Virtual Library

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Background:
Integrated 18F-FDG, PET-CT has shown somewhat variable sensitivity and specificity for nodal staging in tuberculosis endemic areas. This variation is mainly because PET scans show falsely increased 18F-FDG uptake in inflammatory nodes, which may be observed in lymph nodes containing calcification or showing higher attenuations than those of surrounding vessels on unenhanced CT scans. The AIM of the study was to evaluate the efficacy of PET-CT for mediastinal nodal staging in non-small cell lung cancer (NSCLC) patients in a tuberculosis-endemic country.

Methods:
From February 2012 to February 2016, a total of 160 patients underwent surgery for pathologically proven NSCLC. Patients who received neoadjuvant treatment were excluded from the study. Assessment of the diagnostic efficacy of integrated PET- CT for detecting nodal metastasis was performed in 46 patients (Male to Female ratio:4; mean age- 55 years). Patients underwent an integrated PET/CT examination and subsequent surgical nodal staging. Nodes showing greater 18F-FDG uptake at PET without benign calcification or high attenuation >70 household unit (HU) at unenhanced CT were regarded as being positive for malignancy. All patients underwent hilar and mediastinal lymph node dissection according to the AJCC lymph node map after resection of the main tumour. The histologic nodal assessment results were used as reference standards. Of these 46 patients, 10 (20%) had a past medical history of pulmonary tuberculosis as determined by clinical or imaging studies.

Results:
A total of 230 mediastinal nodal stations were evaluated in 46 patients; 5 (2%) stations in 4 (8%) patients proved to be malignant by histopathology. Mean number of lymph node stations were 5. On a per-nodal station basis, PET CT for mediastinal lymph nodes staging has sensitivity: 60%; specificity: 97%; accuracy: 96%; positive predictive value (PPV): 38%; negative predictive value (NPV): 99%.

Conclusions:
Integrated PET-CT provides high specificity and high accuracy, but low sensitivity for mediastinal staging of NSCLC. The high specificity is achieved at the expense of sensitivity by interpreting calcified nodes or nodes with high attenuation at CT, even with high FDG uptake at PET, as benign in a tuberculosis-endemic region.

Clinical trial identification:


Legal entity responsible for the study:
Dr. Bhim Rao Ambedkar Institute Rotary Cancer Hospital, All India Institute Medical Sciences, New Delhi

Funding:
Dr. Bhim Rao Ambedkar Institute Rotary Cancer Hospital, All India Institute Medical Sciences, New Delhi

Disclosure:
All authors have declared no conflicts of interest.

Background:
The newly revised TNM classification shows that cT and pT are to be evaluated with solid component on CT and pathological invasive size in adenocarcinoma of the lung. We evaluated the predictive factors for invasiveness, lymph node metastasis and recurrence in patients with adenocarcinomas of the lung, particularly pathological tumor diameters both of gross and invasive size, and solid component diameter using computed tomography (CT) both with lung window settings and mediastinal window settings.

Methods:
We evaluated 533 patients with lung adenocarcinomas (diameter, 6–132 mm) who underwent surgical resections. 447 of the 533 have been underwent systematic node dissection. Tumors were examined using CT with thin section conditions (1.25-mm thick: high-resolution CT), with tumor dimensions evaluated under two conditions: lung window (LD) and mediastinal window (MD) settings. Both of the tumor size on CT (LD, MD), consolidation component size on CT with lung window settings (C), preoperative serum carcinoembryonic antigen (CEA) levels, pathological tumor diameter both of gross(GS) and invasive lesion(IS), and pathological status [invasion of lymphatic vessels (ly), vascular vessels (v), pleura (pl), lymph node metastasis] were examined. Area under the curve (AUC) of Receiver Operating Characteristic (ROC) was used for evaluation.

Results:
AUCs according to the variables.rnTable: 46PDrn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

Variables	ly(N = 533)	v(N = 533)	pl (N = 533)	Lymph node metastasis (N = 447)	Recurrence (N = 447)	Correlation coefficient IS and other variables
GS	0.73	0.71	0.72	0.72	0.83	N.A
IS	0.86	0.82	0.79	0.79	0.86	N.A
LD	0.71	0.70	0.70	0.71	0.79	0.65
MD	0.87	0.82	0.81	0.80	0.85	0.80
C	0.83	0.80	0.79	0.79	0.83	0.78
C/LD	0.81	0.78	0.75	0.73	0.69	N.A
CEA	0.68	0.66	0.69	0.70	0.72	N.A

rnGS: pathological gross tumor size, IS: Pathological invasive size, LD: Diameter with lung window settings on CT, MD: diameter with mediastinal window settings on CT, C: consolidation diameter with lung window settings on CT.rn

Conclusions:
The solid component evaluated on computed tomography especially by MD and C can predict the invasiveness and lymph node metastasis in lung adenocarcinoma as well as pathological invasive size.

Clinical trial identification:
N.A.

Legal entity responsible for the study:
Yukinori Sakao

Funding:
N/A

Disclosure:
All authors have declared no conflicts of interest.

Abstract not provided

Background:
The presence of solid histologic pattern in lung adenocarcinoma (ADC) is associated with early recurrence. However, individualized prognosis in patients with solid component is still unclear. This study aimed to develop a nomogram predicting the recurrence probability in stage I lung ADC patients with solid component.

Methods:
A total of 5904 patients with stage I lung ADC who underwent curative surgical resection from January 2008 through December 2014 at Shanghai Chest Hospital were retrospectively reviewed. Tumors were subtyped by using the IASCL/ATS/ERS classification. Of these patients, 708 contained a solid component. Prognostic value of gender, age at diagnosis, smoking history, operation type, tumor location, tumor size, pathological subtype, cell differentiation, lymphovascular and visceral pleural invasion were investigated. Multivariate Cox regression analysis of recurrence-free survival (RFS) in patients with solid component was performed and a nomogram to predict RFS was constructed. The nomogram was internally validated.

Results:
The overall recurrence rate in patients with solid component was 25.0% (177/708), with predominant solid subtype and minor solid component in 49.2% (87/177) and 50.8% (90/177) of cases, respectively. Larger tumor size (P = 0.002), predominant solid component (P = 0.003), advanced age at diagnosis (P = 0.015), and visceral pleural invasion (P = 0.040) were associated with an increased risk of recurrence and were included in the nomogram. The predictive model had a concordance index of 0.643 (95% confidence interval, 0.601-0.685) and showed good calibration.

Conclusions:
The nomogram model including identified risk factors for RFS is applicable in treatment decision-making for early stage lung ADC with pathological solid component. External validation is required to recommend this nomogram in routine practice.

Clinical trial identification:


Legal entity responsible for the study:
Shanghai Chest Hospital

Funding:
Shanghai Chest Hospital

Disclosure:
All authors have declared no conflicts of interest.

Background:
The aim of the present study was to assess outcomes among non-small cell lung cancer (NSCLC) patients treated with preoperative chemotherapy followed by Video-Assisted Thoracic Surgery (VATS) lobectomy from a National multi-institutional Registry.

Methods:
A National Registry established in 2013 was used to collect data from 65 Thoracic Surgery Units (>3,700 patients enrolled); only information from Units with >100 VATS lobectomies enrolled were analysed. A retrospective analysis was performed on patients with NSCLC who received preoperative chemotherapy followed VATS lobectomy within one year and compared to a propensity score matched population without preoperative chemotherapy. Propensity score (greedy 5 to 1 digit matching algorithm) estimated with multiple logistic regressions based on selection bias and potential confounding variables produced 221 patients (control group). After propensity score matching, data were compared with the paired Student’s t-test, Pearson’s χ[2] and Fisher’s exact test. Differences were considered to be statistically significant when the p - value was <0.05.

Results:
56/1679 (3.34%) patients met study inclusion criteria. There were no significant differences in baseline characteristics between groups (Table 1a). The majority of patients were clinical stage IIIA, although a small percentage of clinical stage II patients had preoperative therapy. Anatomic distribution of lobectomies and the number of resected lymph nodes not significantly differed between groups. Table 1b presents postoperative histology in the neoadjuvant groups. Table 1c reports short-term perioperative outcomes. No perioperative mortality was recorded in both groups. Overall morbidity (pneumonia, atrial fibrillation) was significantly higher in the neoadjuvant group, but interestingly, all the other variables were not influenced (conversion rate, operative time, blood loss, air leak duration, length of stay).rnTable: 72PDDemographics, postoperative histology/stage of the neoadjuvant group, and selected perioperative/postoperative outcomes. NA = not applicable, SD = standard deviationrn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

Table 1a. Demographics characteristics
Characteristics	Neoadjuvant Group (N = 56)	Propensity Matched Group (N = 221)	p - value
M/F (%)	50	61.99	0.938
Age (mean ± SD)	64.14 ± 10.64	67.44 ± 11.93	0.888
Charlson Index (mean ± SD)	4.19 ± 1.75	4.43 ± 1.85	0.530
ECOG score	0	0	NA
Preoperative stage (N, %) • IIA • IIB • IIIA	13 (23.21) 18 (32.14) 25 (44.64)	64 (28.96) 59 (26.70) 98 (44.34)	0.877 0.568 0.964
Surgical procedure (N, %) • Left upper lobectomy • Left lower lobectomy • Right upper lobectomy • Right lower lobectomy • Lower bilobectomy	8 (14.29) 10 (17.86) 25 (44.64) 12 (21.43) 1 (1.79)	43 (19.46) 29 (13.12) 89 (40.27) 36 (16.29) 3 (1.36)	0.808 0.124 0.628 0.663 0.882

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

Table 1b. Postoperative histology and stage of the Neoadjuvant Group (N, %)
Postoperative histology • Adenocarcinoma • Squamous cell carcinoma • Other	25 (44.64) 12 (21.43) 21 (37.5)
Stage • IA • IB • IIA • IIB • IIIA	7 (12.5) 12 (21.43) 22 (39.29) 1 (1.79) 14 (25)

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

Table 1c. Selected perioperative and postoperative outcomes
Characteristics	Neoadjuvant Group (n = 56)	Propensity Matched Group (n = 221)	p - value
Duration of surgical operation, minutes (mean ± SD)	182.14 ± 65.69	182.76 ± 66.17	0.933
Intraoperative blood loss, mL (mean ± SD)	148.93 ± 153.57	154.19 ± 126.09	0.163
Resected lymph nodes (mean ± SD)	17.54 ± 10.37	14.04 ± 7.65	0.890
Overall conversion to thoracotomy (N, %) • Lymph nodes on pulmonary artery • Anomalies of anatomy • Bleeding	6 (10.71) 3 (5.36) 1 (1.79) 2 (3.57)	20 (9.05) 11 (4.98) 1 (0.45) 8 (3.62)	0.0634 0.193 0.517 0.195
Postoperative air leaks (N, %)	6 (10.71)	12 (5.43)	0.0396
Postoperative complications (N, %) • Pneumonia • Atrial fibrillation	6 (10.71) 4 (7.14)	7 (3.17) 10 (4.52)	0.0398 0.0419
Hospital length of stay, days (mean ± SD)	8.73 ± 6.60	9.54 ± 8.92	0.759

rn

Conclusions:
VATS lobectomy after induction chemotherapy in stage II/IIIA NSCLC is feasible with a favourable profile regarding overall morbidity and mortality. This preliminary report shows that neoadjuvant treatment may not represent per se a contraindication to the VATS approach.

Clinical trial identification:


Legal entity responsible for the study:
Italian VATS Group

Funding:
Italian VATS Group

Disclosure:
All authors have declared no conflicts of interest.

Abstract not provided

Background:
Previous studies have shown that primary tumor metabolic volume (PT-MV) on 18F-FDG-PET/CT could serve as a prognostic factor in patients treated with definitive chemoradiotherapy (CRT). We analyzed a correlation between pre- to posttreatment PT-MV reduction during the course of CRT and overall survival.

Methods:
Sixty-five patients with histologically confirmed NSCLC IIIA-B, treated with definitive CRT in the sequential (21/65 pts, 32.3%) or concurrent (44/65 pts, 67.7%) mode, who underwent 18F-FDG-PET/CT before and about 4-6 weeks after the CRT, were analyzed. Histology yielded 22 (33.8%) adenocarcinomas, 34 (52.3%) squamous cell carcinomas, 8 (12.3%) large-cell carcinomas and 1(1.5%) NSCLC not otherwise specified (NOS). Thirty-five patients (35/65, 53%) were enrolled in randomized clinical trials (16 pts in GILT, 10 pts in InCoDor and 9 pts in BROCAT study CTRT 99/97). The most common concurrent chemotherapy regimen (18/65 pts, 27.7%) consisted of cisplatin given intravenously at a dose of 20mg/m[2] on days 1-4 and oral vinorelbine 50mg/m[2] on days 1, 8 and 15, every 4 weeks for two courses.

Results:
Median OS for the entire cohort was 16 months (95% [CI],11.5–20.5). Complete remission (CR) was reached in 20 (31%), whereas partial and non-response occurred in 31 (48%) and 14 (22%) patients, respectively. The mean change from pre- to posttreatment PT-MV was -51% (range: +125% to -100%). Various cutoffs of PT-MV reduction (100- 90-85-80-70-60 and 50%) after the CRT were analyzed. CR(n = 20) had a median OS of 26 (95 CI: 5-47) vs. 13 months (95 CI:9-16) observed in the rest of treated group (p = 0.01, log-rank test). The results also showed a significant difference in OS favoring patients with PT-MV reduction >90% (n = 18) vs. the rest with a median survival of 24 vs. 13 months, respectively (p = 0.04, log-rank test). However, the association between PT-MV reduction at 80% and 70% and survival showed borderline significance (p = 0.07 and 0.09, log-rank test). PT-MV reduction at 60 and 50% failed to show an effect on OS.

Conclusions:
In this inoperable locally-advanced NSCLC cohort, a PT-MV reduction of at least 90% after definitive CRT correlated with improved OS.

Clinical trial identification:


Legal entity responsible for the study:
Deparment of Radiation Oncology, LMU Munich

Funding:
N/A

Disclosure:
All authors have declared no conflicts of interest.

Background:
Validated outcome prediction models with high discriminative power are important for a cost-effective deployment of proton therapy for locally-advanced non-small cell lung cancer (NSCLC). We validated a model predicting lung toxicity 6 months after radiotherapy (RT) treatment.

Methods:
The model published by Appelt et al. (Acta Oncol 2014) was selected from a literature search. It relies on a review of radiation pneumonitis reports and retained the most important predictors, Mean Lung Dose (MLD) as dosimetric factor and 6 factors influencing the patient’s susceptibility: pre-existing pulmonary comorbidity, age>63 years, mid/inferior tumor location and sequential chemotherapy as risk factors, and current smoking and smoking history as protective factors. A dataset of 109 NSCLC patients treated at MAASTRO Clinic using 1.8 Gy fraction doses (two fractions per day) up to 79.2 Gy was studied. The required parameters were retrospectively collected together with the dyspnea endpoint (CTC 3.0 scoring) at baseline and at 6 months after RT. All treatments were performed using 3D-conformal RT techniques as it was the case in the study of Appelt et al. Odds ratios were calculated using logistic regression modelling.

Results:
19.3% of patients presented post RT dyspnea≥2. Our dataset confirmed current smoking and pulmonary comorbidity as prognostic factors for this endpoint, with similar odds ratios (OR = 0.28 (p = 0.02) and OR = 2.95 (p = 0.02), respectively). Tumor location OR was outside of the reported 95% CI. When predicting the change in dyspnea with respect to the baseline score (delta dyspnea≥1, prevalence of 18.6%), the two prognostic factors were not significant anymore (OR = 0.56 (p = 0.27) and OR = 0.47 (p = 0.21), respectively). Both factors exhibited strong correlation with the baseline patient status: worse baseline dyspnea is often a manifestation of existing comorbidities and it determines the probability to stop smoking. MLD was not associated with outcome in any of our models.

Conclusions:
It is crucial to consider delta toxicity endpoints in prediction modeling in order to obtain meaningful models reflecting radiotherapy outcome. Acknowledgement: This project has received funding from the EU under grant agreement no 601826 (REQUITE).

Clinical trial identification:


Legal entity responsible for the study:
KU Leuven

Funding:
EU

Disclosure:
All authors have declared no conflicts of interest.

Abstract not provided

Background:
Durvalumab (MEDI4736) is an engineered human IgG1 mAb targeting programmed cell death ligand-1 (PD-L1). Treatment with other anti-PD-1/PD-L1 antibodies has demonstrated meaningful clinical benefit in patients with advanced NSCLC.

Methods:
As part of the Lung Master Protocol S1400, a National Clinical Trials Network group "umbrella" trial for second-line SqNSCLC we conducted a phase II study of durvalumab in patients with Stage IV squamous NSCLC (ECOG PS 0-2; ≥1 prior systemic treatment regimens, including one platinum-based), EGFR/ALK wild-type. The primary endpoint was overall response rate (ORR) (RECIST v1.1); secondary endpoints included duration of response (DoR), ORR among PD-L1 positive patients, disease control rate (DCR), investigator-assessed progression-free survival (PFS), overall survival (OS) and safety (CTCAE v4.03). The initial protocol was a randomized phase II/III comparison of durvalumab to docetaxel, and was amended to be a single arm phase II trial of single agent durvalumab. PD-L1 positive tumors were defined by ≥ 25% of tumor cells with membrane staining.

Results:
Of the 68 eligible patients (median [range] age 66 [35-92] years, PS 0/1/2 26%/62%/12%), 11 were responders (16% ORR, 95% Confidence Interval [CI] 7%, 25%). DCR was 54% (CI 43%, 66%), median OS was 11.5 months (CI 10.1-12.4 mos), and median PFS was 2.9 mos (CI 2.0-4.1 mos). Of the 11 responders, median time to response and DoR were 3.6 mos (CI 2.8-4.2 mos) and 18.6 months (95% CI 4.4-NR mos), respectively. Of the 14 PDL1-positive patients, 2 were responders (14% ORR, 95% CI 0%, 33%), DCR was 57% (CI 31%, 83%), median OS and PFS were 10.7 mos (CI 9.2-10.7 mos) and 2.3 mos (CI 1.4-4.9 mos), respectively. Durvalumab showed a manageable safety and tolerability profile; most adverse events (AEs) were low grade and resolved with treatment delay and/or immunosuppressive interventions. Grade ≥3 treatment-related AEs occurred in 23(34%) of patients. Drug-related AEs led to discontinuation in 6 patients (9%, 95% CI 2%, 16%).

Conclusions:
Durvalumab demonstrated clinical benefit and durable responses in a previously treated metastatic NSCLC population with manageable toxicity profile. Results are comparable with other anti-PD-1/PD-L1 therapies in metastatic, relapsed NSCLC.

Clinical trial identification:
NCT02766335

Legal entity responsible for the study:
Southwest Oncology Group/NCTN

Funding:
Lung-MAP supported in part by NIH/NCI grants CA180888, CA180819, CA180820, CA180821, CA180868, and by Amgen, AstraZeneca, Bristol-Myers Squibb Company Genentech and Pfizer through the Foundation for the National Institutes of Health, in partnership with Friends of Cancer Research.

Disclosure:
V. Papadimitrakopoulou: Advisory Role for: Eli Lilly&Co, Genentech, Janssen Global Sevices, Bristol-Myers Squibb, ARIAD, Astra Zeneca Pharmaceuticals, Novartis, Merck Corporate-sponsored research: Novartis, Astra Zeneca, Genentech, Merck, Janssen, ACEA, Bristol-Myers Squibb. H. Borghaei: Advisory: Bristol-Myers Squibb, Lilly, Genentech, Celgene, Pfizer, EMD-Serono, Boerhinger-Ingelheim, Trovagene, AstraZeneca, Research Support: Millenium, Merck, Celgene. K. Kelly: Advisory: Synta, AstraZeneca, Lilly, Clovis Oncology, ARIAD, Boehringer Ingelheim, Bristol-Myers Squibb, Genentech, G1 Therapeutics. Research: Millenium, Novartis, EMD Serono, Lilly, Genentech, Abbvie, Gilead, Celgene, Five Prime Therapeutics, Transgene. All other authors have declared no conflicts of interest.

Background:
Atezolizumab (atezo) inhibits binding of PD-L1 to its receptors, PD-1 and B7.1, restoring tumor-specific T-cell immunity and leaving the PD-L2/PD-1 interaction intact. This single-arm Phase II study (BIRCH; NCT02031458) was designed to evaluate atezo monotherapy in PD-L1–selected patients with advanced NSCLC. A previous analysis (median follow-up, 8.5 months) demonstrated clinical activity in chemotherapy-naive 1L and 2L+ PD-L1–selected patients. Here we present updated efficacy data for 1L patients.

Methods:
Eligible patients had PD-L1–selected advanced-stage NSCLC, with no prior chemotherapy or CNS metastases. PD-L1 was centrally evaluated using the VENTANA SP142 IHC assay. Enrolled patients expressed PD-L1 on ≥ 5% of tumor cells (TC) or tumor-infiltrating immune cells (IC), ie, TC2/3 or IC2/3. Those with EGFR mutation or ALK rearrangement must have had prior treatment with an appropriate TKI. Atezo was administered (1200 mg IV q3w) until radiographic disease progression or unacceptable toxicity. The primary endpoint was independent review facility (IRF)–assessed ORR; secondary endpoints included investigator (INV)-assessed ORR, DOR, PFS (RECIST v1.1) and OS.

Results:
With a median duration of survival follow-up of 22.5 months, INV-assessed ORR was 25% in TC2/3 or IC2/3 (all treated) patients and 34% in TC3 or IC3 patients (Table). Median OS was 23.5 months in all treated patients and 26.9 months in the TC3 or IC3 subgroup. Responses were observed in both EGFR and KRAS mutant and wild-type tumors. The safety profile was consistent with previous atezo NSCLC studies.

Conclusions:
With a median follow-up of 22.5 months, atezo continued to demonstrate durable clinical benefit in 1L NSCLC, in both EGFR and KRAS mutant and wild-type tumors. These results support ongoing Phase III trials evaluating atezo vs chemotherapy in 1L NSCLC.rnTable: 84Orn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

Endpoint (95% CI)	TC3 or IC3[a] (n = 65)	TC2/3 or IC2/3[b] (n = 138)
INV ORR, %	34%	25%
	(22.6-46.7)	(18.4-33.5)
EGFR mutant/wild type, ORR, %	25%/31%	31%/22%
KRAS mutant/wild type, ORR, %	38%/30%	31%/22%
Median DOR, mo	NE	16.5
	(8.5-NE)	(9.9-NE)
Median OS, mo	26.9	23.5
	(12.0-NE)	(18.1-NE)
12-mo OS rate, %	61.5%	66.4%
	(49.0-74.0)	(58.1-74.6)
Median PFS, mo	7.3	7.3
	(4.9-12.0)	(5.7-9.7)
12-mo PFS rate, %	36.5%	32.5%
	(24.0-48.9)	(24.2-40.8)

rnNE, not estimable.rnaTC ≥ 50% or IC ≥ 10% PD-L1–expressing cells.rnbTC or IC ≥ 5% PD-L1–expressing cells.rn

Clinical trial identification:
NCT02031458

Legal entity responsible for the study:
F. Hoffmann-La Roche Ltd/Genentech Inc., a member of the Roche Group

Funding:
F. Hoffmann-La Roche Ltd/Genentech Inc, a member of the Roche Group

Disclosure:
M.C. Garassino: Honoraria, Consulting, Speaker\'s Bureau, research funding, expert testimony, travel expenses: MSD, BMS, AZ, Lilly, Roche. D. Christoph: Honoraria, Speaker\'s Bureau: BI, BMS, Chugai, Novartis, Merck, MSD, Pfizer, Roche; Consulting: BI, BMS, Novartis, Pfizer, Roche; Expert testimony: BI, BMS, Novartis, Pfizer, Roche. J. Chaft: Advisor for Genentech and Astra Zeneca. M.L. Johnson: Consulting: Genentech, Celgene, BI; Research funding: OncoMed, BerGenBio, Lilly, EMD Serono, Kadmon, Janssen, Mirati, Genmab, Pfizer, AZ, Roche/Genentech, Stemcentrix, Novartis, Checkpoint, Array, Regeneron. S. Mocci: Employee, stock: Roche/Genentech. S.N. Gettinger: Consulting: BMS; Research funding: Roche/Genentech, BMS, ARIAD, Incyte, Celldex. E. Felip: Advisory Boards: Lilly, Pfizer, BI, MSD, Roche; Speaker\'s Bureau: AZ, BMS, Novartis. All other authors have declared no conflicts of interest.

Background:
The success of PD-1/PD-L1 immune checkpoint inhibitors strengthens the notion that tumor growth and regression are immune regulated. According to PD-L1 status and number of Tumor Infiltrating Lymphocytes (TILs), the cancer microenvironment can be classified into four types reflecting immune resistance, ignorance, induction and tolerance. However, an extended tissue characterization of the immune contexture and its predictive/prognostic value in NSCLC remain elusive. The aim of the present investigation was to determine whether immunologically defined classes of NSCLC differentially impact on clinical outcome.

Methods:
Histologic sections of NSCLC samples surgically removed to untreated 51 ADC and 69 SCC and 8 ADC and 10 SCC patients receiving Nivolumab were included. PD-L1 (clones 28-8 and SP142) was measured by immunoperoxidase (H-score) and immunofluorescence (QIF). The n/mm[2] and intra-, peri-tumor or invasive margin localization of TILs subpopulations were computed establishing cut off values relative to each phenotype. Immunohistochemical data and clinical records were subjected to Kaplan Meier analysis.

Results:
ADC cases had 2-fold higher CD3[pos] and 1.8-fold lower CD4[pos] cells compared to SCC and TILs-rich ADC had 10 months’ increase in OS vs –poor (p < 0.01). EGFR mutation conditioned a lower intratumor density of TILs. The frequency of type I (PD-L1[high] TILs[high]) contexture was low (14.6%) while > 1/3 of NSCLC samples displayed type II (PD-L1[low/neg] TILs[low]), reflecting immune exhaustion. The proportion of type III (PD-L1[high] TILs[low]) and IV (PD-L1[low/neg] TILs[high]) immune categories, with relatively increased Nks and Tregs, was similar. NSCLC type III had the highest OS (35.5 mos.) and PFS (25.7) while in type II immune ignorant cases OS and PFS were respectively 21.7 and 12. Independently from immune categories, patients with PD-1[low] and high CD8/CD3 ratio had 11 mos. gain in OS (p < 0.01) compared to the reverse counterpart. Accordingly, PD-L1[high] and PD-1[low] characterized 86% of patients responsive to nivolumab.

Conclusions:
In a dynamic PD-L1 milieu a concomitant intrinsic or therapeutically induced decay of PD-1 receptor allows TILs to escape from PD-L1 pressure and delays tumor progression, improving OS.

Clinical trial identification:


Legal entity responsible for the study:
University Hospital of Parma, Medicine

Funding:
University of Parma

Disclosure:
All authors have declared no conflicts of interest.

Abstract not provided

Background:
We assessed self-reported symptoms of advanced non-small cell lung cancer patients treated with osimertinib 80mg or chemotherapy in the AURA3 phase III clinical trial (NCT02151981).

Methods:
Patients completed the European Organisation for Research and Treatment of Cancer QLQ-LC13 questionnaire on disease-specific symptoms and QLQ-C30 on general cancer symptoms, functioning and global health status. QLQ-LC13 was completed at baseline, weekly for 6 weeks, then 3-weekly up to end of study, and at progression. QLQ-C30 was completed at baseline, then 6-weekly up to end of study, and at progression. We compared for differences between treatments in time to deterioration and odds of improvement of symptoms (two assessments ≥18 days apart). A deterioration or improvement was defined as a change in score from baseline of ≥ +/-10. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated using a log-rank test stratified by ethnicity. Odds ratios (OR) and 95% CIs were calculated using logistic regression adjusted for ethnicity.

Results:
At baseline, 215 − 228 of 279 (77 − 82%) patients on osimertinib and 106 − 114 of 140 (76 − 81%) on chemotherapy had QLQ-LC13 scores ≤90 (cut-off to have potential for deterioration) for cough, chest pain and dyspnoea. Time to deterioration of key symptoms was longer with osimertinib than with chemotherapy (Table). The proportion of patients with improvement in global health status was higher with osimertinib (80/215 [37%]) than with chemotherapy (23/105 [22%]; OR: 2.11; 95% CI: 1.24, 3.67; p = 0.007), as it was for appetite loss (OR: 2.50; 95% CI: 1.31, 4.84) and fatigue (OR: 1.96; 95% CI: 1.20, 3.22).rnTable: 85O_PRTime to deterioration of selected key symptomsrn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn rnrn

Symptom	Treatment	Number (%) of patients with event	HR (95% CI)	p-value
Cough	O	99 (46.0)	0.74 (0.53, 1.05)	0.090
	C	58 (54.7)
Chest pain	O	99 (43.8)	0.52 (0.37, 0.73)	<0.001
	C	66 (58.4)
Dyspnoea	O	122 (53.5)	0.42 (0.31, 0.58)	<0.001
	C	84 (73.7)

rnC, chemotherapy; O, osimertinib.rn

Conclusions:
Time to deterioration of key symptoms was longer and more patients had an improvement in global health status with osimertinib treatment than with chemotherapy, demonstrating improved patient outcomes with osimertinib.

Clinical trial identification:
NCT02151981

Legal entity responsible for the study:
AstraZeneca

Funding:
AstraZeneca

Disclosure:
C.K. Lee: Served on advisory boards for AstraZeneca. S. Novello: Served on speaker bureaux for AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Merck Sharp & Dohme Limited and Roche. A. Rydén, H. Mann, S. Ghiorghiu: Employees of AstraZeneca and are AstraZeneca shareholders. A. Templeton, K. Rüdell: Former employees of AstraZeneca and former shareholders. T. Mok: Grant/Research Support from AstraZeneca, BI, Pfizer, Novartis, SFJ, Roche, MSD, Clovis Oncology, BMS, Eisai, Taiho; Speaker’s fees with: AstraZeneca, Roche/Genentech, Pfizer, Eli Lilly, BI, MSD, Novartis, BMS; Major Stock Shareholder in: Sanomics Ltd.; Advisory Board for: AstraZeneca, Roche/Genentech, Pfizer, Eli Lilly, BI, Clovis Oncology, Merck Serono, MSD, Novartis, SFJ Pharmaceutical, ACEA Biosciences, Inc., Vertex Pharmaceuticals, BMS, geneDecode Co., Ltd., OncoGenex Technologies Inc., Celgene, Ignyta, Inc.; Board of Directors: IASLC, Chinese Lung Cancer Research Foundation Ltd., Chinese Society of Clinical Oncology (CSCO), Hong Kong Cancer Therapy Society (HKCTS).

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Background:
TAILOR (NCT00637910) was a phase 3 randomized trial comparing erlotinib over docetaxel in second line treatment in EGFR wild-type patients. The prognostic and predictive value of KRAS is still debatable. Taking advantage from the available tissue samples and clinical data of this trial, we performed sequencing of a customized gene panel in order to identify novel biomarkers and new potential therapeutic targets specifically associated to KRAS.

Methods:
NGS was performed on PGM Ion Torrent platform and involved 111 genes. 5% of frequency was used to define mutations. Association with clinical features or between genes was performed with non-parametric test. Cox regression model was used to define the prognostic role of mutated genes on survival.

Results:
188 out of 222 randomized patients had available tissue. 134 tissues were successfully sequenced. 3 were EGFR mutated, 47 were KRAS mutated (36%). 68 were in erlotinib arm and 63 in docetaxel arm. Biomarkers associated to KRAS were 21 LKB1 (16%) and 61 p53 (46%). Survival results are described in the table.rnTable: LBA3rn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

rn	PFS	PFS	OS	OS
	Univariate	Multivariate	Univariate	Multivariate
	HR (95%CI) p-value	HR (95%CI) p-value	HR (95%CI) p-value	HR (95%CI) p-value
KRASwt/LKB1wt	reference	rn	rn	rn
KRASmut/LKB1wt	0.76 (0.51-1.15) p 0.19	0.79 (0.51-1.22) p 0.29	0.78 (0.51-1.19) p 0.25	0.75 (0.49-1.17) p 0.21
KRASwt/LKB1mut	1.42 (0.68-2.95) p 0.35	1.43 (0.67-3.08) p 0.35	1.69 (0.81-3.52) p 0.16	1.59 (0.75-3.4) p 0.23
KRASmut/LKB1mut	1.18 (0.64-2.19) p 0.59	1.29 (0.67-2.49) p 0.45	1.32 (0.72-2.45) p 0.37	1.34 (0.69-2.61) p 0.38
KRASwt/tp53wt	reference	rn	rn	rn
KRASmut/tp53wt	0.80 (0.49-1.31) p 0.37	0.87 (0.51-1.46) p 0.86	0.88 (0.53-1.46) p 0.61	0.85 (0.50-1.46) p 0.56
KRASwt/tp53mut	1.23 (0.79-1.91) p 0.35	1.41 (0.88-2.26) p 0.15	1.38 (0.88-2.16) p 0.15	1.55 (0.95-2.55) p 0.81
KRASmut/tp53mut	1.07 (0.62-1.82) p 0.81	1.11 (0.64-1.94) p 0.70	1.14 (0.66-1.98) p 0.64	1.12 (0.63-1.97) p 0.70

rn

Conclusions:
In TAILOR, KRAS was not prognostic. The association with LKB1 and p53 does not affect prognosis, while LKB1 mutation could be a negative prognostic factor (HR = 1.39). Further research is needed to prospectively assess this pathway.

Clinical trial identification:


Legal entity responsible for the study:
Fondazione IRCCS Istituto Nazionale Tumori, Milan

Funding:
AIRC (Associazione Italiana Ricerca Cancro)

Disclosure:
All authors have declared no conflicts of interest.

Background:
EGFR inhibitors are approved for use in patients with EGFR mutant NSCLC. The cobas® platform is a companion diagnostic test used in clinical practice on tumour tissue, and is approved for ctDNA EGFR testing from plasma samples. We evaluated the diagnostic performance of this platform on ctDNA in a prospective single centre study.

Methods:
Patients diagnosed with advanced NSCLC at the Royal Marsden Hospital between November 2015 - November 2016 were included. Patients had to have tissue EGFR result available. Peripheral blood samples were collected in cfD tubes (Roche Molecular Solutions CA, USA) and analysed using the cobas® platform following manufacturer’s instructions.

Results:
201 samples were analysed; median age 68 years. 89% of patients had adenocarcinoma, followed by poorly differentiated carcinoma (4%), squamous cell carcinoma (3%), adenosquamous (2.5%) and other (1.5%). 30% of patients were never smokers, whereas 58% were ex-smokers and 10% were current smokers (2% unknown smoking history). 93% of patients had stage 4 disease, whereas 7% had stage 3 disease. 42% of patients had thoracic-only disease. 79% of patients had their tissue EGFR analysis performed on the cobas[®] platform (others included Illumina NGS (7.5%); Therascreen[TM] (1%)). 10% of patients did not have available tumour tissue. All plasma samples were tested on the cobas[®] platform. Concordance rate between ctDNA and tumor was 87%, with sensitivity 65%, specificity 98%, positive predictive value 95%, and negative predictive value 85%. Concordance rate in patients with an EGFR mutation was 55% for thoracic-only disease and 75% for extra-thoracic disease. The mean time to EGFR result was shorter for ctDNA than tissue (8.4 vs 11.2 days; P=0.07) with a ctDNA failure rate of only 2.4%. Two tissue failures resulted in an EGFR mutation being picked up in plasma (Exon 19 and T790M).

Conclusions:
We demonstrate that ctDNA can be used for diagnosis of an EGFR mutation in advanced NSCLC with excellent specificity, despite moderate sensitivity. The cobas[®] test for ctDNA should be incorporated into clinical practice to triage patient care.

Clinical trial identification:
Not applicable

Legal entity responsible for the study:
The Institute of Cancer Research and the Royal Marsden NHS Foundation Trust

Funding:
Roche Molecular Solutions CA, USA

Disclosure:
T.A. Yap: Received research funding from Roche. S. Popat: Consultant to Roche and has received honoraria from Roche. J. Palma: Employee of Roche Molecular Solutions CA, USA. All other authors have declared no conflicts of interest.

Abstract not provided

Background:
PD-1/PD-L1 axis blockade is of pivotal interest in NSCLC, where the relative antibodies have shown relevant activity. Detection of markers able to predict the activity of these treatments is an importance issue. The predictive role of PD-L1 evaluated by IHC is debated in NSCLC, also, considering that small biopsies or cytological specimens represent often the only tumor material available. This study aimed to assess a circulating immuno-profile as predictor of outcome in NSCLC patients treated with nivolumab.

Methods:
A peripheral blood immuno-profile evaluation was performed at three different times: At baseline (T0), after 2 cycles (T1) and after 4 cycles (T2) of Nivolumab in 54 advanced pre-treated NSCLC patients treated in an Expanded Access Program in two Italian institutions. Tumor assessment was performed after 4 cycles and every 2 months. FACS analysis of lymphocyte subpopulations [CD3, CD4, CD8, NK (CD56) and Treg (FOXP3)] was performed. In addition to changes in absolute number and % of these populations, they were also characterized for their functional and proliferative activity. Quali-quantitative leucocyte composition at baseline and its change during therapy were correlated with tumor response and overall survival.

Results:
Baseline Neutrophil-to-Lymphocyte Ratio, baseline NK count, lymphocyte count and CD4 variations during therapy showed a statistically significant prognostic role (p < 0.001; p = 0.012; p < 0.001; p = 0.010, respectively). Preliminary results on 31 patients with all 3 time-points samples, showed: A significant increase of NK cells and a significant decrease of CD4% from T0 to T2; a significant increase from T0 to T2 of NK subpopulation CD56dim (with reduction of CD56bright); considering two groups of patients (19 responders vs. 12 non-responders), a NK increase (overall and of CD56dim) and CD4% decrease was observed in responders and non-responders, respectively; absolute number and % of NK at baseline was statistically different in the two groups; absolute number and % of CD8+/PD1+ at baseline was significantly higher in responders vs. non-responders.

Conclusions:
These data show that it is possible to identify predictive peripheral immuno-biomarkers, such as NK and CD8, for nivolumab therapy in advanced NSCLC.

Clinical trial identification:
NA

Legal entity responsible for the study:
University Hospital of Parma

Funding:
AIRC

Disclosure:
All authors have declared no conflicts of interest.

Background:
Response patterns following cancer immunotherapy (CIT) treatment are often non-classical due to tumor immune infiltration and increase in tumor burden prior to response. As such, RECIST-based endpoints may not adequately predict the potential OS benefit with CIT. In an updated analysis of POPLAR, PFS and ORR improvement with atezolizumab (atezo) vs docetaxel (doc) was restricted to patients (pts) with high PD-L1 expression, whereas OS benefit was seen in the PD-L1–unselected ITT population (HR, 0.69; mOS 12.6 mo vs 9.7 mo with atezo vs doc), suggesting that many pts derive atezo benefit after radiographic progression.

Methods:
287 NSCLC (2L/3L) pts were randomized 1:1 to receive atezo (1200 mg IV q3w) until loss of clinical benefit or doc (75 mg/m[2] IV q3w) until PD per RECIST v1.1 (RECIST). The primary endpoint was OS; secondary endpoints included PFS and ORR per RECIST. Atezo arm pts were evaluated per immune-modified RECIST (imRECIST) and post-PD radiographic changes. Both arms were evaluated for OS post-PD (data cutoff, Dec 1, 2015; minimum follow-up, 20 mo).

Results:
Among 144 pts in the atezo arm, ORR per imRECIST vs RECIST was 17% vs 15%, disease control rate (CR+PR+SD) was 65% (imRECIST) vs 52% (RECIST) and mPFS was 4.3 mo (imRECIST) vs 2.7 mo (RECIST). In 61 atezo pts continuing treatment post-PD, 8% had a subsequent ≥ 30% decrease in target lesions relative to the time of PD, and 74% had a best change between +20% and −30%. mOS post-PD in both arms is shown (Table).

Conclusions:
Increased ORR and PFS per imRECIST vs RECIST in atezo pts highlight the utility of imRECIST as response evaluation criteria to assess efficacy of atezo/CIT. 82% of pts continuing atezo post-PD had subsequent stable ( ≤ +20% change) or decreased target lesions. The nearly 1-yr mOS in pts continuing atezo vs other treatment post-PD is suggestive of atezo benefit lasting beyond PD. Table.rnTable: 96PD_PRrn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

Patients Receiving Therapy Post-PD per RECIST v1.1	n	mOS from time of first RECIST v1.1 PD
Atezo arm pts continuing atezo	61	11.8 mo
Atezo arm pts receiving non-protocol anti-cancer therapy	30	9.2 mo
Doc arm pts receiving non-protocol anti-cancer therapy	46	9.7 mo

rn

Clinical trial identification:
NCT01903993

Legal entity responsible for the study:
F. Hoffmann-La Roche Ltd/Genentech Inc., a member of the Roche Group

Funding:
F. Hoffmann-La Roche Ltd/Genentech Inc., a member of the Roche Group

Disclosure:
A. Artal-Cortes: Advisory boards for Roche, BMS, MSD and travel fees from Roche. A. Rittmeyer: Grants as an advisor or speaker from Astra Zeneca, BMS, Boehringer Ingelheim, Eli Lilly, Pfizer and Roche Genentech. K. Park: Consulting/advisory from Astellas Pharma, AstraZeneca, BI, Clovis Oncology, Lilly, Hanmi, Kyowa Hakko Kirin, Novartis, Ono Pharmaceutical, Roche; speakers\' bureau fees from BI; research funding from AstraZeneca. C. Lewanski: Consulting/Advisory: Roche, Celgene, Amgen. S. Liu: Employed by Genentech and holds Roche stocks. M. Ballinger: A Genentech employee and has Roche stock. All other authors have declared no conflicts of interest.

Abstract not provided

Background:
Lung cancers mostly present at a clinically advanced stage. Predicting the invasiveness of tumors might help in planning treatment and prophylaxis. We aim to investigate the patterns of lung cancer metastasis and determine possible clinical predictors of those patterns in population based study.

Methods:
The records of all patients (2010-2013) were extracted from SEER database of the National Cancer Institute. Data was extracted to SPSS. Univariate and multivariate logistic regression were used for analysis.

Results:
We have identified 76254 (54.5%M, 45.5F) patients with metastatic lung cancer (2010-2013).17% were small cell tumors (SCLC), while 83% of tumors were non-small cell tumors (NSCLC). Tumor was found more frequently on the right (54%) than the left (38%), and was bilateral in 8% of cases. Our analysis showed that NSCLC had higher rates of metastasis to bone than SCLC (37% Vs 34%) (P < 0.001, 95% CI: 1.097-1.190), SCLC had higher rates of metastasis to liver than NSCLC (46% Vs 20%) (p < 0.001 95%CI: 0.27-0.29), both SCLC and NCLC had almost similar rates of brain metastasis at the time of diagnosis (25% Vs 26%, respectively) (P = 0.003, 95% CI: 1.024-1.120). Following adenocarcinoma and small cell, squamous cell histology was 3rd most common tumor histology for both brain and liver metastasis. Although rare, carcinoid tumor had high rates of metastasis to brain at diagnosis (44.8%). Predictors for liver metastasis were small cell and adenocarcinoma histology (p < 0.001, 95% CI:0 .245-0.304), Tumors with upper lobe location (P = 0.028, 95% CI: 0.839-0.990), and high grade tumors (P < 0.001, 95% CI: 1.334-1.618). Predictors for metastasis to brain were advanced age at diagnosis (P < 0.001, 95% CI: 0.970 -0.976), Adenocarcinoma and small cell histology (P < 0.001, 95% CI: 1.254-1.561), Lower lobe, and main bronchus locations (p = 0.004, 95% CI: 0.689-0.930), and tumors of higher grade (P < 0.001, 95%CI: 1.250-1.463).

Conclusions:
Lung cancer subtypes are associated with distinct patterns of metastatic spread. SCLC has significantly higher rates of liver metastases, while NSCLC has higher rates of bone metastasis. Further investigation needed to evaluate the role of primary tumor location on metastatic behavior of lung cancer.

Clinical trial identification:


Legal entity responsible for the study:
Mohamed Hendawi

Funding:
N/A

Disclosure:
The author has declared no conflicts of interest.

Background:
The clinical and radiologic presentation of lung cancer in women differs from men, especially in Asia. The aim of this study is to identify potential gender-related differences of newly developed lung cancer and provide the optimal CT screening intervals for the patients who are managed in the health care center.

Methods:
Between January 2000 and February 2016, a total of 46,766 consecutive patients who underwent screening chest CT in Health Care Center at Asan Medical Center of Korea were retrospectively reviewed. During the study period, 282 patients (M:F=205:77) were finally diagnosed by lung cancer. Among them, lung cancers were detected by initial screening CT in 186 patients (excluded in this study), and 96 patients (M:F=85:11, age range M:F=62.7±12.9: 62.7±8.7 years, respectively) have been clinically diagnosed as newly developed lung cancers using subsequent screening CT follow-up. Using the 96 patients, CT screening intervals, stage, and pathology (with subtype and mutation) of cancers were evaluated and analyzed to find any gender-related differences.

Results:
In the 96 patients, mean age is not significantly different between genders. Most of the men were smokers (74/85, 87%). Adenocarcinoma is the most common type (36/85, 42%), followed by squamous cell carcinoma (35%), small cell lung cancer (18%), and others (5%) in men. All female patients were adenocarcinoma. The mean time interval of screening CT was relatively longer in women (5.6±3.1 years) than in men (3.6±2.4 years) (p=0.02). However, lung cancer stages are significantly higher in men. Minimum CT screening interval to detect stage I lung cancer was longer in women (5 years) compared to that in men (1 year).

Conclusions:
The interval of screening CT could be longer in women than men. To detect most of stage I lung cancer, a 3-year interval screening for women and 1-year interval for men are optimal in our study for minimizing radiation hazard and worry.

Clinical trial identification:


Legal entity responsible for the study:
None

Funding:
N/A

Disclosure:
All authors have declared no conflicts of interest.

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Background:
Rearrangements of the ROS1 gene are found in 1-2% of non-small cell lung cancers (NSCLC) and are regarded as mutually exclusive oncogenic driver mutations. Since targeted therapy has recently been approved for ROS1-positive NSCLC, ROS1 testing is entering diagnostic routine. Fluorescence in situ hybridization (FISH) optionally selected for by immunohistochemistry (IHC) on histological material is regarded as gold standard for detection of ROS1 rearrangements. However, NSCLC is often diagnosed by cytology alone, requiring the option of predictive marker testing on cytological specimens. In this study, we explored the utility of ROS1 immunocytochemistry (ICC) on cytological specimens as a preliminary screening tool for detection of ROS1 rearrangements.

Methods:
ICC using the D4D6 antibody was prospectively performed in the routine diagnostic setting on ethanol-fixed and previously Papanicolaou-stained specimens from 296 patients with NSCLC, including adenocarcinomas (n = 243), NSCLC NOS (n = 47), and others (n = 6) and encompassing specimens from the lung (n = 107), locoregional lymph nodes (n = 87) and distant metastases (n = 102). Cytospin specimens of the cell line HCC-78, known to express ROS1, were used as positive control. Cytoplasmic staining of any intensity was considered positive.

Results:
ICC was positive in 12 cases with ROS1 rearrangements confirmed by FISH. Confirmation of 284 ICC-negative cases was available in 210 cases (FISH in 71 cases, mRNA based fusion NGS in 2 cases and detection of other known, mutually exclusive driver mutations in 137 cases). Only one ICC-negative case showed a ROS1 rearrangement by FISH (sensitivity 92%, specificity 100%, PPV 100%, NPV 99%).

Conclusions:
Our data show that ROS1 ICC is a highly accurate method for detection of ROS1 rearrangements in NSCLC. Given the high costs and technical challenges of FISH and the rarity of ROS1 rearrangements, ICC is cost effective and rapid and therefore well suited as a preliminary screening method. Cases with equivocal or positive findings on ICC can be confirmed by FISH or molecular tests (e.g. NGS).

Clinical trial identification:


Legal entity responsible for the study:
University Hospital Basel

Funding:
University Hospital Basel

Disclosure:
All authors have declared no conflicts of interest.

Background:
Constitutive EGFR signalling due to activating mutations (e.g. EGFRm+ NSCLC) or overexpression (e.g. squamous (sq) NSCLC) is a common mechanism of tumourigenesis. There is biological rationale for combining A + C in tumours with unmet medical need, e.g. EGFRm+ NSCLC with resistance to EGFR-TKIs, and in advanced sqNSCLC. Preclinically, A + C improved anti-tumour activity in EGFRm+ NSCLC vs A or C alone. Here we report outcomes from two Phase Ib studies of A + C: NCT01090011 (study 1) in pts with EGFRm+ NSCLC resistant to gefitinib/erlotinib (G/E); and NCT02020577 (study 2) in unselected pts with heavily pretreated advanced solid tumours, including sqNSCLC.

Methods:
Study 1 comprised a dose-finding phase to identify the maximum tolerated dose (MTD) of A + C, and an expansion phase (EP) in which 126 pts received the MTD of A 40 mg/day + C 500 mg/m[2] Q2W. In study 2, 9 pts received A + C in a dose-escalation phase, and a further 49 pts received the MTD of A 40 mg/day + C 250 mg/m[2]/week in an EP, including 12 pts with sqNSCLC.

Results:
Drug-related adverse events (DR-AEs; all grades) were reported in 99% of pts in the EP of study 1 (n = 126), and 98% of pts in study 2 (n = 58). Grade 3/4/5 DR-AEs were experienced by 44%/2%/2% of pts in study 1 and 31%/3%/0% in study 2. DR-AEs led to discontinuation of treatment in 14% of pts in study 1 and 12% in study 2. Efficacy in pts with NSCLC is shown in the table. In the study 1 EP, ORR was 29% and median PFS was 4.7 months; there were no significant differences in ORR or median PFS in pts with T790M+ vs T790M- tumours (32 vs 25% [p = 0.341] and 4.6 vs 4.8 months [p = 0.643], respectively). Median duration of OR was 5.7 months. In the study 2 EP, 75% of pts with sqNSCLC had SD, and median PFS was 11.9 weeks.

Conclusions:
Based on the demonstrated antitumour activity and a predictable/manageable safety profile, A + C may be a treatment option in pts with EGFRm+ NSCLC and acquired resistance to G/E (irrespective of T790M status) or with pre-treated advanced sqNSCLC. rnTable: 86OEfficacy outcomesrn

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn rnrn

Trial	EGFRm+ NSCLC	sqNSCLC
	Study 1 expansion phase[1] (n = 126)	Study 2 expansion cohort[2] (n = 12)
Disease control (CR/PR/SD), n (%)	89 (71)	9 (75)
OR (CR/PR), n (%)	37 (29)	0
SD	52 (41)	9 (75)
Unconfirmed OR	5 (4)	–
PD, n (%)	27 (21)	3 (25)
Median duration of OR, months (range)	5.7 (1.8–24.4)	–
Median duration of disease control, weeks (standard deviation)	–	17.9 (9.9)
Median PFS (95% CI)	4.7 months (4.3–6.4)	11.9 weeks (5.1–19.4)

rnCR, complete response; PR, partial response; SD, stable disease; PD, progressive disease;rnOR, objective response; PFS, progression-free survival 1. Janjigian YY, et al. Cancer Discov 2014;4:1036-45 2. Gazzah A, et al. Eur J Cancer 2016;68(Suppl. 1);S139:abstract 426rn

Clinical trial identification:
NIH study numbers: 1200.71: NCT01090011 1200.122: NCT02020577

Legal entity responsible for the study:
Boehringer Ingelheim

Funding:
Boehringer Ingelheim

Disclosure:
J-C. Soria: Astrazeneca, Astex, GSK, Gammamabs, Eli Lilly, MSD, Merus, Pfizer, Pharmamar, Pierre Fabre, Roche, Sanofi, Servier, Symphogen, Takeda. Y.Y. Janjigian: Employee of Memorial Sloan Kettering Cancer Center Advisory boards: ASCO GI Corporate-sponsored research: Merck, BMS Consultant: Pfizer Grant/research: BI, Bayer, Amgen, Genentech, Roche. H.J.M. Groen: Participated on advisory boards for Boehringer Ingelheim. W. Pao: Reports that patent rights on EGFR T790M were licensed by MSKCC to Molecular MD. J. Fan: Boehringer Ingelheim employee. M. Ould-Kaci: Employee of Boehringer Ingelheim. L. Horn: Consulting for AbbVie, BMS, Merck, Eli Lilly, Xcovery, Bayer. All other authors have declared no conflicts of interest.

Abstract not provided

Background:
The MET ligand hepatocyte growth factor (HGF) in stromal cells provides an alternative signaling mechanism for EGFR by inducing inter-receptor cross talk with EphA2, CUB domain-containing protein-1 (CDCP1) or AXL. Src-homology 2 domain-containing phosphatase 2 (SHP2) also plays a role in RTK signaling and is required for MAPK activation.

Methods:
We studied the expression levels of stromal HGF and tumor RTKs: AXL, CDCP1, MET, and EphA2, as well as SHP2, and clinical outcome in baseline samples of 64 EGFR-mutant NSCLC patients (pts) treated with first-line EGFR TKI. Paraffin-embedded samples and the discrimination between tumor and stromal cells were performed by a pathologist. mRNA expression was analyzed by qRT-PCR.

Results:
AXL and CDCP1 were among the genes that most significantly influenced treatment outcome. Median progression-free survival (PFS) was 14 months (mo) and 23.4 mo for pts with high and low AXL mRNA, respectively (p = 0.03), (HR: 2.05, p = 0.03). Median PFS was 9 mo and 20.2 mo for pts with high and low CDCP1 mRNA, respectively (p = 0.01), (HR: 2.1, p = 0.02). There were no significant differences in PFS according to levels of EPHA2 and MET in the tumor, and HGF in the stroma. Median PFS was 11.4 mo and 24 mo for pts with high and low SHP2 mRNA, respectively (p = 0.09) (HR: 2.4, p = 0.01). Median overall survival (OS) was 19.1 mo and 40.7 mo for pts with high and low AXL mRNA, respectively (p = 0.009) (HR: 2.9, p = 0.001). Median OS was 19.1 mo and 34.9 mo for pts with high and low CDCP1 mRNA, respectively (p = 0.03) (HR: 2.2, p = 0.03). In addition, differences in OS were noted according to levels of EPHA2, MET and SHP2.

Conclusions:
AXL and CDCP1 are adverse predictive markers of PFS and OS. It has been reported that overexpression of several RTKs can substitute EGFR signaling in EGFR mutant NSCLC pts. The findings show the need to scrutinize RTK levels in EGFR mutant NSCLC patients and, henceforth, the importance of tailored combinatory therapy with AXL inhibitors, or with Src inhibitors, in those pts with elevated CDCP1 (CDCP1 triggers SFK activation).

Clinical trial identification:


Legal entity responsible for the study:
Institut d’Investigació en Ciències Germans Trias i Pujol, Badalona, Spain Institut Català d’Oncologia, Hospital Germans Trias i Pujol, Badalona, Spain

Funding:
Institut d’Investigació en Ciències Germans Trias i Pujol, Badalona, Spain Institut Català d’Oncologia, Hospital Germans Trias i Pujol, Badalona, Spain

Disclosure:
All authors have declared no conflicts of interest.

Background:
The investigational next-generation anaplastic lymphoma kinase (ALK) inhibitor brigatinib is being evaluated in patients (pts) with ALK-positive non–small cell lung cancer (ALK+ NSCLC) in a phase 1/2 trial (Ph1/2) and a pivotal phase 2 trial (ALTA); most of these pts had intracranial central nervous system (CNS) metastases at baseline.

Methods:
In Ph1/2 (NCT01449461), pts with advanced malignancies, including ALK+ NSCLC, received brigatinib (30–300 mg/d). In ALTA (NCT02094573), crizotinib-resistant pts with advanced ALK+ NSCLC received brigatinib at 90 mg qd in arm A or 180 mg qd with a 7-day lead-in at 90 mg in arm B. ALK+ NSCLC pts with baseline brain metastases were analyzed. CNS efficacy in both trials and safety in ALTA are shown.

Results:
In Ph1/2 and ALTA, 50/79 (63%) and 154/222 (69%) ALK+ NSCLC pts had baseline brain metastases based on independent review committee (IRC) and investigator assessment, respectively. Most pts had received chemotherapy (Table). In Ph1/2, 25/50 (50%) pts were receiving brigatinib as of 16 November 2015; in ALTA, 101/154 (66%) pts were receiving brigatinib as of 29 February 2016. In pts with measurable brain lesions, confirmed intracranial objective response rate was 53% in Ph1/2 and 42%/67% in ALTA A/B (Table). In pts with only nonmeasurable brain lesions, 35% (11/31) in Ph1/2 and 7% (4/54)/18% (10/55) in ALTA A/B had confirmed complete resolution of brain lesions. Further data are shown in the table. The most common treatment-emergent adverse events (TEAEs) in the 151 treated ALTA pts with baseline brain metastases were nausea (A/B, 32%/43%), headache (30%/30%), diarrhea (18%/36%), cough (21%/30%), and vomiting (25%/26%); the most common grade ≥3 TEAEs were increased blood creatine phosphokinase (1%/11%), hypertension (4%/7%), increased lipase (3%/3%), and pneumonia (1%/4%).

Conclusions:
Brigatinib yielded substantial clinical activity in ALK+ NSCLC pts with brain metastases in 2 trials.rnTable: 87OBaseline characteristics and IRC-assessed intracranial efficacy of brigatinib in ALK+ NSCLC Pts with brain metastases at baselinern

rnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrnrn

rn	Ph1/2 n = 50	ALTA Arm A n = 80	ALTA Arm B n = 74
Median age, years	53	49	55
Received prior chemotherapy, %	76	74	76
Crizotinib-naive, %	8	0	0
Pts evaluable for intracranial efficacy by IRC, n[a]	46	80	73
Median iPFS, months	14.6	15.6	12.8
Pts with measurable brain lesions, n	15	26	18
iORR, n (%)	8 (53)	11 (42)	12 (67)
iDCR, n (%)	13 (87)	22 (85)	15 (83)
No rad/active[b] subset, n	9	19	15
iORR, n (%)	6 (67)	8 (42)	11 (73)
iDCR, n (%)	8 (89)	16 (84)	14 (93)

rniDCR = intracranial disease control rate, iORR = intracranial objective response rate (confirmed), iPFS = intracranial progression-free survivalrnaLast scan date: 8 October 2015 in Ph1/2; 14 April 2016 in ALTArnbNo prior brain radiotherapy in Ph1/2; active (untreated or treated and progressed) brain lesions in ALTArn

Clinical trial identification:
NCT01449461 and NCT02094573

Legal entity responsible for the study:
ARIAD Pharmaceuticals, Inc.

Funding:
ARIAD Pharmaceuticals, Inc.

Disclosure:
M. Tiseo: Consulting or advisory role (AstraZeneca, BMS, Boehringer Ingelheim, Eli Lilly, Novartis, Otsuka, Pierre Fabre), research funding (ARIAD). R.M. Huber: Honoraria (ARIAD, AstraZeneca, BMS, Boehringer Ingelheim, Pfizer, Roche), consulting or advisory role (BMS, Boehringer Ingelheim, Celgene, Clovis Oncology, Eli Lilly, Novartis, Roche), research funding (Pierre Fabre). L.A. Bazhenova: Stock and other ownership interests (Epic Sciences), honoraria (Novartis), consulting or advisory role (AbbVie, AstraZeneca, BMS, Boehringer Ingelheim, Clovis Oncology, Genoptix, Heat Biologics, Pfizer, Roche/Genentech, Seattle Genetics, Trovagene), speakers bureau (AstraZeneca, Novartis, Pfizer, Roche/Genentech), research funding (AbbVie, ARIAD, Astellas, Astex, AstraZeneca/MedImmune, Boehringer Ingelheim, Chugai, Clovis Oncology, Eisai, Eli Lilly, Heat Biologics, Johnson & Johnson, Merck, Mirati, NanoCarrier, Novartis, Pfizer, Roche/Genentech). S-H.I. Ou: Advisory board (ARIAD). W. Reichmann, J. Haney, D. Kerstein: Employment, stock and other ownership interests (ARIAD). D.R. Camidge: Honoraria (ARIAD), research funding (ARIAD). S.N. Gettinger: Consulting or advisory role (ARIAD, BMS, Janssen), research funding (ARIAD, AstraZeneca/MedImmune, BMS, Boehringer Ingelheim, Incyte, Pfizer, Roche/Genentech). All other authors have declared no conflicts of interest.

Background:
Ensartinib (X-396) is a potent ALK small molecule TKI with additional activity against MET, ABL, Axl, EPHA2, LTK, ROS1 and SLK. In animal studies, CNS concentration of ensartinib in mice given at the therapeutic dose was 4 times higher than the IC50 for growth inhibition of ALK+ cells in vitro. Ensartinib was significantly more effective than crizotinib (C) at inhibiting the intracranial (IC) growth of the SH-SY5Y neuroblastoma model harboring the F1174L mutation. We subsequently evaluated the CNS activity of ensartinib in pts with ALK+ NSCLC.

Methods:
In this multicenter phase I/II study, pts with ALK+ NSCLC were enrolled and given ensartinib orally on a continuous 28-day schedule. 225 mg QD was selected for further evaluation. Pts enrolled with asymptomatic CNS metastases (with or without systemic disease) who were ALK TKI naïve or had received prior C or a 2nd generation ALK TKI were allowed to enroll. Overall and systemic response was assessed using RECIST 1.1. CNS response was assessed using modified RANO criteria. Pts with only CNS disease had to have at least 1 measurable target lesion ≥ 3 mm in diameter.

Results:
26 pts with ALK+ NSCLC and baseline CNS metastases were treated at ≥ 200 mg. Of the 26 pts, 13 pts had baseline target lesions, and 13 pts had only non-target lesions. CNS responses were observed in ALK TKI naïve pts and pts that received prior C or a 2[nd] generation ALK TKI. In the 13 pts with baseline target CNS lesions, IC response rate (RR) was 69%, including 1 CR, and 31% had SD, a 100% disease control rate. In the 13 pts with only non-target baseline lesions, 1 CR was achieved and 8 pts had SD. The IC RR in the 3 ALK TKI naïve pts with baseline target lesions was 100%, and 62% in 8 pts that received prior C only. Of 2 pts with baseline target lesions who received a prior 2[nd] generation ALK TKI, 1 had a PR and 1 SD. Median duration of IC response in the 10 pts who responded (9 with target lesions, 1 with non-target lesions only) is 5.8+ months, with the longest duration being 24 months.

Conclusions:
Our clinical findings support the preclinical results that the use of ensartinib may result in favorable therapeutic outcomes in pts with ALK+ NSCLC with CNS metastases. The ongoing phase III eXalt3 study will assess CNS RR and time to CNS progression in pts receiving 1st-line ensartinib vs C.

Clinical trial identification:
NCT02767804 and NCT01625234

Legal entity responsible for the study:
Xcovery Holding Company

Funding:
Xcovery Holding Company

Disclosure:
K. L. Reckamp: Xcovery Holding Company research funds to institution Ariad consultant and research funds to institution. H. A. Wakelee: Peregrin: Consultant/Independent contractor and honorarium recipient. Novartis: Grants/research support, consultant, honorarium recipient. ACEA: Consultant and honorarium recipient. Pfizer: Grants/research support, consultant, honorarium recipient. BMS: Grants/research support. Xcovery: Grants/research support. Celgene: Grants/research support. Roche/Genentech: Grants/research support. Medimmune: Grants/research support. Lilly: Grants/research support. S. Patel: Research funding from: Bristol-Myers Squibb, Eli Lilly, MedImmune, Pfizer, Roche/Genentech, Xcovery. Speaking fees from: Boehringer Ingelheim, Merck. G. Blumenschein: Grants/Research Support Recipient, Consultant- BMS, Bayer, Merck, Celgene Consultant- Clovis, AbbVie Grants/Research Support Recipient- Novartis, Xcovery, Astrazeneca. J. W. Neal: Consulting or Advisory role: Clovis Oncology, CARET/Physicians Resource Management, Nektar, Boehringer Ingelheim Research Funding- Genentech/Roche, Merck, ArQule, Novartis, Exelixis, Boehringer Ingelheim, Nektar. B. Gitlitz: Speakers Bureau for Lilly Speakers Bureau for Genentech. F. Tan: Manager- Xcovery Holding Company Chief Medical Officer- Betta Pharmaceuticals. K. Harrow: Xcovery Holding Company- Full-time Employee. L. Horn: Consulting for Xcovery Holding Company, BMS, BI, abbvie, Genentech, Merck. All other authors have declared no conflicts of interest.

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided

Abstract not provided