Proteomic Mucin Profiling More Accurate Than Cytology and CEA in Identifying Cystic Precursors of Pancreatic Cancer
Pancreatic cystic lesions are frequently detected radiologic incidentalomas, a considerable proportion of which are pancreatic cancer precursors. In a study reported in the Journal of the National Cancer Institute, Jabbar et al found that proteomic mucin profiling of cyst fluid was more accurate than cytology and cyst fluid carcinoembryonic antigen (CEA) level in detecting malignant potential and predicting malignant transformation of pancreatic cystic lesions.
Study Details
In the study, a proteomic profiling method was devised and prospectively evaluated in consecutive patients referred to a tertiary center for endoscopic ultrasound–guided aspiration of cystic lesions from May 2007 through November 2008 (discovery cohort, n = 28 patients with 29 cysts) and from December 2008 through October 2012 (validation cohort, n = 50 patients with 50 cysts). Among all 78 patients combined, 42 were female and median age was 64 years. Among all 79 cysts, 37 (46.8%) had malignant potential (premalignant or malignant tumors). Among 29 cysts with available histology, 16 exhibited malignant transformation.
Proteomic evidence of any mucin except MUC6 was considered indicative of malignant potential and MUC1 expression was considered predictive of malignancy. For cyst fluid CEA, levels of > 192 ng/mL were defined as premalignancy and > 1,000 ng/mL as malignancy. All diagnostic evaluations were blinded to proteomics results. Histology was required to confirm the presence or absence of malignant transformation.
Malignant Potential
In the pooled cohorts, proteomic mucin profiling had 97.5% accuracy in identifying lesions with malignant potential, a significantly greater rate than the 71.4% accuracy with cytology (P < .001) and the 78.0% accuracy with cyst fluid CEA (P < .001). For proteomics, cytology, and CEA, sensitivity was 97.3%, 59.5% (P < .001), and 68.2% (P = .003); specificity was 97.6%, 82.5% (P = .03), and 83.8% (P = .05); positive predictive value was 97.3%, 75.9% (P = .02), and 73.4% (P = .007); and negative predictive value was 97.6%, 68.8% (P < .001), and 81.6% (P = .02).
The one false-negative result observed with proteomic profiling was in a case in which the cyst aspirate was grossly hemorrhagic. Cyst fluid requirements for CEA quantification (500 μL) precluded CEA analysis in 19 patients (24.4%), of whom 15 had a neoplastic cyst.
Malignant Transformation
Proteomic mucin profiling (MUC1 expression) had 89.7% accuracy in predicting malignant transformation, significantly greater than the 51.7% accuracy with cytology (P = .003) and the 57.1% accuracy with cyst fluid CEA (P = .02). For proteomics, cytology, and CEA, sensitivity was 87.5%, 25.0% (P = .001), and 33.3% (P = .005); specificity was 92.3%, 84.6% (P = 1.00), and 88.9% (P = 1.00); positive predictive value was 93.3%, 66.7% (P = .18), and 80.0% (P = .45); and negative predictive value was 85.7%, 47.8% (P = .04), and 50.0% (P = .06).
The investigators concluded, “Proteomic cyst fluid mucin profiling robustly discriminates benign, premalignant, and malignant [pancreatic cystic lesions]. Consequently, it may improve pancreatic cancer prevention and reduce the morbidity burden of unwarranted pancreatic surgery.”
Karolina Sjöberg Jabbar, MD, of Sahlgrenska University Hospital, Gothenburg, is the corresponding author for the Journal of the National Cancer Institute article.
The study was supported by The Swedish Research Council, Health and Medical Care Executive Board of the Västra Götaland Region, Swedish Society of Medicine, Swedish Cancer Foundation, and others. The study authors reported no potential conflicts of interest.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
