A phase II randomized trial evaluating gefitinib intercalated with pemetrexed/platinum chemotherapy or pemetrexed/platinum chemotherapy alone in unselected patients with advanced non-squamous non-small cell lung cancer

Abstract

Current pemetrexed/platinum chemotherapy does not produce a satisfactory therapeutic response in advanced lung cancer patients. The aim of this study was to determine whether the administration of gefitinib, a tyrosine kinase inhibitor (TKI), intercalated with pemetrexed/platinum could improve the efficacy in chemotherapy-naïve patients with advanced non-squamous NSCLC without subsequent gefitinib maintenance therapy. Treatment-naïve patients with stage IIIB or IV NSCLC were randomly assigned to receive pemetrexed (500 mg/m² d1) and either cisplatin (75 mg/m² d1) or carboplatin (AUC = 5 d1) plus gefitinib (250 mg/d on days 3 to 16 of a 3-week cycle) (PC-G) or pemetrexed–platinum (PC) alone. Randomization was stratified according to the tobacco smoking status and EGFR mutational status of the patients. The primary endpoint was the non-progression rate (NPR) at 12 weeks. Secondary endpoints included progression-free survival (PFS), overall response rate (ORR), overall survival (OS), and biosafety. The NPR at 12 weeks was 84.5% for the PC-G treatment arm and 83.1% for the PC treatment arm (P = 0.87). Median PFS was 7.9 months for the PC-G arm and 7.0 months for the PC arm (P = 0.57). The ORR was 50.0% for the PC-G arm and 47.4% for the PC arm (P = 0.78). Median survival was 25.4 mo for the PC-G arm and 20.8 mo for the PC arm (P = 0.54). The incidence of adverse events was similar between the two treatment arms, except for a higher incidence of skin rash with PC-G. Predefined subgroup analyses demonstrated that PC-G significantly increased the PFS compared with the PC regimen in patients with EGFR mutations (P = 0.017). Although gefitinib intercalated with pemetrexed/platinum chemotherapy did not improve the NPR at 12 weeks compared with chemotherapy, an improvement in the PFS for the intercalated treatment arm was seen in the subgroup of patients with EGFR mutations.

Keywords: :

• non-small cell lung cancer
• non-squamous
• gefitinib
• pemetrexed
• platinum
• EGFR mutations
• phase II randomized trial

Introduction

Non-small cell lung cancer (NSCLC) is one of the malignancies with the highest mortality rate worldwide. In many patients, the diagnosis is first established at an advanced stage of the disease.^1,2 Current treatment for advanced lung cancer does not produce a satisfactory therapeutic response because only limited improvement in survival has been achieved over the past 30 y.³ Much work has been dedicated to improving treatment by combining tyrosine kinase inhibitors of the epidermal growth factor receptor (EGFR-TKIs) with standard platinum-based doublet chemotherapy using a concurrent, intercalated, or maintenance approach.^4,5

Randomized studies using a concurrent combination of EGFR-TKI and chemotherapy demonstrated no advantage in unselected NSCLC patients.^4,5 Two factors that may have contributed to this lack of effect are (1) the potential antagonism of such drug combinations: preclinical studies have suggested that EGFR-TKIs can induce G₁-phase cell cycle arrest, which could have spared the tumor cells from the cytotoxic activity of cell cycle phase-dependent chemotherapeutic agents,⁶ and (2) the study subjects were not selected with regard to biomarkers: recent work has suggested that EGFR-TKIs produce a better clinical outcome in the NSCLC patients with a EGFR mutation than in those without a mutation. A phase II study (FAST-ACT I) on the sequential intercalated combination of erlotinib following chemotherapy with gemcitabine/platinum has shown improvement of PFS and well-tolerated toxicity when compared with gemcitabine/platinum alone as a first-line therapy.⁷ Subsequently, the result of the phase III FAST-ACT II study confirmed that such an intercalated combination therapy could significantly improve the PFS, even the OS in the unselected patients with NSCLC population, especially in the patients with EGFR mutations.⁸ However, in these two trials, erlotinib was administered sequentially during chemotherapy and maintained after the end of chemotherapy, which made it difficult to determine whether the survival benefit was due to the sequential therapy during chemotherapy, the subsequent maintenance treatment, or both. Therefore, it was necessary to design a study to investigate specifically the effect of sequential intercalated EGFR-TKI therapy on chemotherapy without any confounding effects from maintenance treatment.

For patients with non-squamous advanced (stage IIIB/IV) NCSLC, administration of pemetrexed plus cisplatin has demonstrated superior survival and modest toxicity compared with the combination cisplatin/gemcitabine, which has been regarded as the standard regimen for advanced NSCLC.^9,10 However, early studies were always conducted using doublet regimens that did not include pemetrexed as the control therapy arm, which may have led to an overestimation of the benefit of the sequential EGFR-TKI therapy.^11,12 In addition, sequential administration of pemetrexed followed by EGFR-TKI has proven to be promising.

This present randomized, phase II study was therefore conducted to evaluate the efficacy and toxicity of gefitinib intercalated with pemetrexed–platinum without gefitinib maintenance (PC-G) compared with pemetrexed–platinum (PC) alone as the first-line therapy in chemotherapy-naïve patients with advanced non-squamous NSCLC.

Results

Patients

A total of 120 patients were screened to participate in the study, and 117 were randomized into the two treatment groups (58 for the PC-G arm; 59 for the PC alone) between January 2010 and May 2012 (Fig. 1). Baseline characteristics of the study patients were similar between the two groups (Table 1).

Figure 1. Patient disposition up to December 31, 2012, cutoff date of this analysis.

Table 1. Baseline patient characteristics by treatment group

Characteristic	PC-G (n = 58)		PC (n = 59)
	No.	%	No.	%
Age, years
Mean	55.3		54.9
Range	36 to 72		33 to 70
Sex
Male	33	57	25	42
Female	25	43	34	58
Disease stage
IIIB	5	9	5	8
IV	53	91	54	92
Smoking status
Ex-smoker	29	50	20	34
Never-smoker	29	50	39	66
EGFR status
Mutant (19/21 exon)	14	24	18	31
Wild type	26	44	28	47
Unknown	16	28	12	20
20 exon mutation	1	1.7	1	1.7
20 and 21 exon mutation	1	1.7	0	0
ECOG PS
0	8	14	9	15
1	50	86	50	85
Cisplatin or carboplatin
Cisplatin	55	95	56	95
Carboplatin	3	5	3	5
Non-squamous
Large cell	0	0	0	0
Adenocarcinoma	58	100	59	100

Abbreviations: PC, pemetrexed plus either cisplatin or carboplatin; G, gefitinib; ECOG PS, Eastern Cooperation Oncology Group performance status; EGRF, epidermal growth factor receptor.

Efficacy

The non-progression rate (NPR) at 12 wk was found to be 84.5% (49/58) for the PC-G arm and 83.1% (49/59) for the PC arm (P = 0.87). The objective response rate (ORR) was 50.0% (27/54) for the PC-G arm and 47.4% (27/57) for the PC arm (95% CI −1.6% to 2.1%; P = 0.78). Median PFS was 7.9 mo for the PC-G arm and 7.0 mo for the PC arm (HR: 0.88, 95% CI 0.56–1.37; P = 0.57). By the end of the study, 36.2% (21/58) of the patients in the PC-G group and 44.1% (26/59) in the PC group had died. The median OS was 25.4 mo for the PC-G arm compared with 20.8 mo for the PC arm (HR 0.84, 95% CI 0.47–1.48; P = 0.54) (Fig. 2).

Figure 2. Kaplan–Meier plots of progression-free survival (PFS) in both treatment groups. (A) PFS for all NSCLC patients; (B) PFS for never smokers; (C) PFS for current or ex-smokers; (D) PFS for patients with EGFR mutant; (E) PFS for patients with wild-type EGFR; (F) PFS for patients with unknown EGFR status; (G) Kaplan–Meier plots of overall survival (OS) in both treatment groups. Abbreviations: PC, pemetrexed plus either cisplatin or carboplatin; G, gefitinib; HR, hazard ratio.

Predefined subgroup analysis showed that PC-G produced favorable HR (0.20, 95% CI 0.05–0.75; P = 0.017) for patients with EGFR mutations (exons 19/21) compared with PC with regard to PFS. The median PFS was 14.0 mo for the PC arm. However, for the PC-G arm, the median PFS was not reached before the conclusion of the study. For patients with wild-type EGFR (0.89, 95% CI 0.51–1.57; P = 0.698) and those with unknown EGFR genotype (2.21, 95% CI 0.88–5.57; P = 0.092), no significant difference was found between the two treatment groups. Additionally, the PC-G regimen showed a trend for better PFS compared with the PC regimen in the never-smokers (HR: 0.57, 95% CI 0.32–1.02; P = 0.06). In patients with EGFR mutations (exons 19/21), the ORR was 76.9% for the PC-G arm and 50% for the PC arm (P = 0.13). In contrast, in patients with wild-type EGFR, the ORR was 51.9% and 50%, respectively (P = 0.89). In patients with unknown EGFR genotype, the ORR was 21.4% and 36.4%, respectively (P = 0.41) (Table 2).

Table 2. Efficacy analysis

All patients	PC-G (n = 58)	PC (n = 59)	P value
NPR-12 wk	84.5%	83.1%	0.87
Median PFS	7.9 mo	7.0 mo	0.57
ORR	50.0%	47.4%	0.78
Median OS	25.4 mo	20.8 mo	0.54
EGFR mutant	PC-G (n = 13)	PC (n = 18)	P value
Median PFS	Not reached	14 mo	0.0017
ORR	76.9%	50.0%	0.13
EGFR wild type	PC-G (n = 27)	PC (n = 28)	P value
Median PFS	5.3 mo	5.4 mo	0.82
ORR	51.9%	50.0%	0.89
EGFR unknown	PC-G (n = 14)	PC (n = 11)	P value
Median PFS	6.6 mo	10.1 mo	0.09
ORR	21.4%	36.4%	0.41
Never smoker	PC-G (n = 29)	PC (n = 39)	P value
Median PFS	9.2 mo	6.8 mo	0.06
ORR	89.7%	71.8%	0.07
Smoker	PC-G (n = 29)	PC (n = 20)	P value
Median PFS	5.2 mo	7.2 mo	0.25
ORR	37.9%	50.0%	0.4

Safety

The cohort of patients for the safety study included all randomized patients who received at least one dose of the study medications (n = 117). Fifty-five patients in each treatment arm (95% for PC-G, 93% for PC) had at least one AE that was considered to be possibly related to the study treatment (Table 3). Most of the reported AEs were grade 1 or 2 in severity. Grade 3 treatment-related AE was reported in 22% of the PC-G-treated patients and 15% of the PC-treated patients (Chi-square test, P = 0.32).

Table 3. Incidence and severity of treatment related adverse events

Treatment-related AE	PC-G (n = 58)				PC (n = 59)
	Any grade		Grade 3+		Any grade		Grade 3+
	No.	%	No.	%	No.	%	No	%
Total patients with ≥1 AE	55	95	13	22	55	93	9	15
Nonhematologic
Rash	27	47	9	16	6	10	0	0
Nausea	20	34	3	5	22	37	5	8
Anorexia	23	40	1	2	18	30	2	3
Fatigue	19	32	0	0	13	22	1	1
Vomiting	18	31	6	10	17	29	5	8
Dry skin	5	8	0	0	1	1	0	0
Constipation	16	27	3	5	20	33	2	3
Diarrhea	12	20	1	2	5	8	0	0
Stomatitis	20	34	3	5	10	17	1	1
Hematologic
Anemia	30	52	2	3	27	46	1	1
Neutropenia	15	25	6	10	20	34	7	12
Thrombocytopenia	5	8	1	2	8	14	1	1
Biochemical
ALT increased	7	12	1	2	4	7	0	0

Abbreviations: AE, adverse event; PC, pemetrexed plus either cisplatin or carboplatin; G, gefitinib; ALT, alanine aminotransferase.

The most common adverse event in the PC-G group was skin rash, which was noticed in 47% (27/58) of the patients (any grade). More severe skin rash (grade 3 or higher) was observed in 16% (9/58) of the patients. The incidence of hematologic adverse events was similar in both treatment arms. Anemia was the most commonly reported hematologic toxicity. The incidence of anemia was similar in the two treatment arms (52% vs. 46%). Stomatitis was the most common gastrointestinal reaction in the PC-G arm, while constipation was common in the PC arm. No treatment-related death was reported in either arm of the study.

Subsequent systemic anticancer therapy

After progression of the disease, 45 (77.6%) patients in the PC-G arm and 46 (78.0%) patients in the PC arm received anti-cancer treatment following the study period. Docetaxel and erlotinib were the most frequently used drugs as second-line or higher-line treatments (Table 4).

Table 4. Post-study anti-cancer therapy

	PC-G		PC
	(n = 58)	%	(n = 59)	%
Any post-study anti-cancer therapy	45	77.6	46	78.0
Second- or higher line
Docetaxel	31	53.4	32	54.2
Gefitinib	1	1.7	8	13.6
Erlotinib	9	15.5	3	5.0
Radiotherapy	2	3.4	2	3.4
Others	2	3.4	3	5.0

Abbreviations: PC, pemetrexed plus either cisplation or carboplatin; G, Gefitinib; Others included Traditional Chinese Medicine.

Discussion

To our knowledge, this study is the first randomized, phase II trial to evaluate whether gefitinib intercalated with chemotherapy can provide improved benefits to unselected, chemotherapy-naïve patients with advanced non-squamous NSCLC who did not receive gefitinib maintenance therapy. Previous studies have shown that NPR, used as a study endpoint, can predict clinical benefit in cancer patients.^13-15 When designing this study, we presumed that the NPR could be improved from 64% in the PC group to 84% after 12 wk (almost 4 cycles) of PC-G treatment. Our results showed that the intercalated addition of gefitinib did not improve the treatment outcomes not only with regard to NPR but also with regard to ORR, PFS, or OS.

EGFR-TKI has been used in combination with standard platinum-based doublet chemotherapy in the concurrent, intercalated, or maintenance regimens. Studies on the concurrent administration of chemotherapy and EGFR-TKI confirmed a lack of efficacy. Four large-scale phase III clinical trials, including TRIBUTE, TALENT, INTACT-I, and INTACT-II, have demonstrated that platinum-based doublets in combination with TKI did not bring survival benefit to patients with advanced NSCLC.^4,5,11,12 The second approach is to administer EGFR-TKI as maintenance therapy when tumor response is achieved or during stable disease following cytotoxic chemotherapy. Two phase III studies, SATURN^16-20 and INFORM,^21,22 have shown improvement in PFS, but the biomarker-based selection of patients for maintenance therapy is disputable. The third choice is the intercalated administration. Concurrent administration may not be effective because of TKI-induced, G₁-phase cell-cycle arrest.²³ When EGFR-TKI and chemotherapy are given in a sequentially intercalated way, thus achieving pharmacodynamic separation of the two agents, the inhibitory drug interaction could be avoided.²⁴ Mok and colleagues tested this concept in a randomized phase II study (FAST-ACT-I).⁷ A total of 154 treatment-naïve Asian patients were randomized to receive gemcitabine plus cisplatin or carboplatin on day 1 and gemcitabine on day 8 followed by erlotinib from days 15 to 28 or the same chemotherapy followed by placebo on a 4-wk cycle. The primary endpoint was the NPR at 8 wk. The results indicated that the PFS was significantly longer in the intercalated administration arm compared with the control arm (HR 0.47; P = 0.018). The OS was similar because the placebo control arm was unblinded at disease progression and offered the option of erlotinib as a post-study treatment. Thereafter, these authors reported the findings of the phase 3 FAST-ACT-II trial using the same treatment design in The Lancet Oncology.⁸ The results demonstrated that the intercalated regimen improved the PFS (7.6 vs. 6.0 mo, P < 0.0001) and OS (18.3 vs. 15.2 mo, P = 0.042) in an unselected population of East Asian patients. However, in the FAST-ACT-I and FAST-ACT-II trials, the TKI was administered not only intercalated sequentially during chemotherapy but also as maintenance therapy after the end of chemotherapy, which limits the ability to demonstrate whether the survival benefit was due to the intercalated therapy during chemotherapy or the subsequent maintenance treatment because TKI maintenance therapy has been shown to be beneficial in the SATURN, INFORM, and other studies.^16,21,25,26 Our phase II randomized, controlled study partially answered this question. Without confounding effect from gefitinib maintenance therapy, the intercalated regimen of chemotherapy plus TKI provided no added benefit to unselected, chemotherapy-naïve patients with advanced non-squamous NSCLC compared with chemotherapy alone.

In the subgroup of patients with EGFR mutation in our study, an improvement in PFS was observed for the intercalated administration arm (HR, 0.20; median, not reached vs. 14 mo), which was similar to that in FAST-ACT-II. In a predefined subgroup analysis of approximately two-thirds of the patient cohort of FAST-ACT-II, only patients with EGFR-activating mutations benefited significantly from the intercalated regimen (median PFS 16.8 vs. 6.9 mo, P < 0.0001; median overall survival 31.4 vs. 20.6 mo, P = 0.0092). These findings demonstrated that the EGFR mutational status may serve as a biomarker to identify patients who can benefit the most from the intercalated combination therapy. Unfortunately, although more than 70% of the patients had tissue samples available for translational biomarker research, tumors positive for EGFR mutations were identified in less than 30% of the patients. Thus, the exploratory subgroup analysis should not be over interpreted. A larger randomized phase II or phase III study with PFS as the endpoint in the group of patients with EGFR mutations is warranted to define the value of the intercalated combination. It should also be noted that there was a trend toward worse PFS and OS among patients with EGFR unknown status when treated with PC-G vs. PC in our study, even though the difference was not statistically significant. This further emphasizes the need to test the mutational status at the time of the diagnosis.

The JMDB study^27-29 indicated that for patients with non-squamous NSCLC and for patients of Asian descent, the PC regimen had significantly better efficacy than GC, which was used as the control arm in FAST-ACT-I and FAST-ACT-II. Our study chose the pemetrexed-containing regimen as the control arm to avoid overestimating the benefit from the intercalated EGFR-TKI and strengthen the quality of our results. Furthermore, the preclinical study showed that concurrent or sequential administration of pemetrexed with EGFR-TKI has synergistic anti-tumor activity for both TKI-sensitive and TKI-resistant cell lines. This synergism may be related to the activation of the EGFR/PI3K/AKT pathways induced by pemetrexed. These pathways can be inhibited by the subsequent administration of TKI to maximize the inhibition of tumor growth.

In another subgroup analysis, the PFS of never-smokers appears to be longer than that of current or ex-smokers (P = 0.06). This may be because the prevalence of EGFR mutations was 20.4% (10/49) among the smokers but was 32.4% (22/68) among the never-smokers (P = 0.027). The prevalence of EGFR activating mutation in the smokers is acceptable. In NSCLC of any histology, pooled frequencies of EGFR mutations ranged from 8.4% to 35.9% for ever/heavy smokers and from 37.6% to 62.5% for never/light smokers, depending on ethnicity.³⁰ Not only in the Asian but in the western population this different prevalence of EGFR mutation between the smokers and never-smokers does exist. Taking account of the relatively small sample size, further study is required to clarify whether never-smoker status can serve as a biomarker for screening Chinese patients who can benefit the most from intercalated treatment if tumor tissues are not available. For western population, never-smoker status may be more meaningful for predicting the candidates of such an intercalated treatment due to a larger difference in prevalence of EGFR mutations between smokers and never-smokers.

Just as in the FAST-ACT I and II trials, intercalated dosing of TKI with chemotherapy was well tolerated, without additional hematologic toxicity in our study. Dose reduction and dose interruption were uncommon with gefitinib. Skin rash was observed in 47% of the patients in the PC-G arm, and 16% of these skin rash events were grade 3 or higher. However, skin rash could be managed with the best supportive treatment. Diarrhea was more common in the PC-G arm (20% vs. 8% in the PC arm), but almost all these events were grade 1 or 2. No treatment-related death was reported in either group.

Conclusions

This study demonstrates for the first time that intercalated administration of TKI with standard platinum-containing doublet chemotherapy did not improve the clinical outcome of unselected, chemotherapy-naïve patients with advanced non-squamous NSCLC. Additional studies are warranted to evaluate the clinical value of such intercalated combination in patients with EGFR mutations.

Patients and Methods

Study design

This single-center, randomized, controlled, open-label, phase II study was conducted between December 2009 and May 2012 at Fudan University Shanghai Cancer Center in accordance with the Declaration of Helsinki, the international Conference on Harmonization Guidelines for Good Clinical Practice, and applicable local regulations. The study was approved by the ethics committee of Fudan University Shanghai Cancer Center, and written informed consent was obtained from all study patients. This study is registered with ClinicalTrials.gov as number NCT01769066.

Eligibility criteria

Eligible patients were at least 18 y of age, with histologically or cytologically confirmed advanced or recurrent stage IIIB or IV non-squamous NSCLC (Union for International Cancer Control classification, version 6) and an Eastern Cooperation Oncology Group (ECOG) performance status of 0 to 1. These patients also had measurable disease according to Response Evaluation Criteria in Solid Tumors (RECIST 1.0), adequate hematological, hepatic, and renal functions, and a life expectancy of more than 12 wk. Patients were ineligible to participate if they had received previous systemic anticancer therapy or had severe drug allergy, active infection or another serious disease or condition, uncontrolled brain metastases, uncontrolled pleural effusion and/or pericardial effusion, or second malignancy, pregnancy, or lactation.

Procedures

Patients were randomly assigned in a 1:1 ratio to receive gefitinib (250 mg/day, orally) on day 3 to 16 of a 3-wk cycle in combination with pemetrexed (500 mg/m² i.v. infusion on day 1) and either cisplatin (75 mg/m² i.v. infusion on day 1) or carboplatin (AUC = 5, i.v. infusion on day 1) or pemetrexed–platinum alone. Treatment was repeated every 3 wk and continued until disease progression, unacceptable toxicity, or completion of a maximum of 6 cycles. Maintenance therapy was not allowed. The choice of second or higher line anti-cancer therapy was at the investigator’s discretion. Randomization of the study patients was stratified according to the tobacco-smoking status (current or ex-smoker/never-smoker) and EGFR genotype (mutation, wild-type, or unknown). The EGFR mutational status was determined using direct sequencing of exons 18 to 21. The EGFR mutation was defined by the presence of deletions on exon 19 or L858R mutations on exon 21. Patients with other types of mutations or deletions were classified as wild type for convenience of the analysis. Patients were followed up until their death or the study closure.

Baseline characteristics were recorded at the study enrollment. Radiological assessment was conducted at baseline and every 6 wk thereafter. Tumor response (according to RECIST 1.0) was evaluated by CT or magnetic resonance imaging scan of the patients’ chest and abdomen every 6 wk during chemotherapy and after completion of chemotherapy. Brain imaging and bone scans were performed before enrollment. Adverse events were graded according to the National Cancer Institute Common Terminology Criteria (NCI-CTC 3.0) for Adverse Events.

Statistical analysis

The primary endpoint was the NPR at 12 wk (complete response, partial response, or stable disease >12 wk). A two-group Z test with a two-sided level of significance level of 0.1 will have 80% power to detect the difference between the PC-G (NPR 0.84) and PC (NPR 0.64) groups when the sample size was 58 in each group. Secondary endpoints included ORR (complete response or partial response), PFS, and OS. ORR was compared between the two groups using Chi-square tests. Log-rank tests were used to compare the survival between the treatment groups. Cox proportional hazards model was used to calculate the HR for sequential intercalated combination of gefitinib and chemotherapy vs. chemotherapy alone. A two-sided significance level of 0.05 was used for all analyses, which were performed using SPSS (version 19) software.

Abbreviations:
CI	=	confidence interval
ECOG	=	Eastern Cooperative Oncology Group
EGFR	=	epidermal growth factor receptor
HR	=	hazard ratio
NPR	=	non-progression rate
NSCLC	=	non-small cell lung cancer
ORR	=	overall response rate
OS	=	overall survival
PC	=	pemetrexed–platinum
PC-G	=	pemetrexed–platinum plus gefitinib
PFS	=	progression-free survival
TKI	=	tyrosine kinase inhibitor

Disclosure of Potential Conflicts of Interest

The authors declare that they have no competing interests.

Acknowledgments

The authors would like to thank Drs Liu Fang and Xu Qinghua of Fudan University Shanghai Cancer Center-Institut Merieux Laboratory for their outstanding contribution in the translational research such as data analysis and biomarkers analysis.