1012 Background: In ML-3, ribociclib (RIB) demonstrated a significant PFS/OS benefit over placebo (PBO) in pts with HR+/HER2− ABC. We report associations of on-tx ∆ctDNA with OS/PFS and tumor size and of ctDNA detectability with radiological progression. Methods: Pts were randomized 2:1 to RIB or PBO + fulvestrant as 1st- or 2nd-line tx. ctDNA levels were measured by NGS at cycle 1 day 1 (C1D1), C4D1, C8D1, and every 8 cycles until end of tx. Analyses included: Wilcoxon rank sum text to compare ∆ctDNA from C1D1 to C4D1; Kaplan-Meier method to visualize relationship of ctDNA with PFS/OS; Cox proportional hazards model for associations of ctDNA and tumor metrics (objective response rate, best % change, % change at 1st tumor evaluation); Fisher’s exact test to compare ctDNA detectability between arms; and ∆ctDNA across time points to assess lead time between ctDNA level-based and radiological progression. Results: Overall, 534/726 (73.6%) pts had paired C1/C4 measurements. Rates of detectable (+) and undetectable (−) ctDNA at C1 and C4 were: C1+/C4+, 33.0%; C1+/C4−, 34.5%; C1−/C4−, 27.5%; and C1−/C4+, 5.1%. RIB vs PBO had a larger % decrease in ctDNA level from C1 to C4 (P=0.003). Pts with better best overall response also had larger ctDNA decreases from C1 to C4 (P<0.0001). On-tx ∆ctDNA from C1 to C4 correlated with PFS/OS (P<0.0001; Table). C1-/C4-, C1+/C4-, and C1+/C4+ strongly correlated with PFS within response categories (PR, SD, NCRNPD). C1-to-C4 % ∆ctDNA was only moderately correlated with tumor size change (best % change, r=0.41; change at 1st evaluation, r=0.32; both P<0.001). Univariate/multivariate analyses showed significant correlations between ctDNA metrics (% change, detection pattern at C4D1) and PFS/OS. ctDNA detection and % change generally remained significantly associated with PFS/OS even after adjusting for change in tumor size. In a longitudinal analysis of pts who cleared ctDNA at C4D1 (C4–), the median lead time between ctDNA progression and radiological progression was 8.2 mo (IQR 4.5–15.6). Conclusions: On-tx ∆ctDNA from C1 to C4 was correlated with PFS/OS even after adjusting for tumor size changes, suggesting that ctDNA and tumor dynamics capture distinct aspects of tumor tx responses. Decreases in ctDNA were greater with RIB vs PBO, mirroring ML-3 efficacy readouts but at a much earlier timepoint. This longitudinal analysis of ctDNA in ML-3 suggests that detecting ∆ctDNA is a sensitive way to monitor progression and may be a potential surrogate for efficacy in HR+/HER2− ABC. Clinical trial information: NCT02422615 . [Table: see text]

PDF file - 45K, Summary of adverse events suspected to be study-drug related by grade (safety set).

PDF file - 42K, Summary of deaths and adverse events (safety set).

PDF file - 48K, Exposure to buparlisib and trastuzumab (safety set).

PDF file - 67K, Bayesian logistic regression model: dose-response curve and model inference results at the time of determination of the recommended Phase II dose.

2508 Background: DF1001 is a first in class, Tri-specific, NK cell Engager Therapy (TriNKET), that uses HER2 as an anchor to modulate innate and adaptive immunity. DF1001 redirects both NK and CD8 T cells to reprogram the tumor microenvironment, turning cold tumors hot. DF1001-001 (NCT04143711) is an ongoing phase I/II, first in human study investigating the safety, tolerability, clinical and biologic activity of DF1001. The safety, pharmacokinetic (PK), pharmacodynamic (PD) and efficacy data from the phase 1 monotherapy escalation and safety/PK/PD expansion is presented in this report. Methods: The Phase 1 dose escalation of DF1001-001 evaluated doses ranging from 0.000052-15mg/kg in patients with advanced and/or refractory solid tumors. Safety/PK/PD expansions, mandating paired biopsies, enrolled at doses 0.52-15mg/kg for biomarker assessment in patients with at least HER2 1+ expressing tumors. The primary objective of the Phase 1 study was to evaluate the safety of DF1001. The secondary objective was to assess efficacy. Results: As of 8 Dec 2022, 36 patients were treated in the dose escalation and 68 patients were treated in the safety/PK/PD expansion. 63% of the patients were female with a median age of 61 years. Most common tumor types included breast (38%), urothelial carcinoma (10%), esophageal (8%), non-small cell lung (7%) and colorectal (7%) cancer. Treatment Related Adverse Events (TRAE) were reported in 79% of patients. TRAEs were mostly low grade (86% were Grade 1-2) and rarely led to treatment discontinuation. No Grade 4 or 5 TRAEs were reported. The most common TRAEs were infusion related reactions (26%), asthenia (15%) and fatigue (12%). No dose limiting toxicities were detected, and the MTD has not been reached. Best overall response (BoR) of Partial Response (PR) by RECIST 1.1 was observed in 5 patients, with duration of up to 13 months, and 22 patients experienced stable disease (SD), for a Clinical Benefit Rate (CBR) of 39.7%. Responders included patients with refractory lung and breast cancer, whose tumors expressed HER2 1+ to 3+ by IHC. Biomarker analysis in responders demonstrated increases in inflammatory cytokines and chemokines by gene expression profiling, and consonant increases in NK and CD8 cells in the tumor microenvironment, consistent with TriNKET mechanism of action (MOA). Conclusions: In this dose escalation study, DF1001 is shown to be a safe and well-tolerated novel NK cell engager therapy targeting HER2. DF1001 monotherapy demonstrates promising anti-tumor activity in treatment refractory patients with a spectrum of HER2 expression as well as MOA consistent pharmacodynamic changes in the tumor. This promising data set supports further exploration of DF1001 alone, and in combination with other compounds, as a potential novel therapeutic option for patients with solid tumors. Clinical trial information: NCT04143711 .

PDF file - 38K, Patient disposition (full analysis set).

PDF file - 65K, Criteria for defining dose-limiting toxicities.

PDF file - 62K, Summary of buparlisib pharmacokinetic parameters.

TPS1112 Background: The overexpression and/or amplification of human epidermal growth factor receptor 2 (HER2) occurs in approximately 20% of breast cancers (BC) and is a major driver of tumor development and progression. This HER2 subtype confers aggressive tumor behavior and the HER2 receptor remains a valuable target for antibodies, bi-specifics, and antibody drug conjugates (ADC). With advances in targeted therapy, patients with HER2-positive breast cancer (HER2+ BC) may experience an improved prognosis, including survival. Novel HER2-targeted therapies are being investigated to overcome drug resistance and to help mitigate adverse events (e.g., cardiotoxicity). ARX788 is a next-generation ADC using a technology platform whereby a HER2 specific monoclonal antibody is conjugated with Amberstatin269 (AS269), a potent cytotoxic tubulin inhibitor. Site-specificity, high homogeneity, and stable covalent conjugation of ARX788 leads to its slow release and prolonged peak of serum pAF-AS269, which may contribute to the lower systemic toxicity and increased targeted delivery of payload to tumor cells at a lower effective dose compared to other HER2 ADCs. Clinical activity has been seen in Phase I HER2 breast and pan-tumor studies. Methods: Trial Design: ACE-Breast-03 (NCT04829604) is a global, phase 2 study designed to assess anticancer activity and safety of ARX788 in patients with metastatic HER2 positive breast cancer. Patients whose disease is resistant or refractory to T-DM1, and/or T-DXd, and/or tucatinib-containing regimens are eligible. Patients must have adequate organ function. Any brain metastases must be radiographically stable without steroid dependence. Efficacy will be assessed using Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 by imaging every 6 weeks on study. Endpoints include objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), overall survival (OS), best overall response (BOR), duration of response (DOR), and time to response (TTR). The safety and tolerability profile will be evaluated. Blood samples will be collected at specified time points to determine serum concentrations of ARX788, total antibody, and metabolite pAF-AS269. Potential predictive and/or prognostic biomarkers at baseline and on-treatment will be analyzed for exploratory purposes. Descriptive statistics will be used to evaluate anticancer activity, safety, and tolerability. The study is currently recruiting patients. Please contact breast03trialinquiry@ambrx.com for additional information. Clinical trial information: NCT04829604.

526 Background: Patients (pts) with HER2+ metastatic breast cancer (mBC) are at high risk of developing brain metastases (BMs), and treatment options are limited once BMs occur. T-DXd is an antibody-drug conjugate composed of an anti-HER2 antibody, a cleavable tetrapeptide-based linker, and a topoisomerase I inhibitor payload. On the basis of findings of the phase 2 DESTINY-Breast01 trial (NCT03248492), T-DXd was approved for the treatment of adult pts with HER2+ unresectable or mBC who have received ≥2 prior anti-HER2–based regimens (US and Europe) or had prior chemotherapy and are refractory to or intolerant of standard treatments (Japan). Here we describe a subgroup analysis from DESTINY-Breast01 in pts with a history of BMs. Methods: DESTINY-Breast01 is an ongoing, 2-part, multicenter, open-label, phase 2 trial of T-DXd in adult pts with HER2+ unresectable or mBC previously treated with trastuzumab emtansine. Pts with baseline BMs that were treated, asymptomatic, and did not require therapy to control symptoms were eligible for enrollment. All treatment to control symptoms had to be completed ≥60 days before randomization. An MRI of the brain every 6 wks was only required for pts with BMs at baseline. Brain lesions were considered non-target lesions, and thus collection of diameter measurements was not required. This analysis includes pts with a history of BMs at baseline who received T-DXd at the approved dose of 5.4 mg/kg every 3 wks. Results: Twenty-four pts with a history of BMs were included (data cutoff, August 1, 2019). In these pts, the objective response rate (ORR), median progression-free survival (mPFS), and median duration of response (mDoR) per independent central review with T-DXd 5.4 mg/kg were 58.3% (95% CI, 36.6%-77.9%), 18.1 mo (95% CI, 6.7-18.1 mo), and 16.9 mo (95% CI, 5.7-16.9 mo), respectively, and were comparable to those in the total pt population (N = 184) treated at 5.4 mg/kg in the DESTINY-Breast01 study (ORR, 60.9%; mPFS, 16.4 mo; mDoR, 14.8 mo; median follow-up, 11.1 mo). The pattern of disease progression was similar in pts with and without BMs with 8 of 24 pts having progressed (33%; 2 in the brain) in the BMs subgroup and 40 of 160 pts (25%; 2 in the brain) in the non-BMs subgroup, suggesting durable systemic disease control. Baseline diameters of BMs were available for 14 of 24 pts (radiotherapy prior to study enrollment was reported in 12 of 14). Among the pts with information available on baseline BM diameter, the central nervous system response rate per investigators was 50% (7 of 14 pts). Conclusions: T-DXd showed strong clinical activity in both the overall population of pts with HER2+ mBC and the subgroup of pts with BMs. The demonstrable response of BMs to treatment and durable clinical activity of T-DXd in pts with a history of BMs at baseline are promising and warrant further investigation. Clinical trial information: NCT03248492.

510 Background: Trastuzumab (T) increases the incidence of cardiac events (CEs) in patients (pts) with early breast cancer (BC). Dual blockade with P+T improves BC outcomes and is the standard of care for high-risk HER2-positive BC pts following the phase 3 APHINITY trial that evaluated the addition of P or placebo (Pla) to T and chemotherapy (CT). We analyzed the cardiac safety of P+T in APHINITY. Methods: APHINITY eligibility required a left ventricular ejection fraction (LVEF) ≥55% at study entry. LVEF assessment was performed every 3 months (mos) during treatment, every 6 mos up to month 36, and yearly thereafter. Primary CE was defined as heart failure (HF) class III/IV and a significant decrease in LVEF of at least 10 percentage points from baseline and to <50%, or cardiac death. Secondary CE was defined as a confirmed significant decrease in LVEF or CEs confirmed by the cardiac advisory board. Results: The safety analysis population consists of 4,769 pts. With 74 mos median follow-up (FU), CEs were observed in 159 pts (3.3%): 83 (3.5%) in the P+T and 76 (3.2%) in Pla+T arms, respectively. Most CEs occurred during anti-HER2 therapy: 123/159 (77.4%) and were asymptomatic or mildly symptomatic LVEF decrease (133/159; 83.6%) (Table 1). There were 2 cardiac deaths in each arm (0.1%). More CEs occurred in pts receiving an anthracycline-based CT compared to those receiving non-anthracycline CT (139 vs. 20 CEs, respectively). Acute recovery from a CE based on subsequent LVEF values was observed in 127/155 pts (81.9%). Conclusions: Dual blockade with P+T does not increase the risk of CE compared to Pla+T alone. The use of anthracycline-based CT increases the risk of a CE; hence non-anthracycline CT may be considered particularly in pts with other cardiovascular risk factors. Clinical trial information: NCT01358877. [Table: see text]

1001 Background: The Phase III MONALEESA-3 trial (NCT02422615) previously demonstrated a statistically significant improvement in OS with RIB, a cyclin-dependent kinase 4/6 inhibitor (CDK4/6i), plus FUL compared with placebo (PBO) plus FUL as first-line (1L) or second-line (2L) treatment in postmenopausal pts with HR+/HER2− ABC (median, not reached vs 40.0 mo; hazard ratio [HR], 0.72; 95% CI, 0.57-0.92, P =.00455). This analysis was final per the protocol; following the unblinding of the study, pts still on study treatment in the PBO arm were allowed to cross over to the RIB arm. We report an exploratory analysis of OS after an additional median 16.9 mo of follow-up, allowing for further characterization of long-term survival benefits of RIB. Methods: Postmenopausal pts with HR+/HER2− ABC were randomized 2:1 to receive RIB + FUL or PBO + FUL in 1L and 2L settings. Updated OS was evaluated by Cox proportional hazards model and summarized using Kaplan-Meier methods. Additional postprogression endpoints such as progression-free survival 2 (PFS2), time to chemotherapy (CT), and CT-free survival were also evaluated and summarized. Results: At the data cutoff (Oct 30, 2020), the median follow-up was 56.3 mo (min, 52.7 mo) and 68 (14.0%) and 21 (8.7%) patients were still on treatment in the RIB vs PBO arms, respectively. With this extended follow-up, RIB + FUL continued to demonstrate an OS benefit vs PBO + FUL (median, 53.7 vs 41.5 mo; HR, 0.73; 95% CI, 0.59-0.90). RIB + FUL had prolonged OS vs PBO + FUL in the 1L (median, not reached vs 51.8 mo; HR, 0.64; 95% CI, 0.46-0.88) and 2L subgroups (median, 39.7 vs 33.7 mo; HR, 0.78; 95% CI, 0.59-1.04). Subgroup analyses also showed a consistent OS benefit compared with the intent-to-treat (ITT) population for most subgroups. PFS2, time to CT, and CT-free survival for the ITT population favored RIB + FUL (Table). Among pts who discontinued study treatment, 81.9% and 86.4% received a next-line subsequent antineoplastic therapy, while 14.0% and 30.0% received a CDK4/6i as any subsequent line in the RIB vs PBO arms, respectively. No new safety signals were observed. Conclusions: The previously demonstrated robust and clinically meaningful OS benefit with RIB + FUL compared with PBO + FUL was maintained after almost 5 years of follow-up in postmenopausal pts with HR+/HER2− ABC. The OS benefit of RIB was observed in the 1L and 2L subgroups, which further supports the use of RIB in these populations. The results also demonstrated a significant delay in the use of subsequent CT with RIB vs PBO. Clinical trial information: NCT02422615 .[Table: see text]

3519 Background: An understanding of the mechanisms of acquired resistance to CDK4 & 6 inhibitors, either alone or with endocrine therapy (ET), is an unmet need. Abemaciclib is a CDK4 & 6 inhibitor approved for treatment of HR+, HER2- advanced breast cancer (ABC). Here we evaluated acquired genomic alterations detected in circulating tumor DNA (ctDNA) from patients (pts) treated with abemaciclib + nonsteroidal aromatase inhibitor (AI) or placebo + AI in MONARCH 3 or abemaciclib monotherapy in nextMONARCH 1. Methods: MONARCH 3 randomized postmenopausal women with HR+, HER2- ABC with no prior systemic therapy in the advanced setting to abemaciclib (150 mg Q12H) or placebo + AI. nextMONARCH 1 randomized women with HR+, HER2- metastatic breast cancer who had progressed on or after prior ET and CT to abemaciclib (150 mg Q12H) + tamoxifen, abemaciclib (150 mg Q12H), or abemaciclib (200 mg Q12H) + loperamide. Plasma from pts in the abemaciclib or placebo + AI arms (MONARCH 3) or abemaciclib monotherapy arms (nextMONARCH 1) was analyzed by the Guardant360 assay to identify potential tumor-related genomic alterations including point mutations, indels, amplifications, and fusions acquired at EOT in comparison with baseline. Results: For MONARCH 3, commonly acquired alterations at EOT included ESR1 (17%), TP53 (10%), EGFR (8%), FGFR1 (7%), and PDGFRA (7%) in the abemaciclib + AI arm, and ESR1 (31%), TP53 (10%), and BRCA1 (7%) in the placebo + AI arm. Acquired alterations more frequent for abemaciclib + AI pts included RB1 (6%), MYC (5%), and AR (5%), compared to 0% in the placebo + AI arm (p = 0.008 RB1; p = 0.015 MYC or AR). In contrast, acquired ESR1 alterations were less frequent with abemaciclib + AI vs placebo + AI (17% vs 31%, p = 0.038). In nextMONARCH 1, the most commonly acquired alterations with abemaciclib monotherapy were in TP53 (10%), EGFR (9%), RB1 (9%), MYC (9%), and MET (8%). In addition, acquired alterations in ESR1 (6%) and AR (3%) were also found. PIK3CA alterations were not frequently acquired (abemaciclib + AI 1%, placebo + AI 6%, abemaciclib monotherapy 5%). Conclusions: Acquired genomic alterations potentially associated with emerging mechanisms of resistance to abemaciclib alone or in combination with AI may include RB1, MYC, or AR alterations, while the acquisition of ESR1 alterations was less common in pts treated with abemaciclib + AI compared to placebo + AI. These findings are hypothesis-generating and provide insight into mechanisms of resistance to abemaciclib vs ET. Clinical trial information: NCT02246621, NCT02747004 .

e18014 Background: Concomitant cisplatin-based chemoradiotherapy (CCRT), followed by image guided-adaptive brachytherapy (IGABT) is the recommended treatment for patients suffering from locally advanced cervical cancer (LACC). Methods: Between January 2010 and May 2017, 103 patients with LACC (FIGO 2009-stages IB2-IVA) received CCRT followed by IGABT. The objectives of this study were to evaluate the impact of white blood cells (WBC) and polymorphonuclear neutrophils (PMN) counts variations on outcomes. This variable was calculated by substraction between WBC or PMN levels at the first cycle (CB) and the last cycle (CL) of chemotherapy (CT)(DCB-CL). The data were reviewed retrospectively, with Cox regression for univariate and multivariate analysis. Results: The median age at diagnosis was 50 years. The median tumor size at diagnosis was 47mm. The majority of the patients had FIGO stage II (60.2%) or stage I (21.4%) disease with squamous histology (88.3%). Patients received a median dose of external-beam radiotherapy (EBRT) of 45 Gy (range 40-50.4 Gy) by 1.8-2 Gy fractions, with a median cumulative dose of all the radiotherapy of 85 Gy. The median duration of EBRT+IGABT was 51 days (range 31-94). All patients received at least one cycle of cisplatin, but the majority received 5 (40.4%) or 6 (39.4%) cycles. The median follow-up time for all patients was 30.1 months. The overall survival (OS) and recurrence-free survival (RFS) at 3 years was 81,4% and 76,8% respectively. Univariate analysis associated higher DCB-CL WBC and DCB-CL PMN with better OS and RFS. Multivariate analysis confirmed that DCB-CL WBC (HR, 0.856; 95% CI, 0.737-0.986; p = 0.018) and DCB-CL PMN (HR, 0.863; 95% CI, 0.750-0.994; p = 0.041) were associated with better OS and RFS respectively. A linear regression analysis was performed to cross the DCB-CL WBC/PMN and the number of CT cycles. This analysis reveals that an increasing number of CT cycles is linked to an increased DCB-CL WBC/PMN. Conclusions: Our study reveals the impact of DCB-CL WBC and PMN on outcomes. These two tests could become biomarkers during CCRT to discuss adjuvant treatments, but also to adapt our follow-up.

1050Background: Eftilagimod alpha (efti, previously IMP321) is a recombinant LAG-3Ig fusion protein that binds to MHC class II and mediates antigen-presenting cell (APC) activation followed by CD8 T-cell activation. The activation of the dendritic cell network with efti the day after chemotherapy may lead to stronger anti-tumor CD8 T cell responses. We report final results of the safety run-in of a phase IIb trial (NCT02614833) in patients (pts) with hormone receptor positive MBC receiving weekly paclitaxel as first line chemotherapy. Methods: In the safety run-in phase 15 pts with MBC received paclitaxel (80 mg/m2; D1, D8, D15; IV) in a 4-week cycle in conjunction with either 6 mg (n = 6; cohort 1) or 30 mg (n = 9; cohort 2) efti injections (D2 and D16; SC) for 6 cycles. Pts without progressive disease could continue for a maximum of 12 additional efti injections every 4 weeks. Blood samples for pharmacokinetics and …

169 Background: Single-agent durvalumab is being evaluated in patients with advanced squamous and non-squamous NSCLC in an ongoing Phase 1/2 study (NCT01693562). Here we present updated survival and safety data in NSCLC patients. Methods: Treatment-naïve (1L) and previously treated (2L or 3L+) stage IIIB/IV NSCLC patients received durvalumab 10 mg/kg Q2W for up to 12 months. Patients were stratified by tumor PD-L1 expression (Ventana PD-L1 [SP263] Assay [PD-L1 high: ≥25% of tumor cells with membrane staining]), treatment line, and histology. Results: As of 05 September 2017, 304 NSCLC patients received durvalumab monotherapy. Median duration of follow-up was 35.6 (0.3–50.9) months. Investigator-assessed ORR ranged between 23.2% and 30.0% among PD-L1 high patients, and between 3.6% and 8.3% among PD-L1 low/negative patients. Median PFS and median OS were longer in PD-L1 high vs PD-L1 low/negative patients (Table). Any-grade treatment-related AEs (TRAEs) were reported in 57.2% of pts (including fatigue, 17.4%, decreased appetite, 9.2%, diarrhea, 8.9%); in 10.2% of pts these were Grade 3 or 4. TRAEs resulting in treatment discontinuation were reported in 17 patients (5.6%); 1 patient had a Grade 5 TRAE (pneumonia). Conclusions: In this ongoing phase 1 study, OS and safety profile appear encouraging in treatment-naïve and previously treated NSCLC patients, particularly among PD-L1 high patients. Further investigation regarding PD-L1 expression for selection of patients who most likely benefit from durvalumab is needed. Clinical trial information: NCT01693562. [Table: see text]

517 Background: Women with hormone receptor-positive BC have an ongoing risk of relapse. We performed molecular analyses of primary BC from postmenopausal patients (pts) enrolled in the extended adjuvant intermittent letrozole vs. continuous letrozole (SOLE) trial to identify prognostic factors and potential targets. Methods: From 4884 pts enrolled, 3162 had FFPE tumor samples and were eligible. A case-cohort design selected 599 pts, and 499 DNA samples underwent next-generation sequencing of 35 and 19 actionable genes for mutation (SNV) and copy number gain (CNG) analyses. Correlations of SNV/CNG with clinicopathologic factors were analyzed. Associations with breast cancer free-interval (BCFI) and distant recurrence free-interval (DRFI) were assessed using weighted proportional hazards models to obtain unbiased and consistent estimates representing the overall trial population. With 403 pts there was 90% power to detect a hazard ratio (HR) = 2.0 (SNV/CNG+ vs. -; assuming a 40% mutation prevalence), with 0.05 two-sided significance level and adjusted standard error estimate from the weighed Cox model. Results: SNV and CNG data were available for 403 (81%) and 350 (70%) of 499 samples, respectively. 294 samples had ≥1 SNV or CNG; 132 had concurrent alterations. PIK3CA was the most frequently mutated gene (42%). Other genes showed < 5% SNV rate. Recurrent CNGs were seen in CCND1 (15%), ERBB2 (10%), FGFR1 (8%) and MYC (8%). PIK3CA SNVs and MYC CNG were associated with low (p = 0.032) and high (p = 0.004) tumor grade. Ki-67 index was high in tumors with CNG of CCND1, ERBB2 and MYC (each p < 0.01). No association was seen between PIK3CA SNV and recurrence. FGFR1 CNG was associated with worse outcomes (BCFI: HR = 3.2; 95% CI, 1.5-6.9, p = 0.003; and DRFI: HR = 3.5; 95% CI, 1.6-7.7, p = 0.002). The results were consistent in multivariable models adjusting for clinicopathologic factors. Conclusions: We showed that pts with hormone receptor-positive node-positive BC with FGFR1 CNG had an increased risk of late recurrence despite extended therapy. FGFR1 analysis may improve the risk stratification in this population and represent a potential therapeutic target. Clinical trial information: NCT00553410.

1050 Background: Eftilagimod alpha (efti, previously IMP321) is a recombinant LAG-3Ig fusion protein that binds to MHC class II and mediates antigen-presenting cell (APC) activation followed by CD8 T-cell activation. The activation of the dendritic cell network with efti the day after chemotherapy may lead to stronger anti-tumor CD8 T cell responses. We report final results of the safety run-in of a phase IIb trial (NCT02614833) in patients (pts) with hormone receptor positive MBC receiving weekly paclitaxel as first line chemotherapy. Methods: In the safety run-in phase 15 pts with MBC received paclitaxel (80 mg/m2; D1, D8, D15; IV) in a 4-week cycle in conjunction with either 6 mg (n = 6; cohort 1) or 30 mg (n = 9; cohort 2) efti injections (D2 and D16; SC) for 6 cycles. Pts without progressive disease could continue for a maximum of 12 additional efti injections every 4 weeks. Blood samples for pharmacokinetics and immuno-monitoring were taken in cycle 1, 4 and 6. The primary endpoint was determination of the recommended phase 2 dose of this combination. Results: Between Jan and Oct 2016 15 pts (median age 53 years) were enrolled. A majority (67 %) of pts were pre-treated with hormonal therapy. Nine (67 %) pts had a serious adverse event, out of which 1 was related to paclitaxel (dizziness grade 3) and 1 to efti (cytokine release syndrome grade 1). No grade 4 and 4 grade 3 adverse events (AEs) in 4 pts were related to efti. Grade 1 and 2 injection site reactions were the most common efti related AEs and occurred in 14 pts (93 %). Increased number of circulating monocytes, dendritic cells and CD8 T cells as well as increased cellular activation were observed. This sustained (≥ 6 months) activation of the cellular response was associated with increased Th1 markers (IFN-g, CXCL10) levels in the plasma. Seven pts (47 %) had a partial response according to RECIST 1.1 (mean duration of 9 months). The disease control rate was 87 %. Conclusions: Thirty mg efti SC is the recommended phase 2 dose and is currently investigated in the ongoing phase II part of the study. Efti leads to a steady and sustainable APC and T cell activation. Clinical trial information: NCT02614833.

TPS2606 Background: MEN1309/OBT076 is a DM4-loaded Antibody Drug Conjugate (ADC), directed against CD205/Ly75, a type I transmembrane surface protein belonging to the macrophage mannose receptor family. CD205 is therapeutically relevant, because of its broad expression in multiple cancer types, its emerging role at facilitating metastatic invasion and its effective internalization upon antibody binding. Extensive preclinical evidence demonstrated MEN1309 antitumor activity in vitro, in xenograft and patient-derived xenograft (PDX) models. Methods: The CD205-Shuttle study is a two-step, open-label, multicenter, dose-escalation FIH study. Step 1 involves patients affected by different solid tumors, following an Accelerated Titration Design (ATD) whereby 1 patient per dose cohort is enrolled and the dose doubled at each cohort. If grade ≥2 toxicity is observed the trial design is expected to revert to a 3+3 scheme with a Fibonacci ascending dose scheme. Step 2 is designed to test MEN1309/OBT076 at minus 2 levels of the tolerated doses during Step 1 in patients with Non-Hodgkin-Lymphoma (NHL). Each dose level escalation will be subject to the assessment of the Cohort Review Committee (CRC). CD205-positive patients are selected by IHC staining of archived tumor material. Patient with locally advanced or metastatic solid tumor in progression can be included if failed on ≥2 previous cancer treatments and if no standard therapy is available. The primary endpoint is to identify dose limiting toxicities (DLTs) and maximum tolerated dose (MTD) of MEN1309/OBT076 administered by i.v. infusion once every 21-days. The secondary endpoints includes assessment of pharmacokinetics (PK), immunogenicity, preliminary clinical efficacy of MEN1309/OBT076 and correlation with CD205 expression. Adverse Events (AE) will be graded according to NCI CTCAE v. 4.03. Responses will be evaluated according to RECIST v1.1 and Cheson criteria (2014). Study variables will be presented by dose-cohort and overall using appropriate descriptive statistics. The enrollment began in October 2017 in European sites; up to date 3 cohort levels have been achieved. NCT03403725. NOTE: First and Second author equally contributed to this work Clinical trial information: NCT03403725.