Outcomes for patients with metastatic colorectal cancer (mCRC) have been improved by the identification of biomarkers predictive and prognostic of clinical outcome. The present retrospective analysis was undertaken to assess the utility of key biomarkers and clinical parameters in predicting outcomes in Spanish patients with mCRC. We retrospectively analyzed tumor samples from a series of patients aged > 18 years with mCRC who were treated at the Hospital General Universitario Gregorio Marañón Spain. Real-time polymerase chain reaction was used to detect KRAS, NRAS, BRAF, and PIK3CA mutations. The key outcome of interest was overall survival (OS). Survival curves were estimated using the Kaplan–Meier method and stratified by the variables of greatest clinical interest. Differences were tested using the log-rank test. Median OS in the overall population was 24.4 months. Triple WT patients (WT KRAS, NRAS, and BRAF) and quadruple WT patients (WT KRAS, NRAS, BRAF, and PIK3CA) had significantly better OS than those who did not have triple or quadruple WT tumors. OS was significantly better in patients with left- vs. right-sided tumors, patients with resected primary tumors and metastases vs. those without resection, and patients with isolated hepatic and isolated pulmonary metastases. This retrospective, observational study has confirmed the prognostic value of the location and resection status of the primary tumor and metastases in Spanish patients with mCRC. Triple WT status, in particular, was prognostic in this patient population, with PIK3CA adding to the prognostic value in the quadruple WT population.

Background: BO-112 is a synthetic dsRNA formulated with a cationic carrier, polyethylenimine, which prevents its degradation by nucleases. In melanoma (MEL) mouse models, systemic administration of BO-112 activates MDA-5 and NOXA, leading to anti-tumoral activity connected to a sustained and extended expression of interferon (IFN) response genes. IT BO-112 delivery demonstrated in transplanted mouse models a safer and enhanced local and systemic antitumor effects. Final toxicity and biological profile of the first in human clinical trial that explores IT BO112 as an immune-modulatory treatment is presented (NCT02828098). Methods: Patients (pts) with solid malignant tumors and metastases ≥1 cm accessible to IT injection were treated in 3 cohorts (C) in a 3 + 3 phase I design: C1: single IT BO-112 dose of 0.6 mg; C2: 1mg IT BO112 qw x 2-3 doses. C3: 0.6 mg IT BO112 qw x 2-3 doses. Pre & post treatment biopsies from the injected lesion were obtained, analysing necrosis, apoptosis, immune infiltrate and (first time reported for this trial) IFN-gamma mRNA expression (nanostring). PKs, cytokines and circulating immune cells (CICs) were studied in pre and post BO-112 blood samples.Table: LBA20TUMOR, N (%)BLOOD, N (%)COHORT (N)NECROSIS- APOPTOSISCD8CD4IFN-gammaCICs1 (6)4/5 (80)2/5 (40)4/5 (80)3/5 (60)6/6 (100)2 (7)4/5 (80)2/5 (40)4/5 (80)1/1 (100)6/6 (100)3 (3)3/3 (100)0/3 (0)0/2 (0)NA2/3 (67) Open table in a new tab Results: 16 pts (12 more than the first report in ASCO 2017) were treated: 5 MEL, 3 sarcoma, 2 breast, 1 colorrectal, 1 neuroendocrine, 1 ovarian, 1 head & neck, 1 mesothelioma and 1 cystic adenoid. 50% pts had related AEs, being 94% G1-2 and 6% G3-4 (thrombocytopenia in 2 pts in C1 and C2, both considered DLT and recovered) BO112 was not detected in blood in any pts. Table summarizes the biological parameters that increase after BO112 treatment in tumor/blood samples. Conclusions: BO112 demonstrated a manageable safety profile. Its biological activity is consistent with both a direct antitumor effect and with intratumoral and systemic immunity activation, driven by IFN-gamma pathway. 1mg qw x 2-3 doses in combination with anti-PD-1 is the chosen regimen for an expansion cohort in pts refractory to anti-PD-1 therapy. Clinical trial identification: NCT02828098 Release date: June 23, 2016 Legal entity responsible for the study: Bioncotech Theraputics Funding: Bioncotech Theraputics Disclosure: I. Marquez Rodas: Advisory role form BMS, MSD, Novartis, Roche, Pierre Fabre, Amgen and Bioncotech S. Lopez-Tarruella Cobo: Advisory Role (Novartis, Pfizer, Celgene) y Honoraria/lecturer (Novartis, Pfizer, AZ, Celgene, Roche) L. Planelles, M. Quintero: Bioncotech (employment) S. Martin Algarra: Advisory role for BMS, MSD, Novartis, Roche, Amgen I. Melero: Advisory role/consulting AstraZeneca; Bristol-Myers Squibb; Pfizer; Roche/Genentech All other authors have declared no conflicts of interest.

Background: The RMH score has been validated to predict survival in different populations of patients starting phase I clinical trials. Most of the populations where it has been validated are of heavily treated patients that lack other treatment options. The type of phase I trials is changing and we aimed to validate the score in a new cohort with more patients treated on an early line, rahter than the usual classic phase I trials heavily pretreated patient population. Methods: We analyzed the RMH score in the patients treated in our center in a phase I trial between 2012 and 2017. We collected demographics data, overall survival after starting the trial, the RMH score (albumin, LDH, and number of metastatic sites) for all patients and the treatment line. We considered a late line anything over two treatment lines and in any case if the patient did not have any other treatment available depending on the tumor type. An early line was the first or second treatment line when the patient did have further lines available. Results: We treated 77 patients on a phase I trial in our institution, 23 males and 54 females. Mean age was 55 years (26-77). RMH score was (0/1/2/3) in (31/23/20/3) patients. Thirty-three patients were treated on an early line. Median survival for low score (0/1) was 639 days and for high score (2/3) was 327 days p = 0.0834. The mean survival for patients with low RMH score was higher than those with a high RMH score in every treatment line, although due to the low number of patients in some of those categories the difference was not significant. Conclusions: The RMH score did predict well the overall survival in our patients. The survival times in our institution are higher than those previously published, probably due to the inclusion of patients on earlier treatment lines than those used before to calculate and validate the score. Our findings support the use of the RMH score for the selection of patients entering phase I trials irrespectively of the design of the trial (early vs. late line). Legal entity responsible for the study: Medical Oncology Department, Hospital General Universitario Gregorio Marañón Funding: Instituto de Investigación Sanitaria Hospital Gregorio Marañón Disclosure: All authors have declared no conflicts of interest.

PURPOSE:To evaluate the incidence of venous thromboembolism (VTE) in ambulatory pancreas cancer patients receiving chemotherapy and analyze Khorana's predictive model of chemotherapy-associated thrombosis.METHODS/PATIENTS:We performed a retrospective review to determine the incidence of VTE in the gastrointestinal cancer unit of our center. Between 2008 and 2011, 84 consecutives patients diagnosed with pancreas adenocarcinoma were identified and included in the analysis. Pancreatic neuroendocrine tumors were excluded.RESULTS:Thirty patients experienced VTE (35.7 %) and 66 % of the events were diagnosed during the first 6 months after diagnosis. Khorana's score: 33.3 % of the intermediate category patients developed a venous thromboembolic event and 37.5 % in the high-risk category.CONCLUSIONS:The high incidence of VTE observed in this study is consistent with prior reports. Specific predictive model for chemotherapy-associated thrombosis in pancreatic cancer must be investigated.

We investigated the association between skin rash and plasma creatine kinase (CK) levels in oncology phase I trials. We analysed data from 295 patients treated at our institution within 25 phase I trials which included CK measurements in the protocol. Trials involved drugs targeting EGFR/HER2, m-TOR, VEGFR, SRC/ABL, aurora kinase, BRAF/MEK, PARP, CDK, A5B1 integrin, as well as oncolytic viruses and vascular disrupting agents. Creatine kinase measurements were available for 278 patients. The highest levels of plasma CK during the trial were seen among patients with Grade (G) 2/3 rash (median 249 U l−1) compared with G1 (median 81 U l−1) and no rash (median 55 U l−1) (P<0.001). There was a significant reduction in CK after the rash resolved (mean 264.2 vs 100.1; P=0.012) in 25 patients, where serial CK values were available. In vitro exposure of human keratinocytes to EGFR, MEK and a PI3Kinase/m-TOR inhibitor led to the increased expression of CK-brain and not CK-muscle or mitochondrial-CK. Plasma CK elevation is associated with development of skin rash caused by novel anticancer agents. This should be studied further to characterise different isoforms as this will change the way we report adverse events in oncology phase I clinical trials.

ABSTRACT Introduction Standard chemotherapy is increasingly discontinued after successful “induction” in patients (pts) undergoing 1st line treatment for mCRC. MGN1703 is a synthetic DNA-based immunomodulator acting as an agonist of TLR-9 that has shown preclinical activity in mCRC. This study has been conducted to assess clinical efficacy, immunogenicity, and safety of MGN1703 as maintenance vs. placebo. Methods The IMPACT study is an international, multicenter, randomized double-blind placebo-controlled phase II/III study. Pts with mCRC showing disease control (CR, PR or SD) after 4.5 to 6 months of 1st-line standard therapy with FOLFOX/XELOX or FOLFIRI +/- bevacizumab (investigatoŕs choice) were included. Results Interim analysis of the unblinded data revealed a strong therapeutic effect compared to anticipated PFS. Therefore, patients were withheld from further randomization, according to a decision of the steering committee. In the ITT population (N = 55 pts), hazard ratio (HR) was 0.53 in favor of MGN1703 (p = 0.062). In the per-protocol population (excluding screening failures; N = 50), HR was 0.43 (p = 0.015). In the pre-defined target population (2 out of 3 factors: CEA Conclusions Maintenance therapy with MGN1703 after standard chemotherapy with or without bevacizumab, is associated with significantly improved progression-free survival compared to placebo and is accompanied by low toxicity. A confirmatory clinical study in patients with metastatic CRC is currently being planned. Disclosure M. Tschaika: Employment, stock ownership. M. Schmidt: Employment. B. Wittig: Stock ownership. All other authors have declared no conflicts of interest.

Introduction Novel methods of assessing treatment efficacy should prove helpful in drug development and in the management of cancer patients (pts). We have developed a novel method to analyze tumor response to therapy by quantifying the rate of tumor regression (d) and growth (g). We have shown g is slower when pts are on effective therapy and that g correlates with survival. We utilized this novel methodology in pts with a diagnosis of CRC to compare the efficacy of oxaliplatin- and irinotecan-containing regimens (OCR, ICR). Unlike PFS, an incremental measure of efficacy, g is a continuous variable and can more accurately assess differences between treatments. Because calculations of g are indifferent to assessment intervals, estimating a tumor's g allows comparison of efficacy across trials. Methods Using tumor measurements obtained for RECIST assessments or serum CEA levels and a two-phase mathematical equation we determined d and g in 70 pts at each evaluation. Results 70 pts (47 male, 23 female) with metastatic CRC were studied. The median g value with OCR (.00132) was statistically similar (p = 0.573) to that observed with ICR (.00148). Both therapies also had similar effects on d (OCR = 0.00529; ICR = 0.00451; p = 0.716). Similar results were seen using serum CEA levels. Furthermore, in an individual pt, statistically valid g and d values could be estimated long before tumor quantity increased, providing an early indicator of treatment failure. Importantly, in the majority of pts receiving prolonged treatment the growth rate constant did not change, despite rising tumor quantities, indicating resistance is intrinsic. Additional data being gathered should allow us to compare relative effects on g in first and subsequent lines to discriminate between the effects observed in tumor shrinkage and the impact on tumor growth. Conclusions In this cohort of CRC patients the data suggest both OCR and ICR have similar efficacy. The evidence in pts receiving long term therapy demonstrating no change in g over time suggest resistance is intrinsic and not acquired, and would support a strategy of continued treatment with a tolerable regimen given the similar efficacy of the commonly used combinations. Disclosure All authors have declared no conflicts of interest.

ABSTRACT Introduction Cisplatin (CDDP) is a very effective and common treatment in solid malignancies used mostly in breast, ovarian, bladder, esophageal, gastric, head and neck cancer and germ cell tumors. One of the most important side effects of CDDP is the nephrotoxicity, especially with CDDP doses higher than 60 mg/m2, affecting as much as 30% of the patients. Nephrotoxicity is a limitating side effect on treatment with CDDP, preventing patients with limit kidney function of receiving the drug and stopping treatment once kidney function has worsened. Cilastatin (Cls) is a specific inhibitor of renal dedydrodipeptidase I (DHP-I) which prevents hydrolysis of imipenem and its accumulation in the proximal tubule. In this work we hypothesized that Cls acts as a nephroprotector against CDDP-induced damage without compromising antitumor activity. Methods Primary cultures of proximal tubular cells (PTCs) and cell lines of different malignancies (colon, breast, ovarian, bladder) were cultured with different concentrations of CDDP (1, 10 and 30 µM) in the presence or absence of Cls. Cell viability was assessed with MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide) assay. Raft staining was measured with toxine B choleric and FasL by confocal microscopy. Results Cls interfered with CDDP-induced FasL signalling at raft level on PTCs brush border. Concomitant treatment with Cls reduced CDDP-induced changes. Several tumoral cell lines were tested with CDDP and Cls together. Cls did not increase or decreased tumor growth alone or in combination with CDDP. CDDP sensitivity was not affected by Cls. Conclusion By binding a DHP-I, Cls blocks CDDP-induced PTCs apoptosis but it does not affect the CDDP antitumoral activity. Our findings suggest that the affinity of Cls for renal DHP-I makes this effect specific for proximal tubular cells and may be related to a reduction in intracellular drug accumulation. Cls administration might represent a novel strategy in the prevention of CDDP-induced acute renal injury. Cls treatment could potentially increase the number of patients undergoing CDDP treatment or maintaining treatment. Further clinical trials for renal function preservation in cancer patients are on development. Disclosure All authors have declared no conflicts of interest.

8564 Background: Dacarbazine (DTIC) and interferon (IFN) are two agents currently used in the treatment of malignant melanoma (MM). They have modest response rates (RR) and marginal improvement in time to progression (TTP) or survival. The recently published results of targeted agents such as ipilimumab and PLX4032 are encouraging, and will one day replace these agents as front line treatment. We hypothesize that clinical outcomes of patients (pts) with MM treated with targeted therapies on phase I clinical trials are equivalent to those seen with agents such as DTIC or IFN. Methods: We analyzed the outcomes of 65 patients with MM that had undergone treatment in our phase I clinical trials unit between 2005-2010. All pts had received at least one previous line of systemic treatment and acted as theis own controls. RR and TTP on previous treatment was compared to RR and TTP on the phase I trial. Differences were studied by a Chi-squared test and the log-rank test respectively. A Cox analysis was performed to identify clinical variables that could influence TTP. Results: 27 females and 38 males were included. Mean age was 56 years. Performance status (PS) was 0 in 26 pts,1 in 35 pts and 2 in 4 cases. First line of treatment included DTIC or temozolamide in 58 pts, IFN in 5 pts and cisplatin based treatment in 2 pts. 25 pts had received previous treatment on a Phase II/III trial and data regarding best response to previous treatment was not available in 3 pts. There was no statistical significant difference in RR between to the first line treatment with chemotherapy or cytokine therapy vs treatment with a targeted agent in phase I trial (7/65) 11% vs (9/65) 14%, p=0.87. In addition there was no difference in TTP on the first line treatment vs treatment with a targeted agent in phase I trial; 90 days (95% CI 65-120) vs 53 days (95% CI 42-79), p= 0.15. No clinical variable showed influence on TTP. Conclusions: This data emphasize the lack of superiority of DTIC and IFN based regimes over targeted experimental agents for the treatment of MM. In event of progression to treatment with targeted agents as ipilimumab or PLX4032, patients should be considered for clinical trials with investigational agents rather than receiving chemotherapy or cytokine therapy.

2518 Background: N (ABT-263) is a first-in-class orally available Bcl-2 protein inhibitor that in vitro and in vivo potentiates D activity across a variety of solid tumor types. This phase I study was performed to determine the maximum tolerated dose (MTD) and to assess the safety and PK profile of N plus D. Methods: Pts with advanced solid tumors, adequate organ function and ECOG performance status ≤1 were eligible. N at 150 and 200 mg was given PO, QD days (d) 1-3 or d1-5 q21 followed by D 75 mg/ m2 IV on d 1 q21. To compare the PK of either agent alone, N was administered on days 3-5 or 3-7 only in course 2. PK samples were collected for N and D in cycles 1 and 2. Adverse events (AE) were graded by NCI CTCAE V3.0. Results: 15 pts (M/F: 8/7; median age 54 yrs (range 30-71) were treated at 3 dose levels of N: 150 mg and 200mg d1-5 and 200 mg d1-3. DLTs were febrile neutropenia and grade (gr) 4 thrombocytopenia at N 200 mg d1-5 and febrile neutropenia at 200 mg d 1-3. The MTD is set at 150 mg N d1-5 with D 75 mg2 on d1 and at this dose 12 additional pts are currently being enrolled. Other gr 3/4 AEs were neutropenia (43%), fatigue (21%) and thrombocytopenia (14%); frequently occurring grade 1/2 AEs were fatigue (57%), nausea (50%), vomiting (43%), anorexia (36%), constipation (29%), diarrhea (29%), and thrombocytopenia (21%). One pt with NSCLC had a confirmed partial response (PR), one pt with breast cancer had an unconfirmed PR. Five pts showed stable disease; duration of this ranged from 5-18 weeks; 3/5 pts had SD up to 12 weeks, 1/5 pts had SD up to 14 weeks, and 1/5 pts had SD for 18 weeks. Based on limited PK data, there was no apparent PK interaction between N and D (Table). Conclusions: Navitoclax 150 mg for d1-5 can be safely administrated with D 75 mg/m2. This combination is tolerable with no apparent PK interactions. Antitumor activity was observed in 7 pts (2 PR, 5 SD). Cmax/dose (ng/mL/mg)* AUC24/dose (ng*hr/mL/mg)* Navitoclax (150-200 mg, N=7) Navitoclax alone 20.9 ± 10.4 293 ± 174 Navitoclax with D 14.7 ± 11.4 222 ± 166 D (75 mg/m2, N=3) Cmax ( µ g/mL)* AUCinf ( µ g*hr/mL)* D alone 2.6 ± 1.2 2.9 ± 1.2 D with navitoclax 2.3 ± 0.7 2.4 ± 0.7 * Mean +/− SD

3096 Background: AZD8055 is a dual mTORC1/mTORC2 inhibitor with properties that prevent the feedback activation of AKT seen with rapalogues. This is the first-in-man study of AZD8055 to determine the safety, PK and preliminary efficacy of two oral formulations in patients with advanced solid tumours ( NCT00731263 ). Another dual mTORC1/mTORC2 inhibitor, AZD2014, is currently being tested in the clinic. METHODS In an open-label, ascending dose study, cohorts of 3 or 6 patients received a single dose of AZD8055 followed 1 week later by continuous, twice-daily dosing. Starting at 10 mg, each subsequent cohort received an increased dose until a non-tolerated dose was reached. The formulation was switched from oral solution to tablet when the tablet became available. Safety and PK data were collected in addition to FDG-PET and RECIST. Modulation of AKT phosphorylation on serine 473 and of 4EBP1 phosphorylation on threonine 37/46 was measured in peripheral tissue after single administration of AZD8055 to confirm proof-of-mechanism. RESULTS 49 patients (42 evaluable) were dosed in seven cohorts: oral solution; 10, 20, 40 mg BD, tablet; 40, 60, 90, 120 mg BD. Non-tolerated dose was 120 mg BD and so maximum tolerated dose (MTD) was 90 mg BD. Dose limiting toxicities of CTC grade 3 transaminases were reported with 40 mg BD solution (n=1), 90 mg BD tablet (n=1), 120 mg BD tablet (n=3). No cases of Hys law were observed. Onset of increased transaminases was typically between day 35 and 45 but there was no clear relationship with plasma concentration of AZD8055. AZD8055 was rapidly absorbed with median tmax of 0.5 h for both single dose (n=49) and steady state (n=36). Elimination was rapid with mean t1/2 of 2.4 h for single dose (SD=1.1, n=49) and 2.7 h for steady state (SD=0.79, n=36). AZD8055 showed a large apparent volume of distribution and high apparent clearance: mean Vss/F 2400L (SD=2062; 90 mg dose, n=11), mean Cl/F 950 L/h (SD=889; 90 mg dose, n=11). Pharmacodynamic changes in peripheral blood mononuclear cells were observed and results will be presented. CONCLUSIONS The MTD for AZD8055 is 90 mg BD. Toxicities frequently associated with administration of rapalogues such as rash and mucositis were not dose limiting. Updated data will be presented.

2516 Background: M-ICSOLs are often cited as an exclusion criteria in phase I trial protocols. This leads to patients with asymptomatic M-ICSOL being denied entry into phase I trials. We examined clinical outcomes of patients treated in such trials in our institution to determine if this exclusion criteria was justifiable. Methods: Data from consecutive patients who participated in 65 phase I trials from 1998 to 2009 were included in the analysis. For an initial analysis patients were divided into two groups: (A) those that had a known M-ICSOL prior to starting trial therapy and (B) those that did not. A further analysis was performed comparing patients with no known M-ICSOL at the start and at the end of the phase trial—group C—to those who subsequently developed a M-ICSOL—group D. The primary endpoint was time to progression (TTP) calculated for all groups; Kaplan-Meier analysis using a log rank test was used to determine the difference between groups. Results: Data from 885 patients, M:F (479:406), median age 58.6 years (range, 16-89) were studied. There was no significant difference in TTP between patients in group A (n=24) and group B (n=861): 8.0 weeks, (95% CI 7.57-8.43 weeks) versus 14.0 weeks, (95% CI 8.13-19.87 weeks), p = 0.071 respectively. There was no difference in TTP between groups C (n=804) and D (n= 57): 8.0 weeks, (95% CI 7.32-8.67 weeks) versus 8.0 weeks, (95% CI 6.43-9.57 weeks), p = 0.582 respectively. The median time to development of a M-ICSOL in patients in group D was 8 weeks. Conclusions: This analysis suggests that carefully selected patients with M-ICSOLs entered into phase I trials do not have a worse clinical outcome when compared to those who did not have a M-ICSOL at the start of therapy. In addition, patients who developed an M-ICOSL on trial did so after the conventional dose-limiting toxicity window. We recommend that, contrary to current trends, carefully selected patients with M-ICSOLs be included in phase I trials. No significant financial relationships to disclose.

The purpose of this study was to identify clinical and pathological parameters to improve prediction of disease-free survival (DFS) and overall survival (OS) in patients treated with neoadjuvant chemoradiotherapy for rectal cancer.