The last 20 years has seen spectacular advances in our understanding of molecular mechanisms of pain. More than 200 relevant pain genes have been identified [e.g. Wistrom et al. (2022)], the molecular nature of transduction and transmission of sensory signals in primary nociceptors is being unraveled [e.g. Patapoutian et al. (2009)] and many of the neurotransmitters and receptors that modulate pain-related signals in the central nervous system are now known [e.g. Heinricher and Fields (2013)]. Several mutations affecting pain processing have been described, and there are large scale studies underway exploring the genetic variants that might have more subtle effects on pain processing and the epigenetic processes that might regulate them (Dib-Hajj et al., 2010; Bennett et al., 2019; Bali et al., 2021). While all of these mechanistic insights have been dramatic, remarkably little has changed with respect to the treatment of pain, at least in terms of available analgesic approaches, leaving a considerable proportion of chronic pain patients poorly treated (Jensen et al., 2001; Dahlhamer et al., 2018; Pitcher et al., 2019). Consequently, this perceived “failure” of animal studies for analgesic drug development has even raised the question if animal models are sufficient to predict analgesic efficacy in humans (Langley et al., 2008).

The translational gap in pain research may be partially explained by technical limitations to develop animal models that feature realistic approximations of human pathologies (Vierck et al., 2008; Mogil, 2009; Sadler et al., 2022). Unfortunately, these preclinical data may be necessary to find universally accepted objective biomarkers that are crucial to specifically define pathological subsets of pain, evaluate target engagement of new drugs and predict their analgesic efficacy (Davis et al., 2020). The predominant clinical problem of pain patients is suprathreshold, spontaneous pain, and many assays of suprathreshold pain in animals have been available and widely used for decades (e.g. the formalin test). The problem is they do not last very long. Since the focus in preclinical pain research has drifted to the ubiquitous study of “chronic” pain, researchers have been forced to revert to using evoked measures (Mogil and Crager, 2004; Sadler et al., 2022), because in the chronic assays often there are no spontaneous behaviors to measure (Langford et al., 2010; Gregory et al., 2013). On the contrary, it has been suggested that measuring pain thresholds could lack clinical relevance because it may not reflect the major clinical problem which is spontaneous or ongoing pain rather than evoked pain responses (Mogil et al., 2010; Bennett, 2012; Mogil, 2012). This may be particularly important for the development of novel analgesics for neuropathic pain, where a predominant focus on reflexive measures may be oversimplified and thereby potentially explains an apparent mismatch of primary pain assessment endpoints in preclinical vs. clinical trials (Fisher et al., 2020; Schmelz, 2020).

A broad spectrum of translational challenges such as difficulties in experimental designs [e.g. Tappe-Theodor et al. (2019)], species differences [e.g. Mogil (2019)] and potential modulation of evoked sensory reflexes by descending cognitive control [e.g. Wallwork et al. (2017); Dhondt et al. (2019)] have already been extensively discussed. However, despite all these remarkable arguments, another confounding factor which may come with the assessment of pain thresholds has been largely dismissed so far, and this is the temporal integration of axonal excitability or short-term plasticity and its resulting effects on pain. Considering that the general function of the withdrawal reflex is to avoid the pain evoking stimulus, corresponding neuronal activity is concomitantly terminated by the withdrawal action. Under these circumstances, the painful stimulus is potentially even terminated at a level when nociceptive firing rates may just reach clinical relevance. In turn, any mechanism which could be immediately affecting direct or indirect activity-dependent changes in neuronal excitability or synaptic transmission within the assessed nociceptive system will be neglected during this reflexive test. This effect could be particularly relevant for processes of short term plasticity which may be limiting trains of nociceptor activity (Tigerholm et al., 2014). As an example, these mechanisms may play a substantial role for the contrasting accommodation of pain upon prolonged electrical stimulation between healthy subjects and neuropathic pain patients (Jonas et al., 2018), or the sensitization of nociceptors after nerve growth factor (NGF) treatment (Schnakenberg et al., 2021). Interestingly, although both studies clearly demonstrate that the electrical stimulus is painful straight away from the beginning, differences in pain ratings between sensitized and non-sensitized areas have been shown to become only evident after prolonged (20 s) stimulation (Jonas et al., 2018; Schnakenberg et al., 2021). Therefore, albeit generally supporting the need of supra-threshold pain assessment, these studies also indicate that the level of pain perceived in response to a constant stimulation of C-nociceptors changes over time (1 min), and this effect consequently necessitates a prolonged assessment of ongoing pain to study mechanisms of activity dependent short-term plasticity. In turn, similar effects may also explain why quantitative sensory testing (QST) thresholds have been shown not to relate to clinical pain levels (Amiri et al., 2021; Forstenpointner et al., 2021) and therefore appear to provide no additional value for treatment or diagnosis of pain. It is important that Forstenpointner et al. published their results confirming the null hypothesis (Forstenpointner et al., 2021). However, more of resembling results are likely to be drawered (Finnerup et al., 2015), and publication of these data may not only reduce the translational gap in pain research, but also be helpful to overcome a perceived replicability crisis (Baker, 2016; Ventura, 2022).

Taking into account that activity-dependent effects of short-term plasticity on pain are clinically relevant in pain patients, this mismatch may add to the array of confounding variables which are blurring the predictions of mechanistically driven translational approaches based on reflexive measures. This interference may become even more clear for the prediction of postoperative pain from preoperative pain assessments, where a substantial number of confounding variables, such as species differences or technical limitations in pain assessment, are even mitigated. Therefore, in analogy to predicted effects in translational approaches, the preoperative assessment of individual sensory phenotypes should guide predictions about the postoperative outcome. Correspondingly, low preoperative pain thresholds would be expected to indicate higher levels of postoperative pain and vice versa. However, predictive results in patients with postoperative pain are largely heterogeneous, with the most consistent predictive values for postoperative pain outcomes being supra-threshold pain assessments and dynamic parameters comprising temporal summation effects on pain (Abrishami et al., 2011; Sangesland et al., 2017; Petersen et al., 2021).

Taken together, this perspective supports the idea of using differential behavioral read-outs to assess several clinically relevant aspects of pain, i. e. spontaneous pain, lowered pain thresholds and increased supra-threshold encoding. Thus, we not only confirm that even negative results of reflexive tests may be complementing the holistic clinical picture with valuable insights into contributing nociceptive signaling mechanisms (Bordeleau et al., 2021; Vollert et al., 2021), but also accentuate that the assessment of ongoing nociceptive activity and supra-threshold pain responses are emerging as substantial tools to study translational aspects and mechanisms of pain (Mogil and Crager, 2004; Mogil, 2012; Schmelz, 2020, 2022).