Introduction
Pain is a highly common driver of reduced quality of life across rare, musculoskeletal diseases [
4,
21,
32]. Prominent and more accessible features of musculoskeletal disease such as skeletal deformities and lesions, or muscle atrophy and weakness garner extensive clinical and research attention, whereas the neurobiological cause(s) of pain have been less well-studied in this clinical domain. While there have been advances in understanding central abnormalities in musculoskeletal diseases [
26,
50], neurological processes underlying pain in a chronic musculoskeletal state remain largely undefined. Demonstrating that musculoskeletal pathology yields or associates with central, pain-related alteration can serve as a neurobiological basis for implementing specific pharmacological or non-pharmacological interventions. A deeper understanding of the central manifestations of pain in rare musculoskeletal diseases, an area that has received minimal attention to date, may help uncover new targetable pain mechanisms present in musculoskeletal illnesses impacting broader populations.
Here, we aimed to determine whether and how clinical pain is centrally represented in fibrodysplasia ossificans progressiva (FOP; OMIM #135100) patients. FOP is an ultra-rare disease with an estimated 1:800,000–1:3,000,000 prevalence worldwide (285 cases in the United States) [
2,
30,
38]. FOP is an inherited disorder that arises from missense mutations of the type I bone morphogenetic protein (BMP) receptor Activin A receptor type 1 (
ACVR1) [
43]. The mutant form of ACVR1 confers activin-A-dependent osteogenic signaling, which results in the heterotopic ossification (HO) of soft, connective tissue structures such as skeletal muscles, tendons, ligaments, and fascia [
7,
13,
14], often leading to progressive immobilization and persistent pain. Prior observations indicate that the sole presence of the ACVR1 mutation is not sufficient for HO formation, but rather, exogenous events involving soft tissue injury, inflammation, or trauma are necessary to trigger the pathological ACVR1 signaling cascade and HO in FOP [
42,
49].
The induction as well as the expansion of HO lesions are often preceded by soft tissue swellings termed “flare-ups” [
13,
37,
46], which are characterized by local inflammation or edema, redness or warmth of skin, joint stiffness, and pain [
8,
20,
36]. In the event of a flare-up, inflammatory responses alongside sustained pressure on hard or soft tissue structures are factors that can evoke high levels of pain. Interestingly, our prior analysis of data collected from a sample of approximately 100 FOP patients demonstrate that while moderate to severe pain was reported during flare-ups, sub-populations of FOP patients in a quiescent period and showing no objective signs of a flare-up also experienced an identical level of pain [
34]. This finding points to the concept that pain in FOP is associated with but not entirely driven by peripheral, flare-up-related mechanisms. Additionally, FOP patients have reported experiencing recurrent severe headaches (migraine, cluster, or tension-type headaches), neuropathic pain and somatosensory abnormalities (e.g., allodynia, hyperalgesia, numbness, or tingling), which may impact or be driven by central pain processes [
22,
37,
45]. Therefore, we sought to determine if modulated central mechanisms relate to clinical pain in FOP patients using resting-state, functional near-infrared spectroscopy (fNIRS). fNIRS is a non-invasive, flexible, optical neuroimaging method that quantifies the cortical hemodynamics changes based on the neuro-vascular mechanisms. fNIRS revealed for the first time a disruption in the spontaneous oscillations localized to the prefrontal cortex in FOP patients experiencing moderate to severe clinical pain.
Discussion
This work highlights the central representation of a critical element of FOP, namely, pain. We report that the pain experienced by FOP patients was integrated with an individual’s emotional health, which points to central involvement in FOP pain. By utilizing fNIRS, we demonstrated that the severity of clinical pain was negatively associated with the power of low-frequency oscillations localized to the prefrontal cortex, a region implicated in emotional appraisal [
27], reward mechanisms [
29], and clinical conditions including major depressive disorder [
18]. The observed loss of the prefrontal cortex, low-frequency activity within the slow-5 sub-band, strongly points to the disintegration of large-scale CNS networks in a high pain, FOP state. The current findings revealed fNIRS's utility for characterizing central pain mechanisms in FOP and suggests that this approach might be generalized for discovering central features of clinical pain in other musculoskeletal diseases.
Peripheral pathological features (nerve compression) or events (flare-ups) can cause patients to experience pain. However, across rare and also more common musculoskeletal diseases, there is a frequent discrepancy between objective measures of disease burden or peripheral trauma and patient-reported pain levels [
21]. Finding such as those centered around prefrontal cortex functionality in FOP pain patients in conjunction with the elucidation of diminished emotional health may help explain this discordance between pain and tissue pathology, and better direct the implementation of pharmacological and non-pharmacological pain treatment interventions. Therefore, while the current set of results are specific to FOP, there are implications that extend beyond this rare disease and into other clinical conditions characterized by pain and musculoskeletal phenotypes.
In FOP, HO of musculoskeletal tissue causes long-term, severe disability. However, alongside the physical limitations that many FOP patients undergo and which can begin during childhood stages of life [
10,
33,
36], the core pathological elements of FOP such as HO induction and growth may lead to other downstream manifestations [
34]. Pain in FOP, is most often associated with recurrent soft tissue swelling or flare-ups, which in many circumstances precedes new HO lesions or expansion of existing ones [
36,
37,
46]. Yet, as previous work [
34] as well as the current study have revealed, clinical pain is frequently reported by FOP patients during what are considered ‘quiescent’ or non-flare-up states.
We found that pain in FOP can be variable, both across subjects and time. Monitoring patients across a 2–3-week period demonstrated intra-subject fluctuations in pain intensity with some individuals reporting increased levels of pain during this interval, while others showed little to no change in pain intensity. The site of pain often corresponded to the location of known HO lesions; however, there were cases where pain was altogether absent, or was experienced only across a subset of HO lesion sites. A pain level ≥ 4 was moderately associated with gender. A trend of higher pain levels or higher chronic pain prevalence in female FOP patients is in accord with the large body of data demonstrating higher perception of pain and more frequent occurrence of chronic pain conditions in women [
31,
47]. Additionally, slightly higher pain levels were measured in older FOP patients relative to younger patients; an observation that was previously observed in other rare musculoskeletal conditions [
21]. From a battery of clinical questionnaires, the perception of pain in individuals reporting a high level of pain consisted of both sensory and affective elements. Cross-correlation analyses further indicated the interaction between severity of pain, pain quality, physical health, and emotional health; however, two key features stemming from this approach were derived. First, there was clear evidence that the physical (i.e., physical stress) and psychological (i.e., depressive mood or anxiety) challenges that FOP patients harbor are tightly integrated. Second, there was a robust association between higher levels of clinical pain and greater depressive symptoms. While higher pain severity often coincides with a greater depressive symptomology in many clinical conditions [
16], the observation of this relationship in a FOP population has additional implications. Pain in FOP can of course be driven by peripheral pathological events, for example, the soft tissue edema or inflammation occurring during a flare-up episode. However, superimposed upon peripheral, pain-inducing occurrences, are abnormalities anchored within the CNS, which can result in a top-down dysregulation of pain perception and relatedly, heightened pain sensitivity or maintain a persistent state of pain [
17,
24].
Implementation of fNIRS enabled an objective assessment of CNS function in FOP patients and equally important, an understanding of whether and how cortical activity relates to subjective accounts of pain in this disease. Patients with moderate to severe pain were characterized by a reduced amplitude of slow-5 (0.01–0.027 Hz), low-frequency oscillations occurring within the prefrontal cortex, where higher clinical pain severity negatively correlated with more suppressed prefrontal cortex fluctuations. Using resting-state functional magnetic resonance imaging (fMRI), highly similar frequency-dependent changes have been reported in multiple chronic pain conditions. In trigeminal neuralgia [
52], postherpetic neuralgia [
12], and chronic back pain [
28,
51] populations, a loss of slow-5 power was not only noted in prefrontal cortex [
52] but also in subcortical regions such as the nucleus accumbens, which share projections with sub-regions of the frontal cortex areas amongst other (meso-)limbic structures. Thus, in general, the suppression of the frontal cortex (potentially along the frontostriatal pathway) low-frequency oscillation in the slow-5 sub-band may represent a maladaptive cortical process that is derived from persistent pain or mechanisms that sustain a pain state. Further investigation is necessary to determine if and how there is a causal link among suppressed frontal cortex oscillations and chronic pain in FOP.
The alterations in low-frequency dynamics have been previously linked with changes in functional connectivity [
1]. Slow-5 oscillations are more closely associated with spontaneous neuronal oscillation intrinsic to cortical regions, including the prefrontal cortex, while higher frequency activity (e.g., slow-4 sub-band: 0.027–0.073 Hz) involves cortical and subcortical regions such as the basal ganglia [
39,
48]. Slow-5 oscillations enable long-range connectivity, particularly that which is necessary for proper communication amongst CNS hubs (i.e., prefrontal cortex) embedded within large-scale networks (i.e., default-mode or mesolimbic network). A disruption of mechanisms that facilitate long-range CNS communication may underpin the decreased prefrontal, and somatosensory cortex functional connectivity observed in FOP patients with higher self-reported clinical pain levels. The somatosensory cortex encodes the sensory dimension of the pain [
15], while the prefrontal cortex with its connections to the pain modulation centers (periaqueductal grey) is known to exhibit endogenous antinociceptive effects. A decrease in the connectivity of the somatosensory cortex and prefrontal cortex may imply an imbalance in the pain encoding process and the ensuing descending control of pain leading to prolonged pain status. Interestingly, relative to short-range connectivity, there is also a significantly stronger coupling between long-range connectivity and CNS metabolic properties (cerebral blood flow, cerebral metabolic rate for oxygen or glucose [
25], possibly suggesting that cerebral metabolic alterations may especially perturbate slow-5 or slow-4 low-frequency oscillations and in parallel, connectivity within large-scale networks.
We project that the frequency-dependent and pain-dependent effects observed in our FOP cohort of patients are a collective and cyclical outcome of living with a chronic illness such as FOP, withstanding recurrent and severely painful flare-up episodes as well as enduring persistent pain over several months or even years. Nonetheless, the impact of aberrant activin-A- or ACVR1/ALK2-dependent activity with the CNS must be considered. In FOP patients [
19,
41] lesions or focal MRI hyper-intensities along white matter pathways have been noted. Disruption of white matter integrity can indeed modulate CNS functional properties.
This investigation represents the first functional CNS imaging study in FOP and a novel application of fNIRS. fNIRS was implemented as an alternate approach to probe CNS function, as MRI-based assessments are not easily feasible for many FOP patients. This study, in conjunction with other recent efforts, demonstrates fNIRS's utility towards elucidating central pain mechanisms [
9,
35]. However, limitations are noted. fNIRS provides an assessment of cortical (dys-)function, where alterations in the brainstem or subcortical structures and their association with clinical pain cannot be elucidated. Compared to fMRI, fNIRS offers flexibility in terms of implementation, allowing bedside measurement to be made. However, a low signal to noise ratio at specific optode locations, possibly due to hair, was a limiting factor. Some FOP patients presented with HO in the neck, shoulder, and upper back region, which anatomically altered their head position and caused difficultly in fNIRS positioning of optodes, particularly over the somatosensory cortices. The use of wireless fNIRS setups or implementation of advanced light source technology may facilitate improved fNIRS data quality throughout the brain. Additionally, in the current cohort of FOP patients, those categorized in the high pain group were primarily women, which is in line with prior FOP studies [
22]. Nonetheless, future investigations may aim to enroll more male FOP patients with moderate to severe pain. Lastly, some patients continued analgesic treatment during the study, but the varied medication usage between patients makes this unlikely.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.