Premium
HP07
PROTEOMIC IDENTIFICATION OF SERUM MARKERS OF PANCREATIC ADENOCARCINOMA USING SELDI‐TOF MS
Author(s) -
Baxter R. C.,
Smith R. C.,
Scarlett C. J.,
Xue A.,
Butturini G.,
Scarpa A.
Publication year - 2007
Publication title -
anz journal of surgery
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.426
H-Index - 70
eISSN - 1445-2197
pISSN - 1445-1433
DOI - 10.1111/j.1445-2197.2007.04122_7.x
Subject(s) - medicine , biomarker , pancreatic cancer , logistic regression , adenocarcinoma , gastroenterology , pathology , cancer , biology , biochemistry
More accurate biomarkers for pancreatic cancer (PC) would improve early detection and monitoring progress. Aim To develop a biomarker panel for PC, using SELDI‐TOF technology and a bioinformatic approach and compare these markers to Ca19.9. Method Serum samples from patients from the University of Sydney (UoS) were used to derive a diagnostic panel of proteins which was tested blindly on samples from the University of Verona (UoV). From UoS there were 38 PC, 54 disease controls (pancreatico‐biliary) (DC) and 68 healthy volunteers (HV) samples. From UoV there were 40 PC, 21 DC and 19 HV patients. The PC patients from UoV underwent a resection and were therefore in a relatively early stage. The biomarker panel was developed using a logistic regression/10 fold cross validation approach1. Ca 19.9 was also measured. Discriminatory power was assessed by ROC curve analysis. Results The protein panel that best discriminated DC from PC included 7 protein peaks and HV from PC included 4 protein peaks. The discriminatory effect of the protein panel was similar to that of Ca19.9 but the combination of these panels and Ca19.9 was significantly improved P < 0.05, from 0.90 to 0.97 for comparison of PC with DC and from 0.90 to 0.99 for comparison of PC to HV. Results for samples from UoV were similar to those for UoS. Conclusion A panel of serum proteins from PC patients in combination with Ca19.9, significantly improved the ability to identify PC from DC and HV samples. These results were confirmed when tested blindly on blood samples from Verona.