Development of a gene panel for immune status assessment in sepsis

Oral mucosal lesions hold one of the lead places in the structure of dental diseases. Oral lichen planus (OLP) considered as a multifactorial disease is of top priority among dermatoses of oral mucosa and red lip border. Diverse putative concepts behind developing lichen planus pathogenesis are discussed including immunoallergic, viral, genetic and membrane-destructive theories. However, an immune theory is thought to play a crucial role in developing lichen planus. Despite documented induction of immune mechanisms, complex interaction between pathological process and normal defense response as well as stirred interest to it, multiple aspects of immunological conflict in lichen planus remain unclear. Few data describing altered immune parameters depending on clinical picture of lichen planus are currently available that suggested to perform the study aimed at examining immune status in patients with various forms of oral lichen planus. There were enrolled 286 oral lichen planus patients (248 females and 38 males), aged 2784 years. Based on clinical picture, all OLP patients were divided into 6 groups: a typical type, exudative-congestive, erosive-ulcerous, hyperkeratotic, atypical, bullous type. An immune status in peripheral blood samples was evaluated by analyzing innate defense mechanisms as well as humoral and cellular immunity. Multiple altered immune parameters characterized by impaired phagocytic capture and metabolic activity, disimmunoglobulinemia, altered ratio of major immunocompetent cell types and subsets were documented during the study. Moreover, OLP patients with typical, hyperkeratotic, exudative hyperemic, atypical and erosive ulcerous forms were found to have increased amount of CD4+ helper T cells associated with a self-sustained immune response due to suppressed elimination of pathogenic agents consequently resulting in developing autoimmune process. While analyzing immune status in OLP patients, it allowed to find a relationship between dysphagocytosis signs, impaired humoral and cellular immunity as well as various clinical forms of the disease. Thus, it suggests imbalanced mechanisms responsible for pathogen elimination might play a role in OLP pathogenesis, including infectious agents being involved in its development.

In the era of precision therapy, early classification of breast cancer (BRCA) molecular subtypes has clinical significance for disease management and prognosis. We explored the accuracy of machine learning (ML) models for early classification of BRCA molecular subtypes through a systematic review of the literature currently available. We retrieved relevant studies published in PubMed, EMBASE, Cochrane, and Web of Science until 15 April 2022. A prediction model risk of bias assessment tool (PROBAST) was applied for the assessment of risk of bias of a genomics-based ML model, and the Radiomics Quality Score (RQS) was simultaneously used to evaluate the quality of this radiomics-based ML model. A random effects model was adopted to analyze the predictive accuracy of genomics-based ML and radiomics-based ML for Luminal A, Luminal B, Basal-like or triple-negative breast cancer (TNBC), and human epidermal growth factor receptor 2 (HER2). The PROSPERO of our study was prospectively registered (CRD42022333611). Of the 38 studies were selected for analysis, 14 ML models were based on gene-transcriptomic, with only 4 external validations; and 43 ML models were based on radiomics, with only 14 external validations. Meta-analysis results showed that c-statistic values of the ML based on radiomics for the identification of BRCA molecular subtypes Luminal A, Luminal B, Basal-like or TNBC, and HER2 were 0.76 [95% confidence interval (CI): 0.60-0.96], 0.78 (95% CI: 0.69-0.87), 0.89 (95% CI: 0.83-0.91), and 0.83 (95% CI: 0.81-0.86), respectively. The c-statistic values of ML based on the gene-transcriptomic analysis cohort for the identification of the previously described BRCA molecular subtypes were 0.96 (95% CI: 0.93-0.99), 0.96 (95% CI: 0.93-0.99), 0.98 (95% CI: 0.95-1.00), and 0.97 (95% CI: 0.96-0.98) respectively. Additionally, the sensitivity of the ML model based on radiomics for each molecular subtype ranged from 0.79 to 0.85, while the sensitivity of the ML model based on gene-transcriptomic was between 0.92 and 0.99. Both radiomics and gene transcriptomics produced ideal effects on BRCA molecular subtype prediction. Compared with radiomics, gene transcriptomics yielded better prediction results, but radiomics was simpler and more convenient from a clinical point of view.

Estimating whether to treat the rupture risk of small intracranial aneurysms (IAs) with size ≤ 7mm in diameter is difficult but crucial. We aimed to construct and externally validate a convenient machine learning (ML) model for assessing the rupture risk of small IAs. One thousand four patients with small IAs recruited from two hospitals were included in our retrospective research. The patients at hospital 1 were stratified into training (70%) and internal validation set (30%) randomly, and the patients at hospital 2 were used for external validation. We selected predictive features using the least absolute shrinkage and selection operator (LASSO) method and constructed five ML models applying diverse algorithms including random forest classifier (RFC), categorical boosting (CatBoost), support vector machine (SVM) with linear kernel, light gradient boosting machine (LightGBM), and extreme gradient boosting (XGBoost). The Shapley Additive Explanations (SHAP) analysis provided interpretation for the best ML model. The training, internal, and external validation cohorts included 658, 282, and 64 IAs, respectively. The best performance was presented by SVM as AUC of 0.817 in the internal [95% confidence interval (CI), 0.769-0.866] and 0.893 in the external (95% CI, 0.808-0.979) validation cohorts, which overperformed compared with the PHASES score significantly (all P < 0.001). SHAP analysis showed maximum size, location, and irregular shape were the top three important features to predict rupture. Our SVM model based on readily accessible features presented satisfying ability of discrimination in predicting the rupture IAs with small size. Morphological parameters made important contributions to prediction result.

Skeletal muscle loss during treatment is associated with poor survival outcomes in patients with ovarian cancer. Although changes in muscle mass can be assessed on computed tomography (CT) scans, this labour-intensive process can impair its utility in clinical practice. This study aimed to develop a machine learning (ML) model to predict muscle loss based on clinical data and to interpret the ML model by applying SHapley Additive exPlanations (SHAP) method. This study included the data of 617 patients with ovarian cancer who underwent primary debulking surgery and platinum-based chemotherapy at a tertiary centre between 2010 and 2019. The cohort data were split into training and test sets based on the treatment time. External validation was performed using 140 patients from a different tertiary centre. The skeletal muscle index (SMI) was measured from pre- and post-treatment CT scans, and a decrease in SMI≥5% was defined as muscle loss. We evaluated five ML models to predict muscle loss, and their performance was determined using the area under the receiver operating characteristic curve (AUC) and F1 score. The features for analysis included demographic and disease-specific characteristics and relative changes in body mass index (BMI), albumin, neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR). The SHAP method was applied to determine the importance of the features and interpret the ML models. The median (inter-quartile range) age of the cohort was 52 (46-59) years. After treatment, 204 patients (33.1%) experienced muscle loss in the training and test datasets, while 44 (31.4%) patients experienced muscle loss in the external validation dataset. Among the five evaluated ML models, the random forest model achieved the highest AUC (0.856, 95% confidence interval: 0.854-0.859) and F1 score (0.726, 95% confidence interval: 0.722-0.730). In the external validation, the random forest model outperformed all ML models with an AUC of 0.874 and an F1 score of 0.741. The results of the SHAP method showed that the albumin change, BMI change, malignant ascites, NLR change, and PLR change were the most important factors in muscle loss. At the patient level, SHAP force plots demonstrated insightful interpretation of our random forest model to predict muscle loss. Explainable ML model was developed using clinical data to identify patients experiencing muscle loss after treatment and provide information of feature contribution. Using the SHAP method, clinicians may better understand the contributors to muscle loss and target interventions to counteract muscle loss.

Diabetes mellitus (DM) self management can improve health problems due to diabetes mellitus. DM Patients are susceptible to various infections due to decreased immunity. Research on the effect of self management on the immune status of patients with diabetes mellitus is still limited. Increased blood glucose for a long time and not getting good management can have an impact on all body systems, including the immune system which triggers the emergence of infectious diseases and the public perception of DM patients getting sick easily becomes something that is considered true. This study aimed to investigated the relationship between self management and the immune status of diabetes mellitus patients. We use questionnaires concerning self management and seeing the results of laboratory examinations for leukocyte types. This study involved the participation of 40 patients. The median score of self management was 2,5 days in 7 days prior to treatment. the lowest self-management behavior was physical exercise and foot care with a median was 1.3 and 1.25 days, and the highest was diet regulation behavior as much as 2.8 day. The median of leukocytes, lymphocytes and monocytes were 12.4, 25.2, and 4.2 mg/dl. There was no correlation between self management and the immune status of DM. Although there is no relationship between self management and immune status but all patients have self management behavior below 50%. Self management also needs to be improved so that the immune status is better.

Objective To systematically evaluate the clinical value of tenofovir combined with recombinant human interferon α-2b in the treatment of chronic hepatitis B and to provide evidence-based medicine for its popularization and use. Methods The randomized controlled trials (RCTs) of tenofovir combined with recombinant human interferon α-2b in the online database of PubMed, EMBASE, ScienceDirect, Cochrane Library, China knowledge Network (CNKI), China VIP database, Wanfang database, and China Biomedical Literature Database (CBM) were searched. The data included in this study were extracted by two independent researchers. After extracting the data of the study, the Cochrane manual 5.1.0 standard was used to evaluate the bias risk of all the literature included in this study. RevMan5.4 statistical software was used to analyze the collected data by meta. Results Entecavir combined with recombinant human interferon α-2b can inhibit the activity of HBV polymerase and improve the inflammatory response of the liver. Recombinant human interferon α-2b can regulate immune function by inducing T cell differentiation and maturation and enhancing the production of cytokines. The systematic evaluation showed that entecavir combined with recombinant human interferon α-2b had higher serum HBeAg negative conversion rate, higher drug safety compared with entecavir alone, and improved liver function and immune status. Conclusion Tenofovir combined with recombinant human interferon alpha-2b has a high serum HBeAg negative rate and safety profile for the treatment of chronic hepatitis B. The combination treatment can improve liver function and immune status in patients, but more studies with higher methodological quality and longer duration of intervention are needed for further validation.

This study was designed to assess and analyze nutritional status (NS) and immune status in patients with tuberculosis. A retrospective analysis was conducted on 93 TB patients hospitalized in the tuberculosis ward of the West China Hospital of Sichuan University. Subgroup comparisons were made according to age (<65 years and ≥65 years), nutritional risk score 2002 (NRS 2002 <3 and ≥3), tuberculosis location [pulmonary tuberculosis and extrapulmonary tuberculosis (including pulmonary tuberculosis complicated with extrapulmonary tuberculosis)], and prognostic nutrition index (PNI) (<45 vs ≥45). Significantly increased weight loss was associated with extrapulmonary tuberculosis (P =0.0010). Serum albumin (P =0.0214), total lymphocyte count (P = 0.0009) and PNI (P = 0.0033) were significantly decreased in older patients. Neutrophils/lymphocytes (NLR) (P =0.0002), monocytes/lymphocytes (MLR) (P < 0.0001), and platelets/lymphocytes (PLR) (P =0.0107) were higher. According to NRS 2002, higher nutritional risk was associated with lower body weight and body mass index (BMI) (P < 0.0001), higher weight loss (P = 0.0012), longer duration of hospitalization (P =0.0100), lower serum albumin level and hemoglobin concentration (P <0.01), lower creatinine level, and lower PNI (P < 0.01). 0.0001), lower total lymphocyte count (P = 0.0004), higher neutrophil and monocyte counts (P <0.05), and higher NLR (P = 0.0002), MLR (P = 0.0006), and PLR (P = 0.0156). Lower PNI was associated with lower body weight (P = 0.0001) and BMI (P =0.0074), lower total protein, albumin, and hemoglobin concentrations (P < 0.0001), and lower total lymphocyte count (P < 0.0001) and creatinine levels (P = 0.0336), higher age (P =0.0002) and NRS 2002 score, P < 0.0001), longer hos-pital stay (P = 0.0003), higher neutrophil count (P = 0.0042), and NLR, MLR, and PLR (P <0.0001) were significantly correlated. In multivariate logistic regression analysis, weight loss (OR: 0.209, 95% CI: 0.060-0.722; p =0.013) was significantly associated with higher nutritional risk (NRS 2002≥3). In multiple linear regression analysis, the NRS 2002 score was higher (B=2.018; p =0.023), and extrapulmonary tuberculosis (B=-6.205; p =0.007) was linked with a longer duration of hos-pitalization. Older tuberculosis patients are at nutritional risk, and older patients (≥65 years old) need to pay attention to nutritional monitoring and intervention. Older TB patients and those at risk of malnutrition have increased immune ratio and impaired immune function. Management of TB patients using basic diagnostic tools to assess nutritional and immune status and calculate PNI and immunological indexes (NLR, MLR, PLR) to improve treatment outcomes.

Immune state monitoring is an expensive, invasive and sometimes difficult necessity in patients with different disorders. Immune reaction dynamics study in patients with coronary atherosclerosis provides one of the leading components to complication development, clinical course prognosis and treatment and rehabilitation tactics. We've chosen intravenous glucose injection as metabolic irritant in the following four groups of patients: men with proved coronary atherosclerosis (CA), non insulin dependent diabetes mellitus (NIDDM), men hereditary burden by CA and NIDDM and practically healthy persons with longlivers in generation. Immune state parameters such as quantity of leukocytes and lymphocytes, circulating immune complexes levels, serum immunoglobulin levels, HLA antigen markers were studied at 0, 30 and 60 minutes during glucose loading. To obtain continues time function of studied parameters received data were approximated by polynomials of high degree with after going first derivatives. Time functions analyze elucidate principally different dynamics studied parameters in all chosen groups of patients, which couldn't be obtained from discontinuous data compare. Leukocyte and lymphocyte levels dynamics correlated HLA antigen markers in all studied groups. Analytical estimation of immune state in patients with coronary atherosclerosis shows the functional margin of safety of immune system state under glucose disturbance. Proposed method of analytical estimation also can be used in immune system monitoring in other groups of patients.

The aim of the study was to identify differences in the immune system parameters between metastatic melanoma patients who responded and did not respond to dendritic cell vaccination.Material and Methods. The study group included 20 patients with stage III–IV metastatic melanoma, who received vaccine therapy with dendritic cells (DC) in a prophylactic mode. The control groups included 13 patients who had symptoms of disease progression at the time of starting vaccine therapy, and 5 healthy donors. The DC-vaccine was prepared in the form of a suspension of the patient’s autologous dendritic cells loaded with tumor antigens in vitro. A single dose had 2 million dendritic cells in 1 ml of phosphate buffer solution, which was administered intradermally in the nearest site to the regional lymphatic collectors. The immune system status was assessed before starting vaccination. The immune system status was evaluated according to the indices of 25 peripheral blood cell populations using multicolor flow cytometry and integral characteristic in the form of the visual image generated by the visualization method of multidimensional data (NovoSpark, Canada).Results. The immune status in patients with metastatic melanoma at the start of DC-vaccination differed and was associated with the effectiveness of subsequent vaccine therapy. The response to vaccination was observed in patients whose immune system status was similar to that of healthy individuals. Low efficacy of DC-vaccine therapy was shown in patients whose immune system status corresponded to that of patients with disease progression. Alterations of the immune system in patients with metastatic melanoma were registered both at the level of individual immunological parameters and at the level of visualized integral characteristics. The integral characteristics of the immune system associated with the patient’s immunocompromised status can be considered as a criterion for stratification of patients with metastatic melanoma for the effective DC-vaccine therapy.Conclusion. The effectiveness of vaccine therapy with dendritic cells in patients with metastatic melanoma is associated with the immune system state before starting this therapy.

Machine learning (ML) models are being increasingly employed to predict the risk of developing and progressing diabetic kidney disease (DKD) in patients with type 2 diabetes mellitus (T2DM). However, the performance of these models still varies, which limits their widespread adoption and practical application. Therefore, we conducted a systematic review and meta-analysis to summarize and evaluate the performance and clinical applicability of these risk predictive models and to identify key research gaps. We conducted a systematic review and meta-analysis to compare the performance of ML predictive models. We searched PubMed, Embase, the Cochrane Library, and Web of Science for English-language studies using ML algorithms to predict the risk of DKD in patients with T2DM, covering the period from database inception to April 18, 2024. The primary performance metric for the models was the area under the receiver operating characteristic curve (AUC) with a 95% confidence interval (CI). The risk of bias was assessed using the Prediction Model Risk of Bias Assessment Tool (PROBAST) checklist. 26 studies that met the eligibility criteria were included into the meta-analysis. 25 studies performed internal validation, but only 8 studies conducted external validation. A total of 94 ML models were developed, with 81 models evaluated in the internal validation sets and 13 in the external validation sets. The pooled AUC was 0.839 (95% CI 0.787-0.890) in the internal validation and 0.830 (95% CI 0.784-0.877) in the external validation sets. Subgroup analysis based on the type of ML showed that the pooled AUC for traditional regression ML was 0.797 (95% CI 0.777-0.816), for ML was 0.811 (95% CI 0.785-0.836), and for deep learning was 0.863 (95% CI 0.825-0.900). A total of 26 ML models were included, and the AUCs of models that were used three or more times were pooled. Among them, the random forest (RF) models demonstrated the best performance with a pooled AUC of 0.848 (95% CI 0.785-0.911). This meta-analysis demonstrates that ML exhibit high performance in predicting DKD risk in T2DM patients. However, challenges related to data bias during model development and validation still need to be addressed. Future research should focus on enhancing data transparency and standardization, as well as validating these models' generalizability through multicenter studies. https://inplasy.com/inplasy-2024-9-0038/, identifier INPLASY202490038.

Predicting Choroidal Nevus Transformation to Melanoma Using Machine Learning

Introduction Pulmonary vein isolation is superior to antiarrhythmic drugs for rhythm control in patients with atrial fibrillation (AF). Nevertheless, recurrences within the first year are still significant. Several clinical risk scores were developed to predict AF recurrence; however, despite good results in development cohorts, the results in external validation are usually poor, limiting the general applicability of these scores. Purpose We aimed to develop machine learning (ML) models to predict AF recurrences within the first year after catheter ablation and compare their performance with conventional risk scores. Methods We used a retrospective dataset of a tertiary hospital center, including all consecutive patients submitted to the first AF ablation between 2017 and 2021. The dataset included 76 features: patient characteristics, medical history, baseline echocardiography, calcium score, procedure variables, and early recurrence (ER) during the 90-day blanking period. The outcome was defined as the occurrence of an electrocardiographically documented late recurrence (LR) of atrial tachycardia/AF &amp;gt; 30s between 3-12 months. Three different supervised ML models (penalized logistic regression, random forests, and XGBoost) were developed on the training cohort, and hyperparameters were tuned using 10-fold cross-validation. A testing dataset was held out to estimate the final performance (25%). The following risk scores were calculated: APPLE, CAAP-AF, and BASE-AF2 (the last includes ER). Areas under the curve (AUC) of receiver operating characteristic curves were compared using DeLong’s test. Results 679 patients were included: 62.7% males, median age of 59±16 years, and 78.2% with paroxysmal AF. The median time from diagnosis to ablation was 2±4 years. 74.2% underwent radiofrequency ablation and the remaining cryoablation. ER occurred in 12.5%, and most of these patients also experienced LR (68.7 vs 19.6% in those without ER, p&amp;lt;0.001). LR was observed in 25.6%. The XGBoost model showed the best performance with an AUC 0.774, 95% confidence interval (CI) 0.688–0.844) outperforming existing scores (picture 1): APPLE score AUC 0.607, 95% CI 0.562–0.653, p-value&amp;lt;0.001; CAAP-AF score AUC 0.622, 95% CI 0.576–0.668, p-value=0.002; BASE-AF2 AUC 0.628, 95% CI 0.581–0.675, p-value= 0.003. Variable importance analysis showed a significant drop in the performance of the models when ER was not considered, indicating its high importance in predicting LR (picture 2). Conclusions Recurrence during the blanking period was the most important predictor of LR in our population. The ML model was superior to conventional risk scores. ML models might be an essential tool to improve the prediction of outcomes and clinical decision-making for optimal follow-up.ROC curvesFeature Importance

To investigate the association of genetic polymorphisms in MB21D1 (Mab-21 domain-containing 1), TMEM173 (Transmembrane Protein 173), IFNB1 (Interferon beta 1), IFNGR1 (Interferon gamma receptor 1), IFNGR2 (Interferon gamma receptor 2), IRF3 (Interferon Regulatory Factor 3), and IRF8 (Interferon Regulatory Factor 8) with susceptibility to non-tuberculous mycobacteria pulmonary disease (NTM-PD) as well as their correlation with the treatment outcomes and immune status of patients. Forty-four tagSNPs from the candidate genes were genotyped in a 2-phase cohort study including an initial discovery phase involving 707 NTM-PD patients and 726 healthy controls and a replication phase involving 357 NTM-PD patients and 400 controls. The frequencies and distributions of genotypes were compared between the case and control groups. Treatment success rates, sputum culture conversion rates, imaging characteristics, and peripheral blood immunological indices were compared among patients with different genotypes. Individuals with the IFNGR1 rs2234711 A/A genotype were more susceptible to MAC-PD compared to those with the G/G genotype (discovery phase OR = 1.752, P.adj = 0.025; replication phase OR = 2.143, P.adj = 0.019). Patients with the rs2234711 A/A genotype exhibited significantly lower treatment success rates and sputum culture conversion rates, along with elevated levels of peripheral blood heparin-binding protein (HBP), erythrocyte sedimentation rate, and interleukin-10, but significantly decreased interleukin-1β levels (all P < 0.05). Individuals with the IRF8 rs2280378 A/A genotype were more susceptible to MAB-PD (discovery phase OR = 2.302, P.adj = 0.014; replication phase OR = 2.465, P.adj = 0.015). Compared to G/G genotype patients, those with the rs2280378 A/A genotype exhibited lower treatment success rates and sputum culture conversion rates, were more likely to develop pulmonary cavities and multiple lung field involvement, and showed elevated levels of peripheral blood HBP and C-reactive protein, along with significantly reduced levels of serum interleukin-12 P70, tumor necrosis factor-α, and CD8 + T lymphocytes (all P < 0.05). In the Chinese Han population, IFNGR1 genetic polymorphisms are closely associated with MAC-PD susceptibility, while IRF8 genetic polymorphisms are associated with MAB-PD susceptibility. Variants in IFNGR1 and IRF8 significantly affect the immune status and treatment outcomes of MAC-PD and MAB-PD patients, respectively.

The systemic immune inflammation (SII) index provides a comprehensive assessment of inflammatory and immune status in patients with different diseases. However, it remains unclear whether the SII can be considered a valuable prognostic risk factor of all-cause mortality for postmenopausal women. We analyzed data from 1882 postmenopausal women enrolled in the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2018. The Systemic Immune-Inflammation Index (SII) was calculated using peripheral blood cell counts and categorized into quartiles. Multivariable Cox proportional hazards models and restricted cubic spline analyses were employed to assess the association between SII and mortality outcomes. Over a median follow-up period of 8 years, 13.5% individuals died, with 4% deaths attributed to CVD. Patients with extremely high SII levels experienced significantly higher all-cause and CVD mortality. Compared to the low SII group (Q1), the hazard ratio (HR) and 95% confidence interval (CI) for all-cause mortality risk were 0.96 (0.87, 1.07), 0.97 (0.88, 1.08), and 1.28 (1.16, 1.41) for Q2, Q3, and Q4, respectively. Similarly, the HR (95% CI) for CVD mortality in Q2, Q3, and Q4 were 1.02 (0.83, 1.24), 1.11 (0.92, 1.34), and 1.32 (1.10, 1.58), respectively. Including SII in addition to traditional risk factors resulted in a slight enhancement in mortality prediction capability. Among postmenopausal women, extremely high SII levels were identified as an independent risk factor for all-cause and CVD mortality.

The dynamic monitoring of immune status is crucial to the precise and individualized treatment of sepsis. In this study, we aim to introduce a model to describe and monitor the immune status of sepsis and to explore its prognostic value. A prospective observational study was carried out in Zhongshan Hospital, Fudan University, enrolling septic patients admitted between July 2016 and December 2018. Blood samples were collected at days 1 and 3. Serum cytokine levels (e.g., tumor necrosis factor-α [TNF-α], interleukin-10 [IL-10]) and CD14+ monocyte human leukocyte antigen-D-related (HLA-DR) expression were measured to serve as immune markers. Classification of each immune status, namely systemic inflammatory response syndrome (SIRS), compensatory anti-inflammatory response syndrome (CARS), and mixed antagonistic response syndrome (MARS), was defined based on levels of immune markers. Changes of immune status were classified into four groups which were stabilization (SB), deterioration (DT), remission (RM), and non-remission (NR). A total of 174 septic patients were enrolled including 50 non-survivors. Multivariate analysis discovered that IL-10 and HLA-DR expression levels at day 3 were independent prognostic factors. Patients with MARS had the highest mortality rate. Immune status of 46.1% patients changed from day 1 to day 3. Among four groups of immune status changes, DT had the highest mortality rate, followed by NR, RM, and SB with mortality rates of 64.7%, 42.9%, and 11.2%, respectively. Severe immune disorder defined as MARS or deterioration of immune status defined as DT lead to the worst outcomes. The preliminary model of the classification and dynamic monitoring of immune status based on immune markers has prognostic values and is worthy of further investigation.