肺癌放疗和抗PD-1 / PD-L1组合：构建更好的转化研究平台

尽管免疫检查点阻断在为非小细胞肺癌（NSCLC）的某些患者提供持久应答方面获得了无与伦比的成功，但大多数患者并无反应。 PD-L1肿瘤表达和预先存在的肿瘤T细胞浸润与抗PD-1 /抗PD-L1的临床结果改善相关。然而，PD-L1表达阴性的肿瘤患者也可能对治疗有反应。目前正在研究将其他治疗方式如放疗（RT）与免疫检查点抑制剂相结合的策略，作为提高对PD-1 / PD-L1抗体阻断反应率的手段。 RT诱导癌细胞中的免疫原性改变，可以适应性地上调肿瘤细胞PD-L1表达并且可以改善抗PD-1 /抗PD-L1疗法的功效。我们如何设计未来的NSCLC临床试验还取决于实现这些治疗组合的实际考虑因素，如RT剂量，分馏和野外体积，以及免疫检查点阻滞的时间安排。在这里，我们回顾了抗PD-1 /抗PD-L1抗性的原因，以及如何通过建立更好的转化研究平台来将RT与这些药物结合使用以增强其效果。

Despite the unheralded success of immune checkpoint blockade in delivering durable responses for some patients with nonsmall- cell lung cancer (NSCLC), the majority of patients do not respond. PD-L1 tumour expression and pre-existing tumour T-cell infiltration have been correlated with improved clinical outcomes to anti-PD-1/anti-PD-L1. However, patients with tumours that are negative for PD-L1 expression can also respond to treatment. Strategies to combine other treatment modalities like radiotherapy (RT) with immune checkpoint inhibitors are being investigated as means of improving the response rates to PD-1/PD-L1 antibody blockade. RT induces immunogenic changes in cancer cells, can adaptively upregulate tumour cell PD-L1 expression and can improve the efficacy of anti-PD-1/anti-PD-L1 therapy. How we design future clinical trials in NSCLC also depends on practical considerations of delivering these treatment combinations, such as RT dose, fractionation and field volume, as well as scheduling with immune checkpoint blockade. Here, we review reasons for resistance to anti-PD-1/anti-PD-L1 and how RT may be utilised in combination with these drugs to enhance their effect by building better translational research platforms.

引言：非小细胞肺癌（NSCLC）仍然是全球癌症死亡的主要原因。放疗（RT）单独或联合化疗在早期和局部晚期非小细胞肺癌的治疗中发挥重要作用，并且已知通过加强局部疾病控制来改善长期生存[1]。最近，免疫检查点抑制剂（ICI）抗程序性死亡-1（抗-PD-1）抗体派姆单抗和纳夫单抗在晚期NSCLC中获得约20％的客观响应率（ORR），否则用细胞毒性化学疗法治疗[ 2-6]。通过将PD-1阻断剂与T淋巴细胞相关抗原4（CTLA-4）阻断剂结合[7]，这些反应率可以增加至约40％，突出了需要开发ICI治疗的适当组合以改善其效力。 RT免疫修饰肿瘤微环境（TME），包括增强的抗原呈递和肿瘤程序性死亡配体-1（PD-L1）表达的上调[8,9]。 当RT与抗PD-1抗体联合应用时，临床前研究发现原发性和继发性肿瘤的持续反应[8-10]。最近，在局部晚期非小细胞肺癌患者未进行PD-L1选择的III期临床试验中，在确定性化放疗（CRT）后给予durvalumab（抗PD-L1抗体），与安慰剂组相比，无进展生存期（PFS） [11]。除了有限的全身毒性外，RT因此代表着与ICI结合的有吸引力的药物。

Introduction：Non-small-cell lung cancer (NSCLC) remains the leading cause of cancer mortality worldwide. Radiotherapy (RT) alone or in combination with chemotherapy plays a major role in the management of early and locally advanced NSCLC and is known to improve longterm survival by enhancing local disease control [1]. Recently, the immune checkpoint inhibitors (ICIs) anti-programmed death-1 (anti-PD-1) antibodies pembrolizumab and nivolumab have achieved objective response rates (ORRs) of around 20% in advanced NSCLC, which is otherwise treated with cytotoxic chemotherapies [2-6]. These response rates can be increased to around 40% by combining PD-1 blockade with T-lymphocyte-associated antigen 4 (CTLA-4) blockade [7], highlighting the need to develop appropriate combinations of therapy with ICIs to improve their efficacy. RT immunologically modifies the tumour microenvironment (TME), including enhanced antigen presentation and upregulation of tumour programmed death ligand-1 (PD-L1) expression [8, 9]. When RT is combined with anti-PD-1 antibody, durable responses have been seen in primary and secondary tumours in preclinical studies [8-10]. Recently, durvalumab (anti-PD-L1 antibody) administered following definitive chemoradiation (CRT) in a phase III trial of locally advanced NSCLC patients unselected for PD-L1 showed an 11-month improvement of progression-free survival (PFS) compared with placebo [11]. Along with its limited systemic toxicities, RT therefore represents an attractive agent to combine with ICIs.

为什么大多数患者不能对免疫检查点封锁做出反应？

肿瘤细胞PD-L1表达已被假定为预测性生物标志物[4,6,12,13]，然而，即使在表达更高水平PD-L1的患者中，应答率仍低于60％[6]。 此外，PD-L1低或PD-L1阴性肿瘤仍然可以显示出对抗PD-1 /抗PD-L1的一些反应[12,13]。在临床试验中使用的不同PD-L1免疫组织化学测定法（例如Dako对比Ventana测定法）导致关于PD-L1状态的相互矛盾的结果，使用不同的测试抗体克隆，方案，评分系统和阳性阈值截止值[14]。因此，迫切需要考虑和标准化TME中肿瘤和免疫细胞上PD-L1表达的异质性[15]。此外，PD-L2表达，PD-L1对T细胞PD-1亲和力的配体的影响尚不清楚[15]。因此，PD-L1本身可能不是一个可靠的预测性生物标志物，需要考虑肿瘤演变过程中的多种动态因素和治疗抗性的发展。

Why do the majority of patients fail to respond to immune checkpoint blockade?

Tumour cell PD-L1 expression has been postulated as a predictive biomarker [4, 6, 12, 13], however, response rates remain below 60% even in patients expressing higher levels of PD-L1 [6]. Furthermore, PD-L1-low or PD-L1-negative tumours can still show some response to anti-PD-1/anti-PD-L1 [12, 13]. Varying PD-L1 immunohistochemistry assays used within clinical trials (e.g. Dako versus Ventana assays) have led to conflicting results as regards to PD-L1 status, with use of different test antibody clones, protocols, scoring systems, and threshold cut-offs for positivity [14]. Thus, consideration and standardisation of the heterogeneity of PD-L1 expression on both tumours and immune cells in the TME is urgently required [15]. Furthermore, the impact of PD-L2 expression, a ligand that shares affinity for PD-1 on T cells with PD-L1, is less well understood [15]. Consequently, PD-L1 alone is probably not a reliable predictive biomarker, and multiple dynamic factors during tumour evolution and the development of treatment resistance need to be considered.

Blank等人开发的“癌症免疫图”框架。意味着对每种肿瘤反应都至关重要的是效应T细胞（T ^ eff）活性增加。投入这是几个动态生物标志物参数，可能有助于直接个性化的免疫治疗方法。这些包括T细胞浸润和功能，主要组织相容性复合体（MHC）表达状态，新抗原负荷，代谢状态和一般免疫状态因素（如淋巴细胞计数）[16]。对ICI的无反应者可以进一步简化为减少肿瘤特异性体细胞突变负荷，TME内抗原识别机制缺陷（MHC表达）和/或免疫抑制因子增加的患者（表1）。如果肿瘤表现出高突变负荷，但MHC下调或TME仍然抑制（理论患者1和2，表1），则不会发生有效的免疫应答。同样，如果免疫细胞群体，检查点分子和抗原识别机制存在正确的平衡，但肿瘤起始时具有较低的新抗原负荷，治疗耐药性很可能（理论患者3，表1）。 简化ICI反应的障碍可以提供一个框架，用于治疗性地操纵这些因素。

The 'cancer immunogram' framework developed by Blank et al. implies that critical to every tumour response remains increased effector T-cell (T^eff) activity. Feeding into this are several dynamic biomarker parameters that may help direct personalised immunotherapeutic approaches. These include T-cell infiltrate and functionality, major histocompatibility complex (MHC) expression status, neoantigen burden, metabolic status, and general immune status factors (such as lymphocyte count) [16]. Non-responders to ICIs can be further simplified into patients who have reduced tumour-specific somatic mutational burden, defects in antigen recognition machinery (MHC expression) and/or increased immunosuppressive factors within the TME (Table 1). If a tumour exhibits high mutational burden but MHC is downregulated or the TME remains suppressive (theoretical patients 1 and 2, Table 1), an effective immune response will not ensue. Equally, if the right balance of immune cell populations, checkpoint molecules and antigen recognition machinery are present, but a tumour has a low neoantigen burden to begin with, treatment resistance is likely (theoretical patient 3, Table 1). Simplification of barriers to ICI response can provide a framework with which to therapeutically manipulate these factors.