Immune checkpoint inhibitors (ICIs) have transformed cancer treatment, providing significant benefit to patients across various tumour types, including melanoma. However, around 40% of melanoma patients do not benefit from ICI treatment, and accurately predicting ICI response remains challenging. We now describe a novel and simple approach that integrates immune-associated transcriptome signatures and tumour volume burden to better predict ICI response in melanoma patients. RNA sequencing was performed on pre-treatment (PRE) tumour specimens derived from 32 patients with advanced melanoma treated with combination PD1 and CTLA4 inhibitors. Of these 32 patients, 11 also had early during treatment (EDT, 5–15 days after treatment start) tumour samples. Tumour volume was assessed at PRE for all 32 patients, and at first computed tomography (CT) imaging for the 11 patients with EDT samples. Analysis of the Hallmark IFNγ gene set revealed no association with ICI response at PRE (AUC ROC curve = 0.6404, p = 0.24, 63% sensitivity, 71% specificity). When IFNg activity was evaluated with tumour volume (ratio of gene set expression to tumour volume) using logistic regression to predict ICI response, we observed high discriminative power in separating ICI responders from non-responders (AUC = 0.7760, p = 0.02, 88% sensitivity, 67% specificity); this approach was reproduced with other immune-associated transcriptomic gene sets. These findings were further replicated in an independent cohort of 23 melanoma patients treated with PD1 inhibitor. Hence, integrating tumour volume with immune-associated transcriptomic signatures improves the prediction of ICI response, and suggest that higher levels of immune activation relative to tumour burden are required for durable ICI response.

中文翻译：

---

大小很重要：整合肿瘤体积和免疫激活特征可预测免疫治疗反应


免疫检查点抑制剂 （ICI） 改变了癌症治疗，为包括黑色素瘤在内的各种肿瘤类型的患者提供了显着益处。然而，大约 40% 的黑色素瘤患者没有从 ICI 治疗中受益，准确预测 ICI 反应仍然具有挑战性。我们现在描述了一种新颖而简单的方法，该方法整合了免疫相关转录组特征和肿瘤体积负荷，以更好地预测黑色素瘤患者的 ICI 反应。对 32 名接受 PD1 和 CTLA4 抑制剂联合治疗的晚期黑色素瘤患者的治疗前 （PRE） 肿瘤标本进行 RNA 测序。在这 32 名患者中，11 名患者在治疗早期 （EDT，治疗开始后 5-15 天） 也有肿瘤样本。在 PRE 评估所有 32 例患者的肿瘤体积，首先评估 11 例患者的 EDT 样本计算机断层扫描 （CT） 成像。Hallmark IFNγ 基因集的分析显示与 PRE 的 ICI 反应无关 （AUC ROC 曲线 = 0.6404，p = 0.24，敏感性 63%，特异性 71%）。当使用 logistic 回归用肿瘤体积 （基因集表达与肿瘤体积的比值） 评估 IFNg 活性以预测 ICI 反应时，我们观察到区分 ICI 反应者和无反应者的高区分能力 （AUC = 0.7760，p = 0.02,88% 敏感性，67% 特异性）;这种方法与其他免疫相关转录组基因集一起复制。这些发现在接受 PD1 抑制剂治疗的 23 名黑色素瘤患者的独立队列中进一步复制。因此，将肿瘤体积与免疫相关转录组特征相结合可以提高 ICI 反应的预测，并表明持久的 ICI 反应需要相对于肿瘤负荷更高水平的免疫激活。