Towards a Whole‐Body Assessment of Pain Using the 18 kDa Translocator Protein (TSPO) as a Broad‐Spectrum Imaging Marker of Inflammation

ABSTRACT

Background and Objective

Chronic pain is increasingly understood as a systemic condition involving coordinated immune responses across central and peripheral compartments. In this narrative review, we present evidence supporting positron emission tomography (PET) imaging of the translocator protein (TSPO), long considered a marker of neuroinflammation, as a promising tool to visualize inflammation throughout the body.

Databases and Data Treatment

This narrative review synthesizes published evidence from TSPO PET studies in chronic pain. We first provide a concise overview of central nervous system (CNS) applications across various chronic pain conditions, implicating neuroinflammation as a potential pathophysiological mechanism underlying human chronic pain. We then discuss evidence from studies extending this imaging approach beyond the CNS to peripheral tissues, in which TSPO is highly expressed by immune cells. We also review associations between these imaging signals, clinical measures and circulating cytokine levels.

Results

Compared with controls, patients with chronic pain show elevations in TSPO PET signal not only within the CNS, including the brain and spinal cord, but also in peripheral structures such as the neuroforamina and joints. Correlations with clinical measures and systemic cytokine levels underscore the potential clinical significance of these peripheral signals. This expanded view positions TSPO PET as a broad‐spectrum marker of inflammation, enabling a unified imaging modality that bridges traditionally siloed approaches to pain research, while also facilitating the discovery of previously underappreciated pain‐relevant tissues and immune compartments. Of particular relevance, the application of TSPO PET beyond its traditional CNS focus has revealed a previously unrecognized role for CNS‐adjacent immune niches, such as skull bone marrow, in human chronic pain.

Conclusions

With further evaluation, TSPO PET may provide a means of visualizing inflammatory processes across multiple pain‐relevant compartments, paving the way for a whole‐person approach to the study of human chronic pain.

Significance

TSPO PET is not merely a neuroimaging tool but could serve as a systemic marker of inflammation in chronic pain. By enabling a whole‐person framework for studying pain and inflammation, it has the potential to inform precision medicine approaches and guide therapeutic decisions.

1. Introduction

For decades, our understanding of chronic pain has been shaped by a neuron‐centric view, emphasizing synaptic plasticity, altered nociceptive transmission and cortical reorganization. Although these mechanisms remain foundational, the scientific community has increasingly recognized that chronic pain is not merely a disorder of neurons, but one of neuroimmune dysregulation. This realization has prompted a reevaluation of the cellular players involved, emphasizing the role of glial cells (e.g., microglia and astrocytes) and peripheral immune cells, whose interactions with neurons are now thought to contribute to the persistence and amplification of pain (Vergne‐Salle and Bertin ²⁰²¹; Watkins and Maier ²⁰⁰²).

The term ‘neuroinflammation’ has become ubiquitous in this discourse, though its definition remains vague and inconsistently applied (Paolicelli et al. 2022; Woodburn et al. ²⁰²¹). It encompasses a spectrum of responses, from microglial and astrocyte reactivity to peripheral immune cell infiltration into the central nervous system, and the associated release of cytokines, chemokines and other inflammatory mediators. These processes have been implicated in the transition from acute to chronic pain (Beggs and Salter 2016; DeLeo and Yezierski ²⁰⁰¹; Grace et al. ²⁰¹⁴), although mostly shown in preclinical models. Indeed, translating these findings to humans has proven challenging. Some small postmortem studies have provided preliminary evidence of increased glial reactivity in the spinal cord of individuals with complex regional pain syndrome (Woodburn et al. 2021) and HIV‐related neuropathic pain (Shi et al. 2012), whereas cerebrospinal fluid analyses have detected elevated levels of fractalkine and interleukin‐8 in fibromyalgia patients (Bäckryd et al. 2017).

Over the course of the past decade, in vivo imaging has emerged as a promising tool to evaluate the role of pain‐related neuroinflammation. Among the available modalities, positron emission tomography (PET) imaging targeting the 18 kDa translocator protein (TSPO) has recently demonstrated promise in this context. TSPO is a transmembrane protein expressed in various cell populations on the outer mitochondrial membrane. Although constitutively expressed by many cell types (including endothelial cells and some neurons), TSPO is present at low levels in the healthy central nervous system (CNS) but—at least in animal models—becomes markedly upregulated in microglia, astrocytes and infiltrating macrophages under inflammatory or injury‐related conditions (Cheung et al. 2023; Cosenza‐Nashat et al. ²⁰⁰⁹). This context‐dependent upregulation, rather than its basal expression under physiological conditions, has traditionally provided the rationale for considering TSPO a marker of neuroinflammation. Importantly, recent evidence suggests that TSPO behaves quite differently in humans than in preclinical models and does not appear to be upregulated at the level of individual glial or immune cells when these adopt a reactive phenotype. Instead of being a marker of a specific activation phenotype or cellular state, TSPO is now understood to serve more as a marker of glial cell density (Nutma et al. 2021; Owen et al. ²⁰¹⁷; Wijesinghe et al. ²⁰²⁵). This nuance has led to some confusion in the literature, with some interpreting these findings as evidence that TSPO is not, after all, a reliable marker of neuroinflammation. However, it is important to keep in mind that an increase in glial/immune cell density is one of the hallmarks for (neuro‐)inflammation. This distinction is particularly relevant for PET imaging, where each voxel contains tens of thousands of cells and the measured signal necessarily reflects an aggregate tissue‐level property rather than the state of any single cell. Thus, although TSPO may not report a discrete ‘activated’ glial/immune phenotype at the cellular level, its signal can still validly reflect an important component of neuroinflammatory pathology.

Leveraging these expression properties, and the fact that TSPO can be imaged using PET (Table 1), this protein has been used as a surrogate marker of neuroinflammation in numerous conditions, including Alzheimer's disease (Heneka et al. 2015), major depression (Setiawan et al. 2015) and stroke (Schaechter et al. 2021). More recently, our group, and increasingly also others, have utilized TSPO PET imaging to investigate chronic pain, providing supporting evidence for a role of neuroinflammation in pain pathophysiology and highlighting it as a potential therapeutic target (see below).

TABLE 1.

Some of the most commonly used TSPO PET ligands in chronic pain imaging studies.

Ligand	Generation/characteristics	Binding considerations	Representative studies
[11C]PK11195	First‐generation, 11C‐labelled	rs6971‐insensitive; low specific‐to‐nonspecific binding	Used in patients with amputations, or lesions of peripheral or central nervous system‐ all conditions that are likely to be accompanied by chronic pain (Banati 2002; Banati et al. 2001), as well as in fibromyalgia and complex regional pain syndrome (Seo et al. 2021).
[11C]PBR28	Second‐generation, 11C‐labelled	rs6971‐sensitive; affected binding affinity (HAB/MAB/LAB)	Extensively used in brain imaging studies of various conditions, including chronic low back pain (Loggia et al. 2015; Torrado‐Carvajal et al. 2021), fibromyalgia (Albrecht, Forsberg, et al. 2019), migraine (Albrecht, Mainero, et al. 2019) and knee osteoarthritis (Alshelh et al. 2025). Also used to image spinal cord and neuroforamina in individuals with sciatica (Albrecht, Ahmed, et al. 2018), trigeminal nerve in migraine (Tohyama et al. 2025), knee joints in, osteoarthritis (Sandstrom et al. 2024) and skull bone marrow/meninges in migraine (Hadjikhani et al. 2020).
[11C]DPA‐713	Second‐generation, 11C‐labelled	rs6971‐sensitive; higher specific binding compared to PK11195	Used in cervical radiculopathy studies showing elevated TSPO in cervical neuroforamina and spinal segments (Lutke Schipholt et al. 2025).
[18F]DPA‐714	Second‐generation, 18F‐labelled	rs6971‐sensitive; improved signal‐to‐noise relative to PK11195	Used in fibromyalgia studies showing elevated TSPO in widespread cortical regions (Mueller et al. 2023).
[18F]FEPPA	Second‐generation, 18F‐labelled	rs6971‐sensitive; strong correlation with CD68+ macrophage expression (ex vivo and in vivo)	Used in knee osteoarthritis imaging; demonstrates strong relationship between in vivo uptake and macrophage burden; supports biological specificity of joint TSPO signal (Koudys et al. 2024).
[18F]GE‐180	Second‐generation, 18F‐labelled	Limited sensitivity; possible low extraction fraction/brain penetrance issues	Included in rheumatoid arthritis studies where it showed weak correlation with Dynamic Contrast Enhanced (DCE)‐MRI and limited disease discrimination, illustrating tracer‐dependent variability (de Groot et al. 2019).

However, recent evidence suggests that the relevance of TSPO as a tool to image pain‐related inflammation extends beyond the brain. In fact, as illustrated in Figure 1, studies from our group have described TSPO PET signal elevation in individuals with chronic pain across multiple sites, including the spinal cord, neuroforamina, joints, and even skull bone marrow (Albrecht, Ahmed, et al. ²⁰¹⁸; Hadjikhani et al. ²⁰²⁰; Lutke Schipholt et al. ²⁰²⁵; Mohammadian et al. ²⁰²⁶; Sandstrom et al. ²⁰²⁴; Tohyama et al. ²⁰²⁵). These findings suggest that TSPO PET may be a ‘broad‐spectrum’ marker that transcends traditional neurocentric boundaries and offers a whole‐body view of inflammation. In this review, we synthesize the evidence supporting this expanded role of TSPO PET in chronic pain, and explore its implications for diagnosis, patient stratification and therapeutic monitoring.

FIGURE 1.

Graphical overview of TSPO PET signal elevations reported across chronic pain conditions. The figure synthesizes representative findings from the literature, illustrating TSPO signal increases in key central and peripheral sites, including cortical and subcortical brain regions, spinal cord segments, neuroforamina and dorsal root ganglia, skull bone marrow and inflamed joints. This conceptual summary highlights the multisite nature of TSPO PET alterations and supports the whole‐body inflammatory framework discussed in the manuscript. Each panel includes an image adapted from a previously published study (brain: Loggia et al. ²⁰¹⁵; skull bone marrow: Hadjikhani et al. ²⁰²⁰; cranial nerve roots: Tohyama et al. ²⁰²⁵; spinal cord and neuroforamina: Albrecht, Ahmed, et al. ²⁰¹⁸; joints: Sandstrom et al. ²⁰²⁴).

2. TSPO PET Imaging to Assess Nervous System Inflammation in Human Pain

Early studies by Banati and colleagues used the first‐generation radioligand [¹¹C]‐(R)‐PK11195 to report elevated brain TSPO PET signal in patients with amputations, PNS and/or CNS lesions, all conditions likely accompanied by chronic pain (Banati 2002; Banati et al. ²⁰⁰¹). However, the extent to which the observed signal elevations specifically reflected pain‐related neuroinflammatory processes (as opposed to inflammatory responses secondary to the underlying lesions themselves) remains uncertain. The application of TSPO PET imaging to the systematic study of human chronic pain, including primary pain disorders, is relatively recent (Loggia 2024). In 2015, our group demonstrated elevated TSPO signal in the brains of individuals with chronic low back pain (cLBP), using the second‐generation radioligand [¹¹C]PBR28 in combination with integrated positron emission tomography/magnetic resonance imaging (PET/MRI) (Loggia et al. 2015). In that initial study, we observed significantly higher TSPO binding in the thalamus, primary somatosensory cortex (S1) and primary motor cortex (M1) of patients compared to matched healthy controls. Notably, the thalamic signal was elevated in nearly every patient, suggesting the possibility of a robust and reproducible neuroinflammatory signature.

Indeed, in a subsequent study involving an independent cohort of cLBP patients, we replicated the thalamic elevation in TSPO PET signal, further reinforcing the robustness of the original observation (Torrado‐Carvajal et al. ²⁰²¹). Interestingly, the cortical effects observed in the first study were not reproduced in the second. This discrepancy prompted us to consider whether clinical heterogeneity‐specifically, the presence or absence of radicular symptoms‐ might account for the differences. Indeed, when we stratified patients based on pain distribution, we found that those with radicular pain exhibited significantly higher TSPO signal in the S1 representation of the leg and back, as identified through functional MRI‐based somatotopy (Alshelh et al. 2022). Moreover, this signal was proportional to the degree of ‘nociplastic’ pain (Kosek et al. 2021), as measured by the American College of Rheumatology Fibromyalgia Survey Score (Wolfe et al. 2010). These findings suggest that TSPO PET signal in the brain might be spatially and symptomatically specific. The thalamus appears to be a consistent site of elevation across cLBP subtypes, while S1 signal may reflect the somatotopic distribution of pain and its underlying mechanisms.

This spatial distribution of the TSPO PET signal may also be at least partially condition‐specific, thus raising the intriguing possibility that each pain disorder might be accompanied by its own ‘neuroimmune signature’. In fibromyalgia (FM), for example, the TSPO PET signal was distributed very differently than in cLBP, encompassing a much larger portion of the cortex, including in the dorsolateral and dorsomedial prefrontal cortices (dlPFC, dmPFC), S1/M1, precuneus and posterior cingulate cortex (PCC) (Albrecht, Forsberg, et al. ²⁰¹⁹). These effects were independently observed at two study sites (Boston and Stockholm) and were also largely replicated by other groups using different TSPO PET radioligands, such as [¹¹C]PK11195 and [¹⁸F]DPA‐714 (Mueller et al. 2023; Seo et al. ²⁰²¹). Interestingly in FM, the signal was elevated over a large portion of the S1 (possibly mirroring the widespread pain which is a hallmark of FM) whereas in migraine (Albrecht, Mainero, et al. ²⁰¹⁹) and back pain (Loggia et al. 2015; Alshelh et al. ²⁰²²), they were predominantly localized to a ventrolateral subregion compatible with the somatosensory representation of the head and face and of the lumbar spine, respectively. These patterns support the hypothesis that, within S1, TSPO PET signal elevations may reflect the somatotopic distribution of the patients' pain. Because glial/immune cells are highly sensitive to changes in their microenvironment, including those induced by neural activity (Bollinger et al. 2024; Xanthos and Sandkühler ²⁰¹⁴), these patterns may reflect—at least in some cases—aberrant afferent input arising from clinically painful body regions. This interpretation should be viewed cautiously, however, given that only a small proportion of nociceptive spinothalamic projections terminate directly in S1.

Importantly, TSPO signal in various studies has been found to correlate with clinical symptom severity across several studies (e.g., with fatigue in the anterior and posterior middle cingulate cortices in FM; Albrecht, Forsberg, et al. ²⁰¹⁹); with the degree of nociplastic pain in S1 and with depressive symptoms in the pregenual anterior and anterior‐midcingulate cortices in cLBP (Albrecht et al. 2021; Alshelh et al. ²⁰²²); with attack frequency in S1, in individuals with migraine (Albrecht, Mainero, et al. ²⁰¹⁹); and with knee pain severity in the pituitary gland in individuals with knee osteoarthritis (Alshelh et al. 2025). Together, these findings suggest that elevations in TSPO PET signal are not simply epiphenomenal but instead reflect clinically meaningful processes.

While the brain has been the primary focus of TSPO PET imaging in chronic pain, we and others have shown that TSPO signal elevation can be observed in spinal cord segments as well as neuroforamina (which contain nerve roots and dorsal root ganglia). For instance, in patients with lumbar radiculopathy, our group has reported evidence of elevated TSPO PET binding in the T11‐T12 spinal cord segments (compared to T7‐T9 reference segments and the same segments in healthy controls), as well as in the corresponding neuroforamina ipsilateral to the symptomatic leg (L4/L5 and/or L5/S1, compared to the contralateral side and to healthy controls), aligning with dermatomal pain distribution and structural MRI findings (Albrecht, Ahmed, et al. ²⁰¹⁸). Similar findings were reported by another group in patients with painful cervical radiculopathy, demonstrating elevated TSPO PET signal in the affected (in this case cervical) neuroforamina and corresponding spinal cord segment relative to unaffected reference regions (Lutke Schipholt et al. 2025).

Interestingly, these neuroforaminal signals were not only anatomically specific, but also clinically predictive. In our small study with lumbar radiculopathy participants who exhibited elevated TSPO signal in the affected neuroforamen, were significantly more likely to respond to epidural steroid injections, suggesting that TSPO PET may serve as a biomarker for treatment stratification (Albrecht, Ahmed, et al. ²⁰¹⁸).

Taken together, this body of work suggests that TSPO PET signal is elevated in the brain—and, in a smaller set of studies, in the spinal cord and/or dorsal root ganglia/nerve roots—across multiple pain disorders (Loggia 2024), suggesting that neuroinflammation may be part of the pathophysiology of, and perhaps a therapeutic target for, human chronic pain. Furthermore, the ability of TSPO PET to predict responses to an anti‐inflammatory treatment further underscores the link between TSPO signal and inflammation.

3. TSPO PET As a Whole‐Body Inflammation Marker

Although TSPO PET has traditionally been used to study ‘central’ inflammation—as shown above in pain (Loggia 2024) and other disorders with a CNS component (Albrecht, Granziera, et al. ²⁰¹⁸)—emerging evidence suggests that this tool might serve as a whole‐body marker of inflammation, capable of capturing immune processes across central and peripheral compartments. In fact, as previously mentioned, TSPO is not only highly expressed in immunocompetent cells within the CNS (such as microglia and astrocytes), but also in peripheral immune cells, such as monocytes, macrophages and neutrophils (He et al. 2025; Kolabas et al. ²⁰²³; Lacor et al. ¹⁹⁹⁶).

Consistent with this, using the same radioligand used to image TSPO in the CNS (i.e., [¹¹C]PBR28), we were recently able to demonstrate significantly elevated PET signal in the knees of patients with KOA, compared to healthy controls (Figure 2A) (Sandstrom et al. 2024). The signal was significantly elevated in both knees of patients with bilateral KOA symptoms and in the symptomatic knee of patients with unilateral symptoms, but not in their asymptomatic knee (Figure 2B). Furthermore, it was more elevated in the most painful knee compared to the least painful one, and this difference was proportional to the difference in pain ratings across knees (Figure 2C). Importantly, TSPO PET signal did not correlate with Kellgren‐Lawrence grades, highlighting a dissociation between radiographic severity and inflammatory activity. Rather, TSPO PET signal intensity correlated with circulating plasma levels of pro‐inflammatory cytokines (TNF‐α (Figure 2D), IL‐1β, IL‐1RA and IL‐8) and, within the intra‐articular and periarticular compartments, was reduced in patients receiving nonsteroidal anti‐inflammatory drugs (NSAIDs) relative to those not taking these medications, providing additional evidence that TSPO PET signal reflects underlying inflammatory activity.

FIGURE 2.

TSPO PET signal in knee osteoarthritis (KOA) patients. (A) Representative [¹¹C]PBR28 PET/MR scans showing lower signal in a healthy control (top) and elevated signal in a KOA patient (bottom). (B) Boxplot comparing mean PET signal (SUV) across healthy controls (HC), patients with unilateral KOA symptoms, and patients with bilateral KOA symptoms, showing significantly higher signal in symptomatic knees. (C) Scatter plot illustrating a positive correlation between differences in left–right knee pain and differences in PET signal. (D) Scatter plot showing a strong positive correlation between TNF‐α plasma levels and PET signal in KOA patients. All these findings support the use of TSPO PET as a marker of joint inflammation and its association with clinical pain and systemic inflammatory markers. Adapted from Sandstrom et al. (2024).

Later work has imaged knee joints in KOA using [¹⁸F]FEPPA, another second‐generation TSPO ligand (Koudys et al. 2024). In this study, [¹⁸F]FEPPA uptake in the suprapatellar synovium correlated strongly with TSPO expression in CD68⁺ macrophages, both ex vivo and in vivo, supporting its utility as a non‐invasive biomarker of macrophage reactivity and further underscoring the potential of in vivo TSPO imaging to quantify biologically active inflammation in OA beyond structural imaging.

TSPO PET imaging has also demonstrated its utility as a marker of joint inflammation also in other conditions, such as rheumatoid arthritis (RA). In individuals with this condition, TSPO PET signal was significantly elevated in RA pannus tissue compared to healthy joints (Narayan et al. 2018). The highest TSPO expression was observed in activated fibroblast‐like synoviocytes (FLS), undifferentiated (M0) and reparative (M2) macrophages, and to a lesser extent, monocytes and CD4+ T cells. It is, however, important to emphasize that the TSPO PET signal does not necessarily indicate a predominance of cells polarized toward a reparative phenotype. Rather, at least in humans, it is thought to primarily reflect immune cell (or, in the CNS, glial) cell density, rather than a specific ‘activation state’ or phenotype (Nutma et al. 2021; Owen et al. ²⁰¹⁷; Wijesinghe et al. ²⁰²⁵).

While the above studies highlight the promise of radioligands such as [¹¹C]‐PBR28 and [¹⁸F]FEPPA, it is important to note that not all TSPO PET ligands perform equally well across inflammatory joint diseases. For instance, a study using [¹⁸F]GE‐180 (another second‐generation TSPO ligand) has found limited sensitivity and weak correlation with DCE‐MRI parameters in RA, despite its theoretical specificity for TSPO (de Groot et al. ²⁰¹⁹). Together, these studies reinforce the relevance of TSPO PET imaging in OA and RA, while emphasizing the need for disease‐ and tracer‐specific validation.

The use of TSPO PET as a marker of inflammation beyond the CNS not only enables the detection of inflammatory processes in expected regions (e.g., the joints of patients with KOA or RA), but also opens the door to unexpected discoveries. For example, in individuals suffering from migraine, we have recently revealed elevated TSPO PET signal in a region of the occipital skull compatible with the calvarial bone marrow and/or adjacent meninges, compared to healthy volunteers (Figure 3) (Hadjikhani et al. 2020). The skull bone marrow has recently emerged as an important interface for immune‐brain communication (Hadjikhani et al. 2020; Herisson et al. ²⁰¹⁸; Kolabas et al. ²⁰²³; Mills 3rd et al. ²⁰²²). This emerging field has been greatly advanced by the discovery of microscopic channels that perforate the inner skull cortex (Herisson et al. 2018). These structures provide a bidirectional pathway, allowing cerebrospinal fluid‐borne CNS antigens and cues to reach the bone marrow, where they can stimulate local myeloid cell activation and egress toward the meninges and brain, thus contributing to neuroinflammatory responses (Cugurra et al. 2021). This outward migration was found to be accompanied by an increase in total immune cell density within the skull marrow itself, an effect that may underlie the elevated TSPO PET signal we reported in migraine (and that others have subsequently reported in other conditions, including neurodegenerative disorders, multiple sclerosis, depression and autism spectrum disorder; Eiff et al. ²⁰²⁵; Kolabas et al. ²⁰²³; Tseng et al. ²⁰²⁵). While the role and clinical significance of the skull bone marrow in human pain remain to be elucidated, emerging evidence from our laboratory suggests that it may also contribute to musculoskeletal disorders as well (Mohammadian et al. 2026). Collectively, these findings highlight the skull bone marrow as a dynamic interface between the brain and the peripheral immune system, with immune cell accumulation in this region emerging as a promising biomarker of neuroinflammation.

FIGURE 3.

TSPO PET signal reveals elevated uptake in the skull bone marrow and adjacent meninges of migraine patients, compared to healthy controls. (A) Individual PET/MRI image showing a representative case of elevated TSPO signal (white arrows); (B) Group‐level differences in occipital bone marrow TSPO PET signal. Adapted from Hadjikhani et al. (2020).

Taken together, these findings support a conceptual shift in the use of TSPO PET imaging. Rather than being viewed solely as a tool for assessing neuroinflammation within the CNS, TSPO PET may serve as a broad‐spectrum marker of inflammation capable of visualizing immune activity throughout the body. This whole‐person approach offers several potential clinical applications, from diagnosis and stratification to treatment monitoring and outcome prediction.

4. Towards a Whole‐Person Approach to the Study of Pain

While recognizing the CNS' critical role in the development and maintenance of persistent pain has greatly advanced our understanding of chronic pain, this emphasis has, perhaps excessively, overshadowed the importance of peripheral tissues, particularly in musculoskeletal pain research (Langevin 2024). More broadly, pain research has been hampered by a siloed approach, where peripheral mechanisms (e.g., joints and muscles) and central mechanisms (brain and spinal cord) have typically been studied in isolation. We believe that advancing our understanding of chronic pain requires an integrated, whole‐person approach that bridges the brain and body. This approach would support the development of precision medicine strategies, for instance by uncovering distinct inflammatory phenotypes—or ‘(neuro)inflammatory’ signatures—such as ‘predominantly peripheral’, ‘predominantly central’ or ‘mixed’ inflammatory profiles ‐that could be instrumental in guiding the development of personalized treatments and patient stratification strategies. For example, we hypothesize that individuals with a predominantly peripheral inflammatory profile may be more responsive to anti‐inflammatory treatments targeting peripheral mechanisms, such as steroid injections or even knee replacement surgery. In contrast, those with a predominantly central profile may be less responsive to peripheral treatments but more likely to benefit from a centrally acting immunomodulatory treatment, meaning therapies aimed at modulating neuroimmune processes within the CNS (e.g., glial modulators). By providing a unified imaging modality capable of capturing immune activity throughout the entire organism, TSPO PET—perhaps along with the widespread adoption of total‐body PET scanners (Badawi et al. 2019)—is uniquely positioned to enable and accelerate these efforts.

5. Caveats

The execution of TSPO PET studies and interpretation of their results come with several caveats and challenges (Loggia 2024). As noted above, the biological meaning of increased TSPO PET signal is not always straightforward. TSPO is a mitochondrial protein expressed by several cell types, including microglia and astrocytes in the CNS, peripheral immune cells, endothelial cells and other non‐immune populations. Accordingly, elevated PET signal likely reflects (neuro‐)inflammatory processes, but it may in principle also be shaped by changes in cellular metabolism, vascular or tissue remodelling or other mechanisms. Non‐immune cell populations, including endothelial and other cells, can express TSPO constitutively and may contribute to baseline or disease‐related PET signal independently of canonical immune activation. More broadly, because TSPO is a mitochondrial protein implicated in bioenergetic, cholesterol‐handling and oxidative‐stress pathways, TSPO signal may in some contexts reflect cellular stress or metabolic remodelling in addition to inflammatory processes. Despite these interpretive caveats, however, in practice the balance of available evidence supports the view that elevated TSPO signal is most often associated with inflammatory processes, particularly in contexts involving glial or immune‐cell accumulation. Supportive evidence for the clinical relevance of peripheral TSPO PET imaging includes findings that increased knee TSPO PET signal correlates with peripheral inflammatory measures, such as circulating cytokines and chemokines (Sandstrom et al. 2024), as well as ex vivo indices of immune‐cell infiltration/density in synovial or related tissues (Koudys et al. 2024). In addition, as mentioned, elevated TSPO signal has shown prognostic value, including the ability to predict analgesic or anti‐inflammatory treatment responses following epidural steroid injections (Albrecht, Ahmed, et al. ²⁰¹⁸), supporting the interpretation that the PET signal reflects biologically meaningful inflammatory activity. These observations further support the utility of TSPO PET as a biomarker for patient stratification and treatment monitoring in chronic pain and inflammatory musculoskeletal disorders. Importantly, while elevations in TSPO PET signal are most plausibly interpreted in terms of increased glial/immune‐cell density, this does not preclude an indirect link to neural alterations. Indeed, because glial and immune cells are highly sensitive to changes in their microenvironment, including those driven by aberrant activity in nearby neurons, both in the CNS (Bollinger et al. 2024; Xanthos and Sandkühler ²⁰¹⁴) and in the body's periphery (Michoud et al. 2021), TSPO signal increases may in some cases reflect processes consistent with neurogenic (neuro‐)inflammation.

Interpretation of TSPO PET signal requires careful attention to methodological limitations inherent to TSPO quantification (for an in‐dept analysis, refer to Turkheimer et al. ²⁰¹⁵). Both compartment‐based models and reference‐tissue approaches have been used, as well as semiquantitative ratio metrics. While the volume of distribution (V_T; i.e., the ratio of radioligand concentration in tissue to that in plasma, measured in units of mL · cm⁻³, at equilibrium) is often considered the gold standard for quantifying the binding of many radioligands, this metric shows very high inter‐individual variability for TSPO radioligands, due to various factors. Some of them are relatively straightforward to correct, such as the effect of the Ala147Thr TSPO polymorphism (rs6971). This polymorphism strongly affects binding affinity for many TSPO radioligands and typically requires genotyping, exclusion of low‐affinity binders, and statistical adjustment for differences between high‐ and mixed‐affinity binders (HAB and MAB; Nutma et al. ²⁰²¹; Owen et al. ²⁰¹⁷; Wijesinghe et al. ²⁰²⁵). Other factors, arguably more difficult to control, include the highly variable binding to some plasma proteins, whose concentration may be itself altered in inflammatory states (Lockhart et al. 2003), thereby affecting the fraction of radioligand available to reach TSPO at the target sites. In principle this issue could be remedied by using a plasma binding assay to estimate the plasma free fraction (f_P), but in practice reliable measurement of this parameter for TSPO tracers is challenging because it is typically very low (often < 3%–4%). Several groups have reported that semiquantitative ratio metrics, derived by normalizing the PET signal to that of a pseudo‐reference region or to the whole brain (depending on the condition under investigation), may improve sensitivity to biologically meaningful effects and better distinguish patients from controls (Albrecht, Normandin, et al. ²⁰¹⁸; Lyoo et al. ²⁰¹⁵), However, identifying a suitable pseudo‐reference region requires independent validation in each population studied. These factors do not negate the utility of TSPO PET, but they emphasize the need for methodological rigour when interpreting regional differences, group contrasts or correlations with clinical measures.

6. Could TSPO PET Become Part of Routine Clinical Practice?

At present, the use of TSPO PET may lead to important indirect contributions to clinical care, for example by helping to demonstrate a role for neuroinflammation in specific pain conditions, thereby stimulating further research and supporting the development of mechanism‐based treatments, or by helping to identify previously unrecognized peripheral pain generators or inflammatory sources, potentially enabling more personalized therapeutic approaches. However, the routine clinical use of TSPO PET itself is difficult to envision in the foreseeable future, irrespective of how informative and potentially transformative this technique may be from a research perspective. Its high cost, limited accessibility, methodological complexity, and still uncertain clinical specificity make widespread clinical implementation difficult to justify at present. On the other hand, these limitations do not preclude a future clinical role, particularly if ongoing technical and methodological advances can establish clear added value in specific indications. The use of TSPO PET in clinical routine would require several developments beyond those already demonstrated in research settings. Standardization of acquisition and analytical protocols would be essential to ensure reproducibility across centers, scanners and radioligands. Validation in larger, clinically heterogeneous cohorts is also needed to determine the diagnostic or prognostic value of TSPO PET beyond group‐level differences, including whether regional TSPO elevations can meaningfully guide treatment decisions or predict therapeutic response. Importantly, the field will need consensus on the most suitable ligands and modelling approaches for specific clinical indications, together with evidence that TSPO PET provides actionable information not available from conventional imaging or clinical biomarkers. Finally, feasibility, cost and workflow integration must be addressed before TSPO PET can be deployed in clinical practice, particularly if its role expands to whole‐body or multi‐compartment inflammatory phenotyping. Advances in next‐generation PET scanners with enhanced sensitivity, including long‐axial‐field‐of‐view systems, along with the use of artificial intelligence‐driven analysis methods are likely to improve both the detection of binding sites and the quantitative accuracy of PET imaging, while enabling the acquisition of high‐quality data with shorter scan durations and/or lower injected doses, improving clinical applicability (Loggia et al. 2025). Together, these steps define a path toward future translational use while underscoring that TSPO PET currently remains primarily a research tool.

7. Conclusion

The journey of TSPO PET imaging in chronic pain has taken us from the thalamus to the neuroforamina, from cortical networks to the bone marrow. What began as a tool to visualize glial ‘activation’ in the brain has evolved into a modality capable of capturing immune activity across the entire body. This expansion is not merely technical; it reflects a deeper shift in how we conceptualize pain, inflammation and the neuroimmune interface.

We have shown that TSPO PET signal is elevated in multiple regions relevant to pain pathophysiology: sensory and affective brain circuits, spinal cord segments, peripheral nerve roots, inflamed joints and even skull marrow. These signals do not appear to be merely epiphenomenal, but they appear to map onto the patient's symptoms, correlate with clinical measures, and in some cases predict treatment response. Together, these findings reinforce the view that chronic pain is neither confined to the periphery—as traditionally assumed—nor restricted to the brain, as more recent work appears to emphasize. Instead, it represents a systemic condition characterized by coordinated immune responses across central and peripheral compartments.

Of course, challenges remain. TSPO is not a perfect marker; its signal is influenced by cell type, phenotype, genotype and quantification methods (Loggia 2024). But even with these limitations, it has already begun to inform our understanding of chronic pain and may help us move closer to precision pain medicine.

Author Contributions

A.T.‐C. and M.L.L. had a primary role in preparing the manuscript. All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Funding

We would like to thank all the funding agencies who have so generously supported this work through multiple research grant awards (R21‐NS087472‐01A1, R21‐NS082548‐01A1, R01‐NS094306‐01A1, R01‐NS095937‐01A1, R01‐AR079110‐01A1, R01‐DA053316‐01, R01‐DA047088‐01, W81XWH‐14‐1‐0543, IASP Early Career Award).