Abstract Objectives: Prognostication of spinal cord injury (SCI) is vital, especially for critical patients who need intensive care. The study aims to develop machine-learning (ML) classifiers for discharge prediction of SCI patients in the intensive care unit (ICU). Methods: Clinical data of patients diagnosed with SCI were extracted from the publicly available ICU database. A total of 105 ML classifiers were initially developed to predict the discharge destination (dead, further medical care, home), and then the top 3 classifiers with the best performance were stacked into an ensemble classifier (Esb-Clf). To balance the accuracy and the feasibility, the complete Esb-Clf was finally simplified with top 10 features (simplified Esb-Clf). The micro-average area under the curve (AUC) was used to compare the prediction performance of difference ML classifiers and 6 doctors' artificial prediction. Results: A total of 1485 SCI patients were used for the early and the recent prediction of discharge destination. In the early prediction, the micro-average AUC of the Esb-Clf and the simplified Esb-Clf was 0.846 and 0.835 during the independent testing, respectively. While in the recent prediction, the micro-average AUC of the Esb-Clf and the simplified Esb-Clf was 0.898 and 0.892. Performance of both the Esb-Clf and the simplified Esb-Clf were superior to the doctors' in the early and the recent prediction. Conclusions: ML classifiers can discriminate the discharge destination of SCI patients with high accuracy, feasibility and interpretability. Whether the simplified Esb-Clf as an online predictive tool is applicable to guiding clinical management needs further verification.

Background This study aimed to investigate the relationship between signal regulatory protein gamma (SIRPG) and tumor immune microenvironment phenotypes or T cell mediated-adaptive antitumor immunity, and its predictive value for response to PD-1 blockade in cancers.Methods Pan-cancer analysis of SIRPG expression and immune deconvolution was performed using transcriptomic data across 33 tumor types. Transcriptomic and clinical data from 157 patients with non-small-cell lung cancer (NSCLC) and melanoma received PD-1 blockade were analyzed. Expression characteristics of SIRPG were investigated using single-cell RNA sequencing (scRNA-seq) data of 103,599 cells. The effect of SIRPG expression was evaluated via SIRPG knockdown or overexpression in Jurkat T cells.Results The results showed that most cancers with high SIRPG expression had significantly higher abundance of T cells, B cells, NK cells, M1 macrophages and cytotoxic lymphocytes and increased expression level of immunomodulatory factors regulating immune cell recruitment, antigen presentation, T cell activation and cytotoxicity, but markedly lower abundance of neutrophils, M2 macrophages, and myeloid-derived suppressor cells. High SIRPG expression was associated with favorable response to PD-1 blockade in both NSCLC and melanoma. scRNA-seq data suggested SIRPG was mainly expressed in CD8+ exhausted T and CD4+ regulatory T cells, and positively associated with immune checkpoint expression including PDCD1 and CTLA4. In vitro test showed SIRPG expression in T cells could facilitate expression of PDCD1 and CTLA4.Conclusion High SIRPG expression is associated with an inflamed immune phenotype in cancers and favorable response to PD-1 blockade, suggesting it would be a promising predictive biomarker for PD-1 blockade and novel immunotherapeutic target.

KRAS G12D is the most frequently mutated oncogenic KRAS subtype in solid tumors and remains undruggable in clinical settings. Here, we developed a high affinity, selective, long-acting, and non-covalent KRAS G12D inhibitor, HRS-4642, with an affinity constant of 0.083 nM. HRS-4642 demonstrated robust efficacy against KRAS G12D-mutant cancers both in vitro and in vivo. Importantly, in a phase 1 clinical trial, HRS-4642 exhibited promising anti-tumor activity in the escalating dosing cohorts. Furthermore, the sensitization and resistance spectrum for HRS-4642 was deciphered through genome-wide CRISPR-Cas9 screening, which unveiled proteasome as a sensitization target. We further observed that the proteasome inhibitor, carfilzomib, improved the anti-tumor efficacy of HRS-4642. Additionally, HRS-4642, either as a single agent or in combination with carfilzomib, reshaped the tumor microenvironment toward an immune-permissive one. In summary, this study provides potential therapies for patients with KRAS G12D-mutant cancers, for whom effective treatments are currently lacking.

Abstract Background The effective therapeutic approach is still an unmet need for patients diagnosed with both lung cancer and interstitial lung disease (ILD). This is primarily due to the possible risk of ILD exacerbation caused by surgery or radiotherapy. The current study aimed to investigate the efficacy and safety of local ablative therapy (LAT) for this specific population. Methods Consecutive patients with non‐small cell lung cancer (NSCLC) and ILD who received LAT between January 2018 and August 2022 were enrolled, and propensity score matching (PSM) was utilized to match the non‐ILD group. The primary endpoint was recurrence‐free survival (RFS), and secondary endpoints included overall survival (OS), adverse events (AEs) and hospital length of stay (HLOS). Results The PSM algorithm yielded matched pairs in the ILD group (n = 25) and non‐ILD group (n = 72) at a ratio of 1:3. There were no statistically significant differences in RFS (median 16.4 vs. 18 months; HR = 1.452, p = 0.259) and OS (median: not reached vs. 47.9 months; HR = 1.096, p = 0.884) between the two groups. Meanwhile, no acute exacerbation of ILD was observed in the ILD group. However, the incidence of pneumothorax, especially pneumothorax requiring chest tube drainage, was significantly higher (36.0% vs. 11.2%, p = 0.005) among patients with NSCLC and co‐existing ILD, which resulted in longer HLOS (p = 0.045). Conclusion Although ILD was associated with a higher incidence of pneumothorax, the efficacy of LAT for NSCLC patients with ILD was comparable to those without ILD, suggesting that LAT might be a reliable and effective treatment option for this population, particularly in the early stage.

Study Design: A retrospective case-series. Objective: The study aims to use machine-learning (ML) to predict the discharge destination of spinal cord injury (SCI) patients in the intensive care unit (ICU). Summary of Background Data: Prognostication following SCI is vital, especially for critical patients who need intensive care. Methods: Clinical data of patients diagnosed with SCI were extracted from a publicly available ICU database. The firstly recorded data of the included patients were used to develop a total of 98 ML classifiers, seeking to predict discharge destination (e.g. death, further medical care, home). The micro-average area under the curve (AUC) was the main indicator to assess discrimination. The best average-AUC classifier and the best death-sensitivity classifier were integrated into an ensemble classifier. The discrimination of the ensemble classifier was compared with top death-sensitivity classifiers and top average-AUC classifiers. Additionally, prediction consistency and clinical utility were also assessed. Results: A total of 1485 SCI patients were included. The ensemble classifier had a micro-average AUC of 0.851, which was only slightly inferior to the best average-AUC classifier (P=0.10) The best average-AUC classifier death sensitivity was much lower than that of the ensemble classifier. The ensemble classifier had a death sensitivity of 0.452, which was inferior to top 8 death-sensitivity classifiers, whose micro-average AUC were inferior to the ensemble classifier (P<0.05). Additionally, the ensemble classifier demonstrated a comparable Brier score and superior Net benefit in the decision curve analysis, when compared to the performance of the origin classifiers. Conclusions: The ensemble classifier shows an overall superior performance in predicting discharge destination considering discrimination ability, prediction consistency and clinical utility. This classifier system may aid in the clinical management of critical SCI patients in the early phase following injury. Level of Evidence: 3

BACKGROUND:Programmed cell death-ligand 1 (PD-L1) expression was found to be a biomarker of inferior efficacy of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in EGFR-mutated non-small cell lung cancer (NSCLC). However, whether PD-L1 expression could also serve as a similar biomarker in anaplastic lymphoma kinase (ALK)-positive patients, especially for those treated with front-line alectinib, remains unclear. The aim of the study is to investigate the association of PD-L1 expression and efficacy of alectinib in this setting. METHODS:From January 2018 to March 2020, 225 patients with ALK-rearranged lung cancer were consecutively collected at Shanghai Pulmonary Hospital, Tongji University. Baseline PD-L1 expression was detected using immunohistochemistry (IHC) in 56 patients of advanced ALK-rearranged lung cancer who received front-line alectinib. RESULTS:Among the 56 eligible patients, 30 (53.6%) were PD-L1 expression negative, 19 (33.9%) patients had TPS 1%-49% and 7 (12.5%) had TPS ≥ 50%.We found no statistically significant associations between PD-L1 positivity and objective response rate (ORR, 90.0% vs. 80.8%, p = 0.274) or progression-free survival (PFS, not reached vs. not reached, HR: 0.98, 95 %CI: 0.37-2.61, p = 0.97) in patients treated with alectinib. Meanwhile, patients with PD-L1 high expression (TPS ≥ 50%) had a trend of longer PFS (not reached vs. not reached, p = 0.61). CONCLUSIONS:PD-L1 expression might not serve as a predict biomarker for the efficacy of front-line alectinib in ALK-positive NSCLC patients.

To the Editor: In recent years, impressive outcomes have been achieved in patients harboring human epidermal growth factor receptor 2 (HER2) mutations, which accounts for 1% to 4% of cases in non-small cell lung cancer (NSCLC).[1] Previously, we observed that pyrotinib showed potent efficacy in HER2 exon 20 YVMA insertion NSCLC patient-derived organoids and patient-derived xenograft models,[2] and confirmed its efficacy in a phase II clinical trial with an objective response rate (ORR) of 30.0% and a median progression-free survival (PFS) of 6.9 months in the later-line setting,[3] suggesting the potential of pyrotinib to serve as a novel standard of care (SoC) in the later-line setting. Currently, single agent chemotherapy or immune checkpoint inhibitors (ICIs) are the SoC for the later-line treatment of advanced NSCLC, and recent studies have found that the combination of chemo-immunotherapy treatment shows superior efficacy.[4] To compare the efficacy of pyrotinib with current SoC in the second- or third-line settings, we retrospectively collected data of 182 patients from Tongji University School of Medicine Cancer Institute and reported the outcomes here. Patients who were confirmed to have HER2 mutations at Tongji University School of Medicine Cancer Institute from January 2014 to August 2020 were reviewed. The main inclusion criteria were as follows: (1) histologically or cytologically confirmed stage IV NSCLC, (2) complete medical records, (3) disease progression after using first-line chemotherapy with or without bevacizumab, and (4) pyrotinib as a second- or third-line treatment, chemotherapy as a second-line treatment, or ICIs with or without chemotherapy as a second- or third-line treatment. Clinicopathological characteristics were extracted, including age, sex, Eastern Cooperative Oncology Group (ECOG) performance status (PS), smoking history, histological types, metastatic sites, treatment lines, therapeutic regimens, and date of death or last follow-up. The treatment strategy was decided by both the patient and the doctor with full informed consent. This study was approved by the Ethics Committee of Shanghai Pulmonary Hospital (No. K16–223-1). All procedures performed in this study were in accordance with the Declaration of Helsinki (as revised in 2013). Written informed consent forms were obtained from all individual participants. HER2 mutations were detected by the amplification refractory mutation system method as described in our previous publications.[2,3] The categorical variables were compared using the chi-squared test or Fisher's exact test when needed. PFS and overall survival (OS) were estimated by the Kaplan-Meier method, and compared by the log-rank test. Cox proportional hazard models were performed for univariate and multivariable survival analyses. Parameters with P value <0.10 in the univariate analysis were included in the multivariate model. P values were two-sided, and the significance level was set as P < 0.05. All statistical analyses were performed using SPSS statistical software (version 22.0; IBM Corporation, Armonk, NY, USA). Plots were generated by GraphPad Prism (version 7; GraphPad Software Inc., San Diego, CA, USA) and R software (version 4.0.2, https://www.r-project.org). In total, we identified 182 patients with HER2 mutations. Among them, 70 patients who met the inclusion criteria were included in the present study. The workflow of this research is shown in Supplementary Figure 1, https://links.lww.com/CM9/B291. A total of 34, 27, and 9 patients received pyrotinib, chemotherapy alone, and ICI-based immunotherapy, respectively. Chemotherapy and ICI-based immunotherapy were defined as SoC. As a result, patients were dichotomized into pyrotinib and SoC groups. The baseline characteristics are presented in Supplementary Table 1, https://links.lww.com/CM9/B291, and there were no significant differences in the demographic and clinical characteristics between the two groups. The best response to pyrotinib or the SoC is demonstrated in Supplementary Figure 2A, https://links.lww.com/CM9/B291 and Supplementary Table 2, https://links.lww.com/CM9/B291. Patients treated with pyrotinib had a significantly higher ORR (35.3% [12/34] vs. 8.3% [3/36], P = 0.006) and disease control rate (85.3% [29/34] vs. 52.8% [19/36], P = 0.003) than those treated with SoC. The best changes in tumor size from baseline are presented in Supplementary Figure 2B, https://links.lww.com/CM9/B291. The median follow-up time was 15.9 months. Patients treated with pyrotinib showed significantly longer PFS (median, 7.0 vs. 3.6 months, P = 0.036 [Supplementary Figure 2C, https://links.lww.com/CM9/B291]; hazard ratio (HR) = 0.558, 95% confidence interval (CI): 0.320–0.972, P = 0.039) and numerically longer OS (median 26.9 vs. 14.3 months, P = 0.064 [Supplementary Figure 2D, https://links.lww.com/CM9/B291]; HR = 0.512, 95% CI: 0.249–1.052, P = 0.069) than patients treated with SoC. For patients who received ICI-based treatment (n = 9), the ORR was 22.2%, the median PFS was 4.5 months, and the median OS was not reached. In subgroup analysis, pyrotinib showed a superior PFS in male, but not OS [Supplementary Figures 3A and 3B, https://links.lww.com/CM9/B291]. The univariate and multivariable Cox regression analyses are presented in Supplementary Tables 3, https://links.lww.com/CM9/B291 and 4, https://links.lww.com/CM9/B291, respectively. In the univariate analysis, patients with ECOG PS 0–1 and those treated with pyrotinib had significant benefits in regard to PFS, and the significance remained in the multivariable analysis. The simultaneous use of ICIs and targeted therapy might result in a higher rate of immune-related adverse events (AEs). We further evaluated the feasibility of sequential therapy with ICIs and pyrotinib. Among the 70 patients, 14 received pyrotinib and ICI-based therapy (sequential treatment with pyrotinib and ICIs, whichever came first). Individual treatment duration, the best response to pyrotinib and ICIs, HER2 mutation variants, and programmed death-ligand 1 status are presented in Supplementary Figure 2E, https://links.lww.com/CM9/B291. Among the five patients who received ICIs first followed by pyrotinib, the best response to pyrotinib was zero partial response (PR), four stable disease (SD), zero progressive disease, and one patient who discontinued treatment due to grade 3 diarrhea, vomiting, and abnormal hepatic function. Among the nine patients who were treated with pyrotinib first followed by ICIs, five PR and four SD with pyrotinib were observed without grade 3 or higher AEs during follow-up. HER2 mutation has distinct features from HER2 protein overexpression or amplification and is associated with poor prognosis in patients with advanced NSCLC. The targeted drugs, such as lapatinib, neratinib, trastuzumab, and pertuzumab, which are approved for HER2-overexpression or HER2-amplified breast or gastric cancer have poor efficacy in NSCLC. Meanwhile, the pan-ErbB receptor tyrosine kinase inhibitors afatinib and dacomitinib also show very limited efficacy in NSCLC. As the most frequent mutation variants, exon 20 insertions account for 90% of all HER2 mutations. In vitro models revealed that the insertions in HER2 restricted the size of the drug-binding pocket, thus limiting the binding of large, rigid inhibitors.[5] Therefore, an alternative strategy is urgently needed. Fortunately, there are some inspiring results from several agents. Ado-trastuzumab emtansine was confirmed to be associated with an ORR of 44% and a median PFS of 5 months in the later-line setting.[6] Similarly, DS-8201a achieved an ORR of 55%, a median PFS of 8.2 months, and a median OS of 17.8 months in 91 patients harboring HER2 mutations.[7] In our previous phase II study, pyrotinib as a second- or later-line treatment showed an ORR of 30.0%, with a median PFS and OS of 6.9 and 14.4 months, respectively.[3] In addition, poziotinib and mobocertinib also showed promising results in their phase II trials.[8] Nevertheless, all the studies were single-arm clinical designs and lacked comparison with SoC. In the current study, we firstly compared the efficacy of pyrotinib with SoC, and found that pyrotinib was significantly associated with an improved ORR, prolonged PFS, and numerically prolonged OS. The efficacy was in line with the results from prospective clinical trials, suggesting a preferred recommendation of pyrotinib compared with SoC in this setting. Since the simultaneous or sequential use of ICIs and osimertinib has resulted in significantly increased grade 3/4 side effects, we also assessed the efficacy and side effects of the sequential use of ICIs and pyrotinib. The result was consistent with the findings of the sequential use of ICIs and osimertinib and suggested that pyrotinib should precede the use of ICIs. Several limitations must be acknowledged for this study. First, this study is retrospective in nature and includes patients from a single center. Second, a relatively small number of patients, especially only five patients who received the sequential use of ICIs and pyrotinib, were enrolled in the final analysis. Third, antibody-drug conjugates such as T-DM1 and DS-8201a are unavailable in China, thus the findings of this study cannot be generalized to the global population. In conclusion, pyrotinib showed superior ORR, PFS, and OS compared with SoC as a second- or third-line treatment in patients with HER2-mutant advanced NSCLC. Further phase III randomized trial is ongoing to validate the findings of this study. Funding This work was supported by grants from the National Natural Science Foundation of China (Nos. 81972167, 82172869 and 82002419), Shanghai Shenkang Hospital Development Center (No. SHDC12019133), Clinical Research Foundation of Shanghai Pulmonary Hospital (No. FKLY20008) and Shanghai Innovative Collaboration Project (No. 2020CXJQ02). Conflicts of interest None.