1 / 28

Evaluating Change in Hazard in Clinical Trials With Time-to-Event Safety Endpoints

Evaluating Change in Hazard in Clinical Trials With Time-to-Event Safety Endpoints. Rafia Bhore, PhD Statistical Scientist, Novartis Email: rafia.bhore@novartis.com Midwest Biopharmaceutical Statistics Workshop Muncie, Indiana May 21, 2013. Outline. Motivation Metrics of risk

opa
Download Presentation

Evaluating Change in Hazard in Clinical Trials With Time-to-Event Safety Endpoints

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Evaluating Changein Hazard in Clinical Trials With Time-to-Event Safety Endpoints Rafia Bhore, PhD Statistical Scientist, Novartis Email: rafia.bhore@novartis.com Midwest Biopharmaceutical Statistics Workshop Muncie, Indiana May 21, 2013

  2. Outline • Motivation • Metrics of risk • Time-dependency of adverse events • Change-point methodology | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  3. Motivation | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  4. US FDA Regulations FDA regulations created from these laws • Federal Food and Drug Cosmetic (FD&C) Act (1938) • submit evidence of safety to the FDA • Kefauver-Harris Amendments (1962) • Strengthened rules for drug safety • In addition to safety, effectiveness of drug needs to be demonstrated • Food and Drug Administration Amendments Act (FDAAA) (2007) • Enhanced authority on monitoring safety • FDA Safety and Innovation Act (FDASIA) (2012) • Better adapt to truly global supply chain (Chinese and Indian drug suppliers) Safety – an older/consistent regulatory requirement | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  5. Why quantitative methods for evaluation of safety? • Safety evaluation required by regulators • Extensive collection of safety data • E.g., extensive safety data collected in new application (NDA/BLA/PMA) packages comprising several clinical trials • Abundance of descriptive safety analyses • Surprises in post-hoc review of safety data • Descriptive analyses not adequate. No planned inferential analyses. • Top reason why new applications for drugs/biologics/devices go to FDA Advisory Panels • Understand risk of “major” events | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  6. Metrics of risk | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  7. Metrics of Risk • Crude rates • Exposure-adjusted rates • Occurrences (events) per unit time of exposure (akaexposure-adjusted event rate) • Incidences (subjects) per unit time of exposure (akaexposure-adjusted incidence rate) • Cumulative rates • Life table method or Kaplan-Meier method • Hazard rates and functions • Instantaneous measure of risk • Similar to cumulative rates • constant, decreasing, or increasing | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  8. Different Metrics of Risk An overview | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  9. Time-dependency of adverse events | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  10. Drug Exposure vs. Adverse Event Rates Three patterns of AEs – O’Neill, 1988 CUMULATIVE | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  11. Time-to-event Endpoints Time-to-event endpoint is a measure of time for an event from start of treatment until time that event occurs • Safety Outcomes • Invasive breast cancer in Women’s Health Study • CV Thrombotic Events in a large clinical trial • Safety Signals detected through biochemical markers, • Change in grade of Liver Function Tests • Abnormalities in serum creatinine and phosphorus • Abnormal elevations in other lab tests • Efficacy Outcomes • Time-to-Relapse, Overall survival (SCLC), Cessation of Pain (Post-herpetic neuralgia) | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  12. Increased risk of Invasive Breast Cancer? Women’s Health Initiative Study on Estrogen Plus Progestin (JAMA 2002) | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  13. Increased risk of Cardiovascular Thrombotic events? FDA Advisory Committee Meeting – Li, 2001New England Journal of Medicine – Lagakos, 2006 Study 1 Study 2 | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  14. Change-Point Methodology A tool to test and estimate for change in risk | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  15. Definition of the Problem • Risk abruptly changes over time • Define risk using time-to-event outcome • Is there a change in hazard? • Is this statistically significant? • What is the estimated time of change? (aka CHANGE-POINT) Change-point is defined as the time point at which an abrupt change occurs in the risk/benefit due to a treatment | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  16. Change-point models for hazard function • Let (Ti , i)be the observed data (time & censoring variable) with hazard function h(t)and survival function S(t) • Assume hazard is constant piecewise in k intervals of time • Total of k hazard rates l1,...,lkand (k-1) change points t1,...,tk-1 K-piece Piecewise Exponential Two-piece Piecewise Exponential Exponential Model | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  17. Estimation or Hypothesis Testing? Which comes first? (Chicken or Egg) Two-piece Piecewise Exponential Model • Test hypothesis of no change point, H0 ,vs. H1 of one change point. • We can expand statistical methods to more than one change-point • Estimation (Point and 95% Confidence Interval/Region) • Estimate where the change point(s) occurs | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  18. Estimation of hazard rates Known change point • Log likelihood functions for exponential and 2-piece PWE • Maximum likelihood estimates of hazard rates, l’s, given t • Generalized to k (>2) change points (Bhore, Huque 2009) | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  19. Estimation of hazard rates Unknown change point • In real clinical data, change points are unknown • Consider log likelihood functions for 2-piece PWE • Estimate t using a grid search that maximizes profile log likelihood • Substitute MLE of hazard rates into log L and maximize log Lwrtt over a restricted interval [ta,tb]. | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  20. Confidence region/interval for change-point, t • An approximate confidence region for the change point, t, was given by Loader (1991). • Underlying likelihood function is not a smooth function of t. Hence confidence region may be a union of disjoint intervals. • Gardner (2007) developed an efficient parametric bootstrap algorithm to estimate the confidence interval. | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  21. Simulated example of Change-Point λ2= 5 2.5 Change-point? 1.5 1 λ1 = 1 | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  22. Estimation of change-point Simulation example E.g. Result: Change in hazard is estimated to occur at 0.81 units of time (95% CI: 0.64 to 0.99 units of time) | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  23. Testing of Change Points Likelihood Ratio Test (2-piece PWE) • One would think that LRT statistic has χ2 distribution with two degrees of freedom. Not true because of discontinuity at change-point • See Bhore, Huque (2009), Gardner (2007) & Loader (1991) for details on computing significance level | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  24. Goodness-of-fit: Selecting correct CP model Hammerstrom, Bhore, Huque (2006 JSM, 2007 ENAR) Consider 6 time-to-event models • Exponential (constant hazard) • Two-piece PWE with decreasing hazard • Two-piece PWE with increasing hazard • Three-piece PWE with V shape • Three-piece PWE with upside down V shape • Weibull | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  25. Simulation criteria for data True underlying models for change-point Sample size, N = 150 or 40 subjects • 2-piece Piecewise Exponential (15 models) • λ1 = 1 • λ2 = 0.2, 0.5, 1, 2, 5 • Change point,  = 30th, 50th, 70th percentile of λ1 • 3-piece Piecewise Exponential (9 models) • Early:Mid:Late hazard rates = 0.25:1:0.3 or 2:1:2 • Change point,  = 20th:50th, 20th:70th, or 50th:20th percentiles of early and middle hazards • Weibull (25 models) • Shape = 0.25, 0.5, 1, 2, 5 and Scale = 0.5, 2, 3, 3.5, 4 | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  26. True model: 2-piece Piecewise Exponential (N=150) Pairwise comparison of models 2= | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  27. True model: 2-piece Piecewise Exponential (N=40) Pairwise comparison of models 2= | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

  28. Concluding Remarks • Uncontrolled or open-label Phase II/III clinical trials provide a major source of long-term safety/efficacy data for a single group. • Crude incidence rates underestimate the incidence of delayed events • Visual check of Kaplan-Meier curves are not sufficient to detect change in hazard • Change-point methodology (new in application to clinical trials) can be applied to test whether and estimate where a change in hazard occurs. • Piecewise exponential model is robust for modeling change in hazard (Bhore and Huque 2009). • Percentile bootstrap preferred for computing CIs (work not shown) | Change in Hazard | Rafia Bhore | 21 May 2013 | Midwest Biopharmaceutical Statistics Workshop

More Related