Trading Plan Cost for Timeliness in Situated Temporal Planning

If a planning agent is considering taking a bus, for example, the time that passes during its planning can affect the feasibility of its plans, as the bus may depart before the agent has found a complete plan. Previous work on this situated temporal planning setting proposed an abstract deliberation scheduling scheme for maximizing the probability of finding a plan that is still feasible at the time it is found. In this paper, we extend the deliberation scheduling approach to address problems in which plans can differ in their cost. Like the planning deadlines, these costs can be uncertain until a complete plan has been found. We show that finding a deliberation policy that minimizes expected cost is PSPACE-hard and that even for known costs and deadlines the optimal solution is a contingent, rather than sequential, schedule. We then analyze special cases of the problem and use these results to propose a greedy scheme that considers both the uncertain deadlines and costs. Our empirical evaluation shows that the greedy scheme performs well in practice on a variety of problems, including some generated from planner search trees.

A phase I/II study to assess the safety and efficacy of pazopanib (PAZ) and pembrolizumab (PEM) in patients (pts) with advanced renal cell carcinoma (aRCC).

4506 Background: PAZ is indicated for the treatment of aRCC. The combination of an anti-angiogenic agent and immunotherapy may improve anti-tumor activity. We report preliminary safety and efficacy results of the phase I part of the study. Methods: Twenty pts were originally enrolled in cohorts A and B assessing PAZ 800 mg and 600 mg, respectively, both with 2mg/kg (Q2W and then Q3W) PEM to determine the maximum tolerated dose. Due to dose limiting liver toxicity, cohort C was opened to assess if the sequential schedule of 9 weeks PAZ run-in followed by PAZ+PEM would improve safety. Strict safety criteria for initiating PAZ+PEM were set. The data from this ongoing study are presented given the limited information available on the combination of TKI + PD-1 inhibitors in RCC. Results: Overall, 35 pts were treated; 5 out of 15 pts in cohort C received PAZ+PEM at the data cut-off. Three dose-limiting toxicities (DLT) occurred in cohort C in pts receiving PAZ+PEM; updated DLTs in all cohorts are reported in Table. G3/4 AEs were observed in 90% of pts in cohorts A and B and in 80% of pts in cohort C receiving PAZ+PEM. No G3/4 ALT/AST elevation was reported in cohort C PAZ+PEM while they were observed in 70% and 60% in cohorts A and B, respectively. Best overall response (CR+PR) was reported in 6, 2 and 1 pts receiving PAZ+PEM in cohorts A, B and C, respectively. Conclusions: Results from cohorts A and B showed significant hepatotoxicity. The sequential schedule PAZ → PAZ+PEM has shown reduced hepatotoxicity and preliminary signs of efficacy but overall limited tolerability. PAZ+PEM is not suitable to test in a larger cohort. Clinical trial information: 2013-003785-14. [Table: see text]

Sequential Arsenic Trioxide, and All Trans Retinoic Acid with Daunomycin Yield Early Cytogentic and Molecular Response in Newly Diagnosed Acute Promyelocytic Leukemia – Data From a Tertiary Cancer Centre.

Abstract 2628 Acute promyelocytic leukemia (APL) remains the most curable amongst myeloid leukemias though early morbidity and mortality remains an issue. The use of differentiating agents, All Trans Retinoic acid (ATRA) and Arsenic trioxide (ATO) with chemotherapy has been integral to decreasing morbidity and mortality while achieving high cure rates in APL. However, there is some uncertainty about the best combination, sequence and durationfor their use in order to achieve optimal response with minimum toxicity. The understanding that attaining complete cytogenetic and molecular response yields improved long term cures led us to investigate sequential use of ATO followed by ATRA and Daunorubicin in treating APL patients as frontline therapy in order to achieve such response. Consecutive adult patients with APL irrespective of their Sanz risk group were treated using a protocol of ATO (10mg IV infusion over 3 hours daily for 45 days) in the first phase followed by ATRA (45mg/m2 for 60 days) in combination with Daunorubicin (60mg/m2 for 3 days × 3 cycles) in second phase. All patients received maintenance ATRA (45m/m2 for 15 days every 3 months) for a period of 18 months in phase 3. Patients were monitored for cytogenetic and molecular responses after phase 1 and 2. They were monitored for treatment related toxicity and were also followed up for survival post treatment. Between January 2008 and February 2012, 106 patients with APL were treated with the sequential regimen. Median age was 30 years with a male to female ratio of 1.3:1. Only 16% were in Sanz class I risk while 39% and 45% were in class II and III, respectively. Cytogenetic remission was achieved in 71% (data available in 80/106 patients) while molecular remission was seen in 66.6% (data available in 40/106 patients) following phase 1. Molecular remission improved to 96% (data available 74/106) following phase II. Relapse was encountered in 8.4% while 11.3% died early, within 7 days of initiating therapy. One patient died due to unrelated cause during phase III. At a median follow up of 16.5 months, 86.2% are alive with an overall relapse free survival (RFS) of 72%. Class I and class II patients fared better (73% and 77% respectively) versus class III patients who had a 63% RFS. Median survival was not reached at the time of analysis. The regimen was well tolerated with differentiation syndrome occurring in 18% of treated patients and transient QT prolongation occurring in 20%. Only 5% of patients developed peripheral neuropathy (gradeII). Our data involving use of all active agents in APL in a sequential schedule has resulted in significant early cytogenetic and molecular response. The strategy has so far translated into an excellent relapse free survival. Further follow up with molecular monitoring is required to derive conclusions pertaining to late relapses. Early deaths due to disease related complication continue to be a problem. The need to intensify therapy based on early responses (remission versus persistence at the molecular level) needs to be explored. The sequential schedule has shown excellent tolerance and toxicity profile, thereby decreasing morbidity particularly in early phase of treatment of APL. Disclosures: No relevant conflicts of interest to declare.

Phase I study of intravenous decitabine in combination with oral vorinostat in patients with advanced solid tumors and non-Hodgkin's lymphomas (NHL)

3528 Background: Decitabine (D), a hypomethylating agent, and vorinostat (V), a histone deacetylase inhibitor, belong to two different classes of drugs with an epigenetic effect. The ideal dose scheduling of these drugs remains controversial. This phase I study aims to determine the recommended phase II dose (RPTD) of the combination, their toxicity profile, pharmacokinetic (PK) interaction and preliminary clinical activity. Methods: Patients (pts) with advanced solid tumors or relapsed/refractory NHL are eligible. Two different schedules of D and V are being evaluated: sequential administration of D followed by V and concurrent administration of D and V. Dose escalation of D and V on the sequential schedule is described in Table. Results: To date, 27 pts have been entered into dose levels 1, -1, 1a, 1b, -1b, -2b of the sequential schedule. Demographics: median age 61 (range 31–76), F:M = 13:14, ECOG 0:1:2 = 8:16:3, tumor types: 24 solid tumor and 3 NHL. Pts received a total of 77 cycles with a median of 2 cycles (range 1–8). Adverse events (AE) of grade 3 or higher of at least possible attribution to the study treatment were neutropenia (16 pts), thrombocytopenia (4), febrile neutropenia (2), fatigue (2), and 1 pt each for constipation, dehydration, nasal bleeding, elevated alanine aminotransferase, and hyponatremia. Dose limiting toxicities (DLT) consisted mainly of myelosuppression, constitutional and gastrointestinal symptoms occurred in 7/27 (26%) of pts so far. Disease stabilization for 4 or more cycles was observed in 7 out of 22 (31.8%) evaluable pts (two with breast and one each of thymus, colon, pancreatic, appendix and non-small cell lung cancers). Conclusions: The sequential combination of D and V seems to be tolerable after some adjustments in the doses and duration of drug administration. Prolonged disease stabilization has been observed in multiple tumor types. Accrual is ongoing and RPTD will likely be dose level -1b or -2b. ( Supported by NCI Grant No. U01CA132123.) [Table: see text] No significant financial relationships to disclose.

A Phase I Trial of Two Sequence-Specific Schedules of Decitabine and Vorinostat in Patients with Acute Myeloid Leukemia (AML).

Background: Epigenetic silencing of genes has been documented in AML. This phase I trial evaluates the safety, tolerability, and maximum tolerated dose (MTD) of two schedules of administration of the hypomethylating agent decitabine in combination with the pan-selective histone deacetylase inhibitor vorinostat. Methods: Patients receive escalating doses of oral vorinostat administered either sequentially [100 mg bid (n=4), 200 mg bid (n=4), or 200 mg tid (n=8) Days 6–21] or concurrently [100 mg (n=3) or 200 mg (n=6) bid Days 1–21 or 200 mg tid (n=2) Days 1–14] with decitabine (20 mg/m2/d IV Days 1–5) every 28 days. Results: Twenty-seven patients with AML have been treated. Median age was 67 years (range, 32–82 years). Median ECOG status 1 (range, 0 to 2). Eighteen patients (67%) had received prior therapy (median, 1 regimen; range, 0 to 4 regimens); 3 had received a prior allogeneic stem cell transplant. A total of 85 cycles have been administered, with a median of 2 cycles (range, 1 to 13 cycles); 10 patients (37%) have received 3 or more cycles of therapy. One of 7 patients treated at dose level 3 of the sequential schedule developed dose-limiting toxicities (DLT), consisting of grade 3 fatigue, weakness, and mucositis. Therefore, the MTD was not reached in the 3 planned dose levels of the sequential schedule. One DLT (grade 3 fatigue) occurred in 6 patients treated at dose level 2 of the concurrent schedule. Most common drug-related non-hematological toxicities of any grade (all CTCAE grades 1 or 2) were nausea (71%), fatigue (54%), diarrhea (54%), vomiting (42%), anorexia (25%), constipation (13%), abdominal pain (13%), dehydration (13%), and headache (13%). No other non-hematological grade 3 or 4 toxicities were observed. Of the 25 evaluable patients, one patient achieved an incomplete CR (without neutrophil recovery), one a morphologic leukemia-free state (without blood count recovery), and three partial remissions (1 achieved red cell transfusion independency and a second normalization of platelet counts). Seven of these patients remain on study for 2.7 to 13.5+ months. Correlative studies examining histone acetylation and gene promoter methylation in leukemic cells at baseline and after treatment, as well as plasma pharmacokinetic levels for both decitabine and vorinostat are being evaluated. Conclusions: The combination of decitabine and vorinostat is safe, well-tolerated, and has clinical activity in patients with AML. Enrollment is ongoing.

Capecitabine plus oxaliplatin (XELOX) followed by capecitabine plus irinotecan (XELIRI) in a sequential schedule in first-line metastatic colorectal cancer (MCRC): a phase II multicenter study

13516 Background: In phase II trials, XELOX and XELIRI were highly active and well tolerated in first-line MCRC. The aim of this study is to explore the efficacy and safety of XELOX followed by XELIRI as first-line treatment in MCRC. Specifically, we wanted to evaluate the impact of sequential scheduling on the dose-limiting neurotoxicity associated with oxaliplatin accumulation. Methods: Eligible patients (pts) had histologically or cytologically confirmed MCRC, ECOG PS ≤ 2 and adequate bone marrow, renal and hepatic function. Prior chemotherapy for MCRC was not allowed. Pts received 4 cycles of XELOX (capecitabine 1000mg/m2 orally bid d1–14 + oxaliplatin 130mg/m2 i.v. d1, q3w) followed by 4 cycles of XELIRI (capecitabine 1000mg/m2 bid d1–14 + irinotecan 240mg/m2 i.v. d1, q3w). This sequential schedule was repeated until unacceptable toxicity or disease progression. Results: Of the 35 pts analized to date: M/F (69%/31%); median age 68 years (range 41–78); ECOG PS 0–1 (94%); surgery (77%) and adjuvant chemotherapy (31%). 240 cycles (median 6, range 1–16) have been administered. 35 pts received XELOX (123 cycles, median 4), and 21 pts received XELIRI (83 cycles, median 4) in the first sequential schedule. In the second sequential schedule 6 pts received XELOX (22 cycles, median 4) and 4 pts received XELIRI (12 cycles, median 3.5). Median relative dose intensity was 88% for XEL, 96% for OX and 92% for IRI. In 27 efficacy evaluable pts, the ORR was 48% (95% CI, 29–67%). Eight pts were not evaluable due to adverse events (n=6), ongoing treatment (n=1) and lost of follow up (n=1). Conclusions: This sequential schedule is active and well tolerated in first-line MCRC. The improvement/recovery of the oxaliplatin-related neurotoxicity during the XELIRI phase is also promising and allows the re-treatment with oxapliplatin in the next sequence without accumulating neurotoxicity. Final results will be presented at the meeting. [Table: see text] No significant financial relationships to disclose.

Effect of paclitaxel-carboplatin (PC) consolidation chemotherapy after weekly PC concurrent chemo-radiotherapy (CCR) for patients with locally advanced non-small cell lung cancer (LA-NSCLC): 3-year definitive results of the B001-phase III GERCOR-study

7112 Background: Local control rate of LA-NSCLC seems better after concurrent CCR than after sequential schedule; the role of additional chemotherapy is not clearly defined. A multi-institutional phase-III trial was conducted to evaluate the role of chemotherapy consolidation after CCR Methods: Eligibility criteria included mediastinoscopy-controlled unresectable NSCLC stage IIIA (27%) - IIIB (59%) and inoperable mediastinal recurrence after surgery (14%), PS < 3, clinical target volume compatible with a minimal 60 Gy dose radiation. After registration, patients (pts) were treated with combination of weekly P (45 mg/m2), C (AUC 2), and radiotherapy 60–66 Gy (5 × 2 Gy per week). The pts with response or stable disease were randomized either to receive 3 cycles of P (175 mg/m2) and C (AUC 5) consolidation on days 1–22–43 or observation. Primary endpoint was OS (log-rank test), planned sample was 122 pts. Results: Actually, 71 pts were enrolled. Thirty pts (42%) were not randomized because of progression (22%) and disease-related death (22%), toxicity (26%), refusal (9%), protocol violation (21%). Toxicity grade 3–4 per patient: for the PC-TRT sequence; neutropenia 5%, febrile neutropenia 5%, thrombopenia 5%, pneumonitis 5%, oesophagitis 19%. For the PC consolidation sequence; neutropenia 24%, febrile neutropenia 6%, thrombopenia 0% (no treatment-related death). After a minimal follow-up of 3 years, despite poor inclusion rate, OS and PFS were greater in the PC consolidation group with 3-year OS: 29.9% vs 10% (HR = 0.45; CI 0.95: 0.22–0.91) (p = 0.002) and 3-year PFS: 27.8% vs 10% (HR = 0.6; CI 0.95: 0.3–1.3) (p = 0.17). Nine pts developed metastasis in each treatment arm. Conclusions: The addition of PC consolidation to PC-CCR is not easily feasible for all LA-NSCLC. For selected pts (only 58% of pts in this trial), despite premature ending of the trial because of slow accrual, PC consolidation significantly improved the 3-year OS and probably PFS. Because no difference in the metastatic incidence was observed, the effect of chemotherapy dosage in the consolidation arm could be explained by delaying metastasis. No significant financial relationships to disclose.