Background
Lung cancer is the leading cause of cancer death, partially owing to its extensive heterogeneity [
1,
2]. It has been proposed that this extensive heterogeneity results from successive clonal expansion and selection of the fittest clones influenced by genomic accumulation and somatic epigenetic alterations [
3‐
6]. However, tumor evolution may also be shaped by pressure from the immune system, which can prune the most immunogenic branches of the tumor [
7].
T cells play a crucial role in preventing cancer development through antigen-specific detection and destruction of malignant cells, though evolving tumors can eventually escape immune surveillance through a process termed immunoediting [
4,
8‐
11]. Few studies have addressed the impact of the T cell repertoire in shaping metastatic heterogeneity [
12‐
14] with most work to date evaluating longitudinal changes spanning multiple timepoints and therapies. Though these studies offer crucial insights, they do not allow the evaluation of intrinsic intertumor heterogeneity in absence of selective pressure from therapy. Furthermore, to date, the analysis of intertumor heterogeneity has been limited by an inability to concurrently obtain tissue from synchronous metastases unaltered by multiple prior lines of therapy [
15‐
17].
Here, we sought to study the relationship between genomic, epigenomic and T cell repertoire heterogeneity in a rare autopsy case from a 32-year-old female never-smoker with left lung primary late-stage lung adenocarcinoma (LUAD) with more than 20 synchronous metastases. We observed heterogeneous mutation, methylation, and T cell patterns across distinct metastases including a set of prevalent T cell clonotypes which were completely excluded from left-side thoracic tumors. Our work further highlights neoantigen methylation as a potential mechanism driving immunosuppression and some of the hurdles facing the treatment of late-stage lung cancer.
Discussion
Metastasis is an evolutionary process shaped by the dynamic interactions between tumor cells and host factors including immune surveillance [
48]. T cells play a pivotal role in mediating this process by recognizing antigens presented on MHC molecules at the surface of tumors and carrying out cytotoxic responses against tumor cells harboring their cognate antigens [
49]. Accordingly, much importance has been attributed to T cell infiltration in many solid tumors, with more T cells generally associated with a better prognosis [
39,
50,
51]. However, recent studies have highlighted the impact of intratumor heterogeneity (ITH) and bystander T cells [
52‐
54], and suggested that only ~10% of tumor-infiltrating lymphocytes are capable of recognizing antigens presented by the tumor they have infiltrated [
55], prompting deeper investigations into the T cell repertoire. Our understanding of the role of genomic and immune heterogeneity in lung cancer has evolved in recent years, thanks to investigations by our group and others into differences between regions of individual tumors, synchronous metastases and between primary and metastatic tumors [
44,
56‐
59] highlighting potential spatial and temporal factors influencing clinical outcomes [
60,
61]. Here, we assess the characteristics of the T cell repertoire in a treatment-naïve non-smoking patient with synchronous lung metastases and depict the interplay between the primary tumor and synchronous metastases [
11,
44,
62], revealing extensive immunogenomic intertumor heterogeneity across primary and metastatic sites.
In our study, clonal
TP53 mutations were detectable in all tumors, suggestive of an early genomic event, in line with prior reports [
24,
63]. Interestingly, a higher overlap in somatic mutations was observed across proximal tumors suggesting they are more genetically similar, potentially due to metastatic seeding from the primary tumor [
64,
65]. Though our study focused on a single patient, the overlap in mutational burden observed between synchronous metastases is in line with previous reports in lung [
59], melanoma [
66], kidney [
67] and colon cancer [
68]. Considering the role of somatic mutations in triggering T cell responses through the generation of neoantigens, this overlap suggests these somatic mutations may serve as potential therapeutic targets for vaccination or T cell engineering through targeting of unifying antigens present across all synchronous tumors. This is supported by the modest but significant correlation between shared mutations and shared TCRs though additional studies are needed to confirm these hypotheses.
We observed lower T cell repertoire heterogeneity across synchronous metastases in our study than in our prior work assessing multi-region ITH of localized LUAD [
44]. This difference could highlight the distinct resistance mechanisms at play in accelerated progression in our study versus more gradual progression in early-stage LUAD which may have allowed for divergent genomic evolution and immune editing over years. This is reinforced by the absolute restriction of certain T cell clonotypes to metastases surrounding the primary tumor, which could be reflective of the distinct antigenic environments established in distal tumors. Unfortunately, our lack of deep immune phenotyping data precludes our ability to further investigate the role the distinct immune microenvironments, including chemokine gradients and receptors, which may have played a role in establishing these vastly distinct T cell microenvironments. However, the presence of shared T cell clonotypes could also be indicative of common responses against unifying antigens displayed across synchronous metastases.
Aberrant methylation has been reported to be involved in tumorigenesis in a variety of cancers and its impact on anti-tumor immune surveillance are very complex [
69‐
72]. DNA methylation could affect immune surveillance directly by regulating the expression of potential neoantigens and/or immune-related genes thereby suppressing anti-tumor immune responses [
73] or indirectly via modifying chromosomal vulnerability for mutations and copy number alterations (CNAs), both of which are well known to influence the tumor immune microenvironment [
74‐
76]. A high level of hypomethylation could also lead to a high CNA burden which has been associated with a cold tumor immune microenvironment therefore facilitating immune evasion. Conversely, a high level of hypomethylation is associated with an increased mutational burden, which could increase tumor immunogenicity [
74]. However, in the end, the selected cancer cell clones of late-metastases could escape immunosurveillance through the cumulative effects of these molecular aberrations. Recently, considerable progress has been reported in the identification of protein factors with a highly conserved DNA interaction surface, termed the methyl CpG-binding domain (MBD) [
77]. MBD proteins could couple DNA methylation to transcriptional repression therefore silencing tumor suppressor genes that are hypermethylated at their promoter CpG islands in cancer cells [
77]. MBD1 silences the IRF8 gene, which is a central mediator in the IFNγ/STAT1 signaling pathway governing the establishment of immune responses [
78]. Therefore, disrupting MBD1-mediated silencing could enhance immune surveillance. Interestingly, in the current study, the ratio of methylated neoantigen coding mutations was negatively associated with T-cell density, richness and clonality, even at an individual level. One could therefore hypothesize that therapeutic agents modulating methylation could potentially reprogram the immune microenvironment and could exhibit some potential in treating these tumors.
Our study does exhibit certain limitations, including its focus on a single patient. However, analysis of several synchronous tumor sites from a single patient with advanced disease in absence of heavy pre-treatment is rarely possible due to the lack of clinical indication. Unfortunately, deeper analysis of underlying mechanisms, immune cells and soluble factors influencing T cell trafficking and heterogeneity remain unclear due to the archival nature of these samples and will require further investigation. It is well known that DNA methylation is involved in differentiation and activation of naïve CD4
+ T cells [
79,
80]. DNA methylation could release gene suppression upon demethylation of promoter regions regulating differential Th1 and Th2 cytokine secretion and therefore interrupting CD4
+ T-cell polarization with effector phenotypes in these cells [
81,
82]. Interestingly, up to 27% of methylated genes in promoter regions are actively expressed in naïve CD4
+ T cells and associated with immune response and T cell differentiation [
83]. Moreover, a recent study by Bam M
et al [
84] found that the majority of differentially methylated regions (DMRs) in tumor CD4
+ T cells are located in the promoters followed by intergenic regions and introns. One could therefore assume that methylated genes in intergenic regions are also involved in naïve CD4
+ T cell polarization and differentiation into specific lineages. Unfortunately, we were unable to investigate the impact of DNA methylation on the immune system and how genes in promoters or intergenic regions regulate CD4
+ T cell differentiation and activation in our cohort due to a lack of metastasis-paired normal tissue controls.
Despite these limitations, our study provides important evidence of differential tumor-immune responses co-existing in metastases within the same individual, related not only to molecular alterations. As a result, our findings may also partially explain the challenge of treating late-stage lung cancer due to the heterogeneity of metastases. Additional genomic, transcriptomic and immune studies in patients with synchronous metastases could help shed light on these and other mechanisms at play and provide therapeutic insights into late-stage NSCLC. Lastly, our study demonstrates heterogeneous immunogenomic profiles across synchronous metastases and provides important evidence of differential tumor-immune responses co-existing in metastases within the same individual, partially resulting in the challenge of treating late-stage lung cancer. These results advocate for combination regimens with drugs exhibiting distinct mechanisms of action, e.g. concurrent chemotherapy with immunotherapy or anti-PD-1 plus anti-CTLA-4 [
85] to eradicate cancer cells displaying different immunogenomic features to improve survival of patients with synchronous lung cancer metastases [
66]. Ideally, biopsies of multiple sites should be considered when feasible to identify common neoantigens across different disease sites, which could be targeted therapeutically. Interestingly, a shared TP53 mutation was identified across all tumor lesions and the patient was found to express HLA-A*02:01. Recent work has revealed that TP53-derived neoantigens can be targeted using Bispecific T cell Engagers (BiTEs), highlighting a potentially effective therapeutic option for this patient [
86]. Under most circumstances, however, multi-site biopsy is not clinically safe or practical. Alternatively, canonical cancer gene alterations, which are often present in all cancer across difference sites [
87], are appropriate targets. Moreover, with the caveat of variable detection sensitivity, liquid biopsy is less impacted by tumor heterogeneity [
88,
89], and may therefore provide a holistic view of molecular features across different metastatic sites to complement molecular profiling from tumor biopsies.
Declarations
Competing interests
L.A.B. serves on advisory committees for AstraZeneca, AbbVie, GenMab, BergenBio, Pharma Mar SA, Sierra Oncology, Merck, Bristol Myers Squibb, Genentech, and Pfizer and has research support from AbbVie, AstraZeneca, GenMab, Sierra Oncology and Tolero Pharmaceuticals. I.W. reports consulting or advisory roles for AstraZeneca/MedImmune, Bayer, Bristol-Myers Squibb, Genentech/Roche, GlaxoSmithKline, Guardant Health, HTG Molecular Diagnostics, Merck, MSD Oncology, OncoCyte, Jansen, Novartis, Flame Inc, and Pfizer; has received grants and personal fees from Genentech/Roche, Bristol Myers Squibb, AstraZeneca/MedImmune, HTG Molecular, Merck, and Guardant Health; has received personal fees from GlaxoSmithKline and Oncocyte, Daiichi-Sankyo, Roche, Astra Zeneca, Pfizer and Bayer; has received research funding to his institution from 4D Molecular Therapeutics, Adaptimmune, Adaptive Biotechnologies, Akoya Biosciences, Amgen, Bayer, EMD Serono, Genentech, Guardant Health, HTG Molecular Diagnostics, Iovance Biotherapeutics, Johnson & Johnson, Karus Therapeutics, MedImmune, Merck, Novartis, OncoPlex Diagnostics, Pfizer, Takeda, and Novartis. J.J.Z. reports research funding from Merck, Johnson and Johnson, and consultant fees from BMS, Johnson and Johnson, AstraZeneca, Geneplus, OrigMed and Innovent outside the submitted work. J.V.H. reports honorariums from AstraZeneca, Boehringer-Ingelheim, Catalyst, Genentech, GlaxoSmithKline, Guardant Health, Foundation medicine, Hengrui Therapeutics, Eli Lilly, Novartis, Spectrum, EMD Serono, Sanofi, Takeda, Mirati Therapeutics, BMS, BrightPath Biotherapeutics, Janssen Global Services, Nexus Health Systems, EMD Serono, Pneuma Respiratory, Kairos Venture Investments, Roche and Leads Biolabs. A.R. serves on the Scientific Advisory Board and has received honoraria from Adaptive Biotechnologies. A.M. is an employee and receives stock/has ownership interests in Guardant Health. The other authors declare no competing interests.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.