Search

3039 Background: Molecular profiling of tumor tissue is the gold standard for treatment decision making in advanced non-small cell lung cancer. Results may be delayed or unavailable due to insufficient tissue samples or prolonged wait times for biopsy, pathology assessment and testing. We piloted the use of plasma molecular testing as part of the initial diagnostic work-up for patients with suspected advanced lung cancer (NCT04863924). Methods: Patients with radiologic evidence of advanced lung cancer referred to the lung rapid diagnostic program underwent plasma circulating tumor DNA (ctDNA) testing using InVisionFirst-Lung, a next-generation sequencing (NGS) assay targeting 37 genes. Standard tissue testing was performed with comprehensive NGS (Oncomine). The primary endpoint was time to treatment in stage IV NSCLC patients compared to an historical pre-COVID-19 cohort (2018-9). Secondary endpoints included actionable targets identified in plasma, % of patients starting targeted therapy based on liquid biopsy and result turnaround time (TAT). Results: Between July 1 to December 31, 2021, 60 patients were enrolled. Median age was 70 years (range 33-91), 52% were female, 57% Caucasian, 48% never smokers. Of these, 73% had NSCLC, 12% small cell, 10% non-lung pathology and 5% declined tissue biopsy. Of 44 NSCLC patients, 5 (11%) had early-stage disease and underwent curative therapy. Most stage IV patients (79%) had systemic treatment. Median time to treatment initiation in the study cohort was 34 days (n = 31, range 10-90) versus 62 days (n = 101, range 13-159) in the historical cohort (p

e16074 Background: Immune checkpoint inhibition (ICI) is an effective treatment for a subset of patients with inoperable esophagogastric (EG) adenocarcinoma. Robust predictive biomarkers are required to identify these patients and a variety of strategies including immunohistochemical staining of PD-L1 and tumor mutational burden (TMB) assessment have been employed. Here, we explore digital histological (dHis) markers based on routine hematoxylin and eosin (H&E) slides alone or in combination with molecular markers (PD-L1 and TMB) as predictive biomarkers of benefit from maintenance immunotherapy in patients with inoperable EG adenocarcinoma. Methods: We developed a deep learning based algorithm to construct novel digital histological (dHis) markers by summarizing the statistics of all different types of nuclei present in the tumor tissue sections, their morphological features and their colocalization across each of the whole slide image. The dHis markers were then input into a decision-tree based approach to test for prognostic and predictive power alone or in combination with molecular markers. We assessed two cohorts of patients randomized to surveillance (n=38) or maintenance durvalumab (n=35) after 18 weeks of first-line platinum-based chemotherapy in the PLATFORM trial (NCT02678182) according to the 12-week progression-free rate. We measured the accuracy as the area under the receiver operating characteristics curve (AUROC) to determine the prognostic and predictive power of each marker set. We conducted a stratified 3-fold cross-validation, repeated 5 times and report the overall AUROC results. Results: Molecular markers alone yielded an AUROC of 0.5581±0.0939 on the surveillance arm, 0.6671±0.1479 on the treatment arm, and 0.6376±0.0958 for both the arms. Digital histological markers alone yielded an AUROC of 0.8952±0.0638, 0.8995±0.0719 and 0.8488±0.0700 on surveillance, immunotherapy and both arms, respectively. When using these two sets of markers together for both arms, molecular markers offered a limited improvement (around 0.02). Patients with TMB in the highest tertile were associated with lower likelihood of having progressive disease 12 weeks after randomization. Interestingly, dHis markers from morphology of connective and inflammatory nuclei were highly predictive for treatment benefit. Conclusions: Preliminary results suggest digital histological markers offer significant improvement over PD-L1 and TMB markers alone for predicting benefit from immunotherapy in EG adenocarcinoma with the added advantages of scalable, rapid, low-cost and objective quantification on routine histology sections. We are further validating their effectiveness on a larger cohort. Clinical trial information: NCT02678182.

4015 Background: PLATFORM is a prospective, open-label, multicentre, adaptive phase II trial assessing maintenance therapy in patients (pts) with OG adenocarcinoma after platinum-based first-line induction chemotherapy. HER2 negative pts were initially randomised 1:1:1:1:1 to surveillance (A1), capecitabine (A2), durvalumab (A3), with rucaparib (A4) and capecitabine + ramucirumab (A5) added later as per adaptive design. Here we report the primary analysis of durvalumab maintenance vs. surveillance. Methods: Pts were randomised upon achieving response or stable disease following 18 weeks of platinum-based chemotherapy and were stratified by region, disease extent and performance status. A1 pts had 4 weekly surveillance visits and A3 pts received 10mg/kg durvalumab iv Q2W. Target accrual was 154 pts/arm; however, A3 was closed prematurely after cessation of industry support. The primary endpoint was progression-free survival (PFS) from randomisation post induction chemotherapy to disease progression according to RECIST 1.1 criteria, or death. Secondary endpoints were time to treatment failure (TTF), objective response rate (ORR), overall survival (OS) and toxicity. Survival analyses according to PD-L1 using the combined positive score (CPS, Ventana SP263 assay) was conducted. Results: A total of 205 pts were concurrently randomised into A1 (n = 100) and A3 (n = 105). At data cut-off (02 Feb 2021), median follow up was 24.2 months [95% confidence interval (CI) 21.6-36.8]. There was no significant difference in PFS between A1 and A3 [median PFS: 3.2 vs. 4.7 months (hazard ratio (HR) 0.79 (95% CI 0.59-1.06, p = 0.122) respectively]. Median OS was 11.4 vs. 11.3 months (HR 0.92, 95% CI 0.66-1.27, p = 0.60) in A1 and A3 respectively and no significant difference in TTF was detected between both arms (median TTF: 3.2 vs. 3.5 months (HR 0.92, 95% CI 0.69-1.23, p = 0.558)]. ORR was 6% in A3 (n = 2 complete and n = 4 partial responses) whereas no radiological responses were seen in A1. PD-L1 results were available for 76 pts in A1 and 77 pts in A3. PFS was comparable between subgroups using the PD-L1 CPS thresholds of ≥1, ≥5 and ≥10. The safety population consisted of 199 pts (A1: 98 pts and A3: 101 pts). Grade ≥3 treatment-related adverse events were reported in 18% of A3 pts and no new safety signals were identified. Conclusions: Although a survival advantage was not seen with maintenance durvalumab compared to surveillance, a subset of patients who received durvalumab demonstrated incremental radiological responses. Exploratory analysis of PD-L1 expression was not associated with improved survival outcomes. Clinical trial information: NCT02678182.

4500 Background: In the first interim analysis of the randomized, multicenter, open-label, phase 3 KEYNOTE-426 study (NCT02853331), treatment with pembro + axi significantly improved OS, PFS, and ORR vs sunitinib monotherapy in treatment-naive advanced ccRCC. Extended follow-up (median, 30.6 mo) continued to demonstrate the superior efficacy of pembro + axi vs sunitinib monotherapy in this patient population. Here, we present the results of the prespecified final analysis with 42.8-mo median follow-up. Methods: Treatment-naive patients (pts) with advanced ccRCC, KPS ≥70%, and measurable disease (RECIST v1.1) were randomly assigned 1:1 to receive pembro 200 mg IV Q3W for up to 35 doses + axi 5 mg orally BID or sunitinib 50 mg orally QD on a 4-wk on/2-wk off schedule until progression, intolerable toxicity, or withdrawal. Randomization was stratified by IMDC risk (favorable vs intermediate vs poor) and geographic region (North America vs Western Europe vs Rest of World). Dual primary endpoints were OS and PFS. Secondary endpoints were ORR, DOR, and safety. The protocol-specified final analysis was based on a target of 404 OS events. No formal hypothesis testing was performed because all efficacy endpoints were met previously at the first interim analysis; nominal P values are reported. Results: Overall, 861 pts were randomly assigned to receive pembro + axi (n=432) or sunitinib (n=429). Median duration of follow-up, defined as time from randomization to the database cutoff date, was 42.8 mo (range, 35.6-50.6). At data cutoff, 418 pts had died: 193 (44.7%) of 432 pts in the pembro + axi arm vs 225 (52.4%) of 429 pts in the sunitinib arm. Compared with sunitinib, pembro + axi improved OS (median: 45.7 vs 40.1 mo; HR, 0.73 [95% CI, 0.60-0.88]; P

e16561 Background: CheckMate 025 demonstrated favorable efficacy and safety results for nivolumab monotherapy in previously treated advanced or metastatic renal cell carcinoma (aRCC). However, real-world evidence on treatment patterns and clinical outcomes is limited. Methods: This multi-centre, retrospective cohort study examined treatment patterns and overall survival (OS) in aRCC patients treated with nivolumab monotherapy. Eligible patients who initiated nivolumab at second-line (2L) or beyond (index) between 01 March 2016 and 30 June 2018 were sampled from four UK centers. Data were extracted using an electronic case report form from index to earliest of: most recent visit; death; end of follow up (31 May 2019). Results: Overall , 151 patients were included in analyses (mean age at index 66.9 years, 72.2% male, median follow-up from index 15.2 months), with 109 (72.2%) and 42 (27.8%) receiving nivolumab at 2L and ≥ third-line (3L+), respectively. Key clinical characteristics are outlined in Table 1. All 2L nivolumab patients had received first-line (1L) tyrosine kinase inhibitors (TKI), pazopanib (57.8%), sunitinib (30.3%), or both in sequence (10.1%). After 2L nivolumab, 3L cabozantinib (36/52, 69.2%) was most common. Most 3L nivolumab patients received 2L TKI (31/36, 86.1%) - commonly axitinib (70.9%). After 3L nivolumab, most patients received fourth-line cabozantinib (8/12, 66.7%). Median time on line of therapy (LOT) decreased with LOT progression: from 7.8 months at 1L to 4.6 months at fifth-line (5L). The proportion of patients who discontinued treatment due to adverse events decreased by LOT, (28.6%, 22.7%, 16.0% and 0%, and 34.7%, 28.1%, 0% and 0% from 2L to 5L, overall and for nivolumab treatment, respectively). Overall, median OS from nivolumab initiation was 19.2 months [95% CI, 16.9-27.0]. Patients who received 2L nivolumab had longest median OS (23.0 months [95% CI, 17.2, not reached]), comparable to CheckMate 025 (25.8 months [95% CI, 22.2-29.8]). Median OS for 3L+ nivolumab patients was 12.4 months [95% CI, 8.8, 23.2]. Among 2L nivolumab patients, 73.9%, 46.2%, and 33.6% survived 12, 24, and 36 months, respectively. For the same respective timeframes, 52.4%, 24.7%, and 18.6% of 3L+ nivolumab patients survived. Conclusions: This study provides real-world evidence on the characteristics, treatment patterns and effectiveness of 2L or ≥ 3L nivolumab monotherapy in previously treated aRCC patients. OS results from UK routine clinical care were comparable to those found in CheckMate 025.[Table: see text]

4039 Background: Despite potentially curative surgery, long-term survival from OGC remains poor due to high relapse rate. Neoadjuvant (naFLOT) and adjuvant (aFLOT) FLOT is currently standard treatment for resectable OGC based on data from the FLOT-4 trial. We explored whether TRG was associated with FLOT-outcome using RWD. Methods: Pts with OGC treated with naFLOT +/- aFLOT at a tertiary UK centre were identified following institutional board approval. Clinical and laboratory data were extracted from the patient record. TRG was evaluated by a histopathologist. Median overall survival (OS) and median progression-free survival (PFS) were evaluated using Kaplan-Meier and Log-rank tests; time taken from start of naFLOT, and associations between factors with Fisher’s exact (FE) test. Results: 171 pts were identified, median FU 30 mths. 144 (84%) male; median age 66 (32-84); oesophagus 66 (38%), junctional (GOJ) 73 (43%), gastric 32 (19%); stage IB 3 (2%), stage IIB 26 (15%), stage III 91 (53%), stage IVA 47 (28%) and unknown 4 (2%). Pts had median of 2 comorbidities (range 0-6); performance status (PS) 0: 95 (56%), PS 1: 71 (41%), PS 2: 3 (2%) and PS unknown 2 (1%). 132/171 pts completed 4 cycles of naFLOT and this was significantly associated with undergoing surgery (p = 0.02). Those who had surgery (140/171) had significantly improved PFS (not reached (NR) vs. 6 mths; 95% CI 2-10; p < 0.001) and OS (NR vs. 12 mths; 95% CI 6-18; p < 0.001). TRG was reported for 126/140 patients who underwent surgery. TRG 1/2 (42/126) vs. TRG ≥3 was significantly associated with improved PFS (NR vs. 35 mths; 95% CI NR; p < 0.001) and OS (median NR either group; p < 0.001). Pts with TRG 1/2 who commenced aFLOT (≥1 cycle; n = 31/42) or completed 4 cycles of aFLOT (17/31) did not have improved PFS or OS vs. those who did not. Those with TRG ≥3 who commenced aFLOT (≥1 cycle; n = 62/85) had improved PFS (median NR vs. 22 mths; 95% CI 13-31 p = 0.006) and OS (median NR vs. 25 mths; 95% CI 18-32 p = 0.019). Those with TRG ≥3 who completed 4 cycles of aFLOT (n = 38/62) had significantly improved PFS (median NR vs. 25 mths; 95% CI 14-36 p = 0.016) and OS (median NR vs. 36 mths; 95% CI 16-55 p = 0.012). There was no difference in PFS or OS in pts with TRG ≥3 who had a dose reduction at any time during aFLOT. Conclusions: TRG is a predictor of outcome following naFLOT + surgery with superior outcomes in those with TRG 1/2. Our analyses suggest that only pts with TRG >3 following naFLOT + surgery benefit from adjuvant FLOT. Prospective randomised studies are required to confirm whether pts with TRG 1/2 require treatment with aFLOT.

294 Background: The combination of the immune checkpoint inhibitor avelumab and VEGF-targeted, antiangiogenic tyrosine kinase inhibitor axitinib (AA) has demonstrated superior PFS and ORR compared to sunitinib in patients with mRCC and is an option for first line treatment across all IMDC risk scores. In this retrospective review we report our real world experience of this combination in three cancer centres in the Northwest of England. Methods: Treatment naïve mRCC patients receiving AA through the Early Access to Medicine Scheme at 3 cancer sites in the UK between May 2019 and July 2020 were identified. Primary outcomes of interest include overall response rate (ORR), adverse events (AEs) and preliminary survival observations. Results: A total of 44 patients were identified with a median follow up of 6.9 months (0.8-13.5 mo). Median age was 68 (48-81); 68% were male. The patients’ adult comorbidity evaluation score (ACE-27) was calculated: 0 = 43%, 1 = 30%, 2 = 7% and 3 = 20% . 45%, 48% and 7% of patients had favourable (F), intermediate (I) and poor (P) IMDC risk scores respectively. All had clear cell histology with 16% demonstrating sarcomatoid change. Most patients had undergone a nephrectomy (70%) and 36% had a single organ site of metastatic disease. ORR in the whole cohort was 60% (CR 5%, PR 55%, SD 25%, PD 2%, NE 13 %). Median time to first response was 2.6 months (0.6- 8.2mo). At time of data cut-off, 64% of patients remain on treatment (80% F, 48% I and 67% P). 14% of patients discontinued treatment due to disease progression while 22% stopped due to toxicity. The majority of patients (68%) continued axitinib at the starting dose of 5mg BD. Dose escalation of axitinib was possible in 9% patients while 23% needed a dose reduction due to toxicities. AEs were observed in 36 (82%) patients (G3 36%); the commonest being mucositis 30%; hypertension 23% (G3 11%); fatigue 25%; thyroid dysfunction 18%; diarrhoea 20% (G3 5%); hepatitis 20% (G3 11%). 9% of patients experienced an infusion reaction to avelumab. Overall, 9 (20%) patients received steroids for suspected immune related adverse events (irAEs); 6 (14%) were managed as G3≤ irAEs. 9 (20%) patients required inpatient admission due to AEs; 5 (11%) were associated with irAEs. Of the patients who discontinued AA treatment, 50% received subsequent therapy (12.5%, 75% and 12.5% receiving combination checkpoint inhibitor therapy, other VEFG TKi and TKi/MTOR combination respectively). 4 patients remain on active surveillance with no evidence of progression. Conclusions: Our early experience of AA in this real world setting reports comparable clinical responses to the published data. Treatment is well tolerated, with lower than expected levels of G3 or above AEs which is reassuring in a non-trial selected population. Follow-up is ongoing and updated efficacy and safety outcomes will be presented.

295 Background: Ipilimumab and Nivolumab (I+N) is now an established first line option for patients with advanced RCC of intermediate (I) or poor (P) IMDC risk score. In this retrospective review, we review our experience of this combination in two cancer centres in North West England with a focus on immune related adverse events (irAEs) and their impact on the patient pathway. Methods: Treatment naïve mRCC patients starting I+N between May 2019 and July 2020 were identified. Primary outcomes of interest include overall response rate (ORR), the management of irAEs and early survival observations. Results: A total of 69 patients were identified. Median age was 60yr (19-82yr), 77% had clear cell histology. The IMDC risk was 72% I and 28% P. Median follow-up was 11.0 mo (1-22mo). ORR was 45% (CR 9%, PR 36%, SD 28%, PD 23%, NE 4%) Median time to first response was 2.9mo. (1.8- 15.5mo). 10% of patients experienced pseudoprogression. Median PFS and OS are not yet reached with 86% of patients still alive at the time of data cut-off. The majority (75%) of patients completed all 4 doses of I+N. Of the 10% receiving less than 4 doses due to toxicity, 14% continued on single agent N. Overall, 15% discontinued therapy due to toxicity and 28% experienced at least one treatment delay. Any grade irAEs were seen in 74% of patients (G3 35%) with no treatment related deaths. The commonest irAEs were: rash/pruritis 39%; endocrinopathies 30%(G3 7%); diarrhoea 29% (G3 14%); hepatitis 22% (G3 6%); and nephritis 3% (G3 3%). Of the patients developing irAEs, 71% received steroids with 16% requiring additional immunosuppression including infliximab (6%) and mycophenolate mofetil (8%). A third of all patients required admission for irAE management with a total of 37 inpatient episodes across the cohort with a median length of 7 days (1-24). 7% of patients proceeded to surgery for either primary or metastatic disease, which contributed to ongoing disease response in these patients. At the time of data cut-off, 45% of patients were no longer on treatment due to PD (29%), toxicity (15%) or unrelated death (1%). Of those who stopped due to toxicity, 50% subsequently progressed with a median time to progression of 4mo (3-6 mo) and 50% remain on active surveillance with a median follow-up of 7.5mo (1-10). 62% of patients with PD received second line treatment; most frequently, cabozantinib (83%). Conclusions: Our experience of I+N shows comparable efficacy and toxicity profiles to available reports. irAEs requiring intervention are frequent and may be associated with prolonged hospital admission, and patients should be counselled appropriately. Data within mirrors published reports of ongoing responses in a subset of patients who stop treatment due to toxicity and also suggests a possible role for resection of residual or metastatic disease in disease control. Updated survival data will be presented.

327 Background: In the randomized, open-label, phase III KEYNOTE-426 study (NCT02853331), pembrolizumab + axitinib significantly improved overall survival (OS), progression-free survival (PFS), and objective response rate (ORR) versus sunitinib as first-line therapy for advanced RCC. Per protocol, patients could discontinue pembrolizumab or axitinib and continue the other agent. Pembrolizumab was stopped for all patients at 2 years. Axitinib could be continued until progression or toxicity. This exploratory subgroup analysis of KEYNOTE-426 describes outcomes of patients who completed 2 years of pembrolizumab. Methods: Patients included in KEYNOTE-426 were treatment naive, with clear cell RCC, KPS ≥70%, and measurable disease (RECIST v1.1). Patients were randomly assigned 1:1 to receive pembrolizumab 200 mg intravenously every 3 weeks for up to 35 doses + axitinib 5 mg orally twice daily or sunitinib 50 mg once daily (4 weeks on/2 weeks off) until progression, toxicity, or withdrawal. Primary end points of the original analysis were OS and PFS. Key secondary end points were ORR and safety. Results: Of 432 patients treated with pembrolizumab + axitinib, 129 (29.9%) completed 2 years of study therapy. Median (range) age of these patients was 61 (36-82) years, and 72.1% were male; 42 (32.6%) and 87 (67.4%) patients had International mRCC Database Consortium favorable and intermediate/poor risk, respectively, consistent with the intention-to-treat population (31.9% vs 68.1%). Median (range) follow-up (time from randomization to data cutoff) was 31.1 (24.0-37.7) months. For patients who completed 2 years of study therapy, the OS rates at 36 months was 93.8% (95% CI, 85.5%-97.4%). The PFS rates at 24 and 36 months were 72.7% (95% CI, 64.0%-79.7%) and 57.7% (95% CI, 46.3%-67.5%), respectively. The ORR was 85.3%, and the CR rate was 14.0%. 59.7% of patients experienced grade 3-5 treatment-related adverse events and 8.5% experienced grade 3-5 immune-mediated adverse events. Conclusions: In this exploratory analysis, a significant proportion of patients in the pembrolizumab + axitinib arm completed 2 years of pembrolizumab with ongoing clinical benefit. Clinical trial information: NCT02853331 .

e16545 Background: Purpose: The aim of this retrospective study was to assess the efficacy and toxicity of irinotecan +/- capecitabine for patients with metastatic esophago-gastric (EG) adenocarcinomas previously treated with at least two chemotherapy regimens including platinum-based, fluoropyramidines and taxanes. Introduction: Treatment for metastatic EG cancer beyond 2nd line remains controversial. Recently, the TAGS phase III trial (Shitara et al., 2018) showed overall survival benefit with Trifluridine/tipiracil of 2.1 months compared to placebo as a third line. Apatinib is also approved for use as third line in China (Li et al., 2013). Based on minimal evidence (Kang et al., 2013), irinotecan in combination with capecitabine is used in the UK. We performed a retrospective study to assess the efficacy and toxicity of irinotecan as a beyond 2nd line regime in metastatic EG adenocarcinomas in one of Europe’s largest cancer centres, The Christie. Methods: Data was collected retrospectively from all metastatic EG adenocarcinoma patients who were treated with irinotecan +/- capecitabine as a beyond 2nd line palliative treatment at the Christie Hospital from January 2014 to December 2019. Irinotecan was either given 2 or 3 weekly at dose of 180 mg/m² with or without capecitabine 800 mg/m². Response rate (RR) was calculated according to RECIST version 1.1, overall survival measured and toxicity data collected. Results: Fifty-six patients received irinotecan as beyond 2nd line palliative chemotherapy. Out of the thirty-seven patients with measurable disease on their scans, the overall response rate was 8.1% and the disease control rate was 51.4% as per RECIST 1.1. Median overall survival was 5 months (95% CI 4.2-5.8). The presence of grade 3-4 toxicities was 16% (9 patients) and included neutropenia, fatigue, nausea and diarrhea. On average, 3.7 cycles were administered, 32% (18 patients) required dose reductions, mainly secondary to grade 2 diarrhea and fatigue. Conclusions: Irinotecan in combination with capecitabine is efficacious as a beyond 2nd line regime in metastatic EG with an acceptable toxicity profile with a limited role in patients with resistant disease to first line platinum-based chemotherapies with fluoropyramidines.

638 Background: Tivozanib is a selective tyrosine kinase inhibitor (TKI) of vascular endothelial growth factor and has been shown to offer PFS and tolerance advantage compared to sorafenib among patients with mRCC. In this retrospective review, we aim to investigate the real world efficacy and tolerance of tivozanib delivered in four cancer hospital in the Northwest of England. Methods: mRCC patients started tivozanib in first line setting were identified and reviewed. Primary outcomes of interest include overall response rate (ORR), survival (OS, PFS where possible) and treatment tolerance. Results: A total of 113 patients were identified between March 2017 and May 2019. Median follow up was 200 days (15-792). 26% were switched from other prior TKI due to intolerance. 28%, 48% and 24% had Favourable (F), Intermediate (I) and Poor(P) IMDC risk category respectively. ORR was 29% (CR 0%, PR29%, SD38%, PD26%, NE 7%). Median PFS (after 53 events) was 9 months (F = NR, I = 7 months, P = 3 months p value < 0.0001) with estimated 6 and 12 months OS of 80% and 67 % respectively. At cut-off, 26/32 with F IMDC risk remaining on treatment c.f. 24/54 (I) and 6/27 (P). Median treatment received was 5 cycles and 65% still on full dose at end of observation. Dose reduction was necessary in 31% while treatment was stopped in 15% due to toxicity. 46% received subsequent therapy post- progression. The commonest adverse events were fatigue (32%, G3 0%), diarrhoea (15%, G3 1.7%), mucositis (15%, G3 < 1%), and anorexia (7%, G3 1.7%). Conclusions: Preliminary findings from this review suggests similar clinical efficacy of tivozanib compared to agents such as pazopanib or sunitinib in real-world setting particularly among patient with Favourable IMDC category however longer follow up is required to fully evaluate this. Treatment is well tolerated with low incidence of severe grade toxicities and may be a good monotherapy option in patient of Favourable IMDC category unsuitable for combination therapies.

282 Background: Advanced OGA patients (pts) are treated with platinum-based 1st line chemotherapy but the role of maintenance therapy once disease control is obtained is unknown. Methods: PLATFORM is a prospective, open-label, adaptive phase II trial assessing maintenance therapy in OGA. HER2 negative pts with advanced OGA achieving response or stable disease on completion of 1st line platinum-based chemotherapy were initially randomised to surveillance (A1), capecitabine (A2) and durvalumab (A3). Rucaparib (A4) and capecitabine+ramucirumab (A5) were subsequently added as part of adaptive design. Primary endpoint is progression-free survival with target recruitment of 154pts/arm. A pre-planned futility interim analysis (IA) is triggered when 61 pts/arm are recruited and evaluable using 12-week progression-free rate (PFR). Individual arms will continue accrual if the upper limit of 1-sided 95% CI difference is > 0 when compared to A1. Biomarker analyses of A1 and A3 IA patients include: PDL1 as tumour and immune cell combined proportion (TIP), multiplex IHC of tumour infiltrating lymphocytes (TILs), TMB and MSI status by whole exome sequencing. Results: To date, 1053 pts have registered prior to or during 1st line therapy and 356 pts randomised. Primary attrition was due to disease progression. Baseline demographics in the IA population (arms A1 to A3, n = 183) were male (81%); median age 65 yrs; metastatic disease (92%); oesophageal (40%), OG junction (28%) and gastric (32%) primary. 12-week PFR were A1: 51% (95%CI: 38-64%); A2: 52% (95%CI: 40-65%); A3: 49% (95%CI: 36-62%). PFR differences to arm A1 were A2: 1.6% (95%CI: -13.2 to 16.5%); A3: -1.6% (95%CI: -16.5 to 13.3%). In A1, PFR was 53% vs 43% in PDL1 TIP < 10% and ≥10% respectively. In A3, PFR was 51% vs 100% in PDL1 TIP < 10% and ≥10% respectively. 3 pts in A3 had partial response; none were seen in A1 and A2. Density of TILs subsets, TMB and MSI data will be presented. Conclusions: PLATFORM IA did not indicate futility in maintenance capecitabine or durvalumab compared to surveillance in advanced OGA and will continue to target accrual. Incremental radiological responses were observed with maintenance durvalumab only. Clinical trial information: NCT02678182.

Purpose Several microarray studies have reported gene expression signatures that classify non–small-cell lung carcinoma (NSCLC) patients into different prognostic groups. However, the prognostic gene lists reported to date overlap poorly across studies, and few have been validated independently using more quantitative assay methods. Patients and Methods The expression of 158 putative prognostic genes identified in previous microarray studies was analyzed by reverse transcription quantitative polymerase chain reaction in the tumors of 147 NSCLC patients. Concordance indices and risk scores were used to identify a stage-independent set of genes that could classify patients with significantly different prognoses. Results We have identified a three-gene classifier (STX1A, HIF1A, and CCR7) for overall survival (hazard ratio = 3.8; 95% CI, 1.7 to 8.2; P < .001). The classifier was also able to stratify stage I and II patients and further improved the predictive ability of clinical factors such as histology and tumor stage. The predictive value of this three-gene classifier was validated in two large independent microarray data sets from Harvard and Duke Universities. Conclusion We have identified a new three-gene classifier that is independent of and improves on stage to stratify early-stage NSCLC patients with significantly different prognoses. This classifier may be tested further for its potential value to improve the selection of resected NSCLC patients in adjuvant therapy.

e15580Background: Hospital readmission after esophagectomy is a useful metric for measuring resource utilization. Our objective was to evaluate readmission rates within 1 year of discharge home aft...

e14515 Background: There is still controversy whether full lung palpation is required for patients undergoing pulmonary metastasectomy. We aimed to compare pulmonary ipsilateral recurrence (IR) aft...

6547 Background: Molecular testing is now standard of care to guide treatment selection in advanced NSCLC. Population-based implementation of testing remains a challenge in Ontario’spublic healthca...

10597 Background: Isolated lung perfusion (ILP) can enhance the treatment of lung metastases, allowing the localized delivery of drugs to the lungs, without systemic exposure. Previously, short-ter...

40 Background: Cigarette smoking, alcohol consumption and co-morbidities are important determinants of health in lung cancer patients. The gold standard for obtaining accurate data is PRQ. The purpose of this study is to ascertain the accuracy of abstracting health-related behaviour data from retrospective chart review compared to data directly obtained from PRQ in a lung cancer patient population. Methods: 731 lung cancer patients completed a PRQ related to lifetime tobacco use, alcohol consumption and co-morbidity. Relevant smoking, alcohol and co-morbidity data was collected independently from EPR. Results: Ever/never status for smoking showed almost perfect agreement (k=0.95) between PRQ and EPR and surpassed all other health behavioural measures and co-morbidity agreement values. Both the sensitivity and specificity were high (0.94 and 0.99 respectively). The calculation of pack-years from EPR and PRQ showed substantial agreement (k=0.77); However, categorizing the smoking status into current/ former / never, resulted in moderate agreement (k=0.46). Alcohol ever/ never status agreement was moderate (0.43) with high sensitivity (0.90) but low specificity (0.50). Agreement for co-morbidities varied by condition showing moderate to substantial agreement for hypertension (K=0.57), heart attack (K=0.80) and diabetes (K=0.76) while fair to slight agreement (K

7516 Background: Erlotinib has demonstrated major activity in EGFR mutation positive NSCLC, but may also benefit those with wild-type tumours. We conducted a single-arm trial of pre-operative erlotinib in early stage NSCLC to assess radiologic and functional response as well as correlation with known and investigational biomarkers. Methods: Patients with clinical stage IA-IIB NSCLC received erlotinib 150 mg daily for 4 weeks followed by surgical resection. Tumor response was assessed using pre- and post-treatment CT and PET imaging. Pharmacodynamic changes were assessed through comparison of pre- and post-treatment tumour samples (including Sequenom MassARRAY analysis) and measurement of circulating markers/ligands for EGFR activation (TGF-α, amphiregulin, epiregulin, EGFR SNP, EGFR ECD). Secondary endpoints included pathological response, toxicity and progression-free survival. Results: Twenty-five patients were enrolled; 22 received erlotinib treatment with a median follow up of 4.4 years (range 2.2 to 6.4 years). Histology was predominantly adenocarcinoma (14) with smaller numbers of squamous carcinoma (7) and large cell carcinoma (1). PET response (25% SUV reduction) was observed in 2 patients (9%), both with confirmed squamous carcinoma histology. All patients met criteria for stable disease by RECIST and several experienced minor radiographic regression with histologic findings of fibrosis/necrosis, including 2 with squamous histology. The presence of an EGFR exon 19 deletion was detected in one adenocarcinoma case; the patient experienced a minor radiographic response to treatment. Genotyping, functional protein assays and ligand analysis are ongoing. Conclusions: Erlotinib appears to demonstrate some activity in patients with squamous histology. While EGFR mutations have been infrequently demonstrated in squamous NSCLC, the potential exists for other biomarkers predictive of benefit. Clinical trial information: NCT00462995.

e18503 Background: Treatment of locally advanced NSCLC is debated. While survival for inoperable disease ranges between 3-17%, carefully selected patients can be cured when treated surgically in a multimodality concept. Here, we conducted a systematic literature review and pooled data analysis of patients after en bloc resection for pulmonary sulcus NSCLC invading the spine. Outcome and prognostic factors were described. Methods: The MEDLINE database was searched using the PubMed engine to retrieve all relevant articles related to en bloc resection for pulmonary sulcus NSCLC invading the spine. All articles’ corresponding authors were contacted to share their most updated anonymized patient’s data. Data were pooled and analyzed, focusing on outcome and prognostic factors. Results: Search strategy yielded a total of 134 articles. Out of these, only 6 were relevant and non-duplicative. Four out of 6 authors were able to share updated data for a total of 135 patients. All tumors were resected en bloc with the lung, chest wall and spine. Induction treatment was administered in 87 (64%) patients and consisted of chemotherapy (n=32), radiation (n=1) or concurrent chemoradiation (n=54). Spine resections included total (n=23), hemi- (n=94) and partial (n=18) vertebrectomies. Complete resection was achieved in 120 (89%) patients. Five patients died in the postoperative period (4%). Adjuvant treatment was administered in 70 (52%) patients and included chemotherapy (n=16), radiotherapy (n=22) or chemoradiation (n=32). Median follow-up was 26 months. Overall 3-, 5- and 10-year survivals were 57%, 43% and 27%, respectively. Results of the univariate analysis (Cox, Breslow tests) identified incomplete surgical resection (R0 vs R1/2, p

46 Background: Carcinoma of the esophagus has become one of the fastest growing solid tumors in the world over the past 20 years and the 6th most common cause of death in the world. We previously conducted a study to profile the expression of miRNAs in esophageal adenocarcinoma (EAC) pre and post induction therapy. Out of the miRNAs discovered, miR-145, a known tumor suppressor miRNA, was upregulated 8-fold upon induction therapy however its expression was associated with shorter disease-free survival. This unexpected result was explored. Methods: In order to study the role of miR-145 in EAC, miRNA-145 was overexpressed in the EAC cell line OE33 (OE33 miR145). After validation of the expression of miR-145, several hallmarks of cancer such as cell proliferation, resistance to apoptosis and cell adhesion were analyzed. Results: There was no difference in cell proliferation and resistance to radiation between OE33 miR145 and OE33 control. However, there was a significant difference in cell adhesion. OE33 miR145 cells reattach faster than the OE33 control (72.3% cells reattached against 40.5% cells reattached after 3h, n=3, p

17074 Background: Early stage NSCLC is commonly diagnosed in elderly patients who often have significant medical co-morbidity. SBRT is emerging as a treatment alternative that can provide 80–97% local control for peripheral tumors. We performed a literature review and a retrospective review of 24 patients referred to Princess Margaret Hospital for consideration of SBRT to assess what criteria were used to determine patient selection for this new treatment. Methods: Published papers reporting treating stage I NSCLC with SBRT were reviewed. Clinical and demographic characteristics of patients referred for consideration of SBRT over the last 15 months were analyzed. All patients were assessed by experienced thoracic surgeons to determine their operability status. Patients were retrospectively scored as average or high risk for surgery on the basis of accepted PFT guidelines (FEV1 and DLCO 2 associated with impaired survival). Results: Of 13 papers reporting Stage 1 SBRT results, 8 included details of medical inoperability. 2 papers had strict guidelines for PFTs, in addition to other criteria. In 6 studies reporting 276 pts, the most common reason for inoperability was non-specific medical comorbidity (60%). Our cohort (24 patients) referred for SBRT (median age 75, range 57–89) had a mean tumour size of 2.32cm (S.D. 1.11); 50% of patients had a good (0–1) Zubrod Performance status. Based upon PFT tests alone 10 patients (42%) were high risk and 5 (21%) average risk for surgery. The CCI score was ≥3 in 10 patients (42%). 3 patients who scored average risk and CCI ≤2 were deemed medically inoperable due a single severe medical comorbity. In 23/24 pts the surgeon recommended against surgery; one patient refused surgery. Conclusions: With an aging population and increased detection due to incidental finding or screening, it is important to know that there is effective alternative therapy for patients deemed too high risk for surgical resection. Comorbidity is an important factor influencing surgical mortality; defining explicit criteria for inoperability remains a clinical challenge. No significant financial relationships to disclose.

7217 Background Lung cancer remains a leading cause of cancer-related mortality in North America. Despite potentially curative resection, NSCLC patients remain at high risk of relapse and death, wi...