1. Thierry Le Chevalier, MD How do International Regulatory Groups View Clinical Trials and Outcomes Measures?

2. Outline Brief overview of regulatory authorities and their approval processes Discuss their perspectives on clinical trial endpoints in drug approval

3. What Is the Issue? A clinical trial is designed to evaluate (Phase II) and compare (Phase III) a drug/drug combination/therapeutic modality with existing standard treatments A registration-driven study will mainly focus on the drug activity on the disease, while an academic-driven study will mainly focus on the overall effect of the treatment on the patient’s life (duration and QoL) QoL = quality of life.

4. General Introduction and Background How has the field of drug use, manufacture and development evolved over the centuries? To understand the current landscape of all aspects of medicine regulation, we need to know how the story has developed, and country-specific issues To understand how to build a registration-driven study, we need to know the rules To be the patients’ defenders, we need to evaluate the impact of the protocol constraints on the patients This is not essentially “new” but has really come into its own in the past few decades

5. Regulatory Mission Quality, safety and efficacy Protection for users of medicines Adequate and appropriate information for patients and physicians To protect and promote public health

6. Harmonisation The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use is unique in bringing together the regulatory authorities and pharmaceutical industries of Europe, Japan and the US to discuss scientific and technical aspects of drug registration

7. Brief Overview of Regulatory Authorities and Their Approval Processes: FDA, EMA, and MHLW/PMDA

8. FDA: Drivers for Regulation in the 20th Century The Pure Food and Drug Act of 1906 in the US, driven fundamentally by concerns in food industry Harrison Narcotics Tax Act in 1914 Elixir sulfanilamide 1937 Thalidomide 1961 This followed on from a public health mission that began in 1862

9. FDA Organisation http://www.fda.gov/AboutFDA/CentersOffices/OrganisationCharts/ucm393155.htm. Accessed March 04, 2015.

10. FDA: Scope of Products Regulated Drugs (prescription and over-the-counter) Biologics –Vaccines –Blood products –Biotechnology products and gene therapy Medical devices Veterinary products Food Consumer and medical radiation products Cosmetics

11. Seeking Approval: “Menu” of Choice Depending on the drug, disease, and business decision, consider: Type of approval “Regular” approval Accelerated approval Orphan drug (accelerated vs. regular approval) Filing/review type Fast-track designation Priority review Standard review New molecular entity “Me too”/ generics

12. Licensing Approval Scenarios Full data package with sufficient data to demonstrate –Safety and efficacy –Clinical benefit Intended to make promising products for life-threatening diseases available on the basis of preliminary evidence prior to formal demonstration of patient benefit Approval based on a surrogate endpoint Considered provisional approval with a written commitment to complete clinical studies Regular approval on completion of Phase III clinical studies Accelerated approval

13. Filing Types A process designed to facilitate the development and expedite the review of new drugs that are intended to treat serious or life-threatening conditions to fill an unmet medical need Benefits include: –“Rolling submission” for a marketing application –More-frequent meetings with the FDA to discuss development plans and ensure collection of appropriate data needed to support approval Offers major advances in treatment or provides a treatment where no adequate therapy exists Goal for completing a priority review is 6 months (vs standard of 10 months) Request for priority review is made by the sponsor at the time of the new drug application filing Fast track Priority review

14. FDA’s Process Overview Center for Drug Evaluation and Research Courtesy of Thierry Le Chevalier, MD

15. Directives Shaped Regulatory Processes in EU In the UK the whole system was reshaped –Committee on Safety of Medicines in 1963 –Yellow Card scheme in 1964 –EEC Directive 65/65/EEC – law, regulation and administrative action relating to medicinal products Harmonisation took 10 years to develop –Two directives in 1975 (75/318/EEC and 75/319/EEC) –87/22/EEC – introduced a centralised procedure and paved the way for Council Regulation (EEC/2309/93 EMEA) and re-established CHMP to formulate opinion for the agency EEC = European Economic Community Clinical Trial Directive; EMA = European Medicines Agency; CHMP = Committee for Medicinal Products for Human Use.

16. EMA CHMP Committee for Orphan Medicinal Products Committee for Oncology Developments, Advanced Therapies and Paediatrics Coordination of drug licensing in the EU Committees including: Centralised vs decentralised (or mutual recognition) systems of approval Proposed legal basis of applications (full or full-mixed) Generic, hybrid or similar biological applications The role of the Scientific Advisory Committee Approval under exceptional circumstances Conditional marketing authorisation EMA = European Medicines Agency; CHMP = Committee for Medicinal Products for Human Use.

17. EMA/CHMP Organisation HIV/viral disease Anti-infectives (not HIV) Cardiovascular Central nervous system Diabetes and endocrinology Diagnostics Oncology Paediatric WP CHMP Safety WP Joint quality WP Pharmacovigilance WP Biological WP Blood products WP Vaccines WP WP on similar biological medicinal products Cell-therapy WP Gene therapy WP – Cell-based therapy WP Pharmacogenetics WP Scientific advice WP EMA = European Medicines Agency; CHMP = Committee for Medicinal Products for Human Use; WP = working party Quality Review of Documents.

18. EMA: Centralised Approval Procedure Company files application EMA = European Medicines Agency; CPMP = Committee for Proprietary Medicinal Products; EPAR = European Public Assessment Report. Comments from member states EMACPMP Positive opinion issued EU commission EPAR and marketing authorisation Rapporteurs Launch 28 days 210 days

19. EMA: Licensing Approval Scenarios Full data package with sufficient data to demonstrate –Safety and efficacy –Clinical benefit Rarely used by EMA Regular approval on completion of Phase III clinical studies Accelerated assessment procedure

20. Drug Approval Process: EMA vs FDA Source: CDER 21st Century Review Process (www.fda.gov); User Guide for Micro, Small and Medium-sized Enterprises (www.ema.europa.eu). Accessed March 04, 2015. EMA FDA

21. Regulatory History in Japan* First regulation in Japan for the clinical development of anticancer drugs issued in 1991 Results from phase 2 trials using RR were accepted Phase 3 trials needed to be conducted post-approval –Required only the protocol to be submitted at the time of approval However, almost no phase 3 trials were conducted after approval Guidelines revised in 2006 Results from phase 3 studies with survival data is mandatory for major cancers (NSCLC, GC, CRC, BC) Japanese regulatory authorities believe clinical data from foreign patients to be of limited applicability to Japanese patients and require a stand-alone development program *Specific to oncology. Maeda H et al. Ann Oncol. 2015;26:211-6.

22. Regulatory Authorities in Japan MHLW Official: http://www.mhlw.go.jp/english/policy/health-medical/pharmaceuticals/orphan_drug.html. Accessed March 04, 2015. Ministry of Health, Labour and Welfare (MHLW) Pharmaceuticals and Medical Devices Agency (PMDA) National Institute of Biomedical Innovation (NIBIO)

23. Regulatory Authority in Japan: Functions of PMDA PMDA Official: http://www.pmda.go.jp/english/service/outline_s.html. Accessed March 04, 2015. Pmda Its Key Services

24. MHLW/PMDA: Approval Process PMDA Official: http://www.pmda.go.jp/english/service/outline_s.html. Accessed March 04, 2015.

25. Duration of PMDA Review Process PMDA Official: http://www.pmda.go.jp/english/service/outline_s.html. Accessed March 04, 2015.

26. Regulation in Japan: Accelerated Approval and Surrogate Endpoints No formal regulation for accelerated approval program in Japan Priority review for orphan drug in patients with life-threatening disease Orphan-drug designations Public knowledge-based application Special committee on unapproved drugs The regulatory agencies in Japan accept surrogate endpoints as endpoints supporting approval besides survival, but there are no specific guidelines Accelerated approval Expedited programs Maeda H et al. Ann Oncol. 2015;26:211-6. Tanimoto T et al. Invest New Drugs. 2013;31:473-8. PMDA Official: http://www.pmda.go.jp/english/service/outline_s.html. Accessed March 04, 2015. Surrogate endpoints

27. 27 Drug Approvals: EMA vs FDA vs PMDA Application characteristics EU (EMA)USA (FDA)Japan (PMDA) Number approvals NME a 282724 NBE rate14%19%12.5% OD rate b 62%52%37.5% Expedited review rate7%81%42% Expedited approval rate31%33%na ‘First-at-all-approved’ rate18 c 96% d 4% e ‘First-at-all-submitted’ rate18% f 89% g 8% h Approval and review times Median approval time13.3 mts (95% CI: 12.5–14.4 mts)6.0 mts (95% CI: 5.9–8.8 mts)15.0 mts (95% CI: 12.0–18.3 mts) Median active review time6.6 mts (95% CI: 6.5–6.7 mts)6.0 mts (95% CI: 6.0–8.9 mts)nd Median clock-stop time4.2 mts (95% CI: 2.9–4.9 mts)nand Median administrative time2.1 mts (95% CI: 2.0–2.7 mts)na Median ‘Time-to-submission’ delay j –1.7 mts (95% CI: –0.9– –3.2 mts)14.9 mts (95% CI: 6.29–24.1 mts) Median ‘Time-to-market’ delay j –7.2 mts (95% CI: –4.7– –15.0 mts)25.1 mts (95% CI: 6.2–34.1 mts) Expedited approval/review timesExpedited vs normal Median approval time EA 14.6 mts vs 12.8 mts  = -1.8 mts6.0 mts vs 7.7 mts  = -1.7 mts nanana Median approval time ER 9.0 mts vs 13.6 mts  = 4.6 mts6.0 mts vs 9.8 mts  = 3.8 mts14.2 mts vs 15.2 mts  = 1.0 mts  (delta) = time difference; EA = expedited approval (EU: approval under exceptional circumstances and conditional marketing authorization; USA: accelerated approval); ER = expedited review (EU: accelerated assessment; USA: priority review; Japan: priority review and/or expedited review); mts = months; na = not applicable; NBE = new biological entity; nd = no data; NME = new molecular entities; OD = orphan designation. a Eight NME were simultaneously approved between January 2006 and December 2011 in all three regions. NME are defined as active ingredients, never before been marketed in the respective territory in any form; b OD prevalence criteria (cases/inhabitants): EU <5/10,000, US <7.5/10,000, Japan <4/10,000; c First approved in EU: catumaxumab, histamine, mifarmutide, trabectidin, vinflunine; d Only bendamustine had been previously approved elsewhere (EU); e Miriplatin, approved in Japan only so far; f Catumaxumab, histamine, ipilimumab, trabectidin, vinflunine; g Only bendamustine, decitabine, and ipilimumab had been submitted first in the EU; h Miriplatin and eribulin; i First review cycle only; j In comparison to EU based on paired observations, ie, calculated from dossier submission and marketing authorization dates of those drugs which were approved in both regions. Maeda H et al. Ann Oncol. 2015;26:211-216.

28. Differences in Drug Approval: FDA vs EMA vs MHLW/PMDA Tanimoto T et al. Invest New Drugs. 2013;31:473-8. Example: Gemtuzumab ozogamicin in AML

29. Regulatory Authorities’ Perspectives on Clinical Trial Endpoints

30. Clinical Trial Endpoint Selection in Last Two Decades Strict legal requirements to demonstrate benefit –Randomised controlled trials (how many are necessary?) –Primary endpoint: Valid and reliable measure that provides the most clinically relevant and convincing evidence –Wrong design or lack of efficacy the most important reason for rejection The commonly used endpoints are based on –Survival: OS then PFS –Tumour response –Symptom assessment Advantages and disadvantages OS = overall survival; PFS = progression-free survival.

31. Clinical Trial Endpoint Selection in Last Two Decades OS: Historically viewed as the most effective way, as it addresses biology of tumour and the natural history of the disease PFS: Progression is associated with tumour growth, assesses tumour shrinkage and stabilisation of disease

32. 32 Clinical Trial Endpoint Selection: EMA *Variation of established drugs. Eur J Clin Pharmacol. 2002;58(9):573-580. DesignEndpointnReason for Accepting Design Phase IIRR PFS 18 2 Outstanding activity AND No established treatments Phase III RCT RR PFS OS 4* 14 9 DesignEndpointnReason for Rejection Phase IIRR5 Non randomised AND No outstanding activity Phase IIIRR OS 6262 Low level of response Inadequate control Target/size of population Dose justification No effect Wrong comparator Dose justification Design of Pivotal Trials (N=47 approved indications) Rejected/Withdrawn Indications (N=13) RR: 22 (47%) PFS: 16 (34%)OS: 9 (19%)

33. Clinical Trial Endpoint Selection: FDA Primary endpoints for new molecular entities Proportion of clinical studies used to support approval using various endpoints AcceleratedRegular Response rate93%53% Survival0%12% Time to progression7%20% Symptom benefit0%12% Other7%32% Note: Totals are not 100% due to multiple endpoints. S. Hirschfeld, presentation to the CBER Office of Cellular Tissue and Gene Therapy seminar on November 16. Talarico et al. ASCO 2005.

34. 34 Clinical Trial Endpoint Selection: Japan Changes in endpoints supporting approval for oncology drugs before and after the revision of the guideline for oncology drugs in Japan in 2006 a Between the approval of drugs from Arpil 2001 and new drug application by March 2006. b Between new drug application from April 2006 and the approval of drugs by April 2014. Maeda H et al. Ann Oncol. 2015;26:211-216. Endpoint Before the guideline revision a After the guideline revision b N%N% Survival26.32934.9 Response rate2681.32833.7 Progression-free survival00.01518.1 Time to progression26.356.0 Disease-free survival13.122.4 Quality of life00.00 Patient-reported outcomes00.00 Other end points13.144.8 Total32100.083100.0 Evolution in approval endpoints: –↑ OS and PFS –↓ RR –↓ Surrogate endpoints, except PFS

35. Clinical Trial Endpoint Selection Phase III, confirmatory studies (Revision 3) –Interim analyses/data maturity –OS as primary endpoint, not response rate –Possible: PFS when clinically relevant, symptom control Surrogate endpoints –For accelerated approval: must predict clinical benefit over available therapy Overall response rate has been the most commonly used surrogate endpoint in accelerated approval

36. Evolving Landscape in NSCLC Issues: –Pros and cons of OS vs PFS –Patient-related outcomes How do you evaluate clinical benefit? Requires a broader and more balanced view Use of biomarkers in targeting treatment NSCLC = non–small cell lung cancer. Inclusion of biomarkers/companion diagnostics in the label –Erlotinib –Gefitinib –Afatinib

37. Some General Rules and Observations No one size fits all solution –The applicability of PFS is a case-by-case assessment –It varies from health authority to health authority, according to tumour type and even the line of therapy Regulatory authorities increasingly conservative on use of surrogate markers in oncology Q:Has there been a paradigm shift at the FDA on employing surrogates for regulatory approval? A:(Reviewer, Division Hematology Products, FDA): There is an increasing reluctance to rely on surrogates as they often fail to capture significant safety issues, which might impact the ultimate clinical outcome

38. Some Factors Determining Acceptance of PFS Acceptance determined by: Degree of validation of PFS as surrogate for clinical benefit, or, extent of understanding that PFS represents a clinical benefit in itself Tumour type Stage of disease and/or line of therapy (including availability of subsequent therapies) Trial design (e.g. blinded vs open-label) Data quality (including independent review) Proven benefits of comparator therapy Magnitude of effectToxicity of treatment

39. Some Recent Observations and Future Trends PFS is proving more challenging to employ as a regulatory endpoint –However, it will continue to have a future potential role in oncology drug registration if rigorous acceptance criteria and standards are met There will be increasing regulatory pressure to link or associate PFS benefits with other clinical trial outcomes that show direct clinical benefit (e.g. QoL benefits, disease-related symptom benefits, OS positive trends) –PFS may have its best future applications in symptomatic disease settings and/or where delay in disease progression correlates with delay in symptom onset –Benefit of delay of progression may also be measured by assessing QoL benefits pre- vs post-progression, independent of study arm

40. What Are the Differences Between the FDA and EMA Regarding PFS as Primary Endpoint for Registration? FDA is typically more stringent and critical of PFS than EMA in first-line or early metastatic disease settings –Various examples of products in first-/early line settings with easier EU registration, but more difficult US registration Conversely, the FDA is often more accepting of PFS in later-line settings –Delay of disease progression, particularly in symptomatic settings, often seen as a direct clinical benefit in itself –Refractory patients are "a more symptomatic patient population. Hence, a delay in progression in a symptomatic population probably has a little more clinical meaning than simply a radiographic delay in an asymptomatic patient population." R. Pazdur, AVASTIN ODAC Dec 2007 –Various examples of products in later-line settings with easier US registration but more difficult EU registration

41. Metastatic Breast Cancer (mBC) Example: Bevacizumab in the First-line Setting Substance and indicationFDAEMA Bevacizumab in first-line mBC (E2100; add-on to paclitaxel vs paclitaxel, PFS primary endpoint, open-label, cooperative study) Highly complex and difficult regulatory background in US: sBLA withdrawn Sept 2006 Resubmitted Aug 2007 Negative ODAC Dec 2007 Accelerated (provisional) approval Feb 2008 for HER2– AVADO and RIBBON-1 confirmatory studies failed to replicate PFS finding of E2100 Negative ODAC Jul 2010 FDA decision to withdraw mBC indication Dec 2010 Roche currently appealing decision and proposing new biomarker-based study Relatively easily registered in EU (Mar 2007) for HER2+ and HER2– Docetaxel combination registered in Jul 2009 EMA reconfirmed registration of paclitaxel combination in Dec 2010 (but revoked docetaxel, and refused capecitabine combinations) mBC = metastatic breast cancer; sBLA= supplementary biological license application.

42. Further Examples in mBC: Lapatinib and Ixabepilone in the Third-/Fourth-line Setting Substance and indicationFDAEMA Lapatinib for treatment of HER2+ mBC patients after previous anthracycline, taxane and trastuzumab therapy 6-month priority review (full approval Mar 2007) Initial CHMP recommendation after 14-month review. EU Commission requested reassessment. CHMP re-evaluation, with increased liver toxicity warnings. EU Conditional Approval Jul 2008 Ixabepilone for treatment, either as monotherapy or capecitabine combination therapy, in patients failing taxane and anthracycline therapy 6-month priority review (full approval Oct 2007) Marketing authorisation approval withdrawn Jan 2009 after initial rejection by CHMP; "very small increase in survival did not outweigh the concerns over the medicine's safety"

43. mBC Example: Lapatinib in the First-line Setting In contrast to third-line setting, first-line approval was easier in the EU Substance and indicationFDAEMA Lapatinib for combination therapy with letrozole in HER2+ breast cancer Jan 2010 accelerated approval (provisional) Significant post-approval commitments, including OS study of lapatinib/letrozole/ trastuzumab vs letrozole/ trastuzumab Apr 2010 full authorisation

44. Sometimes OS Does Not Suffice…Cetuximab in NSCLC (FLEX Study) Rejected by CHMP in Nov 2009 (after appeal against initial rejection) –Small OS benefit not balanced by significant increase in toxicity, plus negative impact on PFS –Company tried to identify positive subgroup of patients aged <65 years but rejected by CHMP (retrospective analysis, concerns over multiplicity)

45. Can We Observe Any Trends? EMA Example: mBC PFS benefits in pivotal studies in mBC: All lines of therapy (CHMP) PFS primary endpoint unless stated otherwise Courtesy of Thierry Le Chevalier, MD.

46. Can We Observe Any Trends? FDA Example: mBC PFS benefits in pivotal studies in mBC: All lines of therapy (FDA) PFS primary endpoint unless stated otherwise Courtesy of Thierry Le Chevalier, MD.

47. What Did the Previous Slides Conclude? Efficacy Conclusions: CHMP The hurdle to gain approval via PFS has increased (compare 2.9 vs1.9 months of original full approval of bevacizumab + docetaxel in AVADO in Q3 2009) It is not only necessary to have a significant absolute PFS delta, but it is also important to show a significant relative improvement vs active comparator; fold-improvement in RIBBON-1 only 1.51, compared with two for trastuzumab + anastrozole in TAnDEM or bevacizumab + paclitaxel in E2100 Lapatinib might have been rejected in current climate PFS differences of 5–6 months and/or an improvement of PFS by factor of two appear to be good determinants of a full approval PFS benefits of 2 months or less very unlikely to receive approval (see revoked bevacizumab indication based on 1.9 months in AVADO study) The failure of RIBBON-1 with a PFS delta of a robust 2.9 months may suggest that: Conditional approval of lapatinib + capecitabine in second-line setting with similar data to ixabepilone may reflect more favourable safety profile of lapatinib/biomarker- selected population

48. What Did the Previous Slides Conclude? Efficacy Conclusions: FDA Gaining approval in a first-line setting is more difficult compared with second- or third-line (compare with ixabepilone and lapatinib in third- line mBC) Not only necessary to have significant absolute PFS delta, but also important to demonstrate significant relative improvement vs active comparator; improvement in RIBBON-1 was only 1.51-fold Safety profile of test therapy, plus demonstration of other measures of clinical benefit, have crucial role to play in overall assessment PFS differences of 5–6 months and/or an improvement of PFS by a factor of two appear to be good determinants of an accelerated approval unless statistically significant OS differences can be shown (see eribulin) The failure of RIBBON-1 with a PFS delta of 2.9 months shows that:

49. Regulatory Perspective on Symptom Assessment Endpoints Symptomatic improvement is considered a clinical benefit However, measures of global health-related QoL have not served as primary efficacy endpoints in oncology drug approvals For the improvement of signs and symptoms or QoL assessments to be used as primary endpoints to support cancer drug approval, there must be a clear distinction between improvement in tumour symptoms and lack of drug toxicity –An apparent effectiveness advantage based on a global health-related QoL instrument can simply indicate less toxicity rather than effectiveness

50. Problems With Symptom Data Withdrawing treatment because of drug toxicity or tumour progression is one cause of missing symptom data Ideally, when patients stop treatment, data collection forms should continue to gather information to inform the analysis In fact, patients are often lost to follow-up when the treatment is discontinued Missing data and infrequent assessments

51. Biomarkers Predictive biomarkers do not serve as primary endpoints for drug approval Further research is required to establish the validity of available tests and determine which biomarkers may predict clinical benefit May serve as elements of a composite endpoint in the future

52. Summary The function of the regulatory authority is to protect and promote public health There are variations among the regulatory authorities’ organisations, approval processes, as well as their prospective/preferences for approvable clinical study endpoints In order to expedite the process of bringing novel and effective drugs to patients and minimising suboptimal resource utilisation, these idiosyncrasies must be taken into consideration when designing registrational trials and applying for drug approval