Abstract 3187: Concordance of immunohistochemistry (IHC) assay results with gene expression profiling (GEP) methods for diffuse large-B-cell lymphoma (DLBCL) subtype identification

Abstract

Abstract Introduction: The survival of DLBCL patients who receive immunochemotherapy is correlated with the molecular signature of their tumors. The two most common subtypes which are defined by cell of origin (COO) are the germinal center B cell-like (GCB) and activated B cell-like (ABC) subtypes which have 3-year OS rates of 80% and 45%, respectively when treated with R-CHOP. Several methods have been developed to categorize DLBCL into these sub-types as the treatment they receive directly affects their outcome. Microarray GEP was first used to classify DLBCL into ABC or GCB, with a minority unclassified (UNC), however this is difficult to implement widely and has been replaced by more recent technologies. The most commonly used IHC-based Hans method profiles patients into GCB and non-GCB. The primary focus of this study was to evaluate concordance between current GEP and IHC methods to designate DLBCL subtypes. Methods: For this study, the IHC assay used was a standardized IHC assay based on the Hans algorithm and for the GEP-based assay; the NanoString Technologies Lymphoma Subtyping Test (LST) was used. Commercially available DLBCL samples (N = 138) were confirmed for tumor content and sent to a central laboratory for evaluation by the IHC assay. RNA was extracted from each sample and sent to NanoString for GEP evaluations. To compare methods homogeneously, calls for each method were converted to either “GCB” or “non-GCB”. All “ABC” and “UNC” subtype calls from the NanoString platform was converted to “non-GCB” and any “did not pass” (DNP) calls were converted to not available (NA). All NA values were excluded from the tables and concordance calculations. Results: The Positive Percent Agreement [95% CI] in calling non-GCB between IHC and LST was 87% [78% - 96%] (47/54). Overall concordance between IHC and LST was 80% [73% - 87%] (N = 125, 12 LST samples DNP, 1 IHC sample DNP). The Negative Percent Agreement between IHC and LST (i.e., in calling the GCB sub-type) was 75% [65% - 85%] using the LST result as a non-reference standard (53/71). Conclusions: Given that the success of targeted therapy for the treatment of DLBCL is contingent on the correct identification of sub-type of DLBCL, it is important to adopt a method that accurately and consistently identifies the sub-type in the majority of patients. Given the good concordance between the GEP and IHC methods, DLBCL can be stratified into COO sub-types using either platform. The results show that an optimized Hans IHC assay standardized in a central laboratory offers a robust, relatively inexpensive and accessible platform for cell of origin sub-typing in DLBCL and, with clinical validation, may offer an excellent tool for the selection of optimal therapy. Citation Format: Michael Schaffer, Shalini Chaturvedi, Sandy Frans, Regina Aquino, Jaehong Park, Brett Hall, Sriram Balasubramanian. Concordance of immunohistochemistry (IHC) assay results with gene expression profiling (GEP) methods for diffuse large-B-cell lymphoma (DLBCL) subtype identification. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3187.