Erlichman, Charles

551 Background: Sorafenib inhibits various pro-angiogenesis pathways including PDGFR-B, a factor associated with resistance to anti-VEGF therapy. A previous phase II trial in patients with chemorefractory metastatic CRC demonstrated a 63% disease control rate with a combination of bevacizumab (BEV) and sorafenib. This phase I trial sought to determine the MTD of BEV and sorafenib combined with standard cytotoxic therapy for advanced gastrointestinal (GI) cancers. Methods: Patients with advanced GI malignancies appropriate for irinotecan-based therapy were enrolled (14 with CRC, 3 gastroesophageal). A standard 3 + 3 design was used with 3 escalating sorafenib dose levels (DL): (1) 200 mg po daily, days 3-7, 10-14; (2) 200 mg po twice daily, days 3-6, 10-13; and (3) 200 mg po twice daily, days 3-7, 10-14. FOLFIRI: irinotecan 180 mg/m2 d1, leucovorin 400 mg/m2 d1, 5-fluorouracil (FU) bolus 400 mg/m2 d1, 5-FU infusion 2400 mg/m2d1-2 and BEV 5 mg/kg d1. 1 cycle = 14 days. Results: Seventeen pts were enrolled, median age of 56 (range 32 and 81). Two pts were replaced, as they did not complete DLT evaluation, leaving 15 evaluable pts. Four evaluable pts at DL1 and 6 pts at DL2 had no DLTs. At DL 3, the first cohort of 3 pts did not experience any DLTs. In the second cohort of 3 pts, 2 pts experienced DLTs (asymptomatic G3 hypophosphatemia, G3 dehydration and diarrhea). MTD was determined to be DL2: sorafenib 200 mg PO twice daily, days 3-6, 10-13 combined with FOLFIRI and BEV at standard doses. Of the 15 evaluable pts, 4 pts had PR, 8 pts had SD as best response, 1 pt had PD, and 2 pts discontinued treatment prior to first tumor assessment. The median number of cycles was 10 (range 1-37). Three pts with CRC had disease control > 12 months. Conclusions: The MTD of this regimen is sorafenib 200 mg PO twice daily, days 3-6,10-13 combined with standard doses of FOLFIRI and BEV. Dual VEGF inhibition combined with cytotoxic therapy may provide prolonged disease stabilization for select patients with advanced GI malignancies. Supported by CA69912 and CA15083. Clinical trial information: NCT01383343.

245 Background: Gemcitabine (GEM) monotherapy for PDA has modest activity. 17-N-Allylamino-17-demthoxygeldanamycin (Tanespimycin/17AAG), is an HSP90 inhibitor, which results in degradation of a number of client proteins such as RAF and Akt. Arlander et al (J Biol Chem. 2003 278:52572-7) have demonstrated that 17AAG targeting of HSP90 leads to Chk1 degradation. Chk1 is upregulated with GEM treatment which affects cell survival (Karnitz et al. Mol Pharmacol. 2005 68:1636-44). In vitro, the combination demonstrated in vitro synergy. Thus the potential clinical activity in pancreatic cancer of the combination was explored. Methods: A phase II multicenter prospective study was performed using GEM 750 mg/m2 d1,8 and 17AAG 154 mg/m2 d2,9 on a 21 day cycle, as defined in a previous phase I trial. Patients with stage IV PDA, with adequate liver, kidney function and counts, ECOG 0-2 and no prior therapy for metastatic disease were eligible. The primary endpoint of 60% overall survival at 6 months was considered of clinical interest. Sixty-six patients were planned for accrual, and an interim analysis was planned after 25 evaluable patients. Informed, written consent was obtained from all pts. Results: Due to lack of drug availability and failure to achieve the planned endpoint, accrual was halted at 21 pts, enrolled from 5/08 to 9/10, of which 20 were evaluable (1 cancelled prior to treatment). Median age was 61.5 yrs., 55% were male, PS 0 (50%), 1 (45%), 2 (5%); 95% were non-Hispanic white, 1 was African American. Two had received adjuvant therapy for PDA. Tolerability was moderate, with 13 pts (65%) having > grade 3 adverse event (AE), with 15% gr 4 AEs and 0% gr 5. Nausea (20%), vomiting (20%), constipation (15%), dehydration (15%), and anorexia (10%), were the most common non-heme AEs. Lymphopenia (15%), leukopenia (10%) and neutropenia (10%) were common grade 3 hematologic AEs. No complete or partial responses were seen. 40% of patients were alive at 6 months, with a median OS of 5.4 months. Conclusions: HSP 90 inhibitor 17AAG in combination with GEM does not show activity for treatment of PDA. Further trials should not be pursued with this combination. Supported by N01-CM-62205. Clinical trial information: NCT00577889.

708 Background: In June 2015 Cancer Care Ontario convened an expert panel to determine if there is an optimum LV dose in 5FU-LV combinations for the treatment of CRC. This request arose out of observed variation in LV dosages between some cancer centers. The research question was effect of LV dose on overall survival (OS), progression free survival (PFS), disease free survival (DFS), response rate (RR) and adverse events/toxicity, given a constant dose of 5FU. Methods: A systematic search was conducted for guidelines (GL) and comparative trials; eligibility included English language, with > 30 patients, that examined different doses of LV where dose of 5FU was not varied. Assessment of studies for inclusion was completed by 4 reviewers. Results: We identified 5 GL, 0 systematic reviews and 12 trials that defined a LV dose in combination with 5FU. None of the GL informed an optimal dose of LV. RR was assessed in 10 trials; 4 showed trends to higher RR with higher LV dose, but differences were not statistically significant (SS) between arms. PFS or DFS was reported in 6 trials and was similar between arms. Time to recurrence reported in one trial that included bevacizumab (BV) was longer in the high dose LV group that was SS. OS was addressed in 10 studies: no difference found in 7 studies; in one RCT OS was longer with the higher dose LV 55 vs. 45 months (p not reported); in one retrospective study OS was 23 vs. 20 months in favor of high dose LV (p not reported); one study of LV and BV had longer OS vs. lower dose LV at 26 vs. 21 months (SS). Toxicity: higher dose LV was associated with greater toxicity in 3 of 4 studies that reported stomatis, and increased rates of diarrhea in 9 of 11 trials. Meta-analysis was not appropriate as studies were too heterogeneous. Conclusions: There is no convincing evidence to identify an optimum dose of LV to be used in 5-FU/LV combinations. Amongst studies that did show a difference the trend was improved survival in favor of the higher dose. Similarly, differences in toxicities when identified were consistently greater with the higher dose LV. The expert panel concludes that the existing literature provides insufficient data to suggest that chemotherapy protocols should deviate from standard protocol doses.

PURPOSE: Multitargeted antifolate (MTA) is an investigational agent that, like gemcitabine, exhibits broad activity in solid tumors. A phase I trial of MTA and gemcitabine was undertaken, based on the demonstration of preclinical cytotoxic synergy. PATIENTS AND METHODS: Thirty-five patients (group I) received 164 courses (median, four; range, one to 14 courses) of treatment of gemcitabine at doses of 1,000 and 1,250 mg/m2 on days 1 and 8 and MTA at doses of 200, 300, 400, 500, and 600 mg/m2, given 90 minutes after gemcitabine on day 1. Courses were repeated every 3 weeks. Because the day 8 dose of gemcitabine was reduced or omitted in 57% of courses due to neutropenia, 21 patients (group II) were treated on an alternate schedule, with MTA administered on day 8 rather than day 1. This group received 85 treatment courses (median, four; range, one to 10 courses). RESULTS: The most common and dose-limiting toxicity was neutropenia. Other toxicities included nausea, fatigue, rash, and elevated hepatic transaminases. The maximum-tolerated dose was gemcitabine/MTA 1,000/500 mg/m2 for group I and 1,250/500 mg/m2 for group II. Thirteen objective responses were documented (colorectal cancer, n = 3; non–small-cell lung cancer, n = 3; cholangiocarcinoma, n = 2; ovarian carcinoma, n = 2; mesothelioma, n = 1; breast cancer, n = 1; and adenocarcinoma of unknown primary site, n = 1). Gemcitabine had no effect on the disposition of MTA. CONCLUSION: The gemcitabine/MTA combination is broadly active and warrants further evaluation. The sequence of gemcitabine administered on days 1 and 8 with MTA administered on day 8 is better tolerated and is recommended for further study at doses of gemcitabine/MTA 1,250/500 mg/m2.

PURPOSE: The goals of this study were to determine the maximum-tolerated dose and describe the toxicities of the combination of irinotecan and docetaxel administered every 3 weeks to patients with advanced malignancies and, also, to evaluate the effect of irinotecan on the disposition of docetaxel and describe preliminary evidence of antitumor activity. PATIENTS AND METHODS: Eighteen patients received 85 courses (median, two courses; range, one to 15 courses) of treatment with irinotecan, administered over 90 minutes by intravenous infusion, followed by docetaxel, administered over 60 minutes by intravenous infusion. Four escalating dose levels of irinotecan/docetaxel (160/50 mg/m2, 160/65 mg/m2, 200/65 mg/m2, and 200/75 mg/m2) were studied. Pharmacokinetic analyses were performed to evaluate the effect of irinotecan on the disposition of docetaxel. RESULTS: The most common and dose-limiting toxicity was myelosuppression, which consisted of neutropenia that was severe (National Cancer Institute common toxicity criteria [NCI CTC] grade 4) but brief (< 5 days) in 11 patients, with three episodes of febrile neutropenia. Nonhematologic toxicities of anorexia, nausea, and stomatitis were mild to moderate (NCI CTC grades 1 and 2), but there was one incidence each of both CTC grade 3 anorexia and nausea. All patients had total alopecia. Diarrhea was dose-dependent and severe in four patients who failed to take adequate antidiarrhea therapy. Five out of 16 assessable patients, one with cholangiocarcinoma, one with leiomyosarcoma, and three with non–small-cell lung cancer, achieved partial remissions. CONCLUSION: The combination of irinotecan and docetaxel causes significant reversible myelosuppression, which was dose limiting but led to no serious sequelae. There was no evidence of a clinically significant interaction using these two agents in this sequence. The combination showed antitumor activity at all the dose levels tested and should be further studied in a number of tumor types. The recommended phase II dose on this schedule is irinotecan 160 mg/m2 and docetaxel 65 mg/m2.

PURPOSE: To compare the pharmacokinetics of continuous venous infusion (CVI) fluorouracil (5-FU) with that of oral eniluracil/5-FU and to describe toxicities and clinical activity of prolonged oral administration of eniluracil/5-FU. PATIENTS AND METHODS: A randomized, open-label, cross-over study compared CVI 5-FU to an oral 5-FU/eniluracil combination. Seventeen patients (arm A) were randomly assigned to receive eniluracil/5-FU combination tablets (10:1 mg/m 2 BID for 7 days) during the first study period, followed by 5-FU (300 mg/m 2 CVI for 7 days) during period 2, with a 14-day washout between periods. Sixteen patients (arm B) received treatment in the opposite sequence. In period 3, all patients received eniluracil/5-FU tablets BID for 28 days. Plasma levels of 5-FU during CVI and oral administration were analyzed in periods 1 and 2. Dihydropyrimidine dehydrogenase (DPD) activity was determined by measuring plasma uracil, urinary α–fluoro-β-alanine, and peripheral-blood mononuclear cell (PBMC) DPD activity. RESULTS: There were no grade 3 or 4 toxicities in either arm. Partial responses were observed in three patients. Another three patients had stable disease for ≥ 3 months. Eniluracil and 5-FU pharmacokinetics were similar to those observed in previous studies and were unaffected by administration sequence. The mean ± SD steady-state plasma concentration (C P ) and area under the curve (AUC) 144-168h for CVI 5-FU (104 ± 45 ng/mL and 2,350 ± 826 ng·h/mL, respectively) were three-fold greater than those for oral 5-FU (38.1 ± 7.7 ng/mL and 722 ± 182 ng·h/mL, respectively [P < .00001]). Individual 5-FU concentrations during CVI were highly variable, whereas those after eniluracil/5-FU were very reproducible. DPD activity in PBMCs before each study period was normal. CONCLUSION: Both CVI 5-FU and oral eniluracil/5-FU were well tolerated, with moderate activity in these heavily pretreated patients. However, 5-FU steady-state C P and AUCs achieved with oral eniluracil/5-FU were significantly less than with CVI 5-FU.