Ocrelizumab, a humanized anti-CD20 monoclonal antibody, was studied in a first-in-human trial in rheumatoid arthritis (RA) patients receiving concomitant methotrexate (MTX).
Ocrelizumab, a humanized anti-CD20 monoclonal antibody, was studied in a first-in-human trial in rheumatoid arthritis (RA) patients receiving concomitant methotrexate (MTX).
The ACTION trial was a combined phase I/II study of placebo plus MTX versus ocrelizumab plus MTX in 237 RA patients (intent-to-treat population). During phase I, 45 patients were treated with 1 of 5 escalating doses of study drug (infusions on days 1 and 15, 10–1,000 mg per each infusion). An additional 192 patients were randomized during phase II. Eligible patients had active disease, an inadequate response to treatment with at least MTX, rheumatoid factor positivity, and elevated levels of acute-phase reactants. The total study duration was 72 weeks. B cell pharmacodynamics over time was investigated.
Baseline demographics were similar among the treatment groups. Based on the entire 72-week data set, the incidence of serious adverse events in the ocrelizumab-treated patients was 17.9%, as compared with 14.6% in placebo-treated patients. The incidence of serious infections was 2.0% in all ocrelizumab-treated patients and 4.9% in placebo-treated patients. Infusion-associated adverse events were mostly grade 1 or grade 2 and were more frequent around the time of the first infusion. No serious infusion-associated adverse events were reported in the ocrelizumab group. Evidence of clinical activity was observed at all doses evaluated. Peripheral B cell depletion after infusion was rapid at all doses, with earlier repletion of B cells at doses of 10 mg and 50 mg. Human anti-human antibodies were detected in 19% and 10%, respectively, of those receiving 10 mg and 50 mg of ocrelizumab, compared with 0–5% of those receiving 200, 500, and 1,000 mg.
Ocrelizumab therapy in combination with MTX was well tolerated. Doses of 200 mg (2 infusions) and higher showed better clinical responses, better reduction of C-reactive protein levels, and very low immunogenicity.
Rheumatoid arthritis (RA) is a chronic systemic autoimmune disease characterized by symmetric inflammation of affected joints (1, 2). Previous clinical experience with rituximab, a chimeric monoclonal anti-CD20 antibody, has demonstrated clinical benefit in RA, thereby establishing the important role of B cells in the pathogenesis of this disease (3–5). Several critical B cell functions may contribute to this effect, including autoantibody secretion, autoantigen presentation, proinflammatory cytokine production, and regulation of dendritic cell function (6–11).
Ocrelizumab is a novel humanized monoclonal anti-CD20 antibody constructed with recombinant DNA techniques (12) and designed to selectively target CD20+ B cells. In vitro characterization of ocrelizumab demonstrated enhanced antibody-dependent cell-mediated cytotoxicity (ADCC) and reduced complement-dependent cytotoxicity (CDC) as compared with rituximab (13). Ocrelizumab binds to a different, but overlapping, epitope of the extracellular domain of CD20 as compared with rituximab.
To investigate ocrelizumab for the first time in patients, we conducted a combined phase I/II placebo-controlled trial across a wide range of doses in patients with moderate to severe RA. In addition, the study was designed to characterize the relationship between the dose level and clinical activity, as well as the extent and duration of B cell depletion after a single course of treatment.
Patients with moderate to severe RA were recruited from 47 rheumatology clinics in the US. All sites were Institutional Review Board–approved, and all patients provided voluntary written informed consent. Eligible patients were between the ages of 18 and 80 years, fulfilled the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) 1987 revised criteria for RA (14), and had shown an inadequate response to therapy with no more than 6 conventional or biologic disease-modifying antirheumatic drugs (DMARDs) (including methotrexate [MTX] at a dosage of 10–25 mg/week for ≥12 weeks). Active disease was defined as a swollen joint count of ≥8, a tender joint count of ≥8, and either a serum C-reactive protein (CRP) level of ≥1.5 mg/dl or an erythrocyte sedimentation rate (ESR) of ≥28 mm/hour. Enrollment criteria required all patients to be rheumatoid factor (RF) positive (≥20 units/ml).
Key exclusion criteria were additional autoimmune disorders, previous treatment with cell-depleting agents, a neutrophil count of <1.5 × 103/μl, a platelet count of <100,000/μl, an IgG level of <5.65 mg/ml, and an IgM level of <0.55 mg/ml.
The ACTION trial (a randomized, placebo-controlled, blinded, phase I/II study of escalating doses of ocrelizumab in patients with moderate to severe RA on stable doses of concomitant MTX) was a phase I/II randomized, blinded, placebo-controlled, dose-ranging study evaluating the safety of ocrelizumab in combination with MTX in patients with RA. All other DMARDs were withdrawn at least 4 weeks prior to randomization (8 weeks for etanercept, infliximab, adalimumab, and leflunomide). Patients had received 10–25 mg of MTX weekly for at least 12 weeks (stable dosage for at least 4 weeks) before treatment. Stable dosages of oral glucocorticoids (prednisone equivalent of up to 10 mg/day) and nonsteroidal antiinflammatory drugs were permitted.
During phase I, 45 patients were treated in sequential cohorts with escalating dosages of ocrelizumab or placebo, and safety was evaluated. The first cohort received 2 infusions of 10 mg each. Patients began enrolling into the next higher dosage cohort 72 hours after the last patient in the previous dosage cohort received the second infusion of study drug (Figure 1A). An interim safety analysis was conducted after phase I, and subsequently, phase II was opened for an additional 192 patients who were randomized in parallel to receive 1 infusion on day 1 and 1 infusion on day 15 of either placebo or 10, 50, 200, 500, or 1,000 mg of ocrelizumab (Figure 1B). Patients were randomized according to a fixed-block design for phase I, whereas a dynamic allocation algorithm was applied for phase II to balance for the tender joint count at baseline and the study center. Patients and investigators were blinded to the study medication but were unblinded to dosage level because of the differences in the infusion times. It was recommended that patients take both an oral antihistamine and acetaminophen before the infusion, but no intravenous corticosteroids were given as premedication.
Adverse events (AEs) reported during the study period were classified according to Medical Dictionary for Regulatory Activities (MedDRA) terminology and were graded according to the Common Toxicity Criteria of the National Cancer Institute (version 3.0). All treatment-emergent AEs as well as serious AEs (SAEs) were listed and were summarized by treatment group.
Clinical activity was evaluated as a secondary end point via the response according to the ACR criteria for 20% improvement (ACR20), 50% improvement (ACR50), and 70% improvement (ACR70) (15) at week 24 after initiation of treatment with placebo or ocrelizumab. Efficacy assessments performed after week 24 were considered to be of limited utility, because the use of alternative DMARDs was permitted after week 24.
The Disease Activity Score based on a 28-joint assessment (DAS28) (16) was also determined at 24 weeks. Based on these data, DAS28 remission responses (DAS28 score <2.6) as well as European League Against Rheumatism (EULAR) responses categorized as good (DAS28 score ≤3.2 and improvement of >1.2) (17) were assessed. The DAS28 was calculated using measurements of the ESR. In addition, we examined the proportion of patients who had a swollen joint count of zero.
Blood samples were collected at the time of the first infusion and at weeks 2, 4, 12, 24, and 36. Sera were prepared for determination of human anti-human antibodies and analyzed by electrochemiluminescence assay. Circulating B cells were measured by fluorescence-activated cell sorting using labeled antibodies against CD19, a marker also present on B cells.
All data analyses were conducted using descriptive statistics, which were based on the intent-to-treat principle, except where stated otherwise. Results were summarized separately for the placebo group (pooled across all dosage levels) and for each of the 5 ocrelizumab groups. For categorical efficacy response analyses, a nonresponder imputation was applied to all patients who did not complete the visit at week 24, required an increase in the dosage of either MTX or corticosteroids, or required additional therapeutic interventions before week 24. Statistical analyses were performed using SAS (SAS Institute, Cary, NC) or S-Plus (Insightful, Seattle, WA) software.
The sample sizes for the dose-escalation phase (5, 10, 10, 10, and 10 patients receiving 10, 50, 200, 500, and 1,000 mg of ocrelizumab, respectively) were chosen to detect major intolerability signals. The combined phase I/II sample size of ∼40 patients per dose level was selected to provide adequate characterization of the relationship between the dose level and the rates of safety events and efficacy responses. Consistent with the nature of phase II trials, this study was not designed to provide statistical confirmation of efficacy.
A total of 237 patients were randomized and treated, as outlined in Figures 1A–C. Baseline characteristics were similar among the 6 treatment groups (Table 1). Patients had longstanding RA (mean duration 9.3–11.3 years) and active disease, as demonstrated by the elevated tender joint counts, swollen joint counts, DAS28 scores, ESRs, and CRP levels. Approximately 47% of all patients had previously taken tumor necrosis factor (TNF) inhibitors.
|Characteristic||Placebo (n = 41)||Ocrelizumab|
|10 mg (n = 36)||50 mg (n = 40)||200 mg (n = 40)||500 mg (n = 40)||1,000 mg (n = 40)|
|Previous TNF inhibitor use, %||46||50||45||40||50||53|
|Previous no. of DMARDs taken||2.2||2.1||2.1||2.1||2.5||2.9|
|Disease duration, years||10.9||11.1||10.5||11.3||10.5||9.3|
|Swollen joint count (68 joints assessed)||23.7||23.6||21.0||19.9||20.3||21.0|
|Tender joint count (66 joints assessed)||31.6||31.7||33.3||31.7||31.6||31.3|
|Serum RF, units/ml||716||470||525||973||514||480|
|Physician's assessment of disease activity, by VAS (0–100-mm scale)||69.2||70.5||67.4||69.3||66.6||68.4|
|Patient's assessment of disease activity, by VAS (0–100-mm scale)||68.4||61.0||65.0||65.7||62.9||65.5|
|Patient's assessment of pain, by VAS (0–100-mm scale)||62.6||55.1||59.3||64.6||58.7||60.3|
|HAQ score (range 0–3)||1.7||1.6||1.7||1.6||1.6||1.5|
|FACIT fatigue score (range 0–52)||26.9||24.3||25.2||25.5||26.9||25.4|
Of the 237 randomized patients in the intent-to-treat population, 212 (89%) completed the week 24 assessment, and 167 (70%) completed the safety followup period through week 72 (Figure 1C). Among the 25 patients with missing data at week 24, 14 withdrew prior to week 24 and 11 missed the week 24 visit. Of the 14 patients who withdrew prior to the week 24 visit, 2 patients withdrew because of AEs; both were in the group receiving 200 mg. The remaining 12 patients were withdrawn according to the physician's or patient's decision.
Safety data are reported for the entire study period of 72 weeks (Table 2). The incidence of AEs was similar across treatment groups. SAEs were reported in 17.9% of ocrelizumab-treated patients in all 5 dosage groups (35 of 196) and 14.6% of placebo patients (6 of 41), with 1 drug-related SAE in the placebo group and 4 drug-related SAEs in the 5 ocrelizumab groups. No pattern of SAEs was apparent among the total of 41 observed SAEs. Reported drug-related SAEs included 1 patient receiving placebo who experienced supraventricular tachycardia (grade 2) on day 1 and ophthalmic herpes zoster (grade 3) on day 101 after infusion, 1 patient receiving 200 mg of ocrelizumab who had an abdominal wall mass (grade 3) on day 77, and 3 patients receiving 500 mg of ocrelizumab, 1 of whom had basal cell carcinoma (grade 2), 1 who had febrile neutropenia (grade 3), and 1 who had lung adenocarcinoma (grade 3) on day 167 and sepsis (grade 3) on day 229.
|Placebo (n = 41)||Ocrelizumab|
|10 mg (n = 36)||50 mg (n = 40)||200 mg (n = 40)||500 mg (n = 40)||1,000 mg (n = 40)||All doses (n = 196)|
|All AEs||38 (93)||35 (97)||38 (95)||40 (100)||40 (100)||38 (95)||191 (97)|
|Withdrawals due to AEs||0||0||0||2 (5.0)†||0||1 (2.5)||3 (1.5)|
|Any SAEs||6 (14.6)||5 (13.9)||8 (20.0)||7 (17.5)||9 (22.5)||6 (15.0)||35 (17.9)|
|Drug-related SAEs||1 (2.4)||0||0||1 (2.5)||3 (7.5)||0||4 (2.0)|
|Serious infections||2 (4.9)||0||0||3 (7.5)||1 (2.5)||0||4 (2.0)|
|All infections||26 (63)||12 (33)||18 (45)||24 (60)||24 (60)||23 (58)||101 (52)|
|Malignancies||1 (2.4)||1 (2.8)||2 (5.0)||1 (2.5)||3 (7.5)||0||7 (3.6)|
|Infusion-associated SAEs||1 (2)||0||0||0||0||0||0|
|First infusion||11 (27)||19 (53)||18 (45)||18 (45)||26 (65)||19 (48)||100 (51)|
|Second infusion§||8 (20)||3 (8)||9 (23)||8 (20)||8 (20)||5 (13)||33 (17)|
Serious infections were reported by 4.9% of patients receiving placebo (2 of 41) and 2.0% of patients receiving ocrelizumab (4 of 196) (Table 2). All serious infections resolved without sequelae. In general, the most frequently reported infections were upper respiratory tract, urinary tract, and viral infections.
Any AE occurring during or within 24 hours of infusion was defined as an infusion-associated AE (Table 2). No intravenous corticosteroids were given before any infusion. No infusion-associated SAE occurred in the ocrelizumab-treated patients, 1 infusion-associated SAE (supraventricular tachycardia) occurred in a placebo-treated patient. The majority of infusion-associated AEs (98%) were mild to moderate. The incidence of infusion-associated AEs following the first infusion was similar among all ocrelizumab groups and was higher than that in the placebo group (Table 2). The most common infusion-associated AEs included nausea, chills, headache, pyrexia, and dizziness. The rates of infusion-associated AEs after the second infusion were similar in the placebo and ocrelizumab groups. Seven patients did not receive a second infusion. One grade 3 event occurring upon the first infusion was reported in each of the groups receiving 500 mg and 1,000 mg. Two grade 3 events were reported in the group receiving 10 mg of ocrelizumab, both of which were abnormal laboratory results noted in patients who tolerated the infusion and had no clinical symptoms. There were no grade 4 or grade 5 infusion-associated AEs.
Malignancies were observed in 3.6% patients receiving ocrelizumab (7 of 196) and 2.4% patients receiving placebo (1 of 41), as shown in Table 2. No obvious pattern of malignancies was apparent. Observed malignancies included metastatic ovarian cancer in 1 patient receiving placebo, 2 basal cell carcinomas in 1 patient receiving 10 mg of ocrelizumab, laryngeal cancer in 1 patient and breast cancer in 1 patient receiving 50 mg, extranodal marginal lymphoma in 1 patient receiving 200 mg, and lung adenocarcinoma in 1 patient and basal cell carcinoma in 2 patients receiving 500 mg.
One death occurred during the study. The patient, who was in the group receiving 1,000 mg, died in a motor vehicle accident on study day 166.
Serum samples were assessed for the presence of human anti-human antibodies before the first infusion and at weeks 2, 4, 12, 24, and 36. The incidence of patients with detectable human anti-human antibody was inversely related to the ocrelizumab dose: 19% receiving 10 mg (7 of 36), 10% receiving 50 mg (4 of 40), 0% receiving 200 mg and 500 mg (0 of 40 in each dosage group), and 5% receiving 1,000 mg (2 of 40). No relationship was observed between the presence of human anti-human antibodies and safety events or efficacy responses.
B cell depletion was observed immediately after infusion in all ocrelizumab dosage groups (Figures 2A and B). The median CD19+ B cell counts remained close to baseline levels in the placebo-treated patients. The median B cell counts in the group receiving 10 mg and, to a lesser extent, in the group receiving 50 mg fell to similarly low levels as in the higher-dose groups immediately after treatment, but a trend toward B cell recovery was observed by week 24 and continued through week 72.
To further assess the effect of ocrelizumab on circulating B cells, a lower limit of normal was determined for this study, based on all observed pretreatment CD19+ B cell data. A depletion threshold of 40 cells/μl was determined based on the observation that 95% of the patients (225 of 237) had a CD19+ B cell count that was greater than or equal to this level at all available pretreatment visits. The proportion of patients with CD19+ B cell counts ≥40 cells/μl was plotted versus the study visit for each treatment group (Figure 2B). This proportion was close to 0% in all ocrelizumab-treated groups at week 4 but increased to 44% at week 12 in the group receiving 10 mg, while the proportion in the remaining ocrelizumab-treated groups remained between 0% and 2.5% at week 12. This rate increased in all groups by week 24 (11–61%) and continued to increase through week 72 (55–88%). While the B cell repletion curves for the groups receiving 10 mg and 50 mg suggest some dose-dependence, minimal differences were observed among the groups receiving 200 mg, 500 mg, and 1,000 mg. It should be noted that 7 of the study patients received rituximab during the 72-week study period, and all post-rituximab B cell data for these patients were excluded from Figures 2A and B.
The median counts of CD3+, CD4+, and CD8+ T cells, as well as CD56+ (natural killer) cells, remained similar to baseline counts at weeks 24 and 72 (data not shown).
A higher proportion of patients in all of the ocrelizumab groups achieved an ACR20, ACR50, or ACR70 response at week 24 as compared with patients in the placebo group (Figure 3A). ACR20 response rates were 42% in those receiving 10 mg, 35% in those receiving 50 mg, 45% in those receiving 200 mg, and 50% in each of the groups receiving 500 mg and 1,000 mg. The ACR20 response rate in the placebo group was 22%. Ocrelizumab activity was also observed in terms of more difficult-to-achieve end points, such as remission according to the DAS28 (score <2.6), a swollen joint count of zero, and a good response according to the EULAR criteria (Figure 3B). The best response rates in these categories were observed in the groups receiving 200 mg and 1,000 mg.
To assess the effect of ocrelizumab on the CRP level, analyses excluded data from patients who received excluded medications before week 24. At week 24, the median percentage reduction in the CRP level ranged from 31% to 38% at ocrelizumab doses of 200 mg and higher, as compared with 16% in the placebo group and 11–15% in the groups receiving 10 mg and 50 mg of ocrelizumab (Figure 3C).
At week 24, the median reduction in IgM levels from baseline was 10–23% among the 5 ocrelizumab groups, compared with 2% in the placebo group. Similar differences were observed at week 72, with median reductions in IgM levels ranging from 15% to 28% in the ocrelizumab groups and an 8% reduction in the placebo group. Among patients receiving ocrelizumab, 5.1% (10 of 196) demonstrated a drop in IgM levels below the lower limit of normal (0.4 mg/ml) at some time during the study, whereas none of those receiving placebo showed a drop in IgM levels. The lowest IgM level recorded was 0.22 mg/ml on study day 363 in a patient receiving 500 mg of ocrelizumab. Patients were analyzed for the incidence of infections at the time the low IgM levels were noted, and no infections were reported at the time of the low IgM levels.
Minimal differences were observed among the ocrelizumab groups and the placebo group in terms of the median change in both the IgG levels and the IgA levels.
Despite recent advances in the treatment of RA, a significant unmet need remains for therapeutics that can safely provide a larger overall clinical benefit or a different mechanism of action. Rituximab has been established as an effective treatment option in RA that is refractory to other therapies and is currently approved for use in patients with an inadequate response to TNF inhibitors. Given the increased recognition of the involvement of B cells in the initiation and perpetuation of RA, optimization of B cell depletion by newer-generation molecules that could be used at lower dosages and be associated with improvement in either safety, efficacy, or durability parameters is desirable. Ocrelizumab is a novel humanized monoclonal anti-CD20 antibody with modifications of the Fc region that lead to enhanced ADCC and reduced CDC activities. The clinical relevance of ADCC and CDC activities is unclear, but it has been postulated that relative activities of ADCC and CDC may have an impact on infusion-related safety (18).
In this trial, a novel study design was applied that allowed a rapid transition from phase I to phase II and assessed a wide dosing range for both safety and efficacy. A single course of ocrelizumab (2 infusions on days 1 and 15) at doses ranging from 10 mg to 1,000 mg was investigated in humans for the first time. Clinical response data at 24 weeks and safety data at 72 weeks of followup are reported herein.
The safety profile of ocrelizumab in this study was consistent across dosing groups and suggested only slight differences compared with placebo. No increased rates of infections were seen as compared with placebo throughout the trial. All serious infections that were observed resolved without sequelae. There was a trend toward a reduction in the median levels of IgM; however, no association between infections and reduced IgM levels was observed. It is unclear whether the reductions in IgM levels were caused by a decrease in IgM rheumatoid factor levels or whether they represented a more general reduction of IgM levels. In terms of IgG and IgA levels, minimal differences were observed between the placebo-treated and ocrelizumab-treated groups following a single treatment course.
To best characterize the safety of ocrelizumab around the time of the infusions, all AEs during and within 24 hours of infusion were captured, and no intravenous corticosteroids were administered to the patients before infusion of the study drug. No infusion-associated SAEs were observed in any of the ocrelizumab-treated patients during or within 24 hours of infusion. Nonserious infusion-associated AEs were observed predominantly around the first infusion and remained mostly mild to moderate in severity. The incidence of infusion-associated AEs was similar in all treatment groups including placebo following the second infusion. Interestingly, the incidence of infusion-associated AEs was similar for all doses of ocrelizumab. Together with the observation that most peripheral B cells were rapidly depleted at all dose levels of ocrelizumab, this may suggest that the observed set of infusion-related AEs was mostly caused by effects of B cell lysis.
One key objective of this trial was to characterize B cell depletion and repletion across a wide range of doses after a single course of study drug and without subsequent retreatment. Peripheral B cell depletion was observed at all doses immediately after the initial infusion. There was a dose-dependent trend toward B cell repletion over time in the groups receiving 10 mg and 50 mg, but no obvious difference was seen among the 200, 500, and 1,000 mg groups through week 72. The earlier return of peripheral B cells in the low-dose groups may indicate that tissue B cell depletion was not as profound as with doses of 200 mg and higher. The similarity among peripheral B cell repletion profiles at doses of 200–1,000 mg suggests the possibility that a relative maximum of B cell depletion of this molecule was reached in the 200-mg group.
The broad dosing range also allowed us to address for the first time the important question of whether short-term B cell depletion provided the same clinical benefit at week 24 as more profound, longer-lasting depletion. Interestingly, there was clinical activity at 24 weeks in all ocrelizumab groups as compared with placebo. However, doses of 200 mg and higher appeared to provide greater clinical benefit, as documented by more robust efficacy results with regard to other clinically meaningful end points, such as the ACR70 response, remission according to the DAS28, a swollen joint count of zero, and good responses according to the EULAR criteria. Furthermore, greater improvement in median CRP levels was observed, which might have important implications for systemic inflammation and the prevention of erosive joint damage (19). It is notable that ∼50% of the patients had previously had an inadequate response to TNF inhibitors and that, in general, patients had high numbers of swollen and tender joints at baseline. The demonstration of clinical activity in this difficult-to-treat patient population suggests this is a viable therapeutic option even in patients with severe disease that is refractory to treatment.
The incidence of human anti-human antibodies was highest in the groups receiving 10 mg and 50 mg of ocrelizumab. Although there was no obvious relationship between the development of human anti-human antibodies and subsequent safety issues or lack of efficacy in this trial, the issue of antibody formation remains important. Additional experience with multiple retreatments will be required to further understand the clinical significance of human anti-human antibodies and potential advantages of this therapeutic approach over treatment with chimeric antibodies.
Overall, a single course of ocrelizumab with concomitant MTX therapy appeared to be safe over a 72-week period of followup. The wide range of doses we studied suggests that B cell depletion for a longer duration has clinical advantages over shorter-term depletion, possibly due to more effective tissue B cell depletion. In this trial, doses of 200 mg and higher provided the best clinical responses, the best reduction of CRP levels, and minimal immunogenicity.
Dr. Genovese had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Genovese, Dummer.
Acquisition of data. Genovese, Kaine, Lowenstein, Del Giudice, Baldassare, Schechtman, Fudman, Kohen, Trapp, Sweiss, Dummer.
Analysis and interpretation of data. Genovese, Kaine, Lowenstein, Del Giudice, Baldassare, Schechtman, Fudman, Kohen, Gujrathi, Trapp, Sweiss, Spaniolo, Dummer.
Manuscript preparation. Genovese, Baldassare, Schechtman, Fudman, Spaniolo, Dummer.
Statistical analysis. Spaniolo, Dummer.
Genentech, Inc. facilitated the study design, provided writing assistance for the manuscript, and reviewed and approved the manuscript prior to submission. The authors independently collected the data, interpreted the results, and had the final decision to submit the manuscript for publication.
We wish to thank the patients and the principal investigators of the ACTION study group and to acknowledge all of the enrollment contributions by investigators of the ACTION study group, who were as follows: Drs. E. Arthur, A. Baldassare, E. Boling, J. Booth, P. Caldron, D. Carfagno, V. Chindalore, C. Codding, J. Del Giudice, A. Dikranian, J. Forstot, P. Freeman, E. Fudman, N. Gaylis, M. Genovese, D. Halter, M. Heick, A. Kaell, J. Kaine, A. Kivitz, M. Kohen, M. Layton, J. Loveless, M. Lowenstein, M. Luggen, D. Mandel, A. Martin, H. McIlwain, M. Miniter, A. Nussbaum, M. Pearson, S. Reddy, B. Rubin, C. Saadeh, M. Sayers, J. Schechtman, A. Sebba, Y. Sherrer, E. Siegel, M. Stack, T. Swartz, N. Sweiss, J. Tesser, R. Tierney, E. Tindall, R. Trapp, D. Wallace, and S. Wolfe.