Dialog, Dialog, Dialog: Strategies to Facilitate Translation

The development of new and effective analgesics requires ongoing, productive back-and-forth communication between basic scientists and clinicians

Although pain researchers have put great effort into finding new treatments for chronic pain, the translation of basic science discoveries into new therapies for patients has lagged, for the most part. At the inaugural North American Pain School (NAPS), Michael Gold, University of Pittsburgh, US, presented barriers hindering successful translation, and offered potential solutions to overcome them. He made the case that while translation has faltered for a number of reasons—ranging from the experimental approaches and assays that basic scientists use to an under-appreciation of the problems that patients face—a common thread underlying these obstacles is a lack of dialog between basic scientists and clinicians. Building bridges between these all-too-often separate communities is crucial to bring new treatments for chronic pain into the clinic.

What are the barriers?
The first barrier to translation that Gold described is that the model systems used in preclinical pain research to answer similar questions often differ from study to study, making it difficult to interpret experimental results. Here, Gold pointed to work from his group and that of others examining the activity of nicotinic acetylcholine receptors (Talavera et al, 2009 PMID: 19749751). This research has made use of a number of model systems, including heterologous expression systems, mouse nociceptive neurons, rat nociceptive neurons, and in Gold’s unpublished work, human dorsal root ganglion neurons. In the end, findings from studies like these may differ or conflict in large part because of the choice of model system. In the case of nicotinic acetylcholine receptors, there is increasing recognition of species differences in the density and distribution of the receptors.

The second roadblock is that the traditional behavioral assays used in preclinical studies—measures of evoked hypersensitivity—may not be best suited for understanding certain questions, such as whether a drug will be effective for pain or not. For example, Gold pointed to a 2009 rat study where the analgesia produced by clonidine or adenosine in neuropathic pain was assessed using von Frey testing or a conditioned place preference paradigm, which serves as a measure of ongoing pain (King et al., 2009). The study found that while clonidine relieved both evoked hypersensitivity and ongoing pain, adenosine relieved only evoked hypersensitivity. These results are particularly relevant clinically, Gold said, because clonidine has been shown to alleviate neuropathic pain in people, whereas adenosine has not.

On a related note, a third stumbling block to translation is that the actual problems that patients have—like ongoing pain—are not guiding preclinical pain research. For instance, Gold pointed to a clinical study of neuropathic pain finding that 96% of patients reported ongoing pain, but only 64% reported mechanical hypersensitivity, and only 38% reported heat hypersensitivity (Backonja and Stacey, 2004 PMID: 15556827). “Is it really relevant to look at mechanical or thermal hypersensitivity if you are interested in the neuropathic pain patient?” Gold asked the NAPS audience to consider.

The diversity of patient populations also makes translation a difficult proposition. To underscore this challenge, Gold referred to a study of over 2,000 patients comparing the phenotypes of two neuropathic pain conditions, diabetic peripheral neuropathy (DPN) and postherpetic neuralgia (PHN; Baron et al., 2009 PMID: 19592166). The classification of patients into those with PHN, and those with DPN, has been used to “limit” variability in clinical trials in which patients were enrolled with one condition or the other, or to increase the pool of neuropathic pain patients in clinical trials in which patients with either form of neuropathy were enrolled, Gold noted. But, as the authors of the study emphasized, Gold said, both strategies are flawed, because both PHN and DPN patients are highly heterogeneous. A more appropriate strategy, Gold said, is to classify subjects based on the pattern of symptoms they have, which likely reflects different underlying mechanisms—and consequently can be better used for trials and treatment. Gold also stressed another point: underlying mechanisms of neuropathic pain vary as a function of a variety of factors; the cause of the injury is just one. Consequently, developing a strategy to treat PHN pain will likely help only a subpopulation of patients.

Finally, Gold highlighted systemic problems in science, particularly the dwindling numbers of clinician-scientists, as another impediment to translation. Physicians today often do not do research, while basic scientists may have little connection to the clinical world.

What are the solutions?
With regard to the model systems used in preclinical pain research, Gold said they still have an important role to serve, but stressed the benefits of using a variety of model systems in a study (e.g., various heterologous cell lines, and more than one animal species) to rigorously test hypotheses. Still, use of human tissue will have greater translational relevance. Here, Gold emphasized the importance of having a dialog with colleagues at one’s institution, including physicians, researchers, and administrators, which will help researchers obtain human tissue and also spur collaboration.

Similarly, Gold said that while preclinical pain assays have limitations, they still can play an important role in pain research. Because each assay has its own strengths and weaknesses, using multiple assays is important. Researchers should choose assays based on their relevance to the problem of interest and ability to address the right questions, he said.

To show that the third barrier to translation can be overcome—that clinical observations can indeed successfully guide preclinical research—Gold provided examples from his own lab, including studies on the mechanism of action of the anti-migraine agent, sumatriptan. The starting point for these studies was the clinical observation that triptans, which are serotonin receptor agonists, are surprisingly selective for alleviating only migraine pain. Given that serotonin receptors are distributed throughout the body, Gold’s lab became interested in where this selectivity comes from.

The researchers hypothesized that there must be something unique about the dura mater, the outermost layer of the meninges that surround the brain and a pain-sensitive structure thought to contribute to migraine, or about the neurons that innervate it.
Gold’s group discovered that 5-HT1D serotonin receptors were differently distributed; the density of receptors in tissues responsible for migraine pain, including the dura, was much higher than in structures where triptans have no anti-nociceptive effect (Harriott and Gold, 2008, PMID: 18557979). They also observed that the receptor was present in sympathetic neurons innervating the dura, suggesting an additional target for triptan activity. Finally, they documented post-translational modification of the 5-HT1D receptor within the dura, which provided an additional explanation of how triptans could work specifically in that tissue. In sum, this work showed how a clinical observation could successfully direct basic science studies.

Finally, as for the complexity of patient populations, and the disappearance of the clinician-scientist, in both cases Gold again emphasized the importance of dialog and collaboration between clinicians and basic scientists. While this may not make up for the loss of the clinician-scientist, who uses clinical observations to guide the questions asked in the lab and uses observations in the lab to guide treatment strategies, dialog and collaboration do facilitate this process for the clinician and the basic scientist. One way for basic scientists to find clinicians with whom to collaborate is to attend grand rounds, while clinicians can find the basic scientists through journal clubs and lab meetings.

Ram Kandasamy is a fifth-year PhD candidate in neuroscience examining opioid and cannabinoid analgesia at Washington State University, Vancouver, Washington, US.