Chronic musculoskeletal pain is the most frequently reported (in studies using chart review and clinical interview) pain sub-type in MSA patients, with a prevalence of up to 60% in pain-reporting patients. Musculoskeletal pain in MSA may be partly aggravated by the joint and skeletal deformities present in 68% of patients [34]. Neuropathic pain (both central and peripheral) is also prevalent (up to 36%; as declared using The Leeds Assessment of Neuropathic Symptoms and Signs pain scale, LANSS [35], Neuropathic Pain Symptom Inventory, NPSI [36], and clinical interview) [5, 10, 28, 37]. 4–13% of MSA patients have reported experiencing multiple pain types and localization in clinical interviews and the German Pain Questionnaire (DSF) [26, 37, 38].

In studies using chart reviews, DSF, and asking patients to locate their pain on a body map, MSA patients’ most common localizations were limbs, neck, and back [5, 26, 28]. Notably, aching neck pain radiating to the occiput and the bilateral shoulders, or ‘coat-hanger’ pain, was reported by up to 53% of MSA patients via a custom questionnaire and has been linked to both dysautonomia and antecollis (forward flexion of neck) [28, 39, 40]. However, rarely, coat-hanger pain may also occur in PD patients with orthostatic hypotension, so, therefore, it may not be specific for MSA [5]. Similarly, low-back pain affects 25.4% of the general population aged $>$60 and is not MSA-specific [41].

The severity of pain appears to increase with the progression of the disease: while three cohorts of earlier-stage MSA patients (mean disease duration 2.35, 2.4, and 2.72 years) declared mild/moderate intensity of pain via the NMSS and VAS, severe pain (as measured by EQ-5D) predominated in two later-stage European and North American MSA cohorts (mean disease durations 5.9 and 6.3 years) [9, 24, 26, 32, 42]. Interestingly, Kass-Iliyya et al. [5] did not find pain intensity measured by the Short-Form McGill Pain Questionnaire (SF-MPQ) to correlate with disease duration or motor complications, but rather with the Hospital Anxiety and Depression Scale (HADS) scores [43].

The exact pathophysiological changes underlying the heightened experience of pain in MSA are yet to be fully clarified. At the cerebral level, MSA patients exert $\alpha$-synuclein pathology and neurodegeneration in structures engaged in sensory-discriminative and affective-emotional facets of pain, including the anterior cingulate cortex (ACC) substantia nigra, putamen, insula, amygdala, dorsal raphe nuclei, locus coeruleus, and periaqueductal grey matter. It is tempting to speculate whether these changes might compromise the central processing of nociception [44, 45, 46, 47, 48, 49].

Between 10–40% of patients with MSA show signs of peripheral neuropathy (mainly sensorimotor axonal neuropathy) upon nerve conduction studies and electromyography, possibly contributing to unpleasant neuropathic sensations which occur in a subset of MSA patients and might even be a presenting complaint in some [50, 51, 52].

Several studies have employed psychophysical techniques to investigate pain within MSA. Collectively, these studies indicate augmented responses to experimental pain, as summarized in Table 1. Lack of pain-processing abnormalities in early-stage patients may reflect a progressive alteration to central pain processing.

Only 43–67% of MSA patients that declared pain (on EQ-5D, MDS-UPDRS, or VAS) received a pain-relieving treatment, indicating possible under-treatment of pain in this population [5, 26, 53, 54].

Optimizing dopaminergic therapy might alleviate pain intensity to a certain degree in some MSA patients, as shown in studies using DFS and SF-MPQ [5, 26]. Despite this, in two experimental studies, pain sensitivity measured as temporal summation threshold and subjective pain tolerance (time until heat stimulus became unbearable) in MSA patients post-L-DOPA treatment was unchanged or even worsened [8, 10].

In a chart review using a 5-point subjective Clinical Global Impression (CGI) score determined retrospectively based on the clinical notes, botulinum toxin produced pain relief in over 71.4% of 16 MSA patients treated for dystonia, sialorrhea or pain (mainly dystonic pain, off label indication) [55]. Of note, botulinum toxin injections are rarely effective for antecollis-related neck pain [40]. Regular focused physiotherapy improved pain intensity as measured by the DFS in 4 MSA patients [26]. Wiblin et al. [56] recommend pregabalin, gabapentin, and amitriptyline for neuropathic pain, but high-quality evidence is missing.

In PSP, the pain might be both — a presenting complaint and a feature dominating advanced, palliative stages of the disease, being one of the most common triggers for admissions to a palliative care unit (according to a retrospective chart review) [4, 61, 62, 63].

The most common type of pain is musculoskeletal (as reported by Stamelou et al. [11], who used a custom pain questionnaire (100%) and Yust-Katz et al. [37] (57.1%, based on structured interview). Musculoskeletal pain in PSP may be linked to postural changes, recurrent falls, and joint and skeletal deformities [34, 59]. Rigidity-related pain, with possible musculoskeletal involvement, was present in nearly half of all PSP patients admitted to a palliative care unit, according to a retrospective chart review [4]. In terms of localization, Kass-Iliyya et al. [5] noted 100% limb and 25% neck localization on body maps in 4 pain-reporting PSP patients. The lower limbs and neck were most commonly affected [55, 61, 64]. A retrospective chart review noted back pain being an early feature in patients with a PSP-PGF subtype [65].

In PSP patients, in addition to the striatonigral degeneration, the grey matter volume of the medial prefrontal and orbitofrontal cortices is reduced compared to controls [58, 66, 67, 68]. Furthermore, several brainstem structures involved in descending pain modulatory pathways (including periaqueductal grey (PAG), nucleus raphe magnus, and locus coeruleus) undergo neurodegeneration and pathological tau accumulation in PSP [69, 70, 71]. While spinal cord tau pathology is present in PSP patients, it is unlikely to interfere with pain processing given its localization to the anterior horn and intermediate grey matter [72].

The number of experimental studies investigating pain in patients with PSP is limited, and their results mixed. Importantly, the characteristic frontal dysfunction of PSP has been suggested to alter perception and self-estimation of pain [11]. For example, in one study, despite the thresholds of nociceptive flexion reflex, electrical pain, and heat pain comparable to PD patients (who more rarely have frontal dysfunction), significantly fewer PSP patients reported pain on VAS [11] (see Table 1).

A high proportion (up to 88%) of PSP patients receive analgesics or non-pharmacological pain-relieving strategies (e.g., physiotherapy). However, it rarely leads to successful pain relief [4, 5, 53]. Interestingly, dopaminergic medication alleviated pain as measured on the SF-MPQ in only 25% of patients in a cross-sectional study [5]. In 29 pain-triggered palliative care unit admissions where anti-Parkinson drugs, antidepressants, opioids, NSAIDs, antispasmodics, benzodiazepines, and botulinum toxin injections were used to treat pain, successful pain relief was recorded in only 52% of the patients’ medical records [4]. In another retrospective chart review, botulinum toxin injections improved predominantly dystonic pain in 66% of studied PSP patients (n = 6) [55]. A single case report found severe dystonic neck pain markedly improved on the VAS scale (from 10 to 2/10) with intravenous lidocaine; paracetamol, dipyrone, NSAIDs, tramadol, and oxycodone were ineffective [61].

According to several chart reviews and case reports, CBS patients frequently declare limb-localized pain that might be associated with dystonia or rigidity and, more rarely, with cortical sensory loss (impaired stereognosis, graphesthesia, and 2-point discrimination) [76, 77, 78, 79, 80, 81, 82, 83]. Musculoskeletal pain has also been reported as the predominant subtype in 8 CBS patients (identified through the clinical interview) [37, 55]. Interestingly, neuropathic pain was an initial symptom in the clinical history of 29% of 14 pathologically confirmed CBD cases with CBS presentations [84]. 5 of 8 CBS patients in a cross-sectional study reported suffering “pain or paraesthesia” in a clinical interview, while a proportion of CBS patients in a chart review characterized their pain as ‘burning’, potentially reflecting a neuropathic manifestation [13, 78]. Another interesting case is of a CBS patient presenting initially with complex regional pain syndrome - severe paroxysmal pain of the right upper limb, occurring spontaneously or induced by the extension of fingers [85].

According to three chart reviews, dystonic and neuropathic pain are typically localized in the more affected hand and arm [77, 78, 84]. Yet, some patients might have bilateral limb pain, or, in rare cases, pain can also localize to the knees, neck, chest, and orofacial areas [86, 87, 88].

Existing chart reviews and case reports suggest that rigidity, tremor, hand myoclonus and alien limb phenomenon can all co-occur with pain in the affected arm and hand, and patients might suffer concomitant anxiety [77, 81, 89, 90, 91].

The intensity of pain in CBS is largely described by patients as severe and adversely affects activities of daily living, as revealed in chart reviews and case reports [78, 85, 86].

Pain is a presenting symptom in 11–29% of CBS cases in chart reviews [78, 84, 86]. Of note, according to a chart review, in patients with prominent early sensory symptoms and pain, the pain remained severe at follow-up (mean 5.2 years after initial symptoms). Still, new onset of pain was not noted in any of the participants [78].

While the heightened experience of pain in CBS might arise from motor symptoms in the affected limb, it is likely aggravated by central mechanisms [78, 85, 86]. For example, abnormalities in evoked potentials and glucose hypometabolism (measured by PET) have been observed in the S1 and thalamus in pain-reporting CBS patients [13, 90]. However, both studies severely lack statistical power to generalize these findings, and even these limited samples demonstrated heterogeneity.

To date, no experimental electrophysiological studies have investigated pain processing in CBS patients.

Although controlled trials are lacking, prospective studies and retrospective chart reviews (using clinical interview, VAS scores, or CGI) support the analgesic efficacy of botulinum toxin injections for dystonic and musculoskeletal limb pain in CBS, without notable side effects [55, 77, 92, 93]. An exception was a case report of a CBS patient initially presenting with a complex regional pain syndrome, where botulinum toxin injections significantly worsened pain. At the same time, gabapentin produced moderate pain relief [85].

Medical record reviews note the failure of non-steroidal anti-inflammatory drugs, benzodiazepines, muscle relaxants, and high-dose opioids in producing pain relief in CBS patients [77, 86, 90]. Carbamazepine successfully alleviated neuropathic pain in one case report [78]. Physical rehabilitation has both alleviated and worsened pain in recent case reports [94, 95].

DLB patients might experience a range of painful syndromes. For example, musculoskeletal pain (reported in 2 of 8 DLB patients through clinical interviews) might be further aggravated by motor symptoms and recurrent falls that sometimes result in fractures [37, 104]. Reports of neuropathic pain include 2 cases with suspected central neuropathic pain (affecting the lower abdomen, lower back, vulva, and lower limbs) associated with sensations of coldness and numbness, respectively, and a description of continuous pain in both lower limbs of likely peripheral neuropathic origin associated with restless leg syndrome (RLS) and co-occurring with tactile hallucinations [105, 106].

Interestingly, hallucinations, a core feature of DLB, can be closely linked to the experience of pain, modifying and individualizing the painful sensations, such as in a case report of a DLB patient reporting a fishhook stabbing their finger and causing terrible pain [107].

A poorly localized pain (affecting thorax, abdomen, genitals, joints, and/or head) with gastrointestinal disturbances was reported by 17.9% of 162 DLB patients in a prospective, cross-sectional cohort study (using clinical interviews and performing a chart review). It may be present as a prodromal symptom [108]. This particular subtype of pain seems to be more prevalent in DLB than in other neurodegenerative diseases (AD, PSP, MSA, FTD), possibly being a red flag for DLB. However, there are some similarities with the phenotype of an unexplained lower limb pain (persistent pain affecting proximal anterior thigh region with extension occasionally to the pelvic area, ranging from unilateral to bilateral, and, in some cases, associated with whole-body pain) as described by Wallace and Chaudhuri in PD patients [108, 109].

In terms of intensity, DLB patients have described their pain as ‘severe’ or ‘intolerable’ in case reports, while a case-control study recorded “severe” pain in 41% of DLB patients on the EQ-5D tool, adversely affecting the quality of life and activities such as sleeping [100, 106, 107, 110].

Presently, few conclusions can be drawn regarding the pathophysiology of pain in DLB, given the lack of pain-related experimental studies employing psychophysics or neuroimaging methods.

In DLB, widespread neuropathological changes at the cerebral level may affect numerous structures involved in diverse aspects of pain processing pathways, as summarized in Table 1. In addition, Lewy body pathology is present in both the spinal cord dorsal horn and unmyelinated fibers of the cutaneous peripheral nerve of DLB patients, possibly interfering with pain processing on spinal and peripheral levels [111, 112, 113, 114].

In several case reports, calcium channel modulating anti-convulsants gabapentin and pregabalin successfully relieved neuropathic pain in DLB patients and were well tolerated, but high-quality data are lacking [105, 106]. Of note, caution and specialist monitoring is required, as higher doses of pregabalin might exacerbate parkinsonism [115]. In another case report, pramipexole ameliorated central neuropathic pain in a DLB patient [106].

The analgesic efficacy of L-DOPA in DLB is not known, and there is also currently no evidence on treatments for musculoskeletal pain in DLB.

In terms of localization, pain in FTD mostly affects the head, neck, abdomen, and, less commonly, chest and legs [120, 121, 125]. Headaches are common across all FTD subtypes and are reported by 18–26% of patients in studies using chart review and the Autonomic Symptoms Questionnaire [120, 121]. Musculoskeletal pain (14%) and gastrointestinal/genitourinary pain (12%) were also reported [120]. Neck and shoulder pain is more common in svPPA and potentially linked to autonomic dysfunction [121]. Altered temperature responses frequently co-occur with pain [16, 122]. According to the chart reviews, pain in FTD might be coupled with psychotic (such as hallucinations) and affective symptoms that may alter the way patients experience pain [122, 125] (Fig. 1).

12–20% of bvFTD patients additionally develop motor neuron disease and parkinsonism; subsequently, these patients may suffer pain characteristic to these diseases [116, 127, 128]. In rare cases, FTD occurs in the context of inclusion body myopathy associated with Paget disease of bone and frontotemporal dementia (IBMPFD), a rare familial disorder associated with aching bone and joint pain affecting hips, lower back, and pelvis [129, 130].

In a large MRI study (31 FTD patients and 50 healthy controls) investigating the neuroanatomical correlates of altered pain and temperature responses as characterized by a semi-structured caregiver questionnaire, symptoms suggesting abnormalities of pain and/or temperature processing (reported by 53% of FTD patients and 71% of those with bvFTD) were associated with reduced grey matter volumes within mid and posterior insula in all FTD patients, as well as bilateral posterior thalamus in those carrying C9orf72 mutations [16] (Table 1).

As seen using voxel-based morphometry, atrophy has been described in the right cingulo-insulo-amygdalar network previously implicated in deficiencies of interoception (ability to perceive internal bodily sensations) [131]. Interoception is especially impaired in svPPA, which may hold some explanatory value for the higher prevalence of pain symptoms in these patients than in other FTD variants [122, 131].

Interestingly, in a longitudinal study involving clinical and neuropathological characterization, no robust relationships were found between somatic complaints or abnormal pain responses (noted in medical records of 40.2% of the 97 FTD patients) and brain protein pathology, regional pathology, or asymmetric hemispherical atrophy [120].

To date, psychophysical evidence remains scarce, with only one study having employed the robust quantitative sensory testing battery in FTD patients [12]. Both self-reported pain thresholds and tolerance were increased relatively to controls with no divergent responses across FTD variants (Table 1). However, despite being the gold standard for assessing pain thresholds, quantitative sensory testing relies on self-report, which may also be influenced by the loss of pain awareness (more prevalent in bvFTD than svPPA and nfvPPA) or diminished ability to communicate pain in FTD patients [16, 122, 132].

There have been no controlled trials on analgesia specifically for FTD [133]. Pain may be undertreated in FTD outpatients, with a Dutch cross-sectional study reporting analgesic use in just half of the 14 patients who declared pain in clinical interviews [3]. Fluvoxamine, an SSRI used to treat behavioral symptoms in FTD, conferred marked self-reported analgesia for lumbar and abdominal pain in a case report of two FTD patients [134].