利用腫瘤新生抗原疫苗防止腫瘤復發

Abstract

癌症免疫療法最近取得了重要的進展，對已經轉移的晚期癌症也有顯著的治療效果。可惜的是，治療效果僅限於少數病人，大多數患者對免疫治療仍無反應。目前已知，免疫治療的效果和腫瘤內基因突變量以及腫瘤內T細胞數目有正相關，不僅如此，有研究也證實這些腫瘤內的T細胞會辨識經由基因突變產生的腫瘤新生抗原(neoantigen)。我們實驗室過去以放射線合併細胞激素治療腫瘤，在腫瘤組織內可引發大量的CD8+ T 細胞浸潤，使大型腫瘤完全消失。但有些小鼠在長時間觀察後，腫瘤仍會復發。我們認為放射線合併細胞激素治療引發的腫瘤內T細胞對腫瘤新生抗原具有專一性，因此能有效清除腫瘤。但之後，這群腫瘤專一性T細胞數量逐漸減少，造成腫瘤復發。如果應用腫瘤新生抗原製成疫苗免疫小鼠，維持或增加腫瘤專一性T細胞的數量和活性，或許能達到長期抑制腫瘤的目的。 我們選用B16F10 和CT26作為實驗的腫瘤模型，為了找到最適合的腫瘤新生抗原作為疫苗，我們首先從先前文獻搜尋腫瘤內突變點並預測這些短片段突變抗原表位(epitope) 和第一型主要組織相容性複合物 (MHC class I) 的親和力，之後根據親和力的強弱選擇其中21條腫瘤新生抗原作為疫苗。此外，我們進一步確認報告的突變確實存在於我們的腫瘤細胞株，目前已確認B16F10內可表現其中16種點突變，CT26有其中13種點突變。接著，我們根據胺基酸序列將腫瘤新生抗原設計成長片段胜肽疫苗，免疫小鼠後，以ELISPOT技術分析抗原免疫性，發現大部分腫瘤新生抗原疫苗活化的是CD4+ T 細胞反應，較少腫瘤新生抗原引發CD8+ T 細胞反應。另外，我們進一步驗證腫瘤新生抗原疫苗引發的T細胞反應可以有效分區別突變胜肽和野生型胜肽，但有趣的是，這些長片段胜肽疫苗引發的T細胞大部分無法辨識預測的抗原表位，因此，我們認為長片段胜肽疫苗可能不適用於引發抗腫瘤的T細胞反應。我們直接以預測的短片段突變抗原表位作為疫苗 (簡稱短片段胜肽疫苗)，並分析其抗原免疫性。發現其中分別僅4種B16F10短片段胜肽疫苗和3種CT26短片段胜肽疫苗引發T細胞反應，顯示直接注射短片段胜肽疫苗，無法引發有效的T細胞反應，未來需要繼續研發更有效的胜肽疫苗平台。 我們接著測試腫瘤新生抗原疫苗的治療效果。在B16F10腫瘤保護實驗中，活化CD4+ T細胞和活化CD8+ T細胞的長片段腫瘤新生抗原疫苗，對腫瘤就有保護效果。另外，具有很強免疫反應的短片段胜肽疫苗，也可有效抑制腫瘤生長。最後我們觀察疫苗是否可延緩腫瘤的復發，藉由我們實驗室先前研究的合併治療搭配疫苗療法治療B16F10大腫瘤，但無論搭配長片段或短片段胜肽疫苗，其治療效果都不顯著，顯示腫瘤新生抗原疫苗雖然具有潛力，但仍有許多改善的空間來達到更好的治療效果。

Immune checkpoint blockades have shown durable clinical responses in various cancers, but benefit only a small portion of patients. A high mutational load and T cell responses to neoantigens, a class of tumor-specific mutations, are thought to be essential for the success of checkpoint inhibitors. Thus, neoantigens are ideal targets for cancer treatment as they lack expression in normal tissues and can potentially be recognized by T-cell repertoire. Our previous data demonstrated that radiation in combination with cytokines efficiently induced infiltration of CD8+ T cells in the tumor tissue and resulted in tumor regression. However, most of the tumors eventually relapsed. The aim of this project is to investigate whether vaccination with neoantigens can expand neoantigen-specific T-cells that are induced by the combination therapy and improve long-term tumor control. Two independent murine tumor models, B16F10 and CT26, were used in this study. First of all, we identified mutations which were previously reported in these two tumor cell lines and predicted the binding affinity between the mutated epitopes and MHC class I molecules. 21 neoantigens were selected for therapeutic vaccines. Then, we validated the presence of these mutations in our tumor cells and found that 16 and 13 mutations were existed in the B16F10 and CT26 tumor models, respectively. To test the immunogenicity of these mutated epitopes, mice were vaccinated with synthetic long peptides. The results showed dominant CD4+ T cell responses were induced in comparison with CD8+ T cells. We further demonstrated that long peptide-induced T cells can discriminate between the mutated and wild-type peptides, however, these T cells failed to recognized the corresponding short peptides (the predicted T-cell epitopes), suggesting that the long peptide neoantigen vaccines may not be suitable as therapeutic cancer vaccines. We then applied short peptides as cancer vaccines. However, only 4 and 3 short peptides elicited CD8+ T cell responses in B16F10 and CT26 models, respectively. To test the therapeutic effect of neoantigen vaccines, in the tumor protection study, we demonstrated that long neoantigen peptides, which induced CD4+ or CD8+ T cell response, as well as the immunogenic short neoantigen peptide were able to confer antitumor effects. Finally, low response rates were shown in mice vaccinated with long or short peptides after treated with combination therapy of radiation and cytokine as previously described, suggesting that neoantigen vaccines had a potential to obtain anti-tumor effects. However, there were still much improvement for a better therapeutic efficacy.