Rethinking the Application of Pemetrexed for Patients with Renal Impairment: A Pharmacokinetic Analysis

Pemetrexed is an antifolate drug used for the chemotherapeutic treatment of non-small cell lung cancer (NSCLC), mesothelioma and thymoma [1‐3]. A single intravenous dose of 500 mg/m² is administered every 21 days. Although pemetrexed is excreted in the urine [4], dosing recommendations do not include adjustment for renal function. Due to fatal toxicities in a study of pemetrexed in patients with renal impairment, pemetrexed is currently contraindicated when the estimated creatinine clearance (CR_CL) is < 45 mL/min [5].

Approximately 25% of the lung cancer population has a CR_CL <60 mL/min [6]. Since it is unclear what the well tolerated pemetrexed dose is for patients with impaired renal function, a large group is withheld effective treatment. Understanding the relationship between dose, renal function, pharmacokinetics (PK), toxicity and treatment outcome is essential to enable treatment in patients with impaired renal function and to prevent toxicity in patients who are already treated with pemetrexed.

Existing data on the effect of renal dysfunction on pemetrexed PK are conflicting. In phase I studies, the manufacturer showed that 70–90% of the pemetrexed dose is excreted in the urine as unchanged drug within 24 h after administration, through both tubular secretion and glomerular filtration [4, 5]. However, a large population PK study by the manufacturer, published by Latz et al. in 2006, in which the PK data of 10 phase II trials were pooled for analysis, showed that renal elimination contributed only approximately 50% to the clearance of pemetrexed [7]. Notably, this study and other more recent pemetrexed PK studies excluded patients with moderate to severe renal dysfunction (CR_CL <45 mL/min) [8, 9]. Thus, extrapolation of these studies to patients with impaired renal function can be questioned. Therefore, the purpose of our study was to investigate the PK of pemetrexed in patients with renal impairment.

This thorough PK analysis of pemetrexed in a population that included patients with impaired renal function, led to two major findings. First, we found that renal function is more important for pemetrexed clearance than the 50% contribution previously described by the manufacturer [16]. The manufacturer’s analysis did not include patients with impaired renal function (CR_CL < 45 mL/min). We showed, in a representative population, that CL_R accounts for up to 84% of total pemetrexed clearance, which is in line with the manufacturer’s early mass balance studies showing that 70–90% of the administered dose could be recovered in urine [4]. Thus, impaired renal function has a more pronounced impact on pemetrexed exposure than previously thought. Our second important finding relates to the disposition of pemetrexed. To date, it has been found that pemetrexed distributes over two compartments [7‐9]. The data used in this analysis included sampling up to 72 h after administration and this revealed the presence of a third compartment. A third compartment can suggest the presence of extravascular fluid or a difference in redistributing tissues. It is unknown what holds true for pemetrexed, but as this is a hydrophilic drug, extravascular distribution is plausible. However, Dickgreber et al. showed no effect of third space fluid on the PK and toxicity of pemetrexed [17]. This implicates that during drug washout, there can be prolonged exposure to higher concentrations of pemetrexed than previously thought. The driving mechanism for pemetrexed toxicity is the subject of discussion. Mita et al. showed no correlation between renal function (and thus exposure) and non-haematological toxicities [5]. With regard to haematotoxicity, it is hypothesized that neutropenia is associated with the total exposure (AUC) [18, 19]. Based on this, it has been suggested that the dose should be adjusted to reach a target based on renal function, instead of BSA [7, 19, 20]. AUC-based dosing in a 21-day cycle is also routinely applied for carboplatin, where CR_CL and the desired AUC are used to calculate the patient’s individual dose [21]. It is unknown whether this hypothesis holds true for pemetrexed, as it is known that for other antifolate drugs, such as methotrexate, haematological toxicity is threshold-driven [22]. For example, in an early phase I trial, it was shown that prolonged exposure to low concentrations of pemetrexed from daily administration resulted in severe neutropenia as a dose-limiting toxicity, with a maximum tolerated dose of only 4 mg/m² for 5 consecutive days without supplementation of vitamin B12 or folic acid [23]. The predominant role of time above the threshold concentration in determining toxicity is supported by the observation that the maximum tolerated dose of pemetrexed administered in a 21-day cycle, also without vitamin supplementation, was markedly higher (600 mg/m²) [4]. Threshold-driven toxicity will be an issue, particularly in renal impairment, as clearance becomes so low that the pemetrexed plasma concentration exceeds the toxicity threshold for a prolonged period. This would explain why previous studies with pemetrexed in patients with renal impairment [5] were not successful. Currently, the PK determinant for the efficacy of pemetrexed is a topic of discussion. Dose adjustment to reach an AUC target will probably entail toxicity concerns when there is low systemic clearance, for example due to renal dysfunction. To allow safe and effective treatment in renal impairment, innovative interventions are needed to overcome toxicity. For example, rescue therapy with folinic acid, as widely applied with pemetrexed’s structural analogue methotrexate [24], may be a feasible option.

A limitation of this study is that the number of patients with severe renal impairment (eGFR <30 mL/min) was limited due to the toxicity concerns that arose during the conduction of the phase I study. Nonetheless, as it stands, these data are the only currently available data to elucidate the clinical PK of pemetrexed in patients with impaired renal function.

Overall, we found that the contribution of renal function was greater than previously thought and that a third compartment may contribute to prolonged exposure during drug washout. Since both factors may contribute to pemetrexed toxicity, they should be accounted for when developing dosing strategies for pemetrexed in patients with renal impairment.

The present PK model can be used to further unravel the PK–toxicity relationship of pemetrexed. In parallel, we must think of innovative strategies to overcome the haematological toxicity of pemetrexed in patients with impaired renal function, such as rescue therapy with folinic acid.