Determination of therapeutic equivalence of generic products of gentamicin in the neutropenic mouse thigh infection model

Abstract

Background: Drug regulatory agencies (DRA) support prescription of generic products of intravenous antibiotics assuming therapeutic equivalence from pharmaceutical equivalence. Recent reports of deaths associated with generic heparin and metoprolol have raised concerns about the efficacy and safety of DRA-approved drugs.

Methodology/principal findings: To challenge the assumption that pharmaceutical equivalence predicts therapeutic equivalence, we determined in vitro and in vivo the efficacy of the innovator product and 20 pharmaceutically equivalent generics of gentamicin. The data showed that, while only 1 generic product failed in vitro (MIC = 45.3 vs. 0.7 mg/L, P<0.05), 10 products (including gentamicin reference powder) failed in vivo against E. coli due to significantly inferior efficacy (E(max) = 4.81 to 5.32 vs. 5.99 log(10) CFU/g, P<or =0.043). Although the design lacked power to detect differences in survival after thigh infection with P. aeruginosa, dissemination to vital organs was significantly higher in animals treated with generic gentamicin despite 4 days of maximally effective treatment.

Conclusion: Pharmaceutical equivalence does not predict therapeutic equivalence of generic gentamicin. Stricter criteria based on solid experimental evidence should be required before approval for human use.

Figure 1. Repeatability of the animal model to determine therapeutic equivalence of generic gentamicin products.

The graphs illustrate the non-significant differences on the innovator's pharmacodynamic profiles when a single (within-batch variation, Panel A), or a second (between-batch variation, Panel B), or multiple lots (total variation, Panel C) are used to treat E. coli SIG-1 in vivo. In all graphs showing data obtained with the animal model every data-point represents one mouse (mean of two thighs; if standard errors of the mean were to be included, the bars would not be longer than 0.5 logs).

Figure 2. In vivo pharmacodynamics.

Panels illustrate the dose-response curves derived from the neutropenic mouse thigh infection model of 19 generic products and the reference powder with (A) or without (B) therapeutic equivalence respect to the innovator.

Figure 3. Unpredictability of therapeutic equivalence from pharmaceutical equivalence.

The graph illustrates the dose-response curves of gentamicin made by three well-reputed makers: Abbott, Sigma and S. Plough. Abbott and Sigma were indistinguishable from S Plough in terms of concentration and potency of the active pharmaceutical ingredient, MIC, MBC, MBC/MIC ratios but significantly different in terms of therapeutic efficacy, although the same batch of each product was tested in vitro and in vivo.

Figure 4. Results from survival experiments.

Log-rank test curves obtained from neutropenic mice infected in the thighs with P. aeruginosa GRP-0019 and treated during 4 days with placebo (n = 5), GNT-Recipe (n = 10), or the innovator of gentamicin (n = 10) at the dose required for maximal effect (768 mg/kg per day divided q6h), starting 2 h (panel A) or 6 h (panel B) post-infection. Uninfected neutropenic mice serving as toxicity controls received the same treatment and were identical to the other animals but, instead of P. aeruginosa, were mock-inoculated in the thighs with sterile saline (n = 5 mice per gentamicin product). No significant impact on survival was detected between both gentamicin products.

Figure 5. Bacterial dissemination to distant vital organs during survival experiments.

After the fourth day of treatment (same experimental design described in Figure 4) the animals that survived were euthanized and their thighs, lungs, spleens and kidneys processed for bacterial counting (mice found dead were processed immediately). Although thigh counts are not illustrated in the graph, GNT-Recipe left >100,000 CFU/thigh in 80% of the animals, not different from placebo treatment (100%), but significantly different from GNT-S Plough, the innovator product (P<0.0001).