Sorroche Bruna Pereira, de Jesus Teixeira Renan, de Souza Vinicius Gonçalves, Tosi Isabela Cristiane, Tostes Katiane, Laus Ana Carolina, Santana Iara Viana Vidigal, de Lima Vazquez Vinicius, Arantes Lidia Maria Rebolho Batista
Molecular Oncology Research Center, Barretos Cancer Hospital, Barretos, São Paulo, Brazil.
Department of Pathology, Barretos Cancer Hospital, Barretos, São Paulo, Brazil.
J Mol Med (Berl). 2025 May 3. doi: 10.1007/s00109-025-02550-z.
Melanoma poses a significant health concern due to its propensity to metastasize and its high mortality rate. Immunotherapy has emerged as a promising treatment strategy for harnessing the patient's immune system to fight tumor cells. However, not all patients respond equally to immunotherapy, highlighting the need for predictive biomarkers to identify potential responders and optimize treatment strategies. Using data from 579 immunology-related genes evaluated by the NanoString nCounter Human Immunology v2 Panel, we integrated transcriptomic data with the clinical characteristics of 35 individuals to develop a predictive signature for immunotherapy response in melanoma patients. Through comprehensive analysis, we identified 18 genes upregulated in non-responder patients and three upregulated in responder patients. In multivariate analysis, CD24, NFIL3, FN1, and KLRK1 were identified as key predictors with significant potential for forecasting treatment outcomes. We then calculated a score incorporating the expression levels of these genes. The score achieved high accuracy in discriminating responders from non-responders, with an area under the curve of 0.935 (p < 0.001). The signature was also significantly associated with progression-free survival, overall survival, and survival following immunotherapy (p < 0.001). The validation of the signature in two independent cohorts confirmed its robustness and applicability, with areas under the curve of 0.758 (p = 0.036) and 0.833 (p = 0.004), respectively. This study represents a significant advance in precision medicine for melanoma. By identifying patients unlikely to benefit from immunotherapy, our approach could help optimize treatment allocation and improve patient outcomes. KEY MESSAGES: Novel 4-gene signature predicts immunotherapy failure in melanoma. High accuracy for personalized treatment decisions. Signature associated with decreased survival for non-responders. Signature validated in independent cohorts, enhancing generalizability. Potential to tailor treatment strategies and avoid unnecessary burden to patients.
黑色素瘤因其易于转移的倾向和高死亡率而成为一个重大的健康问题。免疫疗法已成为一种有前景的治疗策略，利用患者的免疫系统对抗肿瘤细胞。然而，并非所有患者对免疫疗法的反应都相同，这凸显了需要预测性生物标志物来识别潜在的反应者并优化治疗策略。我们使用由NanoString nCounter人类免疫v2面板评估的579个免疫相关基因的数据，将转录组数据与35名个体的临床特征相结合，以开发黑色素瘤患者免疫疗法反应的预测特征。通过全面分析，我们确定了在无反应患者中上调的18个基因和在有反应患者中上调的3个基因。在多变量分析中，CD24、NFIL3、FN1和KLRK1被确定为具有显著预测治疗结果潜力的关键预测因子。然后，我们计算了一个包含这些基因表达水平的分数。该分数在区分反应者和无反应者方面具有很高的准确性，曲线下面积为0.935（p < 0.001）。该特征还与无进展生存期、总生存期和免疫治疗后的生存期显著相关（p < 0.001）。在两个独立队列中对该特征的验证证实了其稳健性和适用性，曲线下面积分别为0.758（p = 0.036）和0.833（p = 0.004）。这项研究代表了黑色素瘤精准医学的一项重大进展。通过识别不太可能从免疫疗法中获益的患者，我们的方法有助于优化治疗分配并改善患者预后。关键信息：新的4基因特征预测黑色素瘤免疫疗法失败。个性化治疗决策的高准确性。该特征与无反应者生存期缩短相关。该特征在独立队列中得到验证，增强了普遍性。有潜力调整治疗策略并避免给患者带来不必要的负担。
