Long Term and Oncologic Outcomes for Patients Receiving Piperacillin-Tazobactam or Cefoxitin as Antibiotic Prophylaxis for Pancreatoduodenectomy.

OBJECTIVE

To examine long-term and oncologic outcomes of broad-spectrum perioperative antimicrobial prophylaxis, compared with standard care antibiotics, in patients undergoing pancreatoduodenectomy.

BACKGROUND

A registry-linked randomized clinical trial demonstrated a reduction in surgical site infection (SSI) in patients undergoing pancreatoduodenectomy who received piperacillin-tazobactam as perioperative prophylaxis compared to cefoxitin. However, long-term outcomes remain undefined.

METHODS

Participant records from the American College of Surgeons National Surgical Quality Improvement Program database were matched to the National Cancer Database. Associations between treatment arm, perioperative complications, and oncologic outcomes were assessed. Logistic regression models investigated associations with oncologic outcomes. Survival analyses used Kaplan Meier and Cox proportional hazard modeling.

RESULTS

Of 778 patients, 471 (60.5%) were treated at Commission on Cancer accredited hospitals and eligible for linkage. Of those, 426 (90.4%) were matched in the NCDB including 203 (47.7%) treated with piperacillin-tazobactam and 223 (52.3%) with cefoxitin. Among patients with indications for chemotherapy, chemotherapy omission rates were similar between the treatment arms (9.4% in piperacillin-tazobactam vs 15.4% in cefoxitin; P=0.097). There was no association between the receipt of indicated adjuvant chemotherapy and receipt of piperacillin-tazobactam (OR 1.49; 95% CI 0.56-3.95) or presence of SSI (OR 0.58; 95% CI 0.21-1.65). Postoperative SSI was associated with poorer 3-year overall survival (HR 1.69, 95% CI 1.20-2.38).

CONCLUSIONS

While patients experiencing an SSI had poorer survival, piperacillin-tazobactam prophylaxis did not significantly affect oncologic care delivery. These results cement the long-term implications of perioperative complications in patients with cancer and highlight leveraging cancer registries for clinical trial data analysis.