Diabetes article reports that teplizumab significantly reduces C-peptide decline at two years in people with new-onset type 1 diabetes - August 19, 2013

Executive Highlights

  • In the phase 2 AbATE trial, teplizumab treatment significantly reduced C-peptide decline at two years in people with new-onset type 1 diabetes.
  • Subjects who responded to treatment had significantly lower A1cs and insulin usage at baseline.

In a recently published Diabetes article, the renowned Dr. Kevan Herold and colleagues report that teplizumab significantly (p=0.002) reduced two-year C-peptide decline among people with new-onset type 1 diabetes in the phase 2, open-label AbATE study. This study randomized people 2:1 to either two, two-week courses of teplizumab (at diagnosis and again after one year) or to observation. Patients enrolled in the trial (n=77) were between the ages of 8 and 30 and had been diagnosed with type 1 diabetes within the last eight weeks. Those in the teplizumab arm (n=52) had a mean C-peptide level of 0.72 nmol/l at baseline, versus a mean of 0.67 nmol/l for those in the control arm. At two years, mean C-peptide was 75% higher in the teplizumab vs. control arm.

Notably, the study’s authors were able to discern key differences between those who responded to treatment (defined as losing <40% of C-peptide response from baseline; a smaller C-peptide decline than anybody in the control arm) and those who didn’t. Responders represented 45% of subjects in the teplizumab arm (n=49). Interestingly, responders had significantly (p=0.011) lower A1c levels (mean:7.05%; 95% CI: 6.689%, 7.411%) than non-responders (mean: 7.78%; 95% CI: 7.360, 8.203%), and used less insulin at baseline (0.28 U/kg/d vs. 0.49 U/kg/d; p=0.004) after correcting for baseline C-peptide. These findings point to a possible role for metabolic factors in response to immunotherapy. Unfortunately, no clear “threshold” for either A1c or insulin dose that could be used to a priori identify responders was found. The study’s findings also suggest that subjects with less severe disease or those who are not as far along in the natural history of diabetes may be more likely to respond to teplizumab treatment. If teplizumab is indeed more effective earlier in the disease’s progression, it might be that the best time to utilize the drug will be when a person has dysglycemia. TrialNet is currently enrolling a phase 2 trial of teplizumab in relatives of people with type 1 diabetes who have abnormal glucose tolerance and two or more diabetes autoantibodies ( Identifier: NCT01030861). We believe this study’s results will provide a more complete picture of when people are most likely to respond to, and therefore benefit from, teplizumab (or if there is a period of maximal benefit at all).

Looking forward, we feel a case is gradually building for giving teplizumab a “second chance.” As a reminder, teplizumab’s phase 3 trial Protégé did not meet its primary endpoint - a composite of insulin usage and A1c levels; however, in this trial, teplizumab treatment did preserve C-peptide and reduce the need for insulin compared to control. Dr. Herold and his colleagues seem to now be steadily honing in on who will benefit from teplizumab. Still, the challenge remains to develop data-driven guidelines around who is most likely to respond. We are curious if guidelines could be based on a person’s initial response, as they are for a very different compound, Vivus’ Qsymia, for example. As a reminder, the best predictor for long-term weight loss with Vivus’ Qsymia is initial weight loss. As such, Qsymia’s label indicates that people who don’t respond (lose at least 3% of their baseline weight) at week 12 should discontinue further treatment. We are heartened to see the results and will report back on implications.

  • Dr. Herold et al. randomized people 2:1 to either two courses of teplizumab or an open-label control. Participants were eight to 30 years old, diagnosed with type 1 diabetes within eight weeks, and were positive for anti-GAD65, anti-ICA512, or ICA. Fifty-two people were treated with teplizumab for two weeks (day 1: 51 μg/m2, day 2: 103 μg/m2, day 3: 206 μg/m2, Day 4: 413 μg/m2, days 5-14: 826 μg/m2; median cumulative dose=11.6 mg, interquartile range 5.7 mg) at diagnosis and again after one year. After one year, people in the teplizumab arm whohad detectable C-peptide responses at all time-points of the mixed meal tolerance test and had not met stopping criteria for drug administration received a second course of teplizumab (n=40; median cumulative dose of 12.4 mg; interquartile range 5.08 mg). Due to a temporary hold due to drug manufacturing and safety review, four of the first five participants received their second course of drug later than the one-year time point. The control group did not receive a placebo infusion.
  • At baseline people (age 8-30 years) had been diagnosed with type 1 diabetes within eight weeks and had a mean C-peptide level of 0.72 nmol/l in the teplizumab arm and a level of0.67 nmol/l in the control arm.
  • Seventy-seven people were randomized to a treatment and were included in the intention to treat analysis. Of these participants, 52 received all or part of the first drug course and 15 discontinued teplizumab after receiving some of cycle one or two. Of the 15 who discontinued teplizumab, 12 had developed laboratory abnormalities or experienced adverse events leading to discontinuation.
  • At two years, teplizumab-treated subjects had a mean C-peptide decline of 0.28 nmol/l (95% confidence interval: -0.36 nmol/l, -0.20 nmol/l), versus a 0.46 nmol/l (-0.57 nmol/l, -0.35 nmol/l) decline in the placebo group. People treated with teplizumab were estimated to reach the control group’s 6-month C-peptide values 15.9 months later than the control group did (i.e. at 21.9 months). No significant difference in A1c levels was detected between the two groups at month 24; however, at months nine and 15, the control group had a significantly greater change in A1c from baseline (p=0.035, 0.011, respectively). Notably, teplizumab-treated patients used significantly (p = 0.036) less insulin to achieve the stated level of glycemic control.
  • Herold et al. note that responders (those who lost <40% of C-peptide response from baseline) showed both qualitative and quantitative differences in their response to teplizumab.. To examine whether the pattern of change in C-peptide was qualitatively different in responders or was similar to that of the untreated control group but shifted in time, the study’s authors compared C-peptide AUC over time in responders, non-responders, and control subjects. Eighteen months after the initial teplizumab treatment, the average C-peptide level among responders was 113% of that at baseline, and only decreased an average of 6% by month 24. In contrast, in non-responders, even the initial effect of treatment was modest. Adjusting for baseline differences, C-peptide levels were nearly three-fold higher at month 24 among responders than control participants. 24-month C-peptide levels did not differ between teplizumab non-responders and people in the control group.
  • Factors that did not differ significantly between responders and non-responders included age, sex, BMI, duration of disease, baseline autoantibody titers, and baseline C-peptide responses.
  • On the safety side, 11 serious adverse events (SAEs) occurred in 10 drug-treated patients and two SAEs occurred in one control subject. The most common adverse events were rash, transient upper respiratory infections, headache, and nausea. Of the 21 Epstein-Barr virus (EBV) seropositive subjects, nine had a transient increase in viral load detected one month after the first course of drug treatment, and three had an increase after the second course. Five people with an increase in EBV viral load had possible symptoms of a transient EBV infection; however, in all people viral loads were undetectable at two months after treatment. As a reminder, higher rates of transient EBV reactivation were associated with otelixizumab (another anti-CD3 monoclonal antibody) treatment during phase 2 development. In an attempt to avoid this reactivation during phase 3 development, the oteliziumab dose used was reduced to one- sixteenth that of the original. The phase 3 trial found neither EBV reactivation nor significant efficacy.
  • AbATE’s results further fuel questions about whether the negative results of the phase 3 Protégé trial actually indicate that teplizumab is ineffective, or were simply a product of study design. As a reminder, Protégé was a phase 3 trial (n=763) testing the efficacy of three different doses of teplizumab (14-day treatment, ~17 mg; 1/3 regimen, ~5.6 mg; and six-day regimen, ~4.6 mg) given two times, six months apart. Protégé failed to meet its primary endpoint, which was a composite of A1c and insulin dose (A1c <6.5% and insulin dose of<0.5 U/kg/day) after one year (for more on these negative results please see our October 21, 2010 Closer Look at However, the 14-day regimen used in the study had a significant impact on C-peptide preservation at two years (p=0.027). For more details on the year two results, please see page 2 of our EASD 2012 Report – Type 1 Diabetes Treatments and Cure-Based Therapies at Since the one-year results came out, there has been some debate about whether teplizumab truly ineffective, if the wrong endpoint was selected, or if the dose was lowered too much from the dose used in phase 2 trials (which was found to be efficacious but had a poor safety/tolerability profile). For an example of this debate please see our 12th Annual Rachmiel Levine Diabetes and Obesity Symposium report at
  • It’s hard to say where this therapy could go, given the disappointment associated with phase 3 results. It is especially disappointing given that many seem to believe that the failed results had more to do with study design than with deficiencies of the drug. We will come back to this subject after trying to talk to more researchers about their opinions about the potential future of the drug.

-- by Hannah Deming, Lisa Rotenstein, and Kelly Close