Tackling PK Variability in Cancer Chemotherapy
PrecisePK is making precision oncology affordable and easy for your healthcare facility. We are adding the first of many chemotherapy drugs, Carboplatin, in our drug library with other antineoplastic drugs such as Busulfan and Methotrexate soon to be followed.
PrecisePK leverages longitudinal real-world data set, population PK model knowledge, and Bayesian dosing plate form to individualize chemotherapeutic treatment by tailoring the dose to the patient-specific characteristic. Individual differences can alter patients’ responses. This helps maximize efficacy, avoid excessive toxicity, ensure improved patient outcomes, and accelerate life-saving treatment. We build population PK models and employ Bayesian forecast to address interpatient pharmacokinetic variability and to help clinicians quickly calculate an effective individual dose and adjust chemotherapy dose with feedback serum levels, thus enhance clinical decision making in the complex arena of cancer care. Precise PK uses Bayesian approach to facilitate the implementation of TDM in the oncology setting and makes precise dosing practical in a clinical environment. This is critical to quantum leaps in cancer care and creating a future where treatment can be effective for everyone.
There is growing evidence supporting the need for TDM that enables clinicians to provide precise dosing and optimize anticancer therapy [1,2,3]. Unlike antibiotics or immunosuppressive drugs that employ continuous dosing, anti-cancer agents are typically dosed cyclically, and require multiple blood samples to be drawn over time for clinical pharmacokinetic analysis. Individualized precision dosing would benefit cancer patients with unique drug metabolic responses, especially when the intention to treat is curative. It is also especially important in oncology drugs that exhibit steep dose-response relationship, narrow therapeutic index, and a large degree of inter-patient PK and PD variability. Underdosing can lead to therapeutic failure and cancer progression while overdosing results in severe and life-threatening hematologic cytotoxicity such as agranulocytosis, and neutropenia
PrecisePK goes beyond Calvert Formula:
Exclusively using Calvert Formula to dose Carboplatin presents various challenges and may result in inaccurate dosing. Overestimation of renal function could result in a higher than intended carboplatin dose. Furthermore, the applicability of this dosing strategy in a high dose setting is not well-known, wherein protein binding saturation may alter the pharmacokinetics [8,9], and like other static models, it is subject to some degree of imprecision due to the lack of incorporated feedback serum drug concentration. The Calvert Formula is shown below.
Carboplatin Dose (mg) = Target AUC (mg∙min/mL) x (GFR + 25) GFR estimated by calculated creatinine clearance using Cockcroft-Gault Equation
Intravenous (IV) Carboplatin (Paraplatin®) is a platinum-based alkylating compound that works via interfering with DNA synthesis and cell replication. It is one of the most active cytotoxic antineoplastic agents that is widely used in both the adult and pediatric populations [4]. Carboplatin is FDA-approved for use in the initial and secondary treatment of Advanced Ovarian Carcinoma but has a broad spectrum of activities in various malignancies including thoracic cancer, germ cell cancer, head and neck cancers, and bladder cancer [5]. Its dose-limiting toxicity is myelosuppression, which is closely related to the renal clearance of carboplatin [6]. Unlike other conventional cytotoxic agents that are dosed based on body surface area (BSA), carboplatin is dosed to an AUC target based on renal function, commonly described by the Calvert Formula [7]. According to this method, since carboplatin is cleared via glomerular filtration, AUC is the most important pharmacokinetic predictor of clinical response. The Calvert method is most accurate when the GFR is measured using Cr-EDTA, which can be inconvenient to collect in clinical practice and therefore, is rarely used.
Implementation of TDM with Bayesian dosing to optimize Carboplatin exposure
There are well-established literature evidence and prospective clinical evaluation demonstrate that the Bayesian method provides superior accuracy for carboplatin AUC prediction by using feedback concentration levels and pharmacokinetically guiding individualized carboplatin dosing. This method leads to a marked reduction in variability of exposure and allows the target carboplatin AUC to be reached accurately while minimizing toxicities [10]. Peng et al. [11] showed that limited sampling strategies and Bayesian forecasting provide more reliable estimation of AUC values when compared to BSA or renal function-based dosing strategies, resulting in the median bias of 5% and precision of 22%. Similarly, adaptive dosing method when applied to high-dose carboplatin chemotherapy in the pediatric population with soft tissue sarcoma or recurrent primitive neuroectodermal tumor allows carboplatin exposures to be achieved within 84-126% of target AUC values [12].
Duffel et.al published the first sequential Bayesian design that has been used for dose individualization of chemotherapy in the Cancer Chemotherapy and Pharmacology. In this study [13,14], 12 ovarian cancer patients were given carboplatin IV infusion over two courses of treatment to achieve an AUC of 5-7 mg*min/mL. Different compartmental models were fit to the data. The best model determines the priori model is subsequently incorporated into the Bayesian model to estimate the Bayesian posterior, a more accurate description of the patient’s true pharmacokinetic profile. The model was tested prospectively against Calvert’s formula using 2 feedback concentrations and was found to outperform the Calvert formula.
