Can preoperative Carbohydrate Antigen 19-9 predict metastatic pancreatic cancer? Results of a systematic review and meta-analysis

Syed S Raza,Hala Khan,Shahab Hajibandeh,Shahin Hajibandeh,David Bartlett,Nikolaos Chatzizacharias,Keith Roberts,Ravi Marudanayagam,Robert P Sutcliffe,Syed S. Raza,Robert P. Sutcliffe
DOI: https://doi.org/10.1016/j.hpb.2024.01.017
IF: 3.842
2024-02-01
HPB
Abstract:BACKGROUND: To investigate the relationship between preoperative Carbohydrate Antigen19-9(CA19-9)and pancreatic cancer occult metastasis.METHODS: Systematic search of MEDLINE, CENTRAL, Web of Science and bibliographic reference lists were conducted. All comparative observational studies investigating the predictive ability of preoperative CA 19-9 in patients with pancreatic cancer were considered. Mean CA-19-9 value in the pancreatic cancer patients with and without metastasis were evaluated. Best cut-off value of CA 19-9 for metastasis was determined using ROC analysis.RESULTS: Ten comparative observational studies reporting a total of 1431 pancreatic cancer patients with (n = 496) and without (n = 935) metastasis were included. Subsequent meta-analysis demonstrated that mean preoperative CA 19-9 level was significantly higher in patients with metastases compared to those without (MD: 904.4; 95 % CI, 642.08-1166.74, P &lt; 0.0001). The between-study heterogeneity was significant (I<sup>2</sup>: 99 %, P &lt; 0.00001). ROC analysis yielded a cut-off CA 19-9 level of 336 with a sensitivity and specificity for predicting metastasis of 90 % and 80 %, respectively (AUC = 0.90).CONCLUSIONS: CA 19-9 level is significantly higher in patients with metastatic pancreatic cancer. A preoperative CA 19-9 value of 336 should be considered as an acceptable cut-off value to design prospective studies.
gastroenterology & hepatology,surgery
What problem does this paper attempt to address?