The broad-spectrum antibiotic piperacillin-tazobactam used as perioperative prophylaxis significantly reduced surgical site infection (SSI) in patients undergoing open pancreatoduodenectomy (PD, or the Whipple procedure) compared with the standard-of-care cefoxitin, according to results from the first-ever registry-linked surgical clinical trial conducted in North America. The findings, published in JAMA on April 20, support using piperacillin-tazobactam as a standard of care for this patient population.
In the trial, led by Memorial Sloan Kettering Cancer Center (MSK), investigators at 26 hospitals in the United States and Canada randomized 778 patients undergoing an open Whipple procedure between November 2017 and August 2021 to receive either piperacillin-tazobactam or cefoxitin perioperatively. The trial (NCT03269994) was terminated early after reaching its primary end point, according to a predefined stopping rule. (1)
The SSI rate was 19.8% among 400 patients in the piperacillin-tazobactam group, significantly lower than 32.8% among 378 patients in the cefoxitin group. The use of piperacillin-tazobactam versus cefoxitin also resulted in significantly reduced rates of postoperative pancreatic fistula (12.7% versus 19%), and multiple downstream sequelae of SSI, including C. difficile colitis and sepsis. (1)
The study’s senior author is MSK surgical oncologist Michael D’Angelica, MD, FACS, the Enid A. Haupt Chair in Surgery, Attending Surgeon with the Hepatopancreatobiliary Service, and Director of both the Complex General Surgical Oncology Fellowship and Hepatopancreatobiliary Fellowship programs. The co-author is Ryan Ellis, MD, MS, Chief Surgery Fellow in the Complex General Surgical Oncology Fellowship program at MSK.
In this Q&A with Drs. D’Angelica and Ellis, we discuss how the trial came about, the primary and secondary findings and their practice-changing implications, and what’s next for this innovative research methodology.
Dr. D’Angelica: Since it was popularized nearly a century ago, the safety of the Whipple procedure has improved substantially, with a mortality rate of less than 2% at high-volume centers today. However, the operation remains highly morbid. The most common sources of morbidity are SSIs and postoperative pancreatic fistula, which combined occur in about 30% of patients. Postsurgical morbidities negatively affect patients’ quality of life, limit the receipt of adjuvant therapies, and are associated with reduced overall survival.
Dr. D’Angelica: It has been postulated that broad-spectrum antibiotics in patients undergoing PD may be a potential way to reduce SSIs and postoperative complications, but that has not been the standard of care. Dozens of papers have reported retrospective evidence, but there have been no prospective trials since it’s very challenging to get funding for a randomized controlled trial to answer a patient management question in a surgical patient population.
MSK surgeons perform a high volume of pancreatic operations, about 300 annually. The idea for our trial started at MSK several years ago when I was also head of the Clinical Trials Committee of the Americas Hepato-Pancreato-Biliary Association (AHPBA), in discussions with the American College of Surgeons (ACS). We were looking for affordable ways to run surgical trials and utilizing an already-paid-for data collection registry like the ACS National Surgical Quality Improvement Program (ACS NSQIP®) seemed like an ideal mechanism.
Dr. Ellis: I was a clinical scholar at ACS when I became involved with the study about seven years ago. The idea for a randomized controlled trial meshed well with the ACS NSQIP®, one of the most prominent prospective surgical registries in the world. It is used by nearly 1,000 hospitals in North America to collect patient outcome data for internal quality improvement initiatives. Community and academic hospitals were already routinely submitting data for the trial’s end points, and the ACS volunteered to centralize data collection for the randomized trial. This collaboration meant that we could do the trial without additional funding. MSK handled randomizing participating patients to receive piperacillin-tazobactam or cefoxitin, and the rest of the data collection was already paid for and essentially automated.
A reduction in the rate of postoperative SSIs to 19.8% with piperacillin-tazobactam versus 32.8% for the standard of care is a clear result for the trial’s primary end point. Were there any notable nuances in results for secondary end points?
Dr. D’Angelica: Piperacillin-tazobactam also reduced rates of C. difficile colitis and sepsis, which we expected to see with a reduction in SSIs as they are downstream sequelae of SSIs. However, we were surprised to see that something as simple as changing to a broad-spectrum antibiotic reduced the fistula rate.
Dr. Ellis: There is evidence that particular types of bacteria can cause intestinal connections to fail after other surgeries, such as colon surgery, but that association had never been demonstrated in pancreatic surgery before. Our findings contribute additional data to the growing hypothesis that the microbiome is directly involved in the development of postsurgical complications.
Dr. Ellis: To see such strong evidence of a dramatic difference in patient outcomes for such a simple change in patient management is unprecedented, to our knowledge. We’re confident piperacillin-tazobactam will quickly become the standard of care among surgeons now that our results are published.
Dr. D’Angelica: We also anticipate piperacillin-tazobactam will be formally incorporated into the surgical clinical practice guidelines for PD in the future, especially following the correlative studies we plan to publish in the coming months that describe some of the potential bacterial mechanisms driving postsurgical SSIs in greater detail.
Are you planning to conduct more randomized controlled trials with this registry-linked method to answer surgical questions?
Dr. Ellis: Yes. We are currently in the late stages of planning the second ACS NSQIP®-linked trial. It will compare perioperative steroids to no perioperative steroids in patients undergoing major hepatectomy at multiple North American hospitals. The study will serve as a large-scale follow-up to a positive single-center randomized trial conducted recently in Calgary. With the second trial, we are hoping to continue to incorporate diverse patient populations and hospital types into enrollment.
Dr. D’Angelica: We are also exploring a larger initiative that would allow us to execute many registry-linked surgical trials at the same time.