Predictive probability methods for interim monitoring in clinical trials with longitudinal outcomes.

In clinical research and development, interim monitoring is critical for better decision-making and minimizing the risk of exposing patients to possible ineffective therapies. For interim futility or efficacy monitoring, predictive probability methods are widely adopted in practice. Those methods have been well studied for univariate variables. However, for longitudinal studies, predictive probability methods using univariate information from only completers may not be most efficient, and data from on-going subjects can be utilized to improve efficiency. On the other hand, leveraging information from on-going subjects could allow an interim analysis to be potentially conducted once a sufficient number of subjects reach an earlier time point. For longitudinal outcomes, we derive closed-form formulas for predictive probabilities, including Bayesian predictive probability, predictive power, and conditional power and also give closed-form solutions for predictive probability of success in a future trial and the predictive probability of success of the best dose. When predictive probabilities are used for interim monitoring, we study their distributions and discuss their analytical cutoff values or stopping boundaries that have desired operating characteristics. We show that predictive probabilities utilizing all longitudinal information are more efficient for interim monitoring than that using information from completers only. To illustrate their practical application for longitudinal data, we analyze 2 real data examples from clinical trials.

Radiotherapy combined with nivolumab or temozolomide for newly diagnosed glioblastoma with unmethylated MGMT promoter: An international randomized phase III trial.

BACKGROUND: Addition of temozolomide (TMZ) to radiotherapy (RT) improves overall survival (OS) in patients with glioblastoma (GBM), but previous studies suggest that patients with tumors harboring an unmethylated MGMT promoter derive minimal benefit. The aim of this open-label, phase III CheckMate 498 study was to evaluate the efficacy of nivolumab (NIVO) + RT compared with TMZ + RT in newly diagnosed GBM with unmethylated MGMT promoter. METHODS: Patients were randomized 1:1 to standard RT (60 Gy) + NIVO (240 mg every 2 weeks for eight cycles, then 480 mg every 4 weeks) or RT + TMZ (75 mg/m2 daily during RT and 150-200 mg/m2/day 5/28 days during maintenance). The primary endpoint was OS. RESULTS: A total of 560 patients were randomized, 280 to each arm. Median OS (mOS) was 13.4 months (95% CI, 12.6 to 14.3) with NIVO + RT and 14.9 months (95% CI, 13.3 to 16.1) with TMZ + RT (hazard ratio [HR], 1.31; 95% CI, 1.09 to 1.58; P = .0037). Median progression-free survival was 6.0 months (95% CI, 5.7 to 6.2) with NIVO + RT and 6.2 months (95% CI, 5.9 to 6.7) with TMZ + RT (HR, 1.38; 95% CI, 1.15 to 1.65). Response rates were 7.8% (9/116) with NIVO + RT and 7.2% (8/111) with TMZ + RT; grade 3/4 treatment-related adverse event (TRAE) rates were 21.9% and 25.1%, and any-grade serious TRAE rates were 17.3% and 7.6%, respectively. CONCLUSIONS: The study did not meet the primary endpoint of improved OS; TMZ + RT demonstrated a longer mOS than NIVO + RT. No new safety signals were detected with NIVO in this study. The difference between the study treatment arms is consistent with the use of TMZ + RT as the standard of care for GBM.ClinicalTrials.gov NCT02617589.

Potential candidate genomic biomarkers of drug induced vascular injury in the rat.

Drug-induced vascular injury is frequently observed in rats but the relevance and translation to humans present a hurdle for drug development. Numerous structurally diverse pharmacologic agents have been shown to induce mesenteric arterial medial necrosis in rats, but no consistent biomarkers have been identified. To address this need, a novel strategy was developed in rats to identify genes associated with the development of drug-induced mesenteric arterial medial necrosis. Separate groups (n=6/group) of male rats were given 28 different toxicants (30 different treatments) for 1 or 4 days with each toxicant given at 3 different doses (low, mid and high) plus corresponding vehicle (912 total rats). Mesentery was collected, frozen and endothelial and vascular smooth muscle cells were microdissected from each artery. RNA was isolated, amplified and Affymetrix GeneChip® analysis was performed on selectively enriched samples and a novel panel of genes representing those which showed a dose responsive pattern for all treatments in which mesenteric arterial medial necrosis was histologically observed, was developed and verified in individual endothelial cell- and vascular smooth muscle cell-enriched samples. Data were confirmed in samples containing mesentery using quantitative real-time RT-PCR (TaqMan™) gene expression profiling. In addition, the performance of the panel was also confirmed using similarly collected samples obtained from a timecourse study in rats given a well established vascular toxicant (Fenoldopam). Although further validation is still required, a novel gene panel has been developed that represents a strategic opportunity that can potentially be used to help predict the occurrence of drug-induced mesenteric arterial medial necrosis in rats at an early stage in drug development.

Effect of Nivolumab vs Bevacizumab in Patients With Recurrent Glioblastoma: The CheckMate 143 Phase 3 Randomized Clinical Trial.

Importance: Clinical outcomes for glioblastoma remain poor. Treatment with immune checkpoint blockade has shown benefits in many cancer types. To our knowledge, data from a randomized phase 3 clinical trial evaluating a programmed death-1 (PD-1) inhibitor therapy for glioblastoma have not been reported. Objective: To determine whether single-agent PD-1 blockade with nivolumab improves survival in patients with recurrent glioblastoma compared with bevacizumab. Design, Setting, and Participants: In this open-label, randomized, phase 3 clinical trial, 439 patients with glioblastoma at first recurrence following standard radiation and temozolomide therapy were enrolled, and 369 were randomized. Patients were enrolled between September 2014 and May 2015. The median follow-up was 9.5 months at data cutoff of January 20, 2017. The study included 57 multicenter, multinational clinical sites. Interventions: Patients were randomized 1:1 to nivolumab 3 mg/kg or bevacizumab 10 mg/kg every 2 weeks until confirmed disease progression, unacceptable toxic effects, or death. Main Outcomes and Measures: The primary end point was overall survival (OS). Results: A total of 369 patients were randomized to nivolumab (n = 184) or bevacizumab (n = 185). The MGMT promoter was methylated in 23.4% (43/184; nivolumab) and 22.7% (42/185; bevacizumab), unmethylated in 32.1% (59/184; nivolumab) and 36.2% (67/185; bevacizumab), and not reported in remaining patients. At median follow-up of 9.5 months, median OS (mOS) was comparable between groups: nivolumab, 9.8 months (95% CI, 8.2-11.8); bevacizumab, 10.0 months (95% CI, 9.0-11.8); HR, 1.04 (95% CI, 0.83-1.30); P = .76. The 12-month OS was 42% in both groups. The objective response rate was higher with bevacizumab (23.1%; 95% CI, 16.7%-30.5%) vs nivolumab (7.8%; 95% CI, 4.1%-13.3%). Grade 3/4 treatment-related adverse events (TRAEs) were similar between groups (nivolumab, 33/182 [18.1%]; bevacizumab, 25/165 [15.2%]), with no unexpected neurological TRAEs or deaths due to TRAEs. Conclusions and Relevance: Although the primary end point was not met in this randomized clinical trial, mOS was comparable between nivolumab and bevacizumab in the overall patient population with recurrent glioblastoma. The safety profile of nivolumab in patients with glioblastoma was consistent with that in other tumor types. Trial Registration: ClinicalTrials.gov Identifier: NCT02017717.

Weighted least-squares deconvolution method for discovery of group differences between complex biofluid 1H NMR spectra.

Biomarker discovery through analysis of high-throughput NMR data is a challenging, time-consuming process due to the requirement of sophisticated, dataset specific preprocessing techniques and the inherent complexity of the data. Here, we demonstrate the use of weighted, constrained least-squares for fitting a linear mixture of reference standard data to complex urine NMR spectra as an automated way of utilizing current assignment knowledge and the ability to deconvolve confounded spectral regions. Following the least-squares fit, univariate statistics were used to identify metabolites associated with group differences. This method was evaluated through applications on simulated datasets and a murine diabetes dataset. Furthermore, we examined the differential ability of various weighting metrics to correctly identify discriminative markers. Our findings suggest that the weighted least-squares approach is effective for identifying biochemical discriminators of varying physiological states. Additionally, the superiority of specific weighting metrics is demonstrated in particular datasets. An additional strength of this methodology is the ability for individual investigators to couple this analysis with laboratory specific preprocessing techniques.

Diagnostic performance of traditional hepatobiliary biomarkers of drug-induced liver injury in the rat.

Nonclinical studies provide the opportunity to anchor biochemical with morphologic findings; however, liver injury is often complex and heterogeneous, confounding the ability to relate biochemical changes with specific patterns of injury. The aim of the current study was to compare diagnostic performance of hepatobiliary markers for specific manifestations of drug-induced liver injury in rat using data collected in a recent hepatic toxicogenomics initiative in which rats (n = 3205) were given 182 different treatments for 4 or 14 days. Diagnostic accuracy of alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (Tbili), serum bile acids (SBA), alkaline phosphatase (ALP), gamma glutamyl transferase (GGT), total cholesterol (Chol), and triglycerides (Trig) was evaluated for specific types of liver histopathology by Receiver Operating Characteristic (ROC) analysis. To assess the relationship between biochemical and morphologic changes in the absence of hepatocellular necrosis, a second ROC analysis was performed on a subset of rats (n = 2504) given treatments (n = 152) that did not cause hepatocellular necrosis. In the initial analysis, ALT, AST, Tbili, and SBA had the greatest diagnostic utility for manifestations of hepatocellular necrosis and biliary injury, with comparable magnitude of area under the ROC curve and serum hepatobiliary marker changes for both. In the absence of hepatocellular necrosis, ALT increases were observed with biochemical or morphologic evidence of cholestasis. In both analyses, diagnostic utility of ALP and GGT for biliary injury was limited; however, ALP had modest diagnostic value for peroxisome proliferation, and ALT, AST, and total Chol had moderate diagnostic utility for phospholipidosis. None of the eight markers evaluated had diagnostic value for manifestations of hypertrophy, cytoplasmic rarefaction, inflammation, or lipidosis.

Multi-platform investigation of the metabolome in a leptin receptor defective murine model of type 2 diabetes.

We describe a multi-platform ((1)H NMR, LC-MS, microarray) investigation of metabolic disturbances associated with the leptin receptor defective (db/db) mouse model of type 2 diabetes using novel assignment methodologies. For the first time, several urinary metabolites were found to be associated with diabetes and/or diabetes progression and confirmed in both NMR and LC-MS datasets. The confirmed metabolites were trimethylamine-n-oxide (TMAO), creatine, carnitine, and phenylalanine. TMAO and phenylalanine were both elevated in db/db mice and decreased in these mice with age. Levels of both creatine and carnitine increase in diabetic mice with age and creatine was also significantly decreased in db/db mice. Additionally, many metabolic markers were found by either NMR or LC-MS, but could not be found in both, due to instrumental limitations. This indicates that the combined use of NMR and LC-MS instrumentation provides complementary information that would be otherwise unattainable. Pathway analyses of urinary metabolites and liver, muscle, and adipose tissue transcripts from the db/db model were also performed to identify altered biochemical processes in the diabetic mice. Metabolite and liver transcript levels associated with the TCA cycle and steroid processes were altered in db/db mice. In addition, gene expression in muscle and liver associated with fatty acid processing was altered in the diabetic mice and similar evidence was observed in the LC-MS data. Our findings highlight the importance of a number of processes known to be associated with diabetes and reveal tissue specific responses to the condition. When studying metabolic disorders such as diabetes, multiple platform integrated profiling of metabolite alterations in biofluids can provide important insights into the processes underlying the disease.

Interlaboratory evaluation of genomic signatures for predicting carcinogenicity in the rat.

The Critical Path Institute recently established the Predictive Safety Testing Consortium, a collaboration between several companies and the U.S. Food and Drug Administration, aimed at evaluating and qualifying biomarkers for a variety of toxicological endpoints. The Carcinogenicity Working Group of the Predictive Safety Testing Consortium has concentrated on sharing data to test the predictivity of two published hepatic gene expression signatures, including the signature by Fielden et al. (2007, Toxicol. Sci. 99, 90-100) for predicting nongenotoxic hepatocarcinogens, and the signature by Nie et al. (2006, Mol. Carcinog. 45, 914-933) for predicting nongenotoxic carcinogens. Although not a rigorous prospective validation exercise, the consortium approach created an opportunity to perform a meta-analysis to evaluate microarray data from short-term rat studies on over 150 compounds. Despite significant differences in study designs and microarray platforms between laboratories, the signatures proved to be relatively robust and more accurate than expected by chance. The accuracy of the Fielden et al. signature was between 63 and 69%, whereas the accuracy of the Nie et al. signature was between 55 and 64%. As expected, the predictivity was reduced relative to internal validation estimates reported under identical test conditions. Although the signatures were not deemed suitable for use in regulatory decision making, they were deemed worthwhile in the early assessment of drugs to aid decision making in drug development. These results have prompted additional efforts to rederive and evaluate a QPCR-based signature using these samples. When combined with a standardized test procedure and prospective interlaboratory validation, the accuracy and potential utility in preclinical applications can be ascertained.

Histone acetylation and transcriptional regulation in the genome of Saccharomyces cerevisiae.

MOTIVATION: In eukaryotic genomes, histone acetylation and thereafter departure from the chromatin are essential for gene transcription initiation. Because gene transcription is tightly regulated by transcription factors, there are some speculations on the cooperation of histone acetylation and transcription factor binding. However, systematic statistical analyses of this relationship on a genomic scale have not been reported. RESULTS: We apply several statistical methods to explore this relationship on two recent genomic datasets: acetylation levels on 11 histone lysines and binding activities of 203 transcription factors, both in promoter regions across the yeast genome. By canonical correlation analysis, we find that a histone acetylation pattern is correlated with a certain profile of transcription factor binding in the genome. Furthermore, after clustering the genes by their acetylation levels on the 11 histone lysines, the genes within clusters show distinct transcription factor binding profiles, as discovered by principle component analysis. Even after applying fairly stringent statistical measurement, most of these clusters have transcription factors with binding activities significantly deviated from the overall genome. We conclude that in the yeast genome, there is a strong correlation between histone acetylation and transcription factor binding in the promoter regions.