Portal de Periódicos da CAPES

Assessing symptom profiles in neuropathic pain clinical trials: Can it improve outcome?

2011; Wiley; Volume: 15; Issue: 5 Linguagem: Inglês

10.1016/j.ejpain.2011.03.005

1532-2149

Nadine Attall, Didier Bouhassira, Ralf Baron, Jonathan O. Dostrovsky, Robert H. Dworkin, Nanna Brix Finnerup, Geoffrey K. Gourlay, Maija Haanpää, Srinivasa N. Raja, Andrew S.C. Rice, David M. Simpson, Rolf‐Detlef Treede,

Peripheral Neuropathies and Disorders

Neuropathic pain (NP) encompasses a large variety of conditions associated with a lesion or disease of the peripheral or central somatosensory system and remains particularly difficult to treat. Recent meta-analyses have indicated that on the average, about one third more patients with NP respond to active drugs compared to placebo treatment, and in many trials no superiority of the test compound above placebo was demonstrated (Finnerup et al., 2010). Although several reasons may be proposed to account for such findings, including a large placebo response or the fact that the current drugs do not act on the relevant pathophysiological targets, it is also possible that trials failed to identify responder populations because they did not take into account the heterogeneity of NP symptoms and signs that are presumably linked to distinct mechanisms (e.g., spontaneous pain such as burning, the most prominent symptom of NP, but also paroxysmal pains such as electric shocks, or evoked pains to mechanical/thermal stimuli). This could also be one reason for the recent negative results of drug trials in conditions considered as refractory, such as HIV neuropathies or lumbar radiculopathy (Attal et al., 2010). These observations have led to suggest as early as in 1990 that a preferable therapeutic approach to NP might be based on the specific mechanisms underlying the specific NP pain being treated rather than on the aetiological conditions, leading to targeted treatments of these mechanisms (e.g. Attal et al., 2008; Finnerup and Jensen, 2006; Max, 1990; Rowbotham, 2005; Woolf et al., 1998; Woolf, 2004). However such a rational approach is attractive but does not yet seem to be attainable, mainly because of the difficulties of translating in the clinic the pathophysiological mechanisms identified in animal studies (Bouhassira and Attal, 2004; Finnerup and Jensen, 2006; Hansson, 2003). Thus, a realistic and efficient therapeutic approach to overcome these obstacles should focus mainly on clinically observable manifestations of pathophysiology, i.e. on a comprehensive symptomatic assessment of patients with NP. Importantly it has been shown that specific pain symptoms or their combinations may provide with relevant information about underlying mechanisms (e.g. Hatem et al., 2010; Truini et al., 2010). The assessment of symptoms and signs can be best achieved with validated questionnaires such as the Neuropathic Pain Scale (NPS) or the Neuropathic Pain Symptom Inventory (NPSI) as regards symptoms, and with an extension of the clinical examination such as quantitative sensory testing (QST) for sensory signs (Refs. in Haanpaa et al., 2011). The advantage of a symptom based classification of NP has recently been outlined by three large scale clinical studies. One study performed in 482 patients with peripheral or central neuropathic conditions using the NPSI questionnaire failed to identify significant associations between symptoms or dimensions and aetiologies (Attal et al., 2008), suggesting the relevance to group aetiologically distinct conditions for therapeutic trials (Hansson and Dickenson, 2005). The two other studies confirmed the heterogeneity of sensory profiles (symptoms and signs) in patients within the same aetiological condition, which might indicate the involvement of distinct mechanisms. Thus, based on the PainDetect questionnaire, Baron and colleagues identified five different clusters of patients corresponding to various symptom combinations in 2100 patients with postherpetic neuralgia or painful diabetic polyneuropathy (Baron et al., 2009). All clusters occurred in relevant numbers in both diagnoses but their frequencies differed between postherpetic neuralgia and diabetic NP. Using quantitative sensory testing (QST) in 1236 NP patients, Maier and colleagues also found that QST profiles (i.e., loss of mechanoreception, thermoreception or nociception, gain in nociception) were heterogeneous accross the spectrum of NP conditions (Maier et al., 2010). Phenotypic profiling in clinical trials has mainly contributed to date to characterize the effects of treatments on sensory signs and symptoms of NP. Extensive sensory profiling using QST has been achieved mainly with intravenous tests of the sodium channel blocker lidocaine, opiates or the NMDA antagonist ketamine, which are not recommended as first line in clinical use (Attal et al., 2010). These studies found efficacy of these drugs on dynamic mechanical allodynia and less consistent effects on pinprick hyperalgesia, cold allodynia or temporal summation (e.g. Gottrup et al., 2006). In contrast, most trials using first line medications (e.g., tricyclic antidepressants, pregabalin, topical lidocaine) only assessed dynamic mechanical allodynia (Finnerup and Jensen, 2006). Although they found a significant alleviation of this sensory sign, it is not established whether this effect is specific or not in the lack of full sensory profiling. Thus there is evidence from proof of concept studies that NP medications can influence sensory signs in humans, as is the case in animals, but we still have insufficient information as to which first line medications act and on which sensory signs. Regarding self reported symptoms, drugs such as opioids, gabapentin or cannabinoids have been reported to induce differential effects on pain descriptors (with consistent effects on the items "sensitive" or "dull") assessed using the NPS (Refs. in Haanpaa et al., 2011). One study using botulinum toxin type A found that the drug relieved only some dimensions assessed with the NPSI (burning pain, paroxysmal pain, evoked pains but not deep pain or paresthesia) (Ranoux et al., 2008). A recent study comparing the effects of transcranial direct cortical stimulation combined with visual illusion to single treatments in patients with spinal cord injury found that combined treatments alleviated all dimensions of the NPSI (continuous pain, pain paroxysms, evoked pains, paresthesia/dysesthesia) while single treatments improved some dimensions only (continuous pain and paroxysmal pain for the cortical stimulation group, continuous pain and paresthesia/dysesthesia for the visual illusion group) (Soler et al., 2010). Although these data need to be replicated, they suggest the relevance of a comprehensive assessment of the patientś symptom profiles in clinical trials. A second and probably more important contribution of phenotypic profiling is to increase therapeutic prediction. This implies the identification of relevant clinical criteria allowing classification of patients into several subgroups, with the assumption that these groups have different underlying pain mechanisms and hence will respond differentially to treatments. A few randomized controlled studies have identified predictive factors of the responses to treatments of NP generally based on posthoc analysis. In particular it has been found that patients with mechanical (static or dynamic) allodynia were better responders to systemic lidocaine or to lamotrigine than those without allodynia (e.g., Attal et al., 2004; Finnerup et al., 2002; see however Finnerup et al., 2005). A recent study using pregabalin in HIV neuropathy was negative on the primary outcome but posthoc analyses indicated that the drug was superior to placebo in patients with severe mechanical (punctate) hyperalgesia (Simpson et al., 2010). These observations are in keeping with the results of recent studies suggesting that different mechanisms are at play in neuropathic patients with spontaneous pain only and those with superimposed allodynia (e.g., Hatem et al., 2010; Rowbotham, 2005). They suggest the relevance to differentiate patients with/without mechanical allodynia for therapeutic studies. Preservation of thermal sensation has also been associated with a better response to botulinum toxin A in patients with peripheral neuropathic pain (e.g. Ranoux et al., 2008) or to electrical motor cortex stimulation in central pain patients (Drouot et al., 2002); conversely loss of heat pain sensitivity has been found predictive of the response to opioids in patients with postherpetic neuralgia (Edwards et al., 2006). However other studies failed to detect any predictor of the response to drugs such as topical lidocaine (Herrmann et al., 2006) or found unexpected results (Wasner et al., 2005). Since most of these data are based on small trials, large scale studies of QST using standardized assessment are now warranted. From the above observations, we recommend that clinical trials include patients who have different NP aetiologies, but whose clinical phenotypes have been carefully characterized. Nevertheless it appears impossible to determine a priori which of the patients clinical phenotypes may best respond to a treatment. Rather it is more realistic to determine whether profiles of responders may be identified in large scale trials based on posthoc analyses. For this purpose, many secondary endpoints need to be measured and analysed as exploratory. Analysis of the results of these studies should help determine the design of subsequent trials aiming to validate the potential prediction prospectively. The use of validated NP assessment questionnaires and the standardization of sensory testing should constitute major steps forward. In particular it now becomes critical to perform appropriately powered multicenter studies of QST. Since such studies are extremely difficult to implement with a comprehensive range of tests, an optimized range of QST tests suitable for large scale pharmacological studies should be developed (see for example Nurmikko et al., 2007). It will also be important to use new compounds with established mechanisms of action and targeting specific mechanisms. Results obtained with these approaches may be crucial to improve the prediction of a therapeutic response and should contribute to the development of a mechanism based therapeutic strategy of NP. Nadine Attal has reveived consultancy or lecturing fees in the past 12 months from Grunenthal, Pfizer Inc., Eli Lilly, Astellas, Medtronic, Eisai, Biogen. Didier Bouhasira has served on the Speakers' Bureau for Eli Lilly and Company and Pfizer Inc, has worked as a consultant to Eli Lilly, Grünenthal, Newron Pharmaceuticals, Pfizer, Sanofi-Aventis, Sanofi-Pasteur-MSD, Johnson & Johnson, and Schering-Plough and has received research support from Pfizer Inc. Ralf Baron has received consultancy fees from Pfizer, Genzyme, Grunenthal, Mundipharma, Allergan, Sanofi Pasteur, Medtronic, Eisai, UCB Biosciences, Lilly, Boeringher Ingelheim, Astellas, Novartis and research/grant support from Pfizer, Genzyme and Grunenthal. He is a member of the IMI "Europain" collaboratio and industry members of this are: Astra Zeneca, Pfizer, Esteve, UCB-pharmac, Sanofi Aventis, Grunenthal, Eli Lilly, Neuroscience Technologies and Boehringer Ingelheim. Jonathan Dostrovsky has no conflicts of interest. Robert H. Dworkin, PhD, has received in the past 12 months research support from Montel Williams Foundation, Serono, and US Food and Drug Administration, and compensation for consulting from Acura, Analgesic Solutions, Astellas, Dainippon Sumitomo, Endo, Flexion, Forest, Grünenthal, Infinity, Inhibitex, Jazz, Johnson & Johnson, King, Lilly, McNeil, MediciNova, Merck, NeurogesX, Ono, Seikagaku, Smith & Nephew, Spinifex, and US Department of Veterans Affairs. Nanna Finnerup has received in the past 12 months consultancy fees from Grunenthal. She is a member of the IMI "Europain" collaboration and industry members of this are: Astra Zeneca, Pfizer, Esteve, UCB-Pharma, Sanofi Aventis, Grunenthal, Eli Lilly, Neuroscience Technologies and Boehringer Ingelheim. Geoffrey Gourlay has been part of the speakers bureau for Mundipharma in relation to Targin. Maija Haanpää has received consultancy fees and lectures fees from Abbott, Astellas, Eli Lilly, Medtronic, MSD, Mundipharma, Pfizer and Sanofi-Pasteur and travel expenses from Astellas, Boehringer-Inhgelheim, Medtronic and Pfizer. Andrew SC Rice has received consulting fees or honoraria in the past year via Imperial College Consultants from Pfizer, Allergan, Esteve, Spinifex and research support from Pfizer. He is a member of the IMI "Europain "collaboration and industry members of this are: Astra Zeneca, Pfizer, Esteve, UCB-Pharma, Sanofi Aventis, Grunenthal, Eli Lilly, Neuroscience Technologies and Boehringer Ingelheim. David M Simpson MD has served as consultant for Astellas, NeurogeX, Eli Lilly, Astra Zeneca, and Allergan, speaker honoraria from Eli Lilly, Astellas, Pfizer and Allergan, and research grant support from Astra Zeneca, Pfizer, and Allergan. Rolf-Detlef Treede has received research grants from or given lectures for Grunenthal, Astellast, Nycomed, Pfizer, Bohringer Ingelheim, UCB, Lilly (in the years 2009–2010). Christopher Wells in the last 12 months has received fees for speaking at meetings from Napp (sister company to Purdue), Pfizer and Grunenthal. He has serving on a Grunenthal advisory board regarding Tapentadol.