Diabetes after pancreaticoduodenectomy: can we predict it?

There is limited literature about the perioperative factors which can predict endocrine insufficiency after pancreaticoduodenectomy (PD). The primary aim was to correlate percentage pancreatic remnant volume (%RV) after PD in nondiabetic patients with the development of new-onset impaired glucose tolerance/diabetes mellitus (IGT/DM). The secondary aim was to identify the risk factors for new-onset IGT/DM.

In this prospective study, all consecutive patients with resectable periampullary carcinoma and without IGT/DM were evaluated with fasting and postprandial plasma glucose, HbA1c, insulin, and C-peptide levels preoperatively and at 3 mo postoperatively. After that, all patients were followed up with fasting and postprandial plasma glucose level assessed at 3-mo intervals for 24 mo or till death, whichever occurred earlier. The %RV was determined from computed tomography measurements preoperatively.

Of the 50 patients, 11 (22%) patients developed IGT/DM after median follow-up of 32 mo. The patients' with/without IGT/DM were similar in demographic/perioperative variables. The %RV was found to be an independent factor associated with new-onset IGT/DM. A %RV of <48.8% was found to be a predictor of new-onset IGT/DM (sensitivity, 89.7%; specificity, 73.6%). Plasma sugar and glycosylated hemoglobin levels were significantly higher postoperatively after PD than the preoperative levels. Insulin and C-peptide levels were significantly lower after PD, irrespective of new-onset IGT/DM.

The incidence of IGT/DM after PD was 22%, and %RV < 48.8% was found to be a significant risk factor for new-onset IGT/DM. (CTRI/2013/12/004233).