Pharmacokinetic and dose-finding study of osimertinib in patients with impaired renal function and low body weight

Abstract

The safety of osimertinib is limited in patients with severe or moderate renal impairment, or low body weight. This study aimed to investigate the safety, pharmacokinetics (PK) and recommended dose (RD) of osimertinib in patients with epidermal growth factor receptor (EGFR)-mutated non–small cell lung cancer (NSCLC) with impaired renal function and low body weight. Thirty-one eligible patients were enrolled and allocated into four cohorts: A, normal renal function (estimated glomerular filtration rate [eGFR] ≥ 50 mL/min/1.73 m²) and normal body weight (≥45 kg); B, moderate renal impairment (eGFR = 30-50 mL/min/1.73 m²); C, low body weight (<45 kg); and D, severe renal impairment (eGFR <30 mL/min/1.73 m² or undergoing dialysis). PK parameters and safety were evaluated with a starting dose of 80 mg osimertinib administered orally once daily in cohorts A, B, and C and 40 mg once daily in cohort D. The PK parameters in cohorts A, B, and C were found to be similar. No dose-limiting toxicity was observed, and the RD was determined to be 80 mg once daily in patients with moderate renal function and low body weight. Four serious adverse events, acneiform rash, diarrhea, QTc prolongation, and interstitial lung disease, were noted. Although the PK parameters of osimertinib were similar across all cohorts, toxicity occurred more frequently in patients with impaired renal function and low body weight. Clinicians should prescribe osimertinib with caution in NSCLC patients with impaired renal function and low body weight.

Abbreviations
• Ad
• adenocarcinoma
• Adsq
• adenosquamous carcinoma
• AE
• adverse event
• Alb
• albumin
• ALP
• alkaline phosphatase
• ALT
• alanine aminotransferase
• ANC
• absolute neutrophil count
• AST
• aspartate aminotransferase
• AUC
• area under the concentration-time curve
• BMI
• body mass index
• BSA
• body surface area
• BW
• body weight
• CGF
• Cockcroft and Gault formula
• CK
• creatine kinase
• CKD-EPI
• Chronic Kidney Disease Epidemiology Collaboration
• CL/F
• oral clearance
• C_max
• maximum plasma concentration
• COPD
• chronic obstructive pulmonary disease
• Cr
• creatinine
• CrCl
• creatinine clearance
• CTCAE
• Common Terminology Criteria for AEs
• C_trough
• trough concentration
• DLT
• dose-limiting toxicity
• ECG
• electrocardiogram
• ECOG
• Eastern Cooperative Oncology Group
• EGFR
• epidermal growth factor receptor
• eGFR
• estimated glomerular filtration rate
• F
• female
• GFR
• glomerular filtration rate
• H
• hour
• ILD
• interstitial lung disease
• M
• male
• MDRD
• Modification of Diet in Renal Disease
• NRF
• normal renal function
• NSCLC
• non–small cell lung cancer
• PD
• progressive disease
• PK
• pharmacokinetic
• PLT
• platelet count
• PPK
• population pharmacokinetic
• PR
• partial response
• PS
• performance status
• RD
• recommended dose
• RECIST
• Response Evaluation Criteria in Solid Tumors
• SD
• stable disease
• SRI
• severe renal impairment
• TKI
• tyrosine kinase inhibitor
• T_max
• time to maximum concentration
• ULN
• upper limit of normal
• WBC
• white blood cell count

INTRODUCTION

The standard chemotherapy for advanced non–small cell lung cancer (NSCLC) is currently performed based on the molecular subtype of the tumor.^1–4 Epidermal growth factor receptor (EGFR) mutation is the most common druggable mutation in patients with NSCLC, accounting for 30%-40% in East Asians and 10%15% in Westerners. The development of EGFR tyrosine kinase inhibitors (TKIs) is a great success story on how precision medicine can dramatically improve patient outcomes.

Osimertinib, a third-generation oral EGFR TKI, potently and selectively inhibits both EGFR sensitizing and EGFR T790M resistance mutations. Randomized phase 3 trials of AURA3 and FLAURA indicated that osimertinib is the first-line treatment for metastatic NSCLC with EGFR sensitizing mutations and the treatment for NSCLC with T790M mutations that progressed on or after first- or second-generation EGFR TKIs.^5,6 The most common AEs (≥30%) in patients treated with osimertinib were diarrhea (58%), rash (58%), dry skin (36%), and nail toxicity (35%). The most common serious AEs were pneumonia (2.9%) and interstitial lung disease (ILD)/pneumonitis (2.1%). The most frequent AEs leading to dose reductions or interruptions were prolongation of the QTc interval (4.3%), and diarrhea (2.5%). The most frequent AE leading to discontinuation was ILD/pneumonitis (3.9%). In the Japanese subset of FLAURA, AEs with osimertinib included dermatitis anceiform (46.2%), white blood cell count (WBC) decrease (21.5%), prolongation of the QTc interval (21.5%), and ILD/pneumonitis (12.3%).⁷

In a phase 1 study, osimertinib was assessed for safety, pharmacokinetics (PKs), and efficacy at doses of 20-240 mg once daily.⁸ The 20 mg starting dose was selected on the basis of preclinical toxicology data and predictive xenograft models as sufficient to inhibit EGFR T790M, whereas doses equivalent to 80 mg or more were expected to lead to more profound inhibition of tumor growth.⁹ Although no dose-limiting toxicity (DLT) occurred at the doses evaluated, there was an increase in the incidence and severity of AEs at doses of 160 mg and 240 mg. In this study, osimertinib 80 mg was determined as the recommended dose (RD) in patients with EGFR mutations.

In another PK study of osimertinib, the maximum plasma concentration (C_max) and area under the concentration-time curve (AUC) increased in a dose-proportional manner across the range of 20-240 mg.¹⁰ Administration of osimertinib once daily resulted in an approximately threefold accumulation, with steady-state exposures achieved after 15 days of dosing. At steady state, the population-estimated mean half-life of osimertinib was 48 hours and oral clearance (CL/F) was 14.3 L/h. The main metabolic pathways of osimertinib are oxidation (predominantly CYP3A) and dealkylation in vitro. Two pharmacologically active metabolites (AZ5104 and AZ7550) were identified in the plasma. The geometric mean exposure, according to the AUC, of each metabolite was approximately 10% of the exposure to osimertinib at steady state. In biochemical assays of active metabolites, AZD7550 has a comparable potency and selectivity profile to AZD9291 (osimertinib) and AZD5104 is more potent against mutant and wild-type EGFR forms than AZD9291, although AZD5104 exhibits the same overall profile.⁹

Osimertinib and its metabolites are eliminated via both hepatic and renal routes, with 68% of the dose being eliminated in feces and 14% in urine. Unchanged osimertinib accounted for approximately 2%. In the population pharmacokinetic (PPK) model, the largest effect was related to body weight, while serum albumin and ethnicity were also identified as significant covariates that affected the PKs of osimertinib.¹¹ There was no relationship between exposure and efficacy; however, a linear relationship between exposure and safety endpoints was observed. The PPK analysis concluded that renal and hepatic function would not be expected to impact exposure to osimertinib. As eligibility criteria in the prior investigational trial of osimertinib, subjects with a creatinine (Cr) >1.5 × upper limit of normal (ULN) concurrent with creatinine clearance (CrCl) <50 mL/min (measured or calculated using the Cockcroft and Gault formula [CGF]) were excluded. Evidence on the safety of osimertinib is limited in patients with severe or moderate renal and hepatic impairments.

Additionally, Asian female patients, who accounted for the majority of patients with NSCLC with EGFR mutations, had low body weight. Low body weight affects the PKs and toxicity of some oral anticancer agents.^12,13 Therefore, we conducted a PK and dose-finding study in patients with renal impairment with low body weight.

MATERIALS AND METHODS Study design

This is a multicenter, dose-finding study using four cohorts classified by renal function and body weight. Patients were allocated into the following cohorts: cohort A, normal renal function (NRF) (estimated glomerular filtration rate [eGFR] ≥50 mL/min/1.73 m²) and normal body weight; cohort B, moderate renal impairment (eGFR = 30-50); cohort C, low body weight (<45 kg); and cohort D, severe renal impairment (SRI) (eGFR <30 or undergoing dialysis) (Table 1). The protocol was approved by the certified review board of the National Cancer Center Hospital (jRCTs031180232), and the study was conducted at eight institutions in Japan in accordance with the ethical principles stated in the Declaration of Helsinki.

TABLE 1 Cohorts of osimertinib according to renal function and body weight

Cohort	Renal function: BSA-indexed eGFR (mL/min/1.73 m2)
	Normal ≥50	Moderate impairment 30 ≤ <50	Severe impairment <30 or undergoing dialysis
Body weight
≥45 kg	Cohort A	Cohort B	Cohort D
<45 kg	Cohort C

Note: The eGFR was calculated using the following formula: eGFR (mL/min/1.73 m2) = 194 × SCr (−1.094) × Age (−0.287) (×0.739 if female).

Abbreviations: BSA, body surface area; eGFR, estimated glomerular filtration rate.

The primary objective of this study was to investigate the safety, PKs, and RD of EGFR-mutated patients with NSCLC with impaired renal function and low body weight. We used cohort A as a control group and studied cohorts B, C, and D in two parts: dose escalation and expansion. The aim of dose escalation using a standard 3 + 3 design was to investigate the safety, tolerability, and PKs of osimertinib and determine the RD and maximum tolerated dose (MTD). The aim of the expansion section was to investigate the safety, tolerability, and PKs of RD in each cohort.

Adverse events, PK, QTc intervals, and tumor response were evaluated in all patients during a study period of 8 weeks. All AEs were classified and graded using the Common Terminology Criteria for AEs (CTCAE), version 4.0. DLT was assessed during the DLT evaluation period of 4 weeks. DLT was defined as grade 4 neutropenia and thrombocytopenia, any grade ≥3 nonhematological toxicity, prolonged QTcF interval (≥500 ms or an increase of ≥60 ms compared with the baseline interval), any grade of ILD, and grade ≥2 nonhematological toxicity requiring cessation for ≥7 days during the DLT evaluation period. In the dose-escalation part of cohorts B, C, and D, MTD was defined as the previous dose level below any dose level at which ≥2 of up to six evaluable patients experienced a DLT.

Patient eligibility

All eligibility criteria are shown in Table S1. Key inclusion criteria were as follows: histologically confirmed advanced NSCLC harboring activating EGFR mutations, regardless of T790M mutation; osimertinib chemotherapy planned as part of clinical practice; age 20 years or older; Eastern Cooperative Oncology Group performance status (ECOG-PS) 0-2; and adequate hematopoietic and hepatic function (WBC ≥2000/μL, absolute neutrophil count [ANC] ≥1000/μL, platelet count [PLT] ≥75,000/μL, aspartate aminotransferase [AST] and alanine aminotransferase [ALT] ≤100 IUL, and total bilirubin ≤1.8 mg/dL).

Key exclusion criteria were as follows: history of previous osimertinib treatment; concomitant treatment with other anticancer agents; a QTcF prolongation ≥450 ms (males) and ≥470 ms (females); serious or uncontrolled concomitant disorder including active infection, cardiac failure, cardiovascular disease, and diabetes; and a history of ILD or pulmonary fibrosis. All participants provided written informed consent.

Treatment and assessment of renal function and body weights

Based on the principles outlined in the US Food and Drug Administration Renal Impairment Study Guidance for Industry,¹⁴ we used eGFR to estimate renal function for classification.¹⁵ GFR has been considered the most reliable index for assessing overall renal function,¹⁶ the Modification of Diet in Renal Disease (MDRD) study¹⁷ or the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI).¹⁸ The CKD-EPI formula that used age, sex, and race as variables was initially developed in 2009, with the new formula being updated without race in 2021 (CKD-EPI 2021).¹⁹ In 2008, the Japanese Society of Nephrology established an equation for estimating the GFR.¹⁵ Although the eGFR equation was developed based on data obtained from Japanese patients with CKD, it offers higher accuracy for predicting GFR, which was measured by inulin, than the CGF and MDRD in Japanese cancer patients.²⁰ In the prior registration studies of osimertinib, subjects who had Cr >1.5 × ULN concurrent with CrCl <50 mL/min (measured or calculated by the CGF) were excluded.^5,6,8 Therefore, we defined that the subjects in this study were stratified based on renal function using Japanese eGFR formula (Table 1).

Generally, East Asian females with a high EGFR mutation frequency rate have lower body weight than Caucasians. Low body weight has never been defined, but the mean body weight ( ± standard deviation) in adult Japanese males was 67.4 ± 12.0 kg and 53.6 ± 9.2 kg in females.²¹ In this study, we defined low body weight as <45 kg (−1 SD level in Japanese females).

Eligible patients were treated with osimertinib at the dose level according to the allocated cohort (Table 2). In a phase 1 study of osimertinib, partial response was identified from a dose of 20 mg, whereas any AE of grade 3-5 was developed from a dose of 20 mg.⁸ Because tablets of 40 mg or 80 mg are available, we determined that the starting dose was 80 mg once daily in cohorts A, B, and C and 40 mg once daily in cohort D. In the dose-escalation part, a standard 3 + 3 design was planned at the dose levels of 1, 2, −1, and −2 (Table 2) as follows: Three patients were entered at the initial dose level. If a DLT was observed in one-third of the patients at this dose level, an additional three patients were entered at the same dose level. The dose level at which at least two patients experienced DLTs was defined as an unacceptable dose of osimertinib. After a study period of 8 weeks, patients continued to receive therapy until disease progression, clinical deterioration, or intolerable AEs. Serial electrocardiogram (ECG) was performed at the screening visit and during the treatment period at pre-dose, 6 hours, and 24 hours after administration of osimertinib on day 1; at pre-dose and 6 hours on day 15; and at pre-dose on days 29 and 56. The QT values were corrected using Fridericia's (QTcF) formula. Tumor evaluations were performed at baseline and 8 weeks after the start of treatment using the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.

TABLE 2 Dose and treatment schedule at cohort and each level

	Cohort A	Cohort B	Cohort C	Cohort D
Dose level	Normal renal function	Moderate renal impairment	Low body weight	Severe renal impairment
Level −2	40 mg, every other day	40 mg, every other day	40 mg, every other day	NA
Level −1	40 mg, once a day	40 mg, once a day	40 mg, once a day	40 mg, every other day
Level 1	80 mg, once a day	80 mg, once a day	80 mg, once a day	40 mg, once a day
Level 2	NA	NA	NA	80 mg, once a day

Pharmacokinetic evaluation and analysis

Blood samples for PK analysis were collected at pre-dose and 1, 2, 4, 6, 8,9, 10, and 24 hours after administration of osimertinib on day 1; at pre-dose and 1, 2, 4, 6, and 8 hours on day 15; and at pre-dose on days 29 and 56. Plasma levels of osimertinib and its active metabolites (AZ5104 and AZ7550) were determined using liquid chromatography–tandem mass spectrometry. We validated the assay under the Guidance for Industry of the Food and Drug Administration in Bioanalytical Method Validation (Method S2). AUC_0–t, C_max, trough concentration (C_trough), and T_max were estimated by noncompartment analysis using Phoenix WinNonlin version 8.1 (Pharsight Corporation).

Statistical analysis

We considered that six to eight patients were a valid number in each cohort to evaluate PKs. The target number of patients in cohort A was set at 12 and at the most 8 in cohorts B, C, and D. A previous PK report in the package inserted with osimertinib for patients with lung cancer reported that the coefficient of variation of the AUC (14,980 ± 6809 nM/h) on day 15 of osimertinib (N = 32) at a dose of 80 mg/body was 45.4%. Based on these data, we estimated that the 95% CI of the AUC (N = 12) was 11.286-19.865 nM/h and the AUC (N = 8) was 9.268-21.163 nM/h. PK parameters were compared between cohorts using the Mann-Whitney or Kruskal-Wallis test. The patient characteristics and AEs in each cohort were compared using Fisher's exact test. All statistical analyses were performed using GraphPad Prism V.8.0 (GraphPad Software).

RESULTS

A total of 31 patients with NSCLC with EGFR mutations were enrolled from April 2019 to July 2021 and classified into four cohorts according to renal function and body weight (Tables 1 and 2). Patient characteristics are summarized in Table 3, with many variations in baseline patient characteristics, such as sex, age, PS, comorbidity, EGFR mutation status, and prior therapies. The relationships between Japanese eGFR (mL/min/1.73 m²), CrCl using the CGF (mL/min), and CKD-EPI eGFR (mL/min/1.73 m²) are depicted in Figure S1. The CKD-EPI eGFR formula in 2009 and 2021 and the Japanese eGFR formula were used to calculate the BSA-indexed estimated GFR and the CGF to calculate the estimated CrCl. The analysis by Spearman's correlation coefficient demonstrated a high correlation (r > 0.85) among BSA-indexed estimated renal functions or among estimated renal functions.

TABLE 3 Baseline patient characteristics

Patient characteristics	Cohort A	Cohort Aa except early discontinuation	Cohort B	Cohort C	Cohort D	Total
Number of patients	n = 13	n = 12	n = 8	n = 8	n = 2	n = 31
Sex, M/F	4/9	3/9	5/3	1/7	1/1	11/20
Age (median [range])	63 [53-78]	63 [53-77]	77 [58-89]	83 [75-85]	72 [70-74]	72 [53-89]
Height (median [range])	160.0 [147.8-170.5]	157.5 [147.8-170.5]	160.7 [143.8-169]	144.9 [140.7-156.0]	159.1 [155.2-162.9]	154.9 [140.7-170.5]
Body weight (median [range])	62.0 [46.5-85.2]	60.0 [46.5-85.2]	68.9 [50.4-77.6]	39.9 [30.7-44.4]	49.2 [44.7-53.6]	53.6 [30.7-85.2]
BMI (median [range])	26.5 [19.6-30.0]	24.4 [19.6-29.7]	24.6 [20.6-28.2]	17.6 [14.7-20.9]	19.4 [18.6-20.2]	21.9 [14.7-30.0]
BSA (median [range])	1.62 [1.39-1.96]	1.58 [1.39-1.96]	1.63 [1.39-1.88]	1.27 [1.13-1.40]	1.49 [1.40-1.57]	1.53 [1.13-1.96]
Cr (median [range])	0.62 [0.52-1.0]	0.59 [0.52-1.0]	1.22 [0.93-1.61]	0.62 [0.44-1.06]	5.25 [4.77-5.73]	0.8 [0.44-5.73]
eGFR (median [range])	78.0 [55.7-92.3]	81.0 [55.7-92.3]	43.2 [32.4-47.5]	68.5 [37.8-102]	8.0 [7.5-8.5]	56 [7.5-102]
ECOG-PS 0/1/2	6/7/0	6/6/0	3/4/1	2/1/5	0/2/0	11/14/6
Stage IIIB/IVA/IVB/Rec	2/3/6/2	2/3/5/2	0/1/4/3	0/3/3/2	0/0/1/1	2/7/14/8
Smoking history (ex-smoker)	7 (53.8%)	6 (50.0%)	2 (25.0%)	2 (25.0%)	2 (100%)	13 (41.9%)
Lung complication (+)	1 (7.7%)	0	0	1 (12.5%)	1 (50.0%)	3 (9.7%)
Diabetes mellitus (+)	3 (23.1%)	3 (25.0%)	2 (25.0%)	0	1	6 (19.4%)
Hypertension	5 (38.5%)	4 (33.3%)	7 (87.5%)	5 (67.5%)	2 (100.0%)	19 (61.3%)
Macroangiopathy, cerebral infarction	1 (7.7%)	1 (8.3%)	2 (25.0%)	2 (25.0%)	0	5 (16.1%)
Histology Ad/Adsq	12/1	11/1	8/0	8/0	2/0	30/1
EGFR mutation
Ex19del	7	7	4	3	1	15
L858R	6	5	2	4	1	13
Ex19del + L858R	0	0	1	0	0	1
T790Mb	0	0	3	0	0	3
L861Q	0	0	1	1	0	2
Prior surgery for primary or metastatic lesion	3 (23.1%)	3 (25.0%)	4 (50.0%)	2 (25.0%)	1	10 (32.2%)
Prior radiotherapy	2 (15.4%)	2 (16.7%)	3 (37.5%)	3 (37.5%)	0	8 (25.8%)
Prior cytotoxic chemotherapy	2 (15.4%)	2 (16.7%)	3 (37.5%)	1 (12.5%)	0	6 (19.4%)
Prior EGFR TKI	5 (38.5%)	5 (41.7%)	3 (37.5%)	0	0	8 (25.8%)

Abbreviations: Ad, adenocarcinoma; Adsq, adenosquamous carcinoma; BMI, body mass index; BSA, body surface area; Cr, creatinine; eGFR, estimated glomerular filtration rate; EGFR, epidermal growth factor receptor; F, female; M, male; TKI, tyrosine kinase inhibitor.

^aIn cohort A, one patient discontinued osimertinib therapy on day 6 due to an exacerbation of chronic obstructive pulmonary disease.

^bT790M (Ex19del + T790M, n = 1; L858R + T790M, n = 1; Ex19del + L858R + T790M, n = 1).

Due to poor accrual, only two patients were included in cohort D. In cohort A, one patient discontinued osimertinib on day 6 due to exacerbation of chronic obstructive pulmonary disease. The Data and Safety Monitoring Committee judged that this AE was unlikely to be related to osimertinib, and this subject was replaced because PK sampling was not performed on day 15. We could not measure the plasma concentrations of AZD7550 in the second patient in cohort D because the supply of the reference materials was stopped due to the COVID-19 pandemic in 2021. Safety was evaluated in 31 patients and PKs in 29 patients.

Safety

In the dose-escalation part, no DLT was observed during the DLT evaluation period. In six subjects enrolled in cohort B and C, the RD was determined to be 80 mg once and the expansion was completed with two additional subjects. AEs were evaluated in all patients during the study period of 8 weeks (Table 4 and Table S3). Thirteen patients in cohort A developed 25 AEs (seven in grade 2 and two in grade 3). Eight patients in cohort B developed 32 AEs (six in grade 2 and two in grade 3). Eight patients in cohort C developed 44 AEs (12 in grade 2 and one in grade 3). Two patients in cohort D developed six AEs. Four serious AEs occurred in this study, two of which were considered to be related to osimertinib (heart failure grade 3 in cohort B and ILD grade 3 in cohort C). Regarding AEs according to sex, 10 male patients developed 41 AEs (seven in grade 2 and two in grade 3). Twenty female patients developed 65 AEs (18 grade 2 and three in grade 3; Table S4).

TABLE 4 Adverse events during the study period of 8 weeks

	Cohort A		Cohort B	Cohort C	Cohort D	Total
	N = 13	N = 12a	N = 8	N = 8	N = 2	N = 31
Acneiform rash	5 (38.5%)	5 (41.7%)	3 (37.5%)	5 (67.5%)	0	13 (41.9%)
Diarrhea	1 (7.7%)	1 (8.3%)	3 (37.5%)	6 (75.0%)	1 (50.0%)	11 (35.5%)
Anorexia	3 (23.1%)	3 (25.0%)	3 (37.5%)	4 (50.0%)	0	10 (32.3%)
QTc prolongation	2 (15.4%)	2 (16.7%)	3 (37.5%)	2 (25.0%)	1 (50.0%)	8 (25.8%)
Dry skin	1 (7.7%)	1 (8.3%)	0	4 (50.0%)	2 (100.0%)	7 (22.6%)
Dysgeusia	1 (7.7%)	1 (8.3%)	2 (25.0%)	2 (25.0%)	1 (50.0%)	6 (19.4%)
ILD	0	0	3 (37.5%)	1 (12.5%)	1 (50.0%)	5 (16.1%)
Malaise	1 (7.7%)	1 (8.3%)	1 (12.5%)	3 (37.5%)	0	5 (16.1%)
Mucositis	0	0	1 (12.5%)	4 (50.0%)	0	5 (16.1%)
Cr increased	1 (7.7%)	1 (8.3%)	2 (25.0%)	1 (12.5%)	0	4 (12.9%)
CK increased	1 (7.7%)	1 (8.3%)	1 (12.5%)	2 (25.0%)	0	4 (12.9%)
Paronychia	1 (7.7%)	1 (8.3%)	1 (12.5%)	2 (25.0%)	0	4 (12.9%)
PLT decreased	1 (7.7%)	1 (8.3%)	1 (12.5%)	2 (25.0%)	0	4 (12.9%)
Hb decreased	0	0	2 (25.0%)	1 (12.5%)	0	3 (9.7%)
Alb decreased	1 (7.7%)	1 (8.3%)	1 (12.5%)	0	0	2 (6.5%)
Body weight loss	1 (7.7%)	1 (8.3%)	0	1 (12.5%)	0	2 (6.5%)
PS decreased	0	0	1 (12.5%)	1 (12.5%)	0	2 (6.5%)
Nausea	1 (7.7%)	1 (8.3%)	0	1 (12.5%)	0	2 (6.5%)
ALP increased	0	0	1 (12.5%)	1 (12.5%)	0	2 (6.5%)

Abbreviations: Alb, albumin; ALP, alkaline phosphatase; CK, creatine kinase; Cr, creatinine; Hb, hemoglobin; ILD, interstitial lung disease; PLT, platelet count; PS, performance status.

^aIn cohort A, one patient discontinued osimertinib therapy on day six due of exacerbation of chronic obstructive pulmonary disease.

Notable toxicities related to osimertinib included an acneiform rash, diarrhea, QTc prolongation, and ILD. Acneiform rash at any grade was developed in 38.5% of cohort A, 37.5% of cohort B, and 67.5% of cohort C. Diarrhea at any grade was developed in 7.7% of cohort A, 37.5% of cohort B, and 75.0% (25.0% at grade 2) of cohort C. QTc prolongation was developed in 15.4% of cohort A, 37.5% of cohort B, and 25.0% of cohort C. ILD was developed during the study period of 8 weeks in 0% of cohort A, 37.5% of cohort B, 12.5% of cohort C, and 50% of cohort D (Table S5). After the study period, one patient developed grade 4 ILD on day 107. All patients who developed ILD recovered. Overall, toxicity occurred more frequently and at higher grades in cohorts B and C than in cohort A.

Tumor response

Of the eligible patients, 24 had measurable tumor lesions. Six patients had a history of prior EGFR TKI treatment, and 18 were EGFR TKI naïve. Response rate at 8 weeks was 58.3% (95% CI, 38.8-75.6) in all patients with measurable lesions and 72.2% (95% CI, 48.8-87.8) in the TKI-naïve patients (Figure 1 and Table S6).

FIGURE 1. Change in target lesion size from baseline to 8 weeks. Regimental stripe indicates the patients with the history of prior EGFR TKI treatments, and plain indicates the patients with EGFR TKI naïve. *Values in evaluable 24 patients. PD, progressive disease; PR, partial response; SD, stable disease

Pharmacokinetic analysis

All the patients in cohorts A, B, and C received osimertinib 80 mg, and those in cohort D received osimertinib at 40 mg. The PK parameters and time-to-concentration curves on days 1 and 15 are shown in Figure 2 and Table S7. There was no significant difference between the PK parameters of osimertinib in cohorts A, B, and C, although the AUC_0-24h of AZD5104 and AZD7550 in cohort C on day 15 tended to be higher than that in cohort A (P value in the Kruskal-Wallis test = 0.098 and 0.16, respectively; Figure 3). AUC_0-24h at a dose of 40 mg on day 15 in cohort D was about half (5580 nmol*h/L) of that in cohorts A, B, and C.

FIGURE 2. Time to mean concentration curve of osimertinib and its metabolites (A) on day 1 and (B) on day 15. Error bars represent standard deviations. Green line, patients in cohort A; red line, patients in cohort B; blue line, patients in cohort C; brown line, patients in cohort D

FIGURE 3. The AUC (A) from 0 to 24 h on day 1, (B) from 0 to 8 h on day 15, and (C) from 0 to 24 h on day 15 of osimertinib and its metabolites by cohort. The bars represent the mean values ± standard deviation. P values were calculated using the Mann-Whitney or Kruskal-Wallis test. AUC, area under the concentration-time curve

Pharmacokinetic parameters of osimertinib, AZD5104, and AZD7550 according to the Japanese eGFR, CrCl using the CGF, and CKD-EPI eGFR are depicted in Figure S2. We considered that there was no remarkable bias in the PK parameters by renal function using a variety of measurement formulas and body weights. The AUC_0-8h and AUC_0-24h of AZD7550 on day 15 in females were significantly greater than in males (603 ± 279 vs. 353 ± 141; p = 0.009 in the Mann-Whitney test and 1860 ± 811 vs. 1070 ± 422; p = 0.01, respectively).

Association between PK parameters and AEs

We investigated the association between PK parameters and AEs, especially notable toxicity related to osimertinib in patients who received osimertinib 80 mg (Figure 4 and Figure S3). The AUC_0-24h of AZD5104 on day 15 was significantly associated with the grade of diarrhea (grade 0 = 1527 ± 675.2, grade 1 = 2631 ± 1326, and grade 2 = 3472 ± 1260; p = 0.02 in the Kruskal-Wallis test) and grade ≥2 of any AEs (grade 0-1 = 1543 ± 1189 and grade 2-3 = 2311 ± 951.0; p = 0.01 in the Mann-Whitney test). In addition, there was no association between the PK parameters of osimertinib and its metabolites, and notable toxicity related to osimertinib.

FIGURE 4. AUC from 0 to 24 h on days 1 and 15 according to toxicity. (A) diarrhea and (B) any toxicity with grade 2 or higher. The box plot represents quartiles. P values were calculated using the Mann-Whitney or Kruskal-Wallis test. AUC, area under the concentration-time curve

DISCUSSION

In a prior phase 1 study, the PK analysis demonstrated that at the 80 mg dose, the geometric mean AUC_24h on day 22 was 12,120 (3680-38,800) nM*h, and C_max was 635.4 (167-2110) nM. The AUC of the active metabolites AZ5104 and AZ7550 was approximately 10% of the exposure to osimertinib after multiple dosing.⁸ In our dose-finding study, the PK parameters were similar and comparable to these results. The RD was determined to be 80 mg once daily for patients with moderate renal function and low body weight, as well as for those with NRF and normal body weight.

Additionally, renal impairment and low body weight had little effect on the PKs of osimertinib. However, toxicity occurred more frequently and at a higher grade in patients with moderate renal impairment (cohort B) and low body weight (cohort C) than in cohort A. AZD5104 was more potent against mutant and wild-type EGFR forms than osimertinib and was related to AEs including diarrhea and grade ≥2 of any AEs. ILD did not occur in cohort A but occurred in cohorts B (37.5%), C (12.5%), and D (50%). There was no difference in PK parameters including AUC, C_max, and C_trough between patients with and without ILD (data not shown). The high incidence rate of notable toxicities including ILD would be related to biased patient characteristics among the cohorts.

Osimertinib and its metabolites are eliminated via both hepatic and renal routes, with 68% of the dose being eliminated in feces and 14% in urine. Renal clearance is a minor route for osimertinib elimination. On the other hand, the renal excretion of another TKI, crizotinib (unchanged) was considered a minor route of drug elimination, accounting for approximately 2% of the dose administered, and about 22% of total drug-related material excreted in urine. However, the AUC for crizotinib was 1.8-fold greater in patients with SRI (CrCl <30 mL/min) than in those with NRF. Physiologically based PK modeling indicated a similar increase in steady-state AUC after multiple crizotinib dosing.^22,23 The recommended crizotinib dose for patients with SRI who do not require dialysis is 250 mg once daily, which is half of the dose compared with RD for patients with NRF. Thus, renal function may affect the toxicity or PKs of some anticancer agents, even if renal clearance is not the main route of elimination of these agents.

Another PK study of a single dose of osimertinib (80 mg) in patients with solid tumors with SRI (N = 9; CrCl of <30 mL/min) and with NRF (N = 7; CrCl ≥90 mL/min) was conducted by AstraZeneca.²⁴ They also performed PPK analyses from data sets of the AURA (NCT01802632), AURA2 (NCT02094261), AURA3 (NCT02151981), and FLAURA (NCT02296125) studies. In these results, osimertinib exposure based on C_max and AUC 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 versus patients with NRF, although the 90% CIs were wide and included unity (1.00). However, additional analysis showed that CrCl accounted for <18% of the between-patient variability in osimertinib exposure, and the slopes of the log-linear regression showed no relationship between osimertinib exposure and CrCl. They concluded that no dose adjustment for osimertinib was required when treating patients with SRI. They suggested that a proper clinical assessment and continuous monitoring in patients with SRI should be considered.

In a phase 1 study, the C_max and AUC of osimertinib as well as its active metabolites increased in a dose-proportional manner,^8–10 and the incidence and severity of skin and gastrointestinal AEs were increased at higher dose levels. Therefore, the toxicities of osimertinib were considered to be affected not only by the PKs of osimertinib but also by renal function, body weight, and the PKs of its metabolites. The AUC_0-8h and AUC_0-24h of AZD7550 on day 15 in females were significantly greater than those in males (Figure S2). However, the grade and frequency of toxicity in females were similar to those in males. We considered that the difference in the PK parameters of metabolites was related to sex differences in metabolic enzyme activity^25–27 but was not related to osimertinib toxicity.

The dosing of most oral molecular-targeted agents was fixed without adjustment for body weight and body surface area as opposed to the dosing of intravenous cytotoxic agents, although the dose was amended by the AEs and organ function. This is the major reason why the dose adjustment of most oral molecular-targeted agents does not lead to the minimization of inter-individual variability in PK parameters, toxicity, and therapeutic effects. However, the dosing of some molecular-targeted agents, including niraparib and antibody drug conjugates, is adjusted by body weight and body surface area based on their PK parameters and toxicity. In Asians, female patients with low body weight who account for the majority of NSCLC patients with EGFR mutations, the RD and PKs of osimertinib are the same, but toxicity should be considered.

This study had several limitations. First, it evaluated a small number of patients in each cohort. Guidance from the European Medicines Agency suggests describing the relationship between renal function and drug clearance; the inclusion of, for example, six to eight subjects per group is usually sufficient. Therefore, although the sample size was small, it was sufficient to assess the effects of renal function and body weight on the PKs of osimertinib. Second, there were biased patient characteristics among the cohorts, including age, sex, and comorbidities. This was because renal function was related to age and comorbidities and body weight was related to sex. Third, we used the Japanese eGFR to estimate renal function.¹⁵ Historically, GFR has been considered the most reliable index for assessing overall renal function, and the CGF is the most commonly used estimation method in oncological practice. In this study, we analyzed the relationship between the PK parameters and toxicities of osimertinib and renal function using a variety of measurement formulas and serum Cr values. However, there was no remarkable bias in the PK parameters of renal function using a variety of measurement formulas.

In conclusion, the RD of osimertinib was 80 mg once daily for patients with moderate renal function and low body weight, as well as those with NRF and normal body weight. Although the PK parameters of osimertinib were similar across all the cohorts in this study, toxicity occurred more frequently in patients with impaired renal function and low body weight. It is recommended that clinicians prescribe osimertinib with caution in patients with EGFR-mutated NSCLC with impaired renal function and low body weight.

ACKNOWLEDGMENTS

We would like to thank the patients, their families, caregivers, and all the investigators involved in this study.

FUNDING INFORMATION

This research was supported by AMED under grant number 20ck0106464h0003.

CONFLICT OF INTEREST

Yutaka Fujiwara received lecture fees from AstraZeneca and is an ad hoc advisor for Micron. Tetsunari Hase received research funds from AstraZeneca, Chugai, and Novartis. Naozumi Hashimoto received lecture fees from AstraZeneca, Boehringer Ingelheim, GSK, and Novartis and research funds from Boehringer Ingelheim. Yukari Tsubata received lecture fees from AstraZeneca, Chugai, and Daiichi-Sankyo and research funds from Pfizer. Toshiaki Takahashi received lecture fees from AstraZeneca and research funds from Amgen, AstraZeneca, Boehringer Ingelheim, Chugai, Eli Lilly, ONO, Merck Biopharma, MSD, and Pfizer. Yuki Shinno received research fund from Janssen. Yoshitaka Zenke received lecture fees from AstraZeneca, BMS, Chugai, Eli Lilly, and ONO and research funds from Amgen, AstraZeneca, Daiichi-Sankyo, MSD, and Merck. Yukio Hosomi received lecture fees from AstraZeneca. Noboru Yamamoto received lecture fees from Chugai and research funds from Astellas, Chugai, Eisai, Taiho, BMS, Pfizer, Novartis, Eli Lilly, AbbVie, Daiichi-Sankyo, Bayer, Boehringer Ingelheim, Kyowa-Hakko Kirin, Takeda, ONO, Janssen Pharma, MSD, Merck, GSK, Sumitomo Dainippon, Chiome Bioscience, Otsuka, Carna Biosciences, Genmab, Shionogi, and TORAY. The other authors have no conflicts of interest to declare.

ETHICS STATEMENT

Approval of the research protocol by an Institutional Review Board: The protocol was approved by the certified review board of the National Cancer Center Hospital on March 5th, 2019 (jRCTs031180232).

Informed Consent: All participants provided written informed consent.

Registry and the Registration No. of the study/trial: This study was registered in UMIN Clinical Trials Registry (UMIN000033301).

Animal Studies: N/A.