3-month intervals thereafter until recovery to baseline. Median B-cell levels decreased 95% from baseline after the first infusion and after therapy, and most patients (≥44) continued to have samples obtained until at least 4 months after the last infusion (approximately 5 months after baseline) and, in 35 patients, 7 months after last infusion (approximately 8 months after baseline). The onset of B-cell recovery based on the median levels from a smaller subset of 30 patients was 10 months after the last infusion (approximately 11 months after baseline).

Quantitative serum immunoglobulin (Ig) levels were obtained from serum samples that were evaluated at baseline, before infusions, and every 3 months thereafter for at least 1 year. There was no evidence that combination therapy altered Ig levels, because IgG, IgM, IgA, and IgE levels were analyzed separately; and, in each case, both the mean and median percent changes from baseline at each evaluation were small (≤15% at most time points).

Immunogenicity (Human Antihuman Antibody and Human Antichimeric Antibody)

Samples for antiepratuzumab (HAHA) and antirituximab (HACA) analyses were collected at baseline and before the fourth infusion; then, samples were scheduled to be collected at 1 month, 3 months, 6 months, 9 months, 12 months, and 15 months. Of the 49 patients, none (0%) developed positive HAHA, whereas 4 patients developed transient positive HACA of uncertain significance at 1 or 2 evaluations, because no clinical sequelae or apparent pharmacokinetic changes occurred for those patients.

DISCUSSION

Initially, the safety and efficacy of rituximab were established in patients with relapsed/refractory, low-grade, or follicular NHL, with a pivotal trial reporting an OR of 48% and a DR of 11.2 months and with mild-to-moderate side effects.2, 23 Because subsequent studies demonstrated that treatment efficacy improved when rituximab was combined with other agents, particularly chemotherapy,24 this antibody is now used extensively in such regimens in most NHL types and settings.25, 26

Combinations of rituximab with virtually all other therapeutic agents, including interferon-alpha, interleukin-2, proteosome inhibitors, and other antibodies, also have been pursued,26 but most attention in terms of biologicals has been in antibody combinations. The first demonstration in oncology of any antibody combination suggesting acceptable toxicity and the possibility of improved efficacy was with rituximab and epratuzumab, both clinically and preclinically,10, 11, 15 but other antibodies also have been investigated in combination with rituximab27 based on the supposition that different targets and mechanisms of action could have additive or synergistic effects without increased toxicity.

We have demonstrated that combination therapy with 360 mg/m2 epratuzumab and 375 mg/m2 rituximab was tolerated well when administered once weekly for 4 consecutive weeks. All patients were able to complete combination therapy except for 1 patient who had a reaction during the first rituximab infusion and who continued to complete the remaining epratuzumab infusions but did not receive further rituximab infusions. In addition, epratuzumab and rituximab, when administered in combination, did not require longer infusion times than previously reported when administered separately. Infusion reactions in this study occurred predominately during the first infusion, with low-grade (mild-moderate) chills, rigors, fever, and other constitutional events qualitatively similar to those observed previously with either agent alone. No grade 3 or 4 reactions occurred during epratuzumab infusions, and only 1 patient had a more serious grade 3 reaction during the first rituximab infusion. Because grade 3 and 4 reactions have been reported with rituximab when it was administered as a single agent,23 these findings further support the safety of combination therapy with this administration schedule.

The combination of full doses of these 2 antibodies had no apparent effect on safety laboratories, because both standard hematology and serum chemistries as well as quantitative serum Ig levels (IgG, IgM, IgA, IgE) showed no consistent patterns of change from baseline after combination therapy. On the basis of HAHA and HACA findings, none of the 49 patients developed HAHA after treatment, whereas 4 patients developed transient positive HACA of uncertain significance at 1 or 2 evaluation times, because no clinical sequelae or apparent pharmacokinetic changes occurred for those patients. Thus, combination therapy does not appear to increase the immunogenicity of these antibodies.

Combination therapy effectively depletes peripheral B cells, beginning with the first infusion. Thus, epratuzumab, which decreases B-cell levels only modestly when given as a single agent,13, 14 does not interfere with the depletion known to occur when rituximab is given as a single agent.23, 28 Although the data are based on a smaller subset of patients, the recovery of B-cell levels within the first year also appears comparable to results previously obtained with single-agent rituximab.23, 28 Pharmacokinetic findings also determined that the mean serum terminal-phase half-life of both epratuzumab and rituximab was not affected when the 2 antibodies were administered in combination.

This multicenter study demonstrated that, in indolent FL, 22 of 41 patients (54%) achieved an OR according to International Working Group response criteria, including 10 patients (24%) who achieved a CR/CRu. It is reassuring that the current study confirmed the pilot results from the single-center study of the combination of rituximab and epratuzumab, which reported an OR of 67% in 15 patients with FL.15 Although the latter study also produced a higher CR/CRu rate of 56%, all of those patients were rituximab-naive, had received less prior therapy, and had a longer period from the prior therapy until they were accrued to the trial. A European multicenter trial that studied rituximab combined with epratuzumab in patients with low-grade FL and in patients with aggressive NHL reported an OR of 64% and a CR/CRu rate of 24% in patients FL, and those with FL in the FLIPI 0 to 1 group had an 83% OR rate (CR/CRu, 33%).29 Indeed, the FLIPI 0 to 1 group in the current study had an 84.6% OR rate and a 38.5% CR/CRu rate.

The median DRs of 11.2 months and 15 months obtained in patients with relapsed/refractory, indolent NHL after the first application of rituximab2, 23 and after rituximab retreatment,23, 30 respectively, are consistent with the 13.4-month DR observed for all FL responders (with and without prior rituximab therapy) in this study. However, in another, smaller study of 13 patients with indolent NHL who were retreated with rituximab, an OR rate of 38% (all PRs) and a median PFS of 5.1 months were reproted.31 We appreciate, however, that it is difficult to compare results between studies with potentially different patient populations, especially when using data from nonrandomized, phase 2 trials. Nevertheless, based on this study and the other 2 trials that studied the combination of rituximab and epratuzumab,15, 29 totaling over 130 patients with NHL, the overall findings demonstrate that, in indolent lymphoma, combined rituximab plus epratuzumab therapy is effective (with CR/CRu rates that appear to exceed those reported for rituximab alone), ORs are durable, and, most important, our new evidence in the current report that patients who achieved CRs have experienced durable remissions that appear to exceed those reported in other trials in patients with FL who were given the standard 4-week rituximab schedule. Indeed, some remissions currently are continuing for more than 4 years.

There were only 7 patients with SLL in this study, but 4 of those patients (57%) achieved an OR, including 3 patients (43%) with CRs (CR/CRu). For those patients with SLL who achieved an OR, the median DR was 20 months, including 1 patient with a response that continued for at least 3 years. Thus, although the current results were limited by the small number of SLL patients enrolled, combination therapy also appeared to be effective in this subgroup.

Subgroup analysis performed among all patients with low-grade NHL examined the influence of prior rituximab therapy, bone marrow involvement, and FLIPI score (for FL patients only) on response. Exposure to prior rituximab did not appear to decrease the effectiveness of combination therapy to achieve an OR or a CR. In fact, patients with prior exposure to rituximab responded very well to the antibody combination, with an overall response of 50% and a CR/CRu rate of 26%, which appears to be somewhat better than the rates reported (40% and 11%, respectively) from retreatment with rituximab alone.30 Patients with bone marrow involvement or higher risk FLIPI scores generally are more difficult to treat, but combination therapy appeared to be active in these groups, although the rates of OR and particularly CR were higher, as expected, in patients without bone marrow involvement or with low-risk FLIPI 0-1 scores (ie, 84.6% OR and 38.5% CR/CRu).

The interesting results from this trial have lead to the development of Cancer and Leukemia Group B trial 50701, a study of this combination with an extended dosing regimen as initial treatment (without chemotherapy) for FL. In addition to FL, it has been reported that 4 of 6 patients (67%) with DLBCL had an OR, and 3 patients achieved a CR.15 An OR rate of 47% (33% CR/CRu) also was observed in 15 patients DLBCL in the prior multicenter trial of this combination.29 On the basis of these encouraging results, a trial combining these 2 antibodies with conventional chemotherapy (combined cyclophosphamide, doxorubicin, vincristine, and prednisone [CHOP]) indicated that, among 15 previously untreated patients with DLBCL, there was an OR of 87% (67% CR, 20% PR), a 2-year PFS, and an overall survival of 86%.32 Thus, the combination of rituximab with epratuzumab also appears to produce encouraging activity in patients with DLBCL, even in combination with CHOP. A current trial by the North Central Cancer Treatment Group (N0409) currently is evaluating the role of combined epratuzumab, rituximab, and CHOP as initial treatment for DLBCL. These and other ongoing studies will serve to establish more definitively the role of combination antibody therapy with epratuzumab plus rituximab as initial treatment for B-cell lymphoma.

Acknowledgements

We thank the patients who participated in and the other investigators who contributed patients to this study (Mary Ackerman, South Carolina Oncology Associates, Columbia, SC; Linda Bosserman, Wilshire Oncology Medical Group, Pomona, Calif; Andre Goy, M. D. Anderson Cancer Center, Houston, Tex; Stephanie Gregory, Rush-Presbyterian-St. Lukes Medical Center, Chicago, Ill; John Hainsworth, Sarah Cannon Cancer Center, Nashville, Tenn; Richard Herrmann, Royal Perth Hospital, Perth, Western Australia, Australia; Eddie HuCentral Oncology/Hematology Medical Group, Alhambra, Calif; Anthony Mills, Princess Alexandra Hospital, Woolloongabba, Queensland, Australia; Oliver Press, Fred Hutchinson Cancer Research Center, Seattle, Wash; Ian Quirt, Princess Margaret Hospital, Toronto, Canada; Lee Schwartzberg, The West Clinic, Memphis, Tenn; Israel Wiznitzer, Broward General Medical Center, Fort Lauderdale, Fla).

Ancillary