Expression of YAP1 delineates distinct subtypes of pulmonary large cell neuroendocrine carcinoma with divergent therapeutic implications

Introduction

Pulmonary large cell neuroendocrine carcinoma (LCNEC) is a rare, highly aggressive malignancy accounting for approximately 3% of primary pulmonary neoplasms (1). Patients with LCNEC typically present with nonspecific symptoms, frequently leading to advanced-stage diagnosis and poor prognosis (2). Current LCNEC therapeutic strategies lack consensus, with significant inter-institutional heterogeneity in clinical management (3). These challenges highlight the urgent need to elucidate the oncogenic drivers and identify actionable targets through comprehensive molecular profiling. Such efforts may enable personalized treatment approaches and optimize clinical decision-making for this treatment-refractory malignancy.

Next-generation sequencing (NGS) has established LCNEC as a genomically heterogeneous malignancy, segregating into two predominant molecular subtypes. The small cell lung carcinoma-like (SCLC-like) subtype features co-occurring TP53 and RB1 inactivation through mutation or loss of heterozygosity. Conversely, the non-small cell lung carcinoma-like (NSCLC-like) subtype demonstrates a mutational profile overlapping with conventional NSCLC drivers—including STK11, KRAS, and KEAP1 alterations—while lacking concomitant TP53/RB1 aberrations (4). Through integrative genomic and transcriptomic profiling, George et al. refined this classification: Type I (NSCLC-like) tumors harbor TP53 mutations paired with STK11/KEAP1 alterations, whereas Type II (SCLC-like) tumors exhibit TP53 mutations coupled with RB1 loss. Notably, Type I LCNECs display prominent neuroendocrine differentiation marked by upregulated ASCL1/DLL3 expression and suppressed NOTCH signaling. Conversely, Type II LCNECs demonstrate attenuated neuroendocrine marker expression with NOTCH pathway activation (5). This molecular taxonomy holds clinical relevance, as subtype-specific signatures could guide therapeutic strategies and prognostication (6-8).

Recent advances in SCLC classification establish a molecular subtyping system based on four key transcriptional regulators: ASCL1 (SCLC-A), neuronal differentiation factor 1 (NEUROD1; SCLC-N), POU2F3 (SCLC-P), and YAP1 (SCLC-Y) (9). This framework facilitates development of subtype-targeted therapies validated in preclinical and clinical settings (10-14). However, subtype classification discrepancies exist. Gay et al. failed to validate the YAP1-defined subtype, instead proposing an inflamed molecular subtype (SCLC-I) characterized by triple-low expression patterns (ASCL1/NEUROD1/POU2F3) and enhanced immune signatures (10). The biological significance of YAP1 as a definitive SCLC subtype biomarker remains controversial, warranting integrative multi-omics analyses to clarify potential equivalence between putative SCLC-Y and SCLC-I subtypes. Notably, George et al. identified YAP1 co-expression with immune-related genes in Type II LCNEC (5), suggesting molecular parallels between SCLC-Y and this LCNEC subset. Emerging evidence demonstrates that YAP1-driven transcriptional programs in LCNEC harbor clinically actionable features with prognostic relevance (15-18), providing a foundation for precision oncology through YAP1-based stratification in neuroendocrine carcinomas.

Collectively, emerging evidence indicates that molecular profiling of LCNEC, particularly the binary classification framework delineating molecular subtypes (Type I and Type II) coupled with YAP1 expression analysis, offers promise for precision oncology in LCNEC management. Nevertheless, a significant knowledge gap exists: no comparative evidence evaluates the clinical utility of molecular subtyping versus YAP1 status for optimizing therapeutic decisions or prognostication in LCNEC patients. To address this unmet need, we retrospectively analyzed surgically resected LCNEC cases. Using NGS and immunohistochemistry (IHC) on archival specimens, we classified tumors by molecular subtype and YAP1 protein expression, comparing clinicopathological features, survival outcomes, and treatment-relevant biomarkers between these classifications. We present this article in accordance with the REMARK reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-491/rc).

Methods

Patient selection and data collection

This retrospective cohort study analyzed archived clinical specimens from the Peking University Cancer Hospital spanning January 2010 to December 2022. Forty-two surgically resected LCNEC cases met inclusion criteria: (I) pathologically confirmed diagnosis with LCNEC, encompassing both pure histological subtypes and 6 combined subtypes previously characterized in our prior research (19); (II) sufficient tumor tissue availability (minimum diameter ≥0.5 cm) with exclusion of specimens of poor quality for NGS or IHC analyses; (III) complete clinicopathological documentation with overall survival (OS) and disease-free survival (DFS) endpoints extracted from institutional electronic health records or structured telephone follow-up. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking University Cancer Hospital and the local review board (No. 2023KT23), and individual consent for this retrospective analysis was waived.

All archival hematoxylin and eosin (H&E)-stained slides and immunohistochemical preparations (TTF-1, Napsin A, P40, CK5/6, LCA, CD56, CgA, Syn, NUT, INI-1, and BRG) were retrospectively reviewed by two senior thoracic pathologists (Y.Z. and H.W.) to validate LCNEC diagnoses. Discordant cases were resolved through consensus review using multiheaded microscopy. Tumor staging was assigned according to the American Joint Committee on Cancer tumor-node-metastasis (TNM) staging system, 8th edition (20).

IHC staining and scoring criteria

Formalin-fixed paraffin-embedded (FFPE) tissue blocks were processed and serially sectioned at 4 µm thickness for IHC evaluation. Specimens underwent staining for eleven biomarkers: p53, Rb, ASCL1, NEUROD1, POU2F3, YAP1, DLL3, programmed death-ligand (PD-L1), CD8, major histocompatibility complex class I (MHC I), and major histocompatibility complex class II (MHC II) (detailed protocols in Table S1). Whole-tumor-area assessment was performed. For combined LCNECs, FFPE blocks containing ≥50% LCNEC histology were selected, with scoring restricted to the LCNEC component. Rb expression patterns were classified as mutant-type (absent nuclear staining) or wild-type (preserved heterogeneous nuclear staining).

PD-L1 positivity was defined as ≥1% membranous staining on tumor cells or immune cells (21). Tumor Proportion Score (TPS) and Combined Positive Score (CPS) were assessed using standardized quantification methods. TPS represented the percentage of viable tumor cells with partial/complete membrane staining (≥100 viable tumor cells evaluated), whereas CPS was calculated as the ratio of PD-L1-positive cells [tumor cells plus immune cells (lymphocytes, macrophages)] to total viable tumor cell count (22).

CD8 immunostaining was performed using the Pannoramic 250 Flash III scanner (3DHISTECH, Budapest, Hungary) at 200× magnification, with absolute quantification of DAB-positive cells executed through QuPath software (version 0.5.1). Whole-slide images were segmented into three regions: tumor (malignant cell clusters), stroma (peri-tumoral non-neoplastic microenvironment), and region (2-mm peripheral zone abutting tumor frontier). Classification utilized user-defined examples and QuPath’s machine learning features, as previously described (23). Digitally quantified immunopositive cells were aggregated across whole slides prior to statistical analysis. Immune cell infiltration density was stratified into low/high groups using median density thresholding.

Expression of additional biomarkers (ASCL1, NEUROD1, POU2F3, YAP1, DLL3, MHC I/II) was quantified using the histochemical score (H-score) system, calculated as the product of positive cell percentage (0–100%) and staining intensity (1+: weak, 2+: moderate, 3+: strong), yielding a theoretical range of 0–300 (24). As previously reported, specimens with H-scores ≤10 were classified as negative, while scores >10 denoted positivity (25,26).

Tertiary lymphoid structures (TLSs) were identified by morphologic evaluation, defined by three features: (I) organized T-cell zones containing mature dendritic cells, (II) adjacent B-cell follicles with germinal center formation, and (III) distinct lymphocyte aggregates. Transient lymphocyte clusters exhibiting indistinct borders on H&E staining were not classified as TLS.

NGS analyses

A cohort of 39 surgically resected LCNECs underwent whole-exome sequencing (WES), including 36 pure and 3 combined cases. For combined cases with clear morphologic demarcation between subtypes, manual dissection was performed on FFPE specimens followed by H&E verification of dissection precision. Genomic DNA was isolated from tumor regions using QIAGEN DNeasy Blood & Tissue Kit (69504), with subsequent library preparation employing xGen^® Exome Research Panel (Integrated DNA Technologies) and TruePrep DNA Library Prep Kit V2 (TD501, Vazyme Biotech). High-depth sequencing was conducted on Illumina NovaSeq 6000 platforms, achieving mean coverage of 245× for tumor tissues versus 185× for matched normal controls. Bioinformatics processing included BWA alignment to GRCh37 reference genome, sambamba-based PCR duplicate removal, and GATK4.1 base quality score recalibration.

Tumor mutational burden (TMB), defined as the total number of non-synonymous somatic mutations per tumor, was quantified through WES analysis. Tumors were classified as TMB-high using a predetermined threshold above the cohort median TMB value.

Statistical analysis

Statistical analyses were performed using SPSS 26.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA). Categorical variables were expressed as frequencies (percentages) and continuous variables as medians with interquartile ranges. Intergroup comparisons employed chi-squared/Fisher’s exact tests for categorical data and Mann-Whitney U tests for nonparametric continuous variables. Bivariate correlations were assessed through Spearman’s rank correlation coefficients. Survival metrics were defined as follows: OS from pathological diagnosis to death, and DFS from diagnosis to first recurrence/progression or death. Prognostic factors were evaluated using Cox proportional hazards regression models (univariate and multivariate) to calculate hazard ratios with 95% confidence intervals. Survival distributions were compared via Kaplan-Meier curves with log-rank testing. All statistical tests were two-tailed, with P<0.05 defining significance.

Results

Patient and sample characteristics

This study enrolled 42 patients with LCNEC. Demographic and clinicopathological characteristics are summarized in Figure 1 and Table S2. The cohort had a median age of 63 years (range, 52–80 years) with male predominance (88.1%, n=37). Three patients (7.1%) reported no smoking history. Histopathology identified 36 cases (85.7%) as pure LCNEC and 6 cases (14.3%) as combined LCNEC-SCLC (cohort partially overlapping with our prior investigation). Tumor localization revealed central lesions in 18 patients (42.9%) and left lung involvement in 20 cases (47.6%). Disease staging distribution was: stage I (52.4%, n=22), stage II (30.9%, n=13), and stage III (16.7%, n=7). Neoadjuvant therapy was administered to 4 patients (9.5%). Adjuvant therapies included chemotherapy (50.0%), radiotherapy (21.4%), and immunotherapy (7.1%); treatment details were unavailable for 7 cases (16.7%).

Figure 1 Demographic and clinicopathological characteristics of LCNEC molecular subtypes. Genetic profiling identified TP53 mutations in 79.5% (31/39) and RB pathway genomic alterations in 46.2% (18/39). Immunohistochemical analysis revealed p53 positive expression in 84.6% (33/39) and Rb protein loss (mutant phenotype) in 46.2% (18/39). Based on these molecular signatures, cases were classified into molecular subtypes: Type I (61.5%, 24/39) and Type II (38.5%, 15/39). Unsupervised hierarchical clustering of transcription markers (ASCL1, NEUROD1, POU2F3, and YAP1) demonstrated concurrent expression of SCLC-defining lineage markers across both molecular subtypes of LCNEC. LCNEC, large cell neuroendocrine carcinoma; NSCLC, non-small cell lung carcinoma; Rb, retinoblastoma; SCLC, small cell lung carcinoma; TMB, tumor mutational burden; TNM, tumor-node-metastasis.

Classifications of LCNEC

Molecular subtyping was performed in 39 cases; 3 histologically ambiguous tumors exhibiting overlapping LCNEC-SCLC features were excluded. Genetic analysis revealed TP53 mutations in 79.5% (31/39) and RB pathway alterations in 46.2% (18/39). IHC showed p53 expression in 84.6% (33/39) and Rb protein loss (mutant phenotype) in 46.2% (18/39). Accordingly, cases were classified into molecular subtypes: Type I (61.5%, 24/39) and Type II (38.5%, 15/39). Unsupervised hierarchical clustering of transcription markers (ASCL1, NEUROD1, POU2F3, and YAP1) demonstrated concurrent expression of SCLC-defining lineage markers in both LCNEC molecular subtypes (Figure 1).

YAP1 immunoreactivity was positive in 71.4% (30/42) of evaluable specimens, defining YAP1-negative (n=12) and YAP1-positive (n=30) cohorts.

Expression of therapeutic biomarkers

As shown in Table S3, IHC revealed differential expression profiles across molecular markers: ASCL1 (45.2%, 19/42), NEUROD1 (21.4%, 9/42), POU2F3 (40.5%, 17/42), DLL3 (54.8%, 23/42), MHC I (69.0%, 29/42), and MHC II (23.8%, 10/42).

PD-L1 was evaluated on both tumor cells and stromal-infiltrating immune cells. The range of TPS and CPS values for PD-L1 in this cohort were 0–45% and 0–55, respectively. The positive expression rate of PD-L1 on tumor cells was 9.5% (4/42), while PD-L1 expression on the stromal cells was more prevalent, occurring in 57.1% (24/42) of samples (Table S3).

Digital spatial analysis of CD8⁺ T cell revealed distinct infiltration patterns: highest density at tumor-invasive margins (3,388.0±1,810.0 cells/mm²), intermediate in stroma (1,506.6±1,237.2 cells/mm²), and lowest in tumor cores (847.0±826.3 cells/mm²) (Figure S1).

Comparative clinicopathological profiling of LCNEC: molecular subtype stratification and YAP1 expression status

Associations between molecular subtypes/YAP1 status and clinicopathological characteristics are analyzed in Table 1. ASCL1 overexpression (P=0.02) and diffuse prominent nucleoli (P=0.04) were significantly enriched in YAP1-negative tumors, whereas NEUROD1-positive tumors demonstrated a YAP1-positive preference (P=0.04). No significant associations were observed between molecular subtypes/YAP1 status and DLL3, PD-L1, MHC I, or MHC II expression. Notably, Rb exhibited marked differential expression across molecular subtypes (P<0.001), while POU2F3 positivity was predominantly observed in Type II LCNEC (P=0.02). TMB and stromal CD8⁺ cell density showed no intergroup variations (TMB: P=0.11; CD8⁺: P=0.51). Inflammatory marker assessment [neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), platelet-to-neutrophil ratio (PNR)] revealed subtype-dependent PLR variation (P=0.048) but no YAP1-related differences. All baseline clinicopathological parameters (sex, age, smoking history, TNM stage, tumor location) showed no significant subgroup differences.

Intergroup comparative analysis (Figure 2): Type II LCNEC showed higher POU2F3 expression (vs. Type I, P=0.04) and TMB-high status (P=0.002). YAP1-negative tumors correlated with elevated ASCL1 (P=0.003) and DLL3 (P=0.003), while YAP1-positive cases were associated with MHC II overexpression (P=0.04). No significant differences in other biomarkers were observed between molecular subtypes or YAP1-defined cohorts (Figure S2).

Figure 2 Intergroup comparative analysis of therapeutic markers. (A) Type II LCNEC demonstrated elevated POU2F3 expression (vs. Type I, P=0.04) and higher TMB (P=0.002). YAP1-negative tumors correlated with upregulated ASCL1 (P=0.003) and DLL3 (P=0.003), whereas YAP1 positivity associated with MHC II overexpression (P=0.04). (B,C) Histopathological micrographs demonstrating a case of Type I LCNEC, and representative images of a case with LCNEC tumors of Type II. The first column panels depict, from top to bottom: HE staining, p53 expression, and Rb expression. The second column panels show ASCL1 expression, NEUROD1 expression, and POU2F3 expression. The third column panels display YAP1 expression, DLL3 expression, and MHC-II expression (×400). LCNEC, large cell neuroendocrine carcinoma; MHC II, major histocompatibility complex class II; TMB, tumor mutation burden.

Association between YAP1 expression and therapeutic biomarkers

Correlations between YAP1 and therapeutic biomarkers were analyzed (Figure 3). YAP1 levels positively correlated with NEUROD1 expression (rho=0.349, P=0.02) but inversely correlated with ASCL1 (rho=−0.326, P=0.04) and MHC I (rho=−0.338, P=0.03). No significant associations were detected between YAP1 and other evaluated biomarkers.

Figure 3 Association between YAP1 expression and therapeutic biomarkers. Correlations between YAP1 immunoreactivity (H-score) and therapeutic biomarkers were investigated. Quantitative analysis revealed YAP1 levels demonstrated a statistically significant positive correlation with NEUROD1 expression (rho=0.349, P=0.02). Conversely, inverse correlations were observed with ASCL1 (rho=−0.326, P=0.04) and MHC I expression (rho=−0.338, P=0.03). No significant associations were detected between YAP1 and other evaluated biomarkers. LMR, lymphocyte-to-monocyte ratio; MHC II, major histocompatibility complex class II; NLR, neutrophil-to-lymphocyte ratio; PD-L1, programmed death-ligand 1; PLR, platelet-to-lymphocyte ratio; PNR, platelet-to-neutrophil ratio; TMB, tumor mutation burden.

Prognostic survival analysis: correlations with clinicopathological characteristics and molecular subtypes

All patients underwent routine clinical surveillance. At data cutoff (December 31, 2024), tumor recurrence/metastasis occurred in 64.3% (27/42) of cases, while 21.4% (9/42) remained disease-free. All-cause mortality was 52.4% (22/42), 31.0% (13/42) survived and 16.7% (7/42) were lost to follow-up. Median OS and DFS were 34 months and 16 months, respectively. Corresponding 1-, 3-, and 5-year actuarial survival rates were 91.4% vs. 52.8%, 40.0% vs. 22.2%, and 22.9% vs. 19.4% for OS and DFS.

Prognostic analyses employed univariate Cox regression (Table S4), followed by multivariate Cox modeling incorporating variables demonstrating univariate significance (P<0.10, Table 2). Multivariate analysis identified early TNM stage as an independent predictor for both OS (HR 0.148; 95% CI: 0.029–0.751; P=0.02) and DFS (HR 0.184; 95% CI: 0.041–0.827; P=0.03). Notably, spread through air spaces (STAS) independently predicted worse OS (HR 0.188; 95% CI: 0.038–0.942; P=0.042) but not DFS (HR 0.438; 95% CI: 0.109–1.769; P=0.25) (Table 2).

Kaplan-Meier methodology generated risk-stratified survival curves for parameters retained in the final multivariate model (Figure 4). Early-stage disease (OS: P=0.004; DFS: P=0.01), absence of vascular invasion (OS: P=0.02; DFS: P=0.03), and STAS-negative status (OS: P<0.001; DFS: P=0.01) demonstrated significant associations with favorable outcomes, while tumor necrosis exceeding 30% correlated with improved OS (P=0.02). Adjuvant chemotherapy correlated with adverse survival outcomes, with treated patients demonstrating poorer prognosis (P=0.04). ASCL1 protein expression showed positive correlation with DFS advantage (P=0.02), while DLL3 expression and stromal CD8⁺ cell density showed no survival associations.

Figure 4 Kaplan-Meier risk-stratified survival curves for final multivariate model parameters. Kaplan-Meier analysis was used to compare the risk stratification of subgroups according to these prognostic factors, which were included in multivariate analysis. The results showed that early-stage disease (OS: P=0.004; DFS: P=0.01), absence of vascular invasion (OS: P=0.02; DFS: P=0.03), and STAS-negative status (OS: P<0.001; DFS: P=0.01) demonstrated significant associations with favorable outcomes. Paradoxically, tumor necrosis exceeding 30% correlated with improved OS (P=0.02). Adjuvant chemotherapy exhibited adverse survival outcomes, with treated patients demonstrating poorer prognosis (P=0.04). ASCL1 protein expression showed positive correlation with DFS advantage (P=0.02), while DLL3 expression and stromal CD8⁺ lymphocyte infiltration lacked statistical significance in survival outcomes. DFS, disease-free survival; OS, overall survival; STAS, spread through air spaces; TNM, tumor-node-metastasis.

Subgroup analysis revealed no survival differences by molecular subtypes (Type I vs. II) or YAP1 status (Figure 5). Kaplan-Meier analysis of molecular subtypes showed overlapping survival curves for both OS and DFS (OS: P=0.44; DFS: P=0.63). Similarly, YAP1-negative and YAP1-positive cohorts demonstrated comparable survival outcomes (OS: P=0.46; DFS: P=0.85). These findings underscore TNM staging, vascular invasion status, and STAS presence as critical determinants of clinical outcomes in this cohort.

Figure 5 Kaplan-Meier analysis in relation to molecular subtypes (Type I vs. II) and YAP1 expression status. Kaplan-Meier analysis of molecular subtypes showed overlapping survival curves for both OS and DFS (OS: P=0.44; DFS: P=0.63). Similarly, YAP1-negative and YAP1-positive cohorts demonstrated comparable survival outcomes (OS: P=0.46; DFS: P=0.85). DFS, disease-free survival; OS, overall survival.

Discussion

LCNEC represents a rare, aggressive high-grade neuroendocrine malignancy characterized by marked molecular heterogeneity. Current classification stratifies LCNEC into Type I (lacking concomitant TP53/RB1 aberrations) and Type II (TP53/RB1 co-mutated) subtypes (4,5), offering preliminary frameworks for prognostic prediction and targeted therapy. Emerging evidence proposes YAP1 expression-based stratification to refine molecular classification and guide personalized treatment. However, the clinical utility and prognostic significance of these classification systems remain uncertain and require validation through clinical data. Given no consensus therapeutic guidelines, recent clinical investigations demonstrate promising responses to immune checkpoint inhibitors (ICIs) and targeted therapies. The identification of predictive biomarkers and refined molecular classification of treatment-responsive subgroups is essential for optimizing precision oncology in LCNEC.

This retrospective cohort study analyzed archived surgical specimens from pathologically confirmed LCNEC patients. Tumors were stratified into molecular subtypes (Type I/II) and evaluated for YAP1 expression using NGS and IHC. Notably, our findings revealed no significant association between YAP1 expression status and molecular subtypes (P=0.15), contrasting with Stewart et al.’s report of distinct clinicopathological-genomic profiles: YAP1-high tumors exhibited mesenchymal phenotype, inflammatory microenvironment, and genomic features characterized by TP53 mutations with concurrent CDKN2A/B and SMARCA4 alterations—hallmarks of Type I LCNEC, and YAP1-low tumors demonstrated epithelial differentiation, immune-cold microenvironment, and predominant TP53/RB1 co-mutations—features aligning with Type II LCNEC (18). The observed discrepancies may reflect methodological differences: sample source heterogeneity (cell lines/human tumors vs. our human-only cohort) and specimen limitations (core needle biopsies vs. surgical specimens in our IHC analysis). These findings highlight the critical need for standardized evaluation protocols incorporating large-scale surgical specimens with comprehensive molecular profiling to elucidate LCNEC biology. Further independent validation studies are warranted.

Key findings from our study merit emphasis: (I) ASCL1 expression demonstrated significant enrichment in YAP1-negative tumors (P=0.02), while NEUROD1-positive tumors showed YAP1 co-expression (P=0.04). (II) YAP1 and NEUROD1 expression correlated positively (rho=0.349, P=0.02), in contrast to the inverse correlation between YAP1 and ASCL1 expression (rho=−0.326, P=0.04). (III) YAP1-negative tumors exhibited elevated ASCL1 (P=0.003) and DLL3 (P=0.003) expression, whereas YAP1-positive tumors showed upregulated MHC II levels (P=0.04). These differential expression patterns suggest YAP1-positive LCNECs may demonstrate enhanced response to immunotherapies or NEUROD1-targeted therapies, while YAP1-negative tumors could be more responsive to ASCL1/DLL3-targeted approaches. These findings require validation in prospective clinical trials.

Supporting these hypotheses, Stewart et al. proposed that the YAP1-high subtype demonstrates enhanced response to ICIs and molecularly targeted therapies addressing specific genomic alterations, whereas the YAP1-low subtype may show better response to conventional SCLC chemotherapies (18). Complementing this, Owonikoko et al. documented distinct immunogenomic features in SCLC-Y, characterized by elevated interferon-γ response genes expression, T-cell inflammatory signatures, HLA complex components, and T-cell receptor genes (27). Notably, our study revealed paradoxical immunophenotypic associations: MHC I expression inversely correlated with YAP1 levels (rho=−0.338, P=0.03), while MHC II expression positively associated with YAP1 (P=0.04). This discordance is significant given their synergistic roles in antigen presentation—MHC I activates CD8⁺ cytotoxic T-cells, while MHC II engages CD4⁺ helper T-cells and typically drives more robust humoral responses, potentially enhancing immunotherapy efficacy. Technical factors may contribute: MHC I detection used a polyclonal antibody with nonspecific immunoreactivity in IHC validation, whereas the monoclonal MHC II assay demonstrated enhanced specificity. These divergent correlations underscore the need for standardized validation using monoclonal antibodies and multiplex immunoassays, combined with mechanistic studies to elucidating YAP1’s immunomodulatory role in LCNEC pathogenesis.

The incidence of actionable molecular alterations is significantly lower in LCNEC than in classical NSCLC. DLL3, an inhibitory Notch pathway ligand, exhibits restricted normal tissues expression but frequent overexpression in SCLC and LCNEC (28). The recent FDA approval of tarlatamab, a DLL3/CD3 bispecific T-cell engager (BiTE), for relapsed SCLC, underscores DLL3’s therapeutic relevance (29), supporting its exploration in LCNEC. Notably, our analysis revealed no significant associations between DLL3 expression and p53/Rb protein expression or TP53/RB1 mutational status (Table S5). This aligns with Brcic et al.’s finding of no DLL3-TP53/RB1 correlation across neuroendocrine carcinomas (30). Contrastingly, Hermans et al. reported higher DLL3 expression in 89% (8/9) of TP53 wild-type LCNEC versus 50% (29/58) of TP53-mutant cases (P=0.035), though they similarly found no association with RB1 status (31). Critically, we identified a significant inverse correlation between YAP1 expression and DLL3 levels in LCNEC tumors (P=0.003). This finding suggests that YAP1-based molecular classification may better inform DLL3-targeted therapies in LCNEC than mutational subtyping.

Our study demonstrated a 71.4% (30/42) YAP1 positivity rate in LCNEC specimens, significantly exceeding reported SCLC rates (19) and aligning with prior observations (32). This differential expression may reflect two mechanisms: First, YAP1 downregulation has been linked to enhanced neuroendocrine marker expression, suggesting a potential role in suppressing neuroendocrine differentiation (15). This inverse relationship aligns with LCNEC’s comparatively reduced neuroendocrine phenotype versus SCLC. Second, molecular parallels between LCNEC and NSCLC—where YAP1 amplification/overexpression is frequent—may contribute to this phenomenon (15). Notably, we observed an inverse correlation between YAP1 and ASCL1 expressions (rho=−0.326, P=0.04), contrasting with a positive association between YAP1 and NEUROD1 (rho=0.349, P=0.02). However, no significant differences were observed in neuroendocrine marker profiles (synaptophysin/chromogranin/CD56) or TTF-1 expression between YAP1-positive and YAP1-negative cohorts (P>0.05 by χ² test). These findings highlight the necessity for mechanistic studies elucidating YAP1’s precise role in LCNEC neuroendocrine differentiation pathways.

While prospective evidence for ICIs in LCNEC remains limited, retrospective analyses and clinical observations suggest therapeutic potential (33-35). Validated biomarkers for patient stratification are critical for optimizing immunotherapy outcomes. Although PD-L1 expression and TMB have demonstrated predictive value for ICI response in other malignancies (36,37), and CD8⁺ tumor-infiltrating lymphocytes (TILs) show promise as potential biomarkers (38), their prognostic significance in LCNEC remains undefined. Our analysis revealed Type II LCNEC tumors exhibited significantly elevated TMB levels (P=0.002), while PD-L1 expression and YAP1 protein patterns showed no intergroup variation. Survival analyses demonstrated no significant prognostic associations between PD-L1 expression (assessed by TPS/CPS), TMB, or CD8⁺ TIL density and clinical outcomes. Notably, we identified positive correlations between CD8⁺ TIL infiltration and PD-L1 expression (TPS: rho=0.359, P=0.02; CPS: rho=0.331, P=0.03), with no other significant biomarker interrelationships observed (Figure S3). These findings underscore the imperative for prospective validation of PD-L1, TMB, and CD8⁺ TILs to establish their predictive and prognostic utility in LCNEC immunotherapy.

Consistent with established literature, our findings confirm that STAS presence, vascular invasion, and advanced TNM stage independently predict adverse survival—well-characterized prognostic determinants in LCNEC and other pulmonary malignancies (39). Notably, adjuvant chemotherapy was associated with a survival disadvantage (P=0.04), potentially attributable to LCNEC’s intrinsic tumor aggressiveness and refractory nature to conventional cytotoxic regimens. This underscores the urgent need to explore ICIs and molecularly targeted therapies in LCNEC management. While YAP1 overexpression correlates with poor outcomes in other malignancies (16,40-42), our survival analysis (P=0.32) and recent studies reveal divergent findings (17,32). Furthermore, no significant inter-subtype (Type I vs. Type II) survival disparities emerged in our cohort (log-rank P=0.44). However, caution should be taken when interpreting the prognostic significance of molecular subtyping and YAP1 status, given current therapeutic heterogeneity in LCNEC management. Prospective trials incorporating next-generation therapies are critically needed to establish validated prognostic biomarkers.

There are several limitations in this study. First, the single-institutional retrospective design introduces inherent selection bias. Second, although we assembled a substantial cohort of surgically resected LCNEC specimens, the sample size remains modest, potentially limiting statistical power. Third, clinical unavailability of DLL3-targeted agents and ICIs precluded assessment of molecular subtypes and YAP1 expression patterns as predictive biomarkers for treatment response. The potential utility of YAP1-based classification for guiding personalized LCNEC therapy requires validation in rigorously designed clinical trials. Future investigations should prioritize multi-center prospective cohorts and large-scale clinical trials to corroborate these findings.

Conclusions

In summary, our cohort analysis revealed no significant differential YAP1 expression between LCNEC molecular subtypes (Type I vs. Type II). YAP1-based classification offers greater clinical utility than molecular subtyping for guiding DLL3-targeted therapies. Notably, YAP1-positive LCNEC tumors may respond better to immunotherapy or NEUROD1-targeted therapies, while YAP1-negative tumors appear more amenable to ASCL1- or DLL3-targeted approaches. Importantly, stratification by molecular subtypes (Type I/II) or YAP1 expression status showed no significant prognostic significance in our cohort. Validation of these findings requires multi-institutional prospective studies and large-scale clinical trials.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the REMARK reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-491/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-491/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-491/prf

Funding: This work was supported by National Natural Science Foundation of China (82141117), Beijing Natural Science Foundation (L248048), Wu Jieping Medical Foundation (320.6750.2021-16-19), Clinical Research Fund for Distinguished Young Scholars of Beijing Cancer Hospital (QNJJ202212, QNJJ202403), Guangzhou Life Oasis Public Service Center Research and Exchange Program in the Field of Health (1-35), Beijing Xisike Clinical Oncology Research Foundation (Y-pierrefabre202101-0099), Beijing Vlove Charity Foundation (JYKY2024-0200426027) and the Science Foundation of Peking University Cancer Hospital (PY202303).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-491/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 Ethics Committee of Peking University Cancer Hospital and the local review board (No. 2023KT23), and individual consent for this retrospective analysis was waived.

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. Hiroshima K, Mino-Kenudson M. Update on large cell neuroendocrine carcinoma. Transl Lung Cancer Res 2017;6:530-9. [Crossref] [PubMed]
2. Kinslow CJ, May MS, Saqi A, et al. Large-Cell Neuroendocrine Carcinoma of the Lung: A Population-Based Study. Clin Lung Cancer 2020;21:e99-e113. [Crossref] [PubMed]
3. Baine MK, Rekhtman N. Multiple faces of pulmonary large cell neuroendocrine carcinoma: update with a focus on practical approach to diagnosis. Transl Lung Cancer Res 2020;9:860-78. [Crossref] [PubMed]
4. Rekhtman N, Pietanza MC, Hellmann MD, et al. Next-Generation Sequencing of Pulmonary Large Cell Neuroendocrine Carcinoma Reveals Small Cell Carcinoma-like and Non-Small Cell Carcinoma-like Subsets. Clin Cancer Res 2016;22:3618-29. [Crossref] [PubMed]
5. George J, Walter V, Peifer M, et al. Integrative genomic profiling of large-cell neuroendocrine carcinomas reveals distinct subtypes of high-grade neuroendocrine lung tumors. Nat Commun 2018;9:1048. [Crossref] [PubMed]
6. Derks JL, Leblay N, Thunnissen E, et al. Molecular Subtypes of Pulmonary Large-cell Neuroendocrine Carcinoma Predict Chemotherapy Treatment Outcome. Clin Cancer Res 2018;24:33-42. [Crossref] [PubMed]
7. Simbolo M, Barbi S, Fassan M, et al. Gene Expression Profiling of Lung Atypical Carcinoids and Large Cell Neuroendocrine Carcinomas Identifies Three Transcriptomic Subtypes with Specific Genomic Alterations. J Thorac Oncol 2019;14:1651-61. [Crossref] [PubMed]
8. Zhuo M, Guan Y, Yang X, et al. The Prognostic and Therapeutic Role of Genomic Subtyping by Sequencing Tumor or Cell-Free DNA in Pulmonary Large-Cell Neuroendocrine Carcinoma. Clin Cancer Res 2020;26:892-901. [Crossref] [PubMed]
9. Rudin CM, Poirier JT, Byers LA, et al. Molecular subtypes of small cell lung cancer: a synthesis of human and mouse model data. Nat Rev Cancer 2019;19:289-97. [Crossref] [PubMed]
10. Gay CM, Stewart CA, Park EM, et al. Patterns of transcription factor programs and immune pathway activation define four major subtypes of SCLC with distinct therapeutic vulnerabilities. Cancer Cell 2021;39:346-360.e7. [Crossref] [PubMed]
11. Mollaoglu G, Guthrie MR, Böhm S, et al. MYC Drives Progression of Small Cell Lung Cancer to a Variant Neuroendocrine Subtype with Vulnerability to Aurora Kinase Inhibition. Cancer Cell 2017;31:270-85. [Crossref] [PubMed]
12. Poirier JT, George J, Owonikoko TK, et al. New Approaches to SCLC Therapy: From the Laboratory to the Clinic. J Thorac Oncol 2020;15:520-40. [Crossref] [PubMed]
13. Schwendenwein A, Megyesfalvi Z, Barany N, et al. Molecular profiles of small cell lung cancer subtypes: therapeutic implications. Mol Ther Oncolytics 2021;20:470-83. [Crossref] [PubMed]
14. Popper H, Brcic L, Eidenhammer S. Does subtyping of high-grade pulmonary neuroendocrine carcinomas have an impact on therapy selection? Transl Lung Cancer Res 2023;12:2412-26. [Crossref] [PubMed]
15. Ito T, Matsubara D, Tanaka I, et al. Loss of YAP1 defines neuroendocrine differentiation of lung tumors. Cancer Sci 2016;107:1527-38. [Crossref] [PubMed]
16. Sun X, Zhang J, Dong J, et al. Prognostic significance of YAP1 expression and its association with neuroendocrine markers in resected pulmonary large cell neuroendocrine carcinoma (LCNEC). Transl Oncol 2022;25:101538. [Crossref] [PubMed]
17. Liu T, Chen X, Mo S, et al. Molecular subtypes and prognostic factors of lung large cell neuroendocrine carcinoma. Transl Lung Cancer Res 2024;13:2222-35. [Crossref] [PubMed]
18. Stewart CA, Diao L, Xi Y, et al. YAP1 Status Defines Two Intrinsic Subtypes of LCNEC with Distinct Molecular Features and Therapeutic Vulnerabilities. Clin Cancer Res 2024;30:4743-54. [Crossref] [PubMed]
19. Zhu Y, Li S, Wang H, et al. Molecular subtypes, predictive markers and prognosis in small-cell lung carcinoma. J Clin Pathol 2024;78:42-50. [Crossref] [PubMed]
20. Goldstraw P, Chansky K, Crowley J, et al. The IASLC Lung Cancer Staging Project: Proposals for Revision of the TNM Stage Groupings in the Forthcoming (Eighth) Edition of the TNM Classification for Lung Cancer. J Thorac Oncol 2016;11:39-51. [Crossref] [PubMed]
21. Iams WT, Porter J, Horn L. Immunotherapeutic approaches for small-cell lung cancer. Nat Rev Clin Oncol 2020;17:300-12. [Crossref] [PubMed]
22. Kulangara K, Zhang N, Corigliano E, et al. Clinical Utility of the Combined Positive Score for Programmed Death Ligand-1 Expression and the Approval of Pembrolizumab for Treatment of Gastric Cancer. Arch Pathol Lab Med 2019;143:330-7. [Crossref] [PubMed]
23. Bankhead P, Loughrey MB, Fernández JA, et al. QuPath: Open source software for digital pathology image analysis. Sci Rep 2017;7:16878. [Crossref] [PubMed]
24. Fedchenko N, Reifenrath J. Different approaches for interpretation and reporting of immunohistochemistry analysis results in the bone tissue - a review. Diagn Pathol 2014;9:221. [Crossref] [PubMed]
25. Qu S, Fetsch P, Thomas A, et al. Molecular Subtypes of Primary SCLC Tumors and Their Associations With Neuroendocrine and Therapeutic Markers. J Thorac Oncol 2022;17:141-53. [Crossref] [PubMed]
26. Zhu Y, Wu J, Wang H, et al. Whole-section digital analysis of immune profiles in surgically resected small cell lung carcinoma and their associations with molecular subtypes. Transl Lung Cancer Res 2025;14:449-66. [Crossref] [PubMed]
27. Owonikoko TK, Dwivedi B, Chen Z, et al. YAP1 Expression in SCLC Defines a Distinct Subtype With T-cell-Inflamed Phenotype. J Thorac Oncol 2021;16:464-76. [Crossref] [PubMed]
28. Saunders LR, Bankovich AJ, Anderson WC, et al. A DLL3-targeted antibody-drug conjugate eradicates high-grade pulmonary neuroendocrine tumor-initiating cells in vivo. Sci Transl Med 2015;7:302ra136. [Crossref] [PubMed]
29. Dhillon S. Tarlatamab: First Approval. Drugs 2024;84:995-1003. [Crossref] [PubMed]
30. Brcic L, Kuchler C, Eidenhammer S, et al. Comparison of four DLL3 antibodies performance in high grade neuroendocrine lung tumor samples and cell cultures. Diagn Pathol 2019;14:47. [Crossref] [PubMed]
31. Hermans BCM, Derks JL, Thunnissen E, et al. DLL3 expression in large cell neuroendocrine carcinoma (LCNEC) and association with molecular subtypes and neuroendocrine profile. Lung Cancer 2019;138:102-8. [Crossref] [PubMed]
32. Li M, Wang X, Gong J, et al. The analysis of molecular classification of pulmonary neuroendocrine tumors and relationship between YAP1 and efficacy. Invest New Drugs 2025;43:108-17. [Crossref] [PubMed]
33. Vrontis K, Economidou SC, Fotopoulos G. Platinum Doublet plus Atezolizumab as First-line Treatment in Metastatic Large Cell Neuroendocrine Carcinoma: A Single Institution Experience. Cancer Invest 2022;40:124-31. [Crossref] [PubMed]
34. Sherman S, Rotem O, Shochat T, et al. Efficacy of immune check-point inhibitors (ICPi) in large cell neuroendocrine tumors of lung (LCNEC). Lung Cancer 2020;143:40-6. [Crossref] [PubMed]
35. Naganuma K, Imai H, Yamaguchi O, et al. Efficacy and Safety of Anti-Programed Death-1 Blockade in Previously Treated Large-Cell Neuroendocrine Carcinoma. Chemotherapy 2021;66:65-71. [Crossref] [PubMed]
36. Havel JJ, Chowell D, Chan TA. The evolving landscape of biomarkers for checkpoint inhibitor immunotherapy. Nat Rev Cancer 2019;19:133-50. [Crossref] [PubMed]
37. Marabelle A, Fakih M, Lopez J, et al. Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study. Lancet Oncol 2020;21:1353-65. [Crossref] [PubMed]
38. Wong PF, Wei W, Smithy JW, et al. Multiplex Quantitative Analysis of Tumor-Infiltrating Lymphocytes and Immunotherapy Outcome in Metastatic Melanoma. Clin Cancer Res 2019;25:2442-9. [Crossref] [PubMed]
39. Han YB, Kim H, Mino-Kenudson M, et al. Tumor spread through air spaces (STAS): prognostic significance of grading in non-small cell lung cancer. Mod Pathol 2021;34:549-61. [Crossref] [PubMed]
40. Lee KW, Lee SS, Kim SB, et al. Significant association of oncogene YAP1 with poor prognosis and cetuximab resistance in colorectal cancer patients. Clin Cancer Res 2015;21:357-64. [Crossref] [PubMed]
41. Wu Y, Hou Y, Xu P, et al. The prognostic value of YAP1 on clinical outcomes in human cancers. Aging (Albany NY) 2019;11:8681-700. [Crossref] [PubMed]
42. Guo L, Chen Y, Luo J, et al. YAP1 overexpression is associated with poor prognosis of breast cancer patients and induces breast cancer cell growth by inhibiting PTEN. FEBS Open Bio 2019;9:437-45. [Crossref] [PubMed]