Highlight
– Multicenter analysis of 994 total pancreatectomies (TP) from 25 expert centers defines 20 international reference values using a low-risk TP cohort (LR-TP, n=333).
– Key benchmarks for LR-TP include intraoperative blood loss ≤1000 mL, major complication rate ≤37%, 90‑day mortality <6%, and retrieved lymph nodes ≥29.
– TP carries substantially higher postoperative mortality and morbidity than pancreatoduodenectomy (PD), with especially poor outcomes when vascular resection or conversion to TP for high-risk anastomosis is required.
Background
Total pancreatectomy (TP) is an infrequent but sometimes necessary operation for multifocal pancreatic neoplasia, selected locally advanced tumors, or when surgeons elect to avoid a high-risk pancreatoenteric anastomosis and its attendant fistula risk. Although TP eliminates the risk of postoperative pancreatic fistula, it produces immediate and lifelong endocrine and exocrine insufficiency and has historically been associated with high perioperative morbidity and mortality. Contemporary perioperative outcomes have improved at high-volume centers, but international reference benchmarks for TP have not been well established. The multicenter study by Müller et al. in JAMA Surgery (2025) sought to provide such reference values from expert centers and to compare outcomes in predefined higher-risk subgroups and with established PD benchmarks.
Study design
This observational multicenter case-control study pooled consecutive primary TP procedures performed between January 2017 and November 2023 at 25 international expert centers. All indications (benign and malignant) were included. A prespecified low-risk TP (LR-TP) cohort excluded cases with vascular resection and patients with significant comorbidities, yielding 333 patients (33.5% of the total 994). The authors derived 20 reference values using the 75th or 25th percentile of the median values across centers — a benchmarking method intended to reflect realistic, achievable standards in expert practice.
Comparisons were made between LR-TP and three comparator groups: TP with vascular resection, TP performed to avoid a high-risk pancreatic anastomosis, and published benchmark values for low-risk pancreatoduodenectomy (PD).
Primary perioperative endpoints included intraoperative blood loss, major complications (likely defined using Clavien–Dindo grade ≥III), 90‑day postoperative mortality, failure-to-rescue, length of stay, readmission, and oncologic metrics such as number of lymph nodes retrieved.
Key findings
Population and cohorts: Of 994 TPs, 333 patients met LR-TP criteria (median age 66 years, IQR 58–72; 51.4% male). The remaining cohort included TP cases with vascular resections and conversions to TP for high-risk anastomoses.
Reference values from LR-TP cohort
The authors generated 20 reference cutoffs; headline examples include:
- Intraoperative blood loss: ≤1000 mL
- Major complications (presumed Clavien–Dindo ≥III): ≤37%
- 90‑day postoperative mortality (3‑month mortality): <6%
- Retrieved lymph nodes (oncologic yield): ≥29
- Other metrics, by implication, included acceptable ranges for length of stay, reoperation, readmission, and specific complication rates, derived from the 25th/75th percentile methodology (full table available in the manuscript).
Comparisons with higher-risk TP subgroups
TP with vascular resection performed worse than LR-TP on several major outcomes. Key examples:
- Major complication rate: 51% in TP with vascular resection versus the LR-TP reference ≤37% (benchmark not met).
- 90‑day mortality: 11% in TP with vascular resection versus LR-TP ≤6% (benchmark not met).
TP performed to avoid a high-risk pancreatoenteric anastomosis also exhibited worse outcomes:
- Failure-to-rescue rate: 38% versus LR-TP reference ≤6% (substantially worse).
- 90‑day mortality: 11% versus LR-TP ≤6%.
Comparison with pancreatoduodenectomy (PD) benchmarks
When benchmarked against reported low-risk PD reference values, LR-TP had meaningfully different profiles:
- Postoperative mortality: PD benchmark ≤2% versus LR-TP benchmark ≤6% — LR-TP mortality is roughly threefold higher.
- Lymph node yield: PD benchmark ≥16 nodes versus LR-TP ≥29 nodes — TP produces a higher nodal harvest consistent with more extensive nodal clearance but must be interpreted in context of indication and pathology.
Clinical and statistical significance
These results indicate that even in carefully selected low-risk patients treated at high-volume expert centers, TP carries substantial perioperative risk, greater than PD. The adverse outcomes are magnified when vascular resection is required or when TP is performed as a bail-out to avoid a risky anastomosis. The magnitude of differences (e.g., doubling to tripling of mortality/failure-to-rescue in higher-risk subgroups) is clinically significant and should inform patient selection, consent, and institutional benchmarking.
Expert commentary
Strengths of the study include its large international sample from high-volume centers, the predefined low-risk cohort to derive realistic benchmarks, and the direct comparison with clinically relevant higher-risk scenarios. Using percentile-based reference cutoffs reflects achievable standards rather than idealized best-case outcomes.
Limitations to consider:
- Observational design and pooled-center heterogeneity: Practice patterns, perioperative care pathways (enhanced recovery, diabetes management), and selection thresholds likely varied across centers and time (2017–2023).
- Case mix and indication bias: TP indications include benign multifocal disease and malignancy; outcomes may differ by pathology and preoperative therapies. Although the LR-TP cohort excluded vascular resections or major comorbidities, residual confounding is possible.
- Definition and reporting heterogeneity: The manuscript used common outcome definitions, but exact thresholds (e.g., for major complications) and reporting completeness can influence benchmark generation.
- Absence of long-term metabolic and quality-of-life outcomes in the benchmark set: TP’s lifelong endocrine (insulin-dependent diabetes) and exocrine insufficiency have major implications beyond 90 days and warrant integration into composite outcome appraisal.
Clinical implications:
- Patient selection: These benchmarks reinforce that TP should be reserved for patients in whom the anticipated oncologic or anatomic benefit justifies the higher perioperative risk and guaranteed endocrine insufficiency.
- Centralization and expertise: The data support centralizing TP at experienced centers with established multidisciplinary teams (surgical, endocrinology, nutrition, rehabilitation) to approximate LR-TP benchmarks and to manage complex cases such as vascular resections.
- Preoperative counseling: Surgeons should communicate that even in ideal low-risk scenarios, TP carries a meaningful chance of major complications and a nontrivial 90‑day mortality (~up to 6%), substantially higher than PD.
- Quality improvement and auditing: The 20 reference values provide a practical framework for internal and external audit, risk-adjusted outcome monitoring, and benchmarking against peers.
Conclusion
Müller et al. supply urgently needed international reference values for TP, derived from a large cohort of expert centers. TP, even when performed in low-risk patients, carries higher perioperative morbidity and mortality than PD, and outcomes deteriorate markedly when vascular resection or conversion to TP for high-risk anastomosis is needed. These benchmarks can inform patient selection, consent, institutional credentialing, and quality-improvement initiatives. Future work should integrate long-term metabolic outcomes, patient-reported quality-of-life, and risk-adjusted benchmarking tools to guide decision-making for TP versus parenchyma-sparing alternatives when feasible.
Funding and clinicaltrials.gov
Funding: Information as reported in the original JAMA Surgery article (Müller et al., 2025). Please refer to the published manuscript for detailed funding disclosures and conflicts of interest.
ClinicalTrials.gov: Not applicable; this was a multicenter observational surgical outcomes study rather than an interventional trial.
References
1. Müller PC, Berchtold C, Kuemmerli C, Breuer E, Li Z, Vallorani A, et al. International Reference Values for Surgical Outcomes of Total Pancreatectomy. JAMA Surg. 2025 Nov 12:e254941. doi: 10.1001/jamasurg.2025.4941. Epub ahead of print. PMID: 41222926; PMCID: PMC12613087.
AI-friendly thumbnail prompt
A high-resolution, clinical-feeling image of an operating-room scene from above: a surgical team around a patient on the table performing a complex pancreatic operation under bright lights; overlays of subtle infographic-style metrics (blood loss icon, complication percent, lymph node icon) in the corner; cool clinical color palette, professional and slightly dramatic lighting to convey high stakes and expertise.