靶向PD-L1免疫PET/CT影像组学预测NSCLC免疫治疗效果的研究

Immunotherapy has substantially changed the therapeutic strategies for NSCLC. However, the immune checkpoint agents approved for the treatment of NSCLC have relatively low objective response rates that are correlated with PD-L1 expression levels. PD-L1 expression is usually detected through immunohistochemistry (IHC), which is not only time consuming but also limited to heterogeneity of PD-L1 expression within tumors. Thus, novel approach to assess PD-L1 expression level and estimate the potential value of chosen immune inhibitor is crucial in clinical practice..Preclinical and clinical studies of positron-emission tomography (PET) imaging with anti-PD-L1 antibodies suggested that this technique might be an approach to select patients. Our previous study with 89Zr-Df-KN035, an 89Zr labeled anti-PD-L1 domain antibody, demonstrating noninvasive imaging of tumor PD-L1 expression and allowing assessment of inter- and intra-tumoral heterogeneity. .Radiomics is a new technology which involves the analysis and translation of medical images into quantitative data. High-dimentional imaging data allow an in-depth characterisation of tumour phenotypes, which reflect not only macroscopic but also the cellular and molecular properties of tissues. Radiomics features are complementary to biopsies and have the advantage of being non-invasive, which allow evaluation of a tumor and its microenvironment, characterisation of spatial heterogeneity, and longitudinal assessment of disease evolution. Till now, to our best knowledge, there have not been previous articles reporting radiomics based on PD-L1 tracer for evaluation of PD-L1 expression and prediction of immunotherapy. Theoretically, phenotype information provided by imaging examination hides potential connection with PD-L1 expression level, and can be mathematically analyzed. We aimed to develop radiomics signatures of tumor immune environment and to assess the ability of these signatures to evaluate PD-L1 expression levels and predict responses in NSCLC treated with anti-PD-1/PD-L1 immunotherapy. .In this study, 89Zr-Df-KN035 will be employed for PD-L1 immuno-PET/CT imaging in humanized NSG mice bearing NSCLC xenografts. Tumor segmentation was performed to select essential primary lesions of NSCLC xenografts after PET/CT images acquisition. Features were extracted, then filtered with automatic relevance determination based on its relevance of PD-L1 expression status. Next, predictive models with features from the CT, the PET, and the PET/CT images will be built respectively, for assessing different PD-L1 expression status. For evaluation of the signatures' accuracy, the receiver operating characteristic as well as the corresponding area under the curve (AUC) was reckoned for each model. On this basis, the correlation between radiomics and the response to immunotherapies will be further evaluated by PET/CT imaging studies in humanized NSG mice bearing dual NSCLC xenografts. The left xenografts were used to assess PD-L1 expression status pre-immunotherapy while the right xenografts were used to explore the potential predictive value of radiomics features in immunotherapy. Then whether radiomics based on PD-L1 targeted PET/CT imaging superior to PD-L1 expression derived from biopsies for prediction of immunotherapy responses will also be analyzed..We aimed to develop and independently validate a radiomics-based biomarker of PD-L1 and explore the potential predictive value of radiomics features-derived approach in anti-PD-1/PD-L1 immunotherapy in NSCLC. It may be helpful in guiding immunotherapy in clinical practice and deserves further analysis.

依据抗PD-1/PD-L1免疫治疗在非小细胞肺癌（non-small-cell lung cancer，NSCLC）治疗中的应用进展，在本项目组前期构建的有临床转化应用前景的靶向PD-L1免疫PET影像探针89Zr-Df-KN035的基础上，进一步优化标记和显像条件，对多种荷NSCLC（PD-L1表达水平不同）的免疫人源化小鼠进行89Zr-Df-KN035免疫PET/CT显像，进而与抗PD-1/PD-L1免疫治疗后效果相对照，借助影像组学参数特征优于常规PET图像定量指标预测的优势，从免疫PET/CT影像数据中提炼出反映PD-L1表达程度的定量参数和特征，研究它们与肿瘤微环境免疫标志物、疗效之间的定量关系，以筛选NSCLC抗PD-1/PD-L1免疫治疗获益的患者，精准指导NSCLC的免疫治疗，并为进一步挖掘关键免疫分子PD-L1在肿瘤演进、免疫逃逸和治疗中的作用提供技术手段。

近年来，已开发出多种阻断程序性细胞死亡-1/程序性细胞死亡配体-1 (PD-1/PD-L1)通路的单克隆抗体（mAbs）用于肿瘤的免疫治疗。大量研究表明，PD-L1在肿瘤中的表达水平与抗PD-1/PD-L1治疗的疗效密切相关。正电子发射断层扫描（PET）作为一种非侵入性、高灵敏和可量化的成像方式，能在全身范围内对PD-L1的表达进行纵向和重复评估。目前，不同类型的放射性核素标记分子用于PD-L1表达的PET成像。其中，ADX_5322_A02是一种基于Adnectin的靶向PD-L1的基因工程蛋白，在PD-L1 PET成像中显示出良好的前景。本项目采用双功能螯合剂（BFC）马来酰亚胺-DODA-GA对68Ga标记的ADX_5322_A02进行了优化，制备的PET探针68Ga-NODAGA-A的放射化学纯度在99%以上，体外血清120min稳定性>99%，探针对PD-L1亲和力IC50值为8.92±0.90nM。注射68Ga-NODAGA-A后1小时，hPD-L1-B16F10肿瘤相较于野生型B16F10肿瘤（hPD-L1阴性）具有明显增高的放射性聚集，且能被过量的未标记前体（ADX_5322_A02）明显阻断，证实了该探针的PD-L1靶向特异性。体外生物分布研究与PET显像结果一致，注射探针1h和120min后肿瘤/肌肉比分别为7.28±0.63和9.15±1.10。荷瘤小鼠模型和NHPs的PET动态成像显示，探针主要通过肾脏快速清除，体内本底低。该探针还显示出安全的辐射剂量分布，在NHPs中的全身有效剂量为6.34E-03mSv/MBq。因此，68Ga-NODAGA-A是一种可行且安全的PD-L1 PET探针。进一步的研究表明，利用68Ga-NODAGA-A PET/CT显像可有效评估溶瘤病毒治疗前后肿瘤PD-L1表达的变化。此外，我们回顾性地纳入未治疗的非小细胞肺癌（NSCLC）患者，利用18F-FDG PET/CT影像组学评估肿瘤微环境免疫类型（TMITs）的分型。结果表明，基于FDG-PET/CT的影像组学特征和模型可预测NSCLC的免疫分型，尤其是TMIT-I肿瘤，为临床免疫治疗的选择提供了有前景的方法。