Presentation on theme: "Overview of Bayesian Methods for Safety Assessment"— Presentation transcript:
1 Overview of Bayesian Methods for Safety Assessment Karen L. Price, PhDEli Lilly and CompanyOn behalf of the DIA Bayesian Scientific Working Group (BSWG)
2 Outline Brief overview of DIA BSWG Overview of use of Bayesian methods for safety assessmentBayesian network meta-analysis with focus in safety dataBayesian methods for safety trialsConclusion
3 Who are we?Group of representatives from Regulatory, Academia, and Industry, engaging in scientific discussion/collaborationfacilitate appropriate use of the Bayesian approachcontribute to progress of Bayesian methodology throughout medical product development
4 MissionTo facilitate the appropriate use of Bayesian methods and contribute to progress by:Creating a scientific forum for the discussion and development of innovative methods and tools.Providing education on best practices for Bayesian methods.Engaging in dialogue with industry leaders, the scientific community, and regulators.Fostering diversity in membership and leadership.
5 Opportunity Statement Bayesian methods provide framework to leverage prior information and data from diverse sources.Bringing together academic, industrial, and regulatory representatives is essential to overcome hurdles.Provides opportunity to influence proactively by engaging in scientific discussion.Improved patient outcomes.
6 Safety Subteam Opportunity/Goals 3 initial areas of focus Current analytical approaches may be oversimplified and knowledge of/experience with proper methods inadequateSome statistical challenges include: power, multiplicity, complexity of data, continual assessment, signal refinementBayes provides great promise3 initial areas of focusMeta-analysis/Evidence Synthesis: chair David OhlssenSafety Trials: chair Karen PriceSignal Detection: chair Larry GouldInitial deliverables: white papers, publications, sessions
7 Some Advantages of Bayesian Methods Ability to incorporate prior informationNatural for evidence synthesis or meta-analysisHandling multiplicity through borrowing strength and hierarchical modelingAppealing in dealing with rare events as the model modulates the extremesAbility to handle complex problems via unified modeling, taking all the uncertainty into accountAllowing direct probability inferences on different scales
8 “Safety assessment is one area where frequentist strategies have been less applicable. Perhaps Bayesian approaches in this area have more promise.” Chi, Hung, and O’Neill; Pharmaceutical Report, “If I were to predict where Bayesian ideas will have great impact in the years ahead I would highlight drug safety – not only during the development of a drug but also post-marketing.” -- Grieve; Pharmaceutical Statistics, 2007
9 Overview of Some Areas of Implementation Safety signal detectionSafety signal evaluationMeta-analysis for analyzing adverse event dataContinuously monitor an event of interest in an ongoing trialJoint modeling for evaluation of safety/efficacy outcomesEstimating the dose-response relationship of adverse eventsMixed treatment comparisons or network meta-analysis for safety dataSafety Trials
10 Screen shot of Pharmaceutical Statistics Special Issue
11 Recent Publications from DIA BSWG Pharmaceutical Statistics Special Issue: Bayesian Methods in Medical Product Development and Regulatory ReviewThe current state of Bayesian methods in medical product development: Survey results and recommendations from the DIA Bayesian Scientific Working Group: Fanni Natanegara, Beat Neuenschwander, John W. Seaman, Nelson Kinnersley, Cory R. Heilmann, David Ohlssen, George RochesterBayesian Methods for Design and Analysis of Safety Trials: Karen Price, H Amy Xia, Mani Lakshminarayanan, David Madigan, David Manner, John Scott, James Stamey, Laura ThompsonGuidance on the implementation and reporting of a drug safety Bayesian network meta-analysis: David Ohlssen, Karen Price, H Amy Xia, Hwanhee Hong, Jouni Kerman, Haoda Fu, George Quartey, Cory Heilmann, Haijun Ma, Bradley CarlinUse of Historical Control Data for Assessing Treatment Effects in Clinical Trials: Kert Viele, Scott Berry, Beat Neuenschwander, Billy Amzal, Fang Chen, Nathan Enas, Brian Hobbs, Joseph G Ibrahim, Nelson Kinnersley, Stacy Lindborg, Sandrine Micallef, Satrajit Roychoudhury, Laura ThompsonTherapeutic Innovation and Regulatory Science, submittedMethods and Issues to Consider for Detection of Safety Signals from Spontaneous Reporting Databases. Report of the DIA Bayesian Safety Signal Detection Working Group. Larry Gould, Ted Lystig, Yun Lu, Haoda Fu, Haijun Ma, and David Madigan
12 Bayesian Network Meta-analysis with focus in Safety data (based on Ohlssen, et al)
13 Network meta-analysis Study 1Study 2Future studyAPLBCPL vs A: BPL vs COf Interest Cvs AAdditional StudiesAC: Active Comparator13
14 MTC : Random Effects Model (taken from NICE DSU documents) kth arm in study Ik=2,..,KRelative treatment effect between 1st arm and kth armFirst arm in study itreatment effect of 1st armConsistency assumptionbetween trial standard deviation
15 Network meta-analysis Trelle et al (2011) Cardiovascular safety of non-steroidal anti-inflammatory drugsPrimary Endpoint was myocardial infarctionData synthesis 31 trials in patients with more than patient years of follow-up were included.A Network random effects meta-analysis were used in the analysisCritical aspect – the assumptions regarding the consistency of evidence across the networkHow reasonable is it to rank and compare treatments with this technique?Trelle, Reichenbach, Wandel, Hildebrand, Tschannen, Villiger, Egger, and Juni. Cardiovascular safety of non-steroidal anti-inflammatory drugs network meta-analysis. BMJ 2011; 342: c7086. Doi: /bmj.c7086
16 b is the control treatment associated with trial i Poisson network meta-analysis model Based on the work of Lu and Ades (LA) (2006 & 2009)μi is the effect of the baseline treatment b in trial i and δibk is the trial-specific treatment effect of treatment k relative to treatment to b (the baseline treatment associated with trial i)Note baseline treatments can vary from trial to trialDifferent choices for µ’s and ’s. They can be: common (over studies), fixed (unconstrained), or “random”Consistency assumptions required among the treatment effectsPrior distributions required to complete the model specificationb is the control treatment associated with trial i
17 Comments on Trelle et al Drug doses could not be considered (data not available)Average duration of exposure was different for different trialsTherefore, ranking of treatments relies on the strong assumption that the risk ratio is constant across time for all treatmentsThe authors conducted extensive sensitivity analysis and the results appeared to be robust
18 Key Aspects of Ohlssen, et al. Summarizes Bayesian network meta-analysisExtends the Lu and Ades (LA) model via a variety of alternative model parameterizationsParticularly in the context of rare events, estimation of model parameters can be challenging for LA modelOutcomes can be particularly sensitive to the choice of model, emphasizing need for sensitivity analysis and transparency regarding assumptions/limitationsHighlights benefit Bayesian approach provides for decision making (including with multiple outcomes)Provides reporting guidelines
19 Two way layout via MAR assumption An alternative way to parameterize proposed by Jones et al (2011) and Piephoetal et al (2012) uses a classical two-way (TW) linear predictor with main effects for treatment and trial.Both papers focus on using the two-way model in the classical framework. By using the MAR property a general approach to implementation in the Bayesian framework can be formedAll studies can in principle contain every arm, but in practice many arms will be missing. As the network meta-analysis model implicitly assume MAR (Lu and Ades; ) a common (though possibly missing) baseline treatment can be assumed for every study (Hong and Carlin; 2012)Statistical approaches for conducting networkmeta-analysis in drug development†Byron Jones,a* James Roger,b PeterW. Lane,c Andy Lawton,d ChrissieFletcher,e Joseph C. Cappelleri,f Helen Tate,g PatrickMoneuse,h and onbehalf of PSI Health Technology Special Interest Group, EvidenceSynthesis sub-teamModeling between-trial variance structure in mixedtreatment comparisonsGUOBING LU∗, AE ADESBiostatistics (2009), 10, 4, pp. 792–805doi: /biostatistics/kxp032The Use of Two-Way Linear Mixed Models in MultitreatmentMeta-AnalysisH. P. Piepho,1,∗ E. R. Williams,2 and L. V. Madden3Pharmaceut. Statist. 2011, –531DIA
20 Reporting GuidelinesOhlssen et al provides a checklist for use when conducting a safety meta-analysisChecklist includes four main sections: Introduction, Methods, Results, and Interpretation.Each main section includes various items relevant to that sectionThe user of the table should evaluate each item and can utilize the last two columns to confirm whether or not each item has been addressed and to add any relevant commentsThe Introduction section includes items that are specific to the intervention being analyzed along with the meta-analysis objective(s), while the Methods section focuses on aspects associated with the planned modeling and analyses to be conducted. The Results section relates to specific summaries of the data that should be provided. Finally, the Interpretation section enables decision makers utilizing the meta-analyses to appropriately evaluate the results.
21 Bayesian methods for design and analysis of safety trials (based on price, et al)
22 Overview of Paper Reviews challenges associated with safety trials Describes several opportunities for use of Bayesian methods to enhance safety trialsDiscusses several case examples
23 Recommendations: Overview of Bayesian Opportunities for Safety Trials OpportunityKey References Bayesian methods to determine sample sizeAdcock; Wang and Gelfand; Brutti, De Santis, and Gubbiotti; Gaydos et al. Frequent interim analysesConnor and White et al. Bayesian Meta-analysisSpiegelhalter et al.;Stangl and Berry;Sutton et al.
24 Recommendations: Overview of Bayesian Opportunities for Safety Trials OpportunityKey References Sequential meta-analysisCheng and Madigan;Higgins, Whitehead, Simmonds; Ibrahim et al.;Zeggini and Ioannidis  Borrowing historical informationBerry et al.;Hobbs et al. Continuous monitoring of eventsXia et al.;Yao et al. Hierarchical modelingGelman and Hill;Gelman et al.;DuMouchel
25 Recommendations: Overview of Bayesian Opportunities for Safety Trials OpportunityKey References Post approval studies/Surveillance studiesFDA Guidance;Murray, Carlin, and Lystig Logistical planning related to enrollment rates and landmark event rateGajewski, Simon, and Carlson; Bagiella and Heitjan;Ying and Heitjan;Donovan, Elliott, and Heitjan Bayesian interpretations and predictionsSpiegelhalter;Berry et al.
26 Case Example: Sequential Monitoring of AEs Sequential Bayesian methods enable regular updating of knowledge as data accumulateCheng and Madigan illustrated this approach with VioxxPresented a Bayesian sequential meta-analysis of the placebo-controlled trialsThe analysis began with a “family of priors”Proposed a simple graphical summary of the meta-analysis showing the posterior probability over time that the true relative risk of CVT events exceeds two particular thresholdsThe following figure shows the posterior probability that the true relative risk exceeds 1.1 over timePriors ranged from a “skeptical prior” that represents a priori skepticism that Vioxx causes cardiovascular thrombotic (CVT) adverse events, to a “reference prior” the adopts a neutral position, to a “cautious prior” that represents a prior belief that cardiotoxicity is not so implausible.The specific events are acute myocardial infarction, unstable angina pectoris, sudden and/or unexplained death, resuscitated cardiac arrest, cardiac thrombus, pulmonary embolism, peripheral arterial thrombosis, peripheral venous thrombosis, ischemic cerebrovascular stroke, stroke (unknown mechanism), cerebrovascular venous thrombosis, and transient ischemic attack.
27 Case Example: Sequential Monitoring of AEs, cont.
28 Moving ForwardSafety Meta-analysis guidance from FDA (draft published, opportunity to comment)Continued growth in use for signal assessmentOpportunities for increased use for safety trialsExpanded use for evaluation of benefit/risk profile (at least for key benefits/risks)
29 ConclusionSafety assessment is complex with numerous statistical challengesDIA BSWG is actively working to ensure the use of Bayesian methods in the context of safety are appropriately used by increasing awareness and providing best practice guidelinesBayesian methods provide advantages in the context of safety signal assessment