Dr Paul Feuerstadt Explains Findings of PUNCH CD2, CD3 Trials on RBX2660 for Recurrent C difficile

At Digestive Disease Week 2022, we spoke to Paul Feuerstadt, MD, assistant clinical professor at Yale University School of Medicine, gastroenterologist at PACT Gastroenterology Center, who presented part of a pair of posters demonstrating the efficacy of microbiota-based live therapetuc RBX2660 in preducing recurrence of Clostridioides difficile (C difficile) infection.

The American Journal of Managed Care^® (AJMC^®): Can you explain the design and purpose behind the PUNCH CD2 and PUNCH CD3 studies, which evaluated the efficacy of RBX2660 versus placebo in reducing C difficile recurrence?

Feuerstadt: The PUNCH CD2 trial was a trial that was designed as a dose ranging study. It was a prospective, double-blinded, randomized, placebo controlled trial where patients with greater than or equal to 2 recurrences of C difficile infection all received a standard of care antimicrobial. Following that, they were randomized to 1 of 3 arms: 2 doses of placebo separated by a week, a dose of RBX2660 followed a week later by placebo, or 2 doses of RBX2660 separated by a week.

As a result of that trial, a single dose of RBX2660 was decided to be the optimal dose, and that proceeded into the PUNCH CD3 trial. The PUNCH CD3 trial was a prospective, double-blinded, randomized, placebo controlled trial where patients with 1 or more recurrence of C difficile were included. All received a standard of care antimicrobial, but they were randomized to either receive a dose of placebo or a dose of RBX2660, and then follow for 8 weeks for recurrence and 24 weeks or 6 months for safety.

As a result of the similarities between these 2 trials—similar design, similar patient population, the same product, the same efficacy outcome, and similar safety follow up—it was decided that a Bayesian analysis would be performed leveraging the phase 2 data with the phase 3 data. A Bayesian analysis looks at the signals of efficacy of the placebo arm and the signal of efficacy in the intervention arm and, if they are similar, will allow the leveraging of the data to create a larger and heartier study. That's what was done here, and the overall efficacy of RBX2660 was 70.3% versus 58.1% in the placebo arm. That achieves statistical significance with a posterior probability of superiority of .986. That was super exciting because we see in a relatively large population consistent safety and efficacy for this product, especially across these pivotal phase 2 and phase 3 trials.

AJMC^®: What were some of the key findings in these studies, and what are their clinical implications?

Feuerstadt: There are several key findings. Probably the biggest finding for me as a clinician was the phase 2B trial had a 2-year follow-up, or 24 months. That's really huge, because that's the longest follow-up that I'm aware of for any of these microbiota-based live biotherapeutics. The efficacy has been consistent across clinical trials, and a very nice pair of abstracts that were presented at IDWeek 2021 looked at the 3 phase 2 trials and the 2 phase 3 trials for RBX2660 and showed a very consistent safety profile with full follow-ups, but also a very consistent efficacy profile across all of these environments.

In addition, from these trials, we also have seen that RBX2660 in a real-world population is safe and effective. Under enforcement discretion, there was a group of patients that either did not choose to participate in the clinical trial or were not eligible. RBX2660 was administered to them under this FDA overseeing program. In this real-world population, it was presented at the American College of Gastroenterology meeting in 2021 that the efficacy was 82.8% in that population.

This was a retrospective study, but it was a real-world study, with patients that weren't eligible for clinical trials or didn't choose to participate in clinical trials. So this is really reinforcement of the efficacy and the safety. The safety set from that study included 6 months of follow-up and there were no significant signals for concern for any safety. So this is a very safe and effective product.

AJMC®: What did you find in your analysis of both the PUNCH CD2 and PUNCH CD3 trials? Did any of these findings surprise you?

Feuerstadt: What we're presenting here at Digestive Disease Week 2022 is a subgroup analysis where we pool the data from the PUNCH CD2 trial and the PUNCH CD3 trial, whose trial designs we discussed before. Within the PUNCH CD3 trial, though, a Bayesian analysis was performed, leveraging the phase 2 data with the phase 3 data. With this analysis in this abstract considering subgroups, we did a pooled analysis including adults 18 and over who were enrolled in the PUNCH CD2 and the PUNCH CD3 trial who either received only placebo or only a single dose of RBX2660.

The way that we analyzed the data was by comparing odds ratios for treatment success among the subgroups, and the subgroups that we considered were age, gender, number of previous episodes of C difficile, race, ethnicity, site of geographic administration of the product, and antimicrobial duration prior to administration of RBX2660. When we looked at our demographics, we included a total of 221 patients that received RBX2660 and 131 patients who received placebo alone. There were no statistically significant differences when comparing the 2 cohorts for various demographic factors. This study did mimic all the major C difficile studies, including a higher proportion of women in the study, and the majority of the patients did received vancomycin.

When we did this pooled analysis and we looked at overall treatment efficacy or sustained clinical response at 8 weeks, pooling the PUNCH CD2 trial with the PUNCH CD3 trial, the efficacy of RBX2660 was 68.3% versus 55.0% in the placebo arm. Through a chi square analysis, this achieved statistical significance. When we analyzed the specific subgroup odds ratios, there were no differences according to age, gender, number of previous episodes, race, ethnicity, site of geography of administration, as well as antimicrobial duration prior to RBX2660 administration.

What this really implies is that the broader population of patients that have recurrent C difficile, no matter what subgroup they're in based on those categories, will have a consistent efficacy with this product and the safety seems consistent, as well. This is something that shows that this can be given to a wide population and be very effective.

One question you asked about was, when we looked at the odds ratios, we saw that some of them actually did kind of hug the neutral element either favoring RBX2660 versus placebo, and largely that was because of small numbers. Within the eastern section, there were actually specifically smaller numbers and the odds ratio was below 1. In that situation, it's most likely a selection bias, but this is why we do the statistical analysis comparing the subgroups to see if that difference was significant, and it didn't approach anything close to significant, so we believe it was the smaller numbers that played that role there.