Crossover design in confirmatory therapeutic trials

This week, FDA approved a new drug called Aqneursa (levacetylleucine) for the treatment of neurological symptoms associated with Niemann-Pick disease type C (NPC) - a rare disease - in adults and pediatric patients. FDA approval was based on the confirmatory phase 3 study using a crossover design. "The safety and efficacy of Aqneursa for the treatment of NPC were evaluated in a randomized, double-blind, placebo-controlled, two-period, 24-week crossover study. The duration was 12 weeks for each treatment period. The study enrolled 60 patients." The study design was discussed in the paper by Fields et al "N-acetyl-L-leucine for Niemann-Pick type C: a multinational double-blind randomized placebo-controlled crossover study" and the crossover study design was depicted as below. Patients were assessed during a baseline period and then randomized (1:1) to one of two treatment sequences: IB1001 (levacetylleucine) followed by placebo or placebo followed by IB1001 (levacetylleucine). Each sequence consists of a 12-week treatment period. The primary efficacy endpoint is based on the Scale for the Assessment and Rating of Ataxia

The advantages of the crossover design includes:

• It allows a within-patient comparison between treatments. Each patient serves as his/her own control
• Treatment contrasts estimated as a within subject effect
• Fewer subjects (smaller sample size) required for the study

The disadvantages of the crossover design includes: 

• Carryover effect of previous treatment to the next treatment
• Time dependent changes
• Increased time length of the trial
• Drop out issue, too many visits
• May require complicated statistical modeling - for example, mixed model to analyze the data

The crossover design is very common and almost the default design for bioequivalence studies. As stated in FDA's guidance for industry "Bioequivalence Studies With Pharmacokinetic Endpoints for Drugs Submitted Under an ANDA", crossover design was recommended in most of situations. 

"For most dosage forms that release a drug intended to be systemically available, FDA recommends that applicants perform a two-period, two-sequence, two-treatment, single-dose,crossover study using either healthy subjects or other populations, as appropriate. In this design, each subject should receive each treatment (the test and the reference product) in a random order. 

A replicate crossover study design (either partial or fully replicate) is appropriate for drugs  whether the reference product is a highly variable drug or not. A replicate design can have the advantage of using fewer subjects compared to a non-replicate design, although each subject in a replicate design study would receive more treatments."

However, crossover designs are seldom employed in confirmatory therapeutic clinical trials, which typically use clinical endpoints rather than pharmacokinetic ones. The most common design for these trials is the traditional randomized controlled trial (RCT) with parallel groups, where patients are randomly assigned to either an experimental treatment group or a placebo.

In some special situations such as the rare disease drug development, the crossover design may be employed in the phase 3 confirmatory clinical trials. Here are some examples: 

• Samsung Bioepis Co. "A Study to Compare SB12 (Proposed Eculizumab Biosimilar) to Soliris in Subjects With Paroxysmal Nocturnal Haemoglobinuria" - the product" - FDA approved the product based on this study with crossover design
• KalVista Pharmaceuticals "A Phase III, Crossover Trial Evaluating the Efficacy and Safety of KVD900 for On-Demand Treatment of Angioedema Attacks in Adolescent and Adult Patients With Hereditary Angioedema (HAE)" - NDA based on this study is under FDA's review
• GSK "Cross-Over Study in Subjects With COPD, Evaluating Lung Function Response After Treatment With Once Daily Umeclidinium 62.5mcg, Vilanterol 25mcg, and Umeclidinium/​Vilanterol 62.5/​25mcg"
• Wiley et al (2016) "A Crossover Design for Comparative Efficacy: A 36-Week Randomized Trial of Bevacizumab and Ranibizumab for Diabetic Macular Edema"
• Bethoux et al (2024) "A randomized, double-blind, placebo-controlled trial to evaluate the effect of nabiximols oromucosal spray on clinical measures of spasticity in patients with multiple sclerosis"  
• Watz et al (2016) Effects of indacaterol/glycopyrronium (QVA149) on lung hyperinflation and physical activity in patients with moderate to severe COPD: a randomised, placebo-controlled, crossover study (The MOVE Study)
• Zhang et al (2022) Design and analysis of crossover trials for investigating high-risk medical devices: A review

ICH E9 "STATISTICAL PRINCIPLES FOR CLINICAL TRIALS" have a section discussing crossover design:

FDA's Good Review Practice: Clinical Review of Investigational New Drug Applications" also discussed the advantages and disadvantages of the crossover design:

In confirmatory therapeutic trials using a crossover design, an additional concern is the potential for unblinding. As the same patient receives both the experimental drug and the placebo, the patient might be able to identify or guesstimate the treatment group they are assigned to based on slight differences in taste, smell, or side effects.