Predictors of Response to Checkpoint Inhibitors in Melanoma Revealed

Share this content:
A computational tool appears to accurately tease out the portion of patients who might respond to checkpoint inhibitors from those who are likely to be resistant.
A computational tool appears to accurately tease out the portion of patients who might respond to checkpoint inhibitors from those who are likely to be resistant.

Researchers developed a new method to predict whether a patient with melanoma will respond to treatment with a checkpoint inhibitor. The tool, described in Nature Medicine, is reportedly more robust than other regression predictors because it works across many different sets of patients with melanoma, according to the authors of the study.1

With the understanding that checkpoint inhibitors do not work in every patient, investigators analyzed samples across different studies (which included public records of tumor molecular profiles) to test the robustness of their proprietary computational forecasting framework, which they called Tumor Immune Dysfunction and Inclusion (TIDE). They postulated that the TIDE signature could reflect a late stage of T-cell breakdown. They computed reliable TIDE signatures for 5 different cancer types, but chose melanoma to investigate further, as “only melanoma has publicly available data on tumor expression and clinical outcome of patients treated with anti-PD1 or anti-CTLA4.”1

Continue Reading Below

The researchers hypothesized that measurement of the gene expression in tumors coupled with the level of infiltration of cytotoxic T lymphocytes (CTL) could be used to predict survival in those who had been treated with anti-PD1 or anti-CTLA4 agents more accurately than through the measurement of PD-L1 levels, tumor mutation load, and interferon gamma.

To hunt for the potential sources of tumor immune evasion, they reasoned “that by combining profiles of transcriptome profiles of treatment-naive tumors with patient survival outcome,” they could find known regulators of T-cell dysfunction/exclusion: those that work to keep cytotoxic T cells within the tumor from fighting back, and those that prevent T cells from even getting into the tumor.1

The investigators used pretreatment RNA-Seq or NanoString tumor expression profiles ― also known as transcriptome signatures ― to identify the genes that influence the function of cytotoxic T cells.

The authors observed that “[a]mong the data sets, TIDE predicted different numbers of genes to interact with CTL with statistical significance,” and that TIDE was significantly predictive of overall survival.1 To model tumor immune escape in tumors with various levels of CTL, TIDE factored in both T-cell dysfunction and exclusion signatures — whereas other biomarker tests involving drug response only search for one component.

Using TIDE, it was also determined that SERPINB9, which is overexpressed in melanoma, regulates resistance to T-cell killing and therefore, immune evasion. The authors said small-molecule inhibition of SERPINB9 is druggable, hinting that it will be a promising new focus for drug developers.

However, TIDE only worked as a prognostic tool in tumors of patients who were treatment-naive, so it is not a relevant tool for tumors that have progressed after first-line treatment with an immunotherapy, the authors warned.

The impact of TIDE could be widespread, according to the authors. “With additional data, cancer-type-specific regulators may be identified on the basis of the biological variations of T-cell dysfunction scores across different cancer types.”1 In anticipation of this endeavor, the investigators created an online portal for response forecasting using transcriptome profiles.


  1. Jiang P, Gu S, Pan D, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response [published online August 20, 2018]. Nat Med. doi: 10.1038/s41591-018-0136-1

Related Resources

You must be a registered member of Cancer Therapy Advisor to post a comment.

Sign Up for Free e-newsletters

Regimen and Drug Listings


Bone Cancer Regimens Drugs
Brain Cancer Regimens Drugs
Breast Cancer Regimens Drugs
Endocrine Cancer Regimens Drugs
Gastrointestinal Cancer Regimens Drugs
Gynecologic Cancer Regimens Drugs
Head and Neck Cancer Regimens Drugs
Hematologic Cancer Regimens Drugs
Lung Cancer Regimens Drugs
Other Cancers Regimens
Prostate Cancer Regimens Drugs
Rare Cancers Regimens
Renal Cell Carcinoma Regimens Drugs
Skin Cancer Regimens Drugs
Urologic Cancers Regimens Drugs

Cancer Therapy Advisor Articles