Doubly robust weighted log-rank tests and Renyi-type tests under non-random treatment assignment and dependent censoring

2018 ◽  
Vol 28 (9) ◽  
pp. 2649-2664
Author(s):  
Chenxi Li
Keyword(s):  
Dependent Censoring ◽  
Rank Test ◽  
Rank Tests ◽  
Data Set ◽  
Treatment Assignment ◽  
Treatment Groups ◽  
Event Time ◽  
Log Rank Test ◽  
Doubly Robust ◽  
Log Rank Tests

The log-rank test is widely used to test difference in event time distribution between treatment groups. However, if subjects are not randomly assigned to treatment groups, which is often the case in observation studies, the log-rank test is not asymptotically correct for detecting group survival difference due to the imbalance of confounding variables between groups. We develop a class of modified weighted log-rank tests and Renyi-type tests for two-sample survival comparison under non-random treatment assignment. The new tests can also account for non-random censoring that depends on baseline covariates. The proposed methods involve building working models for treatment assignment, cause-specific hazard of dependent censoring, and the time to event. We prove that, when either the models for treatment assignment and dependent censoring or the model for the event time is true, the new tests are asymptotically correct, i.e. being doubly robust. Numerical experiments demonstrate the tests’ double-robustness property in finite samples of realistic sizes, and also show that the doubly robust log-rank test is at least as powerful as the regular log-rank test when the treatment assignment is random and there is no dependent censoring. An application to a kidney transplant data set illustrates the utility of the proposed methods.

scholarly journals Longitudinal analysis of quality of life following treatment with Asunercept plus reirradiation versus reirradiation in progressive glioblastoma patients

2019 ◽  
Vol 145 (3) ◽  
pp. 531-540
Author(s):  
Wolfgang Wick ◽  
Andriy Krendyukov ◽  
Klaus Junge ◽  
Thomas Höger ◽  
Harald Fricke
Keyword(s):  
Quality Of Life ◽  
Neurological Status ◽  
Recurrent Glioblastoma ◽  
Rank Test ◽  
Treatment Groups ◽  
Log Rank Test ◽  
Kaplan Meier ◽  
The Central Nervous System ◽  
Eortc Qlq

Abstract Purpose Glioblastoma is an aggressive malignant cancer of the central nervous system, with disease progression associated with deterioration of neurocognitive function and quality of life (QoL). As such, maintenance of QoL is an important treatment goal. This analysis presents time to deterioration (TtD) of QoL in patients with recurrent glioblastoma receiving Asunercept plus reirradiation (rRT) or rRT alone. Methods Data from patients with a baseline and ≥ 1 post-baseline QoL assessment were included in this analysis. TtD was defined as the time from randomisation to the first deterioration in the EORTC QLQ-C15, PAL EORTC QLQ-BN20 and Medical Research Council (MRC)-Neurological status. Deterioration was defined as a decrease of ≥ 10 points from baseline in the QLQ-C15 PAL overall QoL and functioning scales, an increase of ≥ 10 points from baseline in the QLQ-C15 PAL fatigue scale and the QLQ-BN20 total sum of score, and a rating of “Worse” in the MRC-Neurological status. Patients without a deterioration were censored at the last QoL assessment. Kaplan–Meier estimates were used to describe TtD and treatment groups (Asunercept + rRT or rRT alone) were compared using the log-rank test. Results Treatment with Asunercept + rRT was associated with significant improvement of TtD compared with rRT alone for QLQ-CL15 PAL overall QoL and physical functioning, and MRC Neurological Status (p ≤ 0.05). In the Asunercept + rRT group, QoL was maintained beyond progresison of disease (PoD). Conclusion Treatment with Asunercept plus rRT significantly prolongs TtD and maintains QoL versus rRT alone in recurrent glioblastoma patients.

Radiosurgery alone for solitary and multiple brain metastasis

2006 ◽  
Vol 24 (18_suppl) ◽  
pp. 1541-1541
Author(s):  
J. L. Fox ◽  
L. Kleinberg ◽  
S. Kharkar ◽  
R. E. Clatterbuck ◽  
P. Wang ◽  
...  
Keyword(s):  
Brain Metastases ◽  
Median Number ◽  
Rank Test ◽  
Small Cell Lung ◽  
Treatment Groups ◽  
Time Commitment ◽  
Whole Brain Radiation ◽  
Log Rank Test ◽  
Brain Radiation

1541 Background: Whole-brain radiation (WBRT) in the management of brain metastases treated with radiosurgery (RS) is controversial. Methods: Ninety-eight patients were treated for brain metastases with RS at Johns Hopkins between 4/03 and 7/05. Twenty-eight patients received RS alone after failing WBRT, 33 received RS alone for initial metastases and 37 received RS along with WBRT. Forty-five patients were women and 53 were men, with a median age of 56 (range, 18–92). Histology was: non-small cell lung cancer - 35, breast -14, melanoma -10, renal cell carcinoma - 9, and other - 30. Ninety-two (94%) pts had a KPS of ≥ 70 (median 80). The median number of metastases was 2 (range, 1–14). Results: Follow-up data from date of RS was available for 96 patients. Among those who received RS along with planned WBRT, median survival (MS) was 6.6 months with 1-yr overall survival (OS) 38%. Among patients treated initially with RS alone, MS was 9.7 months with 1-yr OS 42%. Among patients treated with RS for recurrent metastases after prior WBRT, MS was 6.8 months with 1-yr OS 24%. There were no significant differences in survival amongst these 3 treatment groups (p=0.73, log-rank test). For patients with 1–3 metastases (n=66), 1-yr OS was 38% versus 32% for those with ≥ 4 (n=32). Median survivals were 8.4 and 6.7 months, respectively (p=NS). Of patients treated with RS for recurrence, 7 of 25 (28%) with available follow-up data developed recurrent or new metastases whereas 11/27 (41%) treated with RS and planned WBRT and 15/27 (56%) who had RS alone as initial treatment had documented recurrent or new metastases. Conclusions: RS alone may be an effective treatment that preserves survival for those with single or multiple brain metastases at initial presentation or recurrence, but the tradeoff between the marginal increase in risk of brain recurrence versus toxicity and time commitment for WBRT needs further evaluation. The ongoing US Intergroup randomized trial, N0572/Z300, will address some of these questions. No significant financial relationships to disclose.

scholarly journals Neutropenia in Children Treated with Deferiprone or Deferasirox: A Report of the Largest Randomized Trial of Oral Chelators in Transfusion-Dependent Pediatric Patients

Blood ◽  
2019 ◽  
Vol 134 (Supplement_1) ◽  
pp. 3552-3552
Author(s):  
Fernando Tricta ◽  
Mariagrazia Felisi ◽  
Oscar Della Pasqua ◽  
Amal El-Beshlawy ◽  
Hoda Hassab ◽  
...  
Keyword(s):  
Board Of Directors ◽  
Neutrophil Count ◽  
Severe Neutropenia ◽  
Rank Test ◽  
Cumulative Hazard ◽  
Advisory Committees ◽  
Treatment Groups ◽  
Log Rank Test ◽  
Kaplan Meier ◽  
Oral Chelators

Introduction: Agranulocytosis/severe neutropenia is an established adverse event during deferiprone (DFP) use. Less is known about milder episodes, which are frequently transient despite continuous deferiprone use. To provide further insight into this topic, we compared the incidence of neutropenia during DFP or deferasirox (DFX) treatment in the randomized Deferiprone Evaluation in Paediatrics (DEEP-2) trial, where blood neutrophil count was regularly monitored in patients randomized to be treated with DFP or DFX. Methods: DEEP-2 was a multicenter, randomized, 12-month, open-label trial comparing DFP vs DFX in pediatric (<18 years old) patients with transfusion-dependent hemoglobinopathies. 390 patients from 22 centers in 7 countries were randomized (1:1 DFP:DFX) and received at least one dose of the study medication (193 on DFP and 197 on DFX). Neutrophil count was regularly assessed every 7 +/-7 days in all patients. Neutrophil counts <500/mm3 were classified as agranulocytosis/severe neutropenia, 500 - <1,000 /mm3 as moderate neutropenia, and 1,000 - <1,500 /mm3 as mild neutropenia. The incidence of neutrophil counts below the threshold of neutropenia (1,500/mm3) and the rate of reported episodes of neutropenia as identified by the treating physician were compared between the two treatments. An ANOVA model was used to compare the time to neutropenia and time for its resolution between the two treatment groups. To compare the cumulative hazard curves was used the Kaplan-Meier log rank test. Results: 3579 and 4027 neutrophil counts were available for DFP- and DFX-treated patients, respectively. 47 (1.3%) of the total counts in 24 (12.4%) of the 193 DFP-treated patients versus 48 (1.2%) of the total counts in 27 (13.7%) of the 197 DFX-treated patients were below 1,500/mm3. Of these, 28 cases in 23 DFP-treated patients and 15 cases in 11 DFX-treated patients were reported by the treating physician as neutropenia, corresponding to a global incidence rate 11.9% for DFP and 5.6% for DFX (p=0.081, Chi-Square test). 23 (82.1%) of the 28 episodes of neutropenia during DFP use were considered drug related vs 2 (13.3%) of the 15 episodes during DFX use (p-value < 0.001, Fisher Exact test) (table 1). The mean (SD) treatment duration with either DFP or DFX prior to diagnosis of mild or moderated neutropenia was 127 (96.1) days and 101 (85.7) respectively. All those episodes, except for 2, resolved within a mean time of 13 days (1 - 42 days) in DFP-treated patients and 18 days (6 - 46 days) in DFX-treated patients. The 2 cases where neutropenia did not resolve were diagnosed as constitutional neutropenia and bone marrow suppression. There was no significant difference in those results between the two treatment groups as assessed by one-way ANOVA(p = 0.379), Log-Rank test (p = 0.065) or cumulative-hazard and Kaplan-Meier. During the study 3 events of agranulocytosis occurred, all in patients treated with DFP and not included in the present analysis. All 3 episodes resolved. Conclusion: Data from the largest randomized trial of oral chelators in transfusion-dependent pediatric patients provide evidence that, a drop in the neutrophil count below the threshold for mild neutropenia is very common (>10% of patients) for both DFP and DFX treated patients. All episodes of neutropenia, except for 2 with specific etiology, resolved, irrespective of their severity and of chelator used. A causal relationship of neutropenia to chelator use varied based on the chelator; the majority of events observed during DFP use were assessed by the clinician as drug related, whereas the majority of events observed during DFX use were assessed unrelated to its use. These data indicate that while agranulocytosis rate in DFP patients is not changed from previous reports, moderate/mild neutropenia represent discrete events in patients undergoing oral iron chelation therapy. Disclosures Tricta: ApoPharma: Employment. Della Pasqua:GlaxoSmithKline: Employment. Kattamis:Ionis: Membership on an entity's Board of Directors or advisory committees; ViFOR: Membership on an entity's Board of Directors or advisory committees; Vertex: Membership on an entity's Board of Directors or advisory committees; Apopharma: Honoraria; Celgene Corporation: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis Oncology: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Reggiardo:CVBF: Consultancy. Spino:ApoPharma: Employment. Tsang:Apotex Inc.: Employment.

Decrease of Mortality during Deferiprone Treatments: Results from A Large Randomised Cohort of Thalassemia Major Patients Under the Auspices of the Italian Society for Thalassemia and Hemoglobinopathies

Blood ◽  
2008 ◽  
Vol 112 (11) ◽  
pp. 3885-3885 ◽  
Author(s):  
Aurelio Maggio ◽  
Angela Vitrano ◽  
Marcello Capra ◽  
Liana Cuccia ◽  
Francesco Gagliardotto ◽  
...  
Keyword(s):  
Heart Failure ◽  
Cox Regression ◽  
Thalassemia Major ◽  
Randomised Clinical Trial ◽  
Hazard Ratio ◽  
Rank Test ◽  
Cardiac Damage ◽  
Treatment Groups ◽  
Log Rank Test

Abstract Prognosis of thalassemia major patients has dramatically improved in the past two decades. Previous papers, including retrospective and prospective non-randomised clinical trials, suggested that mortality, due mainly to cardiac damage, was less or completely absent in patients treated with DFP alone or with associated chelation treatment. For this reason, the main aim of this study was to evaluate whether the addition of deferiprone treatment was also associated with a mortality decrease among a large randomised cohort of thalassemia major patients. Survival analysis was performed among 264 thalassemia major patients assessed for eligibility from 09/30/2000 to 01/31/2008, during a long-term multicentre randomised clinical trial. The reported chelation therapies included sequential DFP-DFO, associated DFP and DFO, DFP and DFO interventions. The survival curves were compared by gender and treatment groups using the long-rank test. Cox regression models were used to explore association between risk for death among treatments and survival time. All statistical analyses were performed by STATA 9.2. All these patients performed DFO before the date of randomisation. One death was due to a graft versus host disease (GVHD) in a patient underwent bone marrow transplantation and this patient was censored at the time of transplant. The improved survival for sequential DFP-DFO, DFP-alone, and associated DFP-DFO treated patients versus DFO-treated was statistically significant (log-rank test, χ2= 18.64; p≤0.01). In fact, no deaths were reported during DFP-alone and DFP-DFO associated treatments along a 564.5 person-years period of observation. Only one death was reported during DFP-DFO sequential treatment in a patient who had experienced 1-year before an episode of heart failure. All other ten deaths were among patients under DFO treatment. The hazard ratio for death of DFO treatment versus other treatments was 27.78 (p= 0.002). The main factors correlated with increased hazard ratio for death were cirrhosis, arrhythmia, previous episode of heart failure, diabetes, hypogonadism, hypothyroidism. No correlation between serum ferritin levels and hazard ratio for death was found. These results confirm as deferiprone alone or in addition to deferoxamine intervention is able to reduce mortality in thalassemia major patients probably because of its specific cardioprotective effect occurring independently from body iron overloading

A nonparametric test for the association between longitudinal covariates and censored survival data

Biostatistics ◽  
2019 ◽  
Vol 21 (4) ◽  
pp. 727-742
Author(s):  
Ramon Oller ◽  
Guadalupe Gómez Melis
Keyword(s):  
Survival Time ◽  
Joint Model ◽  
Time Dependent ◽  
Rank Test ◽  
Test Statistics ◽  
Time To Event ◽  
Data Set ◽  
Log Rank Test ◽  
Interval Censored ◽  
Empirical Size

Summary Many biomedical studies focus on the association between a longitudinal measurement and a time-to-event outcome while quantifying this association by means of a longitudinal-survival joint model. In this article we propose using the $LLR$ test — a longitudinal extension of the log-rank test statistic given by Peto and Peto (1972) — to provide evidence of a plausible association between a time-to-event outcome (right- or interval-censored) and a time-dependent covariate. As joint model methods are complex and hard to interpret, it is wise to conduct a preliminary test such as $LLR$ for checking the association between both processes. The $LLR$ statistic can be expressed in the form of a weighted difference of hazards, yielding a broad class of weighted log-rank test statistics known as $LWLR$, which allow a specific emphasis along the time axis of the effects of the time-dependent covariate on the survival. The asymptotic distribution of $LLR$ is derived by means of a permutation approach under the assumption that the censoring mechanism is independent of the survival time and the longitudinal covariate. A simulation study is conducted to evaluate the performance of the test statistics $LLR$ and $LWLR$, showing that the empirical size is close to the nominal significance level and that the power of the test depends on the association between the covariates and the survival time. A data set together with a toy example are used to illustrate the $LLR$ test. The data set explores the study Epidemiology of Diabetes Interventions and Complications (Sparling and others, 2006) which includes interval-censored data. A software implementation of our method is available on github (https://github.com/RamonOller/LWLRtest).

A weighted log-rank test for comparing two survival curves

2020 ◽  
Vol 26 (3) ◽  
pp. 253-262
Author(s):  
Seung-Hwan Lee ◽  
Eun-Joo Lee
Keyword(s):  
Weight Function ◽  
Real World ◽  
Null Hypothesis ◽  
New Method ◽  
Rank Test ◽  
Survival Curves ◽  
Rank Tests ◽  
Asymptotically Normal ◽  
Real World Application ◽  
Log Rank Test

AbstractThis paper proposes a weighted log-rank test that maintains sensitivity to realistic alternatives of two survival curves, such as crossing curves, in the presence of heavy censoring. The new test incorporates a weight function that changes over the censoring level, increasing adaptivity and flexibility of the commonly used weighted log-rank tests. The new statistic is asymptotically normal under the null hypothesis that there is no difference in survival between two groups. The performances of the new test are evaluated via simulations under both proportional and non-proportional alternatives. We illustrate the new method with a real-world application.

Sign in / Sign up

Export Citation Format

Share Document