Overall Survival and Prognostic Factors in Metastatic Triple-Negative Breast Cancer: A National Cancer Database Analysis

BACKGROUND

Metastatic triple-negative breast cancer (TNBC) is aggressive with poor median overall survival (OS) ranging from 8 to 13 months. There exists considerable heterogeneity in survival at the individual patient level. To better understand the survival heterogeneity and improve risk stratification, our study aims to identify the factors influencing survival, utilizing a large patient sample from the National Cancer Database (NCDB).

METHODS

Women diagnosed with metastatic TNBC from 2010 to 2020 in the NCDB were included. Demographic, clinicopathological, and treatment data and overall survival (OS) outcomes were collected. Kaplan-Meier curves were used to estimate OS. The log-rank test was used to identify OS differences between groups for each variable in the univariate analysis. For the multivariate analysis, the Cox proportional hazard model with backward elimination was used to identify factors affecting OS. Adjusted hazard ratios and 95% confidence intervals are presented.

RESULTS

In this sample, 2273 women had a median overall survival of 13.6 months. Factors associated with statistically significantly worse OS included older age, higher comorbidity scores, specific histologies, higher number of metastatic sites, presence of liver or other site metastases in those with only one metastatic site (excluding brain metastases), presence of cranial and extra-cranial metastases, lack of chemotherapy, lack of immunotherapy, lack of surgery to distant sites, lack of radiation to distant sites, and receipt of palliative treatment to alleviate symptoms. In the multivariate analysis, comorbidity score, histology, number of metastatic sites, immunotherapy, and chemotherapy had a statistically significant effect on OS.

CONCLUSIONS

Through NCDB analysis, we have identified prognostic factors for metastatic TNBC. These findings will help individualize prognostication at diagnosis, optimize treatment strategies, and facilitate patient stratification in future clinical trials.

Efficacy and safety of tranexamic acid in the treatment of gastric cancer complicated with upper gastrointestinal bleeding

OBJECTIVE

To investigate the efficacy and safety of tranexamic acid (TXA) in preventing upper gastrointestinal (GI) bleeding in patients with gastric cancer.

METHODS

The clinical data of patients with gastric cancer complicated with acute non-operative GI bleeding treated in the Fourth Hospital of Hebei Medical University from 2020 to 2022 were collected and retrospectively analyzed. The survival status of the patients was followed up by telephone. The dataset of 168 patients was divided into a control group (n=85) and a TXA group (n=83), at a 1:1 ratio. The patients in the control group were treated with esomeprazole, and the patients in the TXA group received additional TXA. The hemostatic effect, rebleeding rate, and mortality of patients were compared between the two groups. The Cox proportional hazard model was used to evaluate the overall survival of patients as well as the related risk factors.

RESULTS

The success rate of hemostasis and the normal blood coagulation rate in the TXA group were significantly higher than those in the control group (P=0.003 and P=0.016). The secondary bleeding rate, thrombus formation rate and digestive tract perforation rate in the TXA group were significantly lower than those in the control group (P=0.002, P=0.003 and P=0.035). The improvement of all indicators in the TXA group was better than that in the control group (all P<0.05). For patients with gastric cancer complicated with acute GI bleeding treated with TXA, the Cox proportional hazard model identified III~IV stage, time of TXA treatment, surgical treatment after hemorrhage, and an increase of D-dimer as independent risk factors for upper GI bleeding (all P<0.05).

CONCLUSION

TXA can be an effective treatment for patients with gastric cancer complicated by GI bleeding.

Hazard ratios of second primary malignancy after radioiodine for differentiated thyroid carcinoma: a large-cohort retrospective study

INTRODUCTION

The objective of this study is to evaluate the benefits of radioactive iodine (RAI) treatment and the risk of second primary malignancy (SPM) in RAI-treated patients.

MATERIAL AND METHODS

The cohort for this analysis consisted of individuals diagnosed with a first primary differentiated thyroid carcinoma (DTC), reported by the Surveillance, Epidemiology, and End Results (SEER) database in 1988-2016. Overall survival (OS) difference was estimated by Kaplan-Meier curves and log-rank test, and hazard ratios (HR) were obtained by Cox proportional-hazards model to evaluate the association between RAI and SPM.

RESULTS

Among 130,902 patients, 61,210 received RAI and 69,692 did not, and a total of 8604 patients developed SPM. We found that OS was significantly higher in patients who received RAI than in those who did not (p < 0.001). DTC survivors treated with RAI had increased risk of SPM in females (p = 0.043), particularly for SPM occurring in the ovary (p = 0.039) and leukaemia (p < 0.0001). The risk of developing SPM was higher in the RAI group than in the non-RAI group and the general population, and the incidence increased with age.

CONCLUSIONS

Increased risk of SPM occurs in female DTC survivors treated with RAI, which become more obvious with increasing age. Our research findings were beneficial to the formulation of RAI treatment strategies and the prediction of SPM for patients with thyroid cancer of different genders and different ages.

Understanding correlates of infant mortality in Ethiopia using 2019 Ethiopian mini demographic and health survey data

Infant mortality is one of the most sensitive and commonly used indicators of the social and economic development of a nation. Ethiopia is among the African countries with high infant mortality rates. This study aimed to understand and identify correlates of infant mortality in Ethiopia.

Methods

The data, used in this study, were drawn from 2019 Ethiopian Demographic and Health Survey data. The multivariable Cox proportional hazard analysis was done to identify the correlates of infant mortality.

Results

Infant mortality rate was high in the earlier age of months. Males, higher birth order and rural residences were at higher risk of dying before first birthday compared with respective reference groups whereas health facility deliveries, single births, rich wealth indices and older maternal age were at lower risk of dying before first birthday compared with respective reference groups.

Conclusion

The study found that age of mother, place of residence, wealth index, birth order, type of birth, child sex and place of delivery were statistically significant in affecting the survival of the infants. Thus, health facility deliveries should be encouraged and multiple birth infants should be given special care. Furthermore, younger mothers should better care of their babies to improve the survival of infants in Ethiopia.

The Optimal Machine Learning-Based Missing Data Imputation for the Cox Proportional Hazard Model

An adequate imputation of missing data would significantly preserve the statistical power and avoid erroneous conclusions. In the era of big data, machine learning is a great tool to infer the missing values. The root means square error (RMSE) and the proportion of falsely classified entries (PFC) are two standard statistics to evaluate imputation accuracy. However, the Cox proportional hazards model using various types requires deliberate study, and the validity under different missing mechanisms is unknown. In this research, we propose supervised and unsupervised imputations and examine four machine learning-based imputation strategies. We conducted a simulation study under various scenarios with several parameters, such as sample size, missing rate, and different missing mechanisms. The results revealed the type-I errors according to different imputation techniques in the survival data. The simulation results show that the non-parametric “missForest” based on the unsupervised imputation is the only robust method without inflated type-I errors under all missing mechanisms. In contrast, other methods are not valid to test when the missing pattern is informative. Statistical analysis, which is improperly conducted, with missing data may lead to erroneous conclusions. This research provides a clear guideline for a valid survival analysis using the Cox proportional hazard model with machine learning-based imputations.

An epidemiological approach to the analysis of cribra orbitalia as an indicator of health status and mortality in medieval and post-medieval London under a model of parasitic infection

OBJECTIVES

Many individuals living in medieval and post-medieval London suffered issues with sanitation, food insecurity, infectious disease, and widespread exposure to parasites from a multitude of sources, causing increased risk of death for many inhabitants. We examine this stressful environment and its relationship with various demographic and temporal dimensions, using cribra orbitalia (CO) as an indicator of stress, to model an increased risk of dying under the expectations of our proposed parasitic model of infection.

MATERIALS AND METHODS

We analyze the relationship between CO and mortality across seven medieval and post-medieval cemeteries from London by the covariates of sex, status, and age-at-death. A survival analysis (Cox regression) and a binomial logit estimated hazard and odds ratios of dying with CO across age-at-death, sex, status, and time-period within single statistical models. In addition, we provide new Bayesian age-at-death estimates for post-medieval samples.

RESULTS

The models show the rate of CO decreased over time and age-at-death, regardless of sex or status; post-medieval individuals were ~72% less likely to die with lesions than their medieval counterparts. Further, individuals with CO had ~1% decrease in risk of dying with CO per year of age.

DISCUSSION

These results suggest increased mortality risk for those with lesions indicative of anemia (CO), and selective mortality of younger individuals during the medieval period. Despite sex-specific nutritional and occupational hazards, and status-based access to resources, the prevalence of CO was similar across sex and status, which suggests living with parasitic infection that caused anemia was an everyday reality for medieval and post-medieval Londoners.

Influence of maltreatment, bullying, and neurocognitive impairment on recidivism in adolescents with conduct disorder: A 3-Year prospective study

We sought to clarify the influence of family developmental conditions (e.g., adverse childhood experiences: ACEs), peer relationships (e.g., bullying), and neurocognitive impairment on recidivism in adolescents with conduct disorder (CD). We interviewed 290 adolescents with CD (265 males, 25 females) who had been admitted to a juvenile justice assessment center for the first time about their offense(s), onset of delinquency, ACEs, and peer relationships. The participants also completed tests of neurocognitive activities and self-report symptom measures: the Symbol Digit Modalities Test (SDMT), the Memory Updating test for the assessment of attention, the Adolescent Dissociative Experiences Scale (A-DES) for the assessment of dissociation, and the Barratt Impulsiveness Scale 11th version (BIS-11) for the assessment of impulsivity. After the family court judgement and discharge, the participants were followed for 3 years. Readmissions to the center during the follow-up period were considered cases of "recidivism." During the follow-up, 102 (35.2%) participants were readmitted to the center. A survival analysis (Cox proportional hazard model) showed that lower age, bullying (as the perpetrator), and impaired cognitive function (e.g., a lower SDMT score) significantly increased recidivism, whereas bullying (as both victim and perpetrator) decreased recidivism. ACEs showed no clear effect on recidivism. Lower age and impaired cognitive function appear to predispose adolescent first-time offenders with CD to recidivism. However, the peer bullying analysis showed that being a perpetrator positively influenced recidivism, whereas being both a victim and a perpetrator negatively influenced recidivism. Further research considering the heterogeneity and comorbidity of CD could clarify bullying's influence on recidivism.

Association measures for clustered competing risks

We propose a semiparameteric model for multivariate clustered competing risks data when the cause-specific failure times and the occurrence of competing risk events among subjects within the same cluster are of interest. The cause-specific hazard functions are assumed to follow Cox proportional hazard models, and the associations between failure times given the same or different cause events and the associations between occurrences of competing risk events within the same cluster are investigated through copula models. A cross-odds ratio measure is explored under our proposed models. Two-stage estimation procedure is proposed in which the marginal models are estimated in the first stage, and the dependence parameters are estimated via an expectation-maximization algorithm in the second stage. The proposed estimators are shown to yield consistent and asymptotically normal under mild regularity conditions. Simulation studies are conducted to assess finite sample performance of the proposed method. The proposed technique is demonstrated through an application to a multicenter Bone Marrow transplantation dataset.

Mortality among dentists in Taiwan, 1985-2009

BACKGROUND/PURPOSE

Controversy exists in the literature regarding whether dentists with multiple occupational exposures suffer from premature mortality. A cohort mortality study was conducted to evaluate the survival outcome and determine if potential exposure to harmful agents leads to premature mortality among dentists.

METHODS

Using the Life Table Analysis System, we calculated standardized mortality ratios (SMRs) for a cohort of 11,700 dentists affiliated with the Taiwan Dental Association. These dentists were followed from 1985-2009. Reference rates were derived from cause-, gender-, and age-specific mortality rates of the general population of Taiwan and 18,664 Taiwanese internists, who were considered to be more socioeconomically proximal to dentists. A Cox proportional hazard model was also constructed to determine multiple risk factors associated with mortality.

RESULTS

Compared with the general population, dentists in Taiwan consistently demonstrated reduced from all-cause mortality. However, compared with internists, significant and excess mortality were observed in dentists for overall mortality (SMR=1.13; 95% confidence interval [CI]=1.00-1.26), drowning (SMR=6.62; 95% CI=2.15-15.45), and heart diseases (SMR=1.66; 95% CI=1.22-2.21). After adjusting for other risk factors, the Cox model showed an increased hazard ratio of 1.17 (95% CI=1.01-1.37) for dentists.

CONCLUSION

Taiwanese dentists demonstrated significant elevated SMRs for overall causes, drowning, and heart diseases. Careful precaution should be taken to reduce these trends. Future studies are also needed for in-depth exploration of the mechanisms regarding how professional stress and exposure contribute to the increased risk of mortality in Taiwanese dentists.

Decreased GFR estimated by MDRD or Cockcroft-Gault equation predicts incident CVD: the strong heart study

BACKGROUND

Kidney function, expressed as glomerular filtration rate (GFR), is commonly estimated from serum creatinine (Scr) and, when decreased, may serve as a nonclassical risk factor for incident cardiovascular disease (CVD). The ability of estimated GFR (eGFR) to predict CVD events during 5-10 years of follow-up is assessed using data from the Strong Heart Study (SHS), a large cohort with a high prevalence of diabetes.

METHODS

eGFRs were calculated with the abbreviated Modification of Diet in Renal Disease study (MDRD) and the Cockcroft-Gault (CG) equations. These estimates were compared in participants with normal and abnormal Scr. The association between eGFR and incident CVD was assessed.

RESULTS

More subjects were labeled as having low eGFR (<60 ml/min per 1.73 m2) by the MDRD or CG equation, than by Scr alone. When Scr was in the normal range, both equations labeled similar numbers of participants as having low eGFRs, although concordance between the equations was poor. However, when Scr was elevated, the MDRD equation labeled more subjects as having low eGFR. Persons with low eGFR had increased risk of CVD.

CONCLUSIONS

The MDRD and CG equations labeled more participants as having decreased GFR than did Scr alone. Decreased eGFR was predictive of CVD in this American Indian population with a high prevalence of obesity and type 2 diabetes mellitus.