Concordance of three different PD-L1 diagnostic assays (1)
MPLN’s PD-L1 assay, the Ventana SP263, was included in the concordance study of three PD-L1 Diagnostic Tests by AstraZeneca and results reported at AACR Annual Meeting 2016.
The study conducted by AstraZeneca (2), which complements a Blueprint initiative (3) addressing harmonization of the available PD-L1 assays, compares Ventana SP263 developed for use with AstraZeneca’s immunotherapeutic drug durvalumab, an anti-PD-L1; Dako pharmDx PD-L1 22C, a Companion Diagnostic for Merck’s Keytruda, and Dako pharmDx PD-L1 28-8 a complementary test with Opdivo. All three tests measure the percentage of positive staining for PD-L1 tumor cells with a unique cut-off point for positivity corresponding to potential therapeutic response in lung cancer patients. The three tests had an overall agreement of more than 90% based on testing 500 specimens with squamous and non-squamous histology. In addition, the recently published “Blueprint PD-L1 IHC Assay Comparison Project” demonstrated that SP263, 22C3, and 28-8 clones show close alignment on tumor cell staining (6). The performance characteristics of PD-L1 antibody clones is area of active investigation, and includes a second phase of the Blueprint Project which will compare antibody clones on biopsy material.
The interpretive criteria for each antibody are different (4) and for most labs it is not feasible to maintain all four antibodies. Concordance in PD-L1 testing is an emerging area of interest. NCCN is collaborating with Bristol-Myers Squibb (5) on a study to understand how different assays measure PD-L1 protein expression, to measure the concordance of pathologist interpretation of various test results, and to evaluate the heterogeneity of PD-L1 protein expression within tumor samples. Further, the College of American Pathologists (CAP) will be releasing proficiency testing for PD-L1 testing in 2017 which will indicate further practical concordance between laboratories.
Commenting is turned off for this post.