Comparing Multiple Treatments for Ebola Virus Disease A Discussion of Ethical Ramifications Scott Berry 28 Nov 2017 1 2 Classical Trial Disease Control 3 Treatment Platform Trial perpetual ID: 1037601
Download Presentation The PPT/PDF document "A Randomized Adaptive Platform Trial" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. A Randomized Adaptive Platform Trial Comparing Multiple Treatments for Ebola Virus Disease: A Discussion of Ethical RamificationsScott Berry28 Nov 20171
2. 2
3. Classical TrialDiseaseControl3Treatment
4. Platform Trial (perpetual)DiseaseControl4Tx 1Does the tx 1 work in the disease?Does tx 1 work better than tx 2?What is the best way to treat the disease?Tx 2Tx 3Tx M
5. There is one ProtocolThe master protocol assigns patientsProtocol is disease focusedNo treatment names in protocolTreatment arm appendicesPotentially evolving armsAdaptive Platform Trial Design: Master ProtocolTx 1Tx 2Tx 3Tx 4Master Protocol5
6. EV-003 Adaptive Platform DesignReviewed and approved by:Duke University IRBUniversity of Sierra Leone ethics committeeMaster Protocol dictates trial behavior, each treatment included as an appendixMultiple AgentsSOC?Primary & Secondary agentsCombination + Single agentsResponse Adaptive Randomization (RAR)Run by a single algorithmAssigns treatment regimens that are performing better using collection of primary endpoint dataProtocol is built so trial arms evolve (part of the protocol!), trial is perpetualEndpoint is 14-day mortality6Anti-virals, Sunitinib, Erlotinib, statin, Irbesartan, Azithromycin…
7. EV-003 Adaptive Platform DesignAnalyzeAvailable DataAccrueMoreBurn-In EnrollmentRemove Agent?Revise AllocationRulesNo7Analyze (report)ResultsYesAdd Agents
8. EV-003 Adaptive Platform DesignAnalyzeAvailable DataAccrueMoreBurn-In EnrollmentRemove Agent?Revise AllocationRulesNo8Analyze (report)ResultsYesAdd Agents
9. Example Trial9RegimensAgents1234Agents1234
10. 10
11. 11
12. 12
13. 13
14. 14
15. 15
16. 16
17. 17
18. 18
19. 19
20. 20
21. 21
22. 22
23. 23
24. 24
25. 25
26. 26
27. 27
28. 28
29. 29
30. 30
31. 31
32. 32
33. 33
34. 34
35. 35
36. 36
37. Scenario 3DesignMean DeathsAdapt49.4Fixed69.9ABCDABCD1234123437TruthMean NProb Wins(fixed)Mean N& Fails
38. FDA Interaction: 3-Arm ExampleNot for evaluation of a single drug for FDA/EMA approval as originally designedModifying for regulatory approval with FDA input3 arms (SOC plus treatment) + SOC aloneSOC vs. A, B, C [+SOC]Same trial design!Modifying “success over SOC” probability Was 95%... Use 99.8% for the following…38
39. Null39ArmTrueMean N> SOC> ALLSOC0.4062.5A0.4061.70.0250.013B0.4061.70.0250.013C0.4061.70.0250.013Total0.40247.60.0630.037ArmTrueN>SOCSOC0.4042A0.40420.025ArmTrueN>SOCSOC0.4042B0.40420.025ArmTrueN>SOCSOC0.4042C0.40420.025ArmDeathsN>SOC> ALLTotal0.402520.060?
40. One Nugget40ArmTrueMean N> SOC> ALLSOC0.4029.4A0.4021.60.0040.003B0.4021.60.0030.003C0.1581.10.9700.948Total0.268153.70.9700.948ArmTrueN>SOCSOC0.4042A0.40420.025ArmTrueN>SOCSOC0.4042B0.40420.025ArmTrueN>SOCSOC0.4042C0.15420.71ArmDeathsN>SOC> ALLTotal0.3582520.724?
41. Varying Good41ArmTrueMean N> SOC> ALLSOC0.4026.9A0.3025.00.0170.000B0.2046.90.2020.005C0.1089.60.9510.664Total0.196188.40.9700.969ArmTrueN>SOCSOC0.4042A0.30420.147ArmTrueN>SOCSOC0.4042B0.20420.490ArmTrueN>SOCSOC0.4042C0.10420.892ArmDeathsN>SOC> ALLTotal0.3002520.953?
42. Ethical RamificationsIn a pandemic (LMIC) the struggle over SOC/PBO will be recurringTrial design functions with or without, and removes as soon as answer is knownIn multi-arm setting appropriate RAR improves inference and improves careImproves global ability to find effective therapies among the ‘universe’ of therapiesEthical considerations rarely address type III errorsAre 1:1 trials in a pandemic situation ethical?Complexity trade-off with improved inference + treatmentSpeed of FPI? Speed of 2nd, 3rd arm in…42