EafiE

YBYVa

Cb

c

Md

Te

f

g

h

i

j

Kk

l

m

a

ARA

KEkEpIN

Crns

xy(

h0n

Lung Cancer 104 (2017) 119–125

Contents lists available at ScienceDirect

Lung Cancer

jou rn al hom epage: www.elsev ier .com/ locate / lungcan

fficacy according to blind independent central review: Post-hocnalyses from the phase III, randomized, multicenter, IPASS study ofrst-line gefitinib versus carboplatin/paclitaxel in Asian patients withGFR mutation-positive advanced NSCLC

i-Long Wua,∗, Nagahiro Saijob, Sumitra Thongprasertc, J. C.-H. Yangd, Baohui Hane,enjamin Margonof, Busayamas Chewaskulyongc, Patrapim Sunpaweravongg,uichiro Oheh, Yukito Ichinose i, Jin-Ji Yanga, Tony S.K. Mokj, Helen Youngk,incent Haddadl, Yuri Rukazenkovm, Masahiro Fukuokab

Guangdong Lung Cancer Institute, Guangdong General Hospital, Guangdong Academy of Medical Sciences, 106 Zhongshan Er Road, Guangzhou 510080,hinaMedical Oncology Division, Kinki University School of Medicine, 377-2 Ohno-Higashi, Osaka-Sayama City, Osaka 589 8511, JapanMedical Oncology Unit, Department of Medicine, Maharaj Nakorn Chiang Mai Hospital, Faculty of Medicine, Chiang Mai University, 110 Intavaroros Road,uang, Chiang Mai 50200, ThailandDepartment of Oncology, National Taiwan University Hospital and National Taiwan University Cancer Center, No. 7, Chung-Shan South Road, Taipei 100,aiwanDepartment of Respiratory Disease, Shanghai Chest Hospital, Shanghai Jiaotong University, Shanghai 200030, ChinaJl. Mayjend Prof. Dr. Moestopo No. 6-8 Surabaya, Jawa Timur 60285, IndonesiaSongklanagarind Hospital, Prince of Songkla University, 15 Karnjanavanich Road, Hat Yai, Songkhla 90110, ThailandDivision of Thoracic Oncology, National Cancer Center Hospital, 5-1-1 Tsukiji, Chou-ku, Tokyo 104-0045, JapanNational Hospital Organization Kyushu Cancer Center, 3-1-1, Notame, Minami-ku, Fukuoka 811-1395, JapanDepartment of Clinical Oncology, The Chinese University of Hong Kong, Prince of Wales Hospital, 30-32 Ngan Shing Street, Sha Tin, New Territories, Hongong, China(Formerly of) AstraZeneca, 310 Cambridge Science Park, Milton Road, Cambridge, Cambridgeshire, CB4 0FZ, United KingdomAstraZeneca, Da Vinci Building, Melbourn Science Park, Cambridge Road, Melbourn, Royston, Hertfordshire, SG8 6HB, United KingdomAstraZeneca, Mereside, Alderley Park, Macclesfield, SK10 4TG, United Kingdom

r t i c l e i n f o

rticle history:eceived 31 October 2016ccepted 29 November 2016

eywords:pidermal growth factor receptor tyrosine

a b s t r a c t

Objective: The Phase III, randomized, open-label IPASS study (NCT00322452) of first-line epidermalgrowth factor receptor tyrosine kinase inhibitor (EGFR TKI) gefitinib versus carboplatin/paclitaxel forAsian patients with advanced non-small-cell lung cancer (NSCLC) showed that investigator-assessedprogression-free survival (PFS) and objective response rate (ORR) were significantly prolonged in patientswith EGFR mutation-positive NSCLC who received gefitinib versus patients with EGFR mutation-negative

NSCLC. We report post-hoc analyses of IPASS data by blind independent central review (BICR), performed

inase inhibitorpidermal growth factor receptor-mutationositive

PASS studyon-small-cell lung cancer

at the request of the US FDA, in a subset of patients with EGFR mutation-positive NSCLC.Patients and methods: Eligible patients (aged ≥18 years; histologically/cytologically confirmed Stage IIB/IVadenocarcinoma NSCLC; non- or former light-smokers; treatment-naïve) were randomly assigned 1:1to gefitinib (250 mg/day) or carboplatin. (dose calculated to produce an area under the curve of 5 or6 mg/mL/minute)/paclitaxel (200 mg/m2). Primary endpoint: PFS. BICR analyses included PFS, ORR, andduration of response (DoR)

Abbreviations: ADC, adenocarcinoma; ALK, anaplastic lymphoma kinase; ASR, age-standardized rate; BICR, blind independent central review; CI, confidence interval;NS, central nervous system; CONSORT, Consolidated Standards of Reporting Trials; CT, computed tomography; DoR, duration of response; EGFR, epidermal growth factoreceptor; FDA, Food and Drug Administration; HR, hazard ratio; IPASS, Iressa Pan-ASia Study; ITT, intent-to-treat; NA, not available; NPV, negative predictive value; NSCLC,on-small-cell lung cancer; OR, odds ratio; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; PPV, positive predictive value; PS, performancetatus; RECIST, Response Evaluation Criteria in Solid Tumors; TKI, tyrosine kinase inhibitor; US, United States; WHO, World Health Organization.∗ Corresponding author.

E-mail addresses: syylwu@live.cn (Y.-L. Wu), saijo@jsmo.or.jp (N. Saijo), sumitra95@gmail.com (S. Thongprasert), chihyang@ntu.edu.tw (J.C.-H. Yang),kyyhan@gmail.com (B. Han), bmargono@yahoo.com (B. Margono), bchewask@gmail.com (B. Chewaskulyong), sunpawep@myumanitoba.ca (P. Sunpaweravong),ohe@ncc.go.jp (Y. Ohe), ichinose.yukito@gmail.com (Y. Ichinose), yangjinji2003@163.com (J.-J. Yang), tony@clo.cuhk.edu.hk (T.S.K. Mok), hhyoung@btinternet.comH. Young), vincent.haddad@astrazeneca.com (V. Haddad), yuri.rukazenkov@astrazeneca.com (Y. Rukazenkov), fukuoka@izumi-hp.com (M. Fukuoka).

ttp://dx.doi.org/10.1016/j.lungcan.2016.11.022169-5002/© 2016 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-c-nd/4.0/).

120 Y.-L. Wu et al. / Lung Cancer 104 (2017) 119–125

Results: Scans from 186 IPASS patients (gefitinib n = 88, carboplatin/paclitaxel n = 98) with EGFR mutation-positive NSCLC were available for BICR. Consistent with investigator-assessed results, in patients withEGFR mutation-positive NSCLC: PFS (hazard ratio 0.54; 95% confidence interval [CI] 0.38, 0.79; p = 0.0012)and ORR (odds ratio 3.00; 95% CI 1.63, 5.54; p = 0.0004) were significantly longer with gefitinib versuscarboplatin/paclitaxel. The median DoR by BICR was 9.6 months with gefitinib and 5.5 months withcarboplatin/paclitaxel.Conclusion: BICR analysis of IPASS data support the original, investigator-assessed results. EGFR mutation-positive status remains a significant predictor of response to first-line TKI therapy.

s. Pu

1

n(mbtscApwtsmPTpAwt[

sipmnt[vNnpInsvpv(0tpdi99

wpI

© 2016 The Author

. Introduction

Statistics indicate that, in Asia, lung cancer represents a sig-ificant clinical burden: it is the most common cancer in menage-standardized rate [ASR; per 100,000] 35.2) and the third

ost common cancer in women (ASR 12.7) [1]. As recommendedy clinical guidelines, epidermal growth factor receptor (EGFR)yrosine kinase inhibitors (TKIs) are now widely accepted as atandard-of-care therapy for patients with non-small-cell lung can-er (NSCLC) whose tumors harbor activating EGFR mutations [2,3].lthough EGFR mutations may be detected in the tumor of anyatient with NSCLC, certain clinical characteristics are associatedith a high frequency of activating EGFR mutations and response

o EGFR TKI therapy (Asian ethnicity, female gender, never-mokers, and adenocarcinoma [ADC] histology) [4,5]. The Phase III,ulticenter, randomized, open-label, parallel-group IPASS (Iressa

an-ASia Study) study (NCT00322452), which compared EGFRKI gefitinib (IRESSATM) with carboplatin plus paclitaxel (carbo-latin/paclitaxel) doublet chemotherapy as first-line treatment forsian patients with advanced NSCLC, was the landmark studyhich established that the EGFR mutation status of a patient’s

umor was a strong predictor of response to EGFR TKI therapy6,7].

In support of this conclusion, in IPASS biomarker analy-es, a significant interaction between EGFR mutation status andnvestigator-assessed progression-free survival (PFS; primary end-oint) was observed: in the subgroup of patients with EGFRutation-positive NSCLC (gefitinib n = 132, carboplatin/paclitaxel

= 129) PFS was significantly longer in patients receiving gefi-inib versus patients receiving carboplatin/paclitaxel (hazard ratioHR] 0.48; 95% confidence interval [CI] 0.36, 0.64; p < 0.001). Con-ersely, in the subgroup of patients with EGFR mutation-negativeSCLC (gefitinib n = 91, carboplatin/paclitaxel n = 85), PFS was sig-ificantly longer in patients receiving carboplatin/paclitaxel versusatients receiving gefitinib (HR 2.85; 95% CI 2.05, 3.98; p < 0.001).

nvestigator-assessed objective response rate (ORR) was also sig-ificantly higher with gefitinib versus carboplatin/paclitaxel in theubgroup of patients with EGFR mutation-positive NSCLC (71.2%s. 47.3%, respectively; odds ratio [OR] 2.75; 95% CI 1.65, 4.60;

< 0.001), and significantly higher with carboplatin/paclitaxelersus gefitinib in patients with EGFR mutation-negative NSCLC23.5% vs. 1.1% [one patient], respectively; OR 0.04; 95% CI 0.01,.27; p = 0.001) [6]. Additionally, post-hoc analyses of IPASS foundhat the median duration of response (DoR) with gefitinib foratients with EGFR mutation-positive NSCLC was 8.7 months [8]. Noifferential treatment effect for overall survival (OS) was observed

n the subsets of patients with EGFR mutation-positive (HR 1.00;5% CI 0.76, 1.33; p = 0.990) or EGFR mutation-negative (HR 1.18;5% CI 0.86, 1.63; p = 0.309) NSCLC [7].

As per protocol, the outcomes from the overall IPASS studyere assessed by the local investigators from institutions thatarticipated in the study. In 2015, to support the application of

RESSA with the US Food and Drug Administration (FDA), the FDA

blished by Elsevier Ireland Ltd. This is an open access article under the CCBY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

requested the re-analysis of PFS and other efficacy endpoints ina subgroup of patients from IPASS with EGFR mutation-positiveNSCLC, using blind independent central review (BICR). BICR maybe retrospectively implemented in clinical trials as an auditing stepthat aims to offset the potential measurement variation that mayhave resulted from local evaluation of tumor scans, and validatethe results of investigator-assessed outcomes [9]. BICR is of partic-ular value in non-blinded studies where potential bias cannot beexcluded when assessments are conducted by investigators. FDAguidelines therefore recommend verification of study endpointsbased on tumor measurements by central, independent review-ers blinded to the study treatment [10], and this has been utilizedin previous first-line trials of EGFR TKIs in which investigatorsare not blinded to study treatment [11,12]. The post-hoc analysesrequested by the FDA of efficacy endpoints according to BICR in asubgroup of IPASS patients with EGFR mutation-positive NSCLC arereported here.

2. Patients and methods

2.1. Study design, patients, and treatment

Full details of IPASS have been published previously [6]. Eligi-ble patients were aged ≥18 years; had histologically/cytologicallyconfirmed Stage IIB/IV NSCLC of ADC histology; were non-smokers(had smoked <100 cigarettes in their lifetime) or former lightsmokers (had stopped smoking ≤15 years previously and had≤10 pack-years of smoking); had no prior chemotherapy, bio-logic, or immunologic therapy (adjuvant chemotherapy permittedif not platinum-based and completed >6 months previously); hada World Health Organization (WHO) performance status of 0, 1,or 2; had measurable disease according to Response EvaluationCriteria in Solid Tumors (RECIST) [13]; and had absolute neu-trophil count >2.0 × 109/L with adequate hepatic function. Patientswere randomly assigned 1:1 to gefitinib (250 mg/day) or carbo-platin/paclitaxel (paclitaxel was administered intravenously over3 h on Day 1, followed by carboplatin area under the serumconcentration-time curve 5.0 or 6.0 intravenously over 15–60 minin once-every-3-weeks cycles for ≤6 cycles).

The primary endpoint of IPASS was PFS by investigatorassessment, to indicate non-inferiority of gefitinib relative to carbo-platin/paclitaxel. Secondary endpoints included ORR, OS, quality oflife, reduction in symptoms, safety, and adverse-event profile. Thepost-hoc analyses presented here report PFS, ORR, and DoR data ina subgroup of IPASS patients with EGFR mutation-positive NSCLCaccording to BICR.

All patients provided written, informed consent, and anindependent ethics committee at each participating institu-

tion approved the study protocol. The study was conducted inaccordance with the Declaration of Helsinki, the InternationalConference on Harmonisation/Good Clinical Practice, applicableregulatory requirements, and AstraZeneca’s policy on bioethics.

ancer

2

io6clsws

cpcwfafTeuaelrrI

2

Ca(Haatuct

3

3

iswm(a

3

5gwPwm

Y.-L. Wu et al. / Lung C

.2. Assessments

In IPASS, PFS was assessed by investigators from date of random-zation to the earliest sign of disease progression as per RECIST [13]r death due to any cause, and tumor response was assessed every

weeks until disease progression. Tumor samples from patientsonsenting to biomarker analyses were analyzed at two centralaboratories to determine biomarker status, and patients were con-idered to be positive for the EGFR mutation if 1/29 EGFR mutationsere detected with use of the amplification refractory mutation

ystem and the DxS EGFR29 mutation-detection kit [14,15].For BICR of IPASS data, scheduled computed tomography (CT)

hest and abdomen (including liver and adrenal gland) scans fromatients with EGFR mutation-positive NSCLC were retrospectivelyollected post-IPASS study (magnetic resonance imaging scansere taken where CT scans were medically contraindicated or not

easible). Additional imaging and anatomy (e.g. the brain) werelso reviewed where available. Scans were reviewed in electronicormat or, if not possible, via the original quality hard-copy film.he independent review committee consisted of eight radiologists:ach case was read by a minimum of two radiologists (double-read)sing pre-defined assessment criteria for RECIST progression with

third (or more) acting as adjudicator(s) if necessary. The review-rs were blinded to patient name, investigator site identifiers, siteesion selection for tumor assessments, site determination of tumoresponses, exam dates/reason for exam, and treatment arm. Theeview was performed independently of AstraZeneca (by PAREXELnformaticsTM).

.3. Statistical analyses

For investigator and BICR assessments: PFS was analyzed using aox proportional hazards model adjusted for the following covari-tes: WHO performance status (0 or 1 or 2), smoking historynever-smoker or light ex-smoker), and sex (female or male). TheR (gefitinib:carboplatin/paclitaxel) was estimated with 95% CIsnd p-values. ORR was analyzed using a logistic regression modelnd included the same covariates as for PFS. DoR was the time fromhe first documentation of complete response/partial responsentil the first documentation of progression or death (whicheverame first). Median DoR was calculated using the Kaplan-Meierechnique.

. Results

.1. Patients

Patient disposition is presented in Fig. 1. A total of 261 patientsn IPASS harbored EGFR mutations in their tumors, of which CTcans from 186 were available for BICR. Patient demographicsere consistent between the BICR population, the overall EGFRutation-positive population, and the overall study population

n = 1217; enrolled and randomly assigned between March 2006nd October 2007) (Table 1).

.2. Progression-free survival by BICR

According to BICR, at the time of data cut-off (April 14, 2008),5 (62.5%) and 69 (70.4%) progression events had occurred in theefitinib and carboplatin/paclitaxel arms, respectively. Consistent

ith the original investigator-assessed PFS results (Table 2; Fig. 2A),

FS by BICR demonstrated a statistically significant improvementith gefitinib versus carboplatin/paclitaxel in patients with EGFRutation-positive NSCLC (n = 186; HR 0.54; 95% CI 0.38, 0.79;

104 (2017) 119–125 121

p = 0.0012; median PFS 10.9 vs. 7.4 months, respectively) (Table 2;Fig. 2B).

3.3. Objective response rate and disease progression by BICR

Also consistent with the original investigator-assessed results,ORR by BICR was significantly higher with gefitinib (67.0%) versuscarboplatin/paclitaxel (40.8%) in patients with EGFR mutation-positive NSCLC (OR 3.00; 95% CI 1.63, 5.54; p = 0.0004) (Table 2).

Concordance analysis (Appendix Table A1) of investigator-assessed ORR versus BICR-assessed ORR revealed a concordancerate of 78.0%; equivalent analyses for investigator-assessed diseaseprogression versus BICR-assessed disease progression revealed aconcordance rate of 71.0%.

Furthermore, the median DoR by BICR in patients with EGFRmutation-positive NSCLC was 9.6 months (95% CI 7.4, 12.5) in thegefitinib arm, similar to the original investigator-assessed DoR withgefitinib (median 8.7 months), and 5.5 months (95% CI 4.1, 5.7) inthe carboplatin/paclitaxel arm.

3.4. BICR results by baseline characteristics and EGFR mutationsubtype

No clinically significant differences were noted in PFS or ORRdata by different socio-demographic/clinical subgroups of patientsaccording to BICR (Table 3). PFS and ORR by BICR data by EGFRmutation subtype are presented in Appendix Table A2. PFS byEGFR mutation subtype exon 19 deletion (Fig. 3A) was significantlygreater with gefitinib versus carboplatin/paclitaxel (n = 100; HR0.43, 95% CI 0.25, 0.73; p = 0.0018; median PFS 11.0 vs. 6.9 months,respectively); however, PFS by L858R mutation subtype (Fig. 3B)was similar between treatment arms (benefit of gefitinib versuscarboplatin/paclitaxel was milder with L858R mutation comparedwith exon 19 deletion) (n = 81; HR 0.81, 95% CI 0.47, 1.41; p = 0.453;median PFS 8.6 vs. 7.7 months, respectively).

4. Discussion

The results of BICR analysis of IPASS data in a subset of patientswith EGFR mutation-positive NSCLC, performed at the request ofthe FDA, were highly consistent with the original, investigator-assessed results. PFS was significantly higher in patients withEGFR mutation-positive NSCLC who received gefitinib versus car-boplatin/paclitaxel according to both investigator-assessment (HR0.48) and BICR (HR 0.54). Furthermore, as per the original IPASSstudy [6], crossing of the Kaplan-Meier curves for PFS did not occurbetween the mutation-positive and mutation-negative subgroups,indicating distinct trajectories. A similar pattern was observed withORR (OR 2.75 by investigator assessment and OR 3.00 by BICR),and the median DoR by BICR (9.6 months) was similar to thatof the original IPASS investigation (8.7 months). These indepen-dent review findings, therefore, are aligned with the original IPASSstudy conclusion—and clinical guidelines [2,3]—that EGFR mutationanalysis should be conducted at diagnosis in order to determinean appropriate first-line treatment approach. For patients withEGFR mutation-positive NSCLC, personalized TKI therapy is advo-cated, whereas chemotherapy may be beneficial for those withEGFR mutation-negative NSCLC (noting that patients with anaplas-tic lymphoma kinase [ALK] mutation-positive/EGFR mutation-negative NSCLC may be eligible for ALK-targeted therapy).

As baseline demographics were similar between the overallIPASS EGFR mutation-positive and BICR IPASS populations, it fol-

lows that no clinically significant differences in PFS, ORR, andOS data were noted by different socio-demographic/clinical sub-groups of patients according to BICR. However, there was a trendfor better efficacy by EGFR mutation subtype in terms of PFS and

122 Y.-L. Wu et al. / Lung Cancer 104 (2017) 119–125

Fig. 1. Patient disposition (CONSORT diagram).

Table 1Patient demographics.

Overall IPASS population(n = 1217)

EGFR mutation-positive: originalIPASS population (n = 261)

EGFR mutation-positive: IPASSBICR population (n = 186)

GenderMale 252/1217 (20.7%) 50/261 (19.2%) 32/186 (17.2%)Female 965/1217 (79.3%) 211/261 (80.8%) 154/186 (82.8%)

WHO PS0, 1 1091/1217 (89.6%) 241/261 (92.3%) 175/186 (94.1%)2 126/1217 (10.4%) 20/261 (7.7%) 11/186 (5.9%)

Smoking statusNever smoked 1140/1217 (93.7%) 246/261 (94.3%) 178/186 (95.7%)Ex-smoker 77/1217 (6.3%) 15/261 (5.7%) 8/186 (4.3%)

Data are presented as n/N (%).Abbreviations: BICR, blind independent central review; EGFR, epidermal growth factor receptor; IPASS, Iressa Pan-ASia Study; PS, performance status; WHO, World HealthOrganization.

Y.-L. Wu et al. / Lung Cancer 104 (2017) 119–125 123

Table 2PFS, ORR, and DoR by BICR.

Median PFS, months ORR, n/N (%) Median DoR, months (95% CI)

Gefitinib Carboplatin/paclitaxel

HR (95% CI)p-value

Gefitinib Carboplatin/paclitaxel

OR (95% CI)p-value

Gefitinib Carboplatin/paclitaxel

Overall IPASS population(n = 1217)

5.7 5.8 0.74 (0.65, 0.85)p < 0.001

262/609 (43.0%) 196/608 (32.2%) 1.59 (1.25, 2.01)p < 0.0001

NA NA

EGFR mutation-positive: originalIPASS population (n = 261)

9.5 6.3 0.48 (0.36, 0.64)p < 0.001

94/132 (71.2%) 61/129 (47.3%) 2.75 (1.65, 4.60)p < 0.0001

8.7 NA

EGFR mutation-positive: IPASSBICR population (n = 186)

10.9 7.4 0.54 (0.38, 0.79)p = 0.0012

59/88 (67.0%) 40/98 (40.8%) 3.00 (1.63, 5.54)p = 0.0004

9.6(7.4, 12.5)

5.5(4.1, 5.7)

Abbreviations: BICR, blind independent central review; CI, confidence interval; DoR, duration of response; EGFR, epidermal growth factor receptor; HR, hazard ratio; ITT,intent-to-treat; IPASS, Iressa Pan-ASia Study; NA, not applicable; OR, odds ratio; ORR, objective response rate; PFS, progression-free survival.

Fig. 2. A. PFS by original IPASS investigator-assessment (EGFR mutation-positive ITT population). B. PFS by BICR (EGFR mutation-positive ITT population).

Table 3BICR results by baseline characteristics.

Characteristics Gefitinib; carboplatin/paclitaxel, N ORR, OR (95% CI) PFS, HR (95% CI)

SexMale 12; 20 1.870 (0.389, 8.989) 0.384 (0.152, 0.973)Female 76; 78 3.144 (1.606, 6.155) 0.573 (0.380, 0.864)

Age<65 years 64; 67 2.905 (1.385, 6.090) 0.542 (0.349, 0.841)≥65 years 24; 31 3.442 (0.953, 12.426) 0.559 (0.263, 1.188)

WHO PS0 22; 31 1.888 (0.608, 5.865) 0.448 (0.197, 1.021)1 60; 62 3.569 (1.663, 7.656) 0.625 (0.403, 0.970)2 6; 5 2.667 (0.158, 45.141) NA

Smoking statusNever smoked 85; 93 3.062 (1.639, 5.721) 0.571 (0.391, 0.833)Ex-smoker 3; 5 2.480 (0.060, 101.752) NA

Disease statusLocally advanced 16; 22 3.629 (0.846, 15.573) 0.761 (0.252, 2.303)Metastatic 72; 76 3.010 (1.504, 6.025) 0.519 (0.347, 0.776)

A R, hap n.

Ohi

bbreviations: BICR, blind independent central review; CI, confidence interval; Hrogression-free survival; PS, performance status; WHO, World Health Organizatio

RR, with exon 19 deletion versus L858R mutation. This trendas been observed in several Phase III randomized trials (includ-

ng in the original IPASS study) [7,16,17], and results of a direct

zard ratio; NA, not available; OR, odds ratio; ORR, objective response rate; PFS,

meta-analysis of seven studies of patients with advanced NSCLCreceiving first-line EGFR TKI therapy indicated that patients withtumors harboring an exon 19 deletion were significantly associ-

124 Y.-L. Wu et al. / Lung Cancer 104 (2017) 119–125

letion

am

aBipagiltorhaoasci(ampst

oipttutw

Fig. 3. A. PFS by BICR (EGFR mutation-positive ITT population) − exon 19 de

ted with longer PFS versus patients with tumors harboring L858Rutations (HR 0.75; 95% CI 0.65, 0.85; p < 0.001) [18].Concordance analysis of investigator assessment versus BICR

ssessment of ORR revealed a moderate concordance rate (78%).etween treatment arms, concordance and sensitivity were higher

n the gefitinib arm (83% and 92%, respectively) versus the carbo-latin/paclitaxel arm (74% and 78%, respectively). With sensitivity

measure of the probability of both the original IPASS investi-ators and independent reviewers observing objective responsen a tumor, these data suggest that investigators were moreikely to report objective response in the gefitinib arm versushe chemotherapy arm. On the other hand, specificity (a measuref probability of investigator-assessment detecting no objectiveesponse in a tumor that does not actually show response) wasigher in the carboplatin/paclitaxel arm (71%) versus the gefitinibrm (66%), suggesting that progressive disease was more oftenbserved in the chemotherapy arm by investigators. Supportively,

similar concordance analysis of disease progression indicated thatensitivity was higher in the carboplatin/paclitaxel arm (95.1%)ompared with the gefitinib arm (85.4%), and specificity was highern the gefitinib arm (45.0%) versus the carboplatin/paclitaxel arm40.5%). These findings are likely due to the fact that responsesre seen more frequently with gefitinib whilst progressions areore frequent with chemotherapy. This phenomenon—that the

revalence of disease or detected response directly influences theensitivity and specificity in opposite directions—is well-known inhe field of diagnostic test development [19].

Although acknowledged as a valuable auditing step, BICRf tumor scans is associated with certain limitations: follow-ng investigator-determined progression, further imaging of theatient may have been suspended with their removal from the pro-ocol; however, if independent reviewers determine progression

o occur after investigator-determined progression, they would benable to ascertain time of BICR progression. In IPASS, investiga-ors determined progression to have occurred in 44 patients forhom independent reviewers did not declare progression to have

. B. PFS by BICR (EGFR mutation-positive ITT population) − L858R mutation.

occurred; these patients were included in BICR analyses, yet cen-sored at the last recorded follow-up. In addition, progression wasdetermined to occur in 10 patients by independent reviewers, butnot investigators [9].

5. Conclusion

BICR analysis of IPASS data in a subset of patients withEGFR mutation-positive NSCLC, performed at the request of theFDA, are highly consistent with the original, investigator-assessedresults, further supporting the validity of the IPASS findings: EGFRmutation-positive status remains a statistically and clinically sig-nificant predictor of response to EGFR TKI therapy in the first-linesetting.

Role of the funding source

IPASS was sponsored by AstraZeneca, and designed and coordi-nated by the principal investigators and study steering committeein collaboration with AstraZeneca. All authors had unrestrictedaccess to the original IPASS study data (collected and analyzed byAstraZeneca) and BICR data.

Funding

This study was supported by AstraZeneca.

Acknowledgments

We thank Rosemary Taylor, who provided additional statisticalsupport. We thank Louise Brown, from Complete Medical Com-munications, who provided medical writing support (copyediting,editorial and production assistance) funded by AstraZeneca. IRESSA

is a trademark of the AstraZeneca group of companies.

Appendix.

Y.-L. Wu et al. / Lung Cancer 104 (2017) 119–125 125

Table A1Concordance analysis of investigator-assessed ORR and disease progression versus BICR-assessed ORR and disease progression.

Endpoint Treatmentarm

Concordanceparameter

Concordance 95% CI Sensitivity 95% CI Specificity 95% CI PPV 95% CI NPV 95% CI

ORR Overall 145/186(78.0%)

71.3, 83.7 85/99(85.9%)

77.4, 92.0 60/87(69.0%)

58.1, 78.5 85/112(75.9%)

66.9, 83.5 60/74(81.1%)

70.3, 89.3

Gefitinib 73/88(83.0%)

73.4, 90.1 54/59(91.5%)

81.3, 97.2 19/29(65.5%)

45.7, 82.1 54/64(84.4%)

73.1, 92.2 19/24(79.2%)

57.8, 92.9

Carboplatin/paclitaxel

72/98(73.5%)

63.6, 81.9 31/40(77.5%)

61.5, 89.2 41/58(70.7%)

57.3, 81.9 31/48(64.6%)

49.5, 77.8 41/50(82.0%)

68.6, 91.4

Diseaseprogression

Overall 132/186(71.0%)

63.9, 77.4 99/109(90.8%)

83.8, 95.5 33/77(42.9%)

31.6, 54.6 99/143(69.2%)

61.0, 76.7 33/43(76.7%)

61.4, 88.2

Gefitinib 59/88(67.0%)

56.2, 76.7 41/48(85.4%)

72.2, 93.9 18/40(45.0%)

29.3, 61.5 41/63(65.1%)

52.0, 76.7 18/25(72.0%)

50.6, 87.9

Carboplatin/paclitaxel

73/98(74.5%)

64.7, 82.8 58/61(95.1%)

86.3, 99.0 15/37(40.5%)

24.8, 57.9 58/80(72.5%)

61.4, 81.9 15/18(83.3%)

58.6, 96.4

Data are presented as n/N (%) except for 95% CI.Concordance analysis performed by comparing investigator-assessed versus BICR results of the 186 patient samples available for BICR.Abbreviations: BICR, blind independent central review; CI, confidence interval; NPV, negative predictive value; ORR, objective response rate; PPV, positive predictive value.

Table A2BICR results by EGFR mutation subtype.

EGFR mutation subtype Gefitinib; carboplatin/paclitaxel, N ORR, OR (95% CI) PFS, HR (95% CI)

Exon 19 deletion detected 44; 56 4.194 (1.737, 10.129) 0.430 (0.254, 0.730)Exon 21 L858R detected 45; 36 2.071 (0.826, 5.197) 0.810 (0.466, 1.405)Exon 20 T790M detected 4; 6 1.500 (0.055, 40.633) 1.539 (0.247, 9.593)Other mutations detected 1; 4 NA NA

A R, epidO

R

[

[

[

[

[

[

[

[

[

first-line EGFR-TKIs for advanced non-small cell lung cancer: a meta-analysis,PLoS One 9 (2014) e107161.

bbreviations: BICR, blind independent central review; CI, confidence interval; EGFRR, objective response rate; PFS, progression-free survival.

eferences

[1] World Health Organization. GLOBOCAN 2012: Estimated Cancer IncidenceMortality and Prevalence Worldwide in 2012. http://globocan.iarc.fr/Pages/fact sheets population.aspx (Accessed 2 August 2016).

[2] NICE, EGFR-TK mutation testing in adults with locally advanced or metastaticnon-small-cell lung cancer. http://www.nice.org.uk/guidance/dg9, (Accessed2 August 2016).

[3] National Comprehensive Cancer Network Practice guidelines in oncology−version V.4.2016 (non-small-cell lung cancer). http://www.nccn.org/professionals/physician gls/pdf/nscl.pdf, (Accessed 2 August 2016).

[4] S. Dearden, J. Stevens, Y.L. Wu, D. Blowers, Mutation incidence andcoincidence in non small-cell lung cancer: meta-analyses by ethnicity andhistology (mutMap), Ann. Oncol. 24 (2013) 2371–2376.

[5] Y. Shi, J.S. Au, S. Thongprasert, S. Srinivasan, C.M. Tsai, M.T. Khoa, K. Heeroma,Y. Itoh, G. Cornelio, P.C. Yang, A prospective, molecular epidemiology study ofEGFR mutations in Asian patients with advanced non-small-cell lung cancerof adenocarcinoma histology (PIONEER), J. Thorac. Oncol. 9 (2014) 154–162.

[6] T.S. Mok, Y.-L. Wu, S. Thongprasert, C.-H. Yang, D.-T. Chu, N. Saijo, P.Sunpaweravong, B. Han, B. Margono, Y. Ichinose, Y. Nishiwaki, Y. Ohe, J.-J.Yang, B. Chewaskulyong, H. Jiang, E.L. Duffield, C.L. Watkins, A.A. Armour, M.Fukuoka, Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma,N. Engl. J. Med. 361 (2009) 947–957.

[7] M. Fukuoka, Y.-L. Wu, S. Thongprasert, P. Sunpaweravong, S.-S. Leong, V.Sriuranpong, T.-Y. Chao, K. Nakagawa, D.-T. Chu, N. Saijo, E. Duffield, Y.Rukazenkov, G. Speake, H. Jiang, A.A. Armour, K.F. To, C.-H. Yang, T.S. Mok,Biomarker analyses and final overall survival results from a Phase IIIrandomized, open-label, first-line study of gefitinib versuscarboplatin/paclitaxel in clinically selected patients with advancednon-small-cell lung cancer in Asia (IPASS), J. Clin. Oncol. 29 (2011) 2866–2874.

[8] Y.L. Wu, M. Fukuoka, T.S. Mok, N. Saijo, S. Thongprasert, J.C. Yang, D.T. Chu, J.J.Yang, Y. Rukazenkov, Tumor response and health-related quality of life inclinically selected patients from Asia with advanced non-small-cell lungcancer treated with first-line gefitinib: post hoc analyses from the IPASSstudy, Lung Cancer 81 (2013) 280–287.

[9] L.E. Dodd, E.L. Korn, B. Freidlin, C.C. Jaffe, L.V. Rubinstein, J. Dancey, M.M.Mooney, Blinded independent central review of progression-free survival inphase III clinical trials: important design element or unnecessary expense? J.

Clin. Oncol. 26 (2008) 3791–3796.

10] Food and Drug Administration, Food and drug administration (FDA) guidancefor industry: Clinical trial endpoints for the approval of cancer drugs andbiologics. http://www.fda.gov/downloads/Drug/Guidances/ucm071590.pdf,(Accessed 2 August 2016).

[

ermal growth factor receptor; HR, hazard ratio; NA, not available; OR, odds ratio;

11] L.V. Sequist, J.C. Yang, N. Yamamoto, K. O’Byrne, V. Hirsh, T. Mok, S.L. Geater, S.Orlov, C.M. Tsai, M. Boyer, W.C. Su, J. Bennouna, T. Kato, V. Gorbunova, K.H.Lee, R. Shah, D. Massey, V. Zazulina, M. Shahidi, M. Schuler, Phase III study ofafatinib or cisplatin plus pemetrexed in patients with metastatic lungadenocarcinoma with EGFR mutations, J. Clin. Oncol. 31 (2013) 3327–3334.

12] Y.L. Wu, C. Zhou, C.P. Hu, J. Feng, S. Lu, Y. Huang, W. Li, M. Hou, J.H. Shi, K.Y.Lee, C.R. Xu, D. Massey, M. Kim, Y. Shi, S.L. Geater, Afatinib versus cisplatinplus gemcitabine for first-line treatment of Asian patients with advancednon-small-cell lung cancer harbouring EGFR mutations (LUX-Lung 6): anopen-label, randomised phase 3 trial, Lancet Oncol. 15 (2014) 213–222.

13] E.A. Eisenhauer, P. Therasse, J. Bogaerts, L.H. Schwartz, D. Sargent, R. Ford, J.Dancey, S. Arbuck, S. Gwyther, M. Mooney, L. Rubinstein, L. Shankar, L. Dodd,R. Kaplan, D. Lacombe, J. Verweij, New response evaluation criteria in solidtumours: revised RECIST guideline (version 1.1), Eur. J. Cancer 45 (2009)228–247.

14] C.R. Newton, A. Graham, L.E. Heptinstall, S.J. Powell, C. Summers, N. Kalsheker,J.C. Smith, A.F. Markham, Analysis of any point mutation in DNA. Theamplification refractory mutation system (ARMS), Nucleic Acids Res. 17(1989) 2503–2516.

15] D. Whitcombe, J. Theaker, S.P. Guy, T. Brown, S. Little, Detection of PCRproducts using self-probing amplicons and fluorescence, Nat. Biotechnol. 17(1999) 804–807.

16] M. Maemondo, A. Inoue, K. Kobayashi, S. Sugawara, S. Oizumi, H. Isobe, A.Gemma, M. Harada, H. Yoshizawa, I. Kinoshita, Y. Fujita, S. Okinaga, H. Hirano,K. Yoshimori, T. Harada, T. Ogura, M. Ando, H. Miyazawa, T. Tanaka, Y. Saijo, K.Hagiwara, S. Morita, T. Nukiwa, for the North-East Japan Study Group,Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR,N. Engl. J. Med. 362 (2010) 2380–2388.

17] L.V. Sequist, R.G. Martins, D. Spigel, S.M. Grunberg, A. Spira, P.A. Jänne, V.A.Joshi, D. McCollum, T.L. Evans, A. Muzikansky, G.L. Kuhlmann, M. Han, J.S.Goldberg, J. Settleman, A.J. Iafrate, J.A. Engelman, D.A. Haber, B.E. Johnson, T.J.Lynch, First-line gefitinib in patients with advanced non-small-cell lungcancer harboring somatic EGFR mutations, J. Clin. Oncol. 26 (2008) 2442–2449.

18] Y. Zhang, J. Sheng, S. Kang, W. Fang, Y. Yan, Z. Hu, S. Hong, X. Wu, T. Qin, W.Liang, L. Zhang, Patients with exon 19 deletion were associated with longerprogression-free survival compared to those with L858R mutation after

19] A.C. Janssens, Y. Deng, G.J. Borsboom, M.J. Eijkemans, J.D. Habbema, E.W.Steyerberg, A new logistic regression approach for the evaluation ofdiagnostic test results, Med. Decis. Making 25 (2005) 168–177.