Clinical Development Challenges: Trial Designs and Endpoints Menno de Jong Department of Medical Microbiology Academic Medical Center, University of Amsterdam ISIRV - Options IX for the Control of Influenza Menno de Jong, MD PhD Academic Medical Center University of Amsterdam Amsterdam, the Netherlands I have financial relationship(s) with: Research grant: Crucell Advisory Boards: Crucell, AIMM therapeutics, MedImmune, Avi Biopharma AND My presentation does include discussion of off-label or investigational use of antivirals. Placebo-controlled RCTs of neuraminidase inhibitors: mostly in previously healthy people with mild flu • Symptom relief as primary endpoint Observational studies point to benefits of oseltamivir treatment in hospitalized patients Muthuri SG et al. Antivirals for influenza: current state of affairs • Licensed agents only for treatment of acute uncomplicated flu – Adamantanes: • amantadine, rimantadine • not recommended due to resistance in circulating strains – Neuraminidase inhibitors: • oseltamivir (oral), zanamivir (inhaled) • US, Japan, S Korea: peramivir (IV); Japan: laninamivir (inhaled) • No licensed agents for treatment of serious/hospitalized flu Guidelines recommend (off-license) use of oseltamivir for severe influenza Mind the gap “Rational believers” “Rational non-believers” - Flu is caused by influenza viruses - No evidence from RCTs and can be severe - Oseltamivir inhibits flu viruses - Proven efficacy for uncomplicated flu - Observational studies strongly suggest efficacy for severe flu Treat patients with severe disease Not rational to treat at present and those at risk for severe disease How to determine efficacy of new antivirals in patients with severe influenza? Efficacy studies in severe hospitalized influenza are very complicated • No formal demonstration of clinical efficacy for any antiviral – FDA: active-controlled non-inferiority trial is not an option • Current treatment guidelines prevent placebo controls – FDA: dose-response, or superiority when added to ‘standard of care’ • No validated efficacy endpoints – FDA: endpoints should demonstrate improvement in how the patient feels, functions or survives; primary virological endpoint not appropriate. • Superiority trial: peramivir vs placebo added to standard-of-care (SOC) • Primary efficacy analysis in population not receiving oseltamivir as SOC • Hospitalized patients, broad inclusion criteria, broad geography • Primary endpoint: time to clinical resolution (TTCR) = resolution ≥ 4 of 5 vital signs: Study period: 2009-2012 placebo vs peramivir Study terminated prematurely for futility after preplanned interim analysis The challenges encountered: • Patient enrollment (n = 405) – Long study period (2009-2012), 6 influenza seasons – >300 sites, 21 countries: • no enrollment from > 70% of sites; 6% of sites enrolled 63% of patients • ≈ 90% of non-NAI SOC patients enrolled from India/Eastern Europe • Heterogeneous patient population – broad spectrum of illness severity – ≈ 70% comorbidities and/or age > 60 years – variety of influenza (sub)types • Unvalidated clinical endpoint (TTCR) • 2007 – 2010 • 13 hospitals, 4 countries • 699 screened, 326 randomized primer sand probesto enablethem to detect it by RT-PCR. Furthermore, we wer e able to describe the first worldwide transmission of oseltamivir resi st ant pH1N1 in a com- munity cluster [9]. Laboratory Quality Enhancement Program In partnership with Family Health International (FHI), the SEAICRN initi- ated a clinical laboratory quality improve- ment program in the region. This includ- ed: (i) baseline assessment of each hospital clinical laboratory against international standards, ( ii) inventory, maintenance, and calibration of instruments, (iii) enroll- ment in external quality assurance pro- gram, (iv) assessment of training needs, (v) review of normal reference values used in the laboratory, and (vi) accreditation status of the laboratories by local and interna- tional accreditation bodies. The baseline Figure 1. South East Asia Infectious Disease Clinical Research Network sites and assessments evaluated the overall quality laboratories. doi:10.1371/journal.pmed.1000231.g001 of the clinical laboratory and helped to develop short- and long-term recommen- dations for improvement. Based on the trained in influenza real-time RT-PCR testing, storing, shipping, and document- assessment and recommendations, subse- diagnostics and molecular diagnostics in ing data from trial specimens according to quent follow-up visits and technical sup- general, including principles of unidirec- protocol requirements. A centralized spec- port were provided to each hospital to tional workflow and prevention of carry- imen labeling and database system was set implement those recommendations. over contamination. These trainings were up to track and trace all stored specimens Laboratory management staff in each performed on an ongoing basis at the for the SEAICRN studies. Currently the hospital were trained and encouraged to reference laboratory in HoChi Minh City, SEAICRN uses Freezerworks software implement a quality management system. followed by on-site training. To ensure (Dataworks Development; http:/ / www. This included development and imple- quality, each MDL laboratory was en- freezerworks.com/ ) to track study speci- mentation of quality and technical manu- rolled in at least two different external mens and aliquots. Sites received deep als, standard operating procedures (SOPs), quality assurance programs for molecular freezers for adequate specimen storage and a document control system. To help influenza diagnostics. Furthermore, profi- and power back-up systems were installed to improve laboratory quality it was ciency testing for diagnosing influenza by where needed to ensure an uninterrupted necessary to appoint a senior staff member RT-PCR was performed before sites were power supply. Table 2 summarizes the as a Quality Officer to oversee all aspects allowed to start screening patients for the required laboratory capacity and what was of laboratory quality and the quality studies. done to achieve this. enhancement program. Furthermore, Each hospital is encouraged to use the The laboratory capacity program of the each laboratory was supported with nec- MDL for purposes other than SEAICRN- SEAICRN enabled usto rapidly respond to essary instruments, and staff were encour- related research activities, such as HIV, changes in influenza epidemiology, su ch as aged to establish and monitor an equip- hepatitis, dengue, meningitis, and enceph- thespread of naturally occurringoseltamivir ment maintenance and calibration alitis testing by molecular techniques. The resi st ance in seasonal H1N1 virusesand the program. All staff from each laboratory philosophy is that it is important and recent emergence of the novel influenza A were trained on standards developed by highly beneficial to increase expertise in (H 1N1) st rain (pH1N1) in April 2009. In the technical committee of the Interna- detecting infectious agents by molecular response to the increasi ng rates of oseltami- tional Organisation for Standardisation diagnostics and so stimulate the imple- vir resi st ance in seasonal H1N1 viruses, (ISO15189:2003 and ISO15190:2003, mentation of other relevant molecular before the emer gence of pH1N1, the http:/ / www.iso.org/ ) [10]. diagnostic tests. For example, after imple- laboratory committee decided to test Asi an During the clinical trials it was noticed mentation of a Streptococcus suis–specific isolates and respiratory sp ecimens from that occasionally reported biochemistry molecular test in the MDL of the National influenza patientsin our st udies [7,8]. This results were implausible. In response, a Hospital of Tropical Diseases, Vietnam, it st ep gener ated valuableresi st ance datafrom continuous improvement system was im- wasfound that S. suisisa common cause of theregion, which wereinst antly shared with plemented to ensure that laboratory test bacterial meningitis in adultsin Hanoi [6]. the national and inter national medical and results are reviewed by a qualified person In preparation for the start of the scientific community. In April 2009, when to identify performance issues. In addition, clinical trials, dedicated laboratory trial confronted with a pandemic threat of this advice wasgiven on proper use of internal staff members were appointed to become novel influenza virus, we prepared collabo- controls, how to monitor and investigate staff of the national institutions. They rating laboratories by providing them with the possible cause(s) of controls not received training on processing, labeling, sequence information of this st rain and meeting acceptance criteria, and correc- PLoS Medicine | www.plosmedicine.org 3 April 2010 | Volume 7 | Issue 4 | e1000231 In early development: European Clinical Research organization for Antimicrobial resistance and emerging Infectious Diseases (ECRAID) Smarter trial designs to improve efficiency • Traditional RCTs: – long, slow & expensive to conduct – provide ‘average’ answers • fail to capture the nuances of real-life clinical care • Adaptive design RCTs: – takes advantage of accumulating data during trial – earlier answers by response-adaptive randomisation • more patients randomized to effective intervention • reduce imbalances of subgroups between study arms • detect efficacy ‘signals’ in subgroups – flexible • test several interventions concurrently • add & delete study arms Funded by the European Union The challenges encountered: •