Autologous transplant vs chimeric antigen receptor T-cell therapy for relapsed DLBCL in partial remission.

The relative efficacy of autologous hematopoietic cell transplant (auto-HCT) vs chimeric antigen receptor T-cell (CAR-T) therapy in patients with diffuse large B-cell lymphoma (DLBCL) who achieve a partial remission (PR) after salvage chemotherapy is not known. Using the Center for International Blood & Marrow Transplant Research registry database, we identified adult patients with DLBCL who received either an auto-HCT (2013-2019) or CAR-T treatment with axicabtagene ciloleucel (2018-2019) while in a PR by computed tomography or positron emission tomography scan. We compared the clinical outcomes between the 2 cohorts using univariable and multivariable regression models after adjustment for relevant baseline and clinical factors. In the univariable analysis, the 2-year progression-free survival (52% vs 42%; P = .1) and the rate of 100-day nonrelapse mortality (4% vs 2%; P = .3) were not different between the 2 cohorts, but consolidation with auto-HCT was associated with a lower rate of relapse/progression (40% vs 53%; P = .05) and a superior overall survival (OS) (69% vs 47%; P = .004) at 2 years. In the multivariable regression analysis, treatment with auto-HCT was associated with a significantly lower risk of relapse/progression rate (hazard ratio = 1.49; P = .01) and a superior OS (hazard ratio = 1.63; P = .008). In patients with DLBCL in a PR after salvage therapy, treatment with auto-HCT was associated with a lower incidence of relapse and a superior OS compared with CAR-T. These data support the role of auto-HCT as the standard of care in transplant-eligible patients with relapsed DLBCL in PR after salvage therapy.

Efficient estimation for left-truncated competing risks regression for case-cohort studies.

The case-cohort study design provides a cost-effective study design for a large cohort study with competing risk outcomes. The proportional subdistribution hazards model is widely used to estimate direct covariate effects on the cumulative incidence function for competing risk data. In biomedical studies, left truncation often occurs and brings extra challenges to the analysis. Existing inverse probability weighting methods for case-cohort studies with competing risk data not only have not addressed left truncation, but also are inefficient in regression parameter estimation for fully observed covariates. We propose an augmented inverse probability-weighted estimating equation for left-truncated competing risk data to address these limitations of the current literature. We further propose a more efficient estimator when extra information from the other causes is available. The proposed estimators are consistent and asymptotically normally distributed. Simulation studies show that the proposed estimator is unbiased and leads to estimation efficiency gain in the regression parameter estimation. We analyze the Atherosclerosis Risk in Communities study data using the proposed methods.

A validated pediatric disease risk index for allogeneic hematopoietic cell transplantation.

A disease risk index (DRI) that was developed for adults with hematologic malignancy who were undergoing hematopoietic cell transplantation is also being used to stratify children and adolescents by disease risk. Therefore, to develop and validate a DRI that can be used to stratify those with AML and ALL by their disease risk, we analyzed 2569 patients aged <18 years with acute myeloid (AML; n = 1224) or lymphoblastic (ALL; n = 1345) leukemia who underwent hematopoietic cell transplantation. Training and validation subsets for each disease were generated randomly with 1:1 assignment to the subsets, and separate prognostic models were derived for each disease. For AML, 4 risk groups were identified based on age, cytogenetic risk, and disease status, including minimal residual disease status at transplantation. The 5-year leukemia-free survival for low (0 points), intermediate (2, 3, 5), high (7, 8), and very high (>8) risk groups was 78%, 53%, 40%, and 25%, respectively (P < .0001). For ALL, 3 risk groups were identified based on age and disease status, including minimal residual disease status at transplantation. The 5-year leukemia-free survival for low (0 points), intermediate (2-4), and high (≥5) risk groups was 68%, 51%, and 33%, respectively (P < .0001). We confirmed that the risk groups could be applied to overall survival, with 5-year survival ranging from 80% to 33% and 73% to 42% for AML and ALL, respectively (P < .0001). This validated pediatric DRI, which includes age and residual disease status, can be used to facilitate prognostication and stratification of children with AML and ALL for allogeneic transplantation.

Sample size and power determination for multiparameter evaluation in nonlinear regression models with potential stratification.

Sample size and power determination are crucial design considerations for biomedical studies intending to formally test the effects of key variables on an outcome. Other known prognostic factors may exist, necessitating the use of techniques for covariate adjustment when conducting this evaluation. Moreover, the main interest often includes assessing the impact of more than one variable on an outcome, such as multiple treatments or risk factors. Regression models are frequently employed for these purposes, formalizing this assessment as a test of multiple regression parameters. But, the presence of multiple variables of primary interest and correlation between covariates can complicate sample size/power calculation. Given the paucity of available sample size formulas for this context, these calculations are often performed via simulation, which can be both time-consuming as well as demanding extensive probability modeling. We propose a simpler, general approach to sample size and power determination that may be applied when testing multiple parameters in commonly used regression models, including generalized linear models as well as ordinary and stratified versions of the Cox and Fine-Gray models. Through both rigorous simulations and theoretical derivations, we demonstrate the formulas' accuracy in producing sample sizes that will meet the type I error rate and power specifications of the study design.

Does recipient body mass index inform donor selection for allogeneic haematopoietic cell transplantation?

It is not known whether obesity has a differential effect on allogeneic haematopoietic cell transplantation outcomes with alternative donor types. We report the results of a retrospective registry study examining the effect of obesity [body mass index (BMI) > 30] on outcomes with alternative donors (haploidentical related donor with two or more mismatches and receiving post-transplant cyclophosphamide [haplo] and cord blood (CBU)] versus matched unrelated donor (MUD). Adult patients receiving haematopoietic cell transplantation for haematologic malignancy (2013-2017) (N = 16 182) using MUD (n = 11 801), haplo (n = 2894) and CBU (n = 1487) were included. The primary outcome was non-relapse mortality (NRM). The analysis demonstrated a significant, non-linear interaction between pretransplant BMI and the three donor groups for NRM: NRM risk was significantly higher with CBU compared to haplo at BMI 25-30 [hazard ratio (HR) 1.66-1.71, p < 0.05] and MUD transplants at a BMI of 25-45 (HR, 1.61-3.47, p < 0.05). The results demonstrated that NRM and survival outcomes are worse in overweight and obese transplant recipients (BMI ≥ 25) with one alternative donor type over MUD, although obesity does not appear to confer a uniform differential mortality risk with one donor type over the other. BMI may serve as a criterion for selecting a donor among the three (MUD, haplo and CBU) options, if matched sibling donor is not available.

Marginal semiparametric transformation models for clustered multivariate competing risks data.

Multivariate survival models are often used in studying multiple outcomes for right-censored data. However, the outcomes of interest often have competing risks, where standard multivariate survival models may lead to invalid inferences. For example, patients who had stem cell transplantation may experience multiple types of infections after transplant while reconstituting their immune system, where death without experiencing infections is a competing risk for infections. Such competing risks data often suffer from cluster effects due to a matched pair design or correlation within study centers. The cumulative incidence function (CIF) is widely used to summarize competing risks outcomes. Thus, it is often of interest to study direct covariate effects on the CIF. Most literature on clustered competing risks data analyses is limited to the univariate proportional subdistribution hazards model with inverse probability censoring weighting which requires correctly specifying the censoring distribution. We propose a marginal semiparametric transformation model for multivariate competing risks outcomes. The proposed model does not require modeling the censoring distribution, accommodates nonproportional subdistribution hazards structure, and provides a platform for joint inference of all causes and outcomes.

A calibrated Bayesian method for the stratified proportional hazards model with missing covariates.

Missing covariates are commonly encountered when evaluating covariate effects on survival outcomes. Excluding missing data from the analysis may lead to biased parameter estimation and a misleading conclusion. The inverse probability weighting method is widely used to handle missing covariates. However, obtaining asymptotic variance in frequentist inference is complicated because it involves estimating parameters for propensity scores. In this paper, we propose a new approach based on an approximate Bayesian method without using Taylor expansion to handle missing covariates for survival data. We consider a stratified proportional hazards model so that it can be used for the non-proportional hazards structure. Two cases for missing pattern are studied: a single missing pattern and multiple missing patterns. The proposed estimators are shown to be consistent and asymptotically normal, which matches the frequentist asymptotic properties. Simulation studies show that our proposed estimators are asymptotically unbiased and the credible region obtained from posterior distribution is close to the frequentist confidence interval. The algorithm is straightforward and computationally efficient. We apply the proposed method to a stem cell transplantation data set.

Competing risks regression models with covariates-adjusted censoring weight under the generalized case-cohort design.

A generalized case-cohort design has been used when measuring exposures is expensive and events are not rare in the full cohort. This design collects expensive exposure information from a (stratified) randomly selected subset from the full cohort, called the subcohort, and a fraction of cases outside the subcohort. For the full cohort study with competing risks, He et al. (Scand J Stat 43:103-122, 2016) studied the non-stratified proportional subdistribution hazards model with covariate-dependent censoring to directly evaluate covariate effects on the cumulative incidence function. In this paper, we propose a stratified proportional subdistribution hazards model with covariate-adjusted censoring weights for competing risks data under the generalized case-cohort design. We consider a general class of weight functions to account for the generalized case-cohort design. Then, we derive the optimal weight function which minimizes the asymptotic variance of parameter estimates within the general class of weight functions. The proposed estimator is shown to be consistent and asymptotically normally distributed. The simulation studies show (i) the proposed estimator with covariate-adjusted weight is unbiased when the censoring distribution depends on covariates; and (ii) the proposed estimator with the optimal weight function gains parameter estimation efficiency. We apply the proposed method to stem cell transplantation and diabetes data sets.

Optimal treatment regimes for competing risk data using doubly robust outcome weighted learning with bi-level variable selection.

The goal of the optimal treatment regime is maximizing treatment benefits via personalized treatment assignments based on the observed patient and treatment characteristics. Parametric regression-based outcome learning approaches require exploring complex interplay between the outcome and treatment assignments adjusting for the patient and treatment covariates, yet correctly specifying such relationships is challenging. Thus, a robust method against misspecified models is desirable in practice. Parsimonious models are also desired to pursue a concise interpretation and to avoid including spurious predictors of the outcome or treatment benefits. These issues have not been comprehensively addressed in the presence of competing risks. Recognizing that competing risks and group variables are frequently present, we propose a doubly robust estimation with adaptive L 1 penalties to select important variables at both group and within-group levels for competing risks data. The proposed method is applied to hematopoietic cell transplantation data to personalize the graft source choice for treatment-related mortality (TRM). While the existing medical literature attempts to find a uniform solution ignoring the heterogeneity of the graft source effects on TRM, the analysis results show the effect of the graft source on TRM could be different depending on the patient-specific characteristics.

Pacemaker detected active minutes are superior to pedometer-based step counts in measuring the response to physical activity counseling in sedentary older adults.

BACKGROUND: In patients with permanent pacemakers (PPM), physical activity (PA) can be monitored using embedded accelerometers to measure pacemaker detected active hours (PDAH), a strong predictor of mortality. We examined the impact of a PA Counseling (PAC) intervention on increasing activity as measured by PDAH and daily step counts. METHODS: Thirteen patients (average age 80 ± 6 years, 84.6% women) with implanted Medtronic PPMs with a ≤ 2 PDAH daily average were included in this study. Patients were randomized to Usual Care (UC, N = 6) or a Physical Activity Counseling Intervention (PACI, N = 7) groups. Step count and PDAH data were obtained at baseline, following a 12-week intervention, then 12 weeks after intervention completion. Data were analyzed using independent t-tests, Pearson's r, chi-square, and general linear models for repeated measures. RESULTS: PDAH significantly differed by time point for all subject combined (P = 0.01) but not by study group. Subjects with baseline gait speeds of > 0.8 m/sec were responsible for the increases in PDAH observed. Step counts did not differ over time in the entire cohort or by study group. Step count and PDAH significantly correlated at baseline (r = 0.60, P = 0.03). This correlation disappeared by week 12. CONCLUSION(S): PDAH can be used to monitor PA and PA interventions and may be superior to hip-worn pedometers in detecting activity. A significant increase in PA, regardless of treatment group, suggests that patient awareness of the ability to monitor PA through a PPM increases PA in these patients, particularly in patients with gait speeds of < 0.8 m/sec. TRIAL REGISTRATION: ClincalTrials.gov NCT03052829. Date of Registration: 2/14/2017.

Outcomes after Second Hematopoietic Cell Transplantation in Children and Young Adults with Relapsed Acute Leukemia.

Children with acute leukemia who relapse after hematopoietic cell transplantation (HCT) have few therapeutic options. We studied 251 children and young adults with acute myelogenous or lymphoblastic leukemia who underwent a second HCT for relapse after their first HCT. The median age at second HCT was 11 years, and the median interval between first and second HCT was 17 months. Most of the patients (n = 187; 75%) were in remission, received a myeloablative conditioning regimen (n = 157; 63%), and underwent unrelated donor HCT (n = 230; 92%). The 2-year probability of leukemia-free survival (LFS) was 33% after transplantation in patients in remission, compared with 19% after transplantation in patients not in remission (P = .02). The corresponding 8-year probabilities were 24% and 10% (P = .003). A higher rate of relapse contributed to the difference in LFS. The 2-year probability of relapse after transplantation was 42% in patients in remission and 56% in those in relapse (P = .05). The corresponding 8-year probabilities were 49% and 64% (P = .04). These data extend the findings of others showing that patients with a low disease burden are more likely to benefit from a second transplantation. Late relapse led to a 10% decrement in LFS beyond the second year after second HCT. This differs from first HCT, in which most relapses occur within 2 years after HCT.

Transcriptomic analysis reveals inflammatory and metabolic pathways that are regulated by renal perfusion pressure in the outer medulla of Dahl-S rats.

Studies exploring the development of hypertension have traditionally been unable to distinguish which of the observed changes are underlying causes from those that are a consequence of elevated blood pressure. In this study, a custom-designed servo-control system was utilized to precisely control renal perfusion pressure to the left kidney continuously during the development of hypertension in Dahl salt-sensitive rats. In this way, we maintained the left kidney at control blood pressure while the right kidney was exposed to hypertensive pressures. As each kidney was exposed to the same circulating factors, differences between them represent changes induced by pressure alone. RNA sequencing analysis identified 1,613 differently expressed genes affected by renal perfusion pressure. Three pathway analysis methods were applied, one a novel approach incorporating arterial pressure as an input variable allowing a more direct connection between the expression of genes and pressure. The statistical analysis proposed several novel pathways by which pressure affects renal physiology. We confirmed the effects of pressure on p-Jnk regulation, in which the hypertensive medullas show increased p-Jnk/Jnk ratios relative to the left (0.79 ± 0.11 vs. 0.53 ± 0.10, P < 0.01, n = 8). We also confirmed pathway predictions of mitochondrial function, in which the respiratory control ratio of hypertensive vs. control mitochondria are significantly reduced (7.9 ± 1.2 vs. 10.4 ± 1.8, P < 0.01, n = 6) and metabolomic profile, in which 14 metabolites differed significantly between hypertensive and control medullas ( P < 0.05, n = 5). These findings demonstrate that subtle differences in the transcriptome can be used to predict functional changes of the kidney as a consequence of pressure elevation.

Fludarabine and Busulfan versus Fludarabine, Cyclophosphamide, and Rituximab as Reduced-Intensity Conditioning for Allogeneic Transplantation in Follicular Lymphoma.

Large, multicenter studies comparing commonly used reduced-intensity conditioning (RIC) approaches in follicular lymphoma (FL) have not been performed. Using the Center for International Blood and Marrow Transplant Research database, we report the outcomes of the 2 most commonly used RIC approaches, fludarabine and busulfan (Flu/Bu) versus fludarabine, cyclophosphamide, and rituximab (FCR) in FL patients. We evaluated 200 FL patients undergoing allogeneic hematopoietic cell transplantation (allo-HCT) who received RIC with either Flu/Bu (n = 98) or FCR (n = 102) during 2008 to 2014. All patients received peripheral blood grafts, and graft-versus-host disease (GVHD) prophylaxis was limited to calcineurin inhibitor-based approaches. Median follow-up of survivors in the Flu/Bu and FCR groups was 48 months and 46 months, respectively. On univariate analysis in the Flu/Bu and FCR groups, the 3-year rates of nonrelapse mortality (11% versus 11%, P = .94), relapse/progression (18% versus 15%, P = .54), progression-free survival (PFS) (71% versus 74%, P = .65), and overall survival (OS) (73% versus 81%, P = .18) were not significantly different. On multivariate analysis no difference was seen between the FCR and Flu/Bu cohorts in terms of grades II to IV (relative risk [RR], 1.06; 95% confidence interval [CI], .59 to 1.93; P = .84) or grades III to IV (RR, 1.18; 95% CI, .47 to 2.99; P = .72) acute GVHD, nonrelapse mortality (RR, .83; 95% CI, .38 to 1.82; P = .64), relapse/progression (RR, .99; 95% CI, .49 to 1.98; P = .97), PFS (RR, .92; 95% CI, .55 to 1.54; P = .76), or OS (RR, .70; 95% CI, .40 to 1.23; P = .21) risk. However, RIC with FCR was associated with a significantly reduced chronic GVHD risk (RR, .52; 95% CI, .36 to .77; P = .001). RIC with either Flu/Bu or FCR in patients with FL undergoing allo-HCT provides excellent 3-year OS, with acceptable rates of nonrelapse mortality. FCR-based conditioning was associated with a lower risk of chronic GVHD.

Assessment of Impact of HLA Type on Outcomes of Allogeneic Hematopoietic Stem Cell Transplantation for Chronic Lymphocytic Leukemia.

Chronic lymphocytic leukemia (CLL) is a common hematologic malignancy with many highly effective therapies. Chemorefractory disease, often characterized by deletion of chromosome 17p, has historically been associated with very poor outcomes, leading to the application of allogeneic hematopoietic stem cell transplantation (allo-HCT) for medically fit patients. Although the use of allo-HCT has declined since the introduction of novel targeted therapy for the treatment of CLL, there remains significant interest in understanding factors that may influence the efficacy of allo-HCT, the only known curative treatment for CLL. The potential benefit of transplantation is most likely due to the presence of alloreactive donor T cells that mediate the graft-versus-leukemia (GVL) effect. The recognition of potentially tumor-specific antigens in the context of class I and II major histocompatibility complex on malignant B lymphocytes by donor T cells may be influenced by subtle differences in the highly polymorphic HLA locus. Given previous reports of specific HLA alleles impacting the incidence of CLL and the clinical outcomes of allo-HCT for CLL, we sought to study the overall survival and progression-free survival of a large cohort of patients with CLL who underwent allo-HCT from fully HLA-matched related and unrelated donors at Center for International Blood and Marrow Transplant Research transplantation centers. We found no statistically significant association of allo-HCT outcomes in CLL based on previously reported HLA combinations. Additional study is needed to further define the immunologic features that portend a more favorable GVL effect after allo-HCT for CLL.

Outcomes of Measurable Residual Disease in Pediatric Acute Myeloid Leukemia before and after Hematopoietic Stem Cell Transplant: Validation of Difference from Normal Flow Cytometry with Chimerism Studies and Wilms Tumor 1 Gene Expression.

We enrolled 150 patients in a prospective multicenter study of children with acute myeloid leukemia undergoing hematopoietic stem cell transplantation (HSCT) to compare the detection of measurable residual disease (MRD) by a "difference from normal" flow cytometry (ΔN) approach with assessment of Wilms tumor 1 (WT1) gene expression without access to the diagnostic specimen. Prospective analysis of the specimens using this approach showed that 23% of patients screened for HSCT had detectable residual disease by ΔN (.04% to 53%). Of those patients who proceeded to transplant as being in morphologic remission, 10 had detectable disease (.04% to 14%) by ΔN. The disease-free survival of this group was 10% (0 to 35%) compared with 55% (46% to 64%, P < .001) for those without disease. The ΔN assay was validated using the post-HSCT specimen by sorting abnormal or suspicious cells to confirm recipient or donor origin by chimerism studies. All 15 patients who had confirmation of tumor detection relapsed, whereas the 2 patients with suspicious phenotype cells lacking this confirmation did not. The phenotype of the relapse specimen was then used retrospectively to assess the pre-HSCT specimen, allowing identification of additional samples with low levels of MRD involvement that were previously undetected. Quantitative assessment of WT1 gene expression was not predictive of relapse or other outcomes in either pre- or post-transplant specimens. MRD detected by ΔN was highly specific, but did not identify most relapsing patients. The application of the assay was limited by poor quality among one-third of the specimens and lack of a diagnostic phenotype for comparison.

Autologous transplantation versus allogeneic transplantation in patients with follicular lymphoma experiencing early treatment failure.

BACKGROUND: Early treatment failure (ETF) in follicular lymphoma (FL), defined as relapse or progression within 2 years of frontline chemoimmunotherapy, is a newly recognized marker of poor survival and identifies a high-risk group of patients with an expected 5-year overall survival (OS) rate of approximately 50%. Transplantation is an established option for relapsed FL, but its efficacy in this specific ETF FL population has not been previously evaluated. METHODS: This study compared autologous hematopoietic stem cell transplantation (auto-HCT) with either matched sibling donor (MSD) or matched unrelated donor (MUD) allogeneic hematopoietic cell transplantation (allo-HCT) as the first transplantation approach for patients with ETF FL (age ≥ 18 years) undergoing auto-HCT or allo-HCT between 2002 and 2014. The primary endpoint was OS. The secondary endpoints were progression-free survival, relapse, and nonrelapse mortality (NRM). RESULTS: Four hundred forty FL patients had ETF (auto-HCT, 240; MSD hematopoietic stem cell transplantation [HCT], 105; and MUD HCT, 95). With a median follow-up of 69 to 73 months, the adjusted probability of 5-year OS was significantly higher after auto-HCT (70%) or MSD HCT (73%) versus MUD HCT (49%; P = .0008). The 5-year adjusted probability of NRM was significantly lower for auto-HCT (5%) versus MSD (17%) or MUD HCT (33%; P < .0001). The 5-year adjusted probability of disease relapse was lower with MSD (31%) or MUD HCT (23%) versus auto-HCT (58%; P < .0001). CONCLUSIONS: Patients with high-risk FL, as defined by ETF, undergoing auto-HCT for FL have low NRM and a promising 5-year OS rate (70%). MSD HCT has lower relapse rates than auto-HCT but similar OS. Cancer 2018;124:2541-51. © 2018 American Cancer Society.

Early cytomegalovirus reactivation remains associated with increased transplant-related mortality in the current era: a CIBMTR analysis.

Single-center studies have reported an association between early (before day 100) cytomegalovirus (CMV) reactivation and decreased incidence of relapse for acute myeloid leukemia (AML) following allogeneic hematopoietic cell transplantation. To substantiate these preliminary findings, the Center for International Blood and Marrow Transplant Research (CIBMTR) Database was interrogated to analyze the impact of CMV reactivation on hematologic disease relapse in the current era. Data from 9469 patients transplanted with bone marrow or peripheral blood between 2003 and 2010 were analyzed according to 4 disease categories: AML (n = 5310); acute lymphoblastic leukemia (ALL, n = 1883); chronic myeloid leukemia (CML, n = 1079); and myelodysplastic syndrome (MDS, n = 1197). Median time to initial CMV reactivation was 41 days (range, 1-362 days). CMV reactivation had no preventive effect on hematologic disease relapse irrespective of diagnosis. Moreover, CMV reactivation was associated with higher nonrelapse mortality [relative risk [RR] among disease categories ranged from 1.61 to 1.95 and P values from .0002 to <.0001; 95% confidence interval [CI], 1.14-2.61). As a result, CMV reactivation was associated with lower overall survival for AML (RR = 1.27; 95% CI, 1.17-1.38; P <.0001), ALL (RR = 1.46; 95% CI, 1.25-1.71; P <.0001), CML (RR = 1.49; 95% CI, 1.19-1.88; P = .0005), and MDS (RR = 1.31; 95% CI, 1.09-1.57; P = .003). In conclusion, CMV reactivation continues to remain a risk factor for poor posttransplant outcomes and does not seem to confer protection against hematologic disease relapse.

Variable selection with group structure in competing risks quantile regression.

We study the group bridge and the adaptive group bridge penalties for competing risks quantile regression with group variables. While the group bridge consistently identifies nonzero group variables, the adaptive group bridge consistently selects variables not only at group level but also at within-group level. We allow the number of covariates to diverge as the sample size increases. The oracle property for both methods is also studied. The performance of the group bridge and the adaptive group bridge is compared in simulation and in a real data analysis. The simulation study shows that the adaptive group bridge selects nonzero within-group variables more consistently than the group bridge. A bone marrow transplant study is provided as an example.

Group and within-group variable selection for competing risks data.

Variable selection in the presence of grouped variables is troublesome for competing risks data: while some recent methods deal with group selection only, simultaneous selection of both groups and within-group variables remains largely unexplored. In this context, we propose an adaptive group bridge method, enabling simultaneous selection both within and between groups, for competing risks data. The adaptive group bridge is applicable to independent and clustered data. It also allows the number of variables to diverge as the sample size increases. We show that our new method possesses excellent asymptotic properties, including variable selection consistency at group and within-group levels. We also show superior performance in simulated and real data sets over several competing approaches, including group bridge, adaptive group lasso, and AIC / BIC-based methods.

Screening group variables in the proportional hazards model.

We propose a method to screen group variables under the high dimensional group variable setting for the proportional hazards model. We study the sure screening property of the proposed method for independent and clustered survival data. The simulation study shows that the proposed method performs better for group variable screening than some existing procedures.