Lung metastases occur in 10-20% of patients with metastatic colorectal cancer (mCRC) and represent the 2nd commonest site for distant spread. Comparing the transcriptome of primary tumours and their lung metastases could more accurately define phenotypic heterogeneity and help identify prognostic markers and novel therapeutic targets. The ethically approved retrospective lung resection translational protocol was used to identify and collect clinical data and tissue samples from patients with CRC who underwent lung metastasectomy with curative intent between 1997 and 2012. Formalin-fixed paraffin-embedded tissue was collected from matching primary CRC tumour, lung metastases and representative areas of normal colorectal or lung tissue. RNA was isolated and the global transcriptome was profiled with TruSeq RNA Access RNA Sequencing. The data was analysed using a bioinformatics guided approach where differentially expressed genes were identified with a two-fold cutoff and a false discovery rate < 0.05. The R package CMS caller was used to classify tumour samples based on 530 gene predictors. Sufficient tissue from matching primary CRC and lung metastasis suitable for RNA sequencing was available in 13 patients. Tumour tissue was available from >1 lung metastasis in 2 cases. 69% of patients were male, median age was 64 (range 38-70), 54% were never smokers and 15% had metastatic disease at diagnosis. The primary tumour site was ascending colon in 15%, sigmoid colon in 38% and rectum in 46%. Out of the 19,374 genes analysed, 944 (4.9%) were differentially expressed when the primary tumour was compared to the first lung metastasis. Hierarchical clustering revealed that genes (n=990) were more likely to be over-expressed in metastases samples compared to primary tumours. Genes that were over-expressed were most likely to be involved in immune response. Amongst primary CRC samples, consensus molecular subtype (CMS) 4 (38%) and CMS 2 (31%) were the most common. In metastases samples, CMS 4 (36%) remained the most common followed by CMS 1 (22%). In lung metastases samples, CMS 1 was associated with a favourable prognosis whereas CMS 4 was associated with a poor prognosis. There was evidence of a switch in CMS subtype between the primary tumour and matched lung metastasis sample in 5 out of 9 cases (56%) where prediction was possible. Most patients (80%) with a switch in subtype had received neo-adjuvant chemotherapy. In our small but unique dataset of matched primary CRC with corresponding lung metastases, a switch in CMS subtype may reflect metastatic processes and selection pressure from treatment. In select cases, there may be potential to exploit CMS subtype switching for therapeutic benefit.

Background Checkpoint inhibitors (CPIs) have transformed treatment of many malignancies. However, to date, response to checkpoint blockade in unselected gastrointestinal (GI) cancers has been disappointing. Tumours which display a non-T-cell inflamed phenotype such as mismatch repair proficient (MMRp) oesophagogastric (OG) and colorectal cancers (CRC) respond less frequently to CPIs. Changes in tumour expression of antigens and increased immune infiltrate are associated with improved response to CPIs. Epigenetic modulation of tumours using HDAC inhibitors can lead to increased tumour antigen presentation, immune cell infiltrates and thereby may increase the chance of response to immunotherapy. Trial design EMERGE is designed to evaluate the safety and efficacy of domatinostat, a selective class 1 histone deacetylase inhibitor in combination with avelumab, an anti-PD-L1 monoclonal antibody in patients with previously treated, inoperable or metastatic MMRp OG and CRC. The trial is conducted in 2 stages: Phase IIA will establish a safe and tolerated dose of domatinostat plus avelumab using a 3 + 3 dose finding design and escalating doses of domatinostat will be examined, dosing of avelumab will remain constant. The phase IIB will use a Simon two stage optimal design to assess efficacy of the combination in achieving radiological response according to RECIST v1.1. The primary endpoint of the phase IIB is best objective response rate (ORR) at 6 months. To rule out an ORR of 5% in the CRC cohort, while aiming for 20%, 1/10 and 4/29 responses are required in the 1st and 2nd stages respectively. To rule out an ORR of 15% in the OG cohort while aiming for 35%, 2/9 and 9/34 responses are required in the 1st and 2nd stages respectively with a 1-sided alpha of 5% and 80% power. Secondary endpoints are DoOR, PFS, OS, DCR, safety and tolerability. Exploratory objectives will investigate dynamic changes in expression of tumour associated antigens and immune infiltrates in baseline and on-treatment biopsies and correlate baseline tumour characteristics and circulating biomarkers with tumour response and survival. Recruitment commenced in January 2019, 83 patients will be recruited over 3 years. (ClinTrials.gov. ID: NCT03812796). Legal entity responsible for the study Royal Marsden Hospital NHS Foundation Trust. Funding Royal Marsden Hospital NHS Foundation Trust, 4SC. Disclosure M. Hubank: Research grant / Funding (self), Research grant / Funding (institution): Boehringer Ingelheim; Research grant / Funding (self), Research grant / Funding (institution): Roche Diagnostics; Research grant / Funding (self), Research grant / Funding (institution): Bristol-Myers Squibb; Research grant / Funding (self), Research grant / Funding (institution): Guardant Health; Research grant / Funding (self), Research grant / Funding (institution): Celgene; Research grant / Funding (self), Research grant / Funding (institution): Eli Lilley. I. Chau: Honoraria (self): Eli-Lilly; Research grant / Funding (institution): Eli-Lilly; Research grant / Funding (institution): Janssen-Cilag; Research grant / Funding (institution): Sanofi Oncology; Research grant / Funding (institution): Merck-Serono; Advisory / Consultancy: Eli Lilly; Advisory / Consultancy: Bristol Meyers Squibb; Advisory / Consultancy: MSD; Advisory / Consultancy: Bayer; Advisory / Consultancy: Roche; Advisory / Consultancy: Merck-Serono; Advisory / Consultancy: Five Prime Therapeutics; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Oncologie International; Advisory / Consultancy: Pierre Fabre. N. Starling: Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): BMS; Research grant / Funding (institution): Pfizer; Travel / Accommodation / Expenses: AstraZeneca; Travel / Accommodation / Expenses: BMS; Travel / Accommodation / Expenses: Eli Lilly; Travel / Accommodation / Expenses: Merck; Travel / Accommodation / Expenses: Roche; Honoraria (self): AstraZeneca; Honoraria (self): Eli Lilly; Honoraria (self): Merck; Honoraria (self): Servier; Advisory / Consultancy: Pfizer; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Servier. D. Cunningham: Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): Celgene; Research grant / Funding (institution): MedImmune; Research grant / Funding (institution): Bayer; Research grant / Funding (institution): 4SC; Research grant / Funding (institution): Clovis; Research grant / Funding (institution): Eli Lilly; Research grant / Funding (institution): Janssen; Research grant / Funding (institution): Merck. All other authors have declared no conflicts of interest.

Background 95% of metastatic colorectal cancers (mCRCs) have normal mismatch repair proficient (pMMR) expression and a stable microsatellite phenotype (MSS). As a consequence, checkpoint inhibiting immunotherapy currently plays no role in these tumours. We recently found that RAS/RAF wild type (WT) mCRCs (50% of all CRCs) that first responded to cetuximab and then acquired resistance had converted from an immune-excluded or immune-desert phenotype before treatment to an inflamed phenotype at progression. This progression was characterized by increased CD8+ T cell infiltrates and upregulation of PDL1 and LAG3 immune checkpoints. To assess if the cetuximab induced immune infiltrates can be exploited for therapeutic benefit, the iSCORE trial will treat 25 patients with combined anti-PD1 (nivolumab) and anti-LAG3 (relatlimab) immunotherapy starting ideally within 4 weeks after progression on immunogenic FOLFIRI chemotherapy and cetuximab. Trial design iSCORE is designed to evaluate the efficacy of nivolumab and relatlimab in patients with RAS/RAF WT mCRC who have had radiological response to first line FOLFIRI and cetuximab, but then progressed. Eligible patients will receive nivolumab 480mg and relatlimab 160mg every 4 weeks. The primary endpoint is disease control rate (DCR) at 6 months from treatment initiation. With an A’Hern single stage design for efficacy, 5% significance and a power of 80%, 25 patients would need to be treated and a minimum of 6 would need to be progression free at 6 months in order to support an increase of the DCR at 6 months from <10% to 30% or more. Secondary endpoints include DCR at 12 and 24 months, duration of disease control, best objective response rate at 6, 12 and 24 months, progression free survival and overall survival. Exploratory objectives include investigating dynamic changes in the gut microbiota/metabonome, immune infiltrates, immune checkpoints and molecular tumour characteristics induced by LAG3 and PDL1 blockade. This will be achieved through sequential collection of stool and tumour biopsies and baseline tumour characteristics will be correlated with tumour response and survival. iSCORE is a single-centre phase II trial. Recruitment opened in March 2019 and 25 patients will be recruited in 36 months. Clinical trial identification NCT03867799. Legal entity responsible for the study The Royal Marsden NHS Foundation Trust. Funding The Royal Marsden NHS Foundation Trust, Bristol-Myers Squibb. Disclosure D. Cunningham: Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): Celgene; Research grant / Funding (institution): MedImmune; Research grant / Funding (institution): Bayer; Research grant / Funding (institution): 4SC; Research grant / Funding (institution): Clovis; Research grant / Funding (institution): Eli Lilly; Research grant / Funding (institution): Janssen; Research grant / Funding (institution): Merck. I. Chau: Advisory / Consultancy: Eli Lilly; Advisory / Consultancy: BMS; Advisory / Consultancy: MSD; Advisory / Consultancy: Bayer; Advisory / Consultancy: Roche; Advisory / Consultancy: Merck-Serono; Advisory / Consultancy: Five Prime Therapeutics; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Oncologie International; Advisory / Consultancy: Pierre Fabre; Research grant / Funding (institution): Eli Lilly; Research grant / Funding (institution): Janssen-Cilag; Research grant / Funding (institution): Sanofi Oncology; Research grant / Funding (institution): Merck-Serono; Honoraria (self): Eli-Lilly. M. Gerlinger: Research grant / Funding (institution): BMS; Research grant / Funding (institution): Merck KG. N. Starling: Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): BMS; Research grant / Funding (institution): Pfizer; Travel / Accommodation / Expenses: AstraZeneca; Travel / Accommodation / Expenses: BMS; Travel / Accommodation / Expenses: Eli Lilly; Travel / Accommodation / Expenses: Merck; Travel / Accommodation / Expenses: Roche; Honoraria (self): AstraZeneca; Honoraria (self): Eli Lilly; Honoraria (self): Merck; Honoraria (self): Servier; Advisory / Consultancy: Pfizer; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Servier. All other authors have declared no conflicts of interest.

Background ICONIC is a single arm phase 2 trial investigating the safety and efficacy of 4 cycles pre-operative and 4 cycles post-operative FLOT-A in resectable OGA. We report results from the safety run-in phase and early translational biomarker data. Methods Eligible pts were enrolled into the 3+3 design dose finding stage. Standard dose FLOT was administered with 10mg/kg iv avelumab q2 weeks (dose level 0). Dose limiting toxicities (DLTs) were assessed for 28d. Biopsies were taken at baseline and post cycle 2. Results At data cut-off (12/4/19) 6 pts were enrolled and completed pre-operative treatment. 1/6 pts experienced a DLT (chest pain during 5FU infusion), and dose level 0 was established as the safe dose for the efficacy stage of the trial. During pre-operative FLOT-A all pts experienced at least one grade 1-2 adverse event (AE), most commonly diarrhoea (5/6 pts), fatigue, nausea, peripheral neuropathy and hypokalaemia (all 4/6 pts); 3/6 pts experienced at least one grade 3-4 AE: neutropenia and elevated liver enzymes (1pt), thrombotic event (1pt) and cardiac chest pain (1pt). 4/6 pts reported any grade chest pain and underwent cardiac work up: 1 episode was due to PE, 1 of likely GI origin and 2 cardiac in nature, which then recurred in one pt who was re-exposed to FLOT without avelumab. 3/6 pts completed 4 cycles pre-operative FLOT-A: 1pt discontinued avelumab due to diarrhoea, 2 pts discontinued FLOT-A due to cardiac chest pain and switched to a regimen without 5FU. 5 pts have undergone surgery at data cut-off, without unexpected complications. Immunofluorescence shows changes of CD8-, memory- and regulatory T cell infiltrates between baseline and on-treatment biopsies, and detailed results will be presented. Conclusions This is the first data showing that FLOT can be combined with a PD-L1-inhibitor with FLOT-A having a manageable safety profile at dose level 0 (standard dose FLOT + 10mg/kg avelumab). As chest pain of variable aetiologies was observed in 4/6 patients this will be closely monitored and assessed during the efficacy phase. No unexpected complications have been observed during surgery following FLOT-A treatment. Clinical trial identification 2016-003306-13. Legal entity responsible for the study Royal Marsden Hospitals NHS Foundation Trust. Funding Royal Marsden Hospital NHS Foundation Trust Institute of Cancer Research, Merck Pharmaceuticals. Disclosure M. Davidson: Travel / Accommodation / Expenses: Celgene. N. Starling: Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): BMS; Research grant / Funding (institution): Merck; Honoraria (institution): AstraZeneca. I. Chau: Advisory / Consultancy: Eli-Lilly; Advisory / Consultancy: BMS; Advisory / Consultancy: MSD; Advisory / Consultancy: Bayer; Advisory / Consultancy: Roche; Advisory / Consultancy: Merck-Serono; Advisory / Consultancy: Five Prime Therapeutics; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Oncologie International; Advisory / Consultancy: Pierre Fabre; Research grant / Funding (institution): Eli-Lilly; Research grant / Funding (institution): Janssen Cilag; Research grant / Funding (institution): Sanofi Oncology; Research grant / Funding (institution): Merck-Serono; Honoraria (institution): Eli-Lilly. D. Cunningham: Research grant / Funding (institution): Amgen; Research grant / Funding (institution): AstraZeneca; Research grant / Funding (institution): Bayer; Research grant / Funding (institution): Celgene; Research grant / Funding (institution): Merck-Serono; Research grant / Funding (institution): Medimmune; Research grant / Funding (institution): Merrimack; Research grant / Funding (institution): Novartis; Research grant / Funding (institution): Roche; Research grant / Funding (institution): Sanofi. D. Morganstein: Advisory / Consultancy: MSD; Advisory / Consultancy: BMS; Advisory / Consultancy: Roche. M.D. Forster: Research grant / Funding (institution): Merck; Honoraria (institution): Merck. M. Gerlinger: Research grant / Funding (institution): Merck; Research grant / Funding (institution): BMS. All other authors have declared no conflicts of interest.

Purpose: Leptomeningeal disease (LMD) is an uncommon complication of advanced breast cancer. The prognosis is poor, and although radiotherapy (RT), systemic and intra-thecal (IT) chemotherapy are accepted treatment modalities, efficacy data are limited. This study was designed to evaluate potential predictors of survival in this patient group. Methods: Breast cancer patients with LMD diagnosed by MRI in a 10-year period (2004-2014) were identified from electronic patient records. PFS and OS estimates were calculated using Kaplan-Meier method, with planned sub-group analysis by treatment modality. Cox regression was employed to identify significant prognostic variables. Results: We identified 182 eligible patients; all female, median age at LMD diagnosis 52.5 years (range 23-80). Ninety patients (49.5%) were ER positive/HER2 negative; 48 (26.4%) were HER2 positive, and 27 (14.8%) were triple negative. HER2 status was unknown in 17 (9.3%). Initial management of LMD was most commonly whole or partial brain RT in 62 (34.1%), systemic therapy in 45 (24.7%) or supportive care alone in 37 (20.3%). Fourteen patients (7.7%) underwent IT chemotherapy, of whom two also received IT trastuzumab. From diagnosis of LMD, the median PFS was 3.9 months (95% CI 3.2-5.0) and median OS was 5.4 months (95% CI 4.2-6.6). Patients treated with systemic therapy had the longest OS (median 8.8 months, 95% CI 5.5-11.1), compared to RT; 6.1 months (95% CI 4.2-7.9 months), IT therapy; 2.9 months (95% CI 1.2-5.8) and supportive care; 1.7 months (95% CI 0.9-3.0). On multivariable analysis, triple negative histology, concomitant brain metastases, and LMD involving both the brain and spinal cord were associated with poor OS. Conclusions: Breast cancer patients with triple negative LMD, concomitant brain metastases or LMD affecting both the spine and brain have the poorest prognosis. Clinical trials to identify more effective treatments for these patients are urgently needed. (C) 2017 Elsevier Ltd. All rights reserved.

Background: Treatment with immune checkpoint inhibitors (ICPi) has greatly improved survival for patients with advanced melanoma in recent years. Anti-CTLA-4 and anti-PD1 antibodies have been approved following large Phase III trials. Immunerelated neurological toxicity of varying severity has been reported in the literature. The cumulative incidence of neurotoxicity among ipilimumab, nivolumab and pembrolizumab is reported as< 1% in published clinical trials. We aimed to identify the incidence of neurotoxicity in our institution across anti-CTLA4 and anti-PD-1 antibodies, including the combination of ipilimumab with nivolumab. We also review the existing literature and propose an investigation and management algorithm.Methods: All patients with advanced melanoma treated with ipilimumab, nivolumab, pembrolizumab or the combination of ipilimumab and nivolumab (ipi_nivo), managed at the Royal Marsden Hospital between September 2010 and December 2015, including patients on (published) clinical trials were included. Medical records for each patient were reviewed and information on neurotoxicity recorded. A systematic search strategy was performed to collate existing reports of neurological toxicity.Results: In total, 413 immunotherapy treatment episodes in 352 patients were included, with median follow-up of 26.7 months. Ten cases of neurotoxicity were recorded, affecting 2.8% of patients overall, ranging from grade 1 to 4, affecting both central and peripheral nervous systems. A rate of 14% was noted with ipi_nivo. Three of five patients commenced on corticosteroids responded to these. Six patients had made a full recovery at the time of reporting. A favorable radiological response was found in 7 of the 10 cases. Unusual presentations are described in detail.Conclusions: Neurological toxicity is not uncommon, and may be more frequent in patients treated with combination ipinivo. Patterns of presentation and response to treatment are varied. A prompt and considered approach is required to optimize outcomes in this group of patients.

While treatment of localised AC is well established, there is paucity of data to inform the management of patient (pts) with advanced tumours. Thus we have retrospectively analysed treatment pathways and outcomes of a single institution series of advanced AC pts. Inclusion criteria included epidermoid histology, inoperable locally recurrent or metastatic disease and availability of full medical records. The primary objective was overall survival (OS). Secondary objectives included objective response rate (ORR) and progression-free survival (PFS). Prognostic factors were analysed in a univariate model. From 1997 to 2014, 64 pts were seen at The Royal Marsden NHS Foundation Trust who met the eligibility criteria. Pt characteristics were: females (60.9%), median age 59.2 (IQR: 52.1-66.4), history of HIV infection (7.8%), squamous histology (90.6%), metastatic disease (75%), median time to advanced disease 9.1 months (m) (IQR: 4.1-20.9), prior pelvic radiotherapy (82.8%), prior salvage surgery (15.6%). 51 pts (79.7%) received ≥1 line of systemic CT. Of these, 37% also underwent multimodality treatment including surgery, chemoradiotherapy or radiofrequency ablation. PF was the most common regimen prescribed in the first-line setting (74.5%) with an ORR of 34.4% (95% CI: 18.6-53.2). PTX-based CT (single agent or combination therapy with carboplatin) was used in 15 pts as either front line or salvage treatment and the overall ORR was 53.3% (95% CI: 26.6-78.7). Median PFS after first- and second-line CT was 5.8m (IQR: 2.8-7.6) and 3.2m (IQR: 2.5-7.1), respectively. Median OS in CT-treated pts was 15.4m (IQR: 10.0-45.2) and 13% were alive at 5 years. Age ≤65 years and liver metastases were predictive of better PFS (HR 0.39; 95% CI: 0.16-0.97, p = 0.04) and worse OS (HR 2.25; 95% CI: 1.25-4.03, p = 0.01), respectively. This is the second largest series of advanced AC ever reported. Doublet CT with PF and PTX-based CT are active regimens in this setting. Prospective clinical trials are needed to standardise treatment pathways, investigate the potential of novel therapeutics and ultimately improve the modest survival outcome of this pt population.

Although the potential of detecting ctDNA in solid tumours has been increasingly reported, there are limited data in LARC. We sought to investigate frequency and relevance of KRAS mutations in ctDNA in a randomised phase II trial of CAPOX followed by chemoradiotherapy, surgery and adjuvant CAPOX ± cetuximab in high-risk LARC. RAS (exon 2-4) mutations were previously analysed in the biopsy and resection samples using standard sequencing techniques. ctDNA was isolated from 2 ml of plasma collected prior to treatment start and analysed by digital droplet PCR. Commercially available and validated assays were used to detect KRAS mutations (G12D, G12V and G13D in all pts plus any patient-specific, additional mutation previously detected in the tissue). The sensitivity cut-off for the assay of ctDNA was set at a lower limit of 0.02% mutant alleles. 97/164 study pts (59%) were assessable for ctDNA. G12D, G12V or G13D tissue mutations were previously identified in 28 pts (7 not assessable). KRAS mutations in ctDNA in these codons were found in 13/28 (46%) and 22/62 (35%) of pts who were KRAS mutant and wild type in tissue, respectively. 5/10 pts with G12A, G12C, G12S or A146T tissue mutations had the same mutation in the blood. Among 38 pts with any KRAS tissue mutation, ctDNA was detected in 18 pts (47%) and associated with a higher baseline T stage (p = 0.01). However, no association was found with complete response (CR) (16.7% vs 10.0%, p = 0.65), PFS (HR 0.86, 95% CI: 0.31-2.37), p = 0.77) or OS (HR 0.92, 95% CI: 0.32-2.65, p = 0.88). When tissue and ctDNA mutation data were combined to redefine the mutation status of the assessable EXPERT-C study population (n = 119; 32 RAS wild-type and 87 RAS mutant), no interaction was found between RAS status and cetuximab treatment with regards to CR (p = 0.99), PFS (p = 0.57) or OS (p = 0.98). In this series of high-risk LARC, KRAS mutations in ctDNA were found in up to half and one third of pts with KRAS mutant and KRAS wild-type tumours, respectively, as defined by previous tissue mutation analyses. Larger studies are needed to better address the potential role of ctDNA as a prognostic or predictive tool in LARC.

Abstract Background: Patients diagnosed with primary breast cancer (BC) often have a couple weeks interval between diagnosis and definitive surgery. This time window provides the opportunity for assessing biological drug effects in a treatment naive population. The EPHOS-B trial was designed to measure the effect of pre-operative anti-HER2 therapy on proliferation and apoptosis in HER2+ BC patients. Patients & methods: EPHOS-B is a multicentre, 2-part randomized trial in patients with operable newly diagnosed HER2+ primary BC. In Part 1 patients were randomized (1:2:2) to no perioperative treatment (control), trastuzumab only or lapatinib only (11 days pre-operative therapy). Emerging evidence on the efficacy and safety of combination anti-HER2 therapy led to Part 2 in which patients were allocated to control, perioperative trastuzumab only or lapatinib and trastuzumab (1:1:2). The IDMC have agreed release of data from lapatinib Part 1 patients only. Tissue samples were taken at the time of diagnostic core biopsy and surgery, and analysed centrally for Ki67, apoptosis (activated caspase 3), PgR, HER3 and Bcl2 by immunohistochemistry (IHC). Local ER and PgR status were also recorded. Primary endpoint is change in Ki67 and/or apoptosis. Response is defined by a drop in Ki67 of ≥30% or a rise in apoptosis of ≥30% from baseline. Results: Between Nov-2010 and Jul-2013, 51 patients (pts) were allocated to perioperative lapatinib, with 49 (96.1%) receiving at least 1 dose. All pts were HER2+ (90% 3+ by IHC and 10% amplified by FISH, locally assessed) at entry. Median age was 51 years (IQR 48-60); 65% had tumours >2cm and 51% were grade 3 at surgery. According to local assessment, 61% were ER+ and 43% PgR+. Only 2 pts (4%) had a dose reduction and 1 pt (2%) discontinued lapatinib in the 3 days prior to surgery due to toxicity (rash, 3pts; nausea, 1pt). There were no delays in surgery. Paired samples were valid for analysis in 43/51 (84%); invalid pairs were mostly due to inadequate samples for scoring. Overall, 67% (95% CI: 52% to 81%) of pts demonstrated a ≥30% fall in Ki67 whilst a ≥30% rise in apoptosis was observed only in 30% of pts (95% CI: 17% to 46%). When assessed as continuous variables, Ki67 fell significantly from pre-treatment but there was no significant change in apoptosis detected (results in table). No correlation was observed between Ki67 change and change in apoptosis (p=0.5). Neither HER-3 expression nor BCL2 predicted response. ER- HER2+, n=18ER+ PR- HER2+, n=7ER+PR+HER2+, n=17All, n=43*Ki67 % change from pre treatment-51% (-69% to -21%), p<0.001-53% (-78% to -11%), p=0.02-38% (-58% to -24%), p<0.001-45% (-57% to -32%), p<0.001Apoptosis % change from pre treatment-24% (-37% to +30%), p=0.27-13% (-48% to +125%), p=0.61-13% (-65% to +43%), p=0.19 Median % change (95% confidence interval), p-value: Wilcoxon signed-rank test. *1 pt missing ER status Conclusion: EPHOS-B demonstrates that ∼11 days' lapatinib has a marked anti-proliferative effect in HER2+ve breast cancers. The trial is ongoing and when complete will provide a definitive analysis of the relative biological effects of perioperative treatment with different anti-HER2 therapies (trastuzumab, lapatinib and their combination) in patients with HER2+ BC. Citation Format: Bundred N, Cameron D, Kalaitzaki E, Morley R, Cramer A, Webster-Smith M, Narayanan S, Brunt M, Horgan K, Hanby A, Ooi J, Hong A, Naik J, Evans A, Shaaban A, Bliss J. Effects of perioperative lapatinib in early HER2+ breast cancer - The UK EPHOS-B trial (CRUK/08/002). [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr PD5-06.

Background: Patients diagnosed with primary breast cancer (BC) often have a couple weeks interval between diagnosis and definitive surgery. This time window provides the opportunity for assessing biological drug effects in a treatment naive population. The EPHOS-B trial was designed to measure the effect of pre-operative anti-HER2 therapy on proliferation and apoptosis in HER2+ BC patients. Patients u0026 methods: EPHOS-B is a multicentre, 2-part randomized trial in patients with operable newly diagnosed HER2+ primary BC. In Part 1 patients were randomized (1:2:2) to no perioperative treatment (control), trastuzumab only or lapatinib only (11 days pre-operative therapy). Emerging evidence on the efficacy and safety of combination anti-HER2 therapy led to Part 2 in which patients were allocated to control, perioperative trastuzumab only or lapatinib and trastuzumab (1:1:2). The IDMC have agreed release of data from lapatinib Part 1 patients only. Tissue samples were taken at the time of diagnostic core biopsy and surgery, and analysed centrally for Ki67, apoptosis (activated caspase 3), PgR, HER3 and Bcl2 by immunohistochemistry (IHC). Local ER and PgR status were also recorded. Primary endpoint is change in Ki67 and/or apoptosis. Response is defined by a drop in Ki67 of ≥30% or a rise in apoptosis of ≥30% from baseline. Results: Between Nov-2010 and Jul-2013, 51 patients (pts) were allocated to perioperative lapatinib, with 49 (96.1%) receiving at least 1 dose. All pts were HER2+ (90% 3+ by IHC and 10% amplified by FISH, locally assessed) at entry. Median age was 51 years (IQR 48-60); 65% had tumours u003e2cm and 51% were grade 3 at surgery. According to local assessment, 61% were ER+ and 43% PgR+. Only 2 pts (4%) had a dose reduction and 1 pt (2%) discontinued lapatinib in the 3 days prior to surgery due to toxicity (rash, 3pts; nausea, 1pt). There were no delays in surgery. Paired samples were valid for analysis in 43/51 (84%); invalid pairs were mostly due to inadequate samples for scoring. Overall, 67% (95% CI: 52% to 81%) of pts demonstrated a ≥30% fall in Ki67 whilst a ≥30% rise in apoptosis was observed only in 30% of pts (95% CI: 17% to 46%). When assessed as continuous variables, Ki67 fell significantly from pre-treatment but there was no significant change in apoptosis detected (results in table). No correlation was observed between Ki67 change and change in apoptosis (p=0.5). Neither HER-3 expression nor BCL2 predicted response. Conclusion: EPHOS-B demonstrates that ∼11 days9 lapatinib has a marked anti-proliferative effect in HER2+ve breast cancers. The trial is ongoing and when complete will provide a definitive analysis of the relative biological effects of perioperative treatment with different anti-HER2 therapies (trastuzumab, lapatinib and their combination) in patients with HER2+ BC. Citation Format: Bundred N, Cameron D, Kalaitzaki E, Morley R, Cramer A, Webster-Smith M, Narayanan S, Brunt M, Horgan K, Hanby A, Ooi J, Hong A, Naik J, Evans A, Shaaban A, Bliss J. Effects of perioperative lapatinib in early HER2+ breast cancer - The UK EPHOS-B trial (CRUK/08/002). [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr PD5-06.

BACKGROUNDIntergroup Exemestane Study (IES) was a randomised study that showed a survival benefit of switching adjuvant endocrine therapy after 2-3 years from tamoxifen to exemestane. PathIES aimed to assess the potential prognostic and predictive value of ERβ1 and ERβ2 expression in primary tumours in order to determine benefit in the two treatment arms.PATIENTS AND METHODSPrimary tumour samples were available for 1256 patients (27% IES population). ERβ1 and ERβ2 expression was dichotomised at the median IHC score (high if ERβ1 ≥ 191, ERβ2 ≥ 164). Hazard ratios (HRs) were estimated by multivariable Cox proportional hazards models adjusting for clinicopathological factors. Treatment effects with biomarker expressions were determined by interaction tests. Analysis explored effects of markers both as a continuous variable and with dichotomised cut-offs.RESULTSNeither ERβ1 nor ERβ2 were associated with disease-free survival (DFS) or overall survival (OS) in the whole cohort. In patients treated with continued tamoxifen, high ERβ1 expression compared with low was associated with better DFS [HR = 0.38:95% confidence interval (CI) 0.21-0.68, P = 0.001]. DFS benefit of exemestane over tamoxifen (HR = 0.40:95% CI 0.22-0.70) was found in the low ERβ1 subgroup (interaction P = 0.01). No significant difference with treatment was observed for ERβ2 expression in either DFS or OS.CONCLUSIONIn the PathIES population, exemestane appeared to be superior to tamoxifen among patients with low ERβ1 expression but not in those with high ERβ1 expression. This is the first trial of its kind to report a parameter potentially predicting benefit of an aromatase inhibitor when compared with tamoxifen and an independent validation is warranted.

IES, a randomised study in 4724 postmenopausal patients (pts) with ER + /unknown early breast cancer, showed a survival benefit from switching adjuvant endocrine therapy [2-3 years tamoxifen (T) to exemestane (E)]. PathIES a retrospective translational study assessed if biomarkers can predict response to continuing on T or switching to E. Central assessment of ER, PR, HER2, Ki67 & ERβ1 was conducted on 1256 FFPE primary tumour samples (27% pts). Biomarker expression was measured by IHC. HER2 was confirmed by FISH if ≥IHC2+. ERβ1 was dichotomised at the median. Tumours were deemed positive for ER/PR if IHC ≥1% or Allred ≥3 & for HER2 if IHC 3+ or if FISH amplified. Ki67 was high if > 11% LI (median). Primary endpoints were DFS & OS. Proportional hazards models & interaction tests explored the differential effects of each marker with treatment (trt). Time-to-distant recurrence was used to test the prognostic value of IHC4. Characteristics & trt allocation were comparable to those pts without tumour sample apart from prior chemotherapy (CT) & HRT use. 19% of pts with tumour sample and 37% of pts without had prior CT [p < 0.001], whilst 31% of pts with tumour sample and 14% of pts without had received HRT [p < 0.001]. High IHC4 score was associated with worse prognosis in T & E [2nd tertile v 1st tertile HR 1.4 95%CI (0.7-2.9), 3rd tertile v 1st tertile HR 2.3 95%CI (1.1-4.7), p = 0.04]. Addition of clinical variables to IHC4 made the effect more profound. When adjusting for nodal status, age group, prior CT & HRT use, histological type & grade, HER2, ER, PR & Ki67, high ERβ1 expression was predictive for continued T benefit [DFS HR 0.4 95%CI(0.2-0.7) interaction p = 0.01, OS HR 0.4 95%CI(0.2-0.7) interaction p = 0.004]. High PR expression was associated with significant improvement in DFS for T [HR 0.3 95%CI (0.1-0.7)]. High Ki67 expression was associated with worse prognosis in T & E [HR 1.7 95%CI (1.1-2.7)]. PathIES confirmed the prognostic value of IHC4. High ERβ1/PR expression was associated with improved DFS & OS for T. Results suggest pts with high ERβ1/PR expression should receive 5 years of T rather than switch to E. As this is the first time ERβ1 was examined in the context of a trial, results require independent validation.Disclosure: J. Bliss: Grant income from PfizerR.C. Coombes: Grant income from PfizerAll other authors have declared no conflicts of interest. IES, a randomised study in 4724 postmenopausal patients (pts) with ER + /unknown early breast cancer, showed a survival benefit from switching adjuvant endocrine therapy [2-3 years tamoxifen (T) to exemestane (E)]. PathIES a retrospective translational study assessed if biomarkers can predict response to continuing on T or switching to E. Central assessment of ER, PR, HER2, Ki67 & ERβ1 was conducted on 1256 FFPE primary tumour samples (27% pts). Biomarker expression was measured by IHC. HER2 was confirmed by FISH if ≥IHC2+. ERβ1 was dichotomised at the median. Tumours were deemed positive for ER/PR if IHC ≥1% or Allred ≥3 & for HER2 if IHC 3+ or if FISH amplified. Ki67 was high if > 11% LI (median). Primary endpoints were DFS & OS. Proportional hazards models & interaction tests explored the differential effects of each marker with treatment (trt). Time-to-distant recurrence was used to test the prognostic value of IHC4. Characteristics & trt allocation were comparable to those pts without tumour sample apart from prior chemotherapy (CT) & HRT use. 19% of pts with tumour sample and 37% of pts without had prior CT [p < 0.001], whilst 31% of pts with tumour sample and 14% of pts without had received HRT [p < 0.001]. High IHC4 score was associated with worse prognosis in T & E [2nd tertile v 1st tertile HR 1.4 95%CI (0.7-2.9), 3rd tertile v 1st tertile HR 2.3 95%CI (1.1-4.7), p = 0.04]. Addition of clinical variables to IHC4 made the effect more profound. When adjusting for nodal status, age group, prior CT & HRT use, histological type & grade, HER2, ER, PR & Ki67, high ERβ1 expression was predictive for continued T benefit [DFS HR 0.4 95%CI(0.2-0.7) interaction p = 0.01, OS HR 0.4 95%CI(0.2-0.7) interaction p = 0.004]. High PR expression was associated with significant improvement in DFS for T [HR 0.3 95%CI (0.1-0.7)]. High Ki67 expression was associated with worse prognosis in T & E [HR 1.7 95%CI (1.1-2.7)]. PathIES confirmed the prognostic value of IHC4. High ERβ1/PR expression was associated with improved DFS & OS for T. Results suggest pts with high ERβ1/PR expression should receive 5 years of T rather than switch to E. As this is the first time ERβ1 was examined in the context of a trial, results require independent validation. Disclosure: J. Bliss: Grant income from Pfizer R.C. Coombes: Grant income from Pfizer All other authors have declared no conflicts of interest.

Abstract Background : Plausible underlying treatment effects were investigated in exploratory analysis of the translational component of TACT, a large adjuvant chemotherapy trial (Lancet 2009:373, 1681). Some breast cancer patients are cured by primary therapy and are therefore not at risk of relapse, potentially confounding treatment comparisons using Cox regression, exploratory comparison of outcomes can be undertaken using parametric cure models (PCM) which estimate the proportion of patients cured and the pattern of events in patients not cured. Methods : TACT compared sequential anthracycline-taxane (FEC-D) adjuvant chemotherapy with an anthracycline regimen of similar duration. As part of TransTACT ER Allred scores were assessed centrally on two cores per patient by two pathologists (CG and SP). PCM was implemented using the CUREREGR command in Stata (Buxton A 2007), a lognormal model for time to relapse being fitted for the category of patients who were not cured. The primary endpoint for the analyses was relapse free interval. Results : 2892 patients had Estrogen Receptor (ER) assessed centrally, 1851 tumors were ER positive and 1041 were ER negative, median follow up was 8 years. The proportion of patients cured was related to established prognostic characteristics such as nodal status. In unselected ER+ patients evidence of cure was weak but it could not be ruled out in some subgroups e.g. in node negative ER+ patients the estimate of the proportion cured was approximately 70% (p = ns). In unselected ER- patients the pattern of relapse implied by PCM was found to differ significantly between taxane treated and control patients (p<0.0001), despite the fact there was no overall evidence of a treatment effect (Hazard ratio 0.98, p = 0.71 by Cox Regression). In taxane treated patients the estimate of the proportion of patients cured was 65% (95%CI:60-69), and the estimated median time to relapse in those not cured was 1.9 years (95%CI:1.7-2.1), in the control group the estimates were 45% (95%CI:29-62) cured and in patents not cured median time to relapse 4.2 years (95%CI:2.1-8.4). Conclusions : The concept of cure is of particular interest in tumour subtypes, such as ER negative tumours, in which event rates are high in early years (1-4) but low subsequently. In ER- patients PCM suggested the model-defined cure rate was about 20% higher in docetaxel treated patients, however this was counterbalanced by worse prognosis in the patients without model defined cure. EBCTCG results examining the taxane effect on recurrence were not directly available to support this finding but breast cancer mortality curves in more than 35,000 women with follow up to 8 years (Lancet 379, 432 Webappendix, p15) showed patterns in ER- patients that were consistent with a higher cure rate in taxane treated patients, in addition there was little evidence of cure in unselected ER+ patients. In this study PCM identified differences in the treatment effect between the randomised groups despite the fact there was no evidence of an overall difference based on the hazard ratio by Cox regression, suggesting PCM may have a useful role in exploratory analysis to look for treatment effects in situations where cure is a possibility. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P6-06-03.