AN INDUSTRY PERSPECTIVE ON THE USE OF RWE FOR REGULATORY PURPOSES: EXPERIENCE FROM RARE DISEASES

May 7, 2019 Emily S. Brouwer, MPH PharmD PhD Data Sciences Institute, Global Outcomes Research and Epidemiology Shire Pharmaceuticals, now part of Takeda

1 DISCLOSURES

• Opinions and views expressed here today are my personal views and do not represent Takeda’s collective view or position, nor do they express the views, position or opinions of other industry stakeholders

• Full time employee of Takeda Pharmaceuticals and shareholder

2 | 04/03/2019 WHAT IS THE DEFINITION OF A RARE DISEASE? HOW RARE IS RARE?

• United States FDA definition of rare disease1: – Affects < 200,000 persons in the United States – Affects > 200,000 persons but the Maximum 54-65 persons costs will not be recovered for developing product in this stadium with the disease of interest • EU definition of rare disease2: – < 5 per 10,000

Image: Presenter’s own image. • There are > 6,000 rare diseases2 Michigan Stadium Capacity: ~108,000 1FDA. Developing products for rare diseases and conditions. Available at: https://www.fda.gov/industry/developing-products-rare-diseases-conditions. Accessed on: April, 2019 2Antoni Montserrat Moliner and Jaroslaw Waligora. The European Union Policy in the Field of Rare Diseases. Adv Exp Med Bio.. 2017; 1031:561-587. A FEW EXAMPLES OF RARE DISEASE

Duchenne Muscular Dystrophy

Primary Hereditary Immune Angioedema ~2:100,000 Deficiency persons1

Hunter ~0.62:100,000 Hemophilia male births2 Syndrome

Congenital Protein C Deficiency

1Aasia Ghazi and J Andrew Grant. Hereditary angioedema: epidemiology, management, and role of icatibant. Biologics. 2013; 7: 103-113. 2Raluy-Callado M, et al. The impact of Hunter syndrome (mucopolysaccharidosis type II) on health-related quality of life. Orphanet Journal of Rare Diseases 2013, 8:101. CHALLENGES IN DEVELOPING THERAPEUTICS FOR RARE DISEASE: CURRENT EXPERIENCE AND OPPORTUNITIES FOR USE OF REAL WORLD EVIDENCE (RWE)

Natural history of disease often not well characterized

Potentially not feasible/unethical to conduct randomized placebo controlled study

Challenging to evaluate effectiveness in special populations

Rare diseases often have a long diagnostic journey and frequently it is difficult to enroll patients in trials

Note: Presenter’s own perspective/experience NATURAL HISTORY STUDIES USING REAL WORLD DATA CAN PROVIDE IMPORTANT INFORMATION ON CLINICAL COURSE OF DISEASE1 • Disease progression

• Burden of illness

• Standard of care

• Background rates of comorbid conditions

• Outcomes that are important to the patient, including patient reported outcomes (PROs)

1 Augustine EF et al. Clinical trials in rare disease: challenges and Image 1: Presenter’s own image. Image 2: Obtained from personal friend with permission opportunities. J Child Neurol. 2013; 28(9): 1142-1150. REGISTRIES PROVIDE SOURCE OF DATA FOR NATURAL HISTORY STUDY • Icatibant Outcome Survey (IOS)1 • Hunter Outcome Survey (HOS)2 • Initiated in 2009 as part of a • Initiated in 2005 as part of a commitment to regulatory authorities commitment to regulatory authorities • HAE I/II patients enrolled ➔ Initially • MPS II patients enrolled icatibant only patients but expanded to • In addition to data collected from include patients on C1-esterase normal clinical practice, patients also inhibitors were able to complete a Patient • In addition to data collected from Reported Outcome (HS-FOCUS)* normal clinical practice, IOS also collects • Provides context to newer clinical trial data on HAE attacks, an important data on progression of Hunter clinical outcome that is not captured Syndrome with regularity in existing medical records *HS-FOCUS: Hunter Syndrome - Functional Outcomes for Clinical Understanding Scale • Provides context to newer clinical trial HAE: Hereditary Angioedema data on the progression of HAE disease MPS II: Mucopolysaccharidosis type II

1Longhurst et al. Real-world outcomes in hereditary angioedema: first experience from the icatibant outcomes survey in the United Kingdom. Allergy Asthma Clin Immunol (2018) 14:28. 2Muenzer et al. Ten years of the Hunter Outcomes Survey (HOS): insights acheivements, lessons learned from a global patient registry. Orphanet Journal of Rare Disease. 2017 12:82. LONGITUDINAL COHORT DATA IN PARALLEL TO CLINICAL TRIAL ADDS CONTEXT TO TRIAL DATA1 • Hunter Syndrome longitudinal cohort study conducted in parallel to clinical trial

• Patients aged 2-18 years of age with a concomitant medication for Mucopolysaccharidosis type II enrolled (N=55)

• Measured the same Patient Reported Outcomes (PRO) as the clinical trial: Differential Abilities Scales®, second edition (DAS-II) and Vineland Adaptive Behavior Scales™, second addition (VABS-II) – Association with the HS-FOCUS that was also collected (PRO used in natural history study)

• Evaluated changes in PRO measurements over time

1Yee et al. Analysis of cognitive ability and adaptive behavior assessment tools used in an observational study of patients with Hunter syndrome. 15th Annual WORLD Symposium™, February 4–8, 2019, Orlando, FL, USA CHALLENGES IN DEVELOPING A REGISTRY OR A PROSPECTIVE OBSERVATIONAL STUDY FOR REGULATORY PURPOSES • Critical data elements: which elements are able to be collected reliably, are important to patients, and will be relevant data elements to provide context to ongoing clinical trials

• Timing of registry: must start early enough in development to have enough data and not compete with developmental studies

• Geographical representation: Some countries/sites may have very few patients, if conducting prospective observational study at same time, patient recruitment becomes difficult

• Disease registry (vs. product only registry): The optimal type of registry is a disease registry, however, this can be challenging in a competitive landscape

• Comparison of patients: patients enrolled in trials are likely to be very different (especially if enrolled concurrently); historical controls may also not be similar due to changes in standard of care

Note: Challenges are identified from presenter’s own perspective/experience HISTORICAL CONTROLS HAVE BEEN USED FOR REGULATORY APPROVAL IN RARE DISEASE/CANCER1 Disease Therapy Year Approved Acute hypoammonemia/severe urea cycle disorders sodium phenylacetate and sodium benzoate 2005 Pompe Disease (alpha-glucosidase deficiency) algucosidase alfa 2006 Toxic plasma methotrexate concentrations glucarpidase 2012 Relapsed or refractory Acute Lymphoblastic Leukemia (ALL) blinatumomab 2014 Hypophosphatasia (HPP) asfotase alfa 2015 Lysosomal acid lipase (LAL) deficiency sebelipase alfa 2015 Neuronal ceroid lipofuscinosis type 2 (CLN2 disease) Cerliponase alfa 2017

For the majority of these examples, statistical approaches were not used to compare the single-arm intervention to the historical control Propensity score approaches were used for blinatumomab historical control comparison

1Lim et al. Minimizing patient burden through the use of historical control data in innovative confirmatory trials: review of methods and opportunities. Therapeutic innovation and regulatory science 2018, Vol. 52 (5)546-559. REAL WORLD EVIDENCE USED FOR LABEL EXPANSION TO NEW INDICATIONS AND SPECIAL POPULATIONS IN RARE DISEASE: ECULIZUMAB • Eculizumab pediatric indication for atypical Hemolytic Uremic Syndrome (aHUS)1 • Retrospective registry data of 19 pediatric patients (<18 years of age) supplemented formal clinical trial in this sub-population • Clinical outcomes were nominally compared to pivotal trials, no formal analyses conducted • Extension of eculizumab indication in the EU to patients without history of transfusion2 • Global Paroxysmal Nocturnal Hemoglobinuria (PNH) registry was approved to be used in place of a randomized controlled trial to expand indication 1Soliris® (eculizumab) prescribing information. FDA. November 2011. Available at: https://www.accessdata.fda.gov/drugsatfda_docs/label/2011/125166s172lbl.pdf Accessed: April 2019 2European Medicines Agency. Extension of indication variation assessment report: Soliris. February 2015. Available at: https://www.ema.europa.eu/en/documents/variation- report/soliris-h-c-791-ii-0066-epar-assessment-report-variation_en.pdf. Last accessed April 2019. POTENTIAL EXTENSIONS OF REAL WORLD EVIDENCE USE FOR REGULATORY PURPOSES: IDENTIFICATION OF PATIENTS AND DATA ELEMENTS • Patients with rare diseases have a long diagnostic journey and a high misdiagnosis rate1

• It is often difficult to identify patients for inclusion in clinical trials; a situation not uncommon with most rare diseases or select subgroups1

• New methodologies that leverage: electronic medical records, administrative claims, registries, and chart reviews would increase efficiencies of clinical trials and would be able to2: • Improve identification of patients for clinical trials • Confirm surrogate, physician-reported, Patient Reported Outcomes (PROs) and other alternative endpoints from clinical trials • Perform comparative effectiveness studies using linked PROs as co-primary or secondary endpoints

1Blӧβ S, et al. Diagnostic needs for rare diseases and shared prediagnostic phenomena: results of a German-wide expert Delphi survey. PLoS One. 2017: 12(2): e0172532. 2Presenter’s own perspective/experience ELECTRONIC MEDICAL RECORD (EMR) DATA USING NATURAL LANGUAGE PROCESSING + TRADITIONAL CLAIMS ALGORITHMS FOR HAE COHORT IDENTIFICATION1

Both N=84 (65%)

HAE: Hereditary Angioedema ICD-9/10: International *Diagnosis code only, n=24; notes only, n=70; classification of disease, 9th, 10th codes and notes, n=96. †Diagnosis code only, n=24; revision notes only, n=22; codes and notes, n=84; 1 note mention, n=124.

• Of 130 confirmed cases, 65% were identified through administrative claims and EMR • Such methodology can be used to refine clinical definitions and risk factors which may result in earlier diagnosis and treatment and identification for clinical trials 1Brouwer ES, et al. Identification of HAE cohort in a US Ambulatory EMR Database. Poster presented at the 34th International Conference on Pharmacoepidemiology and Therapeutic Risk Management, Prague 2018. FINAL THOUGHTS

• Real world evidence is used for drug development and regulatory purposes in the rare disease space

• Experience from the study of rare diseases can provide insight for the use of real-world evidence in other therapeutic areas

• Many opportunities for methodologic development in order to insure robust data for regulatory evaluation

Note: Presenter’s own perspective