What are the characteristics of the exposure–response relationships for efficacy?

The results of treatment with an anticancer therapy depend on both pharmacokinetics and pharmacodynamics. Pharmacokinetics refers to the relationship between dose and concentration, whereas pharmacodynamics refers to the relationship between concentration (or exposure) and “response,” with response broadly defined as any measurable effect of the drug related to efficacy, toxicity, or neither. Exposure–response relationships help to define the target exposure range at which the drug is effective without causing severe toxicity (often referred to as the therapeutic index). Many anticancer therapies have a narrow therapeutic index.

In general, any drug may be considered to have a maximal effect and a median dose (the dose required for 50% of the maximal effect). Wagner proposed a generalized sigmoidal model of drug effect (Figure 1), based on the hypothesis that all drug effects require an initial interaction with a receptor.²⁵ Pharmacodynamic models are generally different for agents that only work during certain phases of the cell cycle (phase-specific agents) compared with agents that work during any phase of the cell cycle (nonphase-specific agents). For nonphase-specific agents, such as cyclophosphamide and other alkylating agents, a simple log-linear model can be used as follows:

1

This model has a steep exposure–response curve, as the effect continues to increase proportionally as the concentration (C) increases. For any K (in Equation 1), an increase in C by 2.3/K will result in a 1-log increase in antitumor effect (Figure 2a).^{26, 27}

This is intended to be an illustrative rather than a comprehensive list.

^a http://www.accessdata.fda.gov/drugsatfda_docs/nda/2012/203585Orig1s000ClinPharmR.pdf

^b http://www.accessdata.fda.gov/drugsatfda_docs/nda/2011/202429Orig1s000ClinPharmR.pdf

For phase-specific agents, such as the antimetabolites, the exposure–response relationships are much more complicated. By definition, some cells are out of phase and therefore not sensitive (or relatively insensitive) to the effects of the drug during the period of drug exposure. This cannot be overcome by increasing the dose, but can potentially be overcome by increasing the duration of drug exposure.²⁸ The result is the appearance of a plateau in the exposure–response curve (Figure 2b). To further complicate matters, plasma concentrations may be an inadequate predictor of clinical effect for those agents that undergo intracellular anabolism to active metabolites, as is the case for cytarabine and many other antimetabolites.²⁹

Trastuzumab emtansine (T-DM1), an antibody-drug conjugate consisting of a recombinant humanized monoclonal antibody targeting the extracellular domain of HER2 linked to a microtubule-stabilizing agent, is a recent example of a drug for which the exposure–response relationship for efficacy has been well studied. In the case of T-DM1 for the treatment of HER2-positive metastatic breast cancer, higher trough concentrations on day 21 of cycle 1 are associated with improved OS and PFS after adjusting for baseline risk factors. In fact, patients with exposures below the median level had OS and PFS comparable to the active control arm of lapatinib plus capecitabine, suggesting that dose escalation might improve efficacy for this subset of patients.³⁰ Relationships between exposure and efficacy have also been found for a number of other anticancer therapies that are summarized with references in Table 2. Although still controversial, therapeutic drug monitoring has been suggested for many kinase inhibitors on the basis of these exposure–efficacy relationships.⁴³

What are the characteristics of the exposure–response relationships for safety?

Exposure–response relationships for safety are analogous to those for efficacy, except that the phenotypes of interest are potentially severe toxicities of the drug rather than efficacy measures. Many historical examples with cytotoxic therapies highlight the value of elucidating such relationships. The relationship between toxicity subsequent to high-dose methotrexate and that of delayed methotrexate clearance has led to the routine use of therapeutic drug monitoring of plasma methotrexate concentrations to guide leucovorin dosing.⁴⁴ The relationship between paclitaxel and neutropenia (generalizable to other cytotoxic therapies) is best described using nonlinear models that allow for an indirect effect of plasma concentration on bone marrow suppression, and these types of models have informed the optimization of the dose and schedule for many drugs.^{45, 46} Another example of a drug with a very well-characterized exposure–response relationship for safety is carboplatin, an analogue of cisplatin. Unlike cisplatin, the dose-limiting toxicity of carboplatin is thrombocytopenia, which is a function of drug dose, renal function, pretreatment platelet count, and prior therapy.⁴⁷ The platelet nadir produced by a dose of carboplatin is related to the carboplatin clearance, which is directly proportional to creatinine clearance. Thus, patients at high risk of severe thrombocytopenia following carboplatin therapy can be identified prospectively, and the drug doses can be modified by monitoring creatinine clearance. Although many attempts have been made to avoid severe toxicities of cytotoxic therapies by monitoring plasma drug concentrations during treatment, these attempts have generally not led to changes in clinical practice.

Vemurafenib, an oral small molecule inhibitor of the BRAF kinase harboring the V600E or V600K mutations in patients with metastatic melanoma, provides a more recent example of a drug with a well-established exposure–response relationship for safety in addition to an exposure–response relationship for efficacy. (Of note, although the FDA label indicates that the drug may be administered with or without food, exposure is increased when the drug is taken with food.) When exposure is measured by trough concentration at steady state, there is an increased probability of squamous cell carcinoma (SCC; a well-known, treatment-related adverse effect) with increasing exposure in a logistic regression model (p < 0.0001) (http://www.accessdata.fda.gov/drugsatfda_docs/nda/2011/202429Orig1s000ClinPharmR.pdf). Despite this clear relationship, dose reductions for SCC events are not recommended because the survival benefits outweigh this safety risk.

Does this drug prolong the QT or QTc interval?

Many drugs prolong the QT interval by delaying depolarization and repolarization of the ventricles in the heart. The QT interval is corrected for the heart rate by one of two accepted formulas to calculate the heart-rate corrected QT (QTc) interval. QTc prolongation is associated with a higher incidence of ventricular tachyarrhythmias (particularly torsade de pointes) that can potentially lead to sudden cardiac death. As an increased risk of sudden cardiac death is a major safety issue, FDA and other regulatory authorities require that sponsors adequately assess this risk during clinical development of a new drug. In 2005, international guidelines were developed to guide the design and data interpretation of thorough QT studies, and these guidelines were circulated by FDA in a Guidance to Industry (Guidance for Industry: E14 Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-Antiarrhythmic Drugs. FDA, October 2005). In this guidance document, FDA recommends that the ΔΔQTc (change in QTc compared to baseline, corrected for placebo effect) be used as the endpoint for thorough QT studies, and that the upper bound of the one-sided 95% CI (identical to the upper bound of a two-sided 90% CI) of the largest time-matched mean effect of the drug on the QTc interval should exclude 10 ms in order to conclude that there is no significant safety risk related to QTc prolongation.

Nilotinib is an oral small molecule inhibitor of BCR-ABL and other kinases that is used in the treatment of chronic myeloid leukemia (CML). A thorough QT study in healthy volunteers, in accordance with the E14 guidance, was conducted and demonstrated a clear relationship between steady-state concentration and ΔΔQTc, while concentrations in that study were lower than concentrations that can be expected with the approved therapeutic dose in cancer patients (http://www.accessdata.fda.gov/drugsatfda_docs/nda/2007/022068s000_ClinPharmR.pdf). As a result, it is recommended that the drug not be used in patients with hypokalemia, hypomagnesemia, or long QT syndrome who would be at especially high risk for QTc prolongation leading to sudden cardiac death. Furthermore, the FDA required that the sponsor initiate a Risk Evaluation and Mitigation Strategy (REMS) to “minimize the risk of QT prolongation and its potential cardiac sequelae” (http://www.accessdata.fda.gov/drugsatfda_docs/label/2010/022068s001rems.pdf). As part of the REMS and also in a “black box” warning on the drug label, prescribers and patients are cautioned to avoid concomitant use of drugs known to prolong the QT interval as well as CYP3A4 inhibitors and food, both of which are expected to substantially increase exposure to the drug. ECG (electrocardiogram) monitoring is recommended at baseline, 1 week after starting therapy, and periodically thereafter, including after any dose adjustments (Tasigna prescribing information, http://www.accessdata.fda.gov/drugsatfda_docs/label/2007/022068lbl.pdf).

Is the dose and dosing regimen selected by the sponsor consistent with the known relationship among dose, concentration, and response, and are there any unresolved dosing or administration issues?

While the dose and schedule for most anticancer therapies are selected on the basis of dose–concentration–response relationship, there are some for which there are unresolved dosing issues. One example is omacetaxine, a protein synthesis inhibitor that is approved for use in CML after resistance and/or intolerance of two or more tyrosine kinase inhibitors. Although the drug is labeled to be prescribed at an induction and maintenance dose of 1.25 mg/m² because this is the dose that was used in the pivotal study, the reviewers found no evidence of a correlation between clearance of the drug and BSA. As a result, BSA-based dosing results in lower concentrations in patients with smaller body size, such as women, which was associated with a decrease in efficacy (http://www.accessdata.fda.gov/drugsatfda_docs/nda/2012/203585Orig1s000ClinPharmR.pdf). Another example is cabazitaxel, a microtubule-stabilizing agent that is approved for use in hormone-refractory metastatic prostate cancer after previous therapy with docetaxel. In a phase II trial in advanced breast cancer, the initial dose was 20 mg/m² every 3 weeks, and dose escalation to 25 mg/m² was allowed if no grade > 2 toxicity was observed in cycle 1. Even though only 21 of 71 patients could be escalated in cycle 2, the sponsor chose 25 mg/m² every 3 weeks as the dose for the pivotal phase III trial in prostate cancer. On the basis of the available data, there is no clear exposure–response relationship for efficacy but there is a relationship between AUC and probability of grade 3+ neutropenia, suggesting that some patients could avoid the need for dose reduction and growth factor support by starting at a lower initial dose (http://www.accessdata.fda.gov/drugsatfda_docs/nda/2010/201023s000ClinPharmR.pdf). Finally, correct dosing can often be an issue when a drug is studied in a new disease and gets approved for a new indication. In the case of trastuzumab, the approved dose for breast cancer was used in the randomized phase III trial that demonstrated the efficacy of trastuzumab in combination with chemotherapy compared to chemotherapy alone for metastatic HER2-positive gastric/gastroesophageal junction (GEJ) cancer.⁴⁸ Subsequently, a case report and review of pharmacokinetic data from the sponsor have suggested that clearance of the drug is ∼70% higher in patients with gastric cancer than in those with breast cancer.⁴⁹ One possible explanation is that patients with metastatic gastric cancer have higher tumor burden and higher levels of circulating HER2 extracellular domains, resulting in higher clearance of the drug. Another possibility is that patients with gastric cancer may have lower albumin concentrations, leading to higher clearance of the drug. Standard dose versus higher dose trastuzumab is currently being explored in a randomized phase III trial in patients with HER2-positive metastatic gastric/GEJ cancer (HELOISE study; NCT01450696).

What are the single-dose and multiple-dose pharmacokinetic parameters?

Pharmacokinetics is the study of drug absorption, distribution, metabolism, and excretion. A fundamental concept in pharmacokinetics is drug clearance, that is, elimination of the drug from the body, analogous to the concept of creatinine clearance. In clinical practice, clearance of a drug is rarely measured directly but is calculated as either of the following:

2

3

The AUC represents the total drug exposure integrated over time and is an important parameter for both pharmacokinetic and pharmacodynamic analyses. As indicated in Equation (2), the clearance is simply the ratio of the dose to the AUC, so that the higher the AUC for a given dose is, the lower will be the clearance. If a drug is administered by continuous infusion and a steady state is achieved, the clearance can be estimated from a single measurement of the plasma drug concentration (C_ss) as in Equation (3).

Clearance can conceptually be considered to be a function of both distribution and elimination. In the simplest pharmacokinetic model,

4

where V is the volume of distribution and K is the elimination constant. V is the volume of fluid in which the dose is initially diluted, and thus the higher the V is, the lower will be the initial concentration. K is the elimination constant, which is inversely proportional to the half-life, the period of time that must elapse to reach a 50% decrease in plasma concentration. When the half-life is short, K is high and plasma concentrations decline rapidly. Thus, both a high V and a high K result in high clearance and relatively low plasma concentrations.

In most phase I trials of anticancer therapy, pharmacokinetic sampling is done at multiple time points after the first dose of drug administration and is repeated again after multiple doses of drug administration, the latter of which may or may not correspond to steady state. For single-dose sampling, parameters that are presented typically include the maximum concentration observed (C_max), the time at maximum concentration (T_max), AUC, V, clearance, and the half-life. For multiple-dose sampling, parameters are generally the same except for the potential addition of the minimum concentration (C_min) before a dose. Pharmacokinetic parameters can be described by one of two methods. The first is noncompartmental analysis, a two-step process in which parameters such as C_max and AUC are estimated for each patient and summary statistics are calculated for the population. Noncompartmental analysis can be done quickly and does not require special software. The second is population pharmacokinetic modeling, a one-step process in which parameters are estimated for the population of patients by developing a nonlinear mixed effects model with one or more compartments. Population pharmacokinetic modeling has several advantages: (1) it can be done with sparse (fewer time points) sampling data; (2) it is less susceptible to missing data; and (3) it can easily include covariates that minimize the between patient variability. A disadvantage is that it requires special software and training.⁵⁰ The interested reader is likely to benefit from hands on experience with such software. Several caveats need to be emphasized for the casual reader. The validity of pharmacokinetic modeling depends to a large extent on the quality of the data that are used to develop the model. Thus, drug infusions must be precisely timed, a sufficient number of plasma samples must be drawn, the samples must be obtained on schedule, and analytic methods must be sensitive and specific. The data must be properly weighted to avoid bias due to the increased probability of analytic errors at drug concentrations near the detection limit of the assay. Results obtained using a specific model should be compared with those using noncompartmental analysis. Extrapolation of models outside the known time points must be done with great caution.

What are the characteristics of drug absorption?

Variability in absorption is a major source of interpatient variability in pharmacokinetics. The rate and extent of drug absorption depends on its solubility and permeability across the mucosal lining of the intestine. In pharmacokinetic models, absorption is typically estimated by a first-order rate constant, K_a. Absorption may be saturable, meaning that higher doses do not lead to higher exposure above a certain threshold dose. BCS class 1 drugs with high permeability/high solubility are absorbed quickly and extensively, whereas drugs in BCS classes 2–4 have variable absorption and may be influenced by a number of factors. For example, many small molecule tyrosine kinase inhibitors have their absorption limited by membrane efflux transporters such as P-glycoprotein (also known as ABCB1), breast cancer resistance protein (ABCG2), and members of the multi-drug resistance protein family (ABCC family), the same transporters that may be overexpressed in cancer cells as a resistance mechanism.⁵¹ Prandial conditions are also very important, as a number of oral anticancer therapies have enhanced absorption leading to increased exposure (and potentially increased toxicity) when coadministered with food, particularly a high-fat meal. Examples include erlotinib, lapatinib, nilotinib, pazopanib, vemurafenib, and abiraterone, the last of which can have a 10-fold increase in AUC with a high fat meal (compared to an overnight fast).^52–54 These oral anticancer therapies are generally labeled to be taken fasting despite the fact that exposure is increased with food, in contrast to the typical approach for noncancer therapies; theoretically increasing the risks of the drug if patient adherence to labeling is suboptimal.^{53, 55} Finally, a number of oral anticancer therapies are weak bases that have pH-dependent solubility and may have their absorption decreased by concomitant use of proton-pump inhibitors or other acid-reducing agents that are frequently prescribed and/or used over-the-counter. Examples of therapies with clinically significant decreases in exposure when coadministered with acid-reducing agents include dasatinib, erlotinib, gefitinib, and nilotinib.⁵⁶ When these kinase inhibitors are prescribed, patients should be instructed to abstain from the use of acid-reducing agents that might compromise the efficacy of these drugs.

What are the characteristics of drug distribution?

In pharmacokinetic models, there is a volume of distribution for each compartment in the model. Distribution is impacted by a number of factors, most notably water solubility and protein binding. Hydrophilic compounds, such as methotrexate, can distribute into fluid collections (such as pleural effusions or ascites) and significantly delay clearance of the drug from the central (plasma) compartment.⁵⁷ On the other hand, pegylated liposomal doxorubicin is an example of a drug that has been engineered to be hydrophobic. Compared to free doxorubicin, its volume of distribution is very small and is mostly restricted to the intravascular compartment because the tight junctions of endothelial cells in blood vessels prevent extravasation of the liposomes, a fact that likely explains its reduced cardiotoxicity.⁵⁸ Protein binding is very important because only unbound drug can reach the target site and be eliminated. Vismodegib is an example of a drug that has high-affinity reversible binding to alpha-1-acid glycoprotein (AAG) and albumin, resulting in plasma levels of unbound drug that are <1% of total drug levels and contributing to low clearance of the drug. AAG levels account for approximately 70% of the pharmacokinetic variability between patients, suggesting that this is the single most important factor influencing the pharamacokinetics of the drug.⁵⁹ Concomitant drugs or disease states (e.g., hypoproteinemia) might increase the fraction of unbound drug but the clinical significance of this remains unclear for vismodegib and most other drugs.⁶⁰

Does the mass balance study suggest renal or hepatic as the major route of elimination?

Mass balance studies involve the administration of radiolabeled drug to humans. Radioactivity is measured at different time points in the blood, urine, and feces, in order to understand to what extent hepatic metabolism, biliary excretion, and urinary excretion are contributing to elimination. The biological specimens are also used to identify and quantify the parent drug and its metabolites.⁶¹ FDA guidance recommends that any metabolite whose exposure exceeds 10% of the parent AUC at steady state should be evaluated in a preclinical animal model for toxicology (http://www.fda.gov/OHRMS/DOCKETS/98fr/FDA-2008-D-0065-GDL.pdf).

There are many examples of anticancer therapies that are eliminated primarily by renal or hepatic routes. Carboplatin and methotrexate are examples of drugs that are predominantly excreted unchanged in the kidney. This is relevant to carboplatin dosing, which is calculated by a simple formula based on renal function, as shown in Equation (5).⁶²

5

Related posts:

Aberrant signaling pathways in cancer Disparities in cancer care Monoclonal serotherapy Neoplasms in acquired immunodeficiency syndrome Chemical carcinogenesis Skeletal complications

Stay updated, free articles. Join our Telegram channel

Join