Competing Risk Event Research Articles

Latent class analysis-derived classification improves the cancer-specific death stratification of lymphomas: A large retrospective cohort study.

Lymphomas have diverse etiologies, treatment approaches, and prognoses. Accurate survival estimation is challenging for lymphoma patients due to their heightened susceptibility to non-lymphoma-related mortality. To overcome this challenge, we propose a novel lymphoma classification system that utilizes latent class analysis (LCA) and incorporates demographic and clinicopathological factors as indicators. We conducted LCA using data from 221,812 primary lymphoma patients in the Surveillance, Epidemiology, and End Results (SEER) database and identified four distinct LCA-derived classes. The LCA-derived classification efficiently stratified patients, thereby adjusting the bias induced by competing risk events such as non-lymphoma-related death. This remains effective even in cases of limited availability of cause-of-death information, leading to an enhancement in the accuracy of lymphoma prognosis assessment. Additionally, we validated the LCA-derived classification model in an external cohort and observed its improved prognostic stratification of molecular subtypes. We further explored the molecular characteristics of the LCA subgroups and identified potential driver genes specific to each subgroup. In conclusion, our study introduces a novel LCA-based lymphoma classification system that provides improved prognostic prediction by accounting for competing risk events. The proposed classification system enhances the clinical relevance of molecular subtypes and offers insights into potential therapeutic targets.

Hepatocellular Carcinoma risk scores for non-viral liver disease: a systematic review and meta-analysis

BackgroundHepatocellular carcinoma (HCC) risk prediction models may provide a more personalised approach to surveillance for HCC among patients with cirrhosis. This systematic review aims to summarise the performance of HCC prediction models in patients with non-viral chronic liver disease. MethodThe study was prospectively registered with PROSPERO (ID: CRD42022370078) and reported in accordance with PRISMA guidelines. MEDLINE and EMBASE databases were searched using a validated search filter for prediction model studies. Two reviewers independently assessed studies for inclusion and risk of bias. Model performance (discrimination and calibration) were identified to assess the risk of HCC at specified time points. A random effects meta-analysis was done on a subset of studies that reported performance of the same model. Results7,854 studies were identified. After review, 14 studies with a total of 94,014 participants were included. 45% of patients had viral hepatitis, 27% alcohol related liver disease (ArLD) and 19% metabolic associated steatotic liver disease (MASLD). Follow-up ranged from 15.1 - 138 months. Only one model was developed using a competing risk approach. Age (7 models) and sex (6 models) were the most frequently included predictors. Model discrimination (AUROC or c-statistic) ranged from 0.61 - 0.947. Only the ‘aMAP’ score (age, male sex, albumin,bilirubin, and platelets) had sufficient external validation for quantitative analysis, with a pooled c-statistic of 0.81 (95% CI 0.80-0.83). Calibration was reported in only 9 of 14 studies. All studies were rated high risk of bias. ConclusionStudies describing risk prediction of HCC in non-viral chronic liver disease are poorly reported, have a high risk of bias and do not account for competing risk events. Patients with ArLD and MASLD are underrepresented in development and validation cohorts. These factors remain barriers to the clinical utility and uptake of HCC risk models in persons with the commonest liver diseases. Impact and implicationsThe recent EASL policy statement emphasises the potential of risk-based surveillance to reduce both HCC-related deaths and surveillance costs. This study addresses the gap in understanding the performance of current HCC risk models in patients with non-viral liver diseases, reflecting the epidemiological landscape of liver disease in Western countries.In our review of these models, we identified several key concerns regarding reporting standards and risk of bias and confirmed that patients with alcohol-related liver disease (ArLD) and metabolic-associated steatotic liver disease (MASLD) are underrepresented in model development and validation cohorts. Additionally, most models fail to account for the significant risk of competing events, leading to potential overestimation of true HCC risk.This study highlights these critical issues that may hinder the implementation of risk models in clinical practice and offers key recommendations for future model development studies.

Comparative effectiveness of pneumococcal vaccination strategies to prevent invasive pneumococcal disease: a population-based cohort study at the veterans health administration in the United States

ObjectivesComparative effectiveness of 13-valent pneumococcal conjugate vaccine (PCV13), 23-valent polysaccharide vaccine (PPSV23), and their combinations for adults in preventing invasive pneumococcal disease (IPD) has yet to be thoroughly investigated. We aimed to estimate the comparative effectiveness of preventing IPD, using population-based data from the Veterans Health Administration (VHA) in the United States. MethodsWe included all patients who were between 65 and 74 years, had established primary care within VHA between 2005 and 2021, and had not received any prior pneumococcal vaccination. We measured time-to-event from cohort enrolment to the onset of IPD, considering death a competing risk event, and used Cox regression models to estimate cause-specific hazards. PCV13 only, PPSV23 only, PCV13 after PPSV23, and PPSV23 after PCV13 were incorporated into models as time-dependent covariates. Patient demographics and comorbidities were also included in the model. ResultsA total of 3 044 067 patients were enrolled in the cohort, with 863 958 deaths (28.4%) and 1731 cases of IPD (0.06%) during the study period. The overall incidence rate of IPD in this population was 5.36 per 100 000 patient-years. A total of 921 070 patients (30.3%) received at least one dose of effective pneumococcal vaccine. In multivariate analysis adjusted for comorbidities, PCV13 alone was not associated with the reduced risk of IPD, whereas PPSV23 had protective association with IPD incidence (adjusted hazard ratio [aHR], 0.70; 95% CI, 0.59–0.83). When combined, PCV13 followed by PPSV23 had a stronger protective association (aHR, 0.54; [0.36–0.83]) compared with PPSV23 followed by PCV13 (aHR, 0.73; [0.58–0.91]). Discussion: In this large cohort study at the VHA, the combination of PCV13 and PPSV23, particularly PCV13 followed by PPSV23, was associated with a lower risk of IPD, indicating additional benefits in combined vaccinations with potential importance in vaccination order. Further studies are needed to evaluate the effect of newer pneumococcal vaccines.

Link between remnant cholesterol and the reversion to normoglycemia in Chinese adults with prediabetes: a 5-year cohort study

Despite the clear association between remnant cholesterol (RC)and diabetes risk, no study to date has examined the relationship between RC and reversal of prediabetes to normoglycemia. This retrospective cohort study included a total of 15,023 patients with prediabetes who underwent a physical examination between 2010 and 2016. The link between initial RC levels and the reversion from prediabetes to normoglycemia was analyzed using the Cox proportional-hazards regression model. Additionally, the study explored the possible relationship between RC and the probability of returning normoglycemia by applying Cox proportional hazards regression models with cubic spline functions. To address competing risks, a multivariate Cox regression analysis was undertaken, treating the onset of diabetes as a competing risk event for reversing prediabetes to normoglycemia. Additionally, the study incorporated extensive subgroup analyses alongside multiple sensitivity analyses, enhancing the reliability and robustness of the results. After adjusting for covariates, the findings indicated that RC was inversely associated with the likelihood of reverting to normoglycemia (per 5 mg/dL increase, HR = 0.918, 95% CI 0.909–0.927). The analysis also revealed a nonlinear relationship between RC and normoglycemia reversion, with an inflection point at 51.08 mg/dL. For RC values below this inflection point (RC < 50.08 mg/dL), the HR for the probability of returning to normoglycemia was 0.907 (95% CI 0.897–0.917 per 5 mg/dL). Additionally, the competing risks model demonstrated a negative relationship between RC and the reversal of prediabetes to normoglycemia (SHR = 0.92, 95% CI 0.91–0.93). Sensitivity analyses confirmed the robustness and stability of these results. This study demonstrated a negative and non-linear association between RC and the probability of reversion to normoglycemia in Chinese adults with prediabetes. By actively intervening to reduce RC levels, at least to below 51.08 mg/dL, further reduction of RC may significantly increase the probability of returning to normoglycemia from prediabetes.

A Nomogram for Predicting Cancer-Specific Survival in Young Patients With Advanced Lung Cancer Based on Competing Risk Model.

Young lung cancer is a rare subgroup accounting for 5% of lung cancer. The aim of this study was to compare the causes of death (COD) among lung cancer patients of different age groups and construct a nomogram to predict cancer-specific survival (CSS) in young patients with advanced stage. Lung cancer patients diagnosed between 2004 and 2015 were extracted from the Surveillance, Epidemiology, and End Results (SEER) database and stratified into the young (18-45 years) and old (> 45 years) groups to compare their COD. Young patients diagnosed with advanced stage (IVa and IVb) from 2010 to 2015 were reselected and divided into training and validation cohorts (7:3). Independent prognostic factors were identified through the Fine-Gray's test and further integrated to the competing risk model. The area under the receiver operating characteristic curve (AUC), consistency index (C-index), and calibration curve were applied for validation. The proportion of cancer-specific death (CSD) in young patients was higher than that in old patients with early-stage lung cancer (p < 0.001), while there was no difference in the advanced stage (p = 0.999). Through univariate and multivariate analysis, 10 variables were identified as independent prognostic factors for CSS. The AUC of the 1-, 3-, and 5-year prediction of CSS was 0.688, 0.706, and 0.791 in the training cohortand 0.747, 0.752, and 0.719 in the validation cohort. The calibration curves demonstrated great accuracy. The C-index of the competing risk model was 0.692 (95% CI: 0.636-0.747) in the young patient cohort. Young lung cancer is a distinct entity with a different spectrum of competing risk events. The construction of our nomogramcan provide new insights into the management of young patients with lung cancer.

Constructing a prognostic model for colorectal cancer with synchronous liver metastases after preoperative chemotherapy: a study based on SEER and an external validation cohort.

The combination of preoperative chemotherapy and surgical treatment has been shown to significantly enhance the prognosis of colorectal cancer with liver metastases (CRLM) patients. Nevertheless, as a result of variations in clinicopathological parameters, the prognosis of this particular group of patients differs considerably. This study aimed to develop and evaluate Cox proportional risk regression model and competing risk regression model using two patient cohorts. The goal was to provide a more precise and personalized prognostic evaluation system. We collected information on individuals who had a pathological diagnosis of colorectal cancer between 2000 and 2019 from the Surveillance, Epidemiology, and End Results (SEER) Database. We obtained data from patients who underwent pathological diagnosis of colorectal cancer and got comprehensive therapy at the hospital between January 1, 2010, and June 1, 2022. The SEER data collected after screening according to the inclusion and exclusion criteria were separated into two cohorts: a training cohort (training cohort) and an internal validation cohort (internal validation cohort), using a random 1:1 split. Subgroup Kaplan-Meier (K-M) survival analyses were conducted on each of the three groups. The data that received following screening from the hospital were designated as the external validation cohort. The subsequent variables were chosen for additional examination: age, gender, marital status, race, tumor site, pretreatment carcinoembryonic antigen level, tumor size, T stage, N stage, pathological grade, number of tumor deposits, perineural invasion, number of regional lymph nodes examined, and number of positive regional lymph nodes. The primary endpoint was median overall survival (mOS). In the training cohort, we conducted univariate Cox regression analysis and utilized a stepwise regression approach, employing the Akaike information criterion (AIC) to select variables and create Cox proportional risk regression models. We evaluated the accuracy of the model using calibration curve, receiver operating characteristic curve (ROC), and area under curve (AUC). The effectiveness of the models was assessed using decision curve analysis (DCA). To evaluate the non-cancer-related outcomes, we analyzed variables that had significant impacts using subgroup cumulative incidence function (CIF) and Gray's test. These analyses were used to create competing risk regression models. Nomograms of the two models were constructed separately and prognostic predictions were made for the same patients in SEER database. This study comprised a total of 735 individuals. The mOS of the training cohort, internal validation cohort, and QDU cohort was 55.00months (95%CI 46.97-63.03), 48.00months (95%CI 40.65-55.35), and 68.00months (95%CI 54.91-81.08), respectively. The multivariate Cox regression analysis revealed that age, N stage, presence of perineural infiltration, number of tumor deposits and number of positive regional lymph nodes were identified as independent prognostic risk variables (p < 0.05). In comparison to the conventional TNM staging model, the Cox proportional risk regression model exhibited a higher C-index. After controlling for competing risk events, age, N stage, presence of perineural infiltration, number of tumor deposits, number of regional lymph nodes examined, and number of positive regional lymph nodes were independent predictors of the risk of cancer-specific mortality (p < 0.05). We have developed a prognostic model to predict the survival of patients with synchronous CRLM who undergo preoperative chemotherapy and surgery. This model has been tested internally and externally, confirming its accuracy and reliability.

Association of Circulating 25-Hydroxyvitamin D and Recurrence of Glomerulonephritis in Kidney Transplant Recipients: The Wisconsin Allograft Recipient Database (WisARD).

Recurrence of glomerulonephritis (GN) is a significant contributor to long-term allograft failure among kidney transplant recipients (KTRs) with kidney failure because of GN. Accumulating evidence has revealed the role of vitamin D in both innate and adaptive immunity. Although vitamin D deficiency is common among KTRs, the association between 25-hydroxyvitamin D (25[OH]D) and GN recurrence in KTRs remains unclear. We analyzed data from KTRs with kidney failure caused by GN who received a transplant at our center from 2000 to 2019 and had at least 1 valid posttransplant serum 25(OH)D measurement. Survival analyses were performed using a competing risk regression model considering other causes of allograft failure, including death, as competing risk events. A total of 67 cases of GN recurrence were identified in 947 recipients with GN followed for a median of 7.0 y after transplant. Each 1 ng/mL lower serum 25(OH)D was associated with a 4% higher hazard of recurrence (subdistribution hazard ratio [HR]: 1.04; 95% confidence interval [CI], 1.01-1.06). Vitamin D deficiency (≤20 ng/mL) was associated with a 2.99-fold (subdistribution HR: 2.99; 95% CI, 1.56-5.73) higher hazard of recurrence compared with vitamin D sufficiency (≥30 ng/mL). Results were similar after further adjusting for concurrent urine protein-creatinine ratio, serum albumin, and estimated glomerular filtration rate (eGFR). Posttransplant vitamin D deficiency is associated with a higher hazard of GN recurrence in KTRs. Further prospective observational studies and clinical trials are needed to determine any causal role of vitamin D in the recurrence of GN after kidney transplantation. More in vitro and in vivo experiments would be helpful to understand its effects on autoimmune and inflammation processes.

Competing-risks model for predicting the prognosis of patients with angiosarcoma based on the SEER database of 3905 cases

PurposeTo establish a competing-risks model and compare it with traditional survival analysis, aiming to identify more precise prognostic factors for angiosarcoma. The presence of competing risks suggests that prognostic factors derived from the conventional Cox regression model may exhibit bias.MethodsPatient data pertaining to angiosarcoma cases diagnosed from 2000 to 2019 were extracted from the Surveillance, Epidemiology, and End Results (SEER) database. Multivariate analysis employed both the Cox regression model and the Fine-Gray model, while univariate analysis utilized the cumulative incidence function and Gray’s test.ResultsA total of 3,905 enrolled patients diagnosed with angiosarcoma were included, out of which 2,781 succumbed to their condition: 1,888 fatalities resulted from angiosarcoma itself, and 893 were attributed to other causes. The Fine-Gray model, through multivariable analysis, identified SEER stage, gender, race, surgical status, chemotherapy status, radiotherapy status, and marital status as independent prognostic factors for angiosarcoma. The Cox regression model, due to the occurrence of competing-risk events, could not accurately estimate the effect values and yielded false-negative outcomes. Clearly, when analyzing clinical survival data with multiple endpoints, the competing-risks model demonstrates superior performance.ConclusionThis current investigation may enhance clinicians’ comprehension of angiosarcoma and furnish reference data for making clinical decisions.

Development and validation of a major adverse limb events prediction model for peripheral arterial disease with frailty

ObjectiveTo investigate the risk factors for major limb adverse events (MALE) in peripheral arterial disease (PAD) combined with frailty and to develop and validate a risk prediction model of MALE. MethodsThis prospective study was performed in the vascular surgery department of patients in six hospitals in southwest China. Prospective collection of patients with PAD combined with frailty from February 1 to December 20, 2021, with MALE as the primary outcome, and followed for 1 year. The cohort was divided into a development cohort and a validation cohort. In the development cohort, a multivariate risk prediction model was developed to predict MALE using random forests for variable selection and multivariable Cox regression analysis. The model is represented by a visualized nomogram and a web-based calculator. The model performance was tested with the validation cohort and assessed using the C-statistic and calibration plots. ResultsA total of 1179 patients were prospectively enrolled from February 1 to December 20, 2021. Among 816 patients with PAD who were included in the analysis, the median follow-up period for this study was 9 ± 4.07 months, the mean age was 74.64 ± 9.43 years, and 249 (30.5%) were women. Within 1 year, 222 patients (27.2%) developed MALE. Target lesion revascularizations were performed in 99 patients (12.1%), and amputations were performed in 131 patients (16.1%). The mortality rate within the whole cohort was 108 patients (13.2%). After controlling for competing risk events (death), the cumulative risk of developing MALE was not statistically different. Prealbumin (hazard ratio [HR], 0.6; 95% confidence interval [CI], 0.41-0.89; P = .010), percutaneous coronary intervention (HR, 2.31; 95% CI, 1.26-4.21; P = .006), Rutherford classification (HR, 1.77; 95% CI, 1.36-2.31; P < .001), white blood cell (HR, 1.85; 95% CI, 1.20-2.87; P = .005), high altitude area (HR, 3.1; 95% CI, 1.43-6.75; P = .004), endovascular treatment (HR, 10.2; 95% CI, 1.44-72.50; P = .020), and length of stay (HR, 1.01; 95% CI, 1.00-1.03; P = .012) were risk factors for MALE. The MALE prediction model had a C-statistic of 0.76 (95% CI, 0.70-0.79). The C-statistic was 0.68 for internal validation and 0.66 for external validation for the MALE prediction model. The MALE prediction model for PAD presented an interactive nomogram and a web-based network calculator. ConclusionsIn this study, the MALE prediction model has a discriminative ability to predict MALE among patients with PAD in frailty. The MALE model can optimize clinical decision-making for patients with PAD in frailty.

Association between long-term use of calcium channel blockers (CCB) and the risk of breast cancer: a retrospective longitudinal observational study protocol

IntroductionCalcium channel blockers (CCB), a commonly prescribed antihypertensive (AHT) medicine, may be associated with increased risk of breast cancer. The proposed study aims to examine whether long-term CCB use is...

Joint modeling of longitudinal and competing risks for assessing blood oxygen saturation and its association with survival outcomes in COVID-19 patients.

The objective of the present study is to evaluate the association between longitudinal and survival outcomes in the presence of competing risk events. To illustrate the application of joint modeling in clinical research, we assessed the blood oxygen saturation (SPO2) and its association with survival outcomes in coronavirus disease (COVID-19). In this prospective cohort study, we followed 300 COVID-19 patients, who were diagnosed with severe COVID-19 in the Rohani Hospital in Babol, the north of Iran from October 22, 2020 to March 5, 2021, where death was the event of interest, surviving was the competing risk event and SPO2 was the longitudinal outcome. Joint modeling analyses were compared to separate analyses for these data. The estimation of the association parameter in the joint modeling verified the association between longitudinal outcome SPO2 with survival outcome of death (Hazard Ratio (HR) = 0.33, P = 0.001) and the competing risk outcome of surviving (HR = 4.18, P < 0.001). Based on the joint modeling, longitudinal outcome (SPO2) decreased in hypertension patients (β = -0.28, P = 0.581) and increased in those with a high level of SPO2 on admission (β = 0.75, P = 0.03). Also, in the survival submodel in the joint model, the risk of death survival outcome increased in patients with diabetes comorbidity (HR = 4.38, P = 0.026). The association between longitudinal measurements of SPO2 and survival outcomes of COVID-19 confirms that SPO2 is an important indicator in this disease. Thus, the application of this joint model can provide useful clinical evidence in the different areas of medical sciences.

A Dirichlet process mixture regression model for the analysis of competing risk events

We develop a regression model for the analysis of competing risk events. The joint distribution of the time to these events is flexibly characterized by a random effect which follows a discrete probability distribution drawn from a Dirichlet Process, explaining their variability. This entails an additional layer of flexibility of this joint model, whose inference is robust with respect to the misspecification of the distribution of the random effects. The model is analysed in a fully Bayesian setting, yielding a flexible Dirichlet Process Mixture model for the joint distribution of the time to events. An efficient MCMC sampler is developed for inference. The modelling approach is applied to the empirical analysis of the surrending risk in a US life insurance portfolio previously analysed by Milhaud and Dutang (2018). The approach yields an improved predictive performance of the surrending rates.

Cumulative incidence and risk factors of brain metastases in metastatic non-small cell lung cancer without baseline brain metastasis: Pooled analysis of individualized patient data from IMpower130, IMpower131, and IMpower150.

The objective of this study was to explore the abilities of atezolizumab plus chemotherapy in preventing brain metastases (BMs) among metastatic non-small cell lung cancer (NSCLC) without initial BMs, as well as the risk factors of BMs. Individual patient data from three trials involving first-line atezolizumab for metastatic NSCLC (IMpower130, IMpower131, and IMpower150) were pooled. Among patients without baseline BMs and without epidermal growth factor receptor(EGFR) and/or anaplastic lymphoma kinase (ALK) mutations, those receiving atezolizumab+chemotherapy±bevacizumab were classified as the atezolizumabplus chemotherapy group and those receiving placebo+chemotherapy±bevacizumab were classified as the chemotherapy group. The cumulative incidences of BM (CI-BMs) between the two groups were compared. Other factors associated with the CI-BM were analyzed by Cox regression analyses. With a median follow-up of 17.6 months (range, 0.03-33.64 months), 74 (3.1%) of the 2380 enrolled patients developed BMs, including 50 (3.1%) and 24 (3.0%) in the atezolizumab plus chemotherapy group (n=1589) and the chemotherapy group (n=791), respectively. The CI-BMs at 6, 12, and 24months were 1.7%, 2.8%, and 3.3%, respectively. After taking competing risk events into account, there was no significant difference in the CI-BMs between the two groups (p=.888). Nevertheless, the use of bevacizumab and the histology of nonsquamous NSCLC were found to be independently associated with the risk of BMs. In patients with metastatic EGFR/ALK wild-type NSCLC without baseline BMs, adding atezolizumab in the first-line treatment might not reduce the CI-BM. However, the administration of bevacizumab may reduce the risk of BMs.

Association between beta-blocker utilization and heart failure mortality in the peritoneal dialysis population: a cohort study.

The prognostic significance of beta(β)-blocker therapy in patients at end-stage renal disease, specifically those receiving peritoneal dialysis (PD) and presenting with heart failure, remains inadequately elucidated due to limited research conducted thus far. A retrospective analysis was performed on a cohort comprising 608 patients receiving PD between September 2007 and March 2019, with a subsequent follow-up period extending until December 2020. Cox regression and propensity score matching weighted analysis was used to model adjusted hazard ratios for β-blocker use with heart failure-related mortality. Competing risk analysis and subgroup analysis were carried out to further elucidate the correlation. β-blockers were prescribed for 56.1% of the peritoneal dialysis patients. Heart failure occurred in 43.4% of the total population and 15.5% of deaths were due to heart failure. The prescription of β-blockers was associated with a 43% lower adjusted hazard ratio (HR) for heart failure death within the cohort (95% confidence interval [CI] = 0.36-0.89; P=0.013). Even after accounting for competing risk events, patients in the group using β-blockers demonstrated a significantly lower cumulative risk of heart failure-related mortality compared to those not using β-blockers (P=0.007). This protective effect of β-blockers was also observed in subgroup analyses. Conversely, β-blocker use had no statistically significant associations with all-cause mortality. The use of β-blockers was associated with a reduced risk of heart failure-related mortality in the PD population. Future randomized clinical trials are warranted to confirm the beneficial effect of β-blockers in the context of PD.

Birth Spacing and Working Mothers’ Within-Organization Career Paths

The mechanisms behind mothers’ wage penalties remain unclear. In this article, the authors consider the role of birth spacing and changes in employers after a second birth. Using the 1979 National Longitudinal Survey of Youth and competing risk event history models, the authors investigate how spacing between first and second births influences the likelihood of returning to a pre–second birth employer, changing employers, or remaining outside of the labor force within six months of the second birth. The authors find no differences in the influence of birth spacing on the likelihood of returning to an employer versus changing employers but that shorter birth spacings relate to lower likelihoods of returning to the labor market. There is some evidence that birth spacing and postbirth employment varies by age at first birth, marital status, and occupation. Overall, the results suggest that although birth spacing is relevant for returning postbirth to employment, job changes are unlikely to drive mothers’ wage penalties.

Antidepressant exposure and long‐term dementia risk in a nationwide study on U.S. Veterans with major depressive disorder

AbstractBackgroundMajor depressive disorder (MDD) is associated with dementia risk. The use of antidepressants in MDD has been reported to influence long term risk of Alzheimer’s disease and related dementias (ADRD), but studies are conflicting.MethodsWe used inverse probability weighted Cox models with time‐varying covariates in a retrospective cohort study among Veterans with MDD within the US Veterans Affairs health‐care system from 1/1/2000 to 6/1/2022. We focused on patients receiving their first diagnosis of MDD at the age of 60 or above. Our outcome was incident ADRD and unrelated death was considered as a competing risk event. Exposed intervals were defined using prescription data. Models were weighted for the inverse probability of falling in each of the treatment arms (exposed or not) at the beginning of each treatment period based on demographic and clinical covariates updated with most recent information (1‐year prior). The cumulative effect of antidepressant exposure (No. days) was also studied.ResultA total of 24,964 patients with MDD were identified, of which 45% were on antidepressants (Table 1). Median treatment time was 4.10 years [2.56, 6.10]. No relationship was observed between the number of days on antidepressants and the risk of ADRD (Figure 1). No significant associations were seen for the effect of being exposed to antidepressants (any) vs unexposed (HR = 0.94, 95%; CI: 0.81‐1.09) or exposed to each antidepressant drug classes vs unexposed (SSRI HR = 0.93, 95%; CI: 0.79‐1.10; NDRI HR = 0.80, 95%; CI: 0.42‐1.54; SNRI HR = 0.87, 95%; CI: 0.54‐1.41). (Table II). In sensitivity analysis stratified by sex, a reduction in risk due to antidepressant exposure was suggested in female patients, however the number of cases was small (HR = 0.30, 95%; CI: 0.12‐0.81; P = 0.02)).ConclusionOur study conducted in a large population and after accounting for confounders shows that antidepressant exposure has no effect on ADRD risk. Our result suggesting an association in female patients should be interpreted with caution due to a small sample size and deserves further attention. Replication of results in an external population with a larger representation of female patients is needed for further validation

Predicting de-novo portal vein thrombosis after HCV eradication: A long-term competing risk analysis in the ongoing PITER cohort.

Sustained virological response (SVR) by direct-acting antivirals (DAAs) may reverse the hypercoagulable state of HCV cirrhosis and the portal vein thrombosis (PVT) risk. We evaluated the incidence and predictive factors of de novo, non-tumoral PVT in patients with cirrhosis after HCV eradication. Patients with HCV-related cirrhosis, consecutively enrolled in the multi-center ongoing PITER cohort, who achieved the SVR using DAAs, were prospectively evaluated. Kaplan-Meier and competing risk regression analyses were performed. During a median time of 38.3months (IQR: 25.1-48.7months) after the end of treatment (EOT), among 1609 SVR patients, 32 (2.0%) developed de novo PVT. Aplatelet count ≤120,000/μL, albumin levels ≤3.5mg/dL, bilirubin >1.1mg/dL, a previous liver decompensation, ALBI, Baveno, FIB-4, and RESIST scores were significantly different (p<0.001), among patients who developed PVT versus those who did not. Considering death and liver transplantation as competing risk events, esophageal varices (subHR: 10.40; CI 95% 4.33-24.99) and pre-treatment ALBI grade ≥2 (subHR: 4.32; CI 95% 1.36-13.74) were independent predictors of PVT. After HCV eradication, a significant variation in PLT count, albumin, and bilirubin (p<0.001) versus pre-treatment values was observed in patients who did not develop PVT, whereas no significant differences were observed in those who developed PVT (p>0.05). After the EOT, esophageal varices and ALBI grade ≥2, remained associated with de novo PVT (subHR: 9.32; CI 95% 3.16-27.53 and subHR: 5.50; CI 95% 1.67-18.13, respectively). In patients with HCV-related cirrhosis, a more advanced liver disease and significant portal hypertension are independently associated with the de novo PVT risk after SVR.

Risk-Adapted Donor Lymphocyte Infusion after Hematopoietic Cell Transplant in Pediatric and Young Adult Patients Is Associated with Low Risk of Graft Versus Host Disease

Introduction: Donor lymphocyte infusion (DLI) is a form of adoptive immunotherapy administered to enhance the graft-versus-leukemia (GVL) effect after an allogeneic hematopoietic cell transplantation (HCT). DLI is frequently chosen for its ease of administration and low risk of graft ablation as compared with other chemotherapeutic options in patients experiencing mixed chimerism (MC) or disease relapse. But in patients with acute leukemia experiencing relapse, clinical responses are seen in fewer than 20-30%. Success is also limited by the occurrence of acute and chronic graft-versus-host disease (GVHD). We had previously described our risk-adapted strategy of administering DLIs based on chimerism and donor source (Rujkijyanont P, et al. Blood Cancer J 2013). To determine the efficacy and toxicity of DLIs and to identify potential factors influencing the outcomes, we conducted a retrospective single center study including patients who received DLIs after HCT. Methods: We collected data on all HCTs and conventional DLIs that had been administered to patients from 1993 to 2022 for any indication at our center. Each HCT was considered as an independent event. Numeric variables were represented by median and range and compared using the Wilcoxon-Mann-Whitney test. Categorical variables were compared using the Fisher exact test. Competing-risk regression models using the Fine-Gray method were developed to evaluate risk factors for acute GVHD and chronic GVHD. Death and relapse were considered as competing risk events for these GVHD outcomes. Follow-up data were censored at subsequent HCT, 2 years after the first DLI, 5 years after transplant, or the last follow-up. Logistic regression analysis was performed to assess factors associated with the response to DLI within 180 days after the first DLI. Results: During the study period, 236 patients received 252 HCTs followed by DLIs (median age at HCT 9.2 [range 0.2-25.1] years). An additional 743 patients received 806 HCTs with no subsequent DLI (median age at HCT 10.3 [range 0.1-27.2] years). Dosing most frequently used was 1 × 10 5, 1 × 10 6 and 1 × 10 7 CD3+ cells/kg, for haploidentical donor (HAPLO), matched unrelated donor (MUD) and matched sibling donor (MSD) recipients, respectively. Initial doses were occasionally lower based on risk of GVHD and were escalated as tolerated and based on response. After receiving DLI, 19.4% developed acute GVHD, and 9.2% developed chronic GVHD with no significant differences observed between MSD, MUD and HAPLO recipients. After adjusting for donor source, longer duration between HCT and DLI was associated with a significantly decreased risk of both acute (HR 0.97 P&amp;lt;0.001) and chronic GVHD (HR 0.98 P=0.04). A higher CD3+ cell dose was also significantly associated with an increased risk of acute GVHD (HR 1.03 P&amp;lt;0.001) but not chronic GVHD (HR 0.96 P=0.12). The overall response rate after DLI for treating MC or graft failure was 45.2% for MSD, 32.6% for MUD and 41.4% for HAPLO recipients. Logistic regression analysis revealed leukemia patients had a higher response rate after receiving a DLI for mixed chimerism from MSD (OR=19.5 P=0.002) and HAPLO (OR=9.03 P=0.042), but not from MUD HCT (OR=0.65 P= 0.57), compared to patients with non-malignant disorders. We also evaluated the response to DLI for frank disease relapse in 15 patients. The response rate after DLI for disease relapse was 50% (1/2) for MSD, 44.4% (4/9) for HAPLO, but 0% (0/4) for MUD recipients. Conclusions: In our analysis, the incidence of GVHD was not higher in pediatric and young adult HCT recipients who received DLI as compared to those who did not. Overall, the risk of developing acute and chronic GVHD after DLI was low and comparable between different donor sources. The risk of GVHD decreased further away from HCT and increased with higher DLI dose. The response rate for treating MC and disease relapse remains less than 50%. Thus, DLIs may be safely administered to pediatric and young adult patients after HCT. The therapeutic effect may be modulated by adjusting the dose, timing and frequency of DLI administration.

1160. Comparative Effectiveness of Pneumococcal Vaccination Strategies to Prevent Invasive Pneumococcal Disease: A Cohort Study at the Veterans Health Administration

Abstract Background The direct impact of a combined vaccination strategy pneumococcal 13-valent conjugate vaccine (PCV13) and 23-valent polysaccharide vaccine (PPSV23) for adults in preventing pneumococcal infections has yet to be thoroughly investigated. We aimed to estimate the comparative effectiveness of preventing the most severe form of pneumococcal infections, invasive pneumococcal disease (IPD), by comparing the risk of IPD stratified by the use of PCV13, PPSV23, and their combinations. Methods We included all patients who were 65 years or older, had established primary care within the Veterans Health Administration (VHA) between 2005 and 2021, and had not received any prior pneumococcal vaccination. We measured time-to-event from cohort enrollment to the onset of culture-proven IPD, considering death a competing risk event, and constructed a Cox regression model to estimate cause-specific hazards. PCV13 only, PPSV23 only, and PCV13 + PPSV23 were incorporated into the model as a time-dependent exposure variable, and we estimated hazard ratios (HRs) associated with them. We also considered a waning effect by incorporating time-by-vaccine interaction terms in the model to estimate continuously changing time-dependent hazards. Results 5,069,917 patients were enrolled in the cohort (male: 96.5%; median age: 72 [IQR: 65-79]). 1,217,349 patients received at least one dose of pneumococcal vaccine during the study period. The overall incidence rate of IPD was 9.1 per 100,000 patient-years. In crude analysis, all pneumococcal vaccines were associated with lower HRs for IPD, and there was an additive benefit of PCV13 and PPSV23 (Tables 1 and 2). In multivariate analysis, associations between PCV13 alone and PPSV23 alone were no longer statistically significant after adjusting for age and comorbidities, but PCV13 + PPSV23 was significantly associated with lower HR for IPD. No time-by-vaccine interaction term was significant. Conclusion In this large cohort study at the VHA, neither vaccine by itself appeared to lower the risk of IPD, but the combination of PCV13 and PPSV23 was associated with a significantly lower risk. In addition, we did not observe evidence of a waning effect to prevent IPD. Further studies are needed to evaluate the effect of newer pneumococcal vaccines. Disclosures Michihiko Goto, MD MSCI, Merck &amp; Co.: Grant/Research Support

Association of seafood consumption with cardiovascular disease among adults in Qingdao, China

Background and aimsThe relationship between seafood consumption and cardiovascular disease (CVD) is controversial, and studies have not considered competing risk events. Our study examined the association between a full range of seafood consumption and CVD incidence and mortality based on the Qingdao Diabetes Prevention Program. Methods and resultsWe followed up 5285 participants without CVD at baseline until December 31, 2021. CVD cases and deaths were identified through record linkage with the Qingdao CVD Surveillance System and the Qingdao Death Surveillance System, respectively. Information on seafood consumption was obtained using a food frequency questionnaire. We used the Cox proportional hazard model and the competing risk model to evaluate the association between all types of seafood consumption and CVD incidence and mortality. During a median follow-up of 11.4 years, 122 CVD cases and 75 deaths occurred. After adjustment for potential confounders, compared with nonconsumers, seafood consumption of 300–500 and > 500 g/week was associated with a lower risk of CVD incidence [hazards ratio and 95 % confidence interval (CI): 0.54 (0.29–0.99) and 0.49 (0.26–0.91), respectively]. However, seafood consumption of >500 g/week had a significantly lower risk of CVD mortality [subdistribution hazard ratio and 95 % CI: 0.40 (0.17–0.95)], but it was insignificant in other groups. ConclusionSeafood consumption of 300–500 g/week and >500 g/week was associated with a lower CVD incidence and mortality. Our findings provide evidence of the recommendations of the 2022 Dietary Guidelines for Chinese residents and may guide the promotion of strategies for CVD prevention.