肾细胞癌中酪氨酸激酶抑制剂应答的肿瘤微环境空间转录组学决定因素:一项2期临床试验结果。
完整摘要 · 中文译文
转移性透明细胞肾细胞癌(ccRCC)表现出异质性的临床结局,且可用于预测靶向治疗反应的生物标志物有限。本研究评估了预测酪氨酸激酶抑制剂(TKI)舒尼替尼临床反应的细胞区室特异性转录生物标志物。在这项2期临床试验(NCT00715442)中,46例转移性ccRCC患者在减瘤性肾切除术前和术后接受舒尼替尼治疗,主要终点为进展时间(TTP),次要终点为总生存期(OS)和舒尼替尼毒性。为探究舒尼替尼治疗反应的空间动态,我们对应答者和非应答者的肿瘤组织应用了全转录组数字空间谱分析(DSP)。对于46例试验患者,中位TTP为8.2个月(95% CI:6.1-19.3),OS为36.3个月(95% CI:19.7-70.5)。应答者显示出富含促炎/免疫刺激程序的肿瘤微环境,而非应答者则显示出免疫抑制程序的富集,伴有髓系细胞中PD-L1表达升高和耗竭CD8+ T细胞特征。鉴定出涉及VEGF依赖性和非依赖性血管生成通路的舒尼替尼耐药机制,揭示了潜在的治疗靶点。基于DSP的机器学习模型在外部JAVELIN Renal 101队列中接受舒尼替尼治疗的患者中显示出较强的预后和预测性能,但在阿维鲁单抗联合阿昔替尼队列中性能减弱,表明治疗背景特异性。我们的发现强调了免疫调节 作为转移性透明细胞肾细胞癌(ccRCC)中舒尼替尼(sunitinib)疗效的关键决定因素。空间分析揭示了潜在可靶向的酪氨酸激酶抑制剂(TKI)耐药机制,涉及SPP1相关髓系程序、CD8+ T细胞耗竭及血管生成信号,突出表明血管生成、脂质代谢与肿瘤免疫之间的相互作用是治疗结局的决定因素。
完整摘要 · 英文原文
Metastatic clear cell renal cell carcinoma (ccRCC) exhibits heterogeneous clinical outcomes, and limited biomarkers are available to predict response to targeted therapies. This study evaluated cell compartment-specific transcriptional biomarkers predicting clinical response to the tyrosine kinase inhibitor (TKI) sunitinib. In this phase 2 clinical trial (NCT00715442), 46 patients with metastatic ccRCC received sunitinib pre- and post-cytoreductive nephrectomy, with a primary endpoint of time-to-progression (TTP) and secondary endpoints of overall survival (OS) and sunitinib toxicity. To investigate spatial dynamics of therapeutic response to sunitinib, we applied whole-transcriptome digital spatial profiling (DSP) to tumor tissue from responders and non-responders. For 46 trial patients, median TTP was 8.2 months (95% CI: 6.1-19.3) and OS 36.3 months (95% CI: 19.7-70.5). Responders displayed a tumor microenvironment enriched in pro-inflammatory/immunostimulatory programs, while non-responders displayed enrichment of immunosuppressive programs with elevated PD-L1 expression in myeloid cells and exhausted CD8 + T-cell signatures. Sunitinib resistance mechanisms involving VEGF-dependent and independent angiogenic pathways were identified, revealing potential therapeutic targets. DSP-based machine learning models showed strong prognostic and predictive performance in sunitinib-treated patients in an external JAVELIN Renal 101 cohort, but attenuated performance in the avelumab plus axitinib cohort, indicating treatment-context specificity. Our findings highlight immune regulation as a key determinant of sunitinib response in metastatic ccRCC. Spatial profiling revealed potentially targetable mechanisms of TKI resistance involving SPP1-associated myeloid programs, CD8 + T-cell exhaustion, and angiogenic signaling, highlighting interactions between angiogenesis, lipid metabolism, and tumor immunity as determinants of therapeutic outcome.