Dual-Color CRBN/CD138 Immunohistochemistry Assay for Measuring Cereblon Protein in Samples from Patients with Multiple Myeloma

Background: Cereblon (CRBN) is required for the antiproliferative activity of the IMiDs^® immunomodulatory drugs lenalidomide (LEN) and pomalidomide (POM) in multiple myeloma (MM), but its value as a biomarker of IMiD response remains unknown. We have shown a disconnect between CRBN mRNA and protein regulation, indicating that measurement of CRBN protein is critical. However, accurate quantification of CRBN protein levels is challenging due to scarcity of standardized and validated reagents and assays. Several publications have suggested that CRBN mRNA and/or protein levels detected with commercial reagents and assays correlate with LEN and/or POM treatment outcomes. In an effort to establish a standardized and validated approach for CRBN protein measurement in MM biopsies, we developed a dual bright-field immunohistochemistry (IHC) assay to measure CRBN levels in CD138+ MM cells.

Methods: Using the highly specific CRBN65 rabbit monoclonal antibody against amino acids 65-76 of human cereblon, a CRBN/CD138 dual IHC assay scored using the H-score method was developed and validated. H-scores range from 0 to 300 and are a result of the sum of the products for the percentage of tumor cells (0%-100%) and the intensity of staining (0-3). Since CRBN is a nuclear and cytoplasmic protein, both nuclear and cytoplasmic H-scores were recorded for each sample evaluated with the assay. The dual IHC assay with CRBN65 antibody was used to evaluate the high CRBN–expressing DF15 MM cell line and the low CRBN–expressing DF15R MM cell line (made resistant to POM), as well as bone marrow core biopsies or aspirate clots from 22 patients with MM. For comparison, the same set of samples were stained using a commercially available CRBN antibody in a single IHC assay format. To evaluate interpathologist score variation, H-scores were recorded by 3 pathologists examining the samples independently. Interassay variation was evaluated by staining serial sections of biopsies from the same patient in 3 different experiments and calculating the coefficient of variation (CV) among scores produced for each patient by the same pathologist.

Results: The assay was specific and able to detect high CRBN levels in DF15 cells (cytoplasmic score = 190; nuclear score = 150) and low CRBN levels in DF15R cells (cytoplasmic score = 20; nuclear score = 30). The dual IHC assay was able to detect different CRBN staining levels in biopsies and clots from patients with MM, with average cytoplasmic H-scores ranging from 67 to 240 and nuclear H-scores ranging from 17 to 250. Compared with our dual assay, the single IHC assay performed with a commercial CRBN antibody demonstrated lack of specificity, did not detect nuclear CRBN, and showed similar immunoreactivity in the DF15 and DF15R cells. In addition, the commercial CRBN antibody had a narrow range of detection for CRBN (estimated) in patients with MM (Table). Interpathologist comparison of MM biopsy H-scores from 3 pathologists demonstrated high concordance (r2 = 0.73). Assay precision was shown for both cytoplasmic and nuclear CRBN in MM bone marrow biopsies, with CV values of 2% and 5%, respectively.

Conclusion: A dual CD138/CRBN IHC assay was developed to provide a reliable semiquantitative method to evaluate CRBN protein levels in both bone marrow core biopsies and clots from patients with MM that is specific, precise, and reproducible. The data highlight the requirement for standardization in the evaluation of CRBN expression. This assay will be used in future trials to assess the value of CRBN as a biomarker in MM and other hematologic malignancies.

^a Commercially available antibody for CRBN.

^b Values are estimated because tumor cell cannot be reliably identified.