Article Text
PDF
Abstract
Neoadjuvant chemo-immunotherapy has significantly improved the treatment landscape for patients with ALK/EGFR wt resectable non-small cell lung cancer (NSCLC), offering novel opportunities for translational and clinical investigations. By leveraging deep molecular profiling through spatial transcriptomics integrated with single-cell RNA-sequencing from public atlases, and serum proteomic profiling, we report a case of a patient with resectable NSCLC and a solitary synchronous brain metastasis showing a complete pathologic response after chemo-immunotherapy, and a durable event-free survival. The deep profiling of the tumor immune microenvironment of both pre-treatment brain metastasis and post-treatment primary lung tumor tissues unveiled a key role of B lymphocytes at different maturation state, spanning from naïve to plasma cells, in mediating tumor eradication. Notably, the formation of several tertiary lymphoid structures in the regression bed of post-treatment primary tumor was observed, suggesting the in situ generation of high-affinity antibody and specific immune memory response. This multimodal approach paves the way for the discovery of novel biomarkers at both tissue and systemic levels, fostering improved patient stratification and guiding clinical decisions on post-surgical treatment escalation or de-escalation.
- Immune Checkpoint Inhibitor
- Neoadjuvant
- Tumor microenvironment - TME
- B cell
- Lung Cancer
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Introduction
In recent years, the neoadjuvant setting for resectable non-small cell lung cancer (NSCLC) has been rapidly adopted into clinical practice, with emerging studies on combination treatments showing significant promise.1 2 Mechanistically, neoadjuvant immunotherapy primes the patients’ immune system enhancing its activation against tumor cells.3 Despite these advancements, critical challenges remain with molecular mechanisms underlying response still to be defined, and reliable biomarkers for short- and long-term outcomes lacking. Although complete pathologic response (pCR) is considered a strong prognostic factor, it is still unsatisfactory for defining treatment escalation or de-escalation after surgery.4 5 Understanding the tumor immune microenvironment (TIME) composition and spatial organization is crucial to maximizing the benefits of these therapies.
Herein, we present the case of a patient with resectable NSCLC and a solitary synchronous brain metastasis treated, after the metastasis resection, with an induction of three cycles of chemo-immunotherapy and then lobectomy followed by maintenance immunotherapy. By integrating cutting-edge high-throughput technologies, including spatial transcriptomics and plasma proteomics, we explored the TIME in pre-treatment brain metastasis and in post-treatment primary lung tumor tissue to highlight molecular determinants of the pathological response (pR) and to identify immune-related local and circulating signatures.
Results
An adult patient with a history of heavy smoking was admitted to our institution following the surgical resection of a solitary brain metastasis. Histological and molecular analyses revealed that the brain lesion was a lung adenocarcinoma (ADC) (40% Programmed Death-Ligand 1, PD-L1) with no alterations in ALK, EGFR, or ROS1 genes. Radiological assessments confirmed the presence of a primary lesion in the right upper lobe and the patient was staged as cT2bcN0pM1a. After three cycles of chemo-immunotherapy (cisplatin plus pembrolizumab), the patient underwent lobectomy with the achievement of pCR (online supplemental figure S1A). An adjuvant single agent pembrolizumab for 18 cycles was started and, after 3-year, the patient remains disease-free. The integration of pR with immune-related pathologic response criteria (irPRC)6 7 showed (1) accumulation of immune cells into the regression bed of post-treatment lung tissue with T and B cells organized into tertiary lymphoid structures (TLS), (2) accumulation of cholesterol clefts, (3) features of proliferative fibrosis, (4) macrophages with phagocytic functions (online supplemental figure S1B). To create an in-depth profile of TIME composition and functionality of the pre-treatment resected brain metastasis and the post-treatment resected lung tumor, we performed spatial transcriptomics. Brain metastasis showed high expression of the ADC marker thyroid transcription factor 1, evenly distributed across the tissue in accordance with the pathologic diagnosis (online supplemental figure S1C). Distribution of major stromal and immune cell populations was assessed by microenvironment cell populations (MCP)-counter. Specifically, Visium analysis mapped few T cells, but sparsely distributed, in the tumor tissue, while B cells were exclusively present in hot spots located close to the blood vessels (online supplemental figure S1C). Neutrophils, central-resident microglia, monocyte-derived macrophages and astrocytes were also mapped within the brain tissue (online supplemental figure S1C). To gain deeper insights into the molecular features of the different cell populations, we performed unsupervised clustering analysis. This revealed five spatially and functionally distinct clusters, each characterized by a unique gene expression profile, as shown in the heat map with the top genes of each cluster (figure 1A and online supplemental table 1). Specifically, cluster 0 is solely defined by the selective expression of major facilitator superfamily domain containing 11 (MFSD11), a gene that remains still poorly characterized, while cluster 1, predominantly located at the periphery of the specimen, encompasses ADC cells characterized by the overexpression of genes related to high metastatic potential such as E74 Like ETS transcription factor 3 (ELF3), amphiregulin (AREG), transcriptional and immune response regulator (TCIM) and activating transcription factor 3 (ATF3).8 Conversely, cluster 2, although in close proximity with tumor cells, accounts for activated stromal cells marked by the overexpression of extracellular matrix (ECM)-related transcripts (ACTA2, TAGLN, PDGFRB). The overexpression of ECM-related transcripts such as COL1A1, COL3A1, COL6A1, COL6A3, FN1, SPARC, and LUM, highlights an active reorganized ECM. Cluster 2 also includes immune cells as shown by transcripts related to antigen presentation and immune recognition (C1QC, HLA-DRB1), as well as chemokine genes involved in immune cell recruitment, including CXCL12 and lymphotoxin-β (LTB). Transcripts suggestive of antibody-secreting plasma cells, such as IGHA1, IGKC, and IGHG1, were also detected, highlighting the presence of active humoral immune response. Cluster 3 mostly comprises cancer cells with metastatic adaptation while maintaining lung lineage markers. In contrast, cluster 4 suggests the presence of an active inflammatory response with an enrichment of localized neutrophils and transcripts related to antigen presentation (CD74, HLA-DRA, HLA-DRB1). Focusing on immune infiltration, we found that clusters 4 and 2 exhibited the highest enrichment of monocyte/macrophage populations, FCGR3A (CD16) immune cells, mostly natural killer (NK) cells and neutrophils, and CD4+T cells, with cluster 3 showing a more moderate infiltration. Notably, in accordance with cell deconvolution by MCP-counter and clustering analysis, we identified CD138+plasma cell (marked by MZB1) only within cluster 2, In contrast, clusters 0 and 1 were largely devoid of immune cells (figure 1A). Gene Ontology (GO) analysis displayed the main molecular programs upregulated in each cluster, including signatures of ECM organization, of immune cell activation and metabolic reprogramming in tumor cells (online supplemental figure S1D).
Supplemental material
Supplemental material
Spatial profiling of pre-treatment brain metastasis and post-treatment lung primary tumor tissue by 10x Visium unveils a putative role of B cells in mediating response to treatment. (A) Unsupervised clustering analysis of the patient with non-small cell lung cancer brain metastasis reveals five different clusters (0–4). Heatmap shows the top marker genes for each identified cluster, mainly related to tumor cell metabolism and invasion. Dot plot highlights the presence of the main immune populations within each identified cluster. (B) Unsupervised clustering analysis of the resected post-treatment lung tissue identifies three clusters (0–2), mainly related to B-cell activation, plasma cells presence, antibody secretion and inflammatory fibrosis. Heatmap shows the top marker genes for each identified cluster. Dot plot highlights the presence of the main immune populations within each identified cluster.
The Visium analysis of lung tissue resected after chemo-immunotherapy revealed the absence of residual tumor cells in accordance with pCR (online supplemental figure S2A). Differently from brain metastasis, MCP-counter analysis showed, in this region, an abundance of cancer-associated fibroblasts (CAFs) and immune cells, mostly B cells, but T cells, neutrophils and macrophages, although scarce, were also present (online supplemental figure S2A). Unsupervised clustering analysis identified three clusters, spatially well-defined and exhibiting precise molecular profiles (figure 1B and online supplemental table 2). Notably, cluster 0 is enriched in transcripts related to active antibody production and B-cell/plasma cell infiltration with the presence of immunoglobulin genes (IGHG, IGHA, IGKC), along with the JCHAIN, marker of mucosal immunity with IGHA, suggesting the activation of a robust localized humoral response. Interestingly, this plasma-cell enriched cluster, when integrated with the whole tissue section revealed an area of intense tumor cell death and the presence of TLS as detected by immunohistochemistry (IHC) (figure 2A). Both clusters 1 and 2 comprise immune and stromal cells, with cluster 1 including mostly pro-inflammatory macrophages, as suggested by the overexpression of SPP1, APOE, MMP9, TYROBP, and HLA-DRA transcripts, and CAFs involved in ECM remodeling (FN1, TIMP1, TIMP3, and MMP9) likely shaping a TIME with features of proliferative fibrosis, one of the irPRC.6 7 Cluster 2 exhibits both prominent ECM remodeling features and the presence of genes related to immune activation (C3, EGR1, and IGHM) primarily driven by B cells. In parallel, we performed GO analysis related to each cluster, which highlighted the upregulation of molecular pathways related to cell death and antitumor immunity activation (online supplemental figure S2B). Focusing on immune cell infiltration, we found monocytes/macrophages, T cells, FCGR3A (CD16) immune cells, and plasma cells across all three clusters, while B cells, marked by MS4A1 (CD20), were detected only in clusters 0 and 2 (figure 1B).
Supplemental material
Prominent presence of tertiary lymphoid structures within the regression bed of post-treatment primary lung tumor tissue. (A) IHC of the lung tissue on the whole consecutive Visium slices shows an abundance of tertiary lymphoid structures (TLS) as marked by CD20 and CD3 staining. (B) IHC of the lung tissue on the whole consecutive Visium slices shows TLS at different maturation levels, as evidenced by CD21 and CD23 staining. IHC, immunohistochemistry.
To validate the key role of B cells in mediating the durable immune-driven patient response, we performed IHC analyses confirming the abundance of CD138+ plasma cells both inside and outside TLS as expected in agreement with immunoglobulin (IG) transcripts (online supplemental figure S2C). Furthermore, to confirm that the aggregates were bona fide TLS, with a follicular-like germinal center, we performed an IHC analysis on the entire post-treatment resected tissue, using CD20, CD3, CD21 and CD23 markers. This analysis confirmed the presence of TLS at different stages of maturation (figure 2A, B).
To increase resolution and identify the precise T and B-cell subsets, at single-cell level, that most probably mediated the response to chemo-immunotherapy, we integrated public single-cell T and B cells atlases to our spatial transcriptomic analysis of both brain and lung tissue specimens using the Cytospace tool. This analysis allowed to achieve a granular blueprint of the different lymphocyte subsets populating brain and post-treatment lung tissue specimens, highlighting the expansion and differentiation of B cells (figure 3A and B). Specifically, localized B cell clusters associated with TLS formation, B-cell activation and maturation were selectively observed in post-treatment lung tissue along with sparse antibody-producing plasma cells (figure 3B). Finally, encouraged by the robust immune activation elicited by the treatment at the tumor site, we sought to investigate whether the analysis of the patient’s plasma could inform on the pCR achievement. On the day of lung surgery, we collected patient plasma and used the proximity extension assay-based technology from Olink to perform high-throughput proteomic analysis. Compared with plasma of a gender-matched healthy population (dataset Olink), we observed a relevant number of immune-related protein either upregulated or downregulated in the plasma of the patient (figure 3C). Notably, we identified increased expression of proteins suggestive of enhanced immune activation, tissue remodeling and inflammation along with stress response and metabolic reprogramming in the patient’s plasma. Among the upregulated proteins, we highlight increased interleukin (IL)-5 and IL-5RA expression, an immune network likely associated with eosinophil activation recently implicated in response to immune checkpoint blockade (ICB).9 The chemokines CCL21 and CXCL12, essential for lymphocyte recruitment and dendritic cell activation were also upregulated10 along with ITGB6. Conversely, the downregulation of apoptotic and pro-inflammatory mediators, such as IL-1B, IL-13, and CASP2/CASP8 suggests a reduction in systemic inflammation, favoring a shift from inflammation-driven tissue damage to controlled immune activation.
Spatial transcriptomics, at single-cell resolution, highlights the presence of activated B, memory and plasma cells in post-treatment lung tissue. Mapping of public single-cell RNA sequencing of T and B cells by Cytospace into the pretreatment brain metastasis (A) and post-treatment primary lung tissue (B) of the patient with NSCLC. Main T and B-cell subsets are displayed: CD8_Tn (CD8+naïve T cells), CD8 Teff (CD8+effector T cells), CD8_Tex (CD8+exhausted T cells), CD4_Tn (CD4+naïve T cells), CD4_Treg (CD4+regulatory T cells), CD4_Tfh (CD4+T follicular helper cells), B08_SwBm (switched B memory), B09_AtM (B memory cells), B_plasma (plasma cells), and B15_ASC (antibody secreting plasma cells). (C) Plasma proteomics of the patient after chemo-immunotherapy treatment showing significantly upregulated and downregulated proteins compared with those detected in the plasma of female healthy population. NSCLC, non-small cell lung cancer.
Taken together, this integrated approach suggests that the effectiveness of neoadjuvant chemo-immunotherapy, at least in this patient, is associated with B cells and humoral immune response.
Discussion
Neoadjuvant therapy using chemo-immunotherapy is a new reality in the clinical management of resectable NSCLC, demonstrating significant clinical benefits, including improved long-term survival outcomes for patients.
In this report, we presented a case of a patient with resectable NSCLC and a solitary synchronous brain metastasis showing a pCR after chemo-immunotherapy and a long event-free survival. Taking advantage of spatial transcriptomics and serum proteomics, we deciphered both brain metastasis and primary lung TIME composition and functionality. The insights gathered from this integrated approach to characterize the TIME of the significant, durable response of the patient with NSCLC treated with neoadjuvant chemo-immunotherapy suggest: (1) lack of organized lymphocytes within the pre-treatment brain metastasis; (2) presence of mature TLS within the post-treatment primary lung tumor. Due to the lack of pre-treatment primary lung tumor samples, we cannot draw definitive conclusions, as only validation in a larger cohort can provide. However, this study highlights how the integration of multiomics, TIME, and systemic factors contributes to establishing a framework for neoadjuvant precision immunotherapy.
Our hypothesis that B cells, when organized in TLS may contribute to in situ generation of an high-affinity response with specific immune memory, is in part limited by the lack of pre-treatment primary tumor sample. However, although limited to a single patient, our findings suggest that the abundance of mature TLS in the lung tissue may act as localized immune hubs, in line with recent data by Molina-Alejandre and collaborators that highlighted the important role of TLS in response to neoadjuvant chemo-immunotherapy.11 Notably, while the identification of B cells and TLS provides strong rationale for their role in mediating treatment response, more appropriate preclinical platforms, such as murine models, are needed to draw mechanistic conclusions. Despite this gap, the integration of spatial transcriptomic data and serum proteomic profiling suggests a shift from inflammation-driven tissue damage to controlled immune activation. This multimodal approach paves the way for the identification of both tissue-specific and systemic novel biomarkers to better stratify patients, and guide clinicians’ decision for treatment escalation or de-escalation after surgery. Although being outside the inclusion criteria for clinical trials on neoadjuvant chemo-immunotherapy, this patient treated with radical intent exhibited an extraordinary response, offering insights into the mechanisms underlying the observed effectiveness. Although limited to one patient and lacking the pre-treatment primary tumor sample, our findings underscore how, starting from high-throughput technologies, we could identify biomarkers to be applied in the clinical practice for tailoring precise treatments.
Supplemental material
Supplemental material
Ethics statements
Patient consent for publication
Ethics approval
This study involves a case report based on the medical history and clinical findings of a single patient. As the study is not a clinical trial or interventional research but rather a retrospective description of the patient's case, it did not require approval from an Ethics Committee or Institutional Review Board according to the policies of our institution
Acknowledgments
The authors would like to thank the patient presented in this study. Patients' tissues were available thanks to Biobank IRCCS-Regina Elena National Cancer Institute (BBIRE), Rome, Italy. The authors are grateful to Maria Vincenza Sarcone for secretarial assistance, to Giulia Campo for technical assistance, and to Maria Manuela Rosado for helpful suggestions and critical review of the manuscript.
Footnotes
NP and FG contributed equally.
Contributors NP, FG and PN conceived the study and wrote the paper. GF and LD'A performed computational analysis. VB contributed to IHC and Visium analysis. FDN sequenced the Visium data. FF, LL and FC treated and followed the patient. PV performed the histological diagnosis and evaluated the pathological response according to the IASLC and immune-related pathologic response criteria (irPRC). PN is the guarantor of this work.
Funding This work is supported by AIRC (IG30395), CAL.HUB.RIA Ministero Salute PNRR-POS T4, "Ricerca Corrente 2024" granted by the Italian Ministry of Health.
Competing interests No, there are no competing interests.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.