Machine Learning for Early Discrimination Between Lung Cancer and Benign Nodules Using Routine Clinical and Laboratory Data.

BACKGROUND: Lung cancer poses a global health threat necessitating early detection and precise staging for improved patient outcomes. This study focuses on developing and validating a machine learning-based risk model for early lung cancer screening and staging, using routine clinical data. METHODS: Two medical center, observational, retrospective studies were conducted, involving 2312 lung cancer patients and 653 patients with benign nodules. Machine learning techniques, including differential analysis and feature selection, were employed to identify key factors for modeling. The study focused on variables such as nodule density, carcinoembryonic antigen (CEA), age, and lifestyle habits. The Logistic Regression model was utilized for early diagnoses, and the XGBoost model was utilized for staging based on selected features. RESULTS: For early diagnoses, the Logistic Regression model achieved an area under the curve (AUC) of 0.716 (95% confidence interval [CI] 0.607-0.826), with 0.703 sensitivity and 0.654 specificity. The XGBoost model excelled in distinguishing late-stage from early-stage lung cancer, exhibiting an AUC of 0.913 (95% CI 0.862-0.963), with 0.909 sensitivity and 0.814 specificity. These findings highlight the model's potential for enhancing diagnostic accuracy and staging in lung cancer. CONCLUSION: This study introduces a novel machine learning-based risk model for early lung cancer screening and staging, leveraging routine clinical information and laboratory data. The model shows promise in enhancing accuracy, mitigating overdiagnosis, and improving patient outcomes.

Immunological profiling for short-term predictive analysis in PD-1/PD-L1 therapy for lung cancer.

BACKGROUND: Immune checkpoint inhibitors, such as anti-programmed cell death-1 (PD-1) and PD-1 ligand-1 (PD-L1) antibodies, have achieved breakthrough results in improving long-term survival rates in lung cancer. Although high levels of PD-L1 expression and tumor mutational burden have emerged as pivotal biomarkers, not all patients derive lasting benefits, and resistance to immune checkpoint blockade remains a prevalent issue. Comprehending the immunological intricacies of lung cancer is crucial for uncovering the mechanisms that govern responses and resistance to immunomodulatory treatments. This study aimed to explore the potential of peripheral immune markers in predicting treatment efficiency among lung cancer patients undergoing PD-1/PD-L1 checkpoint inhibitors. METHODS: This study enrolled 71 lung cancer patients undergoing PD-1/PD-L1 inhibitor therapy and 20 healthy controls. Immune cell subsets (CD4 + T cells, CD8 + T cells, B cells, NK cells, and NKT cells), phenotypic analysis of T cells and B cells, and PMA/Ionomycin-stimulated lymphocyte function assay were conducted. RESULTS: Lung cancer patients exhibited significant alterations in immune cell subsets, notably an increased percentage of Treg cells. Post-treatment, there were substantial increases in absolute numbers of CD3 + T cells, CD8 + T cells, and NKT cells, along with heightened HLA-DR expression on CD3 + T and CD8 + T cells. Comparison between complete remission and non-complete remission (NCR) groups showed higher Treg cell percentages and HLA-DR + CD4 + T cells in the NCR group. CONCLUSION: The study findings suggest potential predictive roles for immune cell subsets and phenotypes, particularly Treg cells, HLA-DR + CD4 + T cells, and naïve CD4 + T cells, in evaluating short-term PD-1/PD-L1 therapy efficacy for lung cancer patients. These insights offer valuable prospects for personalized treatment strategies and underscore the importance of immune profiling in lung cancer immunotherapy.

Analysis of risk characteristics for early progression and late progression in locally advanced rectal cancer patients: a large population-based and validated study.

BACKGROUND AND OBJECTIVE: The current study aimed to explore the factors influencing early progression (EP) and late progression (LP) in locally advanced rectal cancer (LARC) patients. METHODS: The patients were classified into EP and LP groups using one year as a cutoff. The random survival forest model was utilized to calculate the probability of time-to-progression. Besides, inverse probability of treatment weighting (IPTW) analysis and the Surveillance, Epidemiology, and End Results (SEER) were conducted to validate our results. RESULTS: Our study revealed that PNI, CEA level, and pathological stage were independent prognostic factors for PFS both in EP group and LP group. For EP group patients, Group 1 had the highest probability of progression at the 9th month of follow-up, while Group 2 exhibited the highest probability at the 6th month. Group 3, on the other hand, showed two peaks of progression at the 4th and 8th months of follow-up. As for LP group patients, Groups 4, 5, and 6 all exhibited peaks of progression between the 18th and 24th months of follow-up. Furthermore, our results suggested that PNI was also an independent prognostic factor affecting OS in both EP group and LP group. Finally, the analysis of IPTW and SEER database further confirmed our findings. CONCLUSIONS: Our results indicated a significant correlation between immune and nutritional status with PFS and OS in both EP and LP groups. These insights can aid healthcare professionals in effectively identifying and evaluating patients' nutritional status, enabling them to develop tailored nutrition plans and interventions.

Prediction of liver metastasis and recommended optimal follow-up nursing in rectal cancer.

We aimed to analyze and investigate the clinical factors that influence the occurrence of liver metastasis in locally advanced rectal cancer patients, with an attempt to assist patients in devising the optimal imaging-based follow-up nursing. Between June 2011 and May 2021, patients with rectal cancer at our hospital were retrospectively analyzed. A random survival forest model was developed to predict the probability of liver metastasis and provide a practical risk-based approach to surveillance. The results indicated that age, perineural invasion, and tumor deposit were significant factors associated with the liver metastasis and survival. The liver metastasis risk of the low-risk group was higher at 6-21 months, with a peak occurrence time in the 15th month. The liver metastasis risk of the high-risk group was higher at 0-24 months, with a peak occurrence time in the 8th month. In general, our clinical model could predict liver metastasis in rectal cancer patients. It provides a visualization tool that can aid physicians and nurses in making clinical decisions, by detecting the probability of liver metastasis.

Kinetics and prognostic significance of laboratory markers in patients with severe fever with thrombocytopenia syndrome: insight from a comprehensive analysis.

Severe Fever with Thrombocytopenia Syndrome (SFTS) is an emerging infectious disease with significant mortality. Identifying prognostic factors that influence patient outcomes is crucial for effective clinical management. In this study, we assessed the dynamic changes of laboratory markers and their association with outcomes in 93 SFTS patients. We found that age and hypertension were significantly associated with poor outcomes in SFTS patients. The deceased group exhibited lower platelet counts, elevated liver and kidney function markers, coagulation profiles, inflammatory markers, and cytokines compared to the survival group. Kinetic analysis showed that these markers gradually normalized in the survival group, while they remained persistently abnormal in the deceased group. Furthermore, hypertension, elevated AST, PCT, and IL-10 were identified as independent risk factors for predicting poor prognosis of SFTS patients. These findings provide valuable insights into the prognostic significance of laboratory markers and highlight the importance of early identification of high-risk SFTS patients.

The clinical and immunological characteristics in fatal severe fever with thrombocytopenia syndrome virus (SFTSV) infection.

OBJECTIVE: This study aimed to make a comprehensive evaluation of peripheral immune profiles for further understanding the immunopathogenesis of severe fever with thrombocytopenia syndrome (SFTS). METHODS: Forty-seven patients with SFTS virus infection were included, of which 24 were deceased. The percentages, absolute numbers, phenotype of lymphocyte subsets were detected by flow cytometry. RESULTS: In patients with SFTS, the numbers of CD3+T, CD4+T, CD8+T and NKT cells were decreased compared with healthy controls (HCs), accompanied with highly active and exhausted phenotypes for T cells, and overproliferating plasmablasts. High inflammatory status, dysregulation of coagulation and host immune response were more obvious in deceased patients than that of survivors. Higher levels of PCT, IL-6, IL-10, TNF-α, APTT, TT and the occurrence of hemophagocytic lymphohistiocytosis were poor prognostic indicators of SFTS. CONCLUSIONS: The evaluation of immunological markers in combination with laboratory tests has critical value for selecting prognostic markers and potential treatment target.

A large-scale study integrating nutritional indicators and clinicopathological parameters to evaluate prognosis, follow-up, and postoperative chemotherapy decisions in rectal cancer patients.

OBJECTIVE: The aim of this study was to evaluate the role of nutritional indicators and clinicopathological parameters in predicting the progression and prognosis for pathological stage II-III rectal cancer (RC) patients without neoadjuvant radiotherapy. In addition, we sought to explore the high-risk population who may require postoperative chemotherapy. METHODS: A total of 894 consecutive RC patients were enrolled in this study. Univariate and multivariate Cox analysis were performed to identify the independent risk factors for PFS and OS. The nomogram and calibration curves were conducted according to multivariable analysis result. Kaplan-Meier survival curves and log-rank tests were performed for different groups. Finally, random survival forest (RSF) model was developed to predict the probability of progression. RESULTS: Our results revealed that CEA level, pathological stage, tumor deposit, and PNI were independently associated with PFS in RC patients. Similarly, the results indicated that CEA level, pathological stage, tumor deposit, PNI, and NRI were independently associated with OS. RSF model revealed that group 1 had the highest risk of progression at the 12th month of follow-up, group 2 had the highest risk of progression at the 15th month of follow-up, while group 3 had the highest risk of progression at the 9th month of follow-up. Besides, subgroup analysis suggested that the high-risk group needs postoperative adjuvant chemotherapy, while patients in the low- and moderate-risk groups may not need postoperative adjuvant chemotherapy. Finally, we validated our results with the SEER database. CONCLUSIONS: In conclusion, we demonstrated that preoperative nutritional indicator and clinicopathological parameters could act as auxiliary prognostication tools for RC patients without neoadjuvant radiotherapy. We also established follow-up strategies for different groups of patients. Collectively, incorporating nutritional assessment into risk stratification for RC resection is crucial and should be an integral part of preoperative planning.

Immunologic memory to SARS-CoV-2 in convalescent COVID-19 patients at 1 year postinfection.

BACKGROUND: Understanding the complexities of immune memory to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is key to gain insights into the durability of protective immunity against reinfection. OBJECTIVE: We sought to evaluate the immune memory to SARS-CoV-2 in convalescent patients with longer follow-up time. METHODS: SARS-CoV-2-specific humoral and cellular responses were assessed in convalescent patients with coronavirus disease 2019 (COVID-19) at 1 year postinfection. RESULTS: A total of 78 convalescent patients with COVID-19 (26 moderate, 43 severe, and 9 critical) were recruited after 1 year of recovery. The positive rates of both anti-receptor-binding domain and antinucleocapsid antibodies were 100%, whereas we did not observe a statistical difference in antibody levels among different severity groups. Accordingly, the prevalence of neutralizing antibodies (nAbs) reached 93.59% in convalescent patients. Although nAb titers displayed an increasing trend in convalescent patients with increased severity, the difference failed to achieve statistical significance. Notably, there was a significant correlation between nAb titers and anti-receptor-binding domain levels. Interestingly, SARS-CoV-2-specific T cells could be robustly maintained in convalescent patients, and their number was positively correlated with both nAb titers and anti-receptor-binding domain levels. Amplified SARS-CoV-2-specific CD4+ T cells mainly produced a single cytokine, accompanying with increased expression of exhaustion markers including PD-1, Tim-3, TIGIT, CTLA-4, and CD39, while the proportion of multifunctional cells was low. CONCLUSIONS: Robust SARS-CoV-2-specific humoral and cellular responses are maintained in convalescent patients with COVID-19 at 1 year postinfection. However, the dysfunction of SARS-CoV-2-specific CD4+ T cells supports the notion that vaccination is needed in convalescent patients for preventing reinfection.

The dynamics of immune response in COVID-19 patients with different illness severity.

This study aimed to analyze the dynamic changes of lymphocyte subsets and specific antibodies in coronavirus disease 2019 (COVID-19) patients with different illness severity. The amounts of lymphocyte subsets and the levels of immunoglobulin M (IgM) and IgG antibody were retrospectively analyzed in 707 COVID-19 cases. The amounts of lymphocyte subsets were significantly decreased with the increased severity of illness and the levels of IgM and IgG were lower in critical cases than severe and moderate cases. In deceased patients, the lymphocytes subsets were significantly lower than recovered patients. However, the relationship between the levels of IgM and IgG and the amounts of lymphocyte subsets were not significantly correlated. During different stages of COVID-19, the total T cell, CD4+ T cell, and CD8+ T cell counts were gradually recovered to the normal levels in severe and critical groups but the changing trend was relatively stable in the moderate group. The production of IgM and IgG antibodies were delayed in critical groups but also could reach the peak levels at one month after illness onset and decreased to background levels. To detect the kinetics of lymphocytes and antibodies has important clinical value in predicting the illness severity and understanding the pathogenesis of COVID-19.

Dynamic changes in peripheral blood lymphocyte subset counts and functions in patients with diffuse large B cell lymphoma during chemotherapy.

BACKGROUND: This study aimed to analyze the lymphocyte subsets, their activities and their dynamic changes during immunochemotherapy in patients newly diagnosed with diffuse large B cell lymphoma (DLBCL). METHODS: Patients with DLBCL (n = 33) were included in the present study. Their peripheral lymphocyte subsets, phenotypes and functions were detected using flow cytometry. The dynamic results of lymphocyte activities were available for 18 patients. RESULTS: Compared with healthy controls (HCs), the counts of CD3+, CD4+, and CD8+ T cells as well as those NK cells decreased in patients newly diagnosed with DLBCL, mainly attributed to patients with high risk of prognosis assessed by International Prognostic Index (IPI) score. Lymphocyte counts didn't present significant difference between high risk (IPI scores 3-5) and low risk patients (IPI scores 0-2), but CD4+ T cells and CD8+ T cells expressed higher levels of CD28 and HLA-DR, respectively, in patients with IPI score ranging from 3 to 5. Patients at high risk harbored higher percentage of regulatory T cells (Tregs), and their CD4+ and CD8+ T cells produced lower levels of IFN-γ, reflecting an impaired cellular immune response. The dynamic changes of lymphocyte numbers and functions during treatment were further investigated. Total counts of CD3+, CD4+, CD8+ T and NK cells progressively decreased because of the cytotoxicity of chemotherapy and then gradually recovered after six cycles treatment (rituximab combined with cyclophosphamide, doxorubicin, vincristine and prednisone, R-CHOP). The functions of CD4+ and CD8+ T cells recovered by the end of two cycles R-CHOP treatment, although NK cell function was not significantly affected throughout treatment. These results suggest that the counts and functions of lymphocytes are significantly decreased in patients with DLBCL, particularly those of CD4+ and CD8+ T cells. CONCLUSIONS: The absolute counts and functions of CD4+, CD8+ T cells, which were significantly lower in patients with DLBCL, gradually recovered after effective treatment. Therefore, combined detection of T cell counts and functions are critically important for administering effective personalized immunotherapy as well as for identifying new prognostic markers or DLBCL.

Clinical characteristics and risk factors associated with COVID-19 disease severity in patients with cancer in Wuhan, China: a multicentre, retrospective, cohort study.

BACKGROUND: COVID-19 has spread globally. Epidemiological susceptibility to COVID-19 has been reported in patients with cancer. We aimed to systematically characterise clinical features and determine risk factors of COVID-19 disease severity for patients with cancer and COVID-19. METHODS: In this multicentre, retrospective, cohort study, we included all adult patients (aged ≥18 years) with any type of malignant solid tumours and haematological malignancy who were admitted to nine hospitals in Wuhan, China, with laboratory-confirmed COVID-19 between Jan 13 and March 18, 2020. Enrolled patients were statistically matched (2:1) with patients admitted with COVID-19 who did not have cancer with propensity score on the basis of age, sex, and comorbidities. Demographic characteristics, laboratory examinations, illness severity, and clinical interventions were compared between patients with COVID-19 with or without cancer as well as between patients with cancer with non-severe or severe COVID-19. COVID-19 disease severity was defined on admission on the basis of the WHO guidelines. Univariable and multivariable logistic regression, adjusted for age, sex, comorbidities, cancer type, tumour stage, and antitumour treatments, were used to explore risk factors associated with COVID-19 disease severity. This study was registered in the Chinese Clinical Trial Register, ChiCTR2000030807. FINDINGS: Between Jan 13 and March 18, 2020, 13 077 patients with COVID-19 were admitted to the nine hospitals in Wuhan and 232 patients with cancer and 519 statistically matched patients without cancer were enrolled. Median follow-up was 29 days (IQR 22-38) in patients with cancer and 27 days (20-35) in patients without cancer. Patients with cancer were more likely to have severe COVID-19 than patients without cancer (148 [64%] of 232 vs 166 [32%] of 519; odds ratio [OR] 3·61 [95% CI 2·59-5·04]; p<0·0001). Risk factors previously reported in patients without cancer, such as older age; elevated interleukin 6, procalcitonin, and D-dimer; and reduced lymphocytes were validated in patients with cancer. We also identified advanced tumour stage (OR 2·60, 95% CI 1·05-6·43; p=0·039), elevated tumour necrosis factor α (1·22, 1·01-1·47; p=0·037), elevated N-terminal pro-B-type natriuretic peptide (1·65, 1·03-2·78; p=0·032), reduced CD4+ T cells (0·84, 0·71-0·98; p=0·031), and reduced albumin-globulin ratio (0·12, 0·02-0·77; p=0·024) as risk factors of COVID-19 severity in patients with cancer. INTERPRETATION: Patients with cancer and COVID-19 were more likely to deteriorate into severe illness than those without cancer. The risk factors identified here could be helpful for early clinical surveillance of disease progression in patients with cancer who present with COVID-19. FUNDING: China National Natural Science Foundation.

Analysis of 2 Reverse Syphilis Testing Algorithms in Diagnosis of Syphilis: A Large-Cohort Prospective Study.

Background: Two serologic syphilis screening algorithms recommended by the US Centers for Disease Control and Prevention (US CDC) and the European Centre for Disease Prevention and Control (ECDC), respectively, are commonly used for syphilis screening; however, which one is optimal remains to be determined. Methods: We conducted a prospective study of 119891 subjects to analyze the consistency of the US CDC- and ECDC-recommended algorithms. The US CDC-recommended algorithm begins with a treponemal immunoassay, followed by a rapid plasma reagin (RPR) test. RPR-nonreactive samples are confirmed by the Treponema pallidum particle agglutination assay (TPPA). The ECDC-recommended algorithm begins with a treponemal immunoassay, followed by a confirmatory treponemal test. If the confirmatory test is reactive, a quantitative nontreponemal assay is used to assess the disease activity and treatment response. In the present study, a total of 119891 serum samples from a large hospital (sixth largest in China) were included, and each sample was screened with a chemiluminescent immunoassay (CIA). CIA-reactive samples were then simultaneously tested with RPR and TPPA. The consistency of these 2 algorithms was determined by calculating the percentage of agreement and κ coefficient. Results: The overall percentage of agreement and κ value between these 2 algorithms were 99.996% and 0.999, respectively. The positivity rate for syphilis as determined by the US CDC- and ECDC-recommended algorithms was 1.43% and 1.42%, respectively. Conclusions: Our results suggest that the US CDC-recommended algorithm and the ECDC-recommended algorithm have comparable performances for syphilis screening in low-prevalence populations.

TIGIT signalling pathway negatively regulates CD4+ T-cell responses in systemic lupus erythematosus.

B-lymphocyte hyperactivity in systemic lupus erythematosus (SLE) is T-cell-dependent, and CD4+ T-cell activation is essential to SLE pathogenesis. However, the mechanism of the deregulation of CD4+ T cells in SLE is largely unknown. T-cell immunoglobulin and ITIM domain (TIGIT) is a new inhibitory receptor preferentially expressed on activated CD4+ T cells. Here, we address the role of TIGIT in the pathogenesis of SLE. Our results showed that TIGIT expression on CD4+ T cells was significantly elevated in patients with SLE and highly correlated with the activity of the disease. TIGIT+ CD4+ T cells from both healthy individuals and patients with SLE had a more activated phenotype than TIGIT- CD4+ T cells. In contrast, the activation, proliferation and cytokine production potential of TIGIT+ CD4+ T cells were significantly lower than those of TIGIT- CD4+ T cells. Furthermore, activation of the TIGIT pathway by using CD155 could substantially down-regulate the activities of CD4+ T cells from SLE patients in vitro, and in vivo administration of CD155 resulted in a delayed development of SLE in MRL/lpr mice. TIGIT is a powerful negative regulator of CD4+ T cells in SLE, which suggests that the TIGIT signalling pathway may be used as a potential therapeutic target for treating this disease.

TIGIT expression levels on human NK cells correlate with functional heterogeneity among healthy individuals.

Human NK cells display extensive phenotypic and functional heterogeneity among healthy individuals, but the mechanism responsible for this variation is still largely unknown. Here, we show that a novel immune receptor, T-cell immunoglobulin and ITIM domain (TIGIT), is expressed preferentially on human NK cells but shows wide variation in its expression levels among healthy individuals. We found that the TIGIT expression level is related to the phenotypic and functional heterogeneity of NK cells, and that NK cells from healthy individuals can be divided into three categories according to TIGIT expression. NK cells with low levels of TIGIT expression show higher cytokine secretion capability, degranulation activity, and cytotoxic potential than NK cells with high levels of TIGIT expression. Blockade of the TIGIT pathway significantly increased NK-cell function, particularly in NK cells with high levels of TIGIT expression. We further observed that the TIGIT expression level was inversely correlated with the IFN-γ secretion capability of NK cells in patients with cancers and autoimmune diseases. Importantly, we propose a novel mechanism that links TIGIT expression with NK-cell functional heterogeneity, and this mechanism might partially explain why individuals have different susceptibilities to infection, autoimmune disease, and cancer.

Tim-3 pathway affects NK cell impairment in patients with active tuberculosis.

Active tuberculosis (TB) patients show impaired NK cell function, and the underlying mechanism remains largely unknown. In this study, we confirmed the decrease in activation, cytokine secretion, and degranulation potential of NK cells in active TB patients. We further investigated whether coinhibitory receptor Tim-3 was involved with impairment of NK cells. Our results revealed that the expression of Tim-3 on NK cells was increased in active TB patients. Tim-3 expression was inversely correlated with IL-12-stimualted IFN-γ production. Moreover, blocking the Tim-3 pathway restored IFN-γ secretion and degranulation of NK cells. Blocking this pathway also increased NK cell cytotoxicity against K562 target cells, and improved the ability of NK cells to control Mtb growth in monocyte-derived macrophages. The Tim-3 expression on NK cells was also observed to be significantly decreased in TB patients post-treatment. In this study, we have identified that Tim-3 is involved with NK cell impairment in TB patients.

Comprehensive evaluation of immune dysregulation in secondary hemophagocytic lymphohistiocytosis.

Host immune dysfunction plays a crucial role in the onset, progression, and outcome of hemophagocytic lymphohistiocytosis (HLH). This study aimed to comprehensively evaluate the peripheral immune profiles in patients with newly diagnosed secondary hemophagocytic lymphohistiocytosis (sHLH), and explore their predictive value for patient prognosis. A total of 77 patients with sHLH were enrolled in this study, with 31 of them experiencing mortality. Flow cytometry was used to assess the percentages, absolute numbers, and phenotypes of lymphocyte subsets. Simultaneously, cytokine levels and routine laboratory indicators were also collected. In sHLH patients, lymphocyte subset absolute numbers were significantly impaired, accompanied by T cell hyperactivation, B cell hyperactivation, and increased plasmablast proliferation. Prognostic analysis revealed that lower CD8+ T cell percentages, elevated APTT, IL-6, IL-10 levels, and increased CD4+CD28null T cell proportions were associated with poor patient outcomes. The study demonstrates dysregulation in the counts and phenotypes of lymphocyte subsets in sHLH patients. Several key factors, including IL-6, IL-10, APTT, and various T cell percentages, have potential as prognostic markers and therapeutic targets in sHLH.

Evaluating the diagnostic and therapeutic significance of KL-6 in patients with interstitial lung diseases.

Background: This study aimed to assess the diagnostic value of Krebs von den Lungen-6 (KL-6), Surfactant protein-A (SP-A), SP-D and molecular matrixmetalloproteinase-7 (MMP-7) in discriminating patients with interstitial lung diseases (ILDs) from disease control subjects. Methods: Serum levels of KL-6, SP-A, SP-D and MMP-7 were measured in both the ILD and non-ILD (NILD) groups. Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the diagnostic potential of these markers and laboratory indices. High-resolution computed tomography (HRCT) fibrosis scores were determined, and their correlation with the serum markers was analyzed. Results: Serum levels of KL-6 and MMP-7 were significantly elevated in the ILD group compared to the control group, while no significant differences were observed for SP-A and SP-D. ROC analysis of KL-6 demonstrated superior diagnostic accuracy, with a sensitivity of 76.36%, specificity of 91.07%, and an area under curve (AUC) of 0.902 (95%CI 0.866-0.945). These findings were consistent across an additional cohort. Correlation analysis revealed a link between KL-6 levels at initial diagnosis and HRCT fibrosis scores, indicating disease severity. Moreover, a negative correlation was found between KL-6 and pulmonary function indices, reflecting disease progression. Patients with increased 12-month HRCT fibrosis score showed higher lactate dehydrogenase (LDH) levels, with LDH exhibiting an AUC of 0.767 (95% CI: 0.520-0.927) as a predictor of progression. Conclusions: Serum KL-6 detection proves to be a valuable tool for accurately distinguishing ILDs from control subjects. While KL-6 shows a correlation with HRCT fibrosis scores and a negative association with pulmonary function indices, its predictive value for ILDs prognosis is limited. Trial registration: This study received retrospective approval from the Ethical Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China (institutional review board ID: TJ-IRB20210331, date: 2021.03.30).

Novel multiclass classification machine learning approach for the early-stage classification of systemic autoimmune rheumatic diseases.

OBJECTIVE: Systemic autoimmune rheumatic diseases (SARDs) encompass a diverse group of complex conditions with overlapping clinical features, making accurate diagnosis challenging. This study aims to develop a multiclass machine learning (ML) model for early-stage SARDs classification using accessible laboratory indicators. METHODS: A total of 925 SARDs patients were included, categorised into SLE, Sjögren's syndrome (SS) and inflammatory myositis (IM). Clinical characteristics and laboratory markers were collected and nine key indicators, including anti-dsDNA, anti-SS-A60, anti-Sm/nRNP, antichromatin, anti-dsDNA (indirect immunofluorescence assay), haemoglobin (Hb), platelet, neutrophil percentage and cytoplasmic patterns (AC-19, AC-20), were selected for model building. Various ML algorithms were used to construct a tripartite classification ML model. RESULTS: Patients were divided into two cohorts, cohort 1 was used to construct a tripartite classification model. Among models assessed, the random forest (RF) model demonstrated superior performance in distinguishing SLE, IM and SS (with area under curve=0.953, 0.903 and 0.836; accuracy= 0.892, 0.869 and 0.857; sensitivity= 0.890, 0.868 and 0.795; specificity= 0.910, 0.836 and 0.748; positive predictive value=0.922, 0.727 and 0.663; and negative predictive value= 0.854, 0.915 and 0.879). The RF model excelled in classifying SLE (precision=0.930, recall=0.985, F1 score=0.957). For IM and SS, RF model outcomes were (precision=0.793, 0.950; recall=0.920, 0.679; F1 score=0.852, 0.792). Cohort 2 served as an external validation set, achieving an overall accuracy of 87.3%. Individual classification performances for SLE, SS and IM were excellent, with precision, recall and F1 scores specified. SHAP analysis highlighted significant contributions from antibody profiles. CONCLUSION: This pioneering multiclass ML model, using basic laboratory indicators, enhances clinical feasibility and demonstrates promising potential for SARDs classification. The collaboration of clinical expertise and ML offers a nuanced approach to SARDs classification, with potential for enhanced patient care.