Chemotherapy dose--response relationships in non-small cell lung cancer and implied resistance mechanisms.

We hypothesized excess resistance factor ("active resistance") gives a dose--response curve (DRC) shoulder, deficiency of a factor required for drug sensitivity ("saturable passive resistance") gives a DRC terminal plateau, and alteration of a factor gives decreased DRC slope. We used response rates from published non-small cell lung cancer (NSCLC) clinical studies to estimate mean percent tumor cell kill in each study (assuming cell kill is proportional to tumor volume change) and performed regression and meta-regression analyses of percent cell survival and patient survival vs planned dose-intensity. As single agents, cell kill approached that of combinations only at highest doses. While DRC shape varied between single agents, DRCs for all combinations tested flattened at higher doses. Patient median survival times also failed to vary significantly with dose for any combination. DRC flattening at higher doses suggests therapy efficacy is limited by deficiency/saturation of factors required for cell killing. Based on this and other clinical observations, we hypothesize: (1) active resistance may modulate cell killing at lower doses, but ability to overcome this by increasing doses is limited by saturable passive resistance (e.g. by non-cycling cells). (2) Cells surviving initial chemotherapy may upregulate active resistance mechanisms (permitting growth despite therapy). (3) If active resistance mechanisms are insufficient for growth/survival, cells may survive until therapy cessation by downregulating metabolism/cycling, becoming temporarily quiescent. This could help explain broad cross-resistance between agents and would imply that improved targeting of non-cycling cells will be required for major improvement in therapy efficacy.