Bioequivalence and Bioavailability Forum

Hi mittyri,

❝ They stated above regarding 90% CI. Don't know what kind of distribution was used for CI calculation.

I think that’s wrong. The upper 90% confidence limit of the pooled CV (0.4357) with 356 degrees of freedom is 0.4601.

library(PowerTOST)
alpha <- c(0.25, 0.20, 0.10)
CV    <- 0.4357
df    <- 356
res <- data.frame(alpha = alpha, df = df, CV = CV, upper.CL = NA)
for (j in seq_along(alpha)) {
  res$upper.CL[j] <- signif(CVCL(CV = CV, df = df, side = "upper",
                                 alpha = alpha[j])[["upper CL"]], 4)
}
cat("CV =", CV, "df =", df, "\n"); print(res[, c(1, 4)], row.names = FALSE)

# CV = 0.4357 df = 356
# alpha upper.CL
#  0.25   0.4485
#  0.20   0.4516
#  0.10   0.4601

❝   – what do you think about statement that males and females (50/50) should be enrolled, since RLD says regarding sex differences in PK? 'Otherwise the marketed formulation could be not BE for other gender'

From the EPAR:

Plasma concentrations were higher in females than in males, without relevant influence on efficacy.
Differences in plasma concentrations were observed, with C_max and AUC being approximately 3- and 2-fold higher, respectively, in females compared to males.

That confirms the observation that highly variable drugs have a flat dose-response curve (no relevant influence on efficacy despite significantly higher concentrations). Since the experts recommended males and females, go for it. Bioanalysts will like it.

❝   – what do you think about FLU results (study 3)?

I don’t believe it. Look at Figure 1 and Table 1.

❝ There's a comment in the text stated that LC-MSMS should be recommended since higher (!) concentrations were achieved with that method and LLOQ 0.5-3.0

Agree. The concentration in study 3 are way lower than what is not only seen in the other studies but also in the originator’s. Furthermore, the LLOQ is 14% of the (mean!) C_max. That’s insufficient.