Date Published: July 24, 2017
Publisher: Public Library of Science
Author(s): Jeannine M. Refos, Alieke G. Vonk, Marian T. ten Kate, Kimberly Eadie, Henri A. Verbrugh, Irma A. J. M. Bakker-Woudenberg, Wendy W. J. van de Sande, Olaf Kniemeyer.
Caspofungin (CAS) which is used as salvage therapy in patients with invasive pulmonary aspergillosis (IPA) inhibits the 1,3-β-D-glucan synthesis in Aspergillus fumigatus. Inhibiting 1,3-β-D-glucan synthesis induces a stress response and in an invertebrate model it was demonstrated that inhibiting this response with geldamycin enhanced the therapeutic efficacy of CAS. Since geldamycin itself is toxic to mammalians, the therapeutic efficacy of combining geldamycin with CAS was not studied in rodent models. Therefore in this study we investigated if the geldamycin derivate 17-(allylamino)-17-demethoxygeldanamycin (17-AAG) was able to enhance the therapeutic efficacy of CAS in vitro and in our IPA model in transiently neutropenic rats. In vitro we confirmed the earlier demonstrated synergy between 17-AAG and CAS in ten A. fumigatus isolates. In vivo we treated A. fumigatus infected neutropenic rats with a sub-optimal dose of 0.75 mg/kg/day CAS and 1 mg/kg/day 17-AAG for ten days. Survival was monitored for 21 days after fungal inoculation. It appeared that the addition 17-AAG delayed death but did not improve overall survival of rats with IPA. Increasing the doses of 17-AAG was not possible due to hepatic toxicity. This study underlines the need to develop less toxic and more fungal specific geldamycin derivatives and the need to test such drugs not only in invertebrate models but also in mammalian models.
Invasive pulmonary aspergillosis (IPA), mainly caused by the fungus Aspergillus fumigatus, is a difficult to treat, life-threatening fungal infection observed in severely immunocompromised patients. Mortality rates up to 50%-90% are observed in these patients . The standard therapy for IPA remains voriconazole, an inhibitor of the ergosterol synthesis. Unfortunately, voriconazole resistance is emerging necessitating the search for novel therapeutic options [1, 2].
To determine the efficacy of a novel therapeutic regimen, it is important to tests it’s efficacy in models in which the features observed in humans are reproduced . Our rat model, characterized by prolonged severe granulocytopenia, inoculation through the respiratory route, fungal broncho- and angio-invasion and dissemination of the fungus from the lung to other organs, closely mimics the pathology observed in human IPA . Within this model, it was possible to study the therapeutic efficacies of voriconazole, CAS and anidulafungin in human pharmacokinetic equivalent dosages, further improving the predictive value of this model [16, 17, 21]. We therefore used this rat model to determine if the therapeutic efficacy of the combination geldamycin and CAS observed in the invertebrate G. mellonella model  could be confirmed in a model mimicking the human disease more closely [16–23]. Although Cowen et al. demonstrated therapeutic efficacy when CAS was combined with Hsp90 inhibitor geldamycin, in their G. mellonella larvae infected with a lethal dose of A. fumiatus, we could not confirm this result in our unilateral IPA model in rats. The difference in outcome could have been the result of the difference of host, but also to the difference in the drugs used in combination. Due to the known hepatoxicity of geldanamycin for mammals  we used its 17-AAG derivative, which differed from the experimental set-up of Cowen et al in the G. mellonella model . Furthermore, the dosages that were used by Cowen in her larval model, greatly exceed the dosages which can be used in mammalian models without toxic side effects. In the larval model dosage as high as 50 mg/kg geldanamycin were tolerated. In our study in rats, the maximum tolerated dose was 1 mg/kg 17-AAG, whilst a five-fold higher dose of 17-AAG resulted in severe toxicity, limiting further studies using higher dosages of 17-AAG. In our study, we did not investigate doses in the range between 1–5 mg/kg, which might increase rat survival without toxic side effects.