Peng Lin

Background: This research focused on evaluating the utility of multimodal radiomics integrated with machine learning to predict pathological complete response (pCR) in a prospective cohort of esophageal squamous cell carcinoma (ESCC) patients undergoing neoadjuvant immunochemotherapy (nICT). Methods: We retrospectively analyzed prospectively collected trial data from 66 ESCC patients. Radiomic features were extracted from computed tomography (CT) and magnetic resonance imaging (MRI) images. Four machine learning algorithms—Random Forest (RF), logistic regression, Support Vector Machine, and Extreme Gradient Boosting (XGBoost)—were applied with leave-one-out cross-validation to predict pCR after nICT. The predictive performance of the models was evaluated using receiver operating characteristic curve analysis. Results: In total, 851 features were identified. Among the four machine learning algorithms, the XGBoost machine learning method demonstrated the best model performance across CT, MRI, and clinical feature-based models. Furthermore, the integrated model demonstrated superior performance compared to individual models based solely on CT, MRI, or clinical features across all machine learning algorithms. Among these, the XGboost-based integrated model achieved the highest performance on the test set, with an AUC of 0.961, a TPR of 84.2%, a TNR of 95.7%, a PPV 88.9% of and a NPV of 93.8%. Decision curve analysis validated the model’s robust clinical utility, with calibration curves demonstrating strong concordance between predicted and observed therapeutic responses. Conclusions: The study demonstrates the potential for predicting pCR in patients with ESCC treated with standardized neoadjuvant chemotherapy and PD-1 inhibitors using machine learning methods that integrate multimodal CT and MRI images with clinical features.

Objectives: This research aimed to examine the relationships between clinicopathological characteristics and the occurrence of Spread Through Air Spaces (STAS) in patients with stage IA lung adenocarcinoma (LUAD) and to develop a preoperative prediction model. Methods: Data from 1,375 patients with stage IA LUAD at Sun Yat-sen University Cancer Center were analyzed. Propensity score matching (PSM) was employed to match 141 STAS-positive patients with 282 STAS-negative patients. Both univariate and multivariate logistic regression analyses were performed to determine independent variables among 16 clinicopathological and 13 CT imaging characteristics. A nomogram prediction model was developed and evaluated via receiver operating characteristic (ROC) and decision curve analyses (DCAs). Results: Multivariate analysis identified several independent risk factors. Irregular nodule shape (OR = 1.817, 95% CI: 1.106–2.986, p = 0.018), irregular margin (OR = 2.050, 95% CI: 1.218–3.449, p = 0.007), lobulation (OR = 2.235, 95% CI: 1.336–3.739, p = 0.002), and vascular convergence (OR = 5.032, 95% CI: 2.050–12.349, p < 0.001) were significantly associated with an increased risk of STAS. Compared with a consolidation tumor ratio (CTR) = 0% (reference), a CTR of 75–100% (OR = 7.086, 95% CI: 2.542–19.750, p < 0.001) and a CTR = 100% (OR = 11.502, 95% CI: 4.752–27.840, p < 0.001) were significantly associated with an increased risk of STAS. The nomogram was developed and internally validated, demonstrating good predictive accuracy (AUC = 0.812, 95% CI: 0.761–0.863) and favorable clinical utility, with a sensitivity of 69.5% and a specificity of 80.2%. Conclusion: The nomogram reliably predicts STAS preoperatively and may assist in guiding surgical decision-making.

Alternative polyadenylation (APA) is critical for shaping transcriptome diversity and modulating cancer therapeutic resistance. While lactate is a well-established metabolic signal in cancer progression, its role in APA regulation remains unclear. Here, we demonstrate that l-lactate-induced lactylation of NUDT21 drives transcriptomic reprogramming through APA modulation. NUDT21 lactylation enhances its interaction with CPSF6, facilitating CFIm complex formation and inducing 3′ untranslated region (UTR) lengthening of FDX1. Extension of the FDX1 3′ UTR attenuates its protein output, thereby conferring resistance to cuproptosis in esophageal squamous cell carcinoma (ESCC). Furthermore, we identify AARS1 as the lactylation “writer” catalyzing NUDT21 K23 lactylation, and HDAC2 as its enzymatic “eraser”. Clinically, elevated levels of both LDHA and NUDT21, as well as increased K23-lactylated NUDT21, are associated with reduced FDX1 expression and worse prognosis in ESCC patients. Notably, combined targeting of the lactate-NUDT21-FDX1-cuproptosis axis with the clinical LDHA inhibitor stiripentol and the copper ionophore elesclomol synergistically suppressed tumor growth. Collectively, our work identifies lactylated NUDT21 as a critical factor linking cellular metabolism to APA and proposes a promising therapeutic strategy for ESCC treatment.

Background: Immune checkpoint inhibitor (ICI) monotherapy and neoadjuvant immunochemotherapy have shown promising results in esophageal carcinoma. However, it is still unclear whether more courses of immunochemotherapy are therapeutically better. We aimed to investigate the safety and efficacy of three courses of neoadjuvant treatment for patients with locally advanced esophageal squamous cell carcinoma (ESCC).Methods: Patients with locally advanced ESCC received three courses of camrelizumab plus nab‐paclitaxel and capecitabine before undergoing surgery. Additionally, patients received safety, computed tomography (CT), and endoscopy (with endoscopic ultrasonography and mucosal biopsy) assessments before and in the second and third courses of treatment. We used the CT and endoscopic assessment results from the second and third courses for comparison.Results: From May 2020 to December 2021, 47 patients were enrolled at Sun Yat‐sen University Cancer Center. In our study, 43 patients completed three courses of preoperative chemotherapy combined with anti‐Programmed cell death‐1 (PD‐1) therapy and radical surgical resection. The toxicity of the third course of immunochemotherapy was mild and well tolerated without increased treatment‐related adverse events (TRAEs) and mortality compared with that of the second course of treatment. In terms of efficacy, an additional course of treatment after the second course of treatment was effective, with increased CT and endoscopy T (clinical T stage) downstaging rates by 16.3% and 25.9%, N (clincial N stage) downstaging rates by 7.0% and 11.1%, and objective response rates (ORRs) by 13.6% and 22.0%, respectively.Conclusions: Regardless of downstaging or ORR, three courses of immunochemotherapy appear to be superior to two courses of treatment without increasing TRAEs.

Introduction: the investigation on the interactions between ferroptosis and lncRNAs for lung squamous cell carcinoma (LUSC) has been scare, and its impact on tumor immune microenvironment remained unknown. We aim to not only identify a ferroptosis-related lncRNAs signature for LUSC prognosis, but also evaluate its correlation to tumor immune evasion.Methods: RNA sequencing data and survival information were obtained from The Cancer Genome Atlas database. A ferroptosis-related lncRNAs signature (FerRLSig) was developed and validated by univariate Cox regression, Least Absolute Shrinkage and Selection Operator regression and multivariate Cox regression. The tumor immune microenvironment and immune evasion were subsequently evaluated based on the FerRLSig stratification.Results: the FerRLSig consisted of 10 ferroptosis-related lncRNAs and significantly associated with overall survival with satisfactory area under curve (HR = 2.240, 95% CI: 1.845–2.720, p < 0.001, 5-years AUC: 0.756). Based on the FerRLSig stratification, the high-risk group demonstrated not only significantly higher immune infiltration, but also more profound T cell dysfunction and immune evasion, which might ultimately lead to the resistance to current immune checkpoint inhibitors.Conclusion: a robust prognostic FerRLSig for LUSC has been developed and validated, demonstrating a close association not only with tumor immune cell infiltration, but also with T cell dysfunction and immune evasion. Further investigation is warranted to better improve the survival of LUSC patients based on the FerRLSig stratification.

Background: Malignant pleural mesothelioma (MPM) is a malignant tumor originating from pleural mesothelial cells and has a high mortality rate worldwide. With the advent of immunotherapy in MPM treatment, there is an urgent need to elucidate the immune-related mechanisms in this caner.Methods: Single-sample gene set enrichment analysis (ssGSEA) was used to score the immunocytes infiltration of data from different database sources. Identification of immunocyte-related genes was performed with weighted gene co-expression network analysis (WGCNA), differentially expressed genes (DEGs) analysis, and correlation analysis. Pan-caner analysis was performed using “DiffExp” and “Correlation” modules in TIMER.Results: T-helper 2 (Th2) cell was found to be a poor prognostic factor for patients with MPM. Then a transcription factor, NFE2L3, was identified as a biomarker that showed a strong positive correlation with Th2 cell infiltration, and was highly expressed in MPM tissues and was related to the poor prognosis of these patients. At the same time, multiple NFE2L3 methylation sites were negatively correlated with Th2 cell infiltration, and patients with a high degree of methylation enjoy a better prognosis. Pan-caner analysis indicated that NFE2L3 might promote the differentiation of Th2 cells through the IL-2/STAT5/NLRP3 signaling pathway in MPM and many other cancers.Conclusion: We believe that NFE2L3 can serve as a potential biomarker related to the diagnosis and prognosis of patients with MPM, and speculate that NFE2L3 could promote Th2 cell differentiation via IL-2/STAT5/NLRP3 signaling pathway in MPM and many other cancers.

Background The role and upstream regulation of mast cells (a crucial constituent of the tumor immune microenvironment) in early-stage lung adenocarcinoma (LUAD) remains unclear. Methods Using CIBERSORT, we analyzed the expression profile downloaded from the NCBI Gene Expression Omnibus (GEO) database. Next, using weighted gene coexpression network analysis (WGCNA) and enrichment analysis, we identified upstream regulators of mast cells. Finally, we verified their relevance using data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases. Results Among patients with early-stage LUAD, those with high mast cell infiltration levels had a better prognosis, and mast cell infiltration levels were closely associated with the mutation status of TP53, STK11 and EGFR. CAMK2D, AURKA, AURKB, PLK1 and EZH2 showed strong correlations with mast cell infiltration and a key growth factor of mast cells (KITLG), and these 5 regulators were closely correlated with the survival of patients with early-stage lung adenocarcinoma. Conclusion Mast cells play a protective role in patients with early-stage LUAD. We speculate that CAMK2D, AURKA, AURKB, PLK1 and EZH2 promote or inhibit the infiltration of mast cells in early-stage lung adenocarcinoma tissues by regulating KITLG, likely serving as novel targets in immunotherapy for LUAD.

Oesophageal squamous cell carcinoma (ESCC) is a malignant tumour of the digestive system that is associated with high morbidity and mortality rates worldwide. With the increased use of immunotherap...

BackgroundEsophageal squamous cell carcinoma (ESCC) is a kind of digestive system malignant tumor with high morbidity and mortality worldwide. With the rise of immunotherapy applied in cancer treatment, immune-related mechanisms in ESCC and other digestive system carcinomas are in urgent need of being detected.MethodsIn our study, single-sample gene set enrichment analysis (ssGSEA) was performed at first to analyze the expression profile downloaded from NCBI Gene Expression Omnibus (GEO) database. Then via a series of bioinformatic analyses, including Mann-Whitney test, weighted gene co-expression network analysis (WGCNA), functional enrichment analysis and differentially expressed genes (DEGs) analysis, we identified targeting immunocyte and related genes. Finally, we validated the results in TIMER database.ResultsOur analyses showed that macrophage infiltrating level is obviously higher in advanced stages in ESCC compared with other types of immunocytes. MEOX2 was detected as a biomarker correlated with macrophage infiltration in ESCC and other types of digestive system carcinomas. And MEOX2 expression was strongly associated with mRNA expression of colony-stimulating factor 1 (CSF-1) and CSF-1 receptor (CSF-1R) in these kinds of carcinomas. ConclusionWe speculated that MEOX2 could facilitate macrophage infiltration via CSF-1/CSF-1R signaling in ESCC and other kinds of digestive system carcinomas, which might serve as a novel target in prospective tumor immunotherapy.

Primary small cell carcinoma of the esophagus (SCCE) is a rare and extremely fatal disease. We aim to evaluate the efficacy of radical surgery for resectable SCCE and to explore potential prognostic factors. We retrospectively reviewed 52 consecutive SCCE patients who underwent radical surgery from February 1993 to November 2014 at a single institution. The Kaplan-Meier estimator with log-rank test was used to assess overall survival (OS), disease-free survival (DFS) and median survival time. Univariate and multivariable analyses were used to evaluate prognostic factors through Cox proportional hazard regression model. Twenty-five (48.1%) patients were treated with surgery alone, whereas 27 (51.9%) patients underwent adjuvant therapy after surgery. The median OS time was 17.4 months (95% CI: 13.5-21.3). The median DFS time was 13.4 months (95% CI: 7.7-19.0). Patients whose tumors were located in the lower part of thoracic esophagus and the esophagogastric junction showed significantly better OS (27.0 vs. 13.2 months, P = 0.016) and DFS (27.0 vs. 11.3 months, P = 0.017) than those located in the upper and middle parts. Patients with N0 status experienced significantly better OS (21.4 vs. 11.6 months, P = 0.012) and DFS (21.4 vs. 8.6 months, P = 0.012) than those with N+ status. Patients whose tumor lengths were shorter than 5 cm had a better OS (17.4 vs. 5.7 months, P = 0.035) than those longer than 5 cm. Patients who underwent chemotherapy experienced a significantly improved OS (21.0 vs. 14.1 months, P = 0.032) compared to surgery alone. Multivariable analysis showed that lower tumor location, shorter tumor length, pN0 status and chemotherapy independently predicted better OS; lower tumor location and pN0 status independently predicted better DFS. Radical surgery in combination with chemotherapy has better outcomes than surgery alone for resectable SCCE. Higher tumor location, longer tumor length, lymph node metastasis and not undergoing chemotherapy independently predict worse prognoses.