Comparison of Biomarker Assay Types in Predicting Response to Anti–PD-1/PD-L1 Treatment


Key Points

  • mIHC/IF had a significantly higher sROC AUC value vs PD-L1 IHC, GEP, and TMB assays in predicting response to treatment.
  • mIHC/IF and multimodality assays appeared to have better overall performance vs other assay types.

In a systematic review and meta-analysis reported in JAMA Oncology, Lu et al found that multiplex immunohistochemistry/immunofluorescence (mIHC/IF) assays appeared to have greater accuracy vs other types of assays in predicting response to anti–programmed cell death protein 1/programmed cell death ligand 1 (PD-1/PD-L1) treatment for solid tumors. Among other biomarker assay types, PD-L1 IHC, tumor mutational burden, and gene-expression profiling appeared to share similar accuracy.

Study Details

The study included data from studies examining use of PD-L1 IHC, tumor mutational burden, gene-expression profiling, and mIHC/IF assays to determine objective responses to anti–PD-1/PD-L1 treatment, including studies that combined more than one biomarker modality. For analysis of PD-L1 IHC, only clinical trials resulting in U.S. Food and Drug Administration approval of anti­–PD-1/PD-L1 treatments were included in the analysis.

A total of 45 eligible reports were identified that used PD-L1 IHC (n = 24), tumor mutational burden (n = 10), gene-expression profiling (n = 9), mIHC/IF (n = 7), or multiple modalities (PD-L1 IHC plus tumor mutational burden or gene-expression profiling, n = 6) and correlated assay results with response to treatment. In the current analysis, summary receiver operating characteristic (sROC) area under the curve (AUC) values for correlation of assay findings with response to treatment were determined for each assay modality.

Assay Accuracy

In total, the analysis included specimens of more than 10 solid tumor types from 8,135 patients. The weighted sROC AUC value for mIHC/IF was 0.79, higher than the values of 0.65 for PD-L1 IHC (P < .001), 0.65 for gene-expression profiling (P = .003), and 0.69 for tumor mutational burden (P = .049); for use of the combined modalities of PD-L1 IHC plus tumor mutational burden or gene-expression profiling, the AUC was 0.74 (P = .48). 

Positive predictive value was higher for mIHC/IF (0.63) vs other individual assays. Multimodality assays (range = 0.34–0.42) and positive likelihood ratios (indicating lower likelihood of false-positive results) were highest for mIHC/IF and multimodality assays.  Negative predictive values (range = 0.75–0.88) and negative likelihood ratios were similar among methods.   

The investigators concluded, “In this meta-analysis, [tumor mutational burden], PD-L1 IHC, and [gene-expression profiling] demonstrated comparable AUCs in predicting response to anti–PD-1/PD-L1 treatment. mIHC/IF and multimodality biomarker strategies appear to be associated with improved performance over PD-L1 IHC, [tumor mutational burden], or [gene-expression profiling] alone. Further studies with mIHC/IF and composite approaches with a larger number of patients will be required to confirm these findings. Additional study is also required to determine the most predictive analyte combinations and to determine whether biomarker modality performance varies by tumor type.”

Janis M. Taube, MD, of the Division of Dermatopathology, Johns Hopkins University, is the corresponding author for the JAMA Oncology article.

Disclosure: The study was supported by the Melanoma Research Alliance, Emerson Collective, National Institutes of Health, National Cancer Institute, Bristol-Myers Squibb, Navigate BioPharma, and others. For full disclosures of the study authors, visit

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.