Pain refers to an “unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage” (Raja et al., 2020). Integrating several dimensions (physiological, sensory, cognitive, and emotional aspects), pain is based on subjective experience and therefore, on conscious processing of the stimulus. Like any subjective experience, communication with the patient is the most appropriate way to assess it. In severely brain injured subjects, such as patients with disorders of consciousness (DoC) and locked-in syndrome (LIS), verbal communication is impaired but does not exclude the possibility that they experience pain. Even more, the absence of behavioral signs of consciousness does not preclude the patient to show (at least a minimum of) cortical activity preservation, suggesting partial preservation of consciousness and pain processing. Therefore, it is important not to neglect the assessment of pain and nociception in these patients with limited or no ability to communicate, regardless of the diagnosis. In the last years, clinicians tended to identify behavioral patterns related to conscious perception of pain, with important ethical and clinical implications in terms of diagnosis, prognosis, and treatment. Nonetheless, the absence of clinical signs of pain does not preclude a conscious (i.e., cortically mediated) pain experience nor a physiological impact of the nociceptive stimuli. Indeed, nociception refers to “neuronal process allowing the encoding and processing of a noxious stimulus” (Loeser and Treede, 2008) and while it does not require conscious perception of the stimulus, it leads to changes in the autonomic control of target organs (e.g., changes in heart rate, sweating, bronchial resistance to air flow, and pupil diameter) and behavioral responses (e.g., flexion withdrawal).

Disorders of consciousness could be due to various traumatic (TBI) or non-traumatic (NTBI) brain injuries (e.g., strokes or anoxia). In the United States, 2.5 million people suffer from a TBI each year (288,000 hospitalizations and 56,800 deaths) and some of them will become unresponsive wakefulness syndrome (UWS) or minimally conscious state (MCS) patients (Capizzi et al., 2020). In the United States, the prevalence of patients with DoC (adults and children) is estimated between 4,000 and 25,000 for patients in UWS and between 112,000 to 280,000 for patients in MCS (The Multi-Society Task Force on PVS, 1994; Pisa et al., 2014). In Europe, the cases of UWS patients are estimated between 4,362 and 58,160 in a population of 727,000,000 (Ashwal, 2004). The current literature does not allow to give the prevalence of LIS patients on a global or national level. Due to the absence of or impaired communication in these patients, pain assessment and management is a major clinical and ethical issue. DoC include different clinical entities based on cognitive level and motor abilities (see Figure 1). For instance, a patient who shows signs of arousal characterized by eye opening periods but no signs of awareness (i.e., only reflexive movement and absence of cortical processes), will be categorized as in UWS (The Multi-Society Task Force on PVS, 1994; Laureys et al., 2010). Patients in a MCS show reproducible, responses without functional communication and have partial cortical processes. They are classified into two main groups based on language preservation: patients in MCS minus (MCS−) showing non-reflexive behaviors (Bruno et al., 2011b; Giacino et al., 2018), and MCS plus (MCS+) who have a preservation of higher level non-reflexive behavior and language abilities (Thibaut et al., 2020). The progress of neuroimaging techniques has also allowed the emergence of new terminologies to classify patients with “atypical” brain activity patterns such as: minimally conscious state star (MCS*, i.e., patients behaviorally diagnosed with UWS but preserving residual brain activity congruent with MCS diagnosis at rest or during a passive or active paradigm; Thibaut et al., 2021), covert cortical processing (CCP or higher-order cortex motor dissociation – HMD, i.e., patient behaviorally diagnosed in a coma, UWS, or MCS− but retaining brain activity upon passive task; Edlow et al., 2017), cognitive motor dissociation (CMD, i.e., patient behaviorally diagnosed in a coma, UWS, or MCS− but retaining brain activity upon active tasks; Schiff, 2015). Finally, when a patient regains functional communication or functional use of objects, he or she is considered to be emerging from MCS (Di Perri et al., 2016). LIS is not considered as a DoC but could be misdiagnosed with coma and UWS (Cistaro et al., 2018). This condition results from a lesion in the corticospinal and corticobulbar pathways of the brainstem due to vascular pathology, traumatic brain injury, masses in the ventral pons, infection, or demyelination (Das et al., 2021). LIS patients suffer from limbs, head, and face paralysis (i.e., quadriparesis) as well as verbalization/vocalization, breathing, and coordination impairments. LIS patients can communicate using eyelid blinks, vertical eye movements, or head movements (i.e., yes/no communication code or letter spelling communication; Lugo et al., 2015). EEG-based brain–computer interfaces also allow LIS patients to communicate through brain signals (Annen et al., 2020). So far, there has been limited scientific research on pain processing in LIS. However, according to a European survey of health professionals, 90% of them considered that patients in LIS are able to feel pain and need to be treated (Demertzi et al., 2014). For patients with DoC, according to a survey, 96% of health professionals believed that MCS patients can feel pain, compared to 56% believing that UWS patients can do so (Demertzi et al., 2009). Nevertheless, as explained above, some behaviorally unresponsive patients can still have a cortical activity preservation suggesting covert consciousness (and potentially a preservation of pain processing). It is therefore important to set up pain assessment tools and treatments that are independent of the clinical diagnosis to avoid mismanagement.

This manuscript aims at reviewing the current knowledge about the assessment and management of pain and nociception in patients with DoC and LIS. We will first give an overview of the physiology of pain and nociception. We will then look more specifically at the possible sources and impact of pain in these populations. Finally, we will describe the tools and treatments currently in place for the assessment and management of pain and nociception for this patient population. This narrative review is based on systematic reviews, meta-analyses, original articles, and case studies.

In order to understand the difference between pain and nociception and to comprehend to what extent severely brain-damaged patients process nociceptive inputs and pain, we must first look at the neurophysiology of these two phenomena. When nociceptive stimulation occurs, following tissue damage for example, a signal will be generated at the endings of the nociceptive Aδ (i.e., thinly myelinated fibers responsible for faster signal transmission, mediate nociceptive inputs but also non-nociceptive heat and cold stimuli) and C-fibers (i.e., non-myelinated, polymodal nociceptors that are sensitive to chemical, mechanical, and thermal stimuli, including nociceptive hot – >48°C – and noxious cold – <11°C). These fibers synapse at the level of the dorsal horn with the second nociceptive neuron that continues its path into the spinothalamic tract to the thalamus (for the majority of the fibers). After the thalamus, the signal arrives at several cortical areas [i.e., the primary and secondary somatosensory cortex and the insula and the anterior cingulate cortex (ACC)]. All these cortical and subcortical structures and their connections form a network which is activated following a painful stimulus. In this review, we will use the term “pain-related neuromatrix” to refer to these brain regions activated following a noxious stimulus. However, it is important to note that this network is not only related to pain processing but has been identified more as a salience detection network (Melzack and Wall, 1965; Ingvar, 1999; Brown et al., 2018). This supports the multidimensional aspect of pain sensation, already highlighted by Melzack and Wall (1965), and subsequently confirmed by neuroimaging studies: pain is not the result of the activation of a single specific region but of a network (Mouraux et al., 2011; Mouraux and Iannetti, 2018). The pain-related neuromatrix can be divided in two parts comprising: (1) the lateral system, involved in the sensory dimension of nociceptive stimulus processing (i.e., localization, duration, and intensity) which includes the lateral thalamic nucleus, the primary somatosensory cortex (S1), the secondary somatosensory cortex (S2), and the insula and the posterior parietal cortex; and (2) the medial system, related to the affective dimension of pain and comprising the medial thalamic nucleus, the prefrontal cortex, the ACC, the posterior cingulate cortex (PCC), and the posterior medial cortex (Bushnell et al., 1999; Hofbauer et al., 2001). These regions (i.e., prefrontal cortex, thalamus, and ACC) are also part of the external and internal networks of consciousness which underlines the importance of these brain areas in the conscious perception of pain. The insula also has an important role in the affective processing of pain because it mediates the signal between the posterior insula (lateral system) and the rostral part of the ACC (medial system) (Coghill et al., 1999; Peyron et al., 2002). According to neuroimaging studies in healthy subjects during acute pain stimulation, the cortical and subcortical regions most involved in pain signal processing are S2, the insula and the ACC [for a review refer to Peyron et al. (2000)]. The use of hypnosis [i.e., a state of consciousness involving attentional focus and reduced peripheral attention, characterized by an increased ability to respond to suggestion (Elkins et al., 2015)] with analgesic suggestions leads to a decrease of brain activity in the ACC, and makes it possible to modulate the affective dimension of pain [for a review refer to Thompson (2019)]. The use of hypnosis has also been shown to be effective in relieving chronic pain. Indeed, in a 1997 study, a positive correlation between the perception of painful sensation and cerebral activity of the ACC was demonstrated (Rainville, 1997). Conversely, when the ACC and the insula are activated just before a nociceptive stimulation, an increase in pain perception is observed (Boly et al., 2007). Altogether, the literature shows that the ACC has an important role in modulating pain perception, notably by interacting with regions of the limbic system like the amygdala, the thalamus, and the hippocampus (Moriarty, 2011; Calabrò et al., 2017). These subcortical and limbic structures participate in the balance of activity between the fronto-temporo-parietal cortex (involved in consciousness; Demertzi et al., 2013; Di Perri et al., 2013, 2016) and the autonomic nervous system.

The ascending pathways described above activate descending pathways responsible for modulating the transmission of peripheral information. Indeed, the descending pathways (cholinergic and serotonergic) to the dorsal horn of the spinal cord are sensitive to peripheral stimuli and not only to nociceptive stimuli. There are also tonic facilitation and inhibition phenomena that originate in the brainstem and respond to peripheral or non-peripheral stimuli (Dunckley, 2005). The activation of these descending pathways starts in the cortex (i.e., in the insula and ACC) and extends through the hypothalamus and the amygdala to finally be transmitted to the brainstem in the PAG, the nucleus of the tractus solitarius and the rostral ventral medulla (Brown et al., 2018). This will result in an inhibition of neurons from the superficial dorsal horn relaying information carried by C-fiber to the deep dorsal horn (Figure 2). The suppression of the C-fibers signal will then facilitate the transmission of sensory-discriminative information conducted by the A-fibers (Heinricher et al., 2009). The hypothalamus, the amygdala and the PAG are also responsible for behavioral changes related to acute pain stimulation (Veinante et al., 2013). An fMRI study of awake subjects undergoing acute thermal pain stimulation has shown a decrease of brain activity in the hypothalamus and amygdala as well as an increase brain activity in the lateral PAG (Robertson et al., 2022). The lateral PAG is involved in the selection of appropriate defensive behaviors in response to the nociceptive stimulus [i.e., increase in motor, autonomic, and endocrine activity, as well as alertness, and inhibitory control of this pain (Bandler et al., 2000)] and is regulated by a number of regions including the hypothalamus and the amygdala. A lesion in these different ascending and descending pathways therefore may lead to a dysfunction in the processing of nociceptive stimuli and pain control (i.e., which can result in the phenomenon of central sensitization).

The conscious perception of pain is supported by the activation of the regions evoked above and the functional connectivity between these different regions and the thalamus (Baars et al., 2003; Dehaene and Changeux, 2011). However, even if some of these regions are essential for pain sensation and sensibility, others will play a subtler role and their lesion may not always lead to any noticeable change in terms of pain perception. As explained above, the pain-related neuromatrix, although well established in the scientific literature, is still subject to debate, not least because the regions of this pain-related neuromatrix are more broadly involved in multimodal processing and not specific to pain processes (Mouraux et al., 2011; Mouraux and Iannetti, 2018). A recent opinion paper discussed the idea that pain perception may originate from the brainstem and not only from the cortex. The authors based this assumption in part on the fact that cortical stimulation of specific regions of the pain-related neuromatrix does not induce pain, unlike other sensory modalities (e.g., primary auditory or visual cortex stimulation evokes respectively sound and light). Few studies have nevertheless shown that the electrical stimulation of regions such as the parietal operculum, the posterior insula, and the ventral caudal nucleus of the thalamus could induce pain sensation (Lenz et al., 1993; Mazzola et al., 2006; Bergeron et al., 2021). However, a lesion of the insula does not make the sensation of pain disappear (Libet, 1973; Mazzola et al., 2006; Afif et al., 2008; Isnard et al., 2011). On the contrary, some patients who have suffered a cortical lesion (i.e., central post-stroke patients) experience an increased sensation of pain (Boivie et al., 1989; Andersen et al., 1995). As pain is necessary for survival, these authors suggest that its conscious perception must have been in place before the expansion of the cerebral cortex and therefore be located in the brainstem. Many brainstem nuclei are involved in nociceptive signal processing [for a review refer to Napadow et al. (2019)]. fMRI in humans show that, upon acute cutaneous or visceral stimulation, the PAG, nucleus cuneiformis, ventral tegmental area, substantia nigra, parabrachial complex, and dorsolateral pons regions of the brainstem become activated (Dunckley, 2005; Fairhurst et al., 2007; Sprenger et al., 2011). The spinal trigeminal nucleus located at the level of the medulla and caudal pons is activated during painful stimulation in the orofacial region (Nash et al., 2009). The brainstem also seems to be involved in the phenomenon of conditioned pain modulation. For instance, inhibition of orofacial pain via painful stimulation of another area (such as the leg) results in a reduction of the fMRI signal in the dorsal reticular nucleus, dorsolateral pons, and spinal trigeminal nucleus (Youssef et al., 2016). PAG and rostral ventral medulla also appear to be necessary for the temporal summation of pain in connection with the phenomenon of nociceptive wind-up [i.e., facilitation of neural discharges caused by repetitive stimulation of primary afferent C-fibers, involved in central sensitization (Mendell, 2022) in humans and animals with chronic pain (Van Oosterwijck et al., 2013; O’Brien et al., 2018)]. The study of the functionality of these different brainstem nuclei is challenging, especially in neuroimaging studies due to the location of these elongated and small cross-sectional nuclei, their proximity to cardiorespiratory noise sources. The role of these nuclei in pain processing has yet to be studied in LIS and DoC, however, these severely brain-injured patients may present cerebral deformations that make the analysis of robust neuroimaging data difficult.

Due to their physical condition and the clinical environment in which they find themselves, patients with DoC and LIS may experience various types of nociceptive insults. For instance, acute nociceptive events can occur after injuries (e.g., fracture, wounds, and soft tissue/solid organ injuries) or during daily care (e.g., catheterization, surgery, or physiotherapy). If pain is present after those injuries, it will act as a protective and adaptive signal for the integrity of the body (Craig, 2003) whereas chronic pain loses the role of warning signal (Varrassi et al., 2010). Chronic pain is persistent and/or recurrent pain lasting for more than 3 months and can result in functional and emotional changes such as depression or anxiety (Grichnick and Ferrante, 1991; Merskey and Bogduk, 1994; Turk et al., 2011). It can be due to muscle contractions, pressure sores, peripheral nerve injury, pain network disruption leading to allodynia, central sensitization, neuropathic pain, or spastic paresis [for a review see Zasler et al. (2022)]. Central sensitization results from a dysfunction of the descending central control system and corresponds to an “increased responsiveness of nociceptive neurons in the central nervous system to their normal or subthreshold afferent input” (Loeser and Treede, 2008). Neuropathic pain is defined by the International Association for the Study of Pain as a type of “pain caused by a lesion or disease of the somatosensory nervous system” as opposed to nociceptive pain occurring following nociceptor stimulation. Neuropathic pain can be of central or peripherical origin depending on the lesion localization (Raja et al., 2020). At present, neuropathic pain may be identified by diagnostic testing [e.g., using questionnaires such as the DN4 (Bouhassira et al., 2005)], sensory testing coupled with self-report or neuroimaging (to locate the lesion). However, the use of questionnaires requires functional communication to express subjective experience and therefore cannot be used in patients with DoC. The difficulty of assessing this type of pain in DoC patients has so far not been addressed by any study. Regarding spastic paresis, a recent study showed that the majority (83%) of patients with DoC experience pain during physiotherapy sessions (Bonin et al., 2022b). This result can be explained in part by the high prevalence of spastic paresis in this population, ranging from 59 to 96% (Martens et al., 2017; Thibaut et al., 2021; Zhang et al., 2021; Bonin et al., 2022a,b). In addition to limiting patients’ motor responses, leading to misdiagnosis (Monti et al., 2010; Cruse et al., 2011), spastic paresis also appears to be related to the presence of nociception phenomena. Indeed, when looking at the scores of behavioral scales which respectively assess nociception and spastic paresis, it seems that both variables are positively correlated especially in wrist and finger muscles (Bonin et al., 2022a). This can greatly affect the patient’s ability to respond to commands and perform other motor-related tasks on which most of the Coma Recovery Scale-Revised [CRS-R, gold standard to assess the level of consciousness in DoC patients (Giacino et al., 2004)] items are based. Due to the fact that LIS patients are bedridden for long periods of time and therefore have limited mobility, they will develop pain mainly in the lower and upper limbs instead of the head, the back, and the abdomen (Bonin et al., 2022a). They will also be prone to develop spastic paresis, which can lead to persistent discomfort in the long run (Cairns and Stein, 2002; Pistoia et al., 2015). A 2022 survey investigated the presence and management of pain in this specific LIS population. The results highlighted that half of the LIS patients surveyed have pain but do not communicate about it and 92% of these patients suffer from chronic pain (Bonin et al., 2022c). Nociception may also have an influence on the autonomic nervous system, provoking an imbalance between sympathetic and parasympathetic activity (Lee et al., 2020). This can have hemodynamic consequences (e.g., increase in blood pressure, tachycardia, and increased heart rate variability), or influence other target organs of the autonomic nervous system (pupils and their diameter, sweat glands, and skin conductance) [for a review see Lee et al. (2020)]. Although no studies exist on this topic in patients with DoC or LIS, it can be assumed that repetitive and/or long-term autonomic nervous system imbalance due to acute or chronic nociception or pain could have consequences for the patient’s wellbeing, and could lead to systemic complications (Leo et al., 2016). For instance, it has been shown that, in moderate to severe traumatic brain injured (TBI) patients, autonomic nervous system dysfunction is correlated with an increase in morbidity (Purkayastha et al., 2019).

The perception of pain can vary according to different factors. Numerous studies carried out over the last few decades have revealed gender differences in terms of prevalence, perception and treatment of pain [for a review see Pieretti et al. (2016)]. Although women seem to report signs of pain more often than men, experimental studies on healthy subjects show mix results depending on the type of stimulation (i.e., mechanical, electrical, thermal, ischemic, and chemical) and the type of investigated parameter [e.g., duration and intensity of the pain sensation or pain tolerance or sensitivity (Labus et al., 2008; Racine et al., 2012)]. However, to our knowledge, no study has investigated gender differences in terms of pain perception in DoC and LIS patients. In LIS patients, the position (lying/sitting) can increase or decrease the pain sensation, depending on each individual (Bonin et al., 2022c). Pain has a direct influence on patients’ quality of life such as sleep quality, cognitive abilities, and emotion (Bonin et al., 2022c). Previous studies showed that the majority of patients with chronic pain have sleep disorders and that poor sleep quality can increase pain perception (O’Brien et al., 2011; Rousseau et al., 2015; Frohnhofen, 2018; Bonin et al., 2022c). In addition, the impact of pain on sleep quality can alter the level of arousal, as well as motivation in patients with DoC or in LIS. In this way, their ability to express signs of consciousness may be impeded, hence compromising the clinical diagnosis (Lanzillo et al., 2016; Estraneo et al., 2022). Consequently, the implementation of treatment to alleviate pain could have a positive impact on sleep quality and allow an improvement in the patients’ level of arousal/vigilance during clinical examinations. Deleterious effects of pain on cognitive abilities (i.e., increase of tiredness and mood swings, and decrease of memory and concentration) and emotional regulation has been observed in patients with LIS (Bonin et al., 2022c). Other surveys found that some patients in LIS claim experiencing anxiety, depression or suicidal thoughts (Bergés et al., 2007; Rousseau et al., 2015). Furthermore, a past study has found an anti-correlated relationship between perceived pain and life satisfaction (Skevington, 1998; León-Carrión et al., 2002). Several variables can be related to the decrease of life satisfaction in patients with LIS such as the loss of mobility during recreational activities or language impairment/speech production (as communication seems to play a key role in the preservation of the quality of life in those patients) (Bruno et al., 2011a; Demertzi et al., 2014). These results underline the importance of identifying the sources of potential pain by using appropriate tools, to propose patient-tailored management.

Regarding the assessment of pain and nociception in patients with DoC or LIS, there are currently no clear guidelines and no clinical consensus. When performing neuroimaging analyses, it is relevant to mention that differences in terms of structures and physiological properties may exist between a severe brain-injured patient and a healthy subject. Therefore, it is essential to perform a multimodal assessment, not only based on neuroimaging but also on pain-related behaviors and physiological changes [(i.e., increase of heart rate and respiratory rhythm, and skin conductance (Cowen et al., 2015; Devalle et al., 2018; Riganello et al., 2019)] to improve pain assessment and indirectly the diagnosis of these patients. From a behavioral perspective, opinions still differ among researchers and clinicians regarding some behaviors that could be reflective of cortical processing (Poulsen et al., 2019). This is particularly the case for facial expression such as grimacing and crying. Indeed, even if grimacing is considered as an indicator of pain, the Multi-Society Task Force on Permanent Vegetative State does not consider it as a necessary sign of conscious perception, as it can occur reflexively through subcortical pathways in the thalamus and limbic system (The Multi-Society Task Force on PVS, 1994). Patients showing no sign of consciousness except for grimaces to nociceptive stimuli can therefore be diagnosed as being in UWS. Moreover, some patients in LIS suffer from cortical lesions. This impacts their cognitive functions by impairing, for example, the recognition of negative facial expressions, or leading to the development of pathological laughter and crying which may distort the assessment of pain (Leonard et al., 2019). Many pain scales take into account the assessment of facial expressions in a more or less developed way (Feldt, 2000; Gélinas et al., 2006; Chanques et al., 2009; Chatelle et al., 2012). However, the facial expression assessment is clinically scored based on gross observation of facial movements in response to a noxious stimulation. A better characterization of facial expressions could be an interesting avenue of research to improve the behavioral assessment of these patients. For instance, the use of the facial action coding system could be developed in these patients. This system allows the coding of different types of emotions (including pain) based on the anatomical analysis of facial movements. It can distinguish 46 different action units produced by a single muscle or a combination of muscles (Kunz et al., 2007, 2008; Bartlett et al., 2014).

Numerous studies have highlighted the relevance of measuring neurophysiological parameters in the assessment of pain and nociception (Riganello et al., 2019; Cortese et al., 2020). At present, very few studies have investigated the clinical utility of GSR and PDR in the assessment of pain in DoC. This is mainly due to the fact that these measures are not suitable for all types of DoC patients, some of whom may suffer from ptosis often associated with the presence of myosis (i.e., pupil constriction) or other pupillary reactivity disorders, which makes it difficult to measure PDR. In addition, it is important to note that there is a gap between research and practice. The scientific literature on LEPs is well developed, but in practice, this technique is more complicated to implement in a systematic way. The device allowing LEP measurement is an expensive non-portable system, difficult to use in a clinical setting, especially with a sensitive population such as patients with DoC. Other less costly and easier to use techniques assessing the integrity of the spinothalamic pathways are used in other populations and deserve to be investigated in patients with DoC and LIS. For instance, pinprick-evoked potentials (PEPs, mechanical stimulation) are useful to assess the functional integrity of mechano-nociceptive pathways and detect central sensitization (Iannetti et al., 2013; Rosner et al., 2020; van den Broeke et al., 2020) but could be difficult to use in non-collaborative population such as DoC patients. Then, cool-evoked potentials (CEPs, thermal stimulation) allow the evaluation of the integrity of the spinothalamic pathways by stimulating Aδ-fibers and participate in the diagnosis of neuropathic pain without inducing pain (De Keyser et al., 2018; Leone et al., 2019). Finally, contact heat-evoked potentials (CHEPs, thermal stimulation) are also used to specifically assess the nociceptive component of a stimulus. These new generation of thermal cutaneous stimulators (i.e., thermodes) are portable and easier alternatives to LEPs for the recording of robust nociceptive (heat) and non-nociceptive (cold) responses in patients with DoC (De Schoenmacker et al., 2021; Lejeune et al., 2022). The aforementioned techniques could allow better understanding of nociception processing and facilitate neuropathic pain detection in patients with DoC and LIS, which is currently understudied. In the future, the NCS-R could be improved by integrating new physiological parameters like other recently developed scales, such as the BINAM for TBI patients or the PAS. Moreover, the measurement of physiological parameters could facilitate the assessment of the nociception/anti-nociception balance after analgesic administration. Indeed, to monitor the effects of analgesics administered during general anesthesia, anesthesiologists can use different types of tools measuring the activity of the autonomic nervous system (De jonckheere et al., 2015). The above-mentioned ANI, for instance, is based on HRV analysis and allows the measurement of the relative parasympathetic tone. Its score ranges from 0 to 100, a low score meaning that the patient is able to process nociceptive stimulus. The Surgical Pleth Index (SPI) is rather based on the measurement of the orthosympathetic hemodynamic response to noxious stimulation, and uses normalized heartbeat intervals (HBIs) and plethysmography wave amplitude for its calculation (Rogobete et al., 2021). The PDR and the GSR are also used in anesthesia to assess the sympathetic tone but have not yet been studied in detail in patients with DoC and LIS. The functional near-infrared spectroscopy (fNIRS) applied to pain detection could also be an avenue of future research to investigate. It is a non-invasive, low cost, easy-to-use, and portable brain imaging technique that allows to measure cortical hemoglobin concentration changes (Barati et al., 2017; Lopez-Martinez et al., 2019). Studies in healthy subjects have shown that fNIRS can provide an objective and robust assessment of pain by measuring changes in hemoglobin in the sensorimotor and prefrontal cortex (Yücel et al., 2015). Its application for pain detection has also been studied in sensitive and non-communicative patient populations such as infants and critically ill patients (Ranger and Gélinas, 2014; Yuan et al., 2022). The fNIRS is also used in patient with DoC to improve diagnosis but there is not, to our knowledge, any study specifically related to the detection of pain in this population (Rupawala et al., 2018). This measurement technique could also be an avenue to develop in post-coma patients given its low cost, ease of use and portability.

In the future, it would also be essential to develop non-pharmacological therapies in order to limit the use of analgesics and thus avoid the side effects such as fatigue or decreased vigilance. Few studies have looked at the effects of music therapy on the level of consciousness of UWS and MCS patients and have shown that it is a safe and effective method that can improve functional outcomes of patients [for a meta-analyses refer to Li et al. (2020)]. The effect of this technique on pain perception in LIS and DoC patients has not, to our knowledge, been studied yet. However, studies carried out in other patient populations have shown interesting effects (by reducing anxiety for instance) which suggest that this may be an interesting avenue to investigate in future research (Lin et al., 2020; Santiváñez-Acosta et al., 2020; Dallı et al., 2022; Seyffert et al., 2022).

This review focuses mainly on the physical pain that DoC and LIS patients may experience. However, LIS patients may also suffer from emotional pain such as depression or anxiety (Bergés et al., 2007; Rousseau et al., 2015; Bonin et al., 2022c). More studies are still needed to better characterize this type of suffering and its impact on patients’ daily lives in order to propose appropriate pharmacological (e.g., antidepressants and anxiolytics) and complementary (e.g., hypnosis and meditation) treatments.

There are still many unknowns in the assessment, management and treatment of pain in DoC and LIS patients. The NCS-R remains the most appropriate way to assess pain in patients with DoC but could be improved by considering the inclusion of physiological parameters in their behavioral assessment. The measurement of pain and nociception should be done with a multimodal approach, also taking into account (neuro)physiological and neuroimaging data as complementary measures. It is known that some behavioral UWS patients may show preservation of cortical areas involved in nociceptive signal processing. Then, pain assessment and analgesic treatments should be applied in a more systematic way, and most importantly, independently of patient’s clinical diagnosis. In particular, titration of analgesic agents should be implemented to determine the optimal dose of the medications. The NCS-R and the BINAM represent relevant assessment tools to find a balance between reduced pain and preserved level of consciousness following analgesic treatment. For the moment, the guidelines of the American Academy of Neurology recommend the use of the NCS-R to assess pain in patient with DoC but these guidelines still need to be developed further and refined (Giacino et al., 2018). Regarding patients in LIS, even if they do not communicate their pain spontaneously, it is important to actively and regularly make an assessment through the use of simple communication codes. When signs of pain are detected, it is essential to identify the source of the physical and emotional pain to be able to propose appropriate treatments, both pharmacological and non-pharmacological.

EB: conceptualization, research, references formatting, writing, and editing. AT: conceptualization, supervision, review and editing, visualization, and resources. NL, ES, VB, CM, OG, and SL: review and resources. All authors contributed to the article and approved the submitted version.