编者按：由中国临床肿瘤学会和北京市希思科临床肿瘤学研究基金会共同主办的第27届全国临床肿瘤学大会暨2024年CSCO学术年会于2024年9月25日至29日在厦门举行。本次大会中，多个肺癌专题论坛汇总了国内外肺癌领域的前沿研究与进展。大会期间，《肿瘤瞭望》特邀香港中文大学医学部莫树锦教授接受采访，分享生物标志物检测在肺癌领域临床实践现状及未来展望。

 

莫树锦教授：这是一个重要的问题。在过去的二十年中，肺癌治疗的发展是革命性的，这是因为生物标志物可以指导我们为患者选择正确的药物。从肺癌驱动基因EGFR突变体的发现开始，后续临床实践中对肺腺癌患者均需进行EGFR突变检测，由此改变了肺癌的治疗格局。此外，自2014年ALK抑制剂问世，临床对ALK基因检测提出了更多需求。目前，肺癌领域用于基因检测的基因数量正在不断增加，共有约10种不同的遗传学异常的基因表型。多数基于驱动基因开发的治疗药物已纳入一线治疗，即使针对于罕见突变，亦在相关一线治疗临床研究数据中显示出临床优势，如RET选择性抑制剂、用于治疗EGFR外显子20插入突变晚期非小细胞肺癌（NSCLC）治疗的EGFR-MET双特异性抗体埃万妥单抗（Amivantamab）等。综上，我认为在患者接受治疗前，对其进行全基因检测，可明确患者病情，为患者临床策略制定提供指导。

 

Dr Tony Mok:This is an important question.The development of lung cancer therapy over the last twenty years is revolutionary，and it is because of biomarkers making it assessable for us to select the right drug for the patient.We start off with the story of the EGFR mutation that we all know and that we all test for now in patients with adenocarcinoma.That changed the paradigm.In addition to that，with the availability of an ALK inhibitor since 2014，we now also test for ALK.The number of genes we are testing for upfront is increasing.Apart from those，we now have a total of around ten different genetic abnormalities that we test for.Most of this already have some first-line information，even for uncommon mutations，like RET.We now have first-line data demonstrating improvement.We also have EGFR exon 20，an uncommon mutation，but then first-line chemotherapy and amivantamab is actually very useful.I think it has become standard practice that we do a whole panel including the ten or eleven genes so we can identify the patient from the beginning.

 

莫树锦教授：这实际上是一个非常重要的研究方向。EGFR突变的发现开启了肺癌靶向治疗时代，随后免疫治疗的出现为肺癌患者长期生存带来新希望。紧随其后的抗体药物偶联药物（ADC）的开发已经开展了大量工作，涉及多个靶点，如HER2、HER3、TROP-2、MET等。但现有研究数据并不像预期那样令人鼓舞，尽管一些数据表现积极，如针对TROP-2靶点的ADC使患者无进展生存期（PFS）略有改善，但远远不够。因此，不禁发问：未来该如何发现有临床价值的生物标志物？

 

在今年的世界肺癌大会（WCLC）上，Marina Chiara Garassino教授团队发表了一项定量连续评分计算的TROP2标准化膜比率预测TROPION-Lung 01的临床结果的相关研究，证实了TROP2标准化膜比率的阳性是预测非小细胞肺癌（NSCLC）患者通过datopotamab deruxtecan（dato-DXd）治疗获得更佳总体缓解率（ORR）和无进展生存（PFS）的预测性标志物。dato-DXd是一款靶向Trop2的抗体偶联药物。通过定量连续评分确定的TROP2膜标准化比率（QCS-NMR）测量的是肿瘤细胞膜中TROP2表达与细胞质的比率。QCS-NMR阳性（+）要求样本中必须至少有75%的肿瘤细胞具有≤0.5585的TROP2膜标准化比率。QCS-NMR阳性（+）则代表使用dato-DXd可最终获得较高的ORR和PFS。我认为这是未来生物标志物探索的方向之一，即使用人工智能数字化技术开发生物标志物。（https://mp.weixin.qq.com/s/gnvqraif8M00lV8dj7hpTw）

 

此外，香港中文大学医学中心开展的相关研究是通过3D量化技术，通过新技术，可以量化三维肿瘤组织块中的蛋白表达，而不是在二维切片上。从三维角度量化肿瘤细胞中蛋白质的表达，亦是未来重要的探索方向。

 

Dr Tony Mok:That is actually a very important research direction.After EGFR，after targeted therapy，we now actually have immunotherapy.After immunotherapy，we now have antibody-drug conjugates(ADCs).A lot of work has been dedicated to the development of antibody-drug conjugates.We have multiple targets-HER2，HER3，TROP-2，MET-that we have ADCs for.But the current data are not as encouraging as we would like them to be.We have some positive responses where we get a better response rate，and slight improvements in progression-free survival with the TROP-2 ADC.However，the significance of the improvement is not high enough.We have to ask the question，how else can we develop biomarkers?At the WCLC this year，they actually reported on the use of new method called the quantitative continuous score，which is a digitalized method of looking at the protein expression of the cell membrane versus the cytoplasm expression.It is a ratio of>75%of the cell that makes this positive.If they are positive，it is quantified that using an ADC with TROP-2 will end up with a high progression-free survival.I think this is one direction-using AI digital technology to look at this protein expression.Other research that we are doing in our center at the Chinese University of Hong Kong is looking at 3D ixl expression.With new technology，we can quantify protein expression not on a two-dimensional slide，but as a tumor block expression in three-dimensions.A lot of work still needs to be done.The quantification in three-dimensions is one important direction.

莫树锦教授：首先，事实证明，基于基因组学的生物标志物探索是成功的研究方向，因为药物的作用有赖于异常基因及其产物。其次，针对ADCs的开发，我们显然需要聚焦于蛋白质组学，其中如何计算蛋白质是很重要的。值得一提的是，肿瘤疫苗、CAR-T及其他细胞疗法创新疗法均利用人体自身的免疫系统来攻击肿瘤细胞。这种情况将更为复杂，因为免疫系统与肿瘤细胞相互作用的同时也与正常细胞相互作用，如何区分二者是十分困难的。我认为此时可以从一个基本概念入手，即HLA分型。由于HLA的功能是呈递新生抗原到细胞表面，供T细胞识别，因此，HLA分型是T细胞对抗原的最基本呈现。当然不同种族之间的HLA分型是不同的，因此对于HLA分型及其与免疫治疗的关系未来仍需深入探索。

 

Dr Tony Mok:It has been proven to be successful with the genomic-based biomarkers，because the drug is targeting the abnormal gene and its product.But a future platform is actually lies beyond molecular genomics.With ADCs，we obviously need to look at the proteins，and how we do the calculation of proteins is important.But the next thing is that we also have other platforms，including cancer vaccines，CAR T-cells or other cell therapies，all of which utilize the immune system.That becomes a lot more complex，because different immune systems interact with each other as well as interacting with normal cells.How do we differentiate between this interaction of the immune system with the cancer cells versus the immune system with normal cells is very difficult.But I think we can start with one basic concept，which is HLA typing.HLA typing is actually the most basic presentation of the antigen to the T-cell.There is different HLA typing between one race versus another race-the answers are totally different-so we need to learn a little bit more，at least about this basic HLA typing and its relationship to immune-related therapy.

 

莫树锦教授：这个问题很广泛，涵盖了诊断和治疗两个方面。首先，众所周知，肺癌早期检测至关重要，因为早期肺癌患者可选择进行根治性手术治疗，极大提升了患者生存率及生活质量，而大多数肺癌患者确诊时已处于晚期，无法手术切除。我们已经确定低剂量CT扫描（LDCT）是肺癌早期检测最有效的工具之一，然而其检测费用高且需借助大型设备操作，所以我认为可以将低剂量CT扫描与基于血浆的生物标志物结合使用，如基于血浆的DNA基因组测试，但之前类似测试的敏感性不高。现阶段基于cfDNA多种特征研发的片段组学技术已成为肿瘤早筛的新方向。一些早期研究已证实该技术是有前景的，对早期肺癌监测具有更高的敏感性，我认为，未来如何将LDCT与片段组学技术将结合是一个有趣的探索方向。

 

其次，预后层面。现阶段已经有很多影响预后因素被发现，并且有许多基因可用于预后预测，但这些对临床疾病管理得帮助并不大，主要原因在于预后基因不一定具有预测性。因此，对我来说，临床相关的生物标志物应具有预测性，如提到的基因组生物标志物和一些正在使用的免疫生物标志物。

 

Dr Tony Mok:The question is quite broad，covering treatment and diagnosis.Maybe I will speak about early detection first.As you probably know，early detection is important because curative surgery has to be done in early stage disease.Most lung cancer patients present at a later stage when surgical resection is not possible.We have established a low-dose CT scan as one of the most effective tools for the early detection of lung cancer.However，this tool is still quite expensive，and requires large equipment to perform，so I think we have to correlate use of this potential tool with plasma-based biomarkers.One possibility is using a plasma-based DNA genomic test for early detection.Previously，there has been a similar test available，but sensitivity is not high.For the future，there is novel technology called fragmentomics，which looks at fragments of DNA from which we can calculate the methylation.This technology has been demonstrated in a number of early studies to be promising with a higher sensitivity for the detection of early stage lung cancer.In the future，how we use the low-dose CT scan and fragmentomics together will be one interesting development.That is for early detection.Then，for prognosis，personally，I would say that there are so many prognostic factors and so many genes can be prognostic，but it doesn’t really help our clinical management as much.A prognostic gene does not necessarily mean it is predictive.So to me，a clinically relevant biomarker should be a predictive one，just like the genomic biomarkers I have mentioned，plus some of the immune biomarkers that we have actually been using.