Vesa K. Kontinen

Animal models and the prediction of efficacy in clinical trials of analgesic drugs : A critical appraisal and call for uniform reporting standards

a Pain Research, Department of Anaesthetics, Pain Medicine and Intensive Care, Faculty of Medicine, Imperial College London, Chelsea & Westminster Hospital Campus, 369 Fulham Road, London SW10 9NH, UK b Mari Lowe Centre for Comparative Oncology Research, School of Veterinary Medicine, University of Pennsylvania, 3900 Delancey Street, Philadelphia, PA 19104-6010, USA c Department of Anesthesia, Wake Forest University School of Medicine, Medical Center Boulevard, Winston Salem, NC 27157, USA d Department of Anaesthesiology and Intensive Care Medicine, Helsinki University Central Hospital, Helsinki, Finland e Pfizer Global Research and Development, Sandwich CT13 9NJ, UK f Department of Psychology, McGill University, 1205 Dr. Penfield Avenue, Montreal, Que., Canada H3A 1B1 g Bitop AG, Stockumerstr. 28, 58453 Witten, Germany

Does the pharmacology of oxycodone justify its increasing use as an analgesic

Oxycodone is a semisynthetic opioid analgesic that is increasingly used for the treatment of acute, cancer, and chronic non-malignant pain. Oxycodone was synthesized in 1917 but its pharmacological properties were not thoroughly studied until recently. Oxycodone is a fairly selective μ-opioid receptor agonist, but there is a striking discrepancy between the relatively low binding potential and G protein activation by oxycodone and its analgesic efficacy. It has been claimed that this is because of active metabolites and enhanced passage to the central nervous system by active transport. We critically review studies on the basic pharmacology of oxycodone and on its pharmacokinetics and pharmacodynamics in humans. In particular, the role of pharmacogenomics and population pharmacokinetics in understanding the properties of oxycodone is discussed in detail. We compare oxycodone with morphine, the standard opioid in clinical use.

The effect of paracetamol and tropisetron on pain: experimental studies and a review of published data.

Experimental studies suggest that paracetamol‐induced analgesia is mediated via central serotonergic pathways and attenuated by 5‐HT3‐antagonists. However, clinical studies do not support this, and 5‐HT3‐antagonists are expected to reduce pain by blocking the descending pronociceptive pathway. The current project tested whether tropisetron attenuates analgesia by paracetamol. Two randomized, double‐blind, crossover studies with 18 healthy male volunteers in each were performed. Pain stimuli were cold water immersion (cold pressor test), contact heat pain (study 1) and electrical stimulation (study 2). In both studies, tropisetron 5 mg i.v. or saline was administered, followed by paracetamol 2 g i.v. 30 min. later. Individual changes in heat and cold pain intensity, cold pain tolerance and unpleasantness were recorded. The same thresholds were also expressed as scores (% of the individual score at baseline). Additionally, previously published findings on the effects of paracetamol and its interaction with 5HT3‐antagonists in human experimental pain models were reviewed. After calculation of the sensory and pain scores (%), tropisetron seemed to amplify the analgesic action of paracetamol. Paracetamol 2 g i.v. did not show any statistically significant analgesia in thermal tests (study 1), or differences in sensory, pain detection or moderate pain thresholds of the electrical stimulus (study 2). As paracetamol did not have a measurable analgesic effect in these tests, no conclusions can be drawn about the interaction between paracetamol and tropisetron. However, tropisetron may have an analgesic effect of its own. Clinicians should not avoid using these drugs together, unless larger clinical studies indicate otherwise.

Why we are proud to publish well-performed negative clinical studies?

In the present issue of the Scandinavian Journal of Pain the aper “Evaluation of the analgesic efficacy of AZD1940, a novel annabinoid agonist, onpost-operative pain after lower thirdmolar xtraction” by Jarkko Kalliomaki et al. [1] describes the effect of peripherally active CB1/CB2-receptor agonist in the most comonly used model of postoperative pain. The compound is novel, he methodology is impeccable, the paper is very well written, and he topic is hot, and yet everyone in the business knows that it is ard to get it published. And this is only because the drug was not ffective. Most researchers have at some point struggled to publish negative study.

From clear reporting to better research models

In clinical research, today guidelines such as the CONSORT for eporting individual randomised controlled trials and the PRISMA tatement on conducting systematic reviews have become elemenary tools to improve reporting quality. In experimental animal esearch, we have been too slow to accept that similar sources f bias are present. However, it has been shown that also in the nimal laboratory methodological shortcomings can have signifiant effects on the findings [1–3]. Some of these methodological roblems can be eliminated by adopting simple research practices, uch as the routine of properly randomising the subjects to the tudy groups, or blinding the observer to the treatment when any ven relatively subjective endpoints are used. The reporting can lso be made significantly more transparent by, e.g. clearly stating ow many repetitions were actually performed, and if there were ny dropouts. These are issues, which have been important for this ournal already from the very beginning [4]. To further commit to this development, editorial board of Scaninavian Journal of Pain has decided to ask authors of experimental nimals studies submitted to the journal to adhere to the ARRIVE uidelines (http://www.nc3rs.org.uk/page.asp?id=1357). We are he first pain journal to do this, among good company, such as the ature and the PLoS journal families and Journal of Physiology. The ajor funding bodies of biomedical research especially in the Unied Kingdom are also endorsing the guideline. The special hybrid ournal format of Scandinavian Journal of Pain makes it feasible to nclude all the necessary details of the experimental conditions ithout problems with space in the print journal or including addiional web resources to an original article. It is important to realise that the ARRIVE guideline does not imit the scientific freedom of researchers, nor force to use any iven practice. The only thing that is required is to clearly and onestly report what has been done. A checklist of twenty items elps authors, editors and readers to get an unobstructed view of hat are the methodological strengths and weak points of research aper. In future, it will also be possible to compare experimental

Treatment of post dural puncture headache: To patch or not to patch?

These have traditionally been treatment with autologous EBP, or the “wait and see” strategy, combined with analgesics, coffee or cola-drinks. Often the headache will gradually be alleviated even with the conservative strategy, and there are certain risks associated with any invasive procedure, even EBP. Are there situations when it is beneficial or even necessary to administer the EBP, or seek alternatives such as the fibrin sealant to stop the leakage of the cerebrospinal fluid?

How good is a model

One of the most famous paintings by Belgian surrealist artistRene Magritte (1898–1967) is an image of a smoking pipe withthe text “Ceci n’est past une pipe” under it. Indeed, an image ofa pipe is not a pipe. When the painter was asked about this, hereplied: “Obviously it is not a pipe, just try to fill it with tobacco!”.This would represent the face validity aspect of a model: does themodel seem to function as the situation modelled? In the presentissue of Scand J Pain there are two papers from the research groupof Per Hartvig Honore that go one level deeper on the validity ofneuropathic pain models. The first one “Neuropathic pain modelsin the development of analgesic drugs” [1] discusses the constructvalidity of the models: are the models biologically representativetotheconditionsstudied?Thesecondpaper[2]ismorespecificallyon the predictive validity of the neuropathic pain models for dif-ferent drug groups. For development of (human) drugs this is thehighestlevelofvalidity:ifsomethingworksinthemodel,itshouldalsoworkinthepatients,andifanothermoleculefailsintheanimaltesting,itmaynotbeapossiblecurewasted.Inrationaldrugdevel-opment,anycompoundthatfailsintheearlypreclinicalphasewillnever be administered to humans. On the other hand, this limitsthe information available on the specificity of these models.In the IASP World Congress in San Diego 2002, we looked atthe predictive validity of the most commonly used neuropathicpain models [3]. In a systematic review of published evidence, thepicture looked rosy, even too good to be true. Since then, the phar-macological horizon for neuropathic pain treatments has widenedtremendously, and with the new mechanisms and biological sys-temsinvolvedthequestionofpredictivevalidityhasbecomemuchmore difficult. In addition, the old problems remain. Clinical effi-cacyisclearlynotayes/no—question,butadichotomousanswerisneeded for quantitative assessment of validity. Most clinical stud-ies of neuropathic pain are performed in postherpetic neuralgiaor painful diabetic neuropathy, but surgical nerve lesions are usedin many laboratory models. Mainly evoked responses are mea-suredintheanimalmodels,butalsoongoingandspontaneouspainare important for the patient. In clinical practice, patients satis-fiedwithtricyclicantidepressantsforneuropathicpainmayreportthat their pain intensity has not changed much, but sleep qualityhas improved significantly. This aspect has been studied very little

Of mice and men : What can we predict from the effects of morphine in a mouse model of bone cancer?

The present issue of Pain brings us news that we did not want to hear [3]. In a mouse model of cancer pain, continuous high-dose infusions of morphine increased evoked, but not spontaneous, pain behaviour in control animals, and both spontaneous and evoked pain behaviour in mice that had been implanted with sarcoma cells. Administration of high doses of morphine increased the expression of substance P and calcitonin G related peptide (CGRP) in the dorsal root ganglion (DRG) in both control and sarcoma mice. In addition, high doses of morphine increased sarcoma-induced expression of activating transcription factor 3 (ATF3), a marker for neuronal damage, in the DRG. Morphine also accelerated sarcoma-induced bone destruction and doubled the incidence of spontaneous bone fractures. However, it did not alter tumour growth in vitro or in vivo. Morphine increased osteoclast activity and up-regulated interleukin-1b in the sarcoma mice suggesting that the bone loss accelerated by morphine resulted from changes in bone metabolism. So, should we shoot the messenger, ignore the study, or blame some technical aspects in order to avoid considering the possible consequences of these results? Should we stop using morphine in the treatment of cancer pain? No. This study should make us re-evaluate the methods that we use in the management of cancer-related pain. However, we must keep in mind that there is a long way from a single experimental animal study to a balanced analysis of the clinical evidence of the efficacy and safety of different analgesics and administration schemes in treating pain caused by a metastatic malignancy. Several questions need to be considered regarding the experimental model and possible species differences, the equivalent doses of morphine for mice and men being one of the most crucial ones. The current study assessed only morphine. However, recent evidence suggests that different l-opioid receptor agonists and their metabolites have interesting pharmacological differences [9,11]. The role of non-opioid analgesics, both alone and in combination with opioids, needs to be studied.