Predicting response to checkpoint inhibitors in melanoma beyond PD-L1 and mutational burden

Abstract

Background: Immune checkpoint inhibitors (ICIs) have changed the clinical management of melanoma. However, not all patients respond, and current biomarkers including PD-L1 and mutational burden show incomplete predictive performance. The clinical validity and utility of complex biomarkers have not been studied in melanoma.

Methods: Cutaneous metastatic melanoma patients at eight institutions were evaluated for PD-L1 expression, CD8⁺ T-cell infiltration pattern, mutational burden, and 394 immune transcript expression. PD-L1 IHC and mutational burden were assessed for association with overall survival (OS) in 94 patients treated prior to ICI approval by the FDA (historical-controls), and in 137 patients treated with ICIs. Unsupervised analysis revealed distinct immune-clusters with separate response rates. This comprehensive immune profiling data were then integrated to generate a continuous Response Score (RS) based upon response criteria (RECIST v.1.1). RS was developed using a single institution training cohort (n = 48) and subsequently tested in a separate eight institution validation cohort (n = 29) to mimic a real-world clinical scenario.

Results: PD-L1 positivity ≥1% correlated with response and OS in ICI-treated patients, but demonstrated limited predictive performance. High mutational burden was associated with response in ICI-treated patients, but not with OS. Comprehensive immune profiling using RS demonstrated higher sensitivity (72.2%) compared to PD-L1 IHC (34.25%) and tumor mutational burden (32.5%), but with similar specificity.

Conclusions: In this study, the response score derived from comprehensive immune profiling in a limited melanoma cohort showed improved predictive performance as compared to PD-L1 IHC and tumor mutational burden.

Keywords: Algorithmic analysis; Borderline; Immune Desert; Inflamed; Ipilimumab; Nivolumab; Pembrolizumab.

Fig. 1

Survival analysis of melanoma patients based on PD-L1 expression levels and mutational burden. a, b Overall survival upon stratification based on PD-L1 positivity (tumor proportion score ≥ 1% versus < 1%) for metastatic melanoma patients treated (a) prior to the introduction of ICIs (historical controls; n = 94) and (b) ICI-treated melanoma patients (n = 137). c, d Overall survival upon stratification based on mutational burden [high (≥ 7.1 mut/Mb) versus non-high] for metastatic melanoma patients treated (c) prior to the introduction of ICIs (historical controls; n = 94) and (d) ICI-treated melanoma patients (n = 137). p-values are indicated

Fig. 2

Immunological landscape of melanoma and its association with multiple variables. a Unsupervised hierarchical clustering (rows = patients, columns = genes) for 394 immune transcripts identified three major groups defined as “inflamed” (group 1), “borderline” (group 2), and “immune desert” (group 3). CD8 (assessed by RNA-seq) and PD-L1 expression levels (assessed by immunohistochemistry), response to ICIs (as per RECIST v.1.1), CD8 infiltration pattern (assessed by a trained pathologist, CM and APS), and mutational burden (assessed by whole-exon sequencing) are depicted. b-d Overall survival for melanoma patients from the (b) inflamed (n = 131), (c) borderline (n = 81), and (d) immune desert (n = 88) groups upon stratification based on treatment (historical controls versus ICI-based immunotherapy). p values are reported

Fig. 3

CD8⁺ T-cell infiltration pattern and clinical benefits from immune checkpoint inhibition. a CD8⁺ T-cell infiltration pattern was assessed by a trained pathologist upon immunohistochemistry with a CD8-specific antibody. Representative images are depicted (scale bar = 500 μm or 1 mm). b, c Overall survival upon stratification based on infiltration pattern (non-infiltrated, infiltrated, excluded) for metastatic melanoma patients treated (b) prior to the introduction of ICIs (historical controls; n = 94) and (c) ICI-treated melanoma patients (n = 137). For all comparisons p > 0.05

Fig. 4

Predictive performance of the RS as compared to PD-L1 expression levels and mutational burden. a Response rates for mutational burden using a scale of very low, low, intermediate, high, and very high. While increasing response rates occur with an increase in mutational burden the range of values from 36% to 45% is also limited for clinical use. b Box plot shows a pair-wise comparison of mutational burden for CR, PR, SD, and PD. c Response rates for PD-L1 IHC using a tumor proportion score (TPS) with values of zero or negative, 1–4%, 5–10%, and > 10%. While increasing response rates occur with an increase in TPS the range of values from 40% to 100% is limited for clinical use. d Box plot shows a pair-wise comparison of PD-L1 IHC TPS for CR, PR, SD, and PD. e Objective response rates in groups of ICI-treated melanoma patients stratified by RS. Linear regression supports a dynamic range from 30% to 100%. f Box plot shows a pair-wise comparison of RS values for CR, PR, SD, and PD. g Overall survival upon stratification based on RS (≥ 50 versus < 50) for ICI-treated melanoma patients (n = 137), p value is indicated, for comprehensive immune profiling using response score (RS) with a bin width of 10. Response rate shows a dynamic range of values from near zero to greater than 95% with increasing RS. Survival curve for patients with a RS ≥ 50 and < 50 shows an improved survival for the former (p = 0.0012)