A pharmacokinetic investigation of chloroquine analogues in cancer autophagy modulation

Abstract

Hydroxychloroquine (HCQ) is currently being investigated for safety and efficacy as an autophagy inhibitor in Phase I/II cancer clinical trials. It is the only clinically-approved autophagy inhibitor for use in cancer clinical trials in the United States. HCQ is used in combination with other chemotherapeutics to augment their efficacy and has shown moderate success in treating patients with late stage cancers. While HCQ has a good safety index and shows promise as an addition to standard of care treatment regimens, it suffers from several critical pharmacologic shortcomings which we take steps to address herein. The primary issues with usage of HCQ addressed in this work are the metrics used to predict patient tumor concentration of the drug following various dosing regimens. HCQ pharmacokinetics (PK) are highly variable in patients, with no correlation between traditional plasma:tumor concentrations. The first step taken to address this problem is to characterize likely sources of interindividual variability in HCQ PK. To do this a physiologically-based pharmacokinetic (PBPK) model was developed to investigate absorption, distribution, metabolism, and toxicity (ADMET) factors relating to HCQ in a mathematical system representative of the human body. This model was developed based on physiological and biochemical parameters relevant to HCQ ADMET in mice and scaled to represent humans. The model was capable of simulating single and multiple dosing regimens in humans, that would be characteristic of a cancer clinical trial. PBPK modeling addressed variability that would be associated with the macrophysiologic scale, but intrinsic and extrinsic factors on the cellular scale needed to be further defined to strengthen the understand of HCQ PK. To investigate factors that affect cellular uptake and sub-compartment localization of HCQ, a base PK model of lysosomotropic agents like HCQ was applied. Model specific parameters were identified for a panel of four human breast cancer cell lines, and a majority of differences in cellular uptake of the drug could be attributed to differences in the relative lysosomal volume fraction of each cell line. The model was able to characterize HCQ PK under different extracellular pH conditions, and identified a positive-feedback loop, related to transcription factor EB (TFEB) activation. This feedback loop caused the cell to increase its lysosomal volume over time of exposure to HCQ, resulting in a continuous increase of HCQ concentration within the cell. Through sensitivity analysis of the model, acidic extracellular pH was identified as a critical limiting factor of HCQ uptake into cells – which is particularly important as the tumor microenvironment is physiologically acidic. HCQ concentrations in cells cultured in an acidic microenvironment are decreased up to 10-fold, which cannot be overcome without the aid of agents that neutralize this pH. Dimeric analogues of HCQ, Lys05 and DC661, have been reported to maintain potency in acidic conditions and so were investigated in a comparative context to HCQ. Lys05 and DC661 were found to behave similarly to HCQ pharmacokinetically – i.e. highly dependent on the lysosomal profile of the cell. These drugs exhibited similar kinetic uptake curves as HCQ, and also induced the lysosomal biogenesis PK feedback loop. Unlike HCQ, Lys05 and DC661 uptake was not completely inhibited by acidic extracellular pH, and they were able to maintain activity under these conditions. PK of these drugs was characterized in a murine model to investigate their potential as in vivo agents, suggesting they could maintain high concentrations for a longer duration than HCQ. Lys05 and DC661 share many pharmacologic similarities to HCQ, while not sharing significant shortcomings such as inactivity under acidic extracellular conditions suggesting they should be investigated for further application as next generation autophagy inhibitors.