Major histocompatibility complex II serves as a prognostic biomarker in resectable pulmonary sarcomatoid carcinoma: development of a prediction model

Introduction

Pulmonary sarcomatoid carcinoma (PSC), a rare lung cancer, constitutes about 0.1–0.4% of all lung cancer (1,2) and about 0.5% of non-small cell lung cancer (NSCLC) (3,4). According to 2015 The World Health Organization (WHO) classification of lung tumor, PSC is divided into pleomorphic carcinoma, giant cell carcinoma, spindle cell carcinoma, carcinosarcoma and pulmonary blastoma (5). Given the high aggressiveness and resistant to platinum-based chemotherapy (6), the prognosis of PSC is much worse than other subtypes of NSCLC (2,4). Even in patients having complete resection, the recurrence rate could be 51.6% (7). The reported 5-year survival rate of PSC ranged from 12% to 25% (1,3,4,8). Thus, further studies on this rare subtype of NSCLC is necessary for understanding this disease profoundly and finding more therapeutic strategies.

Antigen presentation is indispensable in priming anti-tumor immunity. Major histocompatibility complex II (MHC-II) molecules are critical heterodimers in antigen presentation and CD4⁺ T cell priming (9). MHC-II constitutively expresses on the cell membrane of professional antigen presenting cells (APCs) and thymic epithelial cells (10). Although other cells do not express MHC-II, under certain circumstance such as stimulated by interferon-γ (IFN-γ), other cells can express MHC-II and present antigens to activate the immunity (11,12). In previous studies, the MHC-II expression in a variety of cancers, such as colorectal cancer (CRC), esophageal adenocarcinoma (EAC), head and neck squamous cell carcinoma (HNSCC), gastric cancer (GC), NSCLC and pancreatic ductal adenocarcinoma (PDAC), has been investigated on tumor cells, immune cells or other somatic cells, where the correlation between high MHC-II expression and better patient prognosis was also reported (13-19). And higher expression of MHC-II has been linked with more favorable outcome of immune checkpoint blockade (ICB) in melanoma (20). However, MHC-II expression and its role in cancer immunity have been rarely studied in PSC.

In this study, we developed a prognostic prediction model for resectable PSC based on immunohistochemistry (IHC) analysis of MHC-II to evaluate its intratumoral expression and uncover its prognostic role. Specially, the expression of MHC-II on PSC tumor cells and tumor-infiltrating lymphocytes (TILs) was evaluated in IHC sections. For development of a prognostic prediction model, their correlation with the progression-free survival (PFS) and the overall survival (OS) of patients after completed resection were analyzed. To elucidate the potential biological mechanisms, we also explored the potential role of MHC-II in shaping the immune microenvironment of PSC based on the bioinformatics analysis of a public RNA sequencing data. We present this article in accordance with the TRIPOD reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0110/rc).

Methods

Patients

In this retrospective single-center study, we reviewed the records from 2013 to 2017 in Shanghai Pulmonary Hospital and enrolled patients who were diagnosed as PSC after surgery and had available tissue for analysis. All clinical characteristics and preoperative blood test results were extracted from electronic medical records. The major peripheral blood inflammatory markers, including neutrophil to lymphocyte ratio (NLR), lymphocyte to monocyte ratio (LMR), and eosinophil to neutrophil ratio (ENR) were calculated based on complete blood counts. The pathological stage was determined according to the International Association for the Study of Lung Cancer (IASLC) version 8^th tumor-node-metastasis (TNM) staging system. The follow-up period ended in 2022.

Ethical consideration

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Medical Ethics Committee of Shanghai Pulmonary Hospital (No. K17-241). All patients provided written informed consent.

IHC

To analyze MHC-II expression on the PSC tissues, formalin-fixed paraffin-embedded (FFPE) blocks of the PSC patients in our cohort were collected for IHC analysis in 2017. Primary antibodies used in the study were CD3 (1:100, Dako A0452, Glostrup, Denmark), CD4 (1:80, Dako M7310), CD8 (1:100, Dako M7103), and MHC-II (HLA-DR, HLA-DP, HLA-DQ) (1:100, Abcam CR3/43, Cambridge, UK). All FFPE tissue slides were dewaxed with xylene, followed with alcohol, and rinsed with distilled water. Antigen recovering was taken with the target retrieval solution kit (DM828 or DM829, Dako) under hot high pressure for 10 min. After cooling to room temperature, slides were immersed with 0.3% H₂O₂. Then the slides were incubated with the primary antibody for 1 hour at room temperature. After being rinsed with phosphate-buffered saline (PBS), the slides will be incubated with the horseradish peroxidase-conjugated detection antibody for 30 min at room temperature. Rinsed with PBS, and visualizing the antigen with 3,3'-diaminobenzidine (DAB), following with the standard procedures of counterstain for cell nuclear with hematoxylin and mounting of cover slides. Referring to the clinical programmed death-ligand 1 (PD-L1) scoring standard, only cells exhibiting distinct membrane staining were considered positive for MHC-II. The proportion scores were calculated as the percentage of positive tumor cells or TILs out of the total respective cells in the evaluated area. Two pathologists, blinded to clinical outcomes, evaluated the tissue slides together and reached a consensus on the final proportion scores of tumor cells and TILs for each sample.

Determining the cut-off value

The cut-off values for CD3, CD4, CD8, and MHC-II on TILs, MHC-II on tumor cells and peripheral blood variables were calculated by X-tile (version 3.6.1, Yale University, New Haven, CT, USA). The optimal cut-off value for the prognostic score was determined using the R package “survminer”.

Construction of prognostic model

The R package “caret” was used to randomly divide the cohort into train cohort and test cohort according to 1:1. The R package “glmnet” (21) was used for least absolute shrinkage and selection operator (LASSO) regression, where a 10-fold cross-validation was performed to determine the optimal penalty parameter, thereby minimizing the risk of overfitting. We developed two MHC-II based prognostic models for PFS and OS respectively. The prognostic score calculating formula is:

Prognosticscore=∑i=1nCoefi×xi[1]

in which Coef_i means coefficient of the variable, x_i means the value of the variable. The performance of the models was evaluated by time dependent receiver operating characteristic (ROC) curve. Area under the curve (AUC) was calculated for 1-, 3-, and 5-year PFS and OS respectively. Patients were stratified into high- and low-risk groups based on the median value of the prognostic score in the training cohort.

Obtaining the expression profile data of PSC

Because of the rarity of this disease, we only found one available public dataset on Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/gds). GSE110205 contains 17 samples including 14 PSC samples and 3 normal samples. Gene expression profile (GEP) and clinical data were downloaded from GEO. And gene expression was normalized to transcripts per million reads (TPM).

Identification of differentially expressed genes (DEGs)

K-means clustering was applied to classify the samples according to the expression of classical MHC-II genes, HLA-DRA, HLA-DRB1, HLA-DPA1, HLA-DPB1, HLA-DQA1, and HLA-DQB2. Both elbow method and Silhouette coefficient were applied to determine the optimal number of clusters. The R package “DEseq2” was applied to identify DEGs between the two clusters. And |log₂fold change (FC)| higher than 2 and adjust P value less than 0.05 were set as cut-off value.

Gene set enrichment analysis (GSEA)

GSEA was implemented to compare the biological characteristics that differentially enriched in two clusters. c2.cp.kegg.v7.2.entrez.gmt, c5.go.bp.v7.2.entrez.gmt, c5.go.cc.v7.2.entrez.gmt, and c5.go.mf.v7.2.entrez.gmt (version 7.2) were downloaded from Molecular Signature Database (MSigDB; http://www.gsea-msigdb.org/gsea/msigdb/index.jsp) as reference gene sets. R packages “clusterProfiler” and “DOSE” were applied for GSEA (22,23). |Normalized enrichment score (NSE)| >1.5 and q value <0.05 were defined as cutoff value.

Analysis of immune infiltration landscape

We used xCell algorithm (24) (https://xcell.ucsf.edu/) to reveal the immune infiltration of two clusters. 64-cell signature was chosen as reference signature which including various lymphocytes, myeloid cells, stromal cells, and so on. Xcell score represented the relative abundance of each tumor infiltrating cell in each sample.

Comparison of gene signature between two clusters

We collected some gene signature of immune related biological signature including: (I) IFN-γ signature; (II) inflammatory chemokine signature; (III) T cell-inflamed GEP; (IV) CD4⁺ T cell activation; (V) Cytolytic activity; (VI) T cell costimulator signature; (VII) T cell coinhibitor (also known as immune checkpoint) signature; (VIII) programmed cell death protein 1 (PD-1) signaling (Table S1) (20,25-28). Single sample GSEA (ssGSEA) algorithm was conducted to analysis these gene signature in two clusters. R package “GSVA” was applied for ssGSEA score calculation.

Statistical analysis

The statistics was conducted through SPSS 22.0 (SPSS Inc., Chicago, IL, USA) and R software (version 4.0.2). The measurement data were shown as median (quartile) or mean (standard deviation). The enumeration data were shown as number (proportion). Student t-test or Wilcoxon sum rank test was used to identify difference between measurement data. And enumeration data were analyzed by Chi-squared test or Fisher exact test. Spearman correlation was used to analyze the correlation between two variables. Univariable and multivariable logistic regression analysis was applied to identify factors that have potential influence on MHC-II expression. Kaplan-Meier and log-rank test was used to analyze PFS or OS. Cox regression was applied to primarily select the potential prognostic variables. The “forestplot” R package was used to visualize the result of univariate cox regression analysis. Drawing time dependent ROC and calculating AUC value were achieved by R package “timeROC”. P value less than 0.05 was thought as significant.

Results

Patient characteristics

There were 86 patients being enrolled totally (Table 1). The median age was 64 years (range, 39–83 years; quartile, 57–69 years). Among all patients, men constituted much more than women (82.56% vs. 17.44%). And in our cohort, 51.16% patients had smoking history. All participants underwent surgery and more than half patients were at stage I and II (63.96%). Only 5.81% patients were stage IV and the reasons for these patients had surgery were accidentally incomplete examination before surgery and unexpected pleural metastasis observed during surgery. In this cohort, about 44.19% patients had post-operative chemotherapy after surgery.

MHC-II expression and its relationship with clinical and other immune parameters

To explore the clinical and prognostic significance of MHC-II in PSC, we first investigated the correlation between MHC-II expression and various clinicopathological and immune parameters. To evaluate the MHC-II expression levels of tumor cells and TILs, the proportion scores were collected through IHC analysis. Based on the cut-off values determined by X-tile software, high MHC-II expression was defined as proportion score >30% on tumor cells and >10% on TILs. Accordingly, high MHC-II expression was observed on tumor cells (Figure S1A) in 19 (22.09%) patients and on TILs (Figure S1B) in 31 (36.05%) patients.

Next, we next investigated the features associated with MHC-II expression, not only including the basic clinical data, but also the inflammatory cell test of peripheral blood and the IHC of the key T lymphocyte molecule expression in tumor tissue. As results, MHC-II expression on tumor cells was inversely associated with T stage (P=0.02, Table 2). Although MHC-II expression showed no correlation with other clinical parameters or peripheral blood inflammatory markers (Table S2), its correlations with immune features within the tumor microenvironment (TME), particularly CD4⁺ lymphocytes infiltration (P=0.03, Table 3), was found. These results were consistent with the further correlation analysis (Figure 1A). Not surprisingly, CD3 expression had a strong positive correlation with CD4 (R=0.52, P<0.001, Figure 1B) and CD8 (R=0.76, P<0.001, Figure 1C) in the tumor tissue, as expected from the canonical T lymphocyte marker profiles. Notably, MHC-II expression on TILs was positively correlated with its expression on tumor cells (R=0.31, P=0.004, Figure 1D) and showed a weak but statistically significant positive correlation with CD4⁺ T cell infiltration (R=0.27, P=0.01, Figure 1E). This trend was further explored in an independent cohort (GSE110205, n=14), which also supported a positive association between MHC-II and CD4 expression (Figure S2).

Figure 1 Correlation analyses between MHC-II expression and immune markers. (A) Correlation between selected variables; (B) correlation between CD3 and CD4; (C) correlation between CD3 and CD8; (D) correlation between expression of MHC-II on tumor cells and TILs; (E) correlation between expression of MHC-II on TILs and CD4. ENR, eosinophil to neutrophil ratio; LMR, lymphocyte to monocyte ratio; MHC-II, major histocompatibility complex II; MHC-II-T, MHC-II expression on tumor cells; MHC-II-L, MHC-II expression on TILs; NLR, neutrophil to lymphocyte ratio; TILs, tumor-infiltrating lymphocytes.

To confirm that these factors were independent predictors, we performed a multivariate logistic regression analysis. As shown in Table 4, high CD4⁺ T cell infiltration was an independent predictor for high MHC-II expression on TILs [odds ratio (OR) =2.633; 95% confidence interval (CI): 1.012–6.847; P=0.047]; MHC-II expression on tumor cells was strongly associated with its expression on TILs (OR =3.086; 95% CI: 1.050–9.070; P=0.040); higher T stage was an independent predictor for lower MHC-II expression on tumor cells (OR =0.177; 95% CI: 0.036–0.884; P=0.04).

MHC-II expression and survival of PSC patients

At the endpoint of the follow up, 61 (70.93%) patients experienced disease progression or relapse, and the median PFS was about 11 months; 49 (56.98%) patients were deceased and median OS was about 15 months. As expected, advanced stage (III–IV) was associated with significantly shorter PFS (P<0.001) and OS (P=0.003) (Figure 2A). Post-operative conferred a survival benefit only to patients with advanced-stage disease (PFS: P=0.047; OS: P=0.007; Figure 2B,2C).

Figure 2 JenyK-M curves for PFS and OS. (A) Stage; (B) post-operative chemotherapy in whole cohort; (C) post-operative in patients with advanced stage; (D) MHC-II expression on TILs; (E) MHC-II expression on tumor cells; (F) intratumoral MHC-II expression; (G) intratumoral CD4 expression; (H) LMR of peripheral blood. K-M, Kaplan-Meier; LMR, lymphocyte to monocyte ratio; MHC-II, major histocompatibility complex II; OS, overall survival; PFS, progression-free survival; TILs, tumor-infiltrating lymphocytes.

Notably, MHC-II expression emerged as a robust prognostic marker. High MHC-II expression on both TILs (PFS: P=0.042; OS: P=0.03; Figure 2D) and tumor cells (PFS: P=0.005; Figure 2E) was significantly associated with prolonged survival. When combining these, patients in the “MHC-II high” group (high expression on either TILs or tumor cells) had a markedly better prognosis than the “MHC-II low” group (PFS: P=0.003; OS: P=0.01; Figure 2F). Consistent with these results, high intratumoral CD4 expression and a high peripheral LMR were also correlated with improved survival (Figure 2G,2H).

A novel prognostic model integrating MHC-II outperforms clinical stage

To create a more accurate prognostic tool, we first used univariate Cox regression to identify key prognostic variables for PFS and OS (Figure S3). Variables with P<0.10 were then entered into a LASSO regression model to build prognostic scores. And “caret” R package was used to randomly divide the whole cohort into train cohort and test cohort, and the demographic characteristics of two cohorts were summarized in Table S3.

For PFS, the LASSO algorithm optimally selected stage, MHC-II expression on tumor cells, eosinophil level and CD4 expression as variables for model construction (Figure 3A-3C). In the train cohort, the AUC of the model were 0.781 (95% CI: 0.627–0.908), 0.796 (95% CI: 0.622–0.944), and 0.785 (95% CI: 0.472–1.000) for predicting 1-, 3-, and 5-year PFS, respectively (Figure 3D). The results were further confirmed in the test cohort. A shorter PFS was observed in patients with higher prognostic score in the test cohort (P<0.001; Figure 3E). In the test cohort, the AUC of the model were 0.688, 0.708, and 0.784 for predicting 1-, 3-, and 5-year PFS which were slightly higher than single stage (1-, 3-, and 5-year AUC =0.652, 0.618, and 0.714; Figure 3F).

Figure 3 Construction of the prognostic model for PFS prediction. (A-C) LASSO regression was performed, the minimum criteria was chosen and coefficient was calculated; (D) the ROC and AUC values for 1-, 3-, and 5-year PFS in the train cohort; (E) the correlation between prognostic score and PFS in the test cohort; (F) comparing the performance of prognostic score and stage in predicting 1-, 3-, and 5-year PFS in the test cohort. AUC, area under the curve; LASSO, least absolute shrinkage and selection operator; LMR, lymphocyte to monocyte ratio; MHC-II, major histocompatibility complex II; MHC-II-T, MHC-II expression on tumor cells; PFS, progression-free survival; ROC, receiver operating characteristic.

For OS, the LASSO algorithm optimally selected stage, neutrophil level, LMR, MHC-II expression on TILs and CD4 expression for constructing the prognostic model (Figure 4A-4C). In the train cohort, the AUC was 0.769 (95% CI: 0.606–0.904), 0.796 (95% CI: 0.635–0.928), and 0.799 (95% CI: 0.458–1.000) for predicting 1-, 3-, and 5-year OS respectively (Figure 4D). Then we applied the model in the test cohort, survival difference was detected between high and low prognostic score (P=0.02, Figure 4E). The AUC was calculated as 0.735, 0.718 and 0.704 for predicting 1-, 3-, and 5-year OS, respectively. Furthermore, it could be seen that the AUC of prognostic model was higher than that of single stage (1-, 3-, and 5-year AUC =0.619, 0.554, and 0.619; Figure 4F) which highlighted that this model might perform better than single stage in predicting OS of PSC patients.

Figure 4 Construction of the prognostic model for OS prediction. (A-C) LASSO regression was performed, the minimum criteria was chosen and coefficient was calculated; (D) the ROC and AUC values for 1-, 3-, and 5-year OS in the train cohort; (E) the correlation between prognostic score and OS in the test cohort; (F) comparing the performance of prognostic score and stage in predicting 1-, 3-, and 5-year OS in the test cohort. AUC, area under the curve; LASSO, least absolute shrinkage and selection operator; LMR, lymphocyte to monocyte ratio; MHC-II, major histocompatibility complex II; MHC-II-L, MHC-II expression on TILs; OS, overall survival; ROC, receiver operating characteristic; TILs, tumor-infiltrating lymphocytes.

High MHC-II expression defines an inflamed, T cell-rich TME

To elucidate the biological basis for the prognostic role of MHC-II, we analyzed a public PSC transcriptomic dataset (GSE110205) including 14 PSC samples and 3 normal tissues. Based on expression pattern of MHC-II including HLA-DR, HLA-DP and HLA-DQ, samples were stratified into two clusters (Figure S4 and Figure 5A), respectively defined as MHC-II^high and MHC-II^low clusters (Figure 5B), which showed nine DEGs (Figure 5C). Next, we determined if there were coordinated changes at the pathway level using GSEA. The top 6 enriched pathways in relative gene sets were shown in Figure 5D, where the MHC-II^high group was profoundly enriched in antigen processing and presentation, T cell activation, and IFN-γ response pathway. This was substantiated at the cellular level through further deconvolution analysis of the immune landscape, in which MHC-II^high tumors had a significantly higher immune score (P=0.008; Figure 5E). As shown in Figure 5F and Figure S5, the MHC-II^high group was characterized by increased infiltration of dendritic cells (DCs), M1 macrophages, CD8⁺ T-cells, CD4⁺ naïve T-cells, B-cells and regulatory T-cells (Tregs), confirming the inflamed microenvironment.

Figure 5 Bioinformatics analysis for dissecting the tumor immune microenvironment of PSC with high MHC-II expression. (A) Cluster plot for PSC samples analyzed by K-means clustering; (B) boxplot for expression of classical MHC-II molecules of two clusters; (C) volcano plot for DEGs; (D) top 6 enriched pathways in each gene set; (E) general comparing the TME of two clusters; (F) differentially infiltrated immune cells between two clusters; (G) ssGSEA score for immune related gene signature. DEGs, differentially expressed genes; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; MHC-II, major histocompatibility complex II; PSC, pulmonary sarcomatoid carcinoma; ssGSEA, single sample gene set enrichment analysis; TME, tumor microenvironment.

To assess the functional state of this inflamed microenvironment, we further analyzed the immune functional signatures by ssGSEA. Compared to the MHC-II^low group, the signatures for IFN-γ response, immune-recruiting chemokines, T cell inflammation, T cell costimulation, and cytolytic activity were upregulated in the MHC-II^high group (Figure 5G). Interestingly, some coinhibitory signatures like PD-1 signaling were also higher in the MHC-II^high group (Figure 5G), which was considered as the important target for ICBs therapy. In line with this, although MHC-II as a ligand of lymphocyte-activation gene 3 (LAG-3) had no significant correlation with LAG-3 nor PD-L1 in this cohort, it showed significantly positive correlation another immune checkpoint molecule, programmed death-ligand 2 (PD-L2) (Figure 6).

Figure 6 Correlation between MHC-II and immune checkpoints (*, P<0.05; **, P<0.01). MHC-II, major histocompatibility complex II.

Discussion

PSC is a rare malignancy with a dismal prognosis (29-31), a reality reflected in our cohort despite a higher proportion of early-stage patients undergoing complete resection. The high recurrence rate, even in early-stage disease, and the limited efficacy of adjuvant chemotherapy (1,7,31) underscore the urgent need for novel biomarkers and therapeutic targets. MHC-II molecules are pivotal for priming anti-tumor immunity by presenting antigens to CD4⁺ T cells (32-34), whose prognostic significance has been established in various cancers (13-19,35). However, their role in PSC remains largely unexplored.

In our immunohistochemical analysis of PSC tissue, we found a substantial portion of PSC patients with significant intratumoral MHC-II expression and a strong correlation between MHC-II expression and CD4⁺ T cell infiltration. Simulated to other tumors, this MHC-II/CD4 axis may also play an important physiological role in PSC. Crucially, we demonstrated that this biological link has potential of translation into a clinical benefit. High MHC-II expression was a robust indicator of prolonged PFS and OS, echoing findings in oropharyngeal cancer (36) and underscoring its role as a marker of an engaged, tumor-suppressive immune state. To harness this prognostic power, we developed a novel predictive model integrating MHC-II, which demonstrated superior accuracy over the current staging system. While this model requires validation in larger, multi-center cohorts, it represents a significant step towards more personalized risk stratification in PSC.

To dissect the biological underpinnings of MHC-II’s favorable prognosis, our bioinformatics analysis of a public dataset painted a vivid picture of an inflamed TME. The upregulation of genes like C-X-C motif chemokine 9 (CXCL9), a potent T-cell chemoattractant, and ubiquitin D (UBD), crucial for DC maturation (37), in the MHC-II^high group provided the first molecular clues. Corroborated by GSEA, a landscape dominated by T-cell activation and proliferation pathways within TME was revealed. At a cellular level, the MHC-II^high TME was characterized by a significantly higher immune score and was densely infiltrated by a triad of crucial immune players: antigen-presenting cells, effector T cells, and B cells. This aligns with findings in other cancers, where the MHC-II/CD4 axis is crucial for orchestrating T-cell infiltration (13,17). Functionally, multiple scores related to TILs were increased in the MHC-II^high group, in which the heightened cytolytic activity and CD4⁺ T cell activation scores suggest that these infiltrating cells are not mere bystanders but are functionally engaged in an anti-tumor response. In addition, the heightened IFN-γ signature provides a plausible mechanism for maintaining the activated state of serials of immune functions in the MHC-II^high group, which was proven to upregulate expression of MHC-II, CXCL9, and UBD, as well as promote differentiation of M1 macrophage (38-40). Therefore, we propose a positive feedback loop model for the inflamed PSC microenvironment. Initial MHC-II expression primes CD4⁺ T cells, which produce IFN-γ along with the activated CD8⁺ T cells. The IFN-γ further amplifies MHC-II expression on tumor cells, enhances antigen presentation and promotes a tumor-suppressive M1 macrophage phenotype, finally creating the robustly inflamed, "hot" TME of MHC-II^high PSC.

However, this inflamed state is not without opposition. The concurrent upregulation of immunosuppressive elements, including Tregs and PD-1/PD-L2 signaling, indicates a state of “active but counter-regulated” immunity. Given the established link between MHC-II expression and ICBs efficacy in other cancers (17,20), we posit that the anti-tumor effect could be derived from the unleash of pre-existing, yet suppressed, immune response in MHC-II^high PSC by ICBs. Our data, showing a high T-cell-inflamed GEP score—a known predictor of anti-PD-1 efficacy (41)—and the link to PD-L2, also support this hypothesis. Moreover, therapeutic strategies aimed at upregulating MHC-II, such as with HDAC inhibitors, may represent a promising approach to convert “cold” tumors to “hot”, thereby sensitizing them to ICBs therapy (42). Therefore, MHC-II holds promise in the immunotherapy of PSC not only as a biomarker to select potential responders but also as a potential therapeutic target to enhance efficacy.

Our study has limitations, primarily its retrospective nature, the incomplete assessment of peripheral inflammatory markers and the small sample size inherent to studying a rare disease, which carries a potential risk of overfitting and results in relatively wide CIs for the performance metrics. The bioinformatics analysis was also limited by the small number of available public datasets and the fundamental inherent differences between mRNA-level and protein-level expressions. Despite these constraints, our study provides a robust foundation and a compelling rationale for future prospective, multi-center investigations to validate MHC-II as a clinical biomarker in PSC.

Conclusions

In this study of PSC, we establish for the first time the critical prognostic and immunological role of MHC-II expression. Our key findings demonstrate that high MHC-II expression, both on tumor cells and TILs, is a powerful independent predictor of prolonged survival, which was translated into a novel prognostic model outperforming the conventional staging system. Mechanistically, we reveal that a high MHC-II status defines a distinct, immunologically “hot” TME, characterized by an influx of CD4⁺ T cells and an active, T-cell-inflamed gene signature. Collectively, our findings not only provide a deeper understanding of PSC biology but also identify MHC-II as a promising biomarker to refine patient stratification and guide future immunotherapeutic strategies.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0110/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0110/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0110/prf

Funding: This study was supported in part by grants from the National Key Research and Development Program of China (No. 2022YFF0705300), the National Natural Science Foundation of China (Nos. 52272281 and 52473144), the Shanghai Municipal Science and Technology Major Project (No. 2021SHZDZX0100), the Fundamental Research Funds for the Central Universities, Clinical Research Project of Shanghai Pulmonary Hospital (No. FKLY20010), the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No. 2024ZD0521104), the Shanghai Science and Technology Commission (Explorer Program No. 25TS141100), and the Shanghai Leading Talent Program of Eastern Talent Plan.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0110/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Medical Ethics Committee of Shanghai Pulmonary Hospital (No. K17-241). All patients provided written informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Lococo F, Rapicetta C, Cardillo G, et al. Pathologic Findings and Long-Term Results After Surgical Treatment for Pulmonary Sarcomatoid Tumors: A Multicenter Analysis. Ann Thorac Surg 2017;103:1142-50. [Crossref] [PubMed]
2. Yendamuri S, Caty L, Pine M, et al. Outcomes of sarcomatoid carcinoma of the lung: a Surveillance, Epidemiology, and End Results Database analysis. Surgery 2012;152:397-402. [Crossref] [PubMed]
3. Rahouma M, Kamel M, Narula N, et al. Pulmonary sarcomatoid carcinoma: an analysis of a rare cancer from the Surveillance, Epidemiology, and End Results database. Eur J Cardiothorac Surg 2018;53:828-34. [Crossref] [PubMed]
4. Steuer CE, Behera M, Liu Y, et al. Pulmonary Sarcomatoid Carcinoma: An Analysis of the National Cancer Data Base. Clin Lung Cancer 2017;18:286-92. [Crossref] [PubMed]
5. Travis WD, Brambilla E, Nicholson AG, et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of Genetic, Clinical and Radiologic Advances Since the 2004 Classification. J Thorac Oncol 2015;10:1243-60. [Crossref] [PubMed]
6. Li X, Wu D, Liu H, et al. Pulmonary sarcomatoid carcinoma: progress, treatment and expectations. Ther Adv Med Oncol 2020;12:1758835920950207. [Crossref] [PubMed]
7. Zeng Q, Li J, Sun N, et al. Preoperative systemic immune-inflammation index predicts survival and recurrence in patients with resected primary pulmonary sarcomatoid carcinoma. Transl Lung Cancer Res 2021;10:18-31. [Crossref] [PubMed]
8. Sun L, Dai J, Chen Y, et al. Pulmonary Sarcomatoid Carcinoma: Experience From SEER Database and Shanghai Pulmonary Hospital. Ann Thorac Surg 2020;110:406-13. [Crossref] [PubMed]
9. Neefjes J, Jongsma ML, Paul P, et al. Towards a systems understanding of MHC class I and MHC class II antigen presentation. Nat Rev Immunol 2011;11:823-36. [Crossref] [PubMed]
10. Roche PA, Furuta K. The ins and outs of MHC class II-mediated antigen processing and presentation. Nat Rev Immunol 2015;15:203-16. [Crossref] [PubMed]
11. Herkel J, Jagemann B, Wiegard C, et al. MHC class II-expressing hepatocytes function as antigen-presenting cells and activate specific CD4 T lymphocyutes. Hepatology 2003;37:1079-85. [Crossref] [PubMed]
12. Mulder DJ, Pooni A, Mak N, et al. Antigen presentation and MHC class II expression by human esophageal epithelial cells: role in eosinophilic esophagitis. Am J Pathol 2011;178:744-53. [Crossref] [PubMed]
13. Sconocchia G, Eppenberger-Castori S, Zlobec I, et al. HLA class II antigen expression in colorectal carcinoma tumors as a favorable prognostic marker. Neoplasia 2014;16:31-42. [Crossref] [PubMed]
14. Dunne MR, Michielsen AJ, O'Sullivan KE, et al. HLA-DR expression in tumor epithelium is an independent prognostic indicator in esophageal adenocarcinoma patients. Cancer Immunol Immunother 2017;66:841-50. [Crossref] [PubMed]
15. Gameiro SF, Ghasemi F, Barrett JW, et al. High Level Expression of MHC-II in HPV+ Head and Neck Cancers Suggests that Tumor Epithelial Cells Serve an Important Role as Accessory Antigen Presenting Cells. Cancers (Basel) 2019;11:1129. [Crossref] [PubMed]
16. Ghasemi F, Tessier TM, Gameiro SF, et al. High MHC-II expression in Epstein-Barr virus-associated gastric cancers suggests that tumor cells serve an important role in antigen presentation. Sci Rep 2020;10:14786. [Crossref] [PubMed]
17. Johnson AM, Bullock BL, Neuwelt AJ, et al. Cancer Cell-Intrinsic Expression of MHC Class II Regulates the Immune Microenvironment and Response to Anti-PD-1 Therapy in Lung Adenocarcinoma. J Immunol 2020;204:2295-307. [Crossref] [PubMed]
18. Michelakos T, Cai L, Villani V, et al. Tumor Microenvironment Immune Response in Pancreatic Ductal Adenocarcinoma Patients Treated With Neoadjuvant Therapy. J Natl Cancer Inst 2021;113:182-91. [Crossref] [PubMed]
19. He Y, Rozeboom L, Rivard CJ, et al. MHC class II expression in lung cancer. Lung Cancer 2017;112:75-80. [Crossref] [PubMed]
20. Rodig SJ, Gusenleitner D, Jackson DG, et al. MHC proteins confer differential sensitivity to CTLA-4 and PD-1 blockade in untreated metastatic melanoma. Sci Transl Med 2018;10:eaar3342. [Crossref] [PubMed]
21. Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 2010;33:1-22.
22. Yu G, Wang LG, Han Y, et al. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16:284-7. [Crossref] [PubMed]
23. Yu G, Wang LG, Yan GR, et al. DOSE: an R/Bioconductor package for disease ontology semantic and enrichment analysis. Bioinformatics 2015;31:608-9. [Crossref] [PubMed]
24. Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol 2017;18:220. [Crossref] [PubMed]
25. Coppola D, Nebozhyn M, Khalil F, et al. Unique ectopic lymph node-like structures present in human primary colorectal carcinoma are identified by immune gene array profiling. Am J Pathol 2011;179:37-45. [Crossref] [PubMed]
26. Ayers M, Lunceford J, Nebozhyn M, et al. IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest 2017;127:2930-40. [Crossref] [PubMed]
27. Rooney MS, Shukla SA, Wu CJ, et al. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell 2015;160:48-61. [Crossref] [PubMed]
28. Chen L, Flies DB. Molecular mechanisms of T cell co-stimulation and co-inhibition. Nat Rev Immunol 2013;13:227-42. [Crossref] [PubMed]
29. Yang Z, Xu J, Li R, et al. PD-L1 and CD47 co-expression in pulmonary sarcomatoid carcinoma: a predictor of poor prognosis and potential targets of future combined immunotherapy. J Cancer Res Clin Oncol 2019;145:3055-65. [Crossref] [PubMed]
30. Vieira T, Antoine M, Hamard C, et al. Sarcomatoid lung carcinomas show high levels of programmed death ligand-1 (PD-L1) and strong immune-cell infiltration by TCD3 cells and macrophages. Lung Cancer 2016;98:51-8. [Crossref] [PubMed]
31. Lin Y, Yang H, Cai Q, et al. Characteristics and Prognostic Analysis of 69 Patients With Pulmonary Sarcomatoid Carcinoma. Am J Clin Oncol 2016;39:215-22. [Crossref] [PubMed]
32. Haabeth OAW, Fauskanger M, Manzke M, et al. CD4(+) T-cell-Mediated Rejection of MHC Class II-Positive Tumor Cells Is Dependent on Antigen Secretion and Indirect Presentation on Host APCs. Cancer Res 2018;78:4573-85. [Crossref] [PubMed]
33. Borst J, Ahrends T, Bąbała N, et al. CD4(+) T cell help in cancer immunology and immunotherapy. Nat Rev Immunol 2018;18:635-47. [Crossref] [PubMed]
34. Liu Z, Derkach A, Yu KJ, et al. Patterns of Human Leukocyte Antigen Class I and Class II Associations and Cancer. Cancer Res 2021;81:1148-52. [Crossref] [PubMed]
35. Piao XM, Kang HW, Jeong P, et al. A prognostic immune predictor, HLA-DRA, plays diverse roles in non-muscle invasive and muscle invasive bladder cancer. Urol Oncol 2021;39:237.e21-9.
36. Cioni B, Jordanova ES, Hooijberg E, et al. HLA class II expression on tumor cells and low numbers of tumor-associated macrophages predict clinical outcome in oropharyngeal cancer. Head Neck 2019;41:463-78. [Crossref] [PubMed]
37. Ebstein F, Lange N, Urban S, et al. Maturation of human dendritic cells is accompanied by functional remodelling of the ubiquitin-proteasome system. Int J Biochem Cell Biol 2009;41:1205-15. [Crossref] [PubMed]
38. Susek KH, Karvouni M, Alici E, et al. The Role of CXC Chemokine Receptors 1-4 on Immune Cells in the Tumor Microenvironment. Front Immunol 2018;9:2159. [Crossref] [PubMed]
39. Aichem A, Groettrup M. The ubiquitin-like modifier FAT10 in cancer development. Int J Biochem Cell Biol 2016;79:451-61. [Crossref] [PubMed]
40. Müller E, Christopoulos PF, Halder S, et al. Toll-Like Receptor Ligands and Interferon-γ Synergize for Induction of Antitumor M1 Macrophages. Front Immunol 2017;8:1383. [Crossref] [PubMed]
41. Cristescu R, Mogg R, Ayers M, et al. Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy. Science 2018;362:eaar3593. Erratum in: Science 2019;363:eaax1384. [Crossref] [PubMed]
42. Briere D, Sudhakar N, Woods DM, et al. The class I/IV HDAC inhibitor mocetinostat increases tumor antigen presentation, decreases immune suppressive cell types and augments checkpoint inhibitor therapy. Cancer Immunol Immunother 2018;67:381-92. [Crossref] [PubMed]