Subclonal neoantigens present a key roadblock in MMRd tumor immunotherapy success

By Tarun Sai LomteReviewed by Lily Ramsey, LLMSep 20 2023

In a recent study published in Nature Genetics, researchers evaluated sensitivity to immune checkpoint blockade (ICB) therapy in autochthonous mouse models of DNA mismatch repair (MMR) deficiency (MMRd) colon and lung cancers.

Study: Mismatch repair deficiency is not sufficient to elicit tumor immunogenicity. Image Credit: Egoreichenkov Evgenii/Shutterstock.com

Background

Immunotherapy has revolutionized the therapeutic landscape of various cancers, especially those with high tumor mutational burden (TMB). Somatic mutations result in neoantigens, which induce tumor-specific T-cell responses. Increased TMB renders tumors (more) prone to immune attack post-ICB treatment.

Several studies have reported TMB as a strong predictor of ICB response, leading to the approval of pembrolizumab for all tumors based on elevated TMB alone.

Most patients with MMRd tumors lack durable responses, and TMB does not stratify responders, highlighting the need to investigate the factors beyond TMB that mediate efficacy.  

The study and findings

In the present study, researchers developed autochthonous mouse models of MMRd colon and lung cancers and determined their ICB sensitivity. Autochthonous lung tumors were induced in Kras^LSL-G12D Trp53^flox/flox Msh2^flox/flox (KPM) and Kras^LSL-G12D Trp53^flox/floxR26^LSL-Cas9 (KPC) mice. Similarly, autochthonous colon tumors were induced in R26^LSL-Cas9 mice.

Targeted sequencing of MMR genes in these tumors confirmed the predominance of frame-shifting mutations. KPC and KPM models had no significant difference in tumor burden.

Likewise, tumors in the colon model had no difference in growth kinetics. Further, the team performed pre-clinical trials with ICB in the KPM model to assess its sensitivity to immunotherapy.

Additionally, they included a chemotherapy combination (oxaliplatin and cyclophosphamide) alone or with ICB. KPM and KP (Msh2 wildtype) mice had no differences in tumor burden across treatments. Similarly, there were no differences in tumor burden at 16 weeks in KP and KPM mice depleted of T cells.

ICB treatment did not elicit responses in colon tumors. These data suggested that MMRd was insufficient to increase the sensitivity of tumors to ICB.

The researchers noted that TMB was mostly sub-clonal. Most mutations were detected in < 25% of MMRd tumors. The modifications adhered to the neutral evolution model of sub-clonal cancer mutations.

The team performed whole-exome sequencing (WES) of single-cell clones derived from a lung tumor cell line. MMR was restored before sub-cloning by re-expressing MutS homolog 2 (Msh2). These clones had sustained MSH2 expression with stable clonal and mutational architecture. However, more somatic mutations were called in clones than in the parental line.

Phylogenetic analysis of the clones confirmed substantial intratumor heterogeneity (ITH). Next, WES of Msh2^ko lung tumors from ICB-treated mice or those continuously depleted of T cells was performed. T cell depletion did not affect tumor neoantigen burden (TNB) or the overall TMB.

However, a significant increase in clonal TMB and TNB occurred when cancer cell fractions stratified mutations. Results were similar in the colon cancer model.

Next, the team analyzed survival after orthotopic transplantation of lung tumor cell lines and clones. Mice transplanted with parental Msh2^ko cell line or with clones (M1 to M8) derived from it and treated with ICB had 20% to 30% durable responses over 30 weeks.

Mice separately transplanted with five non-immunogenic Msh2^ko clones and subjected to ICB had 30% to 75% durable responses.

However, the remaining clones (M3, M7, and M8) were immunogenic without treatment, driving the disease with T-cell depletion alone. The researchers reasoned that the higher baseline immunogenicity could stem from the increased surface presentation of neoantigens.

Overall, data suggested that increasing mutational clonality could correlate with immunogenicity and response to ICB.

Finally, the researchers re-examined sequencing data from two clinical trials on advanced MMRd colorectal and gastric cancers. Clonal TNB was associated with objective response, while a high ITH index was associated with non-response. Clonal TNB was also associated with longer progression-free survival, whereas the ITH index was associated with shorter progression-free survival.

Conclusions

In sum, the study described sporadic MMRd mouse models of lung and colon cancers that enabled the analysis of mutations acquired continuously in vivo. Tumors in these models had extensive ITH and elevated burden of sub-clonal mutations.

The re-analysis of clinical trials on MMRd cancers revealed a significant association between the ITH index and clonal neoantigen burden with ICB response.

Together, the results raise concerns about therapies that deliberately increase TMB to enhance the immunogenicity of tumors, and these approaches will fail to induce significant immune engagement.

Moreover, collateral mutagenesis may lead to more aggressive cancer, secondary malignancies, or resistance. Further studies with models allowing for the temporal control of tumorigenic effects will help determine the impact of clonal selection on immunotherapy.