Emerging Treatments and Technologies ORIGINAL ARTICLE

Failure to Preserve ␤-Cell Function With Mycophenolate Mofetil and Combined Therapy in Patients With New- Onset

1 6 PETER A. GOTTLIEB, MD DESMOND A. SCHATZ, MD tions Trial (DCCT) demonstrated that im- 2 7 SCOTT QUINLAN, MS ANTOINETTE M. MORAN, MD proved metabolic control reduces chronic 2 2 HEIDI KRAUSE-STEINRAUF, MS JOHN M. LACHIN, SCD 3 8 complications in type 1 diabetes (5). A CARLA J. GREENBAUM, MD JAY S. SKYLER, MD 4 post hoc analysis of DCCT found that DARRELL M. WILSON, MD FOR THE TYPE 1DIABETES TRIALNET 5 those with residual ␤-cell function, man- HENRY RODRIGUEZ, MD MMF/DZB STUDY GROUP* ifested by C-peptide values Ͼ0.2 pmol/ ml, had both less hypoglycemia and fewer complications than those without resid- OBJECTIVE — This trial tested whether mycophenolate mofetil (MMF) alone or with dacli- ␤ ual function (6). Thus an intervention zumab (DZB) could arrest the loss of insulin-producing -cells in subjects with new-onset type that prolongs ␤-cell function would be 1 diabetes. expected to improve metabolic control RESEARCH DESIGN AND METHODS — A multi-center, randomized, placebo- and reduce complications (7). controlled, double-masked trial was initiated by Type 1 Diabetes TrialNet at 13 sites in North (MPA) was dis- America and Europe. Subjects diagnosed with type 1 diabetes and with sufficient C-peptide covered in 1896 and characterized in within 3 months of diagnosis were randomized to either MMF alone, MMF plus DZB, or placebo, 1952. Mycophenolate mofetil (MMF) is and then followed for 2 years. The primary outcome was the geometric mean area under the rapidly absorbed after oral administration curve (AUC) C-peptide from the 2-h mixed meal tolerance test. and hydrolyzed to MPA (8). MPA is a po- tent, selective, noncompetitive, reversible RESULTS — One hundred and twenty-six subjects were randomized and treated during the trial. The geometric mean C-peptide AUC at 2 years was unaffected by MMF alone or MMF plus inhibitor of inosine monophosphate de- DZB versus placebo. Adverse events were more frequent in the active therapy groups relative to hydrogenase that inhibits de novo the control group, but not significantly. guanosine nucleotide synthesis without incorporation into DNA. T- and B- CONCLUSIONS — Neither MMF alone nor MMF in combination with DZB had an effect on lymphocytes depend on de novo synthe- the loss of C-peptide in subjects with new-onset type 1 diabetes. Higher doses or more targeted sis of purines for their proliferation, while may be needed to affect the autoimmune process. other cell types can use salvage pathways. Thus, MMF has potent cytostatic effects Diabetes Care 33:826–832, 2010 on lymphocytes. MMF is effective in au- toimmune diseases (psoriasis and uveitis) ype 1 diabetes is a chronic, slowly toid arthritis, systemic erythemato- (9,10), as anti-rejection therapy in trans- progressive autoimmune disease sus, and . Infusion of an plant recipients (11), and in diabetic ani- T (1). aimed at mod- anti-CD3 monoclonal showed mal models (12,13). ifying the course of disease has been dem- preservation of ␤-cell function in type 1 Daclizumab (DZB) is a humanized onstrated to be successful in a number of diabetes (2–4). that binds to CD25, immune conditions including rheuma- The Diabetes Control and Complica- the ␣ subunit of the -2 (IL-2) ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● receptor expressed on the surface of acti- vated lymphocytes. DZB inhibits IL-2 From the 1Barbara Davis Center for Childhood Diabetes, University of Colorado at Denver, Aurora, Colo- rado; 2The Biostatistics Center, George Washington University, Rockville, Maryland; the 3Benaroya Re- binding and the progression of T- search Institute, Diabetes Clinical Research, Seattle, Washington; the 4Pediatric Endocrinology lymphocytes through the cell cycle. The Department, Stanford University, Stanford, California; the 5Department of Pediatrics, Indiana University, Edmonton protocol used DZB induction Indianapolis, Indiana; the 6Department of Pediatrics, University of Florida, Gainesville, Florida; the therapy in islet transplantation in type 1 7Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota; and the 8Diabetes Research Institute, University of Miami, Miami, Florida. diabetes (14). It has been used in several Corresponding author: Peter A. Gottlieb, [email protected]. autoimmune conditions (multiple sclero- Received 22 July 2009 and accepted 6 January 2010. Published ahead of print at http://care.diabetesjournals. sis and uveitis) (15,16). Recent work in org on 12 January 2010. DOI: 10.2337/dc09-1349. Clinical trial reg. no. NCT00100178, clinicaltrials. the DR-BB rat model demonstrated a syn- gov. ergistic effect of these two drugs when *A complete list of the members of the MMF/DZB study group can be found in the online appendix, available at http://care.diabetesjournals.org/cgi/content/full/dc09-1349/DC1. The members of the TrialNet Study used together (17). Group can be found at www.diabetestrialnet.org. The objective of this study was to de- © 2010 by the American Diabetes Association. Readers may use this article as long as the work is properly termine whether MMF alone or MMF cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. combined with DZB could diminish pro- org/licenses/by-nc-nd/3.0/ for details. ␤ The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby gression of -cell destruction in recent- marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. onset type 1 diabetes.

826 DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 care.diabetesjournals.org Gottlieb and Associates

RESEARCH DESIGN AND virus (EBV) was based on close surveil- Statistics METHODS — This multi-center trial lance rather than active prophylaxis. The prespecified primary analyses were was conducted at 13 sites with subjects Study visits were conducted to assess based on the intention-to-treat cohort aged 8–45 years with autoimmune type 1 safety weekly ϫ 4, biweekly ϫ 2. There- that included all subjects randomized diabetes for less than 3 months and with after, EBV-negative subjects were fol- correctly to the three specified treat- evidence of ␤-cell function evidenced by lowed monthly, and EBV-positive ment groups. The primary outcome was stimulated C-peptide Ͼ0.2 pmol on a 2-h subjects were followed at 3-month inter- the geometric mean difference between mixed meal tolerance test (MMTT). Auto- vals. Visits included assessment of diabe- active- and placebo-treated subjects of immune type 1 diabetes was defined by tes care, adverse events, and laboratory the area under the stimulated C-peptide the presence of any of four islet autoanti- measurements to assess side curve over the first2hofa4-hMMTT bodies within 14 days of diagnosis (GAD, effects. In the case of an acute infection, conducted at the 2-year visit in an insulinoma-associated protein 2, or islet additional studies were performed. ANCOVA model adjusting for the baseline cell autoantibodies [ICAs]). Subjects were All participants received intensive di- C-peptide, age, and sex. The 2-h C- otherwise healthy without major systemic abetes management with the goal of main- peptide area under the curve (AUC) illness nor allergic or autoimmune condi- taining A1C levels Յ7.0%. (pmol/ml/120 min) was computed us- tions requiring treatment with immuno- An independent DSMB met every 6 ing the trapezoidal rule from timed suppressive agents or steroids. The months and had quarterly summary measurements of C-peptide during each protocol was approved by the Type 1 Di- safety reviews. A medical monitor, MMTT (including the basal). The AUC abetes TrialNet Steering Committee, the masked to treatment assignment, re- mean (pmol/ml) equals the AUC di- Data and Safety Monitoring Board vided by the interval of time. The log- viewed all adverse events. An infectious ϩ (DSMB), and regulatory authorities; hu- disease committee developed treatment ([mean C-peptide] 1) transformation man subject approval was obtained at algorithms for common infections and of the baseline and follow-up AUC participating sites prior to study initia- provided consultation as needed. mean was used to allow for mean C- tion. All subjects provided written, in- peptide values close to zero and to nor- formed consent. malize the distribution of the residuals Laboratory assessments (6). Blood samples were analyzed at core lab- Data from all 13 centers contributed Study design oratories. A 4-h MMTT was conducted at to the primary and secondary effective- The study was a three-arm, randomized, ness analyses of the MMF plus DZB com- baseline and 2 years, and a 2-h MMTT at double-masked, placebo-controlled clin- bination and its respective control group. 3, 6, 12, and 18 months with timed sam- ical trial conducted by Type 1 Diabetes However, owing to the randomization er- ple collection at 15–30 min intervals. C- TrialNet. Roche Pharmaceuticals pro- ror, subjects received MMF alone in only peptide levels were measured using a two- vided MMF, DZB, and placebo, but had seven centers. Thus, these analyses com- site immunoenzymometeric assay (Tosoh no involvement in study management, pare the MMF alone subjects only with 600 II analyzer). A1C was measured quar- data collection and analysis, or manu- the concurrently randomized control script preparation. There were 126 sub- terly using ion-exchange high perfor- subjects from these centers. jects randomized to receive MMF alone mance liquid chromatography (Variant II, Secondary analyses include assess- Bio-Rad Diagnostics). The reliability coef- (with DZB placebo), MMF and DZB in Ͼ ment of differences between groups over combination, or control (MMF placebo ficient for each assay was 0.99 from time in a longitudinal normal errors re- and DZB placebo), stratified within clini- split duplicate samples. peated-measures model of the log([mean cal center. Biochemical autoantibodies (GAD- C-peptide] ϩ 1) values. The group geo- By error, among the last six sites to 65, ICA-512, mIAA) were measured us- metric mean C–peptide was obtained us- join the study, 12 subjects assigned to re- ing radio-immunobinding assays; ICAs ing the inverse transformation. The mean ceive MMF alone inadvertently received were measured using indirect immuno- rate of change over 3–24 months was es- DZB-alone, thus resulting in an imbalance fluorescence. Potential participants were timated using a mixed effects random co- in the group sample sizes. The results screened for to hepatitis B sur- efficient model (18) using the log values. from these 12 subjects are not presented face , hepatitis C, and human im- The Cox proportional hazards model as- herein. munodeficiency virus using enzyme sessed the relative risk (hazard ratio) of MMF or matched placebo was admin- immunoassays that, if positive, resulted in the loss of the 2-h C-peptide Ͻ0.2 istered daily at a dose of 600 mg/m2 (max- exclusion from the study. pmol/ml (19). imum 2,000 mg/day) in 2–3 divided Antibodies to CMV EBV were mea- Prespecified secondary outcomes also doses for 2 years. DZB or matched pla- sured using indirect immunofluorescence include: differences in A1C, insulin dose, cebo was given by intravenous infusion at (anti-EBV VCA IgM) and enzyme immu- hypoglycemic episodes, rates of infection, study day 0 and two weeks later at a dose noassay (anti-CMV IgG and IgM; anti- and adverse events over time. of 1 mg/kg. All subjects were to be fol- EBV VCA and EBNA IgG). CMV and EBV For assessment of safety, the two ac- lowed for at least 2 years under the inten- viral load was measured using real time tive groups are compared with the total tion-to-treat principle, including those quantitative PCR (Lightcycler System; control group enrolled. The percents of who did not receive the full course of as- Roche Applied Science). MMF peak and subjects with an event were compared signed therapy. trough levels were determined by MPA among the three groups using Fisher ex- Because both drugs reduce the ability concentrations using HPLC. Flow cytom- act test. The rate of events per subject was to fight viral infections, screening for cy- etry was used to measured T-lymphocyte compared between groups using the Pois- tomegalovirus (CMV), and Epstein-Barr subpopulations, including CD4CD25. son model test (20). care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 827 MMF and DZB in new-onset type 1 diabetes

Table 1—Baseline characteristics of the treatment groups, including all 114 subjects correctly randomized

MMF ϩ DZB MMF alone Active Control Active Control n 41 42 31 28 Age (years) 18.3 Ϯ 9.1 18.8 Ϯ 10.5 17.1 Ϯ 6.7 15.8 Ϯ 8.0 Race (% white) 38 (93) 39 (93) 30 (97) 27 (96) Non-Hispanic (%) 40 (98) 39 (93) 29 (94) 26 (93) Number of Abϩ (%) 1 1 (2) 3 (7) 4 (13) 1 (4) 2 11 (27) 8 (19) 7 (23) 7 (25) 3 12 (29) 16 (38) 6 (19) 11 (39) 4 17 (41) 15 (36) 14 (45) 9 (32) Male sex (%) 23 (56) 25 (60) 20 (65) 16 (57) 2-h C-peptide AUC means (pmol/ml) 0.71 Ϯ 0.36 0.71 Ϯ 0.34 0.65 Ϯ 0.28 0.73 Ϯ 0.36 Baseline A1C (%) 7.5 Ϯ 1.3 7.7 Ϯ 1.6 7.4 Ϯ 1.0 7.5 Ϯ 1.5 Baseline total insulin dose/kg 0.40 Ϯ 0.26 0.36 Ϯ 0.20 0.35 Ϯ 0.15 0.39 Ϯ 0.22 Weight (kg) 58.9 Ϯ 16.6 59.0 Ϯ 16.3 61.3 Ϯ 18.3 57.2 Ϯ 16.7 Height (cm) 162.4 Ϯ 13.5 162.7 Ϯ 13.7 164.4 Ϯ 15.4 160.9 Ϯ 13.4 BMI (kg/m2) 22.0 Ϯ 4.2 21.8 Ϯ 3.6 22.1 Ϯ 4.1 21.6 Ϯ 4.0 zBMI (only on subjects Ͻ20 years) 0.44 Ϯ 1.11 0.66 Ϯ 0.79 0.42 Ϯ 1.02 0.57 Ϯ 0.74 n 30 28 21 23 Mean A1C over 24 months (%) 7.2 Ϯ 1.2 7.2 Ϯ 1.0 7.0 Ϯ 1.2 7.3 Ϯ 0.9 Mean insulin dose/kg over 24 months 0.56 Ϯ 0.29 0.55 Ϯ 0.32 0.59 Ϯ 0.31 0.63 Ϯ 0.34 Mean MPA level over 24 months (mcg/ml)* 4.5 Ϯ 3.4 0.6 Ϯ 0.6 5.8 Ϯ 4.0 0.6 Ϯ 0.2 Received 2 full DZB infusions (%) 40 (98)** 42 (100) 31 (100) 28 (100) % of subjects MMF compliant*** 36 (88) 41 (98) 27 (87) 27 (96) Means Ϯ SD are presented for continuous variables. *Limit of quantitation ϭ 0.5 units. **One subject did not receive the second infusion due to patient decision to continue study treatment. ***80% or greater by capsule count up through last recorded visit starting with month 3.

The target sample size of 120 subjects the conditional power of each compari- years, 31% ages 13–17 years, and 39% (40 per group) provided 85% power to son under the current trend in the data age Ն18 years. Mean time from diagnosis detect a 65% difference in the geometric were less than 0.02% and under the orig- to enrollment was 76 days. mean C-peptide for any one of the three inal design assumptions, termination for Safety data are presented on all 114 possible pairwise comparisons among the futility led to less than a 1% increase in the properly randomized subjects, excluding three treatment groups using a test at probability of a type II error (21). Sites the 12 who received the DZB-alone in- the 0.05 level (one-sided, adjusted for three were notified on April 30, 2008, to imme- stead of MMF alone. Of these, 60 com- comparisons), with 10% loss to follow- diately terminate treatment but continue pleted an MMTT at 2 years and up. Owing to the randomization error, to follow all subjects. This report is based contributed to the primary outcome the protocol was modified to compare on closed and locked data for all visits analysis. the MMF and DZB combination group through April 30, 2008. Nominal one- All randomized subjects received versus all placebo subjects, and to com- sided P values (without adjustment for study treatment, and all but one received pare the MMF only subjects versus the multiple tests) are presented for analyses the two planned DZB/placebo infusions. placebo subjects enrolled within the of primary and secondary effectiveness Median compliance with MMF daily cap- same clinical centers (7 of the original outcomes; 2-sided P values for safety sules was estimated to be 75% in the MMF 13) in which the randomization was not outcomes. plus DZB group, 63% in the MMF alone affected, each using a test at the 0.025 level (one-sided, adjusted for two com- RESULTS group, and 71% in the control group over parisons). The MMF plus DZB versus the treatment period based on capsule Subjects counts. Mean MPA trough levels over 24 placebo comparison, with about 40 per Ϯ group, provided 85% power to detect a Supplemental Figure A1 (available in months were 4.5 3.4 (mcg/ml) and Ϯ 61% difference; and the MMF alone ver- the online appendix at http://care. 5.8 4.0 (mcg/ml), respectively, for sus placebo comparison, with about 30 diabetesjournals.org/cgi/content/full/dc09- MMF plus DZB and MMF alone groups per group, provided 80% power to de- 1349) summarizes subject disposition- with expected trough range of 1.0–3.5 tect a 67% increase, each allowing for screening of 228 subjects, randomization ␮g/ml. Treatment was terminated in 23 10% losses to follow-up. of 126 subjects, and subsequent disposi- subjects due to adverse events (5), ele- In April 2008, based on 41 and 47% tion. Table 1 presents baseline character- vated liver enzymes (2), EBV PCR- of the planned total information, the istics for each active therapy group versus positivity (8), treatment noncompliance DSMB recommended termination of the its respective control group. The groups (3), or loss to follow-up (5) (Supplemen- treatment phase of the study. At that time were well-matched with 30% ages 8–12 tal Figure A1).

828 DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 care.diabetesjournals.org Gottlieb and Associates

Figure 1—Effect of MMF and MMF plus DZB on C-peptide over 2 years. A) The geometric means and 95% CIs for the 2-h AUC stimulated C-peptide levels over time within each group. B) The cumulative incidence of decline in peak C-peptide to Ͻ0.2 pmol/ml within each group. The relative hazard was 0.61 (95% CI: 0.28–1.33, P ϭ 0.11) for MMF plus DZB vs. control, and 1.05 (0.50–2.19, P ϭ 0.83) for MMF alone vs. control. C) Ratio of geometric means for MMF plus DZB vs. control groups, with 95% CIs, within subgroups of subjects defined at baseline. D) Likewise for MMF alone vs. control (A1C 2nd tertile upper 95% confidence limit is 28.9).

C-peptide treated subjects, compared with 0.23 of peak C-peptide below 0.2 pmol/ml did Mean AUC C-peptide at entry was 0.70 Ϯ (0.12–0.35) for their control subjects, not differ between groups (Fig. 1B). 0.33 pmol/ml. Control subjects lost C- P ϭ 0.41. There was no statistical differ- In the primary analysis, the geometric peptide at a rate of 53.5% per year, and ence between treatment and control sub- mean ratio for MMF plus DZB versus con- both the MMF alone and the MMF plus jects over 2 years or during the early phase trol subjects was 1.02 (95% CI 0.65– DZB treatment groups had comparable when DZB would have been more active. 1.59) and for MMF alone versus control rates of loss, 46.4% and 48.1%, respec- Results were similar for 4-h AUC mean subjects was 1.08 (0.57–2.02). Figure 1C tively. In the primary analysis (Fig. 1A)at C-peptide at 2 years. and D show that these mean ratios and 2 years,;the geometric mean stimulated During follow-up, all but eight sub- confidence limits within subgroups, de- C-peptide AUC was 0.28 pmol/ml (95% jects had detectable levels of C-peptide. fined by baseline characteristics, are not CI 0.19–0.37) in those treated with MMF The AUC mean C-peptide fell below 0.2 nominally significantly different from 1, plus DZB, compared with 0.27 (0.18– pmol/ml during follow-up in 12 MMF with the exception of the effect of MMF 0.37) for their control subjects, P ϭ 0.47; plus DZB, 16 MMF alone, and 17 control alone within the 10 subjects in the highest and 0.25 (0.14–0.37) in MMF alone subjects. Cumulative incidence of decline tertile of baseline A1C (P ϭ 0.042). How- care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 829 MMF and DZB in new-onset type 1 diabetes

Figure 2—Effect of MMF and MMF plus DZB on glycemic control over time. A) Mean A1C (%) and 95% confidence limits over time. B) Mean insulin dose and 95% confidence limits over time within each group. ever, comparison of ratios among the A1C More grade 2 or higher AEs occurred effector cells, which can damage and kill tertiles, and among all other subgroups, in MMF plus DZB subjects (167 or 4.1 islets without requiring cell division, the failed to reach significance demonstrating events/subject) compared with MMF primary mode of MMF action. Although that variation among subgroups was alone (117, 3.8 events/subject) or control several studies have reported a potential within the realm of chance. Similar results subjects (133, 3.2 events/subject), P ϭ negative effect of MMF on islet cell func- applied to subgroups defined from the 0.09 (online appendix Table 1). Contrary tion, we did not see any greater loss of mean levels of A1C and insulin dose over to expectations, there was no difference in C-peptide in the MMF alone group com- 24 months. the occurrence of infectious or gastroin- pared with control subjects. For both DZB reduced CD4CD25 T-cell levels testinal events among groups. Eight indi- MMF and DZB, we chose the lowest maximally at 4 weeks (depletion 83.9% viduals had asymptomatic reactivation of known effective doses of each. DZB has and blocking 97.5%) and these recovered previous EBV infection using a sensitive been shown to reduce recurrences in mul- within 6–12 months. The month 24 C- PCR assay (five in MMF plus DZB, one in tiple sclerosis (15) and uveitis (16) when peptide level was not associated with ei- MMF, and two in control). Neutropenia given monthly. Our use of two doses of ther percent reduction in CD4CD25 and leukopenia, both side effects of MMF DZB may not have been sufficient to affect T-cells within the MMF plus DZB group, and DZB, occurred approximately equally activated effectors cells in the pancreas or MMF trough levels in either the MMF among the three groups. A slight excess of even with MMF, despite reasonably good plus DZB or MMF alone groups. elevated liver enzymes occurred in the depletion/coating in the peripheral circu- MMF plus DZB group. Major hypoglyce- lation. Cyclosporine, which also affects A1C and insulin dose mic events were reported for 27 subjects, the IL-2 signaling pathway, was shown to All groups achieved A1C of 7.2–7.3% with an average of two each, with no dif- be effective in past trials if given at high throughout the study (Fig. 2A). Mirroring ference among groups. doses and early enough in the course of changes in C-peptide, daily insulin dose disease (22). This, and the aforemen- slowly rose from below 0.5 units/kg at CONCLUSIONS — The aim of the tioned effect of DZB in two other autoim- baseline to 0.57 units/kg with MMF plus present study was to arrest ␤-cell destruc- mune diseases, suggests that the lower DZB versus 0.61 units/kg among control tion in recently diagnosed type 1 diabetic dose may have played a major part in the subjects (P ϭ 0.17); and to 0.65 units/kg subjects when preservation of existing lack of effect of this therapy. Although with MMF alone versus 0.62 units/kg ␤-cells may have a clinically meaningful higher doses may have greater therapeutic among control subjects (P ϭ 0.68) (Fig. effect on long-term outcomes of type 1 effect, this has to be measured against the 2B). diabetes. We found no treatment benefit increased risk of side effects. Even at the from either MMF alone or from the com- doses used in this study, there was an in- Adverse events bination of MMF and DZB in this ran- crease in AEs when the two drugs were There were 19 serious adverse events domized, masked, placebo-controlled used together in comparison to MMF (AEs) reported for 14 subjects (34%) in trial. alone or placebo. the MMF and DZB group, nine in five sub- Although MMF has been effective in CD4CD25ϩ regulatory T-cells play jects (16%) in the MMF alone group, and combination with other anti-rejection an important role in immune regulation three in three subjects (7%) in the control drugs (such as and ) and a potential problem with an anti-IL-2 group (online appendix Table 1), P Ͻ in a number of transplant protocols (11), receptor antibody is worsening autoim- 0.01. alone it may not have as much effect on munity rather than reducing it if the ef-

830 DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 care.diabetesjournals.org Gottlieb and Associates fects of the drug on the regulatory cell well as our rigorous screening program to ameliorate, and cure type 1 diabetes. Am J population outweigh its effects on the ac- ensure patient safety. Ther 2005;12:481–490 tivated-effector cell population. In this Type 1 Diabetes TrialNet is an Na- 2. Herold KC, Hagopian W, Auger JA, study we saw no worsening of ␤-cell de- tional Institutes of Health (NIH)- Poumian-Ruiz E, Taylor L, Donaldson D, struction or development of other auto- sponsored multicenter trial group formed Gitelman SE, Harlan DM, Xu D, Zivin RA, Bluestone JA. Anti-CD3 monoclonal anti- immune conditions with the use of DZB. to perform intervention trials in new- body in new-onset type 1 diabetes melli- While overall compliance with the onset type 1 diabetes and pre-diabetes, as tus. N Engl J Med 2002;346:1692–1698 MMF and DZB regimens was high, it is well as to develop immunologic and 3. Keymeulen B, Vandemeulebroucke E, possible that the need to withdraw study mechanistic assays to better understand Ziegler AG, Mathieu C, Kaufman L, Hale G, drug, primarily MMF, for various inter- type 1 diabetes pathogenesis. There are Gorus F, Goldman M, Walter M, Candon S, vals due to AEs, may have affected the several advantages of multicenter net- Schandene L, Crenier L, De Block C, Sei- ability to demonstrate a beneficial effect. works. These include consistency in gneurin JM, De Pauw P, Pierard D, Weets I, However, mechanistic assessments study design and study outcomes allow- Rebello P, Bird P, Berrie E, Frewin M, Wald- showed that MMF and DZB each were ing better comparisons between trials. mann H, Bach JF, Pipeleers D, Chatenoud L. bioavailable and had the intended immu- Proposed studies are rigorously reviewed Insulin needs after CD3-antibody therapy in new-onset type 1 diabetes. N Engl J Med nologic effects, but these effects were not for scientific and ethical justification, clin- 2005;352:2598–2608 associated with the C-peptide levels after ical feasibility and prioritization by a di- 4. Herold KC, Gitelman S, Greenbaum C, 2 years. Modified anti-CD3 antibodies verse group of clinicians, basic scientists, Puck J, Hagopian W, Gottlieb P, Sayre P, have been shown to reduce the rate of loss statisticians, and ethicists (25). Trials are Bianchine P, Wong E, Seyfert-Margolis V, of C-peptide in new-onset type 1 diabetic monitored by metabolic, infectious dis- Bourcier K, Bluestone JA, Immune Toler- subjects similar to those studied in this ease, and safety monitoring committees in ance Network ITN007AI Study Group. trial (2–4). MMF and anti-IL-2R are addition to oversight by an independent Treatment of patients with new onset type downstream of the important major his- DSMB. 1 diabetes with a single course of anti- tocompatibility complex-peptide/T-cell– Although this trial was unsuccessful CD3 mAb Teplizumab preserves insulin receptor interaction, which is the driver of at finding new therapies to induce clinical production for up to 5 years. Clin Immu- the autoimmune response. Therapies remission in type 1 diabetes, it showed nol 2009;132:166–173 5. DCCT Research Group. The effect of in- such as anti-CD3 and anti-CD20 as well that our network can successfully design, tensive treatment of diabetes on the devel- as such as GAD, oral insulin, and recruit, and conduct clinical trials of suf- opment and progression of long-term DiaPep277 may have the potential to alter ficient size. The use of novel agents, alone complications in insulin-dependent dia- this critical reaction and blunt the direct or in combination, will be facilitated by betes mellitus. N Engl J Med 1993;329: activation of autoreactive T-cells rather the clinical trial process developed for this 977–986 than limit their activity and division, first trial under the TrialNet mechanism. 6. Palmer JP, Fleming GA, Greenbaum CJ, which is where MMF and DZB are most Herold KC, Jansa LD, Kolb H, Lachin JM, critical. One additional difference be- Polonsky KS, Pozzilli P, Skyler JS, Steffes tween this study and the cyclosporine tri- Acknowledgments— This study was con- MW. C-peptide is the appropriate out- als (22) that had been successful was that ducted by the Type 1 Diabetes TrialNet Study come measure for type 1 diabetes clinical Group, a clinical trials network funded by NIH trials to preserve beta-cell function: report the time to treatment was 56 days with through the National Institute of Diabetes and of an ADA workshop, 21–22 October anti-CD3 versus 76 days in this study. Digestive and Kidney Diseases, the National 2001. Diabetes 2004;53:250–264 Post hoc analysis did not reveal this to be Institute of Allergy and Infectious Diseases, the 7. Sjo¨berg S, Gjo¨tterberg M, Berglund L, a factor in the failure to see an effect and National Institute of Child Health and Human Mo¨ller E, Ostman J. Residual C-peptide ex- probably suggests that more targeted Development, and the General Clinical Re- cretion is associated with a better long-term therapy at sufficient dose is necessary to search Centers Program with support of the glycemic control and slower progress of ret- arrest the diabetes process. New thera- Juvenile Diabetes Research Foundation In- inopathy in type I (insulin-dependent) dia- pies such as DiaPep277 in adults (23) or ternational and the American Diabetes betes mellitus. J Diabet Complications GAD immunization (24) have recently Association. 1991;5:18–22 been shown to slow the rate of loss of P.A.G. reports receiving grants from Bayhill 8. Brazelton TR, Morris RE. Molecular Therapeutics, Macrogenics, and Tolerx; D.W. mechanisms of action of new xenobiotic C-peptide, and others are under study reports serving on an advisory board for Ge- immunosuppressive drugs: tacrolimus such as anti-CD20, , and nentech; H.R. reports serving on an advisory (FK506), sirolimus (rapamycin), myco- thymoglobulin. board for Genentech/Roche and receiving a phenolate mofetil and leflunomide. Curr Although we were concerned with grant from Macrogenics/Eli Lilly; D.A.S. re- Opin Immunol 1996;8:710–720 the number of adverse events that might ports serving on advisory boards for Genen- 9. Epinette WW, Parker CM, Jones EL, Gre- occur from the use of immunosuppres- tech and Roche; J.M.L. reports receiving ist MC. Mycophenolic acid for psoriasis: a sive agents, it is clear from our analysis consulting fees from Tolerx, Bayhill Therapeu- review of pharmacology, long-term effi- that the number and type we detected tics, and Andromeda Biotech; J.S.S. reports re- cacy, and safety. J Am Acad Dermatol were for the most part within our pre- ceiving grants from Bayhill Therapeutics and 1987;17:962–971 study expectations and did not prevent Osiris Therapeutics. No other potential con- 10. Larkin G, Lightman S. Mycophenolate flicts of interest relevant to this article were study subjects from continuing treat- mofetil: a useful immunosuppressive in reported. inflammatory eye disease. Ophthalmol- ment. The finding of asymptomatic low- ogy 1998;106:370–374 level PCR reactivation of EBV in both 11. Sollinger HW. Mycophenolate mofetil for treated and untreated patients was unex- References the prevention of acute rejection in pri- pected and may reflect the differential 1. Aly T, Devendra D, Eisenbarth GS. Immu- mary cadaveric renal allograft recipients: sensitivity of our EBV viral PCR assay as notherapeutic approaches to prevent, U.S. Renal Transplant Mycophenolate care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 831 MMF and DZB in new-onset type 1 diabetes

Mofetil Study Group. Transplantation Rao NA, See RF, Letko E, Buggage RR. BM. Effects of cyclosporine immuno- 1995;60:225–232 Initial evaluation of subcutaneous dacli- suppression in insulin-dependent dia- 12. Hao L, Calcinaro F, Gill RG, Eugui EM, zumab treatments for noninfectious betes mellitus of recent onset. Science Allison AC, Lafferty KJ. Facilitation of uveitis: a multicenter noncomparative in- 1984;223:1362–1367 specific tolerance induction in adult mice terventional case series. Ophthalmology 23. Raz I, Avron A, Tamir M, Metzger M, by RS-61443. Transplantation 1992;53: 2005;112:764–770 Symer L, Eldor R, Cohen I, Elias D. Treat- 590–595 17. Ugrasbul F, Moore WV, Tong PY, Kover ment of new-onset type 1 diabetes with 13. Hao L, Chan SM, Lafferty KJ. Mycophe- KL. Prevention of diabetes: effect of my- peptide DiaPep277 is safe and associated nolate mofetil can prevent the develop- cophenolate mofetil and anti-CD25 on with preserved beta-cell function: exten- ment of diabetes in BB rats. Ann N Y Acad onset of diabetes in the DRBB rat. Pediatr sion of a randomized, double-blind, Sci 1993;696:328–332 Diabetes 2008;9:596–601 phase II trial. Diabete Metab Res Rev 14. Shapiro AM, Lakey JR, Ryan EA, Korbutt 18. Demidenko E. Mixed Models: Theory and 2007;23:292–298 GS, Toth E, Warnock GL, Kneteman NM, Applications. Hoboken, New Jersey, John 24. Ludvigsson J, Faresjo¨ M, Hjorth M, Axels- Rajotte RV. Islet transplantation in seven pa- Wiley & Sons, 2004 son S, Che´ramy M, Pihl M, Vaarala O, tients with type I diabetes mellitus using a 19. Lachin J. The Assessment of Relative Risks. glucocorticoid-free immunosupressive reg- Hoboken, New Jersey, John Wiley and Forsander G, Ivarsson S, Johansson C, imen. N Engl J Med 2000;343:230–238 Sons, 2000 Lindh A, Nilsson NO, Aman J, Ortqvist E, 15. Bielekova B, Richert N, Howard T, 20. Holm S. A simple sequentially rejective Zerhouni P, Casas R. GAD treatment and Blevins G, Markovic-Plese S, McCartin multiple test procedure. Scand J Statist insulin secretion in recent-onset type 1 J, Frank JA, Wu¨ rfel J, Ohayon J, Wald- 1979;6:65–70 diabetes. N Engl J Med 2008;359:1909– mann TA, McFarland HF, Martin R. 21. Lachin JM. Operating characteristics of 1920 Humanized anti-CD25 (daclizumab) sample size re-estimation with futility stop- 25. Skyler JS, Greenbaum CJ, Lachin JM, Le- inhibits disease activity in multiple scle- ping based on conditional power. Statistics schek E, Rafkin-Mervis L, Savage P, Spain rosis patients failing to respond to inter- in Medicine 2006;25:3348–3365 L, Type 1 Diabetes TrialNet Study Group. feron beta. Proc Natl Acad SciUSA 22. Stiller CR, Dupre´ J, Gent M, Jenner MR, Type 1 Diabetes TrialNet: an international 2004;101:8705–8708 Keown PA, Laupacis A, Martell R, collaborative clinical trials network. Ann 16. Nussenblatt RB, Peterson JS, Foster CS, Rodger NW, von Graffenried B, Wolfe N Y Acad Sci 2008;1150:14–24

832 DIABETES CARE, VOLUME 33, NUMBER 4, APRIL 2010 care.diabetesjournals.org