PREDICTING INJECTION SITE MUSCLE DAMAGE .2. EVALUATION OF EXTENDED-RELEASE PARENTERAL FORMULATIONS IN ANIMAL-MODELS

Purpose. The goal of this study was to And a resource sparing alternat ive to the rabbit lesion model (RbLV) for assessing injection site tol eration in extended release (ER) intramuscular (IM) formulation screen ing. Methods. ER formulations (danofloxacin oily and aqueous suspensio ns) were evaluated in RbLV, rat and rabbit plasma creatine phosphokina se (CK), and rat foot edema (RFE) models as described in the companion article. Results. None of the short term models could consistently pr edict acute and chronic effects of the. For example, RFE predicted lit tle muscle damage from aqueous vehicle (0.03 +/- 0.03 g) and 60 mg/ml (0.08 +/- 0.03 g) formulation; while RbLV(days1-3) was marked and grea ter (p < 0.05) for 60 mg/ml (6.0 +/- 3.1) than vehicle (2.2 +/- 2.9) f ormulations. Furthermore, RbLV(days1-3) for vehicle (6.5 +/- 7.5) and 60 mg/ml (4.9 +/- 4.6) danofloxacin oily formulations were worse (p < 0.05) than oil alone (1.4 +/- 2.2); an observation not predicted by CK models, since they apparently reflected only the acute muscle damage of formulation components immediately available to surrounding tissue at the time of injection. Conclusions. The CK models may be useful to screen those ER formulations with unacceptable acute damage due to imm ediately available components. However, to evaluate potential delayed effects from ER formulations, the long-term model RbLV was still recom mended.