Multicentric, KSSG as participating partner
Genetic markers for opioid responses
This work package worked along three lines. The first approach (1) was to select single
nucleotide polymorphisms (SNPs) from genes supported by the literature to be candidates
for affecting opioid response; the second (2) to identify “new” genes of potential relevance
for opioid response by analyzing morphine-induced changes in global gene expression in a neuroblastoma cell line model system, and select SNPs from the most promising
candidates; and the third (3) to perform genome-wide SNP genotyping on pools of DNA
generated from “poor” and “good” opioid responders. Approaches (1) and (3) were based
on the availability of DNA and clinical data from 2294 palliative care cancer patients
treated with opioids for pain (European Pharmacogenetic Opioid Study, EPOS).
The work started by extracting and quantifying DNA from the EPOS samples. An SPSS
master file with patient demographics and clinical data was generated. Literature reviews
were performed to identify genes considered as candidates for affecting opioid response.
The 2262 EPOS samples were genotyped using the SNPlex technology (48-plexing) with
191 SNPs from 37 candidate genes, and the genotypes were entered into the EPOS
masterfile. Subsequent association and statistical analyses on 123 SNPs covering the 25
most revelant candidate genes were carried out on an explorative sample (2/3 of the EPOS
patients), and potentially positive associations were sought replicated in a validation sample
(1/3 of the EPOS patients).
DNA pools of “poor” and “good” opioid responders from the EPOS cohort were selected
on the basis of criteria defined by an expert group ("poor" responders had high pain
intensity scores and a high oral equivalent opioid dose; "good" responders had low pain
scores and a low oral equivalent opioid dose). The two pools were genotyped for >1 million
SNPs to detect differences between the two groups. SNPs showing different allele
frequencies between "good" and "poor" responders were validated by individual
genotyping in two independent cohorts and subject to bioinformatic analysis.
Approach (2) was abandoned as three independent gene expression experiments in cell
lines after morphine exposure showed only minor changes that could not be replicated.
Genetic markers for cachexia
The work started by a systematic literature review on candidate genes associated with
cancer cachexia. DNA was extracted from 855 cachexia samples from Edinburgh,
Montreal, and Edmonton, and a phenotype data-base was composed. DNA was genotyped
for candidate genetic markers for cachexia (129 SNPs in 80 genes). 775 samples (discovery
cohort) were available for bioinformatic analysis. The results were validated in an
independent cohort of 101 patients with cancer cachexia, from St Gallen (validation
cohort). In addition, gene group analysis based on functional similarity according to gene
ontology was performed in the discovery cohort.
Biomarkers for pain perception and treatment response
A protocol was developed for a clinical study to explore serum factors as biomarkers for
pain treatment. Due to the exploratory nature of the investigation it was decided to start
with a limited pilot study, and patients were included at one site only (MUG). The pilot
study showed promising results in 20 included patients, and it was decided to continue
patient recruitment. A total of 45 patients were included at MUG; 38 were eligible for final
analysis. In addition, 20 healthy volunteers were included. Serum samples were analysed
for all candidate biomarkers at BMS and biostatistical analysis was performed at MUG.
WP2: Assessment and classification of pain, depression, and cachexia
The three work packages on assessment and classification started by performing systematic
literature reviews and expert surveys to identify a set of items to classify and assess each of the three symptoms. Results from a Norwegian multi-centre computerised data collection
study on assessment of pain and physical function were used to inform the work. The next
step was a large, international empirical study to test the items. Protocol preparation and
software development for the computerised assessment tool were done in parallel. The
software included a body map for pain localisation.
Qualitative patient interviews were conducted to supplement the experts’ views on domains
The large, multi-centre data collection study (EPCRC-CSA, Computerized Symptom
Assessment of Pain, Depression and Physical Function in Palliative Care) was performed at
16 centres in eight countries and in four languages, yielding a total sample of 1070 patients.
The study closed in December 2009. Data has been analysed with respect to optimal items
and criteria for the assessment and classification of pain, depression, and cachexia. In
addition, data from two other, large data sets (2294 EPOS patients and 1801 patients
included in a prospective study at the Mario Negri Centre in Milan) has been analysed to
validate domains and items of the proposed pain classification.
Based on the software for the CSA study, Verdande Technology has used case-based
reasoning to develop a computerised assessment tool for diagnosis of depression in
A clinical study of 50 patients was performed to validate an activity monitoring system for
palliative care patients. Data from several studies was combined to study activity
monitoring in patients with cancer cachexia.
The work on the EPCRC guidelines was performed according to a common protocol based
on the NICE and SIGN guideline methodology. Systematic literature reviews and expert
opinion input through Delphi procedures (including patient representation) complemented
each other in the process. In addition, feedback from the general palliative care community
was obtained through the EPCRC and EAPC websites and through workshops at the EAPC
A group of 20 European experts were engaged to do systematic reviews for the pain
guidelines. The 23 resulting reviews were presented and discussed at the 5th Bristol Opioid
Conference in February 2010, and form the basis for the recommendations on opioid
All guideline groups have developed patient summaries of the recommendations.
1. Norwegian University of Science and Technology, Trondheim, Norway NTNU
2. Bristol Haematology and Oncology Centre, University of Bristol,
Bristol, UK UNIVBRIS
3. Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy INT
4. King’s College London, London, UK KCL
5. Cantonal Hospital, St. Gallen, Switzerland KSSG
6. RWTH Aachen University, Aachen, Germany RWTH
7. University of Edinburgh, Edinburgh, UK UEDIN
8. Medical University of Graz, Graz, Austria MUG
9. Trollhetta AS, Trondheim, Norway TROLL
10. Bender MedSystems GmbH, Vienna, Austria BMS
11. Verdande Technology AS, Trondheim, Norway VERD
Results and IMpact:
see pdf (request from PI)