Bioequivalence and Bioavailability Forum

Dear Atish!

❝ [...] a four period crossover BE Study with four treatments R1, R2 and T1, T2 comparing R1 with T1 and R2 with T2 separately.

❝ Randomisation sequence would be defined as follows

❝

❝ +-----+----+----+----+----+

❝ | S/P | 1  | 2  | 3  | 4  |

❝ +-----+----+----+----+----+

❝ |  A  | R1 | T1 | R2 | T2 |

❝ |  B  | T1 | R1 | T2 | R2 |

❝ |-----+----+----+----+----+

I guess you 'harvested' the layout from this post dealing with a 4×4×4 Williams's design?

+-----+---+---+---+---+
| S/P | 1 | 2 | 3 | 4 |
+-----+---+---+---+---+
|  1  | D | C | A | B |
|  2  | A | D | B | C |
|  3  | B | A | C | D |
|  4  | C | B | D | A |
+-----+---+---+---+---+

❝ Kindly let me is it correct to conclude Bioequivalence by comparing R1

❝ with T1 and R2 with T2 as if there were two periods.

I highlighted the important point: [...] as if [...].
There's no 'as-if-evaluation'; you should design and evaluate the study properly.

Your design is not balanced any more (T₁/R₁ only in periods 1/2, T₂/R₂ only in periods 3/4). Since any sequence effect can only be avoided by design (single dose, no endogenous compound, sufficient washout, no pre-dose concentrations in periods >1), you may run into questions by assessors. In a Williams' design every treatment is preceeded by all other treatments and is administered once as the first one. In technical terms such a property is called balanced for treatment and carry-over.
In your design both T₂ and R₂ are never administered in period 1 (to the 'treatment-naïve' subject). You may get nasty questions - because obviously in the evaluation you want to split the design into two separate 2×2 'studies':

+-----+----+----+----+----+  #  +-----+----+----+  +  +-----+----+----+
|   actual study layout   |     |    split 1    |     |    split 2    |
+-----+----+----+----+----+     +-----+----+----+     +-----+----+----+
| S/P | I  | II | III| IV |     | S/P | I' | II'|     | S/P | I" | II"|
+-----+----+----+----+----+     +-----+----+----+     +-----+----+----+
|  1  | R1 | T1 | R2 | T2 |     |  1' | R1 | T1 |     |  1" | R2 | T2 |
|  2  | T1 | R1 | T2 | R2 |     |  2' | T1 | R1 |     |  2" | T2 | R2 |
+-----+----+----+----+----+     +-----+----+----+     +-----+----+----+

Whereas the first split is a conventional 2×2×2 cross-over, the second one only looks like one.
I wouldn't recommend it.
Since you are only interested in T₁ vs. R₁ and T₂ vs. R₂, why don't you perform two conventional studies?
The money gained in performing fewer screening/post-study examinations easily can be outweighed by the higher chance of drop-outs (4 periods instead of 2). Furthermore, if you want to submit the study to different countries (let's say the origins of R_1,2) you always would have to submit the entire stuff.

If you really want to stay with your design, be prepared for questions anyhow.