Bioequivalence and Bioavailability Forum

Dear Muneesh!

❝ We are planning a 4-period replicate study for a highly variable drug with

❝ a long washout of 30 days. But the reference product is getting expired in

❝ the month of the last dosing day. Can we go ahead with the study?

Yes. But why are you going for a 4-period replicate design? Both FDA's recommendations (RSABE) and EU's requirements (widening of the acceptance ranage of C_max) also allow for a 3-period study. Are you really interested in the intra-subject CV of the test product? Compared to the 4-period design you will need 50% more subjects to obtain the same statistical power (example: 2-period 64, 3-period 48, 4-period 32), but:

• the number of treatments (and biosamples) will be only 12.5% higher if you don't expect any dropouts:
  • 2-way: 64 subjects × 2 periods = 128 treatments
    
  • 3-way: 48 subjects × 3 periods = 144 treatments (+12.5%)
    
  • 4-way: 32 subjects × 4 periods = 128 treatments
• the chance of drop-outs will be lower (only two washout-phases instead of three):
  expecting a dropout-rate of 5%/washout we get
  • 2-way: 64, 3.2/washout (WA) rounded up to get a balanced design = 4 gives 68 total (136 treatments)
    
  • 3-way: 48, 2.4/WA -> 4.8 total -> 6 = 54 total (162 treatments: +6.6% compared to 4-way)
    
  • 4-way: 32, 1.6/WA -> 4.8 total -> 6 = 38 total (152 treatments)
• In your special case (expiry date approaching) you will gain one month.

❝ And are there any guidelines/document stating that expiry date should be

❝ considered as the last day of the month written on the IP to avoid any

❝ regulatory queries in future?

No one I know of.