Case weighted power priors for hybrid control analyses with time-to-event data.

We develop a method for hybrid analyses that uses external controls to augment internal control arms in randomized controlled trials (RCTs) where the degree of borrowing is determined based on similarity between RCT and external control patients to account for systematic differences (e.g., unmeasured confounders). The method represents a novel extension of the power prior where discounting weights are computed separately for each external control based on compatibility with the randomized control data. The discounting weights are determined using the predictive distribution for the external controls derived via the posterior distribution for time-to-event parameters estimated from the RCT. This method is applied using a proportional hazards regression model with piecewise constant baseline hazard. A simulation study and a real-data example are presented based on a completed trial in non-small cell lung cancer. It is shown that the case weighted power prior provides robust inference under various forms of incompatibility between the external controls and RCT population.

Application of Bayesian approaches in drug development: starting a virtuous cycle.

The pharmaceutical industry and its global regulators have routinely used frequentist statistical methods, such as null hypothesis significance testing and p values, for evaluation and approval of new treatments. The clinical drug development process, however, with its accumulation of data over time, can be well suited for the use of Bayesian statistical approaches that explicitly incorporate existing data into clinical trial design, analysis and decision-making. Such approaches, if used appropriately, have the potential to substantially reduce the time and cost of bringing innovative medicines to patients, as well as to reduce the exposure of patients in clinical trials to ineffective or unsafe treatment regimens. Nevertheless, despite advances in Bayesian methodology, the availability of the necessary computational power and growing amounts of relevant existing data that could be used, Bayesian methods remain underused in the clinical development and regulatory review of new therapies. Here, we highlight the value of Bayesian methods in drug development, discuss barriers to their application and recommend approaches to address them. Our aim is to engage stakeholders in the process of considering when the use of existing data is appropriate and how Bayesian methods can be implemented more routinely as an effective tool for doing so.

The use of phase 2 interim analysis to expedite drug development decisions.

PURPOSE: To expedite drug development, we propose a two-step decision-making process that utilizes interim efficacy results from a comparative phase 2 trial to determine whether to accelerate subsequent phase 3 preparations, and final analysis to ultimately determine whether to conduct phase 3 testing. METHODS: The operational characteristics of this process were evaluated by modeling simulated data of oncology trials and retrospectively analyzing data from historical comparative phase 2 trials. Progression-free survival (PFS) was used as the primary endpoint; the estimated PFS hazard ratios (HRs) of ≤0.60 at interim and of ≤0.65 at final analysis favoring the experimental arm were defined as positive results. The conditional probability of achieving a target PFS HR at final analysis, based on observed interim results, was also estimated by imputing post-interim data with and without the proportional hazard assumption. RESULTS: Simulations of phase 2 trials showed that estimated interim PFS HRs correlated with estimated final PFS HRs, with reasonably low false-positive rates for supporting phase 3 "go" decisions at interim. Using observed historical data, decisions based on interim PFS analyses also matched final phase 3 "go" and "no-go" decisions with a false-positive rate of 16.7% (2 of 12 trials) and a false-negative rate of 9.4% (3 of 32 trials). Analytical modeling accurately predicted final PFS HRs from observed interim data when accounting for appropriate underlying assumptions. CONCLUSIONS: The results support the usefulness of a two-step decision-making process that utilizes interim phase 2 results to reduce the interval between phase 2 completion and phase 3 initiation.

An industry statistician's perspective on PHC drug development.

In the past decade, the cost of drug development has increased significantly. The estimates vary widely but frequently quoted numbers are staggering-it takes 10-15 years and billions of dollars to bring a drug to patients. To a large extent this is due to many long, expensive and ultimately unsuccessful drug trials. While one approach to combat the low yield on investment could be to continue searching for new blockbusters, an alternative method would lead us to focus on testing new targeted treatments that have a strong underlying scientific rationale and are more likely to provide enhanced clinical benefit in population subsets defined by molecular diagnostics. Development of these new treatments, however, cannot follow the usual established path; new strategies and approaches are required for the co-development of novel therapeutics and the diagnostic. In this paper we will review, from the point of view of industry, the approaches to, and challenges of drug development strategies incorporating predictive biomarkers into clinical programs. We will outline the basic concepts behind co-development with predictive biomarkers and summarize the current regulatory paradigm. We will present guiding principles of personalized health care (PHC) development and review the statistical, strategic, regulatory and operational challenges that statisticians regularly encounter on development programs with a PHC component. Some practical recommendations for team statisticians involved in PHC drug development are included. The majority of the examples and recommendations are drawn from oncology but broader concepts apply across all therapeutic areas.

Hormone receptor status does not affect the clinical benefit of trastuzumab therapy for patients with metastatic breast cancer.

BACKGROUND: Hormone receptor (HR) and HER2 signaling pathways are involved in the regulation of breast cancer proliferation. The impact of HR status on the clinical outcome of patients with HER2-overexpressing disease treated with the monoclonal antibody trastuzumab is unknown. PATIENTS AND METHODS: To evaluate this, we conducted a retrospective analysis of 805 patients with metastatic breast cancer enrolled in 3 clinical trials comparing trastuzumab in combination with chemotherapy versus chemotherapy alone or trastuzumab monotherapy as first-, second-, or third-line treatment. Patients whose tumor samples overexpressed HER2 by fluorescence in situ hybridization (FISH) were stratified based on HR status, and clinical outcomes were compared. RESULTS: Tumor samples from 596 of 805 patients were HER2overexpressing by FISH; 45% of these were HR-positive and 43% were HR-negative (HR status was unknown in 12%). Overall response rate (ORR) and time to progression (TTP) were significantly higher in patients treated with chemotherapy plus trastuzumab than in those treated with chemotherapy alone, irrespective of HR status. Median survival was longer for patients with HR-positive tumors receiving combination therapy compared with those with HR-negative tumors. In the trastuzumab monotherapy trials, ORR and TTP were similar for patients with HR-positive and HR-negative tumors. Median survival was longer for patients with HR-positive tumors compared with those with HR-negative tumors. CONCLUSION: Hormone receptor status did not affect the clinical benefit of trastuzumab given as a single agent or combined with chemotherapy. The addition of trastuzumab to chemotherapy provides an improved clinical benefit compared with chemotherapy alone, regardless of HR status.

Population pharmacokinetics of trastuzumab in patients with HER2+ metastatic breast cancer.

PURPOSE: To characterize the population pharmacokinetics of trastuzumab in patients with metastatic breast cancer. METHODS: A nonlinear mixed effect model was based on pharmacokinetic data from phase I, II, and III studies of 476 patients. The phase I study enrolled patients with advanced solid tumors. The phase II and III studies enrolled patients with HER2-positive metastatic breast cancer. Patients in the pivotal phase II and III studies were treated with a 4 mg/kg loading dose of trastuzumab followed by 2 mg/kg weekly for up to 840 days. The model adequately predicted observed trastuzumab concentrations. Model stability and performance were verified using bootstrap simulations. Percentiles, mean, and standard deviation of observed levels were compared with their distributions from 100 replicates of datasets simulated under the model. RESULTS: A two-compartment linear pharmacokinetic model best described the data and accounted for the long-term accumulation observed following weekly administration of trastuzumab. Population estimates from the base model for clearance (CL) and volume of distribution of the central compartment (V1) of trastuzumab were 0.225 L/day, and 2.95 L, respectively. Estimated terminal halflife (t1/2) based on the population estimate was 28.5 days. Interpatient variabilities in clearance and volume were 43 and 29%, respectively. The number of metastatic sites, plasma level of extracellular domain of the HER2 receptor, and patient weight were significant baseline covariates for clearance, volume, or both (P<0.005). However, these covariate effects on trastuzumab exposure were modest and not clinically important in comparison with the large inter-patient variability of CL. Concomitant chemotherapy (anthracycline plus cyclophosphamide, or paclitaxel) did not appear to influence clearance. CONCLUSION: This population pharmacokinetic model can predict trastuzumab exposure in the long-term treatment of patients with metastatic breast cancer and provide comparison of alternative dosage regimens via simulation.

Effect of cardiac dysfunction on treatment outcomes in women receiving trastuzumab for HER2-overexpressing metastatic breast cancer.

Trastuzumab improves time to disease progression (TTP) and survival when added to chemotherapy for HER-positive metastatic breast cancer (MBC), but it is associated with infrequent cardiac dysfunction (CD). We analyzed data from a previous pivotal randomized trial of 469 women with HER2-overexpressing MBC. The aim was to determine the benefit of adding trastuzumab to chemotherapy in terms of TTP that was free of CD, including all CD, moderate or severe (New York Heart Association class III/IV) CD only, or moderate or severe CD that did not improve with cardiac therapy. We also assessed moderate or severe CD-free survival. We assessed the impact of trastuzumab for these indices on the entire cohort and on specific chemotherapy subsets. Median TTP or any CD improved when trastuzumab was added to all chemotherapy (4.6 months vs. 6.6 months with trastuzumab, P = 0.0001), an anthracycline (doxorubicin or epirubicin) plus cyclophosphamide (AC; 6.0 months vs. 6.6 months, P = 0.24), and paclitaxel (2.8 months vs. 6.6 months, P = 0.0001). When defined as time to moderate or severe CD, median TTP improved when trastuzumab was added to all chemotherapy (4.6 months vs. 7.0 months, P = 0.0001), AC (6.0 months vs. 7.2 months, P = 0.02), and paclitaxel (2.8 months vs. 6.9 months, P = 0.0001). There was no statistical difference between moderate and severe CD-free survival with trastuzumab added to chemotherapy. Outcomes improved with trastuzumab despite CD. In particular, the benefit from trastuzumab/paclitaxel outweighed the potential risk of CD in patients with MBC. These types of analyses will be critical for trials assessing trastuzumab as adjuvant therapy.

Circulating D-dimer levels are better predictors of overall survival and disease progression than carcinoembryonic antigen levels in patients with metastatic colorectal carcinoma.

BACKGROUND: Fibrin formation is required for tumor angiogenesis, metastasis, and invasion. D-dimer, a fibrin degradation product, is produced when crosslinked fibrin is degraded by plasmin. The current study prospectively examined D-dimer levels in patients with metastatic colorectal carcinoma treated in a Phase II randomized trial comparing bevacizumab (Avastin, Genentech, South San Francisco, CA) plus 5-fluorouracil/leucovorin (5-FU/LV) with 5-FU/LV alone. METHODS: At least one circulating D-dimer level was evaluable in 98 of the 104 previously untreated patients with metastatic colorectal carcinoma in the current trial. Plasma D-dimer levels were determined using a quantitative immunoassay kit at enrollment, before each treatment, and at the time of trial completion or disease progression. RESULTS: At trial enrollment, 86 of 104 patients (88%) had elevated D-dimer levels (> 20 ng/mL), and 86 of 102 patients (84%) had elevated carcinoembryonic antigen (CEA) levels (> 3 ng/mL). Baseline D-dimer levels were correlated with the following baseline characteristics: CEA (Pearson coefficient, 0.31; P = 0.002), albumin levels (Pearson coefficient, -0.32; P = 0.002), tumor burden (Pearson coefficient, 0.30; P = 0.003), and number of metastatic sites (Pearson coefficient, 0.21; P = 0.04). At the time of progression, plasma D-dimer levels reached a maximum postbaseline value in 51 of 61 patients (84%), whereas the CEA level was at its maximum postbaseline value in 39 of 55 patients (71%). Baseline D-dimer levels were a strong predictor of overall survival on univariate analysis (P = 0.008) and multivariate analysis (P = 0.03). Overall, treatment with bevacizumab (5 mg/kg) and baseline D-dimer levels were the only predictors of overall survival (P < 0.05). CONCLUSIONS: The current study indicates that fibrin remodeling is an important prognostic feature in metastatic colorectal carcinoma. D-dimer levels should be incorporated into prognostic models, and D-dimer may represent a useful biomarker for patients treated with antiangiogenic agents.

Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer.

PURPOSE: This phase II trial investigated the safety and efficacy of two doses of bevacizumab, a monoclonal antibody to vascular endothelial growth factor, plus fluorouracil (FU)/leucovorin (LV) versus FU/LV alone in patients with metastatic colorectal cancer. PATIENTS AND METHODS: One hundred four previously untreated patients with measurable metastatic colorectal cancer were randomly assigned to one of the following three treatment groups: 36 to FU (500 mg/m(2))/LV (500 mg/m(2)) alone, 35 to FU/LV + low-dose bevacizumab (5 mg/kg every 2 weeks), and 33 to FU/LV + high-dose bevacizumab (10 mg/kg every 2 weeks). FU/LV was given weekly for the first 6 weeks of each 8-week cycle. RESULTS: Compared with the FU/LV control arm, treatment with bevacizumab (at both dose levels) plus FU/LV resulted in higher response rates (control arm, 17%, 95% confidence interval [CI], 7% to 34%; low-dose arm, 40%, 95% CI, 24% to 58%; high-dose arm, 24%, 95% CI, 12% to 43%), longer median time to disease progression (control arm, 5.2 months, 95% CI, 3.5 to 5.6 months; low-dose arm, 9.0 months, 95% CI, 5.8 to 10.9 months; high-dose arm, 7.2 months, 95% CI, 3.8 to 9.2 months), and longer median survival (control arm, 13.8 months; 95% CI, 9.1 to 23.0 months; low-dose arm, 21.5 months, 95% CI, 17.3 to undetermined; high-dose arm, 16.1 months; 95% CI, 11.0 to 20.7 months). After cross-over, two of 22 patients had a partial response to bevacizumab alone. Thrombosis was the most significant adverse event and was fatal in one patient. Hypertension, proteinuria, and epistaxis were other potential safety concerns. CONCLUSION: The encouraging results of this randomized trial support further study of bevacizumab 5 mg/kg plus chemotherapy as first-line therapy for metastatic colorectal cancer.