According to Dr Bosco, in order to design biosimilar studies that are fit for purpose the following considerations need to be taken into account:

• Who is the audience? How much certainty is needed?
• If a randomized controlled trial is not required, are existing data accessible and sufficient?
• If not, could they be supplemented/replaced by direct data collection?
• How much follow-up is needed?
o What is the expected induction time for risks and benefits?
o What are stakeholders expectations for follow-up period?
• Are comparators needed?
o Are historical data sufficient?
o Treatment complexity? Single, multiple, etc.?
o Is there reasonable overlap of patient characteristics between treatments?
• Time and budget?

In addition, the intended purpose must be considered, as well as what is acceptable in the region, if there is openness to new models and whether existing data could be sufficient. Studies span a continuum, with varying levels of control, data validation and cost. Once all this has been considered, a decision can be made on whether a randomized or a non-randomized trial is necessary.

Randomized trials include classical randomized controlled trials (RCTs) and pragmatic randomized trials. Classically controlled trials assess the difference between a health intervention and placebo, in a highly controlled setting. Whereas pragmatic randomized trials assess the difference between two or more health interventions, including extraneous factors. They do not include placebo and aim to maximize generalizability to a broader patient population.

Furthermore, outcomes may also determine the appropriate design and data source:
• Immunogenicity outcomes are best evaluated using interventional trial designs
o Anti-drug antibodies (ADAs)
o Trough levels
• Changes in treatment patterns and patient characteristics described using administrative prescription claims data
o Describe changes in treatment patterns of biologicals or biosimilars
o Understand the impact of characteristics of patient populations on treatment initiation or switching to biologicals or biosimilars
• Healthcare utilization outcomes might best be studied in administrative claims data sources
o Switching to a different biological after a switch to a biosimilar
o Treatment discontinuation
• Clinical effectiveness endpoints may best be captured in electronic health records or registry data
o Disease activity scores
o Patient reported outcomes

Dr Bosco also presented some options for novel trial designs in real-world biological and biosimilar studies:

Enriched trials
Primary data collection (agent + comparator) + Secondary data (agent + comparator)
• Improves efficiency of data capture
• More data allows research programme to address many areas of interest

Extended trials
RCT data collections ends (agent + comparator) (LINKED) Direct to patient and/or secondary data
• Low-cost follow-up for long-term safety and effectiveness

Randomized pragmatic trials
Primary data collection (agent + comparator) (MAY BE LINKED) Secondary data (agent + comparator)
• To force exposure to biosimilar or biological of interest with the advantage of randomization
• Uses pragmatic outcomes of interest to clinicians, patients and payers

External comparator trials
Primary data collection (agent) + Secondary data (comparator)
• Comparators for single-arm clinical trials/studies and product registries
• Particularly good for rare diseases when there is a robust data package
• Precedence for label expansion in oncology

Finally, Dr Bosco ended her presentation with the questions:
• Will we be able to do purely database studies?
• In what situations will alternative strategies be needed?
• How will you get comparator data?
• How will we account for channelling bias since there could be substantial clinician preference to prescribe biosimilars only to new users?

Related articles
Uptake of biosimilars in different countries varies

Considerations for real-world research on biosimilars

Impact of biologicals on health care

Reference
1. Bosco J. Biosimilar considerations for real world research and stakeholder questions. DIA Biosimilars Conference; 2019 Sep 23‒24; Bethesda, Maryland, USA.

Permission granted to reproduce for personal and non-commercial use only. All other reproduction, copy or reprinting of all or part of any ‘Content’ found on this website is strictly prohibited without the prior consent of the publisher. Contact the publisher to obtain permission before redistributing.

Copyright – Unless otherwise stated all contents of this website are © 2019 Pro Pharma Communications International. All Rights Reserved.