TOPLINE:

A risk score based on four factors — male sex, tumor size, World Health Organization (WHO) grade, and lymphovascular invasion — predicts recurrence in patients with node-negative pancreatic neuroendocrine tumors with 83% accuracy. With further validation, it could be used to guide post-surgery surveillance.

METHODOLOGY:

• For patients who undergo surgery for pancreatic neuroendocrine tumors (NETs), lymph node metastasis is a key predictor of recurrence. However, some patients are at a high risk for recurrence despite being node-negative, and tools to guide risk-adapted surveillance in this group are lacking.
• To develop and validate a recurrence risk score for patients with node-negative pancreatic NETs, researchers conducted a retrospective case-control analysis across five high-volume US institutions. It involved 770 patients with localized pancreatic NETs who underwent curative pancreatectomy between 2000 and 2023. Median follow-up was 50.6 months.
• The researchers used multivariable logistic regression to identify independent predictors of disease recurrence and construct a 13-point composite risk score.

TAKEAWAY:

• Overall, 82 patients (10.6%) had a recurrence, at a median of 32.4 months after surgery. Independent predictors of recurrence included male sex (odds ratio [OR], 2.21; P = .005), tumor size of at least 3 cm (OR, 2.64; P = .001), WHO grade of 2 or higher (OR, 3.70; P = .01), and lymphovascular invasion (OR, 3.84; P < .001).
• Those four factors were used to develop the composite risk score, which stratified patients into low-, moderate-, and high-risk groups. Tumor recurrence rates in those groups were 2.4%, 9%, and 27.7%, respectively. The scoring system effectively predicted recurrence with an area under the curve of 0.83 (P < .001).
• Kaplan-Meier survival curves demonstrated the validity of the risk classification: 10-year disease-free survival rates were 96.1% among low-risk patients vs 83.6% among moderate-risk patients and 51.3% among those deemed high risk.
• Genomic analyses revealed that somatic mutations in at least one of seven genes (DAXX, CDC42BPB, ERI2, GALNT9, MTOR, NUMAI, and TRPC7) were predictive of recurrence — suggesting a future potential role for genomic profiling to identify high-risk patients, the authors of the study wrote.

IN PRACTICE:

“We propose this four-factor system should be used for tumor recurrence risk assessment and validated on global databases,” the authors of the study wrote. They further suggested risk-based surveillance strategies, including 6-month interval monitoring of high-risk patients for 3 years, followed by annual monitoring for up to 10 years.

An accompanying editorial agreed that the risk score stands as a potential new prognostic model for node-negative patients. As for the authors’ proposed surveillance strategies, “further study will be required to achieve widespread adoption in clinical practice,” wrote Dominic Vitello, MD, and David J. Bentrem, MD, Northwestern University Feinberg School of Medicine, Chicago.

SOURCE:

The study, led by Marco Ventin, MD, Cedars-Sinai Medical Center, Los Angeles, was published online on December 17 in JAMA Surgery.

LIMITATIONS:

Limitations include possible selection bias based on lack of information on clinically relevant subgroups, such as patients undergoing parenchyma-sparing and node-sparing surgery. Further studies are needed to integrate the genetic component and explore its applicability in clinical settings.

DISCLOSURES:

The authors reported receiving no funding source for the study. One author disclosed receiving personal fees and grants outside of the submitted work.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.