P2100 - Utility of KRAS Mutation in the Diagnosis of Pancreatic Adenocarcinoma: A Systematic Review and Meta-Analysis

Katia El Jurdi, MD, MSc¹, Sachin Srinivasan, MD¹, Luke Johnson, MD², Chelsea Wuthnow, MD², Ryan Ford, MD², Wancai Yang, MD³, Jim Lu, MD, PhD³, Kyle Rowe, MD⁴, W. Ransom Kilgore, MD¹, Nathan Tofteland, MD², William J. Salyers, MD, MPH^1
1University of Kansas School of Medicine, Wichita, KS; ²University of Kansas School of Medicine-Wichita, Wichita, KS; ³Go Path Labs, Buffalo Grove, IL; ⁴KU Gastroenterology, Wichita, KS

Introduction: Pancreatic adenocarcinoma (PanCa) is a leading cause of cancer-related deaths worldwide. With recent advances in imaging, the detection rate of pancreatic masses has increased, requiring a need for non-invasive markers to help diagnose and differentiate PanCa from noncancerous lesions. Kirsten RAt Sarcoma (KRAS) gene mutations have recently gained popularity as a diagnostic and prognostic marker in PanCa. The aim of this study is to determine their diagnostic utility in PanCa.

Methods: We performed an electronic search of PubMed, Embase, Web of Science, and Cochrane databases for studies reporting on the diagnostic performance of KRAS mutation in PanCa from inception to March 12, 2021. Case reports/series, editorials and review articles were excluded. Primary outcomes were diagnostic odds ratio (DOR), sensitivity and specificity of KRAS mutation in PanCa. A hierarchal bivariate model was used, sensitivity and specificity were weighted based on prevalence. Subgroup analysis by diagnostic sampling modality [EUS/FNA (endoscopic ultrasound-guided fine-needle aspiration) vs. surgical specimen vs. liquid biopsy] was done. Area under the curve (AUC) and heterogeneity (I²) were reported. Statistical analysis was performed using statistical software R^®.

Results: Our search resulted in 1,399 articles, which were screened by two independent reviewers. A total of 23 studies [2,162 patients, mean age 63 ± 3.7 years, 50% males] were included for analysis. The overall DOR of KRAS for PanCa was 14.7 (95%CI 7.9-27.3, p< 0.001). With a prevalence of 58%, pooled sensitivity and specificity was 66% (95%CI 58-74) and 96% (95%CI 87-99), respectively. AUC was 0.82 and I² was 7.8% (Figure 1). Positive and negative predictive values were 85.6% and 61.5%, respectively. Subgroup analysis showed that the diagnostic accuracy of KRAS mutation for PanCa was maintained independent of the sampling modality. Using EUS/FNA, sensitivity was 74% (95%CI 62-84), specificity was 98% (95%CI 77-100), AUC was 0.86 and I² was 0%. When surgical specimen were analyzed, sensitivity was 65% (95%CI 39-84), specificity was 82% (95%CI 61-93), AUC was 0.74 and I² was 1.1%. With liquid biopsy, sensitivity was 54% (95%CI 39-76), specificity was 99% (95%CI 64-100), AUC was 0.85 and I² was 0%.

Discussion: KRAS mutations appear to have a good diagnostic test performance in detecting PanCa, with a DOR of 14. Once detected in a pancreatic mass specimen, regardless of the sampling modality used, it is highly specific in diagnosing PanCa.



Disclosures:


Katia El Jurdi, MD, MSc¹, Sachin Srinivasan, MD¹, Luke Johnson, MD², Chelsea Wuthnow, MD², Ryan Ford, MD², Wancai Yang, MD³, Jim Lu, MD, PhD³, Kyle Rowe, MD⁴, W. Ransom Kilgore, MD¹, Nathan Tofteland, MD², William J. Salyers, MD, MPH¹. P2100 - Utility of KRAS Mutation in the Diagnosis of Pancreatic Adenocarcinoma: A Systematic Review and Meta-Analysis, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.