Can(Plan)+: Extending the Operational Semantics of the BDI Architecture to deal with Uncertain Information

The BDI architecture, where agents are modelled based on their beliefs, desires and intentions, provides a practical approach to develop large scale systems. However, it is not well suited to model complex Supervisory Control And Data Acquisition (SCADA) systems pervaded by uncertainty. In this paper we address this issue by extending the operational semantics of Can(Plan) into Can(Plan)+. We start by modelling the beliefs of an agent as a set of epistemic states where each state, possibly using a different representation, models part of the agent's beliefs. These epistemic states are stratified to make them commensurable and to reason about the uncertain beliefs of the agent. The syntax and semantics of a BDI agent are extended accordingly and we identify fragments with computationally efficient semantics. Finally, we examine how primitive actions are affected by uncertainty and we define an appropriate form of lookahead planning.

Probabilistic Planning in AgentSpeak using the POMDP framework.

AgentSpeak is a logic-based programming language, based on the Belief-Desire-Intention (BDI) paradigm, suitable for building complex agent-based systems. To limit the computational complexity, agents in AgentSpeak rely on a plan library to reduce the planning problem to the much simpler problem of plan selection. However, such a plan library is often inadequate when an agent is situated in an uncertain environment. In this paper, we propose the AgentSpeak+ framework, which extends AgentSpeak with a mechanism for probabilistic planning. The beliefs of an AgentSpeak+ agent are represented using epistemic states to allow an agent to reason about its uncertain observations and the uncertain effects of its actions. Each epistemic state consists of a POMDP, used to encode the agent’s knowledge of the environment, and its associated probability distribution (or belief state). In addition, the POMDP is used to select the optimal actions for achieving a given goal, even when facing uncertainty.

A new look towards BAC-based array CGH through a comprehensive comparison with oligo-based array CGH

BACKGROUND: Currently, two main technologies are used for screening of DNA copy number; the BAC (Bacterial Artificial Chromosome) and the recently developed oligonucleotide-based CGH (Chromosomal Comparative Genomic Hybridization) arrays which are capable of detecting small genomic regions with amplification or deletion. The correlation as well as the discriminative power of these platforms has never been compared statistically on a significant set of human patient samples.

RESULTS: In this paper, we present an exhaustive comparison between the two CGH platforms, undertaken at two independent sites using the same batch of DNA from 19 advanced prostate cancers. The comparison was performed directly on the raw data and a significant correlation was found between the two platforms. The correlation was greatly improved when the data were averaged over large chromosomic regions using a segmentation algorithm. In addition, this analysis has enabled the development of a statistical model to discriminate BAC outliers that might indicate microevents. These microevents were validated by the oligo platform results.

CONCLUSION: This article presents a genome-wide statistical validation of the oligo array platform on a large set of patient samples and demonstrates statistically its superiority over the BAC platform for the Identification of chromosomic events. Taking advantage of a large set of human samples treated by the two technologies, a statistical model has been developed to show that the BAC platform could also detect microevents.

Radium-223 and concomitant therapies in patients with metastatic castration-resistant prostate cancer: an international, early access, open-label, single-arm phase 3b trial

BACKGROUND: In the previously reported ALSYMPCA trial in patients with castration-resistant prostate cancer and symptomatic bone metastases, overall survival was significantly longer in patients treated with radium-223 dichloride (radium-223) than in patients treated with placebo. In this study, we investigated safety and overall survival in radium-223 treated patients in an early access programme done after the ALSYMPCA study and before regulatory approval of radium-223.

METHODS: We did an international, prospective, interventional, open-label, single-arm, phase 3b study. Enrolled patients were aged 18 years or older with histologically or cytologically confirmed progressive bone-predominant metastatic castration-resistant prostate cancer with two or more skeletal metastases on imaging (with no restriction as to whether they were symptomatic or asymptomatic; without visceral disease but lymph node metastases were allowed). Patients received intravenous injections of radium-223, 50 kBq/kg (current recommendation 55 kBq/kg after implementation of National Institute of Standards and Technology update on April 18, 2016) every 4 weeks for up to six injections. Other concomitant anticancer therapies were allowed. Primary endpoints were safety and overall survival. The safety and efficacy analyses were done on all patients who received at least one dose of the study drug. The study has been completed, and we report the final analysis here. This study is registered with ClinicalTrials.gov, number NCT01618370, and the European Union Clinical Trials Register, EudraCT number 2012-000075-16.

FINDINGS: Between July 22, 2012, and Dec 19, 2013, 839 patients were enrolled from 113 sites in 14 countries. 696 patients received one or more doses of radium-223; 403 (58%) of these patients had all six planned injections. Any-grade treatment-emergent adverse events occurred in 523 (75%) of 696 patients; any-grade treatment-emergent adverse events deemed to be related to treatment were reported in 281 (40%) patients. The most common grade 3 or worse treatment-related treatment-emergent adverse events were anaemia in 32 (5%) patients, thrombocytopenia in 15 (2%) patients, neutropenia in ten (1%) patients, and leucopenia in nine (1%) patients. Any grade of serious adverse events were reported in 243 (35%) patients. Median follow-up was 7·5 months (IQR 5-11) and 210 deaths were reported; median overall survival was 16 months (95% CI 13-not available [NA]). In an exploratory analysis of overall survival with predefined factors, median overall survival was longer for: patients with baseline alkaline phosphatase concentration less than the upper limit of normal (ULN; median NA, 95% CI 16 months-NA) than for patients with an alkaline phosphatase concentration equal to or greater than the ULN (median 12 months, 11-15); patients with baseline haemoglobin levels 10 g/dL or greater (median 17 months, 14-NA) than for patients with haemoglobin levels less than 10 g/dL (median 10 months, 8-14); patients with a baseline Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 (median NA, 17 months-NA) than for patients with an ECOG PS of 1 (median 13 months, 11-NA) or an ECOG PS of 2 or more (median 7 months, 5-11); and for patients with no reported baseline pain (median NA, 16 months-NA) than for those with mild pain (median 14 months, 13-NA) or moderate-severe pain (median 11 months, 9-13). Median overall survival was also longer in patients who received radium-223 plus abiraterone, enzalutamide, or both (median NA, 95% CI 16 months-NA) than in those who did not receive these agents (median 13 months, 12-16), and in patients who received radium-223 plus denosumab (median NA, 15 months-NA) than in patients who received radium-223 without denosumab (median 13 months, 12-NA).

INTERPRETATION: Our findings show that radium-223 can be safely combined with abiraterone or enzalutamide, which are now both part of the standard of care for patients with metastatic castration-resistant prostate cancer. Furthermore, our findings extend to patients who were asymptomatic at baseline, unlike those enrolled in the pivotal ALSYMPCA study. The findings of prolonged survival in patients treated with concomitant abiraterone, enzalutamide, or denosumab require confirmation in prospective randomised trials.

FUNDING: Pharmaceutical Division of Bayer.