`
`Journal of Thoracic Oncology • Volume 8, Supplement 2, November 2013
`
`I, 1 in stage II, 7 in stage III and 18 in stage IV. 16 patients, all with
`metastatic disease, were treated with crizotinib (15 with 250mg 1x2
`and 1 patient with 200mg 1x1). Crizotinib was given to: 0 patients
`as 1st line, 4 patients as 2d line, 6 as 3d line, 5 as 4th line and 1 as 5th
`line. At time of data collection 9 patients had discontinued crizo-
`tinib-therapy. 7 patients had ongoing treatment with an average
`duration of 125 days. 12/16 patients obtained partial remission, 3
`stable disease and 1 disease progression. 3/9 discontinued crizo-
`tinib-therapy due do severe side effects: 1 due to persistant visus
`toxicity grade ≥2, 1 due to pneumonitis occuring at treatment day
`42 and 1 due to liver toxicity with CTCAE grade ≥2 occurring at
`treatment day 37. Only the case with pneumonitis resultated in
`death at day 43. No QTc-syndromes and no hematological toxicity
`CTCAE grade ≥3 occurred.
`
`Conclusion: Identifying patients with ALK-EML4-translocations,
`the predictive factor for crizotinib-treatment, offers new treatment
`options and realistically balanced hope in the severe setting of
`metastatic NSCLC. In our experience, the predictive value of a po-
`sitive ALK-EML4-test is high on histological material and crizotinib
`offers good treatment outcomes after progression on platinum-
`based chemotherapy but for shorter time than what is expected
`for TKIs in patients with activating EGFR-mutations.
`
`Keyword: ALK-EML4-translocations, metastatic NSCLC, predictive
`factors
`
`POSTER SESSION 2 - NSCLC NOVEL THERAPIES
`TUESDAY, OCTOBER 29, 2013 - 09:30-16:30
`
`P2.11-034 INDIRECT COMPARISONS OF HARM/BENEFIT
`PROFILE OF EGFR TYROSINE KINASE INHIBITORS AS
`FIRST LINE TREATMENT IN EGFR MUTATED NSCLC PATI-
`ENTS: A SYSTEMATIC REVIEW AND META-ANALYSIS
`
`Eva R. Haspinger1, Francesco Agustoni1, Francesco Gelsomino1,
`Valter Torri2, Marina C. Garassino1, Michela Cinquini2
`
`1Medical Oncology, Fondazione IRCCS Istituto Nazionale Dei Tumori/
`Italy, 2Laboratorio Di Metodologia Per La Ricerca Biomedica, IRCCS -
`Istituto Di Ricerche Farmacologiche "mario Negri“/Italy
`
`Background: To date, three EGFR Tyrosine Kinase inhibitors (TKIs)
`gefitinib (G), erlotinib (E), and afatinib (A) have been compared to
`standard chemotherapy as first line treatment in patients with ad-
`vanced NSCLC harboring EGFR mutations. We performed a syste-
`matic review and meta -analysis in order to estimate through indi-
`rect comparisons the relative risk benefit associated to each drug.
`
`Methods: The major databases were searched for published and
`unpublished randomized control trial up to March 2013. Data ext-
`raction was performed by two independent reviewers and focused
`on benefit (ORR, PFS) and selected harm outcomes (diarrhea, rash,
`nail disorders, hypertransaminasemia). The adjusted indirect com-
`parisons were performed using the random effect method descri-
`bed by Bucher and Glenny approach for Hazard Ratio (HR) for PFS
`and relative risk (RR) for the other outcome measures.
`
`Results: All EGFR TKIs fared better when compared with chemo-
`therapy in terms of PFS: overall HR 0.40 (95%CI 0.30-0.54); G vs E
`HR 1.34 (95%CI 0.63-2.86), G vs A HR 0.74 (95%CI 0.53-1.04), E vs
`A HR 0.55 (95%CI 0.31-0.99). The relative probability of ORR was
`G vs E 0.96 (95%CI 0.69-1.34), G vs A 0.79 (95%CI 0.49-1.28), E vs
`A 0.82 (95%CI 0.49-1.38). Indirect comparisons for safety showed
`RR for diarrhea G vs E 0.8 (95%CI 0.63-1.01), G vs A 0.32 (95%CI
`
`0.20-0.51), E vs A 0.38 (95%CI 0.24-0.62); for rash G vs E 1.0 (95%CI
`0.82-1.22), G vs A 0.31 (95%CI 0.15-0.65), E vs A 0.31 (95%CI 0.15-
`0.65); for hypertransaminasemia G vs E 2.29 (95%CI 1.63-3.23). Nail
`disorders affected 57% of patients treated with A, 15% with G, and
`4% with E.
`
`Conclusion: Results of our analysis showed that all treatments
`have similar activity and efficacy while the toxicity profile was less
`favorable for A with a significant higher risk of diarrhea, rash, and
`nail disorders. Based on these safety results, we suggest that A
`may not be the first choice for upfront treatment in EGFR mutated
`patients. Confirmation is warranted by ongoing prospective head
`to head RCTs.
`
`Keywords: first line, EGFR-TKI, Metanalysis
`
`POSTER SESSION 2 - NSCLC NOVEL THERAPIES
`TUESDAY, OCTOBER 29, 2013 - 09:30-16:30
`
`P2.11-035 ASSOCIATION OF TUMOR PD-L1 EXPRESSION
`AND IMMUNE BIOMARKERS WITH CLINICAL ACTIVITY
`IN PATIENTS WITH NON-SMALL CELL LUNG CANCER
`(NSCLC) TREATED WITH NIVOLUMAB (ANTI-PD-1; BMS-
`936558; ONO-4538)
`
`Scott J. Antonia1, Joseph F. Grosso2, Christine E. Horak2, Christo-
`pher T. Harbison2, John F. Kurland2, H David Inzunza2, Ashok Gup-
`ta2, Vindira Sankar2, Jong-Soon Park2, Maria Jure-Kunkel2, James
`Novotny2, John Cogswell2, Xiaoling Zhang3, Therese Phillips3, Pau-
`line Simmons3, Jason Simon2
`
`1H. Lee Moffitt Cancer Center & Research Institute/United States Of
`America, 2Bristol-Myers Squibb/United States Of America, 3Dako North
`America/United States Of America
`
`Background: The immune checkpoint receptor programmed
`death-1 (PD-1) negatively regulates T-cell activation upon in-
`teraction with its ligands, PD-L1 and PD-L2. In a Phase 1 dose-
`escalation/cohort expansion study (CA209-003; NCT00730639),
`nivolumab, a fully human PD-1 receptor blocking antibody,
`delivered durable responses in patients with solid tumors, inclu-
`ding advanced NSCLC. Immunohistochemistry (IHC) analysis of
`tumor samples from this study suggested an association between
`pre-treatment tumor PD-L1 expression and clinical response to
`nivolumab in patients with melanoma (Grosso JF J Clin Oncol.
`2013;31(suppl):abs 3016; Topalian SL NEJM 2012;366:2443-54).
`Here we investigate the association between PD-L1 expression
`by IHC and response to nivolumab in patients with NSCLC, and
`patient response with pre-/post-dose absolute lymphocyte counts
`(ALC) and selected lymphocyte cell subsets.
`
`Methods: 129 patients with NSCLC from the CA209-003 trial re-
`ceived nivolumab between 2008 and 2012 (1–10 mg/kg IV every 2
`weeks) during dose escalation and/or cohort expansion. Archived
`formalin-fixed paraffin-embedded pre-treatment tumor tissue and
`pre-treatment and on-treatment peripheral whole blood samples
`were analyzed to explore potential pharmacodynamic/predictive
`biomarkers associated with nivolumab therapy. Pre-treatment tu-
`mor PD-L1 expression was evaluated by IHC using an automated
`assay developed by Dako based on a sensitive and specific anti-
`PD-L1 monoclonal antibody (28-8). Tumors were defined as PD-L1
`positive (PD-L1+) when ≥5% of the tumor cells had membrane stai-
`ning at any intensity. Lymphocyte subsets in the periphery were
`measured using flow cytometry.
`
`Copyright © 2013 by the International Association for the Study of Lung Cancer
`
`S907
`
`
`
`Abstracts
`
`Journal of Thoracic Oncology • Volume 8, Supplement 2, November 2013
`
`Results: Tumor membrane PD-L1 expression was measured in
`63 patients with NSCLC (29 squamous; 34 non-squamous). 31/63
`(49%) NSCLC biopsies were PD-L1+. There was no apparent associ-
`ation between PD-L1 protein expression and NSCLC histology: for
`squamous and non-squamous tumors, 52% (15/29) and 47% (16/34)
`were PD-L1+, respective. Objective response rates for PD-L1+ and
`PD-L1- NSCLC patients with non-squamous and squamous histolo-
`gy are shown in the Table. Objective responses were observed in
`patients with squamous and non-squamous NSCLC who were ne-
`gative for PD-L1 expression. Since increases in on-treatment ALC
`and activated T-cell phenotypes have been shown to positively
`associate with favorable clinical outcomes in ipilimumab monothe-
`rapy (Ku GY Cancer 2010;116:1767-75; Carthon BC Clin Cancer Res
`2010;16:2861-71), results from an analysis correlating patient res-
`ponse with pre-/post-dose ALC and T-cell populations in patients
`with NSCLC receiving nivolumab will be presented.
`
`Table. Patient response according to PD-L1 expression status in
`patients with NSCLC
`
`Tumor type
`
`NSCLC
`(all patients)
`
`NSCLC
`(squamous)
`
`NSCLC (non-
`squamous)
`
`PD-L1 expression
`status
`
`+
`–
`+
`–
`+
`–
`
`Objective
`response rate,
`n/N (%)
`5/31 (16.1)
`4/32 (12.5)
`2/15 (13.3)
`3/14 (21.4)
`3/16 (18.8)
`1/18 (5.6)
`
`Conclusion: Further evaluation of PD-L1 as a molecular marker
`of nivolumab therapy is required. Association of PD-L1 protein
`expression with clinical outcome is currently being prospectively
`assessed in ongoing Phase 3 trials. Clinical Trial Registration Num-
`ber: NCT00730639
`
`Keywords: nivolumab, non-small cell lung cancer, programmed
`death-1 receptor ligand 1, biomarker
`
`type tumors. The relationship between EGFR activating mutations
`and their primary tumor location in the lungs was not reported
`before.
`
`Methods: We retrospectively reviewed the data of our pulmonary
`adenocarcinoma patients who had received complete staging and
`received tumor EGFR mutation analysis. The association between
`EGFR mutation status, patients smoking status, patient’s gender
`and primary tumor location were analyzed.
`
`Results: 205 cases were reviewed. There are 126 patients with
`tumor EGFR mutations, including 115 patients with classic EGFR
`mutations (exon 19 deletions or L858R), and 79 patients were wit-
`hout EGFR mutation. There are statistically significant association
`between tumor EGFR mutations and primary tumor location in
`right upper lobe (P=0.007); especially in RB1 segment (P=0.018),
`and primary tumor location of exon 19 deletions occurred more
`frequently in right upper lobe (P<0.001). There were no significant
`associations between patients smoking status and primary tumor
`location(P=0.659), nor was patients gender and primary tumor
`location (P=0.473).
`
`Conclusion: There are statistically significant association between
`EGFR mutation and primary tumor location in right upper lobe of
`patients with adenocarcinoma
`of the lungs.
`
`Keywords: adenocarcinoma of lung, Tumor EGFR Mutation, tumor
`location
`
`POSTER SESSION 2 - NSCLC NOVEL THERAPIES
`TUESDAY, OCTOBER 29, 2013 - 09:30-16:30
`
`P2.11-037 EGFR ACTIVATING MUTATION AND BONE
`METASTASIS HAVE ASSOCIATION WITH CNS METASTA-
`SIS AT DIAGNOSIS IN PATIENTS WITH NON-SMALL CELL
`LUNG CANCER
`
`Ki Hwan Kim, Hyo-Jeong Lim, Jin Hyun Park, Jin-Soo Kim, In Sil
`Choi
`
`POSTER SESSION 2 - NSCLC NOVEL THERAPIES
`TUESDAY, OCTOBER 29, 2013 - 09:30-16:30
`
`Department Of Internal Medicine, Seoul National University Boramae
`Medical Center/Korea
`
`P2.11-036 ASSOCIATION BETWEEN TUMOR EGFR MU-
`TATION AND PRIMARY TUMOR LOCATION IN PATIENTS
`WITH ADENOCARCINOMA OF THE LUNGS
`
`Chu-Yun Huang1, Yung-Hung Luo1, Chieh-Hung Wu1, Chih-Wei
`Wu1, Wen-Shuo Wu1, Chun-Ming Tsai2, Yu-Chin Lee2, Reury-Perng
`Perng1, Jacqueline Whang-Peng3, Yuh-Min Chen2
`
`1Department Of Chest Medicine, Taipei Veterans General Hospital/
`Taiwan, 2Division Of Thoracic Oncology, Department Of Chest Medi-
`cine, Taipei Veterans General Hospital, School Of Medicine, National
`Yang-Ming University/Taiwan, 3Centre Of Excellence Cancer Research
`(CECR), Taipei Medical University/Taiwan
`
`Background: Lung cancer is the leading cause of cancer death in
`the world, and the non-small cell lung cancer accounts for more
`than 80% of the lung cancer. Among patients with non-small cell
`lung cancer, tumor epidermal growth factor receptor (EGFR) acti-
`vating mutations were mostly found in patients with adenocarcino-
`ma and were associated with a better prognosis than EGFR wild-
`
`Background: CNS metastasis happens not infrequently in patients
`with non-small cell lung cancer, reaching upto 25%. The presence
`of CNS metastasis is a major decision factor of planning primary
`treatment. Furthermore, predictors for CNS metastasis could be
`used for selecting patients who may get the potential benefit of
`prophylactic cranial irradiation.
`
`Methods: We retrospectively analyzed the clinicopathologic data
`of 233 patients with non-small cell lung cancer who underwent
`brain MRI at the time of diagnosis between Jan 2008 and Dec
`2012. Chi-square analysis and multivariate logistic regression mo-
`del was used to find risk factors for CNS metastasis.
`
`Results: Forty-five (19.3%) patients had initial CNS metastasis (41
`brain parenchymal metastasis and 4 leptomeningeal seeding).
`Chi-square analysis revealed that never-smoking (28.7% vs. 13.7%,
`P=0.005), lung metastasis (29.6% vs 14.8%, P=0.009), bone me-
`tastasis (35.7% vs 13.1%, P<0.001), adenocarcinoma (24.8% vs
`10.9%, P=0.008), and EGFR activating mutation (44.4% vs 18.3%,
`P=0.004) were associated with CNS metastasis. However, pleural,
`
`S908
`
`Journal of Thoracic Oncology • Volume 8, Supplement 2, November 2013
`
`