External validation of a machine learning based algorithm to differentiate hepatic mucinous cystic neoplasms from benign hepatic cysts

Diagnosis and treatment of hepatic cysts. Usefulness of intracystic tumor markers (CEA and CA 19.9.)

INTRODUCTION

To decide treatment of hepatic cysts diagnosis between simple hepatic cyst (SHC) and cystic mucinous neoplasm (CMN). Radiological features are not patognomonic. Some studies have suggested the utility of intracystic tumor markers.

METHODS

Retrospective analysis of our prospective database including patients treated due to symptomatic SHC from 2003 to 2021. The aim of the study was to evaluate the results of treatment of symptomatic SHC and the usefulness of the determination of intracystic "carcinoembryonic antigen" (CEA) and "carbohydrate antigen" CA 19.9.

RESULTS

50 patients diagnosed and treated for symptomatic SHC were included. In 15 patients the first treatment was percutaneous drainage. In 35 patients the first treatment was laparoscopic fenestration. Four patients were diagnosed of premalignant or malignant liver cystic lesions (MCN, IPMN, lymphoma B); three of them required surgery after initial fenestration and pathological diagnosis. Median CEA and CA 19-9 were 196 μg/L and 227.321 U/mL respectively. Patients with malignant or premalignant pathology did not have higher levels of intracystic tumor markers. Positive predictive value was 0% for both markers, and negative predictive value was 89% and 91% respectively.

CONCLUSION

Values of intracystic tumor markers CEA and CA 19-9 do not allow distinguishing simple cysts from cystic liver neoplasms. The most effective treatment for symptomatic simple liver cysts is surgical fenestration. The pathological analysis of the wall of the cysts enables the correct diagnosis, allowing to indicate a surgical reintervention in cases of hepatic cyst neoplasia.