A phase 1b/2 study of PF-06747775 as monotherapy or in combination with Palbociclib in patients with epidermal growth factor receptor mutant advanced non-small cell lung cancer

ABSTRACT

Introduction

This Phase 1/2 study (NCT02349633) explored the safety and antitumor activity of PF-06747775 (oral, third-generation epidermal growth factor receptor [EGFR] tyrosine kinase inhibitor) in patients with advanced non-small cell lung cancer after progression on an EGFR inhibitor.

Methods

Phase 1 was a dose-escalation study of PF-06747775 monotherapy (starting dose: 25 mg once daily [QD]). Phase 1b/2 evaluated PF-06747775 monotherapy at recommended Phase 2 dose (RP2D; Cohort 1); PF-06747775 200 mg QD plus palbociclib (starting dose: 100 mg QD orally; Cohort 2A); and PF-06747775 monotherapy at RP2D in a Japanese lead-in cohort.

Results

Sixty-five patients were treated. Median treatment duration was 40.1 weeks. Monotherapy maximum tolerated dose was not determined. Two patients in Cohort 2A had dose-limiting toxicities. The monotherapy RP2D was estimated to be 200 mg QD. Most frequently reported adverse events (AEs) were diarrhea (69.2%), paronychia (69.2%), and rash (60.0%). Most AEs were grades 1–3. Overall, objective response rate (90% confidence interval [CI]) was 41.5% (31.2–52.5%). Median (range) duration of response was 11.09 (2.70–34.57) months. Median progression-free survival (90% CI) was 8.1 (5.4–23.3) months.

Conclusions

PF-06747775 had a manageable safety profile and the study design highlights important considerations for future anti-EGFR agent development.

KEYWORDS:

• (5-8): advanced non-small cell lung cancer
• epidermal growth factor receptor
• PF-06747775
• phase 1/2 study
• tyrosine kinase inhibitor

1. Introduction

Epidermal growth factor receptor (EGFR) mutations occur in approximately 20% of patients with NSCLC, with higher prevalence of approximately 40% in Asian patients (range, 20–76%) [1–4]. Approximately 40–60% of the patients with acquired resistance to first- and second-generation EGFR tyrosine kinase inhibitors (TKIs) can develop a secondary EGFR exon 20 p.T790M mutation [5–7]. Prior to the first-line approval of osimertinib (a third-generation TKI), acquisition of a second site EGFR mutation (p.T790M) was observed in almost half of patients at the time of progression. Therefore, targeting these double-mutant EGFR variants (exon 19 deletion/p.T790M and exon 21 p.L858R/p.T790M) was an unmet need.

PF-06747775 is a highly potent and selective third-generation EGFR TKI, with strong activity against EGFR single and double mutants, and weak activity against wild-type EGFR [8–10]. In xenograft mouse models, PF-06747775 demonstrated tumor growth inhibition and regression at well-tolerated doses driven by EGFR single and double mutants (data on file) [9].

This study was started prior to the US Food and Drug Administration approval of osimertinib in the first- and second-line settings for EGFR p.T790M disease, and at the time there was a significant unmet need to identify effective treatment options for EGFR-mutated (EGFRm) NSCLC. We hypothesized that combining PF-06747775 with palbociclib, a potent and highly selective reversible cyclin-dependent kinase (CDK)4/6 inhibitor, could further impede tumor growth, with increased efficacy, compared with PF-06747775 alone [11–14]. In vitro and in vivo assessments of PF-06747775 plus palbociclib showed increased efficacy versus PF-06747775 alone in NSCLC models representing first-line EGFRm and second-line p.T790M-resistant patient populations (data on file).

The purpose of this Phase 1/2 study was to explore the safety and antitumor activity of PF-06747775 as monotherapy and in combination with palbociclib or avelumab in patients with EGFRm (exon 19 deletion or p.L858R, with or without p.T790M) advanced NSCLC. Avelumab is a human monoclonal antibody that selectively binds to programmed death-ligand 1 [15,16]. However, the avelumab cohort of this study was never initiated following the results of the TATTON study that reported increased rates of interstitial lung disease with osimertinib plus durvalumab [17,18].

2. Methods

2.1. Study design

This Phase 1/2 study (NCT02349633) was performed at 11 centers in Australia, Japan, Republic of Korea, and the United States between 14 May 2015 and 28 May 2020. The Phase 1 portion was an open-label, multicenter, multiple-dose, safety, pharmacokinetics (PK), pharmacodynamics, and dose-escalation study of PF-06747775 monotherapy in patients with EGFRm advanced NSCLC. PF-06747775 was administered orally in successive cohorts as a single agent with a starting dose of 25 mg once daily (QD) continuously in 21-day cycles. This dose-escalation portion included a single-dose, lead-in period to assess single-dose PK of PF-06747775, followed by continuous QD dosing in a 21-day cycle. The Continual Reassessment Method was used to guide dose assignment (Supplementary Materials).

The Phase 1 portion included a series of PK sub-studies to investigate potential drug–drug (DDI) and drug–food interactions (Figure 1): (1) a two-period sub-study of PF-06747775 in combination with sildenafil (cytochrome P450 [CYP]3A4 probe substrate) to investigate modulation of CYP3A4 by PF-06747775; (2) a three-period sub-study of PF-06747775 in combination with esomeprazole (an antacid) and in combination with itraconazole (CYP3A4 probe inhibitor); and (3) a three-period sub-study of PF-06747775 administered with food and in combination with rifampin (CYP3A4 probe inducer). The third sub-study was planned, but was not initiated.

Figure 1. Original and planned study design.

* Approximate enrollment totals indicated. The following Cohorts were planned, but not performed: food-drug and CYP3A4 inducer drug–drug interaction PK sub-study (combination with rifampin); Cohort 2B; and Cohort 3. A Japanese patient lead-in cohort evaluating PF-06747775 in Japanese patients with advanced EGFRm NSCLC was also included, and comprised of two Cohorts: RP2D tolerability cohort and PK cohort. CYP, Cytochrome P450; EGFRm, epidermal growth factor receptor mutant; NSCLC, non-small cell lung cancer; PK, pharmacokinetic; RP2D, Recommended Phase 2 Dose

The Phase 1b/2 portion included four cohorts (Figure 1). Cohort 1 evaluated PF-06747775 monotherapy at the recommended Phase 2 dose (RP2D) in previously untreated patients with EGFRm advanced NSCLC (exon 19 deletion or p.L858R with/without p.T790M). Cohort 2 evaluated PF-06747775 plus palbociclib in previously treated patients with EGFRm advanced NSCLC (exon 19 deletion/p.T790M and p.L858R /p.T790M). There were two sub-cohorts within Cohort 2, A and B. Cohort 2A was a dose-finding evaluation of oral PF-06747775 200 mg QD plus oral palbociclib (starting dose: 100 mg QD) in 21-day cycles. Dose-finding followed the modified toxicity probability interval (mTPI) method with adjustments using dose-limiting toxicity (DLT) rate. Cohort 2B was a randomized evaluation of PF-06747775 plus palbociclib versus PF-06747775 monotherapy at the RP2D. Cohort 3 was a dose-finding and expansion evaluation of oral PF-06747775 (starting dose: 200 mg QD) plus avelumab 10 mg/kg every 2 weeks in 28-day (4-week) cycles in patients with EGFRm advanced NSCLC (exon 19 deletion/p.T790M and p.L858R /p.T790M). Dose-finding followed the mTPI design. Finally, there was a Japanese patient-only lead-in cohort (LIC) to evaluate PF-06747775 monotherapy at the RP2D in Japanese patients with EGFRm advanced NSCLC. The Japanese-only LIC comprised a RP2D tolerability cohort and a PK cohort.

The study was terminated on 5 April 2018 because of an internal decision to discontinue further development of PF-06747775 based on changes in the treatment landscape with the approval of a first-line third-generation EGFR TKI (osimertinib) [19]. As a result, no patients were enrolled in Cohort 2B and Cohort 3. Patients on treatment in other cohorts were allowed to remain on study and received PF-06747775 as a single agent or in combination with palbociclib.

The study was approved by either an institutional review board or an independent ethics committee at each study center. The study was conducted in accordance with all local legal and regulatory requirements, as well as the general principles set forth in the International Ethical Guidelines for Biomedical Research Involving Human Patients, Guidelines for Good Clinical Practice, and the Declaration of Helsinki. All patients provided written informed consent.

2.2. Patients

Eligible patients were aged ≥18 years with an Eastern Cooperative Oncology Group performance status (ECOG PS) 0 or 1, adequate liver, renal, and bone marrow function, and had histologically or cytologically confirmed locally advanced or metastatic EGFRm (exon 19 deletion or p.L858R) NSCLC.

In the Phase 1 portion of the study, patients were eligible regardless of EGFR p.T790M status following EGFR TKI progression. In the PK sub-studies involving food/antacid and CYP3A4 effects, patients with EGFRm (exon 19 deletion or p.L858R) with any T790M status were eligible. Patients had progressed after at least one prior line of therapy, including EGFR TKI. Reasons for including patients regardless of p.T790M status were to allow for a flexible study design and based on information from preclinical analyses about potent on-target inhibition against EGFR exon 19 deletion and p.L858R [8] and the possible activity against non-p.T790M mechanisms of resistance.

In Cohort 1 of the Phase 1b/2 portion, treatment-naïve patients with EGFRm NSCLC regardless of p.T790M status were eligible. In Cohort 2A, patients were required to have EGFRm (exon 19 deletion AND p.T790M or p.L858R and p.T790M) NSCLC tumors and progression on an approved first- or second-generation EGFR TKI. Patients previously treated with a third-generation EGFR TKI were ineligible. Patients with prior multiple lines of therapy were only eligible if the last therapy prior to study treatment was an EGFR TKI received ≤6 weeks prior to study registration. In the Japanese-only LIC, eligible patients had locally advanced or metastatic EGFRm NSCLC who had progressed on or were intolerant to standard therapy, or for whom no standard therapy was available, or who declined standard therapy. Prior progression must have included a first- or second-generation EGFR TKI.

Across all cohorts, patients with brain metastases were eligible if they had completed the appropriate radiation treatment that was clinically indicated, if any, and had recovered from the acute effects of any treatment that was delivered prior to study registration, had discontinued corticosteroid treatment for these metastases prior to registration, and were neurologically stable. Exclusion criteria included major surgery within 2 weeks or radiation therapy within 1 week prior to registration, systemic anticancer therapy within 2 weeks or five half-lives prior to registration, or prior irradiation to >25% of the bone marrow. The following concurrent medications were not permitted: strong CYP3A4 inhibitors, strong CYP3A4 inducers, CYP3A4 substrates with narrow therapeutic indices, and strong P-glycoprotein inhibitors. Patients treated with proton pump inhibitors had to switch to a histamine 2 receptor antagonist with staggered dosing.

2.3. Endpoints

2.3.1. Phase 1

The primary endpoint was Cycle 1 DLTs (defined in the Supplementary Materials). Secondary endpoints included objective response rate (ORR) per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1; safety parameters, e.g. adverse events (AEs), serious AEs (SAEs), and laboratory test evaluations; PK parameters for PF-06747775 and sildenafil; and the presence/absence of EGFR mutations in tumor and plasma.

2.3.2. Phase 1b/2 cohort 1

The primary endpoint was ORR per RECIST v1.1. Secondary endpoints were duration of response (DOR); overall survival (OS) at 24 months; progression-free survival (PFS); safety parameters, e.g. AEs, SAEs, and laboratory test evaluations; PK parameters for PF-06747775; and the presence/absence of EGFR mutations in tumor and plasma.

2.3.3. Phase 1b/2 cohort 2A

The primary endpoint was Cycles 1 and 2 DLTs. Secondary endpoints were DOR; OS at 24 months; ORR; PFS; safety parameters, e.g. AEs, SAEs, and laboratory test evaluations; PK parameters for PF-06747775 and palbociclib; and the presence/absence of EGFR mutations in tumor and plasma.

2.3.4. Japanese LIC

The endpoints were Cycle 1 DLTs; safety parameters, e.g. AEs, SAEs, and laboratory test evaluations; and PK parameters for PF-06747775.

2.4. Statistical analyses

Full details on sample size and analysis populations are described in the Supplementary Materials. ORR was defined as the proportion of patients with a confirmed (per RECIST v1.1) complete response (CR) or partial response (PR) relative to the total number of response evaluable patients. ORR was based on confirmed responses and two-sided 90% confidence intervals (CIs) were calculated using the Clopper-Pearson exact binomial method. PFS was defined as the time from Cycle 1 Day 1 (first dose) to the date that objective progressive disease was documented or death due to any cause, whichever occurred first. OS was defined as the time from the date of the first dose of study treatment until death. DOR, PFS, and OS were estimated using the Kaplan–Meier method; two-sided 90% CIs were calculated using the Brookmeyer and Crowley method.

Safety endpoints were summarized descriptively. All AEs and laboratory test abnormalities were graded using Common Terminology Criteria for Adverse Events (CTCAE) v4.03.

Blood samples for PK analyses were collected pre-dose and at regular intervals throughout the study and were assayed for all the analytes using a validated analytical method. Plasma PK parameters were estimated using noncompartmental analysis, and PK parameters were summarized descriptively unless otherwise stated. For DDI sub-studies, natural log-transformed area under the plasma concentration–time profile (AUC) from time 0 extrapolated to infinite time (AUC_inf) and to time tau (AUC_tau), and maximum observed plasma concentration (C_max) values were analyzed using a mixed-effects model with treatment as fixed effects and patient as random effect to estimate the effect PF-06747775 on sildenafil exposure, as well as the effect of esomeprazole and itraconazole on PF-06747775 exposure. Estimates of the adjusted mean differences and 90% CI were obtained from the model and were exponentiated to provide estimates and 90% CI of the ratios of adjusted geometric means. PK parameters were calculated using a Pfizer-validated software system (eNCA, v2.2.4).

Full details on the analysis of EGFR mutations are described in Supplementary Materials. Outcomes were summarized descriptively unless otherwise stated.

Statistical analyses were conducted using SAS v9.4 (Cary, NC). The data cutoff date for these analyses was 28 May 2020.

3. Results

3.1. Patient disposition and demography

Across cohorts, a total of 65 patients were enrolled and treated. Twenty-six patients were treated in the Phase 1 dose-escalation part (PF-06747775 doses: 25 mg to 600 mg QD). Twenty-nine patients were treated with PF-06747775 200 mg QD as monotherapy in Cohort 1 and Japanese-only LIC or in combination with sildenafil 25 mg or esomeprazole/itraconazole. Five patients were treated with PF-06747775 300 mg QD plus sildenafil 25 mg in Phase 1 and five patients were treated with PF-06747775 200 mg QD plus palbociclib 100 mg QD in Cohort 2A.

The cohort consisted of 25 male and 40 female patients (mean age, 60.8 years) and included 47 (72.3%) Asian patients. All patients, except one, had an ECOG PS 0 or 1 (Table 1) and 26 (40.0%) patients had brain metastases at baseline. Overall, 33 (50.8%), 21 (32.3%), and 40 (61.5%) patients had a positive plasma/tumor mutation status for EGFR p.T790M, p.L858R, and exon 19 deletion, respectively.

Table 1. Baseline demographics and characteristics

	PF-06747775 QD
	25 mg (N = 4)	50 mg (N = 3)	150 mg (N = 4)	275 mg (N= 4)	300 mg (N = 3)	450 mg (N = 4)	600 mg (N = 4)	200 mg (N = 29)	300 mg + Sildenafil 25 mg (N = 5)	200 mg + Palbociclib 100 mg QD (N = 5)	Total (N = 65)
Gender, n
Male	1 (25.0)	1 (33.3)	1 (25.0)	0	3 (100.0)	4 (100.0)	1 (25.0)	10 (34.5)	2 (40.0)	2 (40.0)	25 (38.5)
Female	3 (75.0)	2 (66.7)	3 (75.0)	4 (100.0)	0	0	3 (75.0)	19 (65.5)	3 (60.0)	3 (60.0)	40 (61.5)
Age, years
Mean (SD)	70.5 (10.8)	72.0 (1.0)	62.0 (7.3)	60.3 (3.9)	55.7 (19.4)	63.5 (10.2)	61.0 (12.2)	58.8 (11.9)	63.6 (7.9)	55.0 (10.5)	60.8 (11.1)
Range	57–81	71–73	52–69	55–64	43–78	53–77	46–74	37–76	51–70	47–73	37–81
Race, n (%)
White	1 (25.0)	2 (66.7)	1 (25.0)	2 (50.0)	0	1 (25.0)	2 (50.0)	4 (13.8)	1 (20.0)	2 (40.0)	16 (24.6)
Asian	3 (75.0)	1 (33.3)	2 (50.0)	2 (50.0)	3 (100.0)	3 (75.0)	2 (50.0)	24 (82.8)	4 (80.0)	3 (60.0)	47 (72.3)
Other	0	0	1 (25.0)	0	0	0	0	0	0	0	1 (1.5)
Unspecified	0	0	0	0	0	0	0	1 (3.4)	0	0	1 (1.5)
ECOG PS, n (%)
0	1 (25.0)	0	1 (25.0)	1 (25.0)	2 (66.7)	3 (75.0)	3 (75.0)	18 (62.1)	3 (60.0)	1 (20.0)	33 (50.8)
1	2 (50.0)	3 (100.0)	3 (75.0)	3 (75.0)	1 (33.3)	1 (25.0)	1 (25.0)	11 (37.9)	2 (40.0)	4 (80.0)	31 (47.7)
2	1 (25.0)	0	0	0	0	0	0	0	0	0	1 (1.5)

ECOG PS, Eastern Cooperative Oncology Group performance status; QD, once daily

The median (range) duration of treatment for the overall population was 40.1 (4–195) weeks, with most patients (n = 62) receiving treatment for >4 weeks. At the time of study termination, 35 (53.8%) patients had completed the study (patients who completed follow-up visits 28–35 days after permanent treatment discontinuation) and 30 (46.2%) patients had permanently discontinued from the study. Reasons for permanent discontinuation were consent withdrawal (n = 6), no longer willing to participate (n = 6), transition to compassionate use (n = 5; therapy unknown), early termination of survival follow-up (n = 5), death (n = 5), and study termination (n = 3). Due to early study termination, enrollment into the food effect and rifampin DDI PK sub-study, Cohort 2B (PF-06747775 plus palbociclib and single-agent PF-06747775) and Cohort 3 (PF-06747775 plus avelumab) were not initiated.

3.2. Safety

Overall, 37 patients were evaluable for DLTs: 26 patients in the dose-escalation cohort, six patients in the Japanese-only LIC (PF-06747775 200 mg QD), and five patients in Cohort 2A (PF-06747775 200 mg QD + palbociclib 100 mg QD).

No DLTs were observed in the dose-escalation and the Japanese cohorts. Two patients in Cohort 2A treated with PF-06747775 plus palbociclib had DLTs. One of these patients experienced AEs of grade 3 neutrophil count decreased, grade 2 platelet count decreased, and grade 3 white blood cell count decreased in the first cycle. These AEs were considered to be related to palbociclib and resulted in a 13-day temporary discontinuation of palbociclib. The other patient had an SAE of grade 3 alanine aminotransferase (ALT) increased and an AE of grade 1 aspartate aminotransferase (AST) increased (worsened to grade 3) at the beginning of Cycle 2 treatment (started on study day 22). Both AEs were considered to be related to PF-06747775 and palbociclib.

Since no DLTs were observed in the dose-escalation cohorts (maximum dose tested: PF-06747775 600 mg QD), the MTD was not determined for single-agent PF-06747775. An RP2D of 200 mg QD was selected for PF-06747775 monotherapy based on the AEs reported for both PF-06747775 monotherapy and combination therapy in the study.

All patients (N = 65) had at least one all-causality treatment-emergent AE and 61 patients had at least one treatment-related AE. The most frequently reported all-causality and treatment-related AEs were diarrhea, paronychia, and rash (Table 2). Most AEs were grades 1–3 in severity. Thirty-four (52.3%) and one (1.5%) patients experienced all-causality grade 3 and 4 AEs (not classed as DLTs), respectively, and 27 (41.5%) patients had treatment-related grade 3 AEs; no treatment-related grade 4 AEs were reported. Three (4.6%) grade 5 AEs (up to 28 days after last dose of study drug; one each of disease progression, pneumonia, and aspiration pneumonia) occurred, none of which were considered treatment related.

Table 2. Summary of most frequent all-causality TEAEs (≥15% of the patients) and treatment-related TEAEs

n (%)	Total (N = 65)
	All-causality		Treatment-related
	All-grade	≥Grade 3	All-grade	≥Grade 3
Any AEs	65 (100.0)	38 (58.5)	61 (93.8)	27 (41.5)
Diarrhea	45 (69.2)	8 (12.3)	45 (69.2)	8 (12.3)
Paronychia	45 (69.2)	2 (3.1)	45 (69.2)	2 (3.1)
Rash	39 (60.0)	7 (10.8)	39 (60.0)	7 (10.8)
Stomatitis	30 (46.2)	0	30 (46.2)	0
Dermatitis acneiform	28 (43.1)	3 (4.6)	28 (43.1)	3 (4.6)
Dry skin	27 (41.5)	1 (1.5)	27 (41.5)	1 (1.5)
Blood creatinine increased	24 (36.9)	3 (4.6)	21 (32.3)	1 (1.5)
Rhinorrhea	22 (33.8)	0	21 (32.3)	0
Pruritus	21 (32.3)	0	20 (30.8)	0
Decreased appetite	16 (24.6)	0	15 (23.1)	0
Nausea	16 (24.6)	0	10 (15.4)	0
Alopecia	15 (23.1)	0	13 (20.0)	0
Anemia	15 (23.1)	7 (10.8)	7 (10.8)	3 (4.6)
Dyspnea	15 (23.1)	2 (3.1)	1 (1.5)	0
Mucosal inflammation	14 (21.5)	2 (3.1)	14 (21.5)	2 (3.1)
Dizziness	13 (20.0)	0	6 (9.2)	0
Pneumonia	13 (20.0)	5 (7.7)	0	0
Rash maculo-papular	12 (18.5)	1 (1.5)	12 (18.5)	1 (1.5)
ALT increased	11 (16.9)	2 (3.1)	9 (13.8)	1 (1.5)
Headache	11 (16.9)	0	2 (3.1)	0
Cough	10 (15.4)	1 (1.5)	0	0
Dehydration	10 (15.4)	2 (3.1)	4 (6.2)	0
Epistaxis	10 (15.4)	0	6 (9.2)	0
Nasal dryness	10 (15.4)	0	10 (15.4)	0

Included data up to 28 days after last dose of study drug. NCI CTCAE (version 4.03) was applied. MedDRA (version 23.0) coding dictionary was applied.

AE, adverse event; ALT, alanine aminotransferase; CTCAE, Common Terminology Criteria for Adverse Events; MedDRA, Medical Dictionary for Regulatory Activities; TEAE, treatment-emergent adverse event

Twelve (18.5%) patients permanently discontinued from study treatment due to all-causality AEs. The most frequent of these were pneumonia (4.6%), rash (3.1%), and maculo-papular rash (3.1%). Seven (10.8%) patients permanently discontinued due to treatment-related AEs of rash and maculo-papular rash (n = 2 each), and cardiac failure, dermatitis acneiform, and skin infection (n = 1 each).

Twenty-seven (41.5%) and 25 (38.5%) patients had dose reductions due to all-causality and treatment-related AEs, respectively. The most frequent treatment-related AE leading to dose reduction was rash (16.9%). Two patients died <28 days after the last dose due to pneumonia and aspiration pneumonia (not considered treatment-related).

The most frequent shifts (≥5 patients) from grade ≤2 at baseline to grade 3 or 4 post-baseline for hematology laboratory results were lymphopenia (n = 8; 12.3%) and anemia (n = 5; 7.7%), and for chemistry results was hyponatremia (n = 8; 12.3%).

3.3. Efficacy

The ORR (90% CI) for the overall population regardless of EGFR p.T790M mutation status was 41.5% (31.2–52.5%) (Table 3), all of which were PRs (Figure 2). The median (range) DOR was 11.09 (2.70–34.57) months.

Figure 2. Waterfall plot of tumor size change data with mutation status based on tumor and plasma.

Largest decrease or smallest increase representing best response to treatment. Central assessments for plasma and tumor mutation at screening are displayed. A, archival; F, fresh; In, indeterminate; N, not done; PD, progressive disease; PR, partial response; SD, stable disease; U, uninformative.

Table 3. Summary of efficacy results

Patients, n (%)	PF-06747775 QD
	25 mg (N = 4)	50 mg (N = 3)	150 mg (N = 4)	275 mg (N = 4)	300 mg (N = 3)	450 mg (N = 4)	600 mg (N = 4)	200 mg (N = 29)	300 mg + Sildenafil 25 mg (N = 5)	200 mg + Palbociclib 100 mg QD (N = 5)	Total (N = 65)
CR	0	0	0	0	0	0	0	0	0	0	0
PR	2 (50.0)	0	3 (75.0)	3 (75.0)	0	2 (50.0)	1 (25.0)	11 (37.9)	3 (60.0)	2 (40.0)	27 (41.5)
Stable/no response	1 (25.0)	1 (33.3)	1 (25.0)	1 (25.0)	3 (100)	2 (50.0)	3 (75.0)	15 (51.7)	1 (20.0)	3 (60.0)	31 (47.7)
Objective progression	1 (25.0)	1 (33.3)	0	0	0	0	0	3 (10.3)	1 (20.0)	0	6 (9.2)
Indeterminate	0	1 (33.3)	0	0	0	0	0	0	0	0	1 (1.5)
Objective response rate (CR + PR)	2 (50.0)	0	3 (75.0)	3 (75.0)	0	2 (50.0)	1 (25.0)	11 (37.9)	3 (60.0)	2 (40.0)	27 (41.5)
90% exact CI^a	9.8–90.2	0–63.2	24.9–98.7	24.9–98.7	0–63.2	9.8–90.2	1.3–75.1	22.9–54.9	18.9–92.4	7.6–81.1	31.2–52.5
Duration of response, median (range), months	11.10 (2.70–19.51)	–	8.52 (8.49–10.79)	24.84 (8.09–27.57)	–	34.56 (34.54–34.57)	22.11 (22.11–22.11)	8.32 (2.80–33.59)	23.49 (9.87–26.25)	11.07 (9.70–12.43)	11.09 (2.70–34.57)

^aUsing exact method based on binomial distribution.

CI, confidence interval; CR, complete response; QD, once daily; PR, partial response; QD, once daily

The 29 patients who had plasma EGFR p.T790M-positive status achieved an ORR (90% CI) of 34.5% (20.0–51.4%), with 10 patients achieving PR, 15 stable disease (SD), three progressive disease (PD), and one an indeterminate response. The 36 patients with plasma EGFR p.T790M-negative status had an ORR (90% CI) of 47.2% (32.8–62.0%). The 13 patients who had a tumor EGFR p.T790M-positive status had an ORR (90% CI) of 38.5% (16.6–64.5%). Five patients had PR, six had SD, one had PD, and one had an indeterminate response. The 38 patients with tumor EGFR p.T790M-negative status achieved an ORR (90% CI) of 44.7% (30.9–59.3%). Among patients who were EGFR p.T790M-negative and had been previously treated with EGFR TKIs (n = 23), 11 had PR for an ORR (90% CI) of 47.8% (29.6–66.5). Of the 21 patients with an EGFR p.L858R mutation detected in either plasma or tumor, five had PR, 13 had SD, two had PD, and one had an indeterminate response. Of the 40 patients with an EGFR exon 19 deletion in either plasma or tumor, 21 had PR, 16 had SD, and three had PD. Overall, seven patients experienced new tumor lesions (lung, pleural effusion, lymph node, cervical spine, adrenal gland, and pleura).

In Cohort 1 (n = 29), the ORR (90% CI) was 37.9% (22.9–54.9%), with 11 patients achieving a PR and a median (range) DOR of 8.32 (2.80–33.59) months. In Cohort 2A (n = 5), the ORR (90% CI) was 40.0% (7.6–81.1%), with two patients achieving a PR, with a median (range) DOR of 11.07 (9.70–12.43) months.

The PFS analysis included 29 patients in the PF-06747775 200 mg QD group (per changes in planned analyses). Of these patients, 19 (65.5%) had objective progression and 10 (34.5%) were censored (two patients were given new anticancer treatment and eight patients permanently discontinued treatment prior to tumor progression). The median (90% CI) PFS was 8.1 (5.4–23.3) months. The probability of being event-free at 24 months (90% CI) was 22.8% (9.6–39.5%).

The OS analysis included nine patients from Cohort 1 and the Japanese-only LIC (per changes in planned analyses). Of these nine patients, one (11.1%) died and eight (88.9%) were censored due to loss of follow-up. Therefore, the median OS was not estimable since most patients were censored. Survival probability at 12 months (90% CI) was 87.5% (50.0–97.5%) but survival probability at 24 months was not available.

3.4. Pharmacokinetics

In the Phase 1 dose-escalation cohort and the Phase 2 Cohort 1, plasma PF-06747775 exposure generally increased in a dose-related manner over the 25–600 mg QD dose levels (Figure 3, Supplementary Table S1). Following single-dose administration, mean terminal elimination half-life (t_½) values ranged from 4.11 to 13.1 hours across all doses; there was no major accumulation or evidence of non-linearity. Following 11 days of PF-06747775 dosing (200 mg QD and 300 mg QD) plus a single oral dose of sildenafil 25 mg on Cycle 1 Day 11, a decrease in sildenafil AUC_inf and C_max by approximately 34% and 17%, respectively, was observed (Supplementary Figure S1, Supplementary Table S2). Co-administration with esomeprazole decreased PF-06747775 AUC_tau and C_max by approximately 17% and 22%, respectively. Co-administration with itraconazole increased PF-06747775 AUC_tau and C_max by approximately 16% and 22%, respectively (Supplementary Figures S2 and S3, Supplementary Table S1). No changes in the PK of PF-06747775 were evident when administered in combination with palbociclib compared with PF-06747775 alone. Following single- and multiple-dose administration, plasma PF-06747775 PK parameters were similar between the Japanese LIC RP2D and PK cohorts, and no clinically meaningful differences in the PK parameters (AUC_inf, AUC_tau, C_max) were observed.

Figure 3. Plasma PF-06747775 (a): AUC_tau(dn) and (b): C_max(dn) following multiple oral administration on cycle 1 day 11 (Phase 1 dose-escalation).

Circle represents individual value and star represents geometric mean. AUC_tau, area under the concentration-time profile from time 0 to time tau (τ), the dosing interval, where tau = 24 h for QD dosing; C_max, maximum observed plasma concentration; (dn), dose normalized; MTD, maximum tolerated dose

3.5. Biomarkers

EGFR mutation status was monitored using both formalin fixed paraffin embedded (FFPE) tumor tissue and plasma (Figure 2). DNA extraction from FFPE tissue was conducted using the QIAamp DSP DNA FFPE Tissue Kit for therascreen EGFR RGQ PCR Kit. Extracted DNA was stored at −20°C. Tumor mutation analysis from FFPE tumor tissue was conducted using the Qiagen Therascreen EGFR PCR assay established for in vitro diagnostic use. The Therascreen EGFR PCR assay detects a total of 7 mutations including EGFR p.L858R, exon 19 deletion and p.T790M.

For mutation analysis from plasma, blood was drawn into K2EDTA tubes and plasma isolated by a double centrifugation of blood first at 3000 rpm followed by 14,000 rpm for 10 minutes each. Plasma was stored at −70°C until analysis. EGFR mutation analysis from plasma was conducted using the Sysmex Inostics EGFR panel with the BEAMing method. The panel detected eight different molecular alterations including EGFR p.L858R, exon 19 deletion and p.T790M. The cut-off for variant calling was 0.04% for EGFR exon 19 deletions and p.L858R, and 0.06% for EGFR p.T790M mutation. Clinical response or lack thereof was not dependent on presence or absence of EGFR p.T790M mutation.

4. Discussion

This Phase 1/2 study evaluated the safety, tolerability, PK, efficacy, and pharmacodynamics of PF-06747775 in patients with EGFRm advanced NSCLC. Overall, 65 patients were treated, including patients from PF-06747775 Phase 1 dose-escalation, Phase 1 PK DDI sub-studies (combination with sildenafil, esomeprazole, or itraconazole), single-agent expansion Cohort 1 (PF-06747775 200 mg QD), combination Cohort 2A (PF-06747775 200 mg QD + palbociclib 100 mg QD), and Japanese-only LIC (PF-06747775 200 mg QD). The study was prematurely terminated because of an internal decision to discontinue further development of PF-06747775 and was not due to drug safety. Due to early study termination, enrollment into the food effect and rifampin DDI PK sub-study, Phase 2 Cohort 2B (PF-06747775 plus palbociclib and single-agent PF-06747775) and Phase 1b Cohort 3 (PF-06747775 plus avelumab) were not initiated.

The median (range) duration of treatment for the overall population was 40.1 (4–195) weeks. No patients in the dose-escalation cohorts experienced DLTs, with PF-06747775 600 mg QD the maximum dose tested; therefore, the MTD was not determined. Based on the overall AE data from the study, the longer-term tolerability of the higher doses was limited by persistent EGFR-wild type driven toxicities, i.e. grades 1 and 2 skin toxicities and diarrhea. Based on the safety profile and antitumor activity of PF-06747775, an RP2D of 200 mg QD monotherapy was selected, as this dose provided good tolerability and clinical activity (a high rate of ORR). Additionally, no DLTs were observed in the Japanese-only LIC. Two patients in Cohort 2A experienced DLTs: one patient had DLTs of neutrophil count decreased, platelet count decreased, and white blood cell count decreased; and the other patient had DLTs of ALT increased and AST increased. These DLTs were attributable to either palbociclib only or PF-06747775 plus palbociclib.

The most frequently reported AEs were diarrhea and paronychia. Other frequent AEs (>40% patients) were rash, stomatitis, dermatitis acneiform, and dry skin. Grade 5 AEs were reported for three patients, none of which were related to study treatment: disease progression, pneumonia, and pneumonia aspiration. No notable findings in laboratory test parameters and electrocardiogram data were reported. The AEs reported with PF-06747775 are consistent with those reported in Phase 1 trials with the third-generation EGFR TKIs osimertinib and nazartinib, which suggests gastrointestinal and dermatologic AEs are a class-wide effect of these third-generation EGFR TKIs [18,20,21].

ORR was 41.5% for the overall population (including treatment-naïve and previously treated patients), with a median DOR of 11.1 months. No CRs were observed. PRs were observed in patients with both EGFR exon 19 deletion and p.L858R mutation, and with p.T790M mutation detected in either plasma or tumor. PRs were also evident among EGFR p.T790M-negative patients previously treated with EGFR TKIs, with an ORR (90% CI) of 47.8% (29.6–66.5). In future studies, molecular profiling conducted at baseline and end-of-treatment should be performed to shed light on patterns of response and resistance. Survival probability at 12 months (N = 9) was 87.5%, and median PFS (N = 29) was 8.1 months. In previously treated patients, the ORR rate with osimertinib ranged from 61% to 71%, and PFS ranged from 9.6 to 12.3 months [21–24]. In the Phase 1 dose-expansion trial of osimertinib, ORR was 61% and median PFS was 9.6 months in patients with EGFR p.T790M-positive NSCLC who had progressed on prior therapy [21]. It is important to note that EGFR p.T790M selection was required in that study but not in our study of PF-06747775.

PF-06747775 PK was generally dose proportional over the range of 25 mg to 600 mg following both single- and multiple-dose administration; there was no major accumulation or evidence of non-linearity. A minor decrease in sildenafil exposure suggests a potential weak induction of CYP3A with multiple doses of PF-06747775. A slight increase in PF-06747775 exposure when co-administered with itraconazole suggests CYP3A-based metabolism is a contributor to the elimination of PF-06747775; however, the increase in PF-06747775 exposure is not expected to be clinically meaningful. A slight decrease in PF-06747775 exposure when co-administered with esomeprazole suggests an increase in gut pH may slightly decrease the oral absorption of PF-06747775, but not to a clinically meaningful extent. Based on these PK results, no dose adjustment of PF-06747775 is warranted in these clinical scenarios.

At the time of study design, there were several third-generation EGFR inhibitors in development. The study protocol and subsequent amendments were intentionally designed to be flexible and have various cohorts and outcomes assessed including the inclusion of patients with or without p.T790M in the Phase 1 portion. The protocol allowed for de-escalation of either PF-06747775 or palbociclib when given in combination. These study design considerations, particularly the inclusion of cohorts that combined PF-06747775 with other agents, were an attempt for efficient development in a rapidly evolving therapy area, as well as to provide an early clinical pharmacology package of PF-06747775 (multiple DDI investigations, food effect, and PK-electrocardiogram relationships) in the event that PF-06747775 continued clinical development. Therefore, the current study could be used as an example of an optimized and flexible study design and development approach.

Promising efficacy and a tolerable safety profile have been reported with nazartinib in EGFRm NSCLC, which is still in clinical development [20,25,26]. Other third-generation EGFR TKIs, besides osimertinib, which were in clinical development at the same time as PF-06747775 were rocelitinib, olmutinib, and naquotinib, all of which are no longer in development. Despite the discontinuation of these third-generation EGFR TKIs, it is important to continue to explore other anti-EGFR strategies to address the unmet needs around osimertinib drug resistance [27,28]. Current paradigms in overcoming osimertinib drug resistance may include EGFR TKI combinations with chemotherapy, CDK4/6 inhibitors, anti-angiogenesis inhibitors, mesenchymal–epithelial transition (MET) inhibitors, antibody–drug conjugates, or dual EGFR–MET targeting bispecific antibodies, among others [27–29].

As is common with early-phase studies, there are limitations that need to be taken into consideration when interpreting study results, including small sample size and a lack of a comparison arm. The early termination of this study further compounded these limitations, placing potential restrictions on the interpretation of any data from the PF-06747775 and palbociclib combination cohort as only five patients were treated and there was no randomized evaluation.

5. Conclusion

PF-06747775 as monotherapy and in combination with palbociclib had a safety profile that was generally tolerable and manageable in patients with EGFRm advanced NSCLC. Antitumor activity of PF-06747775 was observed across all dose levels and cohorts. Durable responses and disease stabilization were achieved by patients regardless of EGFR p.T790M mutation status. Plasma PF-06747775 exposures were generally dose proportional over the range of 25–600 mg QD. No adjustment in dose would be recommended for co-administration with CYP3A4 substrates, CYP3A4 modulators, or antacids. No further clinical evaluation of PF-06747775 as monotherapy or in a combination regimen is planned. However, the design of this study has provided an important understanding of how quickly the treatment landscape has changed and serves as an example of an optimized and flexible approach in the development of future agents for EGFRm NSCLC, e.g. the multi-cohort approach, broad assessment of endpoints, allowance for dose de-escalation, and comprehensive clinical pharmacology assessment.

6. Data sharing

Upon request, and subject to review, Pfizer will provide the data that support the findings of this study. Subject to certain criteria, conditions and exceptions, Pfizer may also provide access to the related individual de-identified participant data. See https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information.

Declaration of interest

Byoung Chul Cho: institutional research funding from Novartis, Bayer, AstraZeneca, MOGAM Institute, Dong-A ST, Champions Oncology, Janssen, Yuhan, Ono, Dizal Pharma, and MSD; consulting roles for Novartis, AstraZeneca, Boehringer-Ingelheim, Roche, BMS, Ono, Yuhan, Pfizer, Eli Lilly, Janssen, Takeda, and MSD; stock ownership of TheraCanVac Inc, Gencurix Inc, Bridgebio Therapeutics; and royalties from Champions Oncology.

Sarah B Goldberg: research funding from AstraZeneca and Boehringer Ingelheim; consulting/advisory board for AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Amgen, Spectrum, Blueprint Medicine, Sanofi Genzyme, Daiichi-Sankyo, Regeneron, Takeda and Janssen.

Dong-Wan Kim: institutional research funding from Alpha Biopharma, Amgen, AstraZeneca/Medimmune, Boehringer-Ingelheim, Daiichi-Sankyo, Hanmi, Janssen, Merus, Mirati Therapeutics, MSD, Novartis, ONO Pharmaceutical, Pfizer, Roche/Genentech, Takeda, TP Therapeutics, Xcovery, Yuhan, Chong Keun Dang, Bridge BioTherapeutics, GSK; Travel and accommodation support for advisory board meeting attendance from Amgen, Daiichi-Sankyo.

Timothy F Burns: consulting and/or advisory board roles for Novartis, Foundation Medicine Inc., Amgen, and Mirati Therapeutics Inc.

Zarnie Lwin: honoraria for Astra Zeneca, Roche; advisory board for AbbVie; consulting role for MSD, Merck; travel and educational support from BMS.

Nuzhat Pathan, Wei Dong Ma, Joanna C Masters, Nandini Cossons, and Keith Wilner: employees and shareholders of Pfizer.

Makoto Nishio: grants and/or personal fees from Ono Pharmaceutical, Bristol Myers Squibb, Pfizer, Chugai Pharmaceutical, Eli Lilly, Taiho Pharmaceutical, AstraZeneca, Boehringer-Ingelheim, MSD, Novartis, Merck Biopharma, Daiichi Sankyo, Takeda Pharmaceutical Company Limited, TEIJIN PHARMA LIMITED, and AbbVie.

Hatim Husain: institutional research funding from BMS, Eli Lilly, Roche Sequencing Solutions, Pfizer; consulting or advisory roles for Takeda, Blueprints Medicine, AstraZeneca, Mirati, Foundation Medicine, Philips, Coherus, Merck; Honoraria from AstraZeneca, Blueprints Medicine, Janssen.

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Author contributions

All authors were involved in the trial conception/design, or the acquisition, analysis, or interpretation of data. All authors contributed to the drafting of the manuscript. The manuscript has been read and approved by all the authors, and all the conditions as previously stated by the ICMJE have been met. All authors agree to be accountable for all aspects of the work.

Reviewer disclosures

Peer reviewers in this manuscript have no relevant financial or other relationships to disclose

Acknowledgments

Medical writing support was provided by Anne Marie McGonigal, PhD, of Engage Scientific Solutions and funded by Pfizer.

Data from this study has been published/presented at the following congresses as an abstract and/or poster/oral presentation: European Society for Medical Oncology 42nd Congress – ESMO 2017; UC San Diego Moores Cancer Center – 13th Industry/Academia Translational Oncology Symposium 2017; IASLC Targeted Therapies of Lung Cancer 2017.

Supplementary material

Supplemental data for this article can be accessed here

Additional information

Funding

This paper was funded by Pfizer.