Clonal expansion of shared T cell receptors reveals the existence of immune commonality among different lesions of synchronous multiple primary lung cancer

Lung cancer remains the leading cause of cancer-associated deaths worldwide [1]. With the widespread application of computed tomography in lung cancer screening, the detection rate of synchronous multiple primary lung cancer (MPLC) is continually increasing in clinical practice [2]. The underlying mechanisms associated with the initiation and development of MPLC remain unknown. Compared to solitary lung cancer nodule (SN), MPLC has posed critical challenges to clinicians in the fields of diagnosis and treatment [3].

The introduction of next-generation sequencing (NGS) has greatly broadened our understanding of MPLC, particularly in the differential diagnosis between MPLC and intrapulmonary metastasis [4]. Genomic analyses revealed substantial heterogeneity in the mutational landscape between different lesions of the same patient with MPLC, while metastatic lesions always harbored the same driver mutations compared to the primary tumor [5]. Many investigators consider that the occurrence of MPLC is a consequence of multiple independent and parallel evolutionary events [6]. However, given the shared germline mutations and environmental exposure, there might be an unexplored link among different lesions of the same patient with MPLC.

The latest research reveals that air pollution promotes lung cancer by acting on healthy lung cells with pre-existing oncogenic mutations, indicating genomic alterations and environmental exposure act synergistically in promoting tumorigenesis [7]. In addition, the proliferation and metastasis of tumor are shaped by the interaction between tumor cells and host immune surveillance. T cells can recognize tumor-associated antigens via the T cell receptor (TCR) and in turn play an important role in antitumor immunity [8]. Several studies have revealed that TCR sequencing could provide valuable information about the diagnosis, prognosis, and prediction of treatment response in lung cancer [9, 10]. Therefore, systematically assessing genomic alterations and TCR repertoire of MPLC and comparing them with those of SN will further deepen our understanding of MPLC.

In the current study, targeted NGS using a 1,021-gene panel was performed to assess the mutational landscape and tumor mutation burden (TMB). In parallel, the activation status of antitumor immune response was evaluated using programmed cell death-ligand 1 (PD-L1) expression and TCR repertoire. Comprehensive comparisons across genomic and immune features between MPLC and SN were performed. Furthermore, the relationship between different lesions of the same patient with MPLC was explored.

We aim for this work to broaden our knowledge of MPLC, particularly focusing on its immunological features, and attract more researchers to actively delve into the research of this increasingly prevalent type of lung cancer.

In recent years, the incidence of MPLC is significantly increased due to the widespread implementation of lung cancer screening [2]. Despite the substantial advances in surgery and radiotherapy, removing all of the lesions may not be feasible for all patients, especially for those with limited pulmonary function or a large number of lung cancer lesions [20, 21]. Consequently, the management of MPLC is much more complex than that of SN and remains an unmet medical need. The pathogenesis of MPLC is currently not well understood. Potentially susceptible germline mutations may serve as early events contributing to the development of a small subset of MPLC, yet further in vivo experimental validation is needed to verify the hypothesis [12]. More extensive research, particularly comparative analyses with SN, is indispensable for advancing our understanding of MPLC. In this study, mutational landscapes and immunologic characterizations of MPLC and SN were comprehensively assessed and compared using various integrative approaches.

The histopathologic and mutational heterogeneity among multiple lesions of the same patient with MPLC observed in our research confirmed their nonmetastatic nature. When it comes to the mutational landscape, one recent study with a large sample size suggested that EGFR mutation was more commonly mutated in SN cases, while BRAF mutation was more frequently observed in MPLC cases [14]. Similar differences in the mutation rate between MPLC and SN were found in our study, indicating that distinct mutational patterns did exist in MPLC. Notably, differentially mutated genes were exactly located in the upstream and downstream of RTK/RAS signaling pathway [17]. Therefore, MPLC might not be a simple combination of multiple solitary lung cancer lesions due to the influence of identical germline and environmental background. The specific causes and clinical significance of this finding have not been clearly elucidated. Recent studies have indicated that the genetic heterogeneity of MPLC was compatible with the convergent evolution, and only a limiting number of signaling pathways were mutated [22, 23]. However, evolutionary tree reconstruction in another study reached the opposite conclusion indicating that divergent rather than convergent evolution at the levels of both individual genes and functional pathways [24]. Overall, most of the previous research only included gene mutation information and relatively small sample sizes, making it premature to draw any definitive conclusion.

In clinical practice, the approval of adjuvant targeted therapy after surgery in localized lung cancer affords us an opportunity to evaluate the efficacy of targeted therapy in residual lesions in patients with MPLC [25]. However, postoperative EGFR tyrosine kinase inhibitors only showed limited efficacy on residual GGNs after the surgery of major lesion(s) harboring EGFR mutation [26]. Although combination with a radiomics model that can predict EGFR mutation status could further improve treatment efficiency, the overall response rate was 53.8%, which happens to be consistent with the rate of EGFR mutation in East Asian populations [27]. The significant mutational heterogeneity across different lesions of MPLC in our study could account for the limited response rate to EGFR inhibitors. It is reasonable to infer that a significant reduction in diameter may only be observed in residual GGNs harboring sensitive EGFR mutations. A previous case report showed that the combination of osimertinib and alectinib resulted in disease stability in a patient diagnosed with MPLC harboring distinct EGFR and RET alterations [28]. However, it should be noted that the available evidence is insufficient to support the use of routine targeted therapy in MPLC.

In recent years, immunotherapy has reshaped the therapeutic landscape for patients with lung cancer. Nevertheless, whether this strategy can also be applied to MPLC remains unexplored. In our study, the TMB and PD-L1 expression were relatively low both in MPLC and SN, in accordance with the high prevalence of early-stage lung cancer in our cohort. Previous studies have shown that MPLC often exhibited low TMB, low PD-L1 expression, and heterogeneous immune cell infiltration [23, 29]. It is noteworthy that disparities in immunological features have also been documented between MPLC and SN. The PD-L1 expression and the density CD8⁺ tumor-infiltrating lymphocytes of MPLC were significantly lower than those of traditional lung cancer [12]. Additionally, more immune-related biological processes and pathways, such as B cell activation and the IL-17 signaling pathway, were identified in the MPLC cohort[14]. Furthermore, high mutation rate of EGFR in East Asian populations together with the inert tumor immune microenvironment in lesions presenting as pGGN collectively implied a heterogeneous and limited response to immunotherapy in patients with MPLC [30]. Several small-scale clinical studies have revealed that no obvious reduction in diameter was observed after receiving immunotherapy in most high-risk synchronous GGNs [31, 32]. Moreover, the heterogeneous drug response to neoadjuvant immunotherapy in different lesions of MPLC has been reported in a patient presenting with one solid nodule and two subsolid nodule [33]. Multi-omics analyses indicated the proportions of tissue-resident memory T cells and CD8⁺ T cells varied greatly between responding and non-responding lesions and impaired T lymphocyte immunity may lead to the inferior response to immunotherapy [33]. Altogether, perioperative immunotherapy may be not suitable for patients with MPLC due to the inert tumor immune microenvironment.

Given the crucial role of T cells in mediating the antitumor response, TCR sequencing could provide us with an opportunity to comprehensively assess the activation status of immune system. The higher TMB exhibits a trend associated with increased TCR clonality in both cohorts, supporting the critical role of somatic mutations in boosting tumor immunogenicity and eliciting T cell responses. The similar positive correlation was observed in American smokers, implying that the association between TMB and TCR clonality may not be influenced by ethnicity or tobacco exposure [18]. Besides the previously mentioned heterogeneity in the mutational landscape, considerable intrapatient and interpatient heterogeneity were also observed in TCR diversity and expansion [34]. Early carcinogenesis of lung cancer has been confirmed to be a gradual process shaped by host immune surveillance [35]. TCR repertoire displayed a gradual increase in diversity and a progressive decrease in expansion along with the progression of invasiveness from preneoplasia to invasive lung adenocarcinoma [36]. In this study, similar results were present in the MPLC cohort but not in the SN cohort. Furthermore, invasive lesions of MPLC harbored more diverse T cell infiltration than that of SN. The interaction between tumor and the host’s immune system is crucial for the control of tumor growth and spread. Therefore, as the tumor progresses from preneoplasia to IAC, the immune system is believed to gradually lose its ability to kill tumor cells and acquire suppressive or ineffective subsets of T cells [37]. Several studies have shown that TCR clonality was positively correlated with CD8⁺ T cells but negatively correlated with CD4⁺ T cells [36, 38]. Meanwhile, Dejima et al. found TCR clonality was not associated with T cell density in their study, suggesting that lower T cell clonality in the invasive lesions of MPLC may reflect reduced T cell expansion rather than T cell exclusion [36]. The most likely reason is that although the immune system recruited a greater number of T cells, the majority of them were bystander T cells and did not exert a suppressive effect on the tumor [39, 40]. Taken together, MPLC may represent a distinct subtype of lung cancer characterized by unique pathogenesis and immune response patterns, rather than simply being a combination of multiple solitary lung cancer nodules. Notably, differences in TCR repertoire between the two cohorts could be masked by the inclusion of MIA and lung adenocarcinoma precursor lesions. One possible reason is that lesions at those stages have a limited capacity to invade through the basement membrane and release the tumor antigens into the immune microenvironment [41].

Tumor-specific neoantigens derived from somatic mutations could induce T cell activation and provide a critical link between tumor and adaptive immune system [42]. Therefore, the marked heterogeneity of TCR repertoire in different lesions of MPLC could be the result of the interaction between diverse neoantigens in each lesion and shared immune contexture. The impact of shared environmental exposure and immune contexture on the occurrence and development of multiple lesions of the same patient may make them share a degree of similarity in biological features. Indeed, unlike the remarkable heterogeneity in the mutational landscape, it was recently shown that the DNA methylation patterns of multiple lesions of the same patient were very similar to each other [14]. In this study, lesions of the same patient always share a certain proportion of TCR clones. More importantly, significant clonal expansion could be observed in all 100% shared TCR clones, indicating that shared TCR clones might reflect the immune commonality among lesions of the same patient and play a critical role in the antitumor immune response [34].

The interaction between tumor cells and host immune surveillance is a dynamic process, and TCR on the surface of T cells specifically recognizes the neoantigens presented by major histocompatibility complex molecules is the first step [41]. Previous studies have shown that the immune system could exert evolutionary pressure through the process of immunoediting to eliminate highly immunogenic tumor clones, and ultimately leads to the outgrowth of the fittest clones [35, 43]. Notably, the vast majority of lesions in the MPLC cohort are at the early stages of lung cancer manifesting as pGGNs, which have been widely considered as indolent, asymptomatic, slow-growing tumors and can remain clinically silent for years [44]. Therefore, the tumor microenvironment at this time is still in the equilibrium phase, where the adaptive immune system cannot eliminate tumor cells, but keeps them in a state of immune-mediated tumor dormancy [30, 45]. Similar to the selective expansion of dominant tumor clones, T cell populations may also experience evolutionary selection pressure and eventually lead to the selective survival and clonal expansion of neoantigen-specific T cells to maintain immune equilibrium, following the Darwinian rule of natural selection [46]. One recent study has indicated that universally present neoantigens across multi-regions of the same tumor should be targeted to maximize the potential for uniform T cell responses due to the presence of neoantigen intratumor heterogeneity [47, 48]. In light of these findings, the identification of shared TCR clones that were significantly expanded across multiple lesions of the same patient may have important implications for the development of therapeutic approaches targeting all lesions simultaneously to solve the treatment dilemma of MPLC. Preclinical work has suggested that adoptive therapy with neoantigen-specific TCR-transduced T cells could significantly inhibit tumor growth in the mice xenograft model [49]. Tebentafusp, a T cell receptor-bispecific molecule has been confirmed to confer a long-term survival benefit and been approved for previously untreated patients with metastatic uveal melanoma [49]. Taken together, our findings are important for deepening the understanding of MPLC. We hope that this study could inspire an increasing number of researchers to focus on and actively engage in scientific research on MPLC to move the field forward.

To the best of our knowledge, this is the first systemic study to reveal the existence of immune commonality among different lesions of the same patient with MPLC via TCR sequencing. However, there are still some deficiencies in our research. First, the retrospective and single-center nature may inevitably introduce some selection bias in this study, and further validation of these results in prospective multi-center cohorts is warranted. Second, the scarcity and small sizes of lesions hindered more comprehensive multi-omics analyses. Moreover, tumor-specific neoantigens responsible for TCR clonal expansion and the correspondence between neoantigens and tumor-specific T cells should be further investigated. More in-depth studies, especially functional studies using humanized animal models are warranted to elucidate the detailed mechanisms.

In conclusion, significant differences in the mutational landscape, activation status of oncogenic signaling pathways, and TCR repertoire between MPLC and SN collectively implied that MPLC might not be a simple combination of multiple solitary lung cancer. More importantly, the significant clonal expansion of shared TCR clonotypes demonstrated the existence of immune commonality among different lesions of the same patient and provided a new idea for resolving the treatment dilemma of MPLC.