Dr Holubar, from the Cleveland Clinic, discusses his research into the question of whether patients who take biologic therapies prior to surgery for IBD are more likely to develop postoperative complications.

Stefan Holubar, MD, is chief of the inflammatory bowel disease surgical section and director of research at the Cleveland Clinic in Cleveland, Ohio.

TRANSCRIPT

Good afternoon. My name is Stefan Holubar. I'm a colorectal surgeon in the Department of Colorectal Surgery in Cleveland Clinic, where I'm the director of research and also the section chief of the inflammatory bowel disease surgical section.

I presented an abstract on "Biologics First Before Surgery for Inflammatory Bowel Disease, Are They Associated with Postoperative Infectious Complications?"

This is a NSQIP IBD collaborative study in over 1500 patients. Many of you are aware that the advent of the biologics at the turn of the century has revolutionized the medical care for patients with inflammatory bowel disease.

However, there's been some initial concerns from the Mayo Clinic and Cleveland Clinic and more studies that were published in 2007 and ‘08 that these biologic medications, which were still relatively new at the time, were associated with postoperative complications. The bottom line, the take-home message, is that it remains very controversial in the surgical IBD literature whether or not exposure to biologics before surgery has any impact on the postoperative outcomes.

Part of the problem is that there's a lot of confounding. There's many different variables that may be at play here, and it's very difficult to tease those out. We aim to use a relatively new data set, which is the National Surgical Quality Improvement Program. It's a national data set that's sponsored by the American College of Surgeons. It's been around since about 2005.

Several years ago, my colleague Sam Eisenstein from University of California at San Diego, we started an IBD surgical collaborative using this platform. In 2017 and '18, we collected additional granular IBD data on about 1500 patients. This included the biologic medication, and this was not previously in the NSQIP data set.

We also had steroid use as well as whether or not the patient was diverted with a stoma and whether or not they had immunomodulators. This was really a nice upgrade. The success of this project is due to the many centers that collaborate in the NSQIP IBD collaborative.

Right now, we have about 10, including Cleveland Clinic, UCSD, Mount Sinai in New York City, Mass General Hospital, Lahey Clinic, Beth Israel Deaconess, Penn State, Emory, Wash U, Stanford. Since then, multiple other study sites have joined.

Getting back to the study at hand, this study included both Crohn's disease and ulcerative colitis patients. We wanted to tease out if biologic use within 60 days of surgery, as it was defined in our NSQIP study, was associated with postoperative surgical-site infections primarily, or secondarily, any surgical-site infections.

Some of the high-level findings were that about half of the patients, 47% or 730 patients, had been exposed to biologics before surgery, which shows, at these IBD centers, at least, that biologics are really being used in about half of the patients who require surgery. The univariate analysis did show some confounding. The bottom line on the univariate analysis for the endpoint of surgical-site infection or infectious complications was that biologic exposure did not seem to be associated.

Now, when we did the multivariate analysis, these same findings held true that biologics did not appear to be associated with any infectious complications, nor any surgical-site infections. However, Crohn's disease and proctectomy both were independently associated with infectious complications, as was the operative length.

Now we did a secondary analysis of risk factors associated with anastomotic leak after proctectomy because the proctectomy did come out as associated with infectious complications. In that subgroup analysis, we actually did find that biologics were associated with an increase in anastomotic leak rate. The statistical significance of this was only at a univariate level of 0.02, so it remains to be seen if this is really a robust effect or not. Presently, we are gathering an additional year of data to increase the size of this cohort to see if we can't suss that more subtle secondary finding, but I would not hang my hat on that at this time on at all.

Some of the limitations of this study were that it is only limited to 30-day follow-up, but for most surgical outcomes, that's OK. As I mentioned, although we had 1500 patients and this is the largest study— larger than the PUCCINI study, actually— it's still relatively low numbers in subgroups, such as those with pouches or certain types of ileocolic resection.

Subgroup analysis, as mentioned, it was somewhat limited due to limited numbers. But the strength of the study, as mentioned, it's one of the largest studies to date, if not the largest. Looking at this question, it's been vexing us for the last 20 years, basically.

There are multivariate analyses with robust adjustment for monitoring confounders, such as stoma use and concurrent immunomodulator or corticosteroid use, anemia and malnutrition, all of which we believe are associated with postoperative infectious complications. When adjusting for all of those, biologics did not seem to be associated with postoperative infectious complications, nor overall surgical-site infections.

The interesting hypothesis that got generated out of this work is that biologics before proctectomy may be associated with an increased anastomotic leak. Hopefully, we'll be back next year to share more data on that.

I want to thank all the different collaborators and NSQIP champions and surgical clinical reviewers from the 10 institutions that helped contribute data to this project and make it possible. Thank you.