Clinical predictors of resectability of pancreatic adenocarcinoma.

BACKGROUND/

aimsIdentifying patient-related factors as well as symptoms and signs that can predict pancreatic cancer at a resectable stage, which could be used in an attempt to identify patients at an early stage of pancreatic cancer that would be appropriate for surgical resection and those at an unresectable stage be sparred unnecessary surgery. MATERIALS AND

methodsA retrospective chart review was conducted at a major tertiary care, university hospital in Riyadh, Saudi Arabia. The study population included individuals who underwent a computed tomography and a pancreatic mass was reported as well as the endoscopic reporting database of endoscopic procedures where the indication was a pancreatic mass, between April 1996 and April 2012. Any patient with a histologically confirmed diagnosis of adenocarcinoma of the pancreas was included in the analysis. We included patients' demographic information (age, gender), height, weight, body mass index, historical data (smoking, comorbidities), symptoms (abdominal pain and its duration, anorexia and its duration, weight loss and its amount, and over what duration, vomiting, abdominal distention, itching and its duration, change in bowel movements, change in urine color), jaundice and its duration. Other variables were also collected including laboratory values, location of the mass, the investigation undertaken, and the stage of the tumor.

resultsA total of 61 patients were included, the mean age was 61.2 ± 1.51 years, 25 (41%) were females. The tumors were located in the head (83.6%), body (10.9%), tail (1.8%), and in multiple locations (3.6%) of the pancreas. Half of the patients (50%) had Stage IV, 16.7% stages IIB and III, and only 8.3% were stages IB and IIA. On univariable analysis a lower hemoglobin level predicted resectability odds ratio 0.65 (95% confidence interval, 0.42-0.98), whereas on multivariable regression none of the variables included in the model could predict resectability of pancreatic cancer. A CA 19-9 cutoff level of 166 ng/mL had a sensitivity of 89%, specificity of 75%, positive likelihood ratio of 3.6, and a negative likelihood ratio of 0.15 for resectability of pancreatic adenocarcinoma.

conclusionThis study describes the clinical characteristics of patients with pancreatic adenocarcinoma in Saudi Arabia. None of the clinical or laboratory variables that were included in our study could independently predict resectability of pancreatic adenocarcinoma. Further studies are warranted to validate these results.