Groove pancreatitis, a masquerading yet distinct clinicopathological entity: Analysis of risk factors and differentiation

Objectives Our objective was to delineate predictive factors differentiating groove pancreatitis (GP) from other lesions involving the head of the pancreas (HOP). Methods A case-control study of patients older than 10 years was performed comparing patients with GP to those with other surgically resected HOP lesions. Results Thirteen patients with GP (mean ± SD age, 51.9 ± 10.5 years; 11 males [84.6%]), all with a history of smoking (mean, 37.54 ± 17.8 pack-years), were identified. Twelve patients (92.3%) had a history of heavy alcohol drinking (heavy alcohol [EtOH]). The mean lesion size was 2.6 ± 1.1 cm, and the CA 19-9 was elevated (>37 IU/mL) in 5 patients (45.5%). The most common histopathologic condition was duodenal wall cyst with myofibroblastic proliferation and changes of chronic pancreatitis in the HOP. Univariate analysis revealed decreasing age, male sex, weight loss, nausea/vomiting, heavy EtOH, smoking, and a history of chronic pancreatitis were predictive of GP. A multivariate analysis among smokers demonstrated that weight loss (P = 0.006; odds ratio, 11.96; 95% confidence interval, 2.1-70.2), and heavy EtOH (P < 0.001; odds ratio, 82.2; 95% confidence interval, 9.16-738.1) were most predictive of GP. Compared to pancreatic adenocarcinoma (n = 183), weight loss and heavy EtOH remained predictive of GP. Conclusion Groove pancreatitis in the HOP is associated with a history of heavy EtOH and weight loss. In the absence of these symptoms, it is essential to rule out a malignant lesion.