A multicenter, phase I, pharmacokinetic study of osimertinib in cancer patients with normal renal function or severe renal impairment

Abstract Osimertinib is a third‐generation, irreversible, oral epidermal growth factor receptor (EGFR)‐tyrosine kinase inhibitor (TKI) that potently and selectively inhibits both EGFR‐TKI sensitizing and EGFR T790M and has demonstrated efficacy in non‐small cell lung cancer (NSCLC) central nervous system metastases. In this phase I study, we assessed the effects of normal renal function (NRF) and severe renal impairment (SRI) on the pharmacokinetics (PK) of osimertinib in patients with solid tumors. Part A: patients with NRF (creatinine clearance [CrCL] ≥90 mL/min), and SRI, (CrCL <30 mL/min), received a single 80‐mg oral dose of osimertinib and standard PK measures were assessed. Part B: patients with SRI were treated for 3 months to obtain safety data, if deemed clinically appropriate. The geometric mean osimertinib plasma concentrations were higher in patients with SRI (n = 7) vs NRF (n = 8) and were highly variable. Osimertinib exposure based on C max and area under the plasma concentration‐time curve, was 1.19‐fold (90% CI: 0.6, 2.0) and 1.85‐fold (90% CI: 0.9, 3.6), respectively, higher for patients with SRI vs patients with NRF, with no clear correlation between CrCL and exposure. No new safety signals were identified after 12 weeks of osimertinib 80 mg continuous dosing. PK parameters pooled across this study and other phase I, II, and III osimertinib clinical studies (exploratory population PK analysis), showed minimal correlation between CrCL and total clearance. In conclusion, no dose adjustment is required for osimertinib for patients with SRI.


| INTRODUC TI ON
Osimertinib is a third-generation, irreversible, oral epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitor (TKI) that potently and selectively inhibits both EGFR-TKI sensitizing (EGFRm) and EGFR T790M and has demonstrated efficacy in non-small cell lung cancer (NSCLC) CNS metastases. [1][2][3][4][5][6] Osimertinib is currently approved in 84 countries, for the treatment of patients with locally advanced or metastatic EGFR T790M mutation-positive NSCLC, and in 75 countries for use as first-line treatment of patients with locally advanced or metastatic NSCLC whose tumors have EGFR exon 19 deletion or p.Leu858Arg EGFR mutations (with country-specific variations). [7][8][9] In patients with EGFRm NSCLC, osimertinib exposure, maximum plasma concentration (C max ), and area under the concentration-time curve (AUC) increase with dose proportionally from 20 to 240 mg/ day after single and multiple dosing. [10][11] The mean half-life of osimertinib is ~48 hours and visual observations of trough levels indicate steady-state is generally achieved by 15 days of dosing, consistent with single-dose pharmacokinetics (PK). 10 In vitro reaction phenotyping studies indicate that CYP3A4/5 are , with a smaller contribution by renal clearance. 12 Both metabolites are potentially active; however, they each circulate at levels ~10% of that seen with osimertinib. 10 In a 14 C-osimertinib mass balance study, after a single 20 mg oral dose, ~68% of the dose was eliminated in feces and 14% in urine, with unchanged osimertinib accounting for <2% (0.8% in urine and 1.2% in feces) of the dose. [8][9]12 In a hepatic impairment study, osimertinib exposure was not increased due to mild or moderate hepatic impairment (Child Pugh A or B). 13 Although osimertinib urinary excretion is low and renal impairment is not expected to have a significant impact on the PK of osimertinib, it has been observed that severe renal impairment (SRI) can impact the exposure of many compounds that are not primarily eliminated renally. 14 In a previous population PK analysis of osimertinib in 593 patients with mild renal impairment (creatinine clearance [CrCL] 60 to <90 mL/min), 254 patients with moderate renal impairment (CrCL 30 to <60 mL/min), five patients with severe renal impairment (CrCL 15 to <30 mL/min) and 502 patients with normal renal function (NRF; CrCL ≥90 mL/min), osimertinib exposures were similar. 8 Across osimertinib clinical trials, which were part of this population PK analysis, data from patients with SRI (n = 5) were very limited and at the time of the initiation of this clinical study, it was even lower (n = 3). As osimertinib may be used by patients who suffer from varying degrees of renal impairment, it is important to define the effects on the PK of osimertinib, to determine whether it is necessary to develop dose adjustment recommendations and, thereby, ensure appropriate use.
Here, we report the results of Part A (single-dose PK phase) and Part B (continued dosing 12-week safety phase in patients with SRI) of a three-part phase I trial (NCT02923947) designed to characterize the impact of SRI on the PK of osimertinib and its metabolites (AZ5104 and AZ7550) in patients with advanced solid tumors. Severe renal impairment group: n = 7

Excluded (n = 2)
Terminated due to physician decision: n = 2 Excluded (n = 2) Terminated due to AE: n = 3   Hepatic function was evaluated at baseline and throughout the study, by analysis of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and bilirubin levels.

| Safety and tolerability
Safety assessments included adverse event (AE) reporting graded by Common Terminology Criteria for Adverse Events (version 4.0), physical examination, vital signs, electrocardiogram (ECG), ophthalmological examination, clinical chemistry, hematology, and urinalysis.

| Sample collection and bioanalysis
Plasma samples were collected for PK analysis predose and at 1, 2,

| Pharmacokinetic analysis
The PK parameters were derived using non-compartmental methods

| Statistical methods
To provide adequate PK information and to assess the effects of

| Population pharmacokinetic analysis
An additional assessment of the impact of renal impairment on the PK of osimertinib was performed using a population PK analysis and the surrogate marker of CrCL as determined by the C-G formula. Previously published population PK analysis methods were updated with additional clinical data from this study. 17 The population PK data set included patient data from phase I, II, and III osimertinib trials AURA (NCT01802632), 11 this study, and five as mentioned previously). For each clinical trial dataset, analysis of variance (ANOVA) statistical analysis was performed and mean differences between groups and respective confidence intervals for each dataset were calculated.

| Patients
In this analysis, 26 patients were screened, and 16 were enrolled and assigned to treatment    Figure S1 Part A and B, respectively. On average, PK parameters showed a higher AUC and C max for osimertinib for patients with SRI vs those with NRF Table 2. Osimertinib exposure, based on C max and AUC, was 1.19-fold and 1.85-fold, respectively, for patients with SRI relative to patients with NRF.

| Osimertinib pharmacokinetics
Because of the limited sample size, the 90% CIs were wide and included unity (1.00 Table 2).
Terminal half-life (t 1/2λZ ) and time of maximum concentration (t max ) were longer in patients with SRI relative to patients with NRF

| Osimertinib metabolite pharmacokinetic parameters
The metabolite to parent ratios for C max and AUC were similar for patients with SRI and for patients with NRF, and overall amounted to less than 11% of osimertinib exposure

| Population pharmacokinetic analysis
A graphical and a statistical analysis was performed which included osimertinib AUC was plotted as a function of renal impairment.
Inclusion of the present results along with the population PK analysis results showed that the data in the present study had a similar range of exposures Figure 3. Individual osimertinib AUC ss was plotted as a function of baseline CrCL which showed no clear relationship between baseline CrCL and osimertinib exposure Figure 4. A comparison of osimertinib PK parameters after single-dose administration across the clinical study program is shown in Supplementary Section and Table S2.
The linear regression analysis showed that the correlation between CL/F (total apparent clearance) and baseline CrCL is weak (R 2 less than .02) and this correlation appears to be similar while including and excluding data from this study. This suggests that the PK analysis regarding the influence of CrCL to the overall clearance of osimertinib.
A combined statistical analysis of 12 patients with SRI vs NRF (n = 508) shows that osimertinib exposure, based on AUC ss , was 1.48-fold higher for patients with SRI. This was lower than the 1.85fold increase in patients with SRI, which was observed based on this study alone.
A matched comparison analysis confirmed that the increase in exposure observed with SRI shows a median of 1.26-fold increase in patients with SRI, vs patients with NRF, when a random dataset of 12 patients with NRF were selected from the population PK dataset (age, sex, and body weight were matched as per study criteria)

| Safety
In Part B, the mean total treatment duration was 53.7 days (standard

| D ISCUSS I ON
The present study was designed to assess the impact of SRI on the PK of osimertinib in patients with advanced solid tumors. The study was designed in accordance with Food and Drug Administration (FDA) and EMA guidance on the assessment of PK in patients with impaired renal function. 16,20 In line with the FDA and EMA guidance, a reduced PK study design involving single-dose PK analysis in patients with SRI and those with NRF was considered appropriate.
This was based on the results of prior PK analyses showing that osimertinib and its active metabolites exhibit linear, dose-proportional (20-240 mg) and time-independent PK with steady state predictable from a single dose. 10,17 Furthermore, based on the population PK analysis of patients across the osimertinib clinical studies, mild-tomoderate renal impairment had no impact on the plasma clearance of osimertinib, therefore it was considered unlikely that SRI would have a significant impact on the PK of osimertinib. 17 Due to limited safety data available in patients with SRI, Part B (3 months' daily dosing) of the study was included to enable further understanding of the safety characteristics of this patient population. A comprehensive review of the concomitant medications showed no patients were taking strong inducer of CYP3A4/5 or any other drug that could impact the analysis during the PK phase of the study.
Our results showed that osimertinib and AZ5104 exposure in patients with SRI, based on AUC, was 1.85-fold and 1.62-fold higher relative to that of patients with NRF. In contrast, AZ7550 AUC was approximately 0.74-fold relative to patients with NRF. Osimertinib C max in patients with SRI was 1.19-fold higher while AZ5104 and AZ7550 C max were 0.86-fold and 0.57-fold relative to patients with normal function. For all analytes, between-patient variability in exposure was high; the 90% CIs in the ANOVA comparison were wide and, with the exception of AZ7550 C max , all 90% CIs included unity.
Renal clearance for all three analytes was negligible in patients with SRI and in patients with NRF. Hence, changes in renal clearance are unlikely to account for the higher exposure observed in SRI patients.
Similar to other studies which have shown renal impairment to affect the function of CYP3A enzymes, here the changes to AZ5104 and AZ7550 (which are primarily metabolized by CYP3A) appear to be qualitatively similar to that seen with the strong CYP3A inhibitor, itraconazole. [21][22] The mean exposure to osimertinib and its metabolites (AZ5104

Severe 12 201
Note: Relationship between renal function (CrCL calculated on baseline) and natural log-transformed osimertinib PK parameters AUC ss is presented here.
Results are based on an ANOVA model with a fixed effect for renal function group.
a Three subjects without PK exposure are not included in this analysis. In patients with SRI, the higher osimertinib exposure cannot be explained by changes in protein binding. In general, renal impairment may be associated with an increase in unbound drug con- In patients with SRI, the higher osimertinib exposure may possibly be due to the limited number of patients evaluated in this study (seven in this study and five in the population PK analysis, of a total of ~1400). This was due to difficulties in recruiting eligible patients and is a common limitation of such trials. Guidance from the EMA suggest that a population of 6-8 patients per group is required to provide adequate PK data, 16 thus, while the study population was small, it was sufficient to assess the effects of SRI on the PK of osimertinib and the results we observed were reflected in the overall population PK population. Renal impairment did not meaningfully alter AZ5104 metabolite/ parent ratios for C max and AUC, which accounted for less than 11% of osimertinib exposure. AZ7550 metabolite/parent ratios for C max and AUC were lower in patients with SRI; metabolite AUC amounted to less than 2% (SRI) or 8% (NRF) of the exposure to osimertinib.
As AZ7550 accounts for <10% of the exposure of osimertinib with similar in vitro pharmacological properties, the changes in AZ7550 metabolite/parent exposure ratio are not considered of clinical relevance.
Based on the AURA phase I study, clinical activity was demonstrated at all doses studied (20 to 240 mg in T790M population and at both 80 and 160 mg in the first-line population), with no maximum tolerated dose reached at the 240 mg dose. 11,17 It is also clinical practice to reduce osimertinib dose to 40 mg to manage drug-related toxicities, while keeping adequate osimertinib efficacy dose levels. As such, it has been established that increases in mean osimertinib exposure of less than 2-fold (ie, less than exposure equivalent to a 160 mg dose) and decreases of no more than 50% (ie, greater than that achieved at a 40 mg dose) would require no dose adjustments as it will unlikely have any clinically meaningful impact on efficacy or safety. 17 Osimertinib shows inter-patient PK variability on CL/F (45% between subject variability) 23 and, as such, any changes in exposure that is less than 2-fold (lower than that observed with the 160 mg dose) is unlikely to alter the benefit:risk ratio.
In conclusion, mean osimertinib PK exposure was higher in patients with SRI vs patients with NRF. However, a lack of correlation between exposure and the CrCL, together with a similar and consistent known safety profile of osimertinib after single and multiple dosing in patients with SRI, indicate that no dose adjustment for osimertinib is required when treating patients with SRI.
Nevertheless, as the mean exposure change due to SRI approached an almost 2-fold increase in the present study, a proper clinical assessment and continuous monitoring in patients with severe and end stage renal disease should be considered.

DATA S H A R I N G A N D ACCE SS I B I LIT Y
Data underlying the findings described in this article may be ob-