Cox Logistic Regression Models Research Articles

Early presence of antiangiogenesis-related adverse events as a potential biomarker of antitumor efficacy in patients with metastatic gastric cancer treated with apatinib.

4052 Background: Reliable biomarkers of apatinib response in gastric cancer (GC) are lacking. We investigated the association between early presence of common adverse events (AEs) and clinical outcomes in metastatic GC patients. Methods: We conducted a retrospective cohort study using data on 269 apatinib-treated GC patients in two clinical trials. AEs were assessed at baseline until 28 days after the last dose of apatinib. Clinical outcomes were compared between patients with and without hypertension (HTN), proteinuria or hand and foot syndrome (HFS) in the first 4 weeks using Kaplan–Meier methods, Cox proportional hazard regression and logistic regression models. Landmark analyses were performed as sensitivity analyses. Predictive model and risk scores were analyzed to predict overall survival (OS). Results: Presence of AEs in the first 4 weeks was associated with prolonged median OS (169 vs. 103 days, log-rank p=0.0039; adjusted hazard ratio [HR] 0.64, 95% confidence interval [CI] 0.64-0.84, p=0.001), prolonged median progression-free survival (PFS, 87 vs. 62 days, log-rank p=0.0309; adjusted HR 0.69, 95%CI 0.53-0.91, p=0.007), and increased disease control rate (54.67% vs. 32.77%; adjusted odds ratio 2.67, p<0.001). Results remained significant in landmark analyses. An AE-based prediction model and subsequently derived scoring system showed high calibration and discrimination in predicting OS. Conclusions: Presence of HTN, proteinuria or HFS during the first cycle of apatinib treatment was a viable biomarker of antitumor efficacy in metastatic GC patients. [Table: see text]

Polypharmacy and potentially inappropriate medication use in older patients with aggressive non-Hodgkin lymphoma (NHL) leads to inferior survival and increased treatment-related toxicities.

10039 Background: Survival outcomes for older patients with aggressive NHL are disproportionally inferior to those of younger patients. While differences in tumor biology may play a role, older patients are often frail with comorbidities, polypharmacy, and use potentially inappropriate medications (PIM) such as anticholinergics and benzodiazepines. Methods: Using Cox proportional hazard and logistic regression models, we analyzed all aggressive NHL patients age ≥60 treated at our two affiliated hospitals from 2009-2014 to examine the association of polypharmacy and PIM use with progression-free survival (PFS), overall survival (OS), and treatment-related toxicities. Results: In this updated and final analysis, we included 171 patients with complete records from these two hospitals. They share similar demographic, clinical, and laboratory characteristics except for higher International Prognostic Index (IPI) in patients from one hospital. The median age was 70 years (range 65-77). At the time of diagnosis, 46% of patients used more than 4 medications (polypharmacy) and 47% used at least one PIM. Only 43% of patients received first-line chemotherapy of adequate relative dose intensity (>85% dosage), and 65% experienced ≥grade 3 toxicities. Polypharmacy and PIM use were associated with shortened PFS and OS by log-rank test. Most importantly, PIM use remained an independent predictor of PFS, OS, and ≥ grade 3 toxicities in multivariable analyses (Table). Conclusions: This is the first report of significantly adverse survival impacts of polypharmacy and PIM use in older patients with aggressive NHL, presumably from drug-drug interactions that increase toxicities and impair the delivery of adequate chemotherapy dosage. Our findings support the use of evidence-based geriatric principles to guide meticulous medication management to improve outcome disparity for these patients. [Table: see text]

The role of uric acid in chronic kidney disease patients.

Chronic kidney disease (CKD) is always associated with hyperuricaemia. However, the studies evaluating the clinical implications of hyperuricaemia have shown conflicting results in these patients. A retrospective observational study was conducted in 2408 stage 3-5 CKD patients. Instead of one baseline uric acid (UA) level, the averaged level of the two consecutive measurements for each participant was used as the predictor for the outcomes of the study, which included mortality, renal outcomes, and hospitalization risk. A multivariate Cox proportional hazards model and logistic regression model were performed to determine the independent risk factor. The mean UA level was 0.46 ± 0.106 mmol/L. Of the 2408 patients, there were 563 (23.3%) deaths, 143 (5.9%) cardiovascular deaths, 652 (27%) subjects commencing renal replacement therapy (RRT), 664 (27.5%) subjects with rapid renal progression, 1937 (58%) patients requiring hospitalization and 404 (16.7%) patients with CVD hospitalization during a mean follow-up of approximately 3.03 years. After multivariate adjustments, a 1-mg/dL increase in uric acid level was associated with a hazard ratio (HR) of 1.26 for RRT (P = 0.002), an odds ratio (OR) of 1.27 for rapid renal progression (P = 0.001), an HR of 1.19 for all-cause hospitalization (P < 0.001), and an HR of 1.12 for cardiovascular disease (CVD) hospitalization (P = 0.02), but not significantly with all-cause mortality and cardiovascular death at the end of follow-up. In stage 3-5 CKD patients, hyperuricaemia was associated with a higher risk of renal replacement therapy, rapid renal progression and hospitalization for all causes or CVD, but not with all-cause mortality or cardiovascular mortality.

Emerging Trends in Family History of Breast Cancer and Associated Risk

Abstract The impact of the rise in breast cancer incidence associated with mammography diffusion on the prevalence of breast cancer family history is unknown, and may attenuate risk associations between family history and breast cancer. Methods: The proportions of women 40–74 years reporting a first-degree family history of breast cancer were estimated in the Breast Cancer Surveillance Consortium (BCSC, N = 1,170,900) from 1996–2012 and Collaborative Breast Cancer Study (CBCS; cases N = 23,400; controls N = 25,460) from 1987–2007. Breast cancer (ductal carcinoma in situ and invasive) relative risk estimates associated with family history and 95% confidence intervals (CI) were calculated using multivariable Cox proportional hazard (BCSC) and logistic regression (CBCS) models. Results: The proportion of women reporting a family history increased from 11% in the 1980s to 16% in 2010–13. Family history was associated with a 60% increased risk of breast cancer in the BCSC (hazard ratio = 1.61, 95% CI = 1.55–1.66) and CBCS (odds ratio = 1.66, 95% CI = 1.58–1.74), with relative risks decreasing with age. Trends in relative risks were not evident over time or stage of disease at diagnosis, except among older women (60–74) where estimates were attenuated in more recent years (P-trend = 0.08 for both cohorts). Conclusion: The proportion of women with a first-degree family history of breast cancer increased over time and by age, nonetheless breast cancer risk associations with family history were constant over time for women 40–59. First-degree family history of breast cancer remains an important breast cancer risk factor, especially for younger women, despite its increasing prevalence in the mammography screening era

Persistence and Adherence with Mirabegron versus Antimuscarinic Agents in Patients with Overactive Bladder: A Retrospective Observational Study in UK Clinical Practice

BackgroundPersistence with antimuscarinic therapy in overactive bladder (OAB) is poor, but may be different for mirabegron, a β3-adrenoceptor agonist with a different adverse event profile. ObjectiveTo compare persistence and adherence with mirabegron versus tolterodine extended release (ER) and other antimuscarinics in routine clinical practice over a 12-mo period. Design, setting, and participantsRetrospective, longitudinal, observational study of anonymised data from the UK Clinical Practice Research Datalink GOLD database. Eligibility: age ≥18 yr, ≥1 prescription for target OAB drug (between May 1, 2013 and June 29, 2014), and 12-mo continuous enrolment before and after the index prescription date. InterventionsMirabegron, darifenacin, fesoterodine, flavoxate, oxybutynin ER or immediate-release (IR), propiverine, solifenacin, tolterodine ER or IR, and trospium chloride. Outcome measurements and statistical analysisThe primary endpoint was persistence (time to discontinuation). Secondary endpoints included 12-mo persistence rates and adherence (assessed using medication possession ratio, MPR). Cox proportional-hazards regression models and logistic regression models adjusted for potential confounding factors were used to compare cohorts. Analyses were repeated after 1:1 matching. Results and limitationsThe study population included 21996 eligible patients. In the unmatched analysis, the median time-to-discontinuation was significantly longer for mirabegron (169 d, interquartile range [IQR] 41–not reached) compared to tolterodine ER (56 d, IQR 28–254; adjusted hazard ratio [HR] 1.55, 95% confidence interval 1.41–1.71; p<0.0001) and other antimuscarinics (range 30–78 d; adjusted HR range 1.24–2.26, p<0.0001 for all comparisons). The 12-mo persistence rates and MPR were also significantly greater with mirabegron than with all the antimuscarinics. Limitations include the retrospective design, use of prescription records to estimate outcomes, and inability to capture reasons for discontinuation. ConclusionsPersistence and adherence were statistically significantly greater with mirabegron than with tolterodine ER and other antimuscarinics prescribed for OAB in the UK. Patient summaryThis study assessed persistence and adherence (or compliance) with medications prescribed for OAB in a large UK population. We found that patients prescribed mirabegron remained on treatment for longer and showed greater adherence than those prescribed traditional antimuscarinics.

Early-life and adult socioeconomic determinants of myocardial infarction incidence and fatality

Social inequalities in coronary heart disease mortality have roots in childhood conditions, but it is unknown whether they are associated both with the incidence of the disease and the following survival. We studied how several different early-life socioeconomic factors, together with later socioeconomic attainment, were associated with myocardial infarction (MI) incidence and fatality in Finland. The data was based on a register-based sample of households from a census in 1950 that also provided information on childhood circumstances. MI hospitalizations and mortality in 1988–2010 were studied in those who were up to 14 years of age at the time of the census and resident in Finland in 1987 (n = 94,501). Parental education, occupation, household crowding, home ownership, and family type were examined together with adulthood education and income. Hazard and odds ratios with 95% confidence intervals (CI) were calculated using Cox regression (incidence and long-term fatality) and logistic regression (short-term fatality) models. Lower parental education, occupational background and greater household crowding were associated with MI incidence. In models adjusted for adulthood variables, crowding increased the risk by 16% (95% CI 5–29%) in men and 25% (95% CI 3–50%) in women. Short-term survival was more favourable in sons of white-collar parents and daughters of owner-occupied households, but most aspects of childhood circumstances did not strongly influence long-term fatality risk. Socioeconomic attainment in adulthood accounted for a substantial part of the effects of childhood conditions, but the measured childhood factors explained little of the disparities by adulthood education and income. Moreover, income and education remained associated with MI incidence when adjusted for unobserved shared family factors in siblings. Though social and economic development in society seems to have mitigated the disease burden associated with poor childhood living conditions in Finland, low adult socioeconomic resources have remained a strong determinant of MI incidence and fatality.

Lymph node yield in node-negative patients predicts cancer specific survival following radical cystectomy for transitional cell carcinoma.

PurposeTo determine the oncological implications of increased nodal dissection in node-negative bladder cancer during radical cystectomy in a contemporary Australian series.Materials and MethodsWe performed a multicenter retrospective study, including more than 40 surgeons across 5 sites over a 10-year period. We identified 353 patients with primary bladder cancer undergoing radical cystectomy. Extent of lymphadenectomy was defined as follows; limited pelvic lymph node dissection (PLND) (perivesical, pelvic, and obturator), standard PLND (internal and external iliac) and extended PLND (common iliac). Multivariable cox proportional hazards and logistic regression models were used to determine LNY effect on cancer-specific survival.ResultsOver the study period, the extent of dissection and lymph node yield increased considerably. In node-negative patients, lymph node yield (LNY) conferred a significantly improved cancer-specific survival. Compared to cases where LNY of 1 to 5 nodes were taken, the hazard ratio (HR) for 6 to 15 nodes harvested was 0.78 (95% confidence interval [CI], 0.43–1.39) and for greater than 15 nodes the HR was 0.31 (95% CI, 0.17–0.57), adjusted for age, sex, T stage, margin status, and year of surgery. The predicted probability of cancer-specific death within 2 years of cystectomy was 16% (95% CI, 13%–19%) with 10 nodes harvested, falling to 5.5% (95% CI, 0%–12%) with 30 nodes taken. Increasing harvest in all PLND templates conferred a survival benefit.ConclusionsThe findings of the current study highlight the improved oncological outcomes with increased LNY, irrespective of the dissection template. Further prospective research is needed to aid LND data interpretation.

The Association of Work Characteristics With Ovarian Cancer Risk and Mortality

Ovarian cancer (OvCA) is a leading cause of cancer death for women. Depression and social isolation have been associated with a higher OvCA risk and poorer survival, but other forms of chronic psychosocial stress, including work-related characteristics, remain understudied. Women from three prospective cohorts (Nurses' Health Study: n = 31,754; Nurses' Health Study II: n = 74,260; Northern Sweden Health and Disease Study: nnested case-control study = 196) completed a job questionnaire, assessing demand and control at work, social support provided by coworkers and supervisor, and job security. Multivariate Cox and conditional logistic regression models estimated hazard ratios (Nurses' Health Study/Nurses' Health Study II) and odd ratios (Northern Sweden Health and Disease Study) of OvCA risk and mortality among cases. Random coefficient models were used for meta-analyses. There were 396 OvCA cases and 186 deaths during follow-up. Overall, job strain, strain chronicity, social support, and job security were not significantly associated with OvCA risk (e.g., pooled relative risk [RR]high demand/low control = 1.06, confidence interval [CI] = 0.72-1.55) or mortality (e.g., pooled RRhigh demand/low control = 1.08, CI = 0.64-1.82). When considered individually, compared with low levels, only moderate levels of demand were associated with a reduced OvCA risk (pooled RR = 0.66, CI = 0.49-0.90). Social support provided by the coworker or the supervisor did not moderate the association of job strain with either OvCA risk or overall mortality. We did not observe clear associations between work characteristics and OvCA incidence or mortality, but further research with diverse populations is warranted.

Long-Term Outcomes after Abdominal Wall Reconstruction with Acellular Dermal Matrix

Long-term outcomes data for hernia recurrence rates after abdominal wall reconstruction (AWR) with acellular dermal matrix (ADM) are lacking. The aim of this study was to assess the long-term durability of AWR using ADM. We studied patients who underwent AWR with ADM at a single center in 2005 to 2015 with a minimum follow-up of 36 months. Hernia recurrence was the primary end point and surgical site occurrence (SSO) was a secondary end point. The recurrence-free survival curves were estimated by Kaplan-Meier product limit method. Univariate and multivariable Cox proportional hazards regression models and logistic regression models were used to evaluate the associations of risk factors at surgery with subsequent risks for hernia recurrence and SSO, respectively. A total of 512 patients underwent AWR with ADM. After excluding those with follow-up less than 36 months, 191 patients were included, with a median follow-up of 52.9 months (range 36 to 104 months). Twenty-six of 191 patients had a hernia recurrence documented in the study. The cumulative recurrence rates were 11.5% at 3 years and 14.6% by 5 years. Factors significantly predictive of hernia recurrence developing included bridged repair, wound skin dehiscence, use of human cadaveric ADM, and coronary disease; component separation was protective. In a subset analysis excluding bridged repairs and human cadaveric ADM patients, cumulative hernia recurrence rates were 6.4% by 3 years and 8.3% by 5 years. The crude rate of SSO was 25.1% (48 of 191). Factors significantly predictive of the incidence of SSO included at least 1 comorbidity, BMI ≥30 kg/m2, and defect width >15 cm. Use of ADM for AWR was associated with 11.5% and 14.6% hernia recurrence rates at 3- and 5-years follow-up, respectively. Avoiding bridged repairs and human cadaveric ADM can improve long-term AWR outcomes using ADM.

Abdominal Wall Reconstruction with Concomitant Ostomy-Associated Hernia Repair: Outcomes and Propensity Score Analysis

The optimal strategy for abdominal wall reconstruction in the presence of a stomal-site hernia is unclear. We hypothesized that the rate of ventral hernia recurrence in patients undergoing a combined ventral hernia repair and stomal-site herniorraphy would not differ clinically from the ventral hernia recurrence rate in patients undergoing an isolated ventral hernia repair. We also hypothesized that bridged ventral hernia repairs result in worse outcomes compared with reinforced repairs, regardless of stomal hernia. We retrospectively reviewed prospectively collected data from consecutive abdominal wall reconstructions performed with acellular dermal matrix (ADM) at a single center between 2000 and 2015. We compared patients who underwent a ventral hernia repair alone (AWR) and those who underwent both a ventral hernia repair and ostomy-associated herniorraphy (AWR+O). We conducted a propensity score matched analysis to compare the outcomes between the 2 groups. Multivariable Cox proportional hazards and logistic regression models were used to study associations between potential predictive or protective reconstructive strategies and surgical outcomes. We included 499 patients (median follow-up 27.2 months; interquartile range [IQR] 12.4 to 46.6 months), 118 AWR+O and 381 AWR. After propensity score matching, 91 pairs were obtained. Ventral hernia recurrence was not statistically associated with ostomy-associated herniorraphy (adjusted hazard ratio [HR] 0.7; 95% CI 0.3 to 1.5; p= 0.34). However, the AWR+O group experienced a significantly higher percentage of surgical site occurrences (34.1%) than the AWR group (18.7%; adjusted odds ratio 2.3; 95% CI 1.4 to 3.7; p < 0.001). In the AWR group, there were significantly fewer ventral hernia recurrences when the repair was reinforced compared with bridged (5.3% vs 38.5%; p < 0.001). There was no statistically significant difference in ventral hernia recurrence between the AWR and AWR+O groups. Bridging was associated with an increased rate of hernia recurrence andshould be avoided if possible.

Prognosis of patients with BRCA1-associated ovarian carcinomas depends on TP53 accumulation status in tumor cells

ObjectiveTP53 mutation is the most frequent molecular event in BRCA1-associated ovarian carcinomas. TP53 status may be a confounding factor in the evaluation of clinical importance of other proteins. We aimed to evaluate the clinical significance of BRCA1 mutations with respect to the TP53 accumulation status in 159 high-grade ovarian carcinomas. MethodsStatistical analyses were done with the Kaplan-Meier method, log-rank test, the Cox's and logistic regression models for all patients, and in subgroups with and without TP53 accumulation (TP53+ and TP53−, respectively). ResultsForty of 159 ovarian carcinomas (25.2%) were diagnosed in patients with BRCA1 germline mutations; 102 tumors (64.2%) were TP53+ and 57 (37.8%) were TP53−. Among patients with TP53+ carcinomas, BRCA1 carriers had increased odds of recurrence compared with sporadic cases (HR 2.25, P=0.003; median disease-free survival time 7.7 vs. 18.4months, respectively). In the smaller TP53− subgroup, BRCA1 mutation reduced the risk of death by 46% (HR 0.54, P=0.099, median overall survival time 42.7 vs. 28.1months), but beyond the border of significance. When the TP53 status was not taken into account, BRCA1 mutations did not show any significance, however, there was a trend toward increased odds of complete remission for women with BRCA1 mutations compared to non-carriers (OR 2.47, P=0.064). Taxane-platinum therapy showed advantage over the platinum-cyclophosphamide one in the entire group of patients and in the TP53+ subgroup. ConclusionsOur results suggest that the TP53 accumulation status determines the prognosis of BRCA1 mutation carriers with high-grade ovarian carcinomas.

Contribution of polypharmacy and potentially inappropriate medication use to inferior survival in older patients with aggressive lymphoma.

129 Background: Survival outcomes for older patients with aggressive non-Hodgkin’s lymphoma (NHL) are disproportionally inferior to those of younger patients. While differences in tumor biology may play a role, older patients are often frail with comorbidities, polypharmacy, and use potentially inappropriate medications (PIM). Methods: Using Cox proportional hazard and logistic regression models, we retrospectively analyzed all aggressive NHL patients age 60 and older diagnosed and treated at our institution from 2009-2014 to examine the effect of polypharmacy and PIM use on progression-free survival (PFS), overall survival (OS), and treatment-related toxicities. Results: We included 141 patients with evaluable data after excluding patients with incomplete record. The median age was 71 years. At the time of diagnosis, 44% of patients used more than 4 medications and 47% used at least one PIM. During first-line treatment, only 43% of patients received chemotherapy of adequate relative dose intensity (&gt;85% scheduled dose), and 63% experienced grade 3 or greater toxicities. Age, International Prognostic Index, and PIM use correlated with each other. Number of medications (p = 0.005) and PIM use (p &lt; 0.001) were associated with shortened PFS by log-rank test, and PIM use remained a strong independent predictor of PFS in multivariable analysis (HR 1.84, p = 0.005). Number of medications (p = 0.003) and PIM use (p = 0.009) were also associated with shortened OS by log-rank test, although only albumin level predicted OS in multivariable analysis. Most importantly, PIM use was strongly associated with grade 3 or greater toxicities in multivariable analysis (OR 7.4, p = 0.001). Conclusions: We report here for the first time adverse impacts of polypharmacy and PIM use in older patients with aggressive lymphoma. We suggest that drug-drug interactions may significantly impair the delivery of adequate chemotherapy dosage and increase toxicities thus resulting in inferior survival outcome. Our findings support the use of evidence-based geriatric and palliative care principles to guide meticulous medication management to eliminate outcome disparity in older lymphoma patients.

Impact of Incidental Cardiac Radiation on Cardiopulmonary Toxicity and Survival for Locally Advanced Non-Small Cell Lung Cancer: Reanalysis of NRG Oncology/RTOG 0617 With Centrally Contoured Cardiac Structures

RTOG 0617 showed high dose RT had a greater risk of death than standard dose. Higher heart dose was associated with worse overall survival (OS) on multivariate analysis. However, the variability in submitted heart contours and no contouring of substructures limited the ability to accurately quantify the role of heart dose and ultimately set meaningful dose constraints. Five thoracic radiation oncologists collectively contoured the cardiac structures for each available case, guided by a common atlas. Anatomic structures defined were pericardium (P), ventricles (V), atria (A), and coronary space (CS). Volumes of each structure and overlap with PTV, distribution of V5, V10, V20, V30, V40, V45, V60, maximum dose (Dmax), and mean dose (MD) were analyzed. Volumes and doses were compared between centrally contoured P and submitted heart contours. Impact of volume and dose to each structure and combined structures on OS, local failure (LF), distant failure (DF), progression free survival (PFS), and treatment related pulmonary, esophageal, cardiovascular and esophageal toxicity were evaluated using Cox PH model and logistic regression model with multiplicity adjustments. Among 495 analyzable cases, 57 (11.5%) were not available for re-contouring and excluded. Re-contoured P volumes, PTV overlap with P, and dose volumes were larger than submitted heart contours. On univariate analysis all analyzed dose volumes and MD to P were associated with increased risk of death, P < 0.001. Similar relationships were found with V5-60 and MD to A and V and Dmax to the V, P < 0.01. On multivariate analysis, larger PTV volume and Gr3+ esophagitis remained associated with worse OS (P < 0.05); in addition, MD to P (HR = 1.019, P = .007) and V45 for A (HR = 1.007, P = 0.022), V (HR = 1.015, P = 0.0043), and CS (HR = 1.005, P = 0.0022) each had a detrimental impact on OS. On Univariate analysis MD to P and V5-45 were associated with all Gr3+ pulmonary toxicity (P < 0.05) and MD and V30-60 with Gr3+ pneumonitis (<0.05). On multivariate analysis MD to P (OR = 1.044, P = 0.0372) and stage IIIB disease (OR = 2.51, P = 0.0342) were associated with Grd3+ pneumonitis. Excessive RT doses to cardiac structures including pericardium MD, V45 to Atria, ventricles, and CS have a detrimental effect on overall survival. Pericardium MD was associated with severe treatment related pneumonitis.

A Moment Matching Approach for Generating Synthetic Data.

Synthetic data are becoming increasingly important mechanisms for sharing data among collaborators and with the public. Multiple methods for the generation of synthetic data have been proposed, but many have short comings with respect to maintaining the statistical properties of the original data. We propose a new method for fully synthetic data generation that leverages linear and integer mathematical programming models in order to match the moments of the original data in the synthetic data. This method has no inherent disclosure risk and does not require parametric or distributional assumptions. We demonstrate this methodology using the Framingham Heart Study. Existing synthetic data methods that use chained equations were compared with our approach. We fit Cox proportional hazards, logistic regression, and nonparametric models to synthetic data and compared with models fitted to the original data. True coverage, the proportion of synthetic data parameter confidence intervals that include the original data's parameter estimate, was 100% for parametric models when up to four moments were matched, and consistently outperformed the chained equations approach. The area under the curve and accuracy of the nonparametric models trained on synthetic data marginally differed when tested on the full original data. Models were also trained on synthetic data and a partition of original data and were tested on a held-out portion of original data. Fourth-order moment matched synthetic data outperformed others with respect to fitted parametric models but did not always outperform other methods with fitted nonparametric models. No single synthetic data method consistently outperformed others when assessing the performance of nonparametric models. The performance of fourth-order moment matched synthetic data in fitting parametric models suggests its use in these cases. Our empirical results also suggest that the performance of synthetic data generation techniques, including the moment matching approach, is less stable for use with nonparametric models. The benefits of the moment matching approach should be weighed against additional computational costs. In summary, our results demonstrate that the introduced moment matching approach may be considered as an alternative to existing synthetic data generation methods.

The Added Prognostic Value of Metabolic Tumor Size on FDG‐PET at First Suspected Recurrence of Glioblastoma Multiforme

Patients with glioblastoma multiforme (GBM) face a dismal prognosis, with an average survival of 6-7 months after recurrence. There remains no consensus for managing these patients due to the heterogeneity of these tumors. Imaging may affect treatment decisions by helping to stratify patient prognosis. The purpose of this analysis was to evaluate the added utility of 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) over magnetic resonance (MR) imaging metrics in predicting survival. Forty-four consecutive patients who underwent FDG-PET for first recurrence of GBM were included in this analysis. Tumor sizes, using cross products, and volumes on FDG-PET and MRI, maximum standardized uptake value (SUV), minimum apparent diffusion coefficient (ADC) value, presence of satellite lesions, presence of multifocal lesions, and presence of bilateral tumor were considered as prognostic variables. Survival was assessed using Cox hazard and logistic regression models based on the time interval between the PET scan and the patient's date of death. Tumor volumes on FDG-PET (P = .046), tumor cross products on FDG-PET (P = .017), and tumor cross products on MRI (P = .031) were significant prognostic variables, adjusting for the extent of the initial resection. Enhancing tumor volume, tumor cross product on a T2-weighted MRI sequence, maximum SUV on FDG-PET, minimum ADC value, presence of satellites, multifocality, and bilaterality were not prognostic (P > .5). Prognostic accuracy of predicting short survival increased from 58% with tumor cross product on MRI alone to 74% after including tumor cross product from PET. Tumor size on FDG-PET adds prognostic information to enhancing tumor size on MRI at first suspected recurrence of GBM.

High-Sensitive C-Reactive Protein Predicts Recurrent Stroke and Poor Functional Outcome: Subanalysis of the Clopidogrel in High-Risk Patients With Acute Nondisabling Cerebrovascular Events Trial.

Minor stroke and transient ischemic attack are common disorders with high rate of subsequent disabling stroke. We aim to investigate the role of high-sensitive C-reactive protein (hsCRP) in predicting recurrent stroke and poor functional outcome. In the Clopidogrel in High-Risk Patients With Acute Nondisabling Cerebrovascular Events (CHANCE) trial, 3044 (59%) consecutive patients from 73 (64%) prespecified centers had hsCRP levels measured. The primary outcome was any stroke within 90 days. The secondary outcome included combined vascular events and dependence or death defined as modified Rankin Scale score of 2 to 6 at 90 days and a new vascular event during 1-year follow-up. The associations of hsCRP with recurrent stroke and functional outcome were analyzed by using Cox proportional hazards and logistic regression models. Elevated hsCRP (>3.0 mg/L) was observed in 32% of the study population. Patients with hsCRP >3 mg/L had an increased risk of recurrent stroke (adjusted hazard ratio, 1.46; 95% confidence interval, 1.08-1.98; P=0.039), ischemic stroke and combined vascular events, and poor functional outcome (adjusted odds ratio, 1.68; 95% confidence interval, 1.22-2.32; P=0.002) compared with those with hsCRP <1 mg/L within 90-day follow-up period. High hsCRP levels also independently predicted recurrent stroke during 1-year follow-up. There was no interaction of hsCRP levels with randomized antiplatelet therapy. High hsCRP levels predict recurrent stroke and poor functional outcome in acute patients with minor stroke or transient ischemic attack. URL: http://www.clinicaltrials.gov. Unique identifier: NCT00979589.

Prenatal maternal distress affects atopic dermatitis in offspring mediated by oxidative stress

Recent evidence suggests that prenatal maternal distress increases the risk of allergic diseases in offspring. However, the effect of prenatal maternal depression and anxiety on atopic dermatitis (AD) risk remains poorly understood. We investigated whether prenatal maternal distress is associated with AD risk in offspring and whether the mechanism is mediated by reactive oxygen species. Two general population-based birth cohorts formed the study. One cohort (Cohort for Childhood Origin of Asthma and Allergic Diseases [COCOA]) consisted of 973 mother-baby dyads, and the other (Panel Study on Korean Children [PSKC]) consisted of 1531 mother-baby dyads. The association between prenatal distress and AD was assessed by using Cox proportional hazards and logistic regression models. In COCOA placental 11β-hydroxysteroid dehydrogenase type 2 and glutathione levels and serum IgE levels in 1-year-old children were measured. In COCOA and PSKC AD occurred in 30.6% (lifetime prevalence) and 11.6% (1year prevalence) of offspring, respectively. Prenatal maternal distress increased the risk of AD in offspring, both in COCOA (hazard ratio for depression, 1.31 [95% CI, 1.02-1.69]; hazard ratio for anxiety, 1.41 [95% CI, 1.06-1.89]) and PSKC (odds ratio for distress, 1.85 [95% CI, 1.06-3.25]). In COCOA both prenatal maternal depression and anxiety scores were positively related to the predicted probability of AD (P<.001 in both). Prenatal distress decreased placental glutathione to glutathione disulfideratios (P=.037) and, especially in those who later had AD, decreased placental 11β-hydroxysteroid dehydrogenase type 2 levels (P=.010) and increased IgE levels at 1year of age (P=.005). Prenatal maternal depression and anxiety promote risk of AD in offspring. Maternal distress increases the predicted probability of AD. The mechanism might involve chronic stress, abnormal steroid levels, and reactive oxygen species.

Sex and Ethnic Differences in Outcomes of Acute Coronary Syndrome and Stable Angina Patients With Obstructive Coronary Artery Disease.

The joint contribution of sex, ethnicity, and initial clinical presentation to the long-term outcomes of patients undergoing coronary angiography for acute coronary syndrome (ACS) or stable angina, in whom there is angiographic evidence for obstructive coronary artery disease, remains unknown. We conducted a population-based cohort study on 49 556 adult ACS or stable angina patients with angiographic evidence of obstructive coronary artery disease (≥ 50% stenosis) in British Columbia. The 2-year composite outcome was all-cause death and hospital readmissions for myocardial infarction, heart failure, cerebrovascular accident, or angina after the index angiography. Sex and ethnic differences in the composite outcome were examined by clinical presentation using the Cox proportional-hazards and logistic regression models. Overall, 25.6% were women, 9.5% were South Asians, 3.0% were Chinese, and 65.9% presented with ACS. Regardless of ethnicity, women were more likely than men to have adverse outcomes, but the magnitude of the sex difference was greater in the ACS patients (P(interaction) for sex and clinical presentation=0.03). Angina readmission accounted for 45% of the composite outcome and was the main component for all groups with the exception of Chinese women with ACS. Furthermore, women were more likely than men to be readmitted for angina (odds ratio [95% confidence interval], 1.13 [1.04-1.22]). Higher rates of adverse events among women with obstructive coronary artery disease, regardless of ethnicity, as well as high rates of angina readmission, highlight the need for more targeted interventions to reduce the burden of angina because this presentation is clearly not benign.

Genetic loci for serum lipid fractions and intracerebral hemorrhage

BackgroundSerum total cholesterol and its fractions are inversely associated with intracerebral hemorrhages (ICH) and their potential subclinical precursor, cerebral microbleeds. To ascertain whether there is a genetic basis for this inverse association, we studied established genetic loci for serum total, LDL, and HDL cholesterol, and triglycerides in their association with ICH and microbleeds. MethodsData on 161 genetic variants for serum lipids was collected in 9011 stroke-free participants (mean age 65.8, SD 10.2; 57.9% women) of the population-based Rotterdam Study. Participants were followed from baseline (1997–2005) up to 2013 for the occurrence of ICH. A subset of 4179 participants underwent brain MRI for microbleed assessment between 2005 and 2011. We computed genetic risk scores (GRS) for the joint effect of lipid variants. Cox proportional hazards and logistic regression models were used to investigate the association of GRS of lipid fractions with ICH and microbleeds. ResultsAfter a mean follow-up of 8.7 (SD 4.1) years, 67 (0.7%) participants suffered an ICH. Microbleed prevalence was 19.6%. Higher genetic load for high serum total and LDL cholesterol was associated with an increased risk of ICH. Higher genetic load for high serum LDL cholesterol was borderline associated with a higher prevalence of multiple lobar microbleeds. ConclusionsGenetic susceptibility for high serum total and LDL cholesterol is positively associated with incident ICH and borderline associated with multiple lobar microbleeds. We did not find a genetic basis for the previously reported inverse association between serum lipid levels and ICH.

Weekly versus every-three-weeks platinum-based chemoradiation regimens for head and neck cancer.

BackgroundThe majority of chemoradiation (CRT) trials for locally advanced head and neck squamous cell carcinoma (HNSCC) have relied on platinum-based chemotherapy regimens administered every-3-weeks. However, given the increased utilization of weekly platinum regimens, it remains unclear how different chemotherapy schedules compare regarding efficacy and toxicity.MethodsWe retrospectively identified 212 patients with HNSCC who were treated at a single academic medical center with concurrent platinum-based CRT given weekly (N = 68) or every-three-weeks (N = 144). JMP version 10 (SAS Institute) was used for statistical analysis. Discrete variables were compared with the chi-square test and differences in the medians were assessed using the Wilcoxon test. Survival curves were constructed using the Kaplan-Meier method and significance was assessed using the log rank test. For univariate analysis and multivariate analysis, we used Cox proportional hazard or logistic regression models to compare differences in survival or differences in categorical variables, respectively.ResultsPatients receiving weekly platinum regimens were more likely to be older (median age 61.4 vs. 55.5 y; P < .001), have high or very high Charlson comorbidity index (45.6% vs. 27.8%; P = .01), and receive carboplatin-based chemotherapy (6.3% vs. 76.5%; P < .001). Weekly and every-3-week platinum regimens had similar locoregional control (HR 1.10; 95% CI 0.63–1.88; P = .72), progression-free survival (HR 1.13; 95% CI 0.75–1.69; P = .55), and overall survival (HR 1.11; 95% CI 0.64–1.86; P = .71). Every-3-weeks platinum regimens were associated with increased days of hospitalization (median: 3 days vs. 0 days; P = .03) and acute kidney injury (AKI) during radiotherapy (50.0% vs. 22.1%; P < .001). On multivariate analysis, AKI was significantly associated with every-3-weeks regimens (OR: 24.38; 95% CI 3.00–198.03; P = .003) and high comorbidity scores (OR: 2.74; 95% CI 2.15–5.99; P = .01).ConclusionsOur results suggest that every-3-weeks and weekly platinum-containing CRT regimens have similar disease control but weekly platinum regimens are associated with less acute toxicity.Electronic supplementary materialThe online version of this article (doi:10.1186/s40463-016-0175-x) contains supplementary material, which is available to authorized users.