Back to Annual Meeting Program

Objective: Diabetes mellitus (DM) is postulated to be both a risk factor and manifestation of pancreatic neoplasia. Pancreaticoduodenectomy (PD) is often the treatment of choice for resectable tumors. Current data suggests that maintaining euglycemia following abdominal surgery modestly decreases certain adverse events. This study sought to evaluate the effects of long-term preoperative glycemic control as determined by Hemoglobin A1c (HbA1c) on outcomes following PD.
Design: Retrospective cohort review of prospective database
Setting: Academic tertiary center
Patients: 202 patients underwent PD by a single surgeon from 9/2006 to 1/2013. The mean age was 65.5 years and 76.2% of cases were performed for malignancy. HbA1c levels were assessed preoperatively; serial blood glucose levels and insulin requirements were recorded for 48 hours postoperatively.
Intervention: Not applicable
Outcome Measures: The primary outcome measure was overall operative morbidity as determined by the Clavien-Dindo scoring system. Secondary outcomes included individual adverse events of interest, time to dietary resumption, and length of stay.
Results: Preoperative HbA1c levels ranged from 4.0-13.5 mmol/mol. 57 patients (28.2%) carried a clinical diagnosis of DM and demonstrated increased preoperative HbA1C and elevated postoperative mean blood glucose levels and insulin requirements compared to non-DM patients (p<0.05 for all). 28 (13.9%) patients experienced a ‘major’ (Clavien>2) complication; mean HbA1c, blood glucose, and insulin requirements among these patients did not differ from those with minor or no adverse events (p=NS for all). Similarly, these glycemic parameters did not differ between most secondary outcomes assessed, including delayed gastric emptying and wound infection.
Conclusions: PD can be safely performed in patients with HbA1c levels suggestive of poor long-term preoperative glycemic control. Medical efforts to optimize glycemic control should not delay resection.