SPARC Expression in Pancreatic Adenocarcinoma: Development of a Robust, Predictive Immunohistochemical Assay and Scoring Method
Peter B Illei, E Conde, N Dominguez, C Plaza, P Redondo, A Suarez-Gauthier, Carla Heise, Manuel Hidalgo, Fernando Lopez-Rios. Johns Hopkins Medical Institutions, Baltimore, MD; Laboratorio de Dianas Terapeuticas, Madrid, Spain; Celgene Corporation, San Francisco, CA
Background: SPARC is a glycoprotein involved in the regulation of extracellular matrix deposition and remodeling. Our aim was to develop a robust SPARC IHC assay and scoring method that would be adaptable and reproducible in pathology laboratories worldwide. Overexpression of SPARC assessed by IHC has been associated with poorer prognosis and improved survival in patients with advanced pancreratic carcinoma (PC) treated with nab-paclitaxel plus gemcitabine, an albumin-containing formulation of paclitaxel (Celgene, Summit,NJ), in a single arm phase I/II trial (VonHoff et al., 2012). Another objective was to assess value for predicting response to this therapy using tissue of the phase I/II clinical trial.
Design: Three SPARC antibodies were evaluated on formalin-fixed paraffin-embedded tissues of 50 resected PC (Invitrogen clone ON-1; Novocastra clone 15G12; and Sygma-Aldrich polyclonal). Staining was assessed in tumor and stromal cells. Based on these results, a novel scoring method was developed and 20 samples from the nab-paclitaxel plus gemcitabine phase I/II trial stained using the Invitrogen antibody and a Ventana Benchmark Ultra autostainer by two laboratories (Baltimore and Madrid). Slides were scored using a published method (Infante et al., 2007) and a novel method similar to HER2 assessment in gastric carcinomas (Ruschoff J et al., 2010) All readers were blinded to each other and to the clinical data.
Results: The Invitrogen antibody had the most intense and specific staining in the carcinoma and stromal cells. Concordance was high (85-95%) for all scoring criteria between the different pathologists. Kaplan-Meier survival analysis of overall survival (OS) using the Infante method showed no significant difference between positive and negative tumors. In contrast, Kaplan-Meier survival analysis of OS according to stromal SPARC expression using the novel scoring method showed significant survival benfit (21.2 vs. 6.1 months) in tumors with SPARC positive stroma.
Conclusions: The Invitrogen antibody produced the most specific stain and was the easiest to interpret. These preliminary results also suggest that SPARC expression analysis using IHC and a novel scoring method is reproducible. This IHC methodology will be used to assess the correlation of clinical outcome with SPARC expression from a randomized Phase 3 trial of nab-paclitaxel followed by gemcitabine versus gemcitabine alone in metastatic PC.
Tuesday, March 5, 2013 1:00 PM
Poster Session IV # 166, Tuesday Afternoon