Background: The presence of elevated systemic inflammation in people with advanced non-small cell lung cancer (NSCLC) is associated with significantly shorter survival following carboplatin-based chemotherapy.
Objective: This study investigated whether novel factors, such as systemic inflammation [platelet-lymphocyte ratio (PLR) and neutrophil-lymphocyte ratio (NLR)], impact carboplatin pharmacokinetics and drug utilisation. The study also examined the ability of current and alternate dosing regimens to meet therapeutic targets.
Methods: Seventy-two people with advanced NSCLC treated with carboplatin-based (460-1050 mg) doublet chemotherapy were recruited and pharmacokinetic data (n = 61) were analysed using non-linear mixed modelling. Covariate analysis was performed to investigate the impact of standard and novel patient characteristics of carboplatin pharmacokinetics. A Monte Carlo simulation of 100,000 representative NSCLC patients evaluated the ability of the Calvert formula and novel dosing strategies to achieve the targeted therapeutic range. The associations between systemic inflammation and chemotherapy drug utilisation (cycles received, relative dose intensity (RDI) and second-line uptake) and clinical endpoints were also investigated in the pharmacokinetic cohort, and two independent cohorts of people with advanced NSCLC from the Chemotherapy Dosing in Cancer-Related Inflammation (CDCRI) database that were administered carboplatin-paclitaxel (n = 37) or carboplatin-gemcitabine (n = 358).
Results: In all cohorts, 25-53% of people had elevated systemic inflammation (NLR > 5 or PLR > 300). In the pharmacokinetic cohort, no patients achieved the desired therapeutic target of carboplatin. Carboplatin exposure was related to renal function, as estimated using the Cockcroft-Gault formula, albumin and inflammation (NLR). In the pharmacokinetic cohort, increasing carboplatin area under the curve (AUC) correlated with greater reductions in red blood cells and haemoglobin. In this cohort, the average measured AUC of partial responders was 2.4 mg·min/mL. Also in the pharmacokinetic cohort, only 12% of people with an NLR > 5 received four or more cycles of chemotherapy, compared with 62% of patients with an NLR ≤ 5 (p < 0.001). For people in the CDCRI cohort receiving carboplatin-gemcitabine, those with an NLR > 5 also received less cycles (four or more cycles, 41% vs. 60%; p < 0.01) as well as less second-line chemotherapy (46% vs. 60%; p = 0.02) compared with patients without inflammation. People in the pharmacokinetic cohort with an NLR > 5 had 12 months less median survival compared with people with an NLR ≤ 5 (6.5 vs. 18 months; p = 0.08). Similarly, overall survival was significantly shortened in people in the CDCRI cohort receiving carboplatin-gemcitabine with an NLR > 5 compared with those with an NLR ≤ 5 (7 vs. 12 months; p < 0.001), and Cox regression analysis showed a 1.5-fold (1.3-2.1; p < 0.001) increased hazard of death associated with the increased systemic inflammation. Simulations of the newly developed model-based and Calvert dosing assessed the ability to reach this study's proposed actual target AUC of 2.2-2.6 mg·min/mL. These showed current Calvert dosing was predicted to result in substantial overexposure in patients with high systemic inflammation. The newly developed model showed equivalent levels of carboplatin therapeutic target achievement across the spectrum of inflammation observed in the lung cancer population.
Conclusion: An alternate model-based dosing strategy for carboplatin was developed and is predicted to result in consistent drug exposure across the population and improve attainment of therapeutic targets. Further studies of this new model are warranted in people with advanced NSCLC.