New-onset diabetes mellitus after distal pancreatectomy: Incidence, predictors and clinical impact

Table 1 Pathophysiological and diagnostic differences between new-onset diabetes mellitus, type 3c diabetes mellitus, and postoperative hyperglycemia

Feature	NODM	T3cDM	Postoperative hyperglycemia
Definition	De novo diabetes in previously normoglycemic patients	Diabetes secondary to pancreatic disease/surgery	Transient elevation in blood glucose
Primary driver	Surgical loss of β-cell mass	Pancreatic pathology + exocrine loss	Stress response/inflammatory cytokines
Exocrine status	May be normal or impaired	Must be impaired	Usually, unaffected
Imaging	Post-surgical anatomy	Structural damage (atrophy/fibrosis)	Typically, normal

NODM: New-onset diabetes mellitus; T3cDM: Type 3c diabetes mellitus.

Table 2 Key prospective and observational studies on diabetes and pancreatic function in patients with pancreatic disease and post-pancreatectomy outcomes

Ref.	Procedures	Outcomes
Dugnani et al[31], 2016	Prospective observational study in a tertiary referral center. Inclusion of 364 patients with PDAC. Assessment of diabetes status and timing relative to PDAC diagnosis. Classification of diabetes as: Long-duration, short-duration, concomitant with PDAC, early-onset after PDAC diagnosis, late-onset after PDAC diagnosis. Biochemical evaluation: Fasting blood glucose, insulin and/or C-peptide, HbA1c, anti-islet antibodies	Overall diabetes prevalence in PDAC patients: 67%. Diabetes present at PDAC diagnosis: 174 patients (47.8%). Clinical and biochemical profile similar to T2D. Associated with classic T2D risk factors (age, sex, BMI, family history). Evidence of both insulin resistance and β-cell dysfunction. Diabetes developing after PDAC diagnosis: 70 patients (19.2%). Strongly associated with β-cell mass loss after pancreatectomy. Occurred mainly in patients with pre-existing T2D risk factors. Overall interpretation: PDAC-associated diabetes largely overlaps with T2D phenotypes. Diabetes onset alone is a weak marker for early, asymptomatic PDAC. Identification of diabetes-independent PDAC biomarkers is needed
Maignan et al[32], 2018	Prospective follow-up of 91 patients undergoing pancreatic resection (2014-2015). Surgical procedures: PD: 72%, LP: 28%. Assessment of pancreatic insufficiency at 6 months: ExoPI: Fecal elastase < 200 μg/g. EndoPI: Fasting glucose > 126 mg/dL or worsening of preexisting diabetes. Measurement of residual pancreatic volume using standardized volumetric methods	6-month incidence: ExoPI: 75.9%, EndoPI: 30.8%. ExoPI significantly more frequent after PD vs LP (98% vs 21%; P < 0.001). EndoPI rates similar between PD and LP (28% vs 38.5%; P = 0.412). No difference in ExoPI between reconstruction techniques (PG vs PJ). Residual pancreatic volume < 39.5% predictive of ExoPI. ExoPI occurs almost universally after PD, regardless of reconstruction method
Illés et al[33], 2020	Prospective, multicenter observational cohort study. Inclusion of patients ≥ 60 years with new-onset diabetes (≤ 6 months). Exclusion criteria: Alcohol abuse, chronic pancreatitis, prior pancreatic surgery, pregnancy, active malignancy, type 1 diabetes. Follow-up every 6 months for up to 36 months. Data collection at each visit: Questionnaires, clinical symptoms and body weight, fasting blood samples. Every second visit: CA 19-9 measurement, blood sampling for biobank, plasma analysis using mass spectrometry-based metabolomics. Patients with suspicious findings undergo MRI and/or endoscopic ultrasound, with surgical referral if indicated	Validation of metabolomic biomarkers for early PDAC detection in new-onset diabetes. Identification of PDAC at an early, operable stage. Estimation of PDAC incidence in patients with newly diagnosed diabetes. Potential reduction in PDAC-related morbidity, mortality, and healthcare burden
Latenstein et al[34], 2021	Cross-sectional study in six Dutch centers. Inclusion of patients ≥ 3 years after PD or LP. Indications: Premalignant and benign (non-pancreatitis) pancreatic disease. Assessment tools: EQ-5D-5 L (generic quality of life), EORTC QLQ-C30. Questionnaires on exocrine and endocrine pancreatic insufficiency. PAID20 (diabetes-related distress)	153/183 patients responded (84% response rate); median follow-up 6.3 years. Surgery-related complaints in 47% of patients, 8.4% would not choose surgery again. EQ-5D-5 L VAS lower than general population (76 vs 82; P < 0.001). Global health status (QLQ-C30) similar to general population. Exocrine pancreatic insufficiency in 41%, with symptom relief from enzymes in 48%. New-onset diabetes in 14%. Low diabetes-related distress (median PAID20 6.9/20). Despite preserved overall quality of life, persistent symptoms highlight need for long-term follow-up and counseling
Lee et al[35], 2023	Prospective observational study, inclusion of 224 patients after partial pancreatectomy. PD: 152 (67.9%), DP: 72 (32.1%). Comprehensive glucose assessment: 75 g oral glucose tolerance test. Performed preoperatively, and at 1 week, 12 weeks, and 52 weeks postoperatively. Long-term follow-up up to 3 years. Outcomes monitored: NODM in patients without baseline DM; worsening of glucose regulation (≥ 1% increase in HbA1c) in patients with preexisting DM	Disposition index decreased at 1 week postoperatively in both groups. Recovery and improvement of β-cell function above baseline in PD group. Persistent β-cell dysfunction in DP group, with significant between-group difference at 1 year (P < 0.001). DP associated with higher risk of: NODM (HR = 4.29, 95%CI: 1.49-12.3). Worsening glucose regulation (HR = 2.15, 95%CI: 1.09-4.24). Distal DP and spleen preservation linked to better glucose outcomes. DP showed stronger association with glucose dysregulation than PD
Liu et al[36], 2025	Multicenter study in six centers in China, inclusion of 1211 PDAC patients undergoing R0 resection. Patients categorized by diabetes status: Non-diabetes mellitus (non-DM): 602 (49.7%), long-term diabetes (LTD): 127 (10.5%), NOD: 482 (39.8%). Comparison of perioperative outcomes (e.g., pancreatic fistula) and oncological outcomes (overall and recurrence-free survival)	NOD patients had higher rates of fatty pancreas and POPF (P < 0.05). Median OS: Non-DM: 29.4 months, NOD: 24.6 months (P < 0.001), LTD: 25.2 months (P = 0.041 vs non-DM). Median RFS: Non-DM: 15.8 months, NOD: 13.3 months (P < 0.001), LTD: 13.8 months (P = 0.007 vs non-DM). No significant survival difference between NOD and LTD groups. Multivariate analysis: NOD, LTD, larger tumor size, and poor tumor differentiation independently associated with worse OS and RFS. Conclusion: NOD increases risk of POPF and both NOD and LTD are linked to worse long-term survival after PDAC resection

PDAC: Pancreatic ductal adenocarcinoma; HbA1c: Glycated hemoglobin; T2D: Type 2 diabetes; BMI: Body mass index; PD: Pancreatoduodenectomy; LP: Left pancreatectomy; ExoPI: Exocrine insufficiency; EndoPI: Endocrine insufficiency; CA 19-9: Carbohydrate antigen 19-9; MRI: Magnetic resonance imaging; EQ-5D-5 L: EuroQol 5-Dimensions 5-Levels; EORTC QLQ-C30: European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30; PAID20: Problem Areas in Diabetes-20; DP: Distal pancreatectomy; NODM: New-onset diabetes mellitus; DM: Diabetes mellitus; HbA1c: Glycated hemoglobin; LTD: Long-term diabetes; NOD: New-onset diabetes; PG: Pancreaticogastrostomy; PJ: Pancreaticojejunostomy; VAS: Visual Analogue Scale; QLQ-C30: Cancer Quality of Life Questionnaire Core 30; HR: Hazard ratio; CI: Confidence interval; POPF: Postoperative pancreatic fistula; OS: Overall survival; RFS: Recurrence-free survival.