Background: P-glycoprotein (Pgp), a member of the ATP-binding cassette (ABC) transporter family, is overexpressed in a number of cancers and some studies show that Pgp overexpression can be correlated to poor prognosis or therapeutic resistance. We aimed to elucidate if the efficacy of PF-03758309, a novel p-21 activated kinase inhibitor, could be predicted based on tumor Pgp/MDR1 expression.
Methods: Based on in vitro proliferation data, a panel of cell lines were ranked as sensitive (IC50<15nM) or resistant (IC50> 1μM) and MDR1 (Pgp) gene array data evaluated for the cell lines. Pgp expression was determined by western blot. We then evaluated if Pgp inhibition would restore PF-03758309 activity in vitro. PF-03758309 activity was evaluated in vivo in murine cell line xenograft models and in primary patient derived explants (PDX). Mice were implanted with either tumor cell lines or a fragment of patient tumor tissue and when tumors reached 200–300 mm3, mice were treated with 25 mg/kg PF-03758309 orally, twice daily. On the last day of treatment, tumor and plasma were collected for PF-03758309 analysis. Gene array data and tumor and plasma concentrations were evaluated.
Results: MDR1 gene expression correlated (spearman, p<0.01) with resistance to PF-03758309 treatment in vitro and the expression of Pgp on resistant cells was verified by western blotting. Furthermore, the inhibition of Pgp increased the sensitivity of “resistant” cells, bringing IC50s down to ∼2–200 nM. Eleven cell line xenografts were treated and the MDR1 gene expression was found to correlate (spearman, p = 0.02) with resistance measured by an increased %T/C (treated volume/control volume × 100%). Thirteen PDX models were treated and only two were found to be “sensitive” (%T/C < 50%). Gene array analysis has not yet been performed on PDX treated specimens and baseline specimen data cannot be used because of drift in MDR1 and ABCG2 expression found in these models at different passages. Preliminary tumor and plasma concentration data has been gathered for 2 sensitive and 3 resistant models (1 PDX sensitive, 1 cell line xenograft sensitive, 2 PDX resistant, 1 cell line xenograft resistant). Tumors that responded to treatment generally exhibited a higher tumor concentration level and a higher tumor:plasa ratio (p=0.07, p=0.04). Finally, a preliminary in vivo study demonstrated that inhibition of Pgp prior to administration of PF-03758309 increased the mean tumor:plasma ratio of PF-03758309.
Conclusions: Numerous agents, approved and in development, are substrates for transporters, and it may not be clear yet the role that drug resistance transporters play in resistance to these therapies. In vitro and in vivo data strongly suggests that PF-03758309 efficacy may be influenced by the expression of the tumor multidrug resistance (MDR) protein, Pgp. Pgp mediated resistance has been studied for decades and efforts have been made at sensitizing tumors to compounds that are Pgp substrates by co-administration of Pgp inhibitors, to no avail. A possible alternative to Pgp reversal is simply using patient Pgp status as a negative selective parameter for therapy.