Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 May;32(5):661-672.
doi: 10.1016/j.annonc.2021.02.006. Epub 2021 Mar 15.

High tumor mutation burden fails to predict immune checkpoint blockade response across all cancer types

D J McGrail  1 , P G Pilié  2 , N U Rashid  3 , L Voorwerk  4 , M Slagter  5 , M Kok  6 , E Jonasch  2 , M Khasraw  7 , A B Heimberger  8 , B Lim  9 , N T Ueno  9 , J K Litton  9 , R Ferrarotto  10 , J T Chang  11 , S L Moulder  9 , S-Y Lin  12
Affiliations

High tumor mutation burden fails to predict immune checkpoint blockade response across all cancer types

D J McGrail et al. Ann Oncol. 2021 May.

Abstract

Background: High tumor mutation burden (TMB-H) has been proposed as a predictive biomarker for response to immune checkpoint blockade (ICB), largely due to the potential for tumor mutations to generate immunogenic neoantigens. Despite recent pan-cancer approval of ICB treatment for any TMB-H tumor, as assessed by the targeted FoundationOne CDx assay in nine tumor types, the utility of this biomarker has not been fully demonstrated across all cancers.

Patients and methods: Data from over 10 000 patient tumors included in The Cancer Genome Atlas were used to compare approaches to determine TMB and identify the correlation between predicted neoantigen load and CD8 T cells. Association of TMB with ICB treatment outcomes was analyzed by both objective response rates (ORRs, N = 1551) and overall survival (OS, N = 1936).

Results: In cancer types where CD8 T-cell levels positively correlated with neoantigen load, such as melanoma, lung, and bladder cancers, TMB-H tumors exhibited a 39.8% ORR to ICB [95% confidence interval (CI) 34.9-44.8], which was significantly higher than that observed in low TMB (TMB-L) tumors [odds ratio (OR) = 4.1, 95% CI 2.9-5.8, P < 2 × 10-16]. In cancer types that showed no relationship between CD8 T-cell levels and neoantigen load, such as breast cancer, prostate cancer, and glioma, TMB-H tumors failed to achieve a 20% ORR (ORR = 15.3%, 95% CI 9.2-23.4, P = 0.95), and exhibited a significantly lower ORR relative to TMB-L tumors (OR = 0.46, 95% CI 0.24-0.88, P = 0.02). Bulk ORRs were not significantly different between the two categories of tumors (P = 0.10) for patient cohorts assessed. Equivalent results were obtained by analyzing OS and by treating TMB as a continuous variable.

Conclusions: Our analysis failed to support application of TMB-H as a biomarker for treatment with ICB in all solid cancer types. Further tumor type-specific studies are warranted.

Keywords: biomarker; immune checkpoint blockade; tumor mutation burden.

PubMed Disclaimer

Conflict of interest statement

Disclosure DJM, PGP, EJ, and S-YL have a pending patent on a gene expression signature to predict response to immune checkpoint blockade. MK has served as a consultant or advisory roles for Janssen, AbbVie, Ipsen, Pfizer, Roche, and Jackson Laboratory for Genomic Medicine, received research funding from AbbVie, Bristol Myers Squibb (BMS), and Specialized Therapeutics, and has an advisory role for BMS, Roche, MSD, and Daiichi Sankyo, and the institute receives research funding from AstraZeneca, BMS, and Roche outside the submitted work. ABH has stock options, is an advisory board member of Caris Life Sciences, serves on the advisory board of WCG Oncology, has received licensing fees from Celldex Therapeutics and DNAtrix, and received research funding from Merck. The remaining authors have declared no conflicts of interest.

Figures

Figure 1 |
Figure 1 |. Most new cases of cancer in the U.S. arise from tumor types where CD8 T cell infiltration is not associated with neoantigen load.
(A) Spearman correlation between CD8 T cell infiltration determined from RNAseq and neoantigen load across cancer types identifies two Categories of tumor types, those where increasing neoantigen load significantly correlates with increased CD8 T cell infiltration (Category I, red), and those where increasing neoantigen does not correspond with increased CD8 T cell infiltration (Category II, blue). Example correlation plots are shown to right. (B) Common types of newly diagnosed cancer in the U.S., representing over 80% of solid tumors,. Red bars indicate Category I cancers where CD8 T cell levels positively correlate with neoantigen load and blue bars indicate Category II cancers where CD8 T cells are not associated with neoantigen load. Inset percentages indicate estimated percent of new cases. Gray bars indicate tumor types that could not be classified due to lack of samples in TCGA. # indicates tumor types included in FDA approval. ACC, Adrenocortical carcinoma; BLCA, Bladder Urothelial Carcinoma; BRCA, Breast invasive carcinoma; CESC, Cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, Cholangiocarcinoma; COAD, Colon adenocarcinoma; DLBC, Lymphoid Neoplasm Diffuse Large B-cell Lymphoma; ESCA, Esophageal carcinoma; GBM, Glioblastoma multiforme; HNSC, Head and Neck squamous cell carcinoma; KICH, Kidney Chromophobe; KIRC, Kidney renal clear cell carcinoma; KIRP, Kidney renal papillary cell carcinoma; LAML, Acute Myeloid Leukemia; LGG, Brain Lower Grade Glioma; LIHC, Liver hepatocellular carcinoma; LUAD, Lung adenocarcinoma; LUSC, Lung squamous cell carcinoma; MESO, Mesothelioma; OV, Ovarian serous cystadenocarcinoma; NSCLC, Non-small cell lung cancer; PAAD, Pancreatic adenocarcinoma; PCPG, Pheochromocytoma and Paraganglioma; PRAD, Prostate adenocarcinoma; READ, Rectum adenocarcinoma; SCLC, Small cell lung cancer; SARC, Sarcoma; SKCM, Skin Cutaneous Melanoma; STAD, Stomach adenocarcinoma; TGCT, Testicular Germ Cell Tumors; THCA, Thyroid carcinoma; THYM, Thymoma; UCEC, Uterine Corpus Endometrial Carcinoma; UCS, Uterine Carcinosarcoma; UVM, Uveal Melanoma
Figure 2 |
Figure 2 |. High TMB predicts ICB response in Category I tumors where CD8 T cells positively correlate with neoantigen load.
(A) ICB response rate in metastatic endometrial cancer from KEYNOTE158 stratified by TMB. (B) ICB response rate in metastatic cervical cancer from KEYNOTE158 stratified by TMB. (C) ICB response rate in microsatellite stable (MSS) colorectal cancer from Chalabi et al. (localized) and Goodman et al. (metastatic) stratified by TMB. Cochran-Mantel-Haenszel. (D) ICB response rate in metastatic melanoma from Goodman et al. stratified by TMB. (E) ICB response rate in metastatic melanoma from Hugo et al. stratified by TMB. (F) ICB response rate in metastatic melanoma from Miao et al. stratified by TMB. (G) ICB response rate in metastatic bladder from Mariathasan et al. stratified by TMB. (H) ICB response rate in metastatic bladder from Snyder et al. stratified by TMB. (I) ICB response rate in metastatic bladder from Miao et al. stratified by TMB. (J) ICB response rate in metastatic non-small cell lung cancer (NSCLC) adenocarcinomas from Goodman et al. stratified by TMB. (K) ICB response rate in NSCLC adenocarcinoma from Rizvi et al. (JCO) stratified by TMB. (L) ICB response rate in metastatic NSCLC adenocarcinoma from Rizvi et al. (Science) stratified by TMB. (M) ICB response rate in metastatic NSCLC adenocarcinoma from Hellmann et al. stratified by TMB. (N) Hazard ratio for overall survival following ICB treatment stratified by TMB in various cancer types from Samstein et al., with negative log2(Hazard Ratio) representing better outcomes in patients with TMB-H tumors. Comparisons of response rates made using Fisher’s exact test unless otherwise specified, inset value indicates (# of responses/total # of cases). Survival comparisons made using log-rank test.
Figure 3 |
Figure 3 |. High TMB does not predict ICB response in Category II tumors where neoantigen load is not associated with increased CD8 T cell levels.
(A) ICB response rate in metastatic anal cancer from KEYNOTE158 stratified by TMB. (B) ICB response rate in MSS metastatic gastric cancer from Kim et al. stratified by TMB. (C) ICB response rate in metastatic head and neck squamous cell carcinoma (HNSCC) from Goodman et al. and Miao et al., stratified by TMB. (D) ICB response rate in HNSCC from MDACC patients stratified by TMB. (E) ICB response rate in lung squamous cell carcinomas from Hellmann et al. stratified by TMB. (F) ICB response rate in lung squamous cell carcinomas from Goodman, Rizvi, and Miao cohorts stratified by TMB. (G) ICB response rate in mixed metastatic squamous cell carcinoma (SCC) (head & neck, lung, urethral, cervical, and unknown) from Goodman et al. stratified by TMB. (H) ICB response rate in metastatic clear cell renal cell carcinoma from Braun et al. stratified by TMB. (I) ICB response rate in metastatic triple negative breast cancer from Voorwerk et al. stratified by TMB. (J) ICB response rate in adriamycin/cyclophosphamide-resistant TMB-H TNBC from ARTEMIS trial treated with either ICB compared to other targeted therapeutics. (K) ICB clinical benefit rate in metastatic prostate cancer from Subudhi et al. stratified by TMB. No objective responses were observed. (L) Kaplan-Meier curve showing overall survival following ICB treatment stratified by TMB in metastatic prostate cancer from Subudhi et al. (M) Hazard ratio for overall survival following ICB treatment stratified by TMB in various cancer types from Samstein et al., with negative log2(Hazard Ratio) representing better outcomes in patients with TMB-H tumors. (N) Hazard ratio for overall survival following ICB or non-ICB treatment stratified in high-TMB metastatic breast cancer from Samstein et al. N = 259. (O) Hazard ratio for overall survival following ICB or non-ICB treatment stratified in high-TMB glioma from Samstein et al. N = 207. Comparisons of response rates made using Fisher’s exact test, inset value indicates (# of responses/total # of cases). Survival comparisons made using log-rank test.
Figure 4 |
Figure 4 |. High TMB does not predict ICB response across all cancer types.
(A) ICB response rate in all Category I cohorts where neoantigen load correlates with CD8 T cell levels from Figure 2. Inset numbers indicate (number of responders / total number), error bars indicate 95% confidence interval. Odds ratio and significance determined by the Cochran-Mantel-Haenszel method. (B) ICB response rate in all Category II cohorts where neoantigen load does positively not correlate with CD8 T cell levels from Figure 3. Inset numbers indicate (number of responders / total number), error bars indicate 95% confidence interval. Odds ratio and significance determined by the Cochran-Mantel-Haenszel method. (C) ICB response rate in all merged TMB-H cohorts. Inset numbers indicate (number of responders / total number), error bars indicate 95% confidence interval. P-values over bars are for alternative hypothesis that the response rate is different than 20%. Significance between groups assessed by Fisher’s exact test. (D) Area under receiver-operating characteristic (AUROC) curve values for Category I and Category II tumors representing prediction of ICB response rate across TMB threshold values, where 0.5 represents random chance and 1.0 represents perfect prediction. Rank-sum test. See Figure S6 for individual plots. (E) Logistic regression for ability of TMB to predict response to ICB in Category I and Category II tumors, where a positive coefficient represents improved response in TMB-H tumors. TMB was taken as a continuous variable, and individual cohorts treated as random effects. (F) Hazard ratio for overall survival following ICB treatment in TMB-H vs. TMB-L tumors, with negative log2(Hazard Ratio) representing better outcomes in patients with TMB-H tumors. Cox proportional hazards model with cohort used as a stratification variable. (G) Hazard ratio for overall survival following ICB treatment treating TMB as a continuous variable, with negative log2(Hazard Ratio) representing better outcomes in patients with higher TMB tumors. Cox proportional hazards model with cohort used as a stratification variable.
See this image and copyright information in PMC

Comment in

References

    1. Wei SC, Duffy CR & Allison JP Fundamental mechanisms of immune checkpoint blockade therapy. Cancer Discov. 8, 1069–1086 (2018). - PubMed
    1. Chan TA et al. Development of tumor mutation burden as an immunotherapy biomarker: Utility for the oncology clinic. Ann. Oncol 30, 44–56 (2019). - PMC - PubMed
    1. Gubin MM, Artyomov MN, Mardis ER & Schreiber RD Tumor neoantigens: building a framework for personalized cancer immunotherapy. J. Clin. Invest 125, 3413–3421 (2015). - PMC - PubMed
    1. Cristescu R et al. Pan-tumor genomic biomarkers for PD-1 checkpoint blockade–based immunotherapy. Science (80-. ). 362, eaar3593 (2018). - PMC - PubMed
    1. Alexandrov LB et al. The repertoire of mutational signatures in human cancer. Nature 578, 94–101 (2020). - PMC - PubMed

Publication types

Substances