Carbohydrate antigen 19·9 accurately selects patients for laparoscopic assessment to determine resectability of pancreatic malignancy

Abstract

Abstract Background Laparoscopy with laparoscopic ultrasonography (L–LUS) may be useful in the selection of patients for surgery to resect peripancreatic malignancy in addition to contrast-enhanced computed tomography (CE–CT). The present prospective study assessed the strategy of using carbohydrate antigen 19·9 (CA19·9) levels to select patients for L–LUS. Methods Patients with suspected peripancreatic malignancy that appeared resectable on CE–CT were selected for immediate surgery if CA19·9 was low (up to 150 kU/l, or up to 300 kU/l if serum bilirubin was above 35 µmol/l), or to L–LUS if CA19·9 was high (over 150 kU/l, or over 300 kU/l if serum bilirubin was above 35 µmol/l). Data were assessed to determine the clinical utility of this strategy. Results A total of 94 patients went straight to surgery, of whom 65 proved resectable: 63 of 80 with a low CA19·9 level but only two of 14 with a high CA19·9 level and gastric outlet obstruction. From 55 patients with high CA19·9 levels, L–LUS correctly identified 26 of 31 resectable tumours and eight of 24 unresectable tumours. Conclusion Using CA19·9 levels to help select patients with pancreatic malignancy for immediate surgery or L–LUS for further assessment of resectability effectively increased resection rates and reduced unnecessary laparotomies.