Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

Indels Provide Highly Immunogenic Neoantigens and May Be Potential Biomarkers of Immunotherapy Response

The response to immunotherapy associates with the burden of insertion and deletion mutations across several tumour types
31 Aug 2017
Translational Research;  Immunotherapy

A pan-cancer assessment of indel load across 5777 tumour samples spanning 19 cancer types highlight the importance of frameshift neoantigens alongside non-synonymous single nucleotide variants (nsSNVs) neoantigens as determinants of immunotherapy efficacy and potentially crucial targets for vaccine and cell therapy interventions. The observations in renal cell cancer (RCC) might reconcile the observed immunogenicity of this tumour type despite its low overall mutational burden.

Investigators analysed whole-exome sequencing data from solid tumours from The Cancer Genome Atlas in a search whether the frameshift nature of insertion and deletion (indel) mutations might contribute to the immunogenic phenotype. They showed that tumour-specific neoantigens generated by indel mutations associated with the upregulation of antigen presentation genes and T-cell activation, as measured by CD8-positive expression in RCC. In addition, frame shift indel count significantly associated with the response to checkpoint inhibitors across separate three melanoma cohorts, according to an article published in the August 2017 issue of Lancet Oncology.

Checkpoint inhibitor therapies have shown clinical benefit and received regulatory approval for treatment of several types of solid tumours, including clear renal cell carcinoma (RCC) and melanoma. It has been hypothesised that tumour-specific T-cell responses are pre-existing in some patients and kept in check by immune modulatory mechanisms that may be activated by inhibiting molecular interactions at key checkpoints.

Joint first authors Samra Turajlic and Kevin Litchfield and colleagues investigated the role of indels as the source of the tumour mutations providing the key substrate to activate anticancer immunity. Thus far, the focus of such analyses has remained on nsSNVs; however, the authors studied the less commonly characterised effect of small scale indels on antitumour immunity in 19 types of cancer, including RCC, which is known to have a high level of tumour-infiltrating immune cells despite a nsSNV burden approximately 10 times lower than melanoma. 

The investigators studied the contribution to the immunogenic phenotype of frame shift indel mutations, which create novel open reading frames and mutagenic non-self peptides, by analysing whole-exome sequencing data from 5777 solid tumours from the Cancer Genome Atlas by PyClone and ASCAT to determine the clonal status of variants. They compared the indel burden and treatment response across four datasets wherein the checkpoint inhibitors anti-CTLA-4 and anti-PD-L1 were used. RNAseq profiling of 392 RCC cases was used to compare immune gene expression across patient subgroups.

Highest indel burden seen in RCC

The investigators determined that RCC have more than two-fold the proportion (0.12) and number of indel mutations across the pan-cancer cohort (p < 2.2 × 10−16) of all cancer types examined. Papillary and chromophobe RCC had the second and third highest indel proportion with a median indel number of 10 (95% confidence interval [CI] 9, 11) and 8 (95% CI 7, 10), respectively, suggesting a possible tissue-specific mutational process contributing to the acquisition of indels in RCC.

Indels provide neoantigens that are high-affinity binders and associate with T cell activation and response

Their analysis of tumour-specific neoantigens showed that enrichment of indel mutations for high-affinity binders was three times that of non-synonymous SNV mutations. The neoantigens derived from indel mutations were nine times enriched for mutant specific binding compared to neoantigens derived from non-synonymous SNVs.

Further immune gene expression analysis of the RCC cohort revealed that the detection of mutant-specific neoantigens was associated with antigen presentation gene upregulation, which also associated with T-cell activation measured by CD8-positive expression (r=0.78).

Response to checkpoint inhibition significantly associates with frame shift indel mutations 

Response to checkpoint inhibitors analysis revealed that the frame shift indel count significantly associated with response across three separate melanoma cohorts (p = 4.7 × 10−4). In the melanoma cohort, the median indel proportion value was 0.05 and the median indel count was 4.

Although further evaluation of exome sequencing results from 38 patients treated with anti-PD-1 for melanoma showed no association between response and nsSNVs (p = 0.27) or in-frame indels (p = 0.19), a significant association was found between anti-PD-1 response and frame shift indel mutations (p = 0.023). Furthermore, 7 of the 8 (88%) patients in the upper quartile with the highest burden of frame shift indels demonstrated a response to anti-PD-1 therapy, compared with the response of 43% in patients in the lower three quartiles, odds ratio 9.5; 95% CI 1.02, 89.23.

Additional checkpoint inhibitor response data from 162 patients in two separate trials1, 2 receiving anti-CTLA-4 treated confirmed that frame shift indel burden significantly associated with checkpoint inhibitor response, hazard ratio [HR] 3.4; 95% CI 1.05, 11.27 (p = 0.0074) and HR 2.9; 95% CI 1.15, 7.55 (p = 0.032)for the respective cohorts.

Pooled data from all three of these cohorts were evaluated in a meta-analysis that confirmed the frame shift indel count more significantly associated with checkpoint inhibitor response (p = 4.7 × 10−4) than did the nsSNV count (p = 4.8 × 10−3).


The authors stated that RCC has the highest pan-cancer proportion and number of indel mutations, which may suggest a tissue-specific mutational process.

They cited the differential benefit from checkpoint inhibitors among patients and the cost of these drugs as crucial to identifying biomarkers of response. Thus far PDL-1 expression and microsatellite instability-high status are the only biomarkers linked to drug approval. Although mutational and neoantigen burdens have been shown to correlate with clinical outcomes from checkpoint inhibitor therapy in patients with advanced melanoma, colorectal cancer, and non-small cell lung cancer, some patients with cutaneous melanoma and UV-protected mucosal melanomas have a low nsSNV burden still derive benefit from checkpoint inhibitors. Therefore, the investigators analysed two melanoma datasets having both response and mutational data, which demonstrated frame shift indel burden was a better predictor of response to either anti-PD-1 or anti-CTLA-4 therapy than nsSNV burden. Analysis of a third study showed the nsSNV burden and frame shift burden both significantly associated with checkpoint inhibitor response.

These findings suggest that indels are a highly immunogenic mutational class, which can trigger an increased abundance of neoantigens and greater mutant-binding specificity that figure prominently in the response to immunotherapy in RCC and melanoma and may provide a broader indication of response than nsSNV burden.


This trial was sponsored by Cancer Research UK, UK National Institute for Health Research (NIHR) at the Royal Marsden Hospital National Health Service Foundation Trust, Institute of Cancer Research and University College London Hospitals Biomedical Research Centres, the UK Medical Research Council, the Rosetrees Trust, Novo Nordisk Foundation, the Prostate Cancer Foundation, the Breast Cancer Research Foundation, the European Research Council.


Turajlic S, Litchfield K, Xu H, et al. Insertion-and-deletion-derived tumour-specific neoantigens and the immunogenic phenotype: a pan-cancer analysis. Lancet Oncology 2017;18(8):1009-1021.


  1. Van Allen EM, et al. Science 2015; 350:207–211.
  2. Snyder A, et al. N Engl J Med 2014; 371:2189–2199.
Last update: 31 Aug 2017

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings
  • Necessary cookies enable core functionality. The website cannot function properly without these cookies, and you can only disable them by changing your browser preferences.