Bioequivalence and Bioavailability Forum

Dear Group Members,

I would like to draw your attention on a potential bug found in Winnonlin 5.2.1 software. Here are the details:

During Winnonlin PK data analysis, Lambda_z range needs to be selected. This could be done by one of the following methods:

1. Setting up the Start & End times (terminal phase) or,

2. Manually deselecting points one by one (by holding Ctrl key while selecting or deselecting a point) from a set of all points.

Although there were no differences between the two methods in Lambda_z, Half Life, Rsq, Cmax, Tmax, Tlast & Clast values, significant differences were observed in C0 & its related parameters such as, Vz, Vss, AUCINF, MRT, etc. It appears that WinNonlin is doing the log-linear regression with the first two points when no exclusions are applied; however, it is defaulting to the rule of using the first non-zero concentration for C0 when the first two points are excluded.

Pharsight Corporation admits that this is a bug & states that a potential fix would be developed in the next release. Until then, it is recommended to use the first option (selection of Start & End times of terminal phase).

Best regards,
Dr. Harish L. Rao
India.