Case Histories - Freeze-drying of Pharmaceuticals and Biopharmaceuticals - page 170

Figure 3 Activity assays of a product, stabilised with mannitol, and stored for up to 26
weeks at 20 and þ 371C. For details, see text
stability tests. A set of random results, similar to that shown in Figure 3,
was obtained.
It was pointed out to the client that during freeze-drying, great care
had been exercised to ensure that the preparation, including the man-
nitol, remained in the amorphous state. On the other hand, stability
trials carried out at 371C ensured that mannitol would crystallise sooner
or later, with the water remaining in the amorphous product phase, thus
reducing its T g and the stability of the product, but in an erratic manner,
as was indeed observed. The curious stability data did not represent a
recovery of activity in vials that had earlier given a poor result. Stability
assays had been performed on vials, chosen randomly from a batch.
Because of the random nature of mannitol crystal nucleation, any assay
result depended entirely on the physical state of the dried product in a
particular vial at the time of testing that, in turn, was governed by the
absence or occurrence of mannitol crystallisation and the degree to
which mannitol had already crystallised in the vial under test at the time.
If stability at 371C was essential, then a major reformulation was
required.
12.6 Unfamiliarity with Freeze-Drying Parameters
A freeze-drying process was developed for a client for a preparation that
contained 2 mg of a bioactive (cytotoxic) material and 51 mg of
excipient, made up to 10 ml with water; process parameters are shown
in Column 2 of Table 3. Starting with a measurement of T g , and the