The use of SPECT-CT in the evaluation of lumbar spine pathology - A Narrative Review

Abstract

Introduction

Chronic low back pain is frequently associated with multilevel degenerative abnormalities that may not identify the symptomatic structure. Single-photon emission computed tomography combined with computed tomography (SPECT-CT) combines metabolic and anatomical imaging and has been proposed as an adjunct where conventional imaging is inconclusive.

Research question

To critically assess current evidence concerning the diagnostic and clinical utility of SPECT-CT in lumbar degenerative pathology.

Materials and methods

PubMed was searched for English-language publications dated from 2000 to 2025 concerning SPECT-CT, lumbar pathology and low back pain. Primary clinical studies, technical studies, relevant reviews and guidelines were synthesised narratively, with direct and contextual evidence distinguished.

Results

The evidence base is heterogeneous and predominantly observational. SPECT-CT can anatomically localize sites of increased osseous tracer uptake and may provide additional information in selected diagnostically challenging cases. However, interpretation is limited by non-standardised positivity thresholds, interobserver variability, discordance with MRI, imperfect agreement with diagnostic blocks, absence of a validated reference standard and limited controlled evidence that SPECT-CT-guided management improves outcomes.

Discussion and conclusion

SPECT-CT may have an adjunctive role in carefully selected patients with unresolved diagnostic uncertainty where the result is likely to alter management. Current evidence is insufficient to support routine use or to conclude that SPECT-CT independently identifies the true pain source, improves treatment selection or improves outcomes over conventional diagnostic pathways.

Highlights

• •SPECT-CT offers insight not provided by conventional imaging. However, lack of standardised guidelines on result interpretation hinders its use in surgical decision making. Management of lower back pain involves multiple different imagining methods.

Abbreviations:

SPECT-CT -

single-photon emission computed tomography

CLBP -

chronic lower back pain

MRI

magnetic resonance imaging

SI

sacroiliac

VAS

Visual Analogue Scale

ODI

Oswestry Disability Index

1. Introduction

Chronic low back pain (CBLP) poses a significant strain to healthcare systems worldwide, affecting 619 million individuals globally (GBD, 2021 Low Back Pain Collaborators, 2023). Aside from individual patient burden, CLBP poses a substantial, global socioeconomic burden through healthcare expenditure and lost productivity estimated to cost 40 billion dollars annually in the United States (Chang et al., 2024). 90-95% of CLBP cases are classified as non-specific, with degenerative disc disease cited as a major contributor to pathological phenotypes (Romaniyanto et al., 2022). It is therefore key that clinical guidelines are tailored to most efficient identification of symptomatic areas. Recently, single-photon emission computed tomography combined with computed tomography (SPECT-CT) has emerged as a hybrid imaging technique to evaluate lumbar spine pathologies, particularly in patients with CLBP. SPECT-CT provides both metabolic and anatomical information by coupling a radiotracer bone scan with cross-sectional CT. The technique can pinpoint areas of active bone turnover, such as inflamed facet joints, degenerative disc endplates, or sacroiliac joints, and directly localize them on the accompanying CT images. This ability to correlate “hot spots” of tracer uptake with specific bony structures offers an advantage in identifying the putative pain generators in degenerative spinal conditions, especially when magnetic resonance imaging (MRI) or plain CT findings are equivocal (Hirsch et al., 2024).

Despite its promise, significant limitations surround the role of SPECT-CT in spine disease. As clinicians seek to integrate this modality into practice, they face several challenges: (1) There is a lack of standardized thresholding or quantitative criteria for interpreting SPECT-CT signal intensity in the spine, making results operator-dependent. (2) The biological significance of a “hot” SPECT-CT signal in degenerative spine conditions is unclear, particularly how tracer uptake correlates with underlying pathological processes of pain. (3) SPECT-CT findings have had limited validation against histopathological gold standards or definitive diagnostic benchmarks, leaving uncertainty about their true positive and negative predictive value. (4) Integrating SPECT-CT into treatment algorithms is challenging, as evidence-based guidelines are lacking and the potential clinical benefit is not established. In this structured review, we examine each of these areas in detail and compare SPECT-CT with conventional and novel MRI sequences.

1.1. Principles of SPECT-CT

SPECT-CT bone imaging relies on radiotracers (most commonly ^99mTc-methylene diphosphonate) that accumulate in areas of bone remodelling and osteoblastic activity. Delayed-phase SPECT of the spine (performed a few hours after tracer injection) highlights regions of increased tracer uptake, reflecting increased bone turnover (Fig. 1). As the tracer decays, the releasing gamma rays are detected by a camera rotating 360°. When these SPECT images are fused with CT, the exact anatomical site of uptake can be determined with high precision, however, this should not be equated with confirmation that the identified structure is the source of pain. In the lumbar spine, this means SPECT-CT can differentiate whether a focal hot spot lies in a facet joint, a vertebral endplate, a pars interarticularis, or another bony structure, which cannot be accomplished with planar bone scans or SPECT alone. This improved localisation means that SPECT-CT detects more lesions or active degenerative sites than planar scintigraphy, as one analysis found that SPECT identified 24% more spinal lesions than planar bone scan in low back pain evaluation (Mulconrey et al., 2006).

Fig. 1.

SPECT-CT protocol. (SPECT) = single-photon emission computed tomography. (CT) = computed tomography. (SPECT-CT) = single-photon emission computed tomography-computed tomography. (IV) = intravenous. (99m-Tc DPD) = 99m-technetium diphosphonate.

1.2. Pathophysiology of lower back pain

Pain is a complex neurobiological phenomenon comprising a complex interplay between physiological and psychological processes. Regarding back pain, multiple potential culprits may be identified as causative in pain generation including ligament, muscle and disc injury (Mosabbir, 2023). Peripheral nociceptive signals from lumbar structures are carried by primary afferent neurons whose cell bodies lie in the dorsal root ganglia; their central processes enter the dorsal horn and are transmitted through spinal nociceptive pathways. Persistent nociceptive input may contribute to central sensitisation through spinal and supraspinal mechanisms. SPECT-CT does not image pain, inflammation or central sensitisation directly; it depicts radiotracer uptake associated principally with osteoblastic activity and bone turnover, which may or may not correspond to symptoms.

2. Methodology

A structured narrative literature search was conducted in PubMed. The final search was performed on 4/1/2026 and was restricted to English-language publications dated from 1 January 2000 to 31 December 2025. The primary search combined controlled vocabulary and title/abstract terms relating to single-photon emission computed tomography with terms relating to the lumbar spine and low back pain. The exact search strategy was: ((“Tomography, Emission-Computed, Single-Photon"[Mesh] OR SPECT[Title/Abstract] OR “SPECT/CT"[Title/Abstract] OR “SPECT-CT"[Title/Abstract] OR “single photon emission computed tomography"[Title/Abstract] OR “single-photon emission computed tomography"[Title/Abstract]) AND (“Lumbar Vertebrae"[Mesh] OR “Low Back Pain"[Mesh] OR lumbar[Title/Abstract] OR “lumbar spine"[Title/Abstract] OR “low back pain"[Title/Abstract] OR “lower back pain"[Title/Abstract])) AND (“2000/01/01"[Date - Publication]: “2025/12/31"[Date - Publication]) AND English[Language]. A supplementary targeted search was performed for emerging MRI techniques, including T1rho and T2 mapping, gagCEST, ultrashort and zero echo-time imaging, and synthetic CT. Reference lists of relevant reviews and included studies were also hand-searched for additional publications.

A supplementary targeted search was performed for emerging MRI techniques, including T1rho and T2 mapping, gagCEST, ultrashort and zero echo-time imaging, and synthetic CT. Reference lists of relevant reviews and included studies were also hand-searched for additional publications.

Eligible publications included diagnostic-accuracy studies, prospective and retrospective observational studies, case series, technical or quantitative imaging studies, reviews and relevant clinical guidelines. Eligible populations included adults with chronic low back pain, lumbar degenerative disease, facet arthropathy, disc degeneration or persistent pain following lumbar surgery. Outcomes of interest included detection and anatomical localisation of tracer uptake, association with symptoms or reference procedures, reader reliability, diagnostic performance, modification of management and clinical outcomes. Indirect evidence from sacroiliac, cervical, metastatic or peripheral-joint populations was retained only where it informed a specific technical or interpretive issue and was explicitly identified as contextual evidence.

Non-English publications, conference abstracts, editorials, publications outside the prespecified date range and studies without relevance to spinal pain assessment or SPECT-CT interpretation were excluded. One author performed the initial title and abstract screening and reviewed potentially eligible full texts. A second author reviewed the candidate publications and final included set, with disagreements resolved by discussion and consensus. The primary search yielded 350 records, of which 147 underwent full-text assessment. 33 publications were included in the final narrative synthesis, including 12 primary studies summarised in Table 1.

Table 1.

Characteristics, principal findings, directness and major limitations of primary studies included in the narrative review. Indirect or contextual evidence is identified separately and is not interpreted as direct evidence for lumbar degenerative pain.

Study	Design	Patient population and cohort (n)	Clinical indication	Index test	Principle Result	Directness	Major Limitation
Russo et al. (2017)	Prospective observational diagnostic/comparative study	Adults with CLBP (n = 99)	Assessing scintigraphic/osteoblastic activity of lumbar facet joints	Bone hydroxydiphosphonate (HDP) SPECT/CT,	Agreement between Parthria grade 3 and SPECT/CT positivity P < 0.001	Directly relevant to imaging degenerative lumbar facet disease	Lack of independent reference for pain generation, lack of diagnostic facet joint blocks
Harisankar et al. (2012)	Prospective observational	Adults with CLBP > 3 months (n = 30)	Evaluation of CLBP	99mTc bone scintigraphy with hybrid SPECT/CT	Incidence of neurologic symptoms in SPECT positive patients vs SPECT negative (p = 0.002)	Directly relevant	Lack of independent reference for pain generation
Cusi et al. (2013)	Prospective diagnostic accuracy study	Patients with SIJ post trauma or peripatrum period (n = 100)	Evaluation of SIJ dysfunction	99mTc bone SPECT/CT	Sensitivity and specificity of SPECT-CT in sacroiliac joint (95% and 99% respectively; inter-/intra-observer agreement κ = 0.85)	While not directly relevant to lumbar pathology, relevant study with regards to sensitivity and specificity of SPECT-CT	Lack of diagnostic SIJ injection
Tofuku et al. (2015)	Prospective observational diagnostic study	Patients with severe SIJ dysfunction refractory to conservative treatment (n = 32)	Evaluation of SIJ in patients with chronic pain	99mTC bone SPECT-CT using SUV	Mean SUV of symptom dominant side vs nondominant side in patients with lateralised symptom intensity (n = 32) (P = 0.02)	Low, but relevant study as it uses SUV in evaluation of SPECT-CT uptake	No independent reference standard
Lehman et al. (2016)	Retrospective observational cohort study	Adults with CLBP post inconclusive conventional imaging (n = 60)	Localisation of CLBP	99mTc bone SPECT-CT vs MRI	Agreement between SPECT/CT and MRI (n = 60, κ = −0.026, 95% confidence interval: −0.051, 0.000)	Directly relevant	Retrospective, no independent reference standard
Freiermuth et al. (2015)	Prospective randomised, placebo-controlled diagnostic accuracy	Patients with CLBP and suspected lumbar facet joint pain (n = 29)	Value of high-resolution SPECT/CT of lumbar spine prior to any diagnostic infiltration of the medial branches with respect to sensitivity and specificity	99mTc-DPD SPECT-CT	Value of high-resolution SPECT/CT of lumbar spine prior to any diagnostic infiltration of the medial branches with respect to sensitivity and specificity sensitivity 0.57 (95% CI: 0.18–0.90); specificity was 0.77 (CI 95% 0.55–0.92); odds ratio was 4.53 (CI 95% 0.75–27.40); diagnostic accuracy was 0.72.	Directly relevant	Small sample size
Malham et al. (2015)	Retrospective cross- sectional imaging correlation study	Patients with anterior column lumbar degenerative spine disease (n = 20)	Evaluation of reliability of bone SPECT-CT and conventional SPECT in identifying facet disease	99mTc-MDP SPECT-CT	Comparison of reliability of bone SPECT-CT and conventional bone SPECT scans (n = 20 scans, κ = 0.72 and 0.59 respectively). Reliability of bone SPECT-CT and conventional bone SPECT scans in disc pathology (κ = 0.81 and 0.72 respectively. Reliability of bone SPECT-CT and conventional bone in identifying facet disease (κ = 0.81 and 0.60 respectively)	High, assesses agreement between SPECT-CT and conventional bone SPECT	Major limitations: MRI use as comparator as opposed to reference standard
Mulconrey et al. (2006)	Retrospective interobserver reliability study	Adults with CLBP (n = 17)	Evaluation of imaging modalities with regards to interobserver reliability of SPECT-CT/bone scan	Lumbar SPECT with planar bone scintagraphy	SPECT-CT detection capability; 24% more lesions identified vs bone scan	High, evaluation of reliability of image interpretation	Small sample size, no independent reference standard
Brusko et al. (2019)	Retrospective single centre case series	Patients with cervical or lumbar spondylosis (n = 23)	Evaluation of SPECT-CT as a pre-operative assessment tool	SPECT-CT or bone SPECT	Symptom relief 12 months post-operatively (n = 23, 82.6%)	Moderate evaluation of preoperative localisations prior to surgical intervention, however mixed cervical/lumbar evidence	Small sample, no independent reference standard, retrospective design
Tender et al. (2019)	Retrospective observational case series	Patients with degenerative spinal disease and axial spine pain (lumbar subgroup, n = 23)	Preoperative localisation of pain generators when other imaging methods were insufficient	SPECT- CT	Patient outcome following surgery on SPECT/CT identified pain generators measured by Visual Analogue Scale (VAS) (n = 48). Cervical (n = 25): VAS pre-surgery. (8.8 ± 1.8) vs post (3.92 ± 2.2) (p = 0.019). Lumbar fusion (n = 23): VAS pre-surgery (9.35 ± 0.7) vs post (4.87 ± 2.3) (p = 0.008)	High clinical relevance	No independent reference standard, small sample size,
Kato et al. (2019)	Retrospective case series	Patients with elderly patient with degenerative lumbar spine disease and CLBP (n = 5)	Evaluation of primary pain generators	99mTc SPECT-CT	Application of SPECT-CT to guide conservative/minor surgical treatment decision making (n = 5; NRS pain score decreased from 8.4 ± 0.5, before treatment, to 2.2 ± 1.3, post-treatment (P < 0.05).)	High clinical relevance	Small sample size, retrospective case series
Kaiser et al. (2023)	Prospective observational cohort study	Patients with CLBP and single level lumbar degenerative disc disease (n = 38)	Patient outcome following surgery on SPECT-/CT identified pain generators measured by Visual Analogue Scale (VAS) and Oswestry Disability (ODI)	99mTc SPECT-CT	30% reduction in VAS and ODI → minimum clinically important difference (MCID) Secondary outcomes: patient reported satisfaction, return to work, willingness to re-undergo surgery and surgical complications, VAS pre-surgery 8.4 (±1.1), VAS post-surgery 3.2 (±2.5, p < 0.001); ODI pre-surgery 51.5 (±7.3) ODI post-surgery 20.7 (±14.68, p < 0.001); MCDI achieved in 84.2%; 71% patients satisfied with surgery, 89.4% reported willingness to re-undergo surgery; 82.9% returned to work	High clinical relevance	No control group, no independent confirmation

As this was a narrative rather than systematic review, study selection was purposive and intended to represent the principal clinical, diagnostic and technical issues. No formal risk-of-bias instrument or quantitative meta-analysis was undertaken. Evidence was interpreted according to study design, comparator or reference standard, directness to lumbar degenerative pain and major methodological limitations. Uncontrolled observational outcome studies were considered hypothesis-generating and were not interpreted as demonstrating causal treatment benefit.

3. SPECT-CT in the lumbar spine: current applications and rationale

Four distinct endpoints are considered throughout this review: (1) detection and anatomical localisation of increased tracer uptake; (2) association between uptake and symptoms or another clinical reference procedure; (3) modification of diagnostic or treatment decisions; and (4) improvement in patient outcomes. Detection of uptake does not by itself establish symptomatic relevance, a change in management does not necessarily represent improved management, and favourable outcomes in selected patients do not establish benefit over a conventional diagnostic pathway.

The primary application of SPECT-CT in lumbar spine disease has been in investigating chronic axial low back pain where conventional imaging is inconclusive. Clinicians have used SPECT-CT to identify active facet joint arthropathy, degenerative disc disease, sacroiliac (SI) joint dysfunction, or failed fusion (pseudarthrosis) as potential pain sources where other conventional imaging is inconclusive. For example, in facetogenic back pain, a facet joint showing high uptake on SPECT suggests active osteoarthritic remodelling, potentially indicating facets contribute to symptomatic pain SPECT-CT has been reported to detect facet joint disease in cases where MRI showed only mild degeneration, thereby potentially identifying painful facets that may otherwise be missed. In a series of 99 patients, Russo et al. noted that while many facets had degenerative changes on CT, only a subset showed SPECT uptake. The highest-grade arthropathies were most likely to be SPECT-positive, for example 69% of grade 3 facets showed uptake compared with ∼5% of grade 0–1 (Russo et al., 2017). This implies SPECT-CT might help distinguish “active” degeneration from quiescent, age-related changes. Similarly, for discogenic pain, SPECT-CT can reveal increased uptake at vertebral endplates adjacent to degenerated discs. Fig. 2 in Hirsch et al. described a patient with normal postoperative MRI but persistent pain whose SPECT showed uptake at the L3–L4 disc, leading to the diagnosis of an active discogenic pain source (Hirsch et al., 2024). Other reports have found that SPECT-CT uptake in anterior vertebral bodies correlates with degenerative disc findings on MRI/CT, though the clinical significance of this remains under study (Harisankar et al., 2012). Contextually, for sacroiliac joint (SIJ) pain, SPECT-CT can be particularly useful: Cusi et al. reported SPECT-CT achieved 95% sensitivity and 99% specificity in diagnosing SIJ dysfunction in postpartum or post-traumatic cases, far exceeding MRI sensitivity (Cusi et al., 2013). The high accuracy was attributed to SPECT-CT's ability to detect chronic repetitive stress changes due to increased tracer uptake in areas of ligamentous injury and subsequent calcium deposition, which can otherwise be missed on MRI in absence of acute oedema. Spine SPECT-CT also has niche roles in detecting occult injuries (e.g. pars fractures or insufficiency fractures), pseudoarthrosis after fusion, or loosening of hardware. It has also been used for postoperative back pain evaluation when MRI is obscured by metal: for instance, differentiating symptomatic adjacent segment disease or facet arthritis from soft-tissue causes (Awosika et al., 2024). In a case described by Scharf, a patient with a unilateral graft after spinal fusion underwent SPECT-CT, which revealed a lack of tracer uptake on one side showing failure of bone remodelling and graft incorporation; where other imaging studies were inconclusive (Scharf, 2015). While fusion of SPECT with CT may allow more precise anatomical localisation of increased osseous tracer uptake in spinal components, this should not be equated with confirmation that the identified structure is the source of pain without further clinical investigation.

Fig. 2.

A) Sagittal (I) SPECT and (II) SPECT-CT image depicting spondylolisthesis. B) Coronal (I) SPECT (II) CT (III) SPECT-CT highlighting spondylolysis. C) (I + II) Sagittal and (III + IV) Coronal CT and SPECT-CT slices showing instrumentation failure. D) (I) Lateral X-ray (II) Axial CT (III) Axial MRI (IV) Axial SPECT-CT highlighting instrumentation loosening. Adapted from Kato et al.(Kato et al., 2019), Israel et al. (Israel et al., 2019), Al-Riyami et al. (Al-Riyami et al., 2019) under the Creative Commons Attribution License. Note: orange arrows indicate pathology.

4. Challenges integrating SPECT-CT into clinical guidelines and practice

4.1. Lack of consensus on indications

Professional guidelines (e.g., North American Spine Society or neurosurgery associations) have not incorporated SPECT-CT as a required or routine part of preoperative workup for low back pain. For instance, the AANS/CNS guidelines for low back pain focus on clinical and traditional imaging factors for recommending fusion in degenerative disc disease, without mention of SPECT-CT -reflecting the guideline committees’ view that evidence was insufficient at that time to include it. Similarly, according to American College of Radiology Appropriateness Criteria SPECT-CT should not be used as part of standard initial diagnostic workup for lower back pain, but rather as part of a more targeted diagnostic effort (Expert Panel on Neurological et al., 2021/11).

4.2. Lack of standardized protocols for SPECT-CT interpretation

A fundamental limitation of spine SPECT-CT is the absence of standardized quantitative criteria or threshold values to define a “positive” finding. There is currently no universally accepted uptake cut-off (analogous to a standardised uptake value (SUV) threshold in PET) for calling a spinal facet joint or disc abnormal on SPECT. Instead, many practices use subjective comparison, for example a facet joint is considered SPECT-positive if it shows visibly greater tracer accumulation than adjacent levels or the surrounding bone. This subjectivity introduces variability and impedes comparisons between studies or institutions. As Ross et al. observed, bone scan reporting has traditionally been visual, and while quantitative SPECT technology (e.g. computing SUV) is emerging, “the rate of clinical implementation of quantitative bone SPECT remains slow” and substantial validation is needed (Ross et al., 2019). In practice, one clinician or institution might label a facet joint with moderate uptake as significant, whereas another might only consider it significant if uptake is above a certain fold increase over background, reflecting the lack of consensus on threshold values.

Technical factors also contribute to this issue. SPECT-CT scans from different cameras or reconstruction algorithms may yield different apparent intensities for the same biological uptake. Interoperator variability can also arise due to differences in readers with greater expertise in nuclear spine imaging being able to recognize subtle, but clinically relevant patterns with greater accuracy. The use of variable colour scales in SPECT-CT can exaggerate or diminish perceived uptake depending on how images are generated as well. Additionally, patient-specific factors such as body habitus can affect tracer distribution, and therefore compound outcome variability (Damian et al., 2024). One study confirmed that normalizing quantitative SPECT uptake to patient body weight significantly altered the “normal” reference values, with obese patients otherwise showing artifactually higher SUVs (Damian et al., 2024), further contributing to inter-operator discrepancies.

The lack of thresholding further contributes to clinical decision-making uncertainty. For example, in facet joint pain, some clinicians have proposed only treating facets that show dramatically increased uptake, assuming those that are contributing to symptomatic pain. However, there is no clear consensus on the level of uptake that should be deemed diagnostically significant. For context, in a prospective study of SI joint pain, Tofuku et al. attempted a semi-quantitative approach: they measured the degree of tracer accumulation in SI joints and found that higher uptake correlated with more severe symptoms and the need for advanced treatments (Tofuku et al., 2015). While this suggests a dose-response relationship for SI joints, such quantitative grading has not been widely translated to lumbar facet or disc disease. Indeed, non-standardized reporting makes comparisons across patients and studies difficult and complicates research efforts to establish the sensitivity or specificity of SPECT-CT at any given threshold of positivity.

Research is ongoing to produce SUVs that can be objectively compared: tangentially, efforts to calibrate SPECT/CT for bone have shown that metastatic lesions have significantly higher SUVs than benign bone lesions, and normal vertebrae have a relatively narrow SUV range (around 6 ± 1 in one series) (Lin et al., 2023). This led to proposals for SPECT SUV cut-offs of >8 or >11 to distinguish metastases from benign lesions in vertebrae. However, for degenerative spine disease, no such cut-off has been validated. These advances hint at the possibility of setting numeric criteria for “abnormal” uptake in the spine. However, experts caution that such numbers must be interpreted carefully, as technical limitations (partial-volume effects, resolution) and the complex bone biology can confound a simplistic SUV cut-off (Ross et al., 2019).

Another important limitation of SPECT-CT in practice is interobserver variability - different readers may interpret the same scan differently. A number of studies have quantified this phenomenon.

Reliability of identifying uptake lesions: Malham et al. conducted a surgeon survey where seven spine surgeons independently read a set of lumbar SPECT images (some fused with CT fusion, others not) to identify levels of “diagnostically relevant” uptake in discs or facets (Malham et al., 2015). The interobserver agreement was measured by kappa statistics. They found that agreement was only moderate for plain SPECT tomograms (κ = 0.59 for overall identification of an uptake site) but improved to substantial with SPECT-CT (κ = 0.72). In particular, identifying facet joint pathology was significantly more reliable with the addition of CT: κ = 0.81 for SPECT-CT vs 0.60 for SPECT alone. This confirms that the CT component helps readers agree on which joint is active, presumably by providing precise anatomical context. For planar bone scans or SPECT without CT, the reliability was worse, reflecting ambiguity in localizing the uptake. Thus, SPECT-CT has better interobserver reliability than planar or SPECT-alone.

Reliability vs. MRI: Interestingly, when compared to MRI interpretation, nuclear imaging tends to show lower interobserver agreement. A study by Mulconrey et al. had radiologists and orthopaedic surgeons interpret MRI, planar bone scans, and SPECT for the same patients (Mulconrey et al., 2006). They reported high κ values for identifying degenerative changes on MRI (0.73–0.77 for disc degeneration, spondylolisthesis, etc.), whereas bone scan and SPECT showed lower κ (∼0.54 for planar, ∼0.46 for SPECT). These values indicate only moderate agreement among readers for SPECT, significantly less consistent than MRI interpretations. This disparity likely arises because MRI offers clearer anatomic criteria (which are more objectively defined, like a Modic change presence/absence, disc height loss, etc.), while SPECT involves interpreting intensity gradients and deciding if uptake is above normal, which is more subjective. Additionally, biases come into play where some interpreters might “see” what they expect, that is if MRI is degenerated at L4–L5, they might be biased to call the SPECT hot at that level. In that same study, the readers’ ability to predict SPECT findings based on MRI was fairly high (positive predictive value ∼66%, negative ∼84%), suggesting MRI clues often hint at where SPECT would be positive. This could imply that truly new information from SPECT (beyond MRI) may be limited, or that readers unconsciously integrate MRI info into SPECT interpretation, which can skew objectivity.

4.3. Limited histopathological validation and gold standard comparison

How SPECT-CT findings correlate with actual histopathology of spinal tissues remains unknown as direct validation with a gold standard is scant. In degenerative spine disease, one rarely obtains tissue biopsies of “hot” lesions purely for diagnostic confirmation. The true-positive and false-positive rate of SPECT-CT is not definitively known.

Traditionally, the closest proxy for a gold standard in back pain diagnostics is often a diagnostic block or surgical outcome. For facets, medial branch blocks serve as the reference test for pain origin, as histologically a facet joint will virtually always show osteoarthritis in older adults regardless of pain (Curtis et al., 2026), although they are themselves imperfect. Their apparent diagnostic performance is influenced by procedural technique, local anaesthetic spread, the threshold used to define a positive response, placebo effects and false-positive or false-negative results. Agreement with diagnostic blocks is therefore clinically informative but does not represent validation against an unquestionable biological gold standard.

A randomised controlled trial by Freiermuth et al. is illustrative of this: it used diagnostic blocks as the comparator and found moderate accuracy of SPECT-CT (sensitivity ∼57%, specificity ∼77%) (Freiermuth et al., 2015). This implies that many SPECT-identified facets did not correlate with the “gold standard” of block-confirmed pain (and vice versa), highlighting limited validation. Similarly, for discogenic pain, provocation discography is sometimes considered a diagnostic gold standard (albeit a controversial one). Few studies have directly matched SPECT findings with discography results or post-fusion outcomes. One older study suggested that operating on SPECT-positive levels in multi-level degeneration led to better pain improvement than operating on SPECT-negative levels, but such data are retrospective and not uniformly confirmed (Tender et al., 2019).

When it comes to true histopathology data are even more sparse. In surgical cases, one could theoretically analyse resected facet joints or endplates from levels that were SPECT-positive to see what features are present (e.g. inflammation, microfractures, bone turnover markers). However, the literature on such correlations is minimal. There has been analogous research in peripheral joints: for example, in end-stage ankle osteoarthritis, researchers correlated ^99mTc bone scan uptake with histologic changes in the subchondral bone (Paul et al., 2015). They found that increased tracer uptake corresponded to areas of subchondral fibrosis and microfracture on histology. By analogy, one might suspect that a SPECT-hot facet joint has active subchondral bone changes (oedema, woven bone formation, microcracks) whereas a cold-but-arthritic facet might have sclerotic bone without ongoing turnover. Yet this remains an extrapolation until directly studied in the spine.

The lack of a clear gold standard is explicitly noted in comparative studies. In the facet SPECT vs. MRI comparison by Lehman et al., the authors caution that since neither modality had an independent gold standard for “painful facet,” it was impossible to conclude which modality was truly more accurate (Lehman et al., 2016). They found MRI identified far more facets with oedema (20.7% of facets) than SPECT did with uptake (7.3%). Without a criterion standard (like systematic facet blocks for all or pathology), we cannot be certain.

Moreover, the outcome studies on SPECT-guided treatment are limited. Small uncontrolled observational studies have reported favourable outcomes following treatment directed at SPECT-CT-positive sites; however, the absence of SPECT-negative or conventionally selected control groups precludes attribution of these outcomes to SPECT-CT-guided selection (Kato et al., 2019; Kaiser et al., 2023).

This hints that SPECT can help target the correct level for fusion in degenerative disc disease. However, these studies may be inherently biased, as surgeons likely chose the level based on SPECT, so the comparison isn't randomized. A more robust validation could consist of a trial whereby surgical planning is conducted with vs. without SPECT-CT and outcomes compared. To date, such trials have not been conducted, and thus strong evidence for improved outcomes with SPECT-informed treatment is lacking.

4.4. Uncertain correlation of SPECT uptake with pathology

Even when a focal hot spot is identified on SPECT-CT, the pathophysiological meaning of that finding is often uncertain. Brighter SPECT uptake may reflect a more inflamed or painful lesion, but it could also represent chronic remodelling that is clinically silent.

This dilemma is well illustrated in facet joint osteoarthritis. A facet joint that is “hot” on a bone scan suggests ongoing subchondral bone turnover (possibly due to stress or microfractures in the arthritic facet). One might assume this correlates with active inflammation or pain in that joint, yet, comparative studies cast doubt on a one-to-one correlation. Lehman et al. directly compared lumbar facet joint findings on SPECT-CT vs. MRI fat-suppressed sequences in the same patients (Lehman et al., 2016). Interestingly, they found no overall agreement between SPECT uptake and MRI evidence of inflammation (bone oedema or gadolinium enhancement) in facets - the κ statistic was ∼–0.03, essentially indicating random correspondence (Lehman et al., 2016). In other words, many facets that showed increased tracer uptake had no MRI oedema, and vice versa. This implies that the biological processes captured by MRI (oedema/inflammation in bone or soft tissue) and SPECT (osteoblastic activity) can diverge. A facet joint might have chronic bony remodelling (SPECT-positive) without acute inflammatory changes on MRI, or conversely, an oedematous painful facet (MRI-positive) might not yet have significant bone turnover to be SPECT-positive (Lehman et al., 2016). Thus, SPECT-CT and MRI appear to identify different facets as “abnormal” in many cases, and they should not be considered interchangeable tests. This inter-modality discordance underscores our poor understanding of what a SPECT hot spot truly signifies at the tissue level. Correlative studies with clinical outcomes compound uncertainty. Early reports suggested that a positive SPECT facet might predict better response to facet injections or radiofrequency ablation, on the premise that it marks the painful joint. Indeed, a review noted that “a positive finding on SPECT in facet arthropathy is associated with a significantly higher facet blockade effect” in some studies (Varga et al., 2023). However, higher-level evidence paints a more nuanced picture. In a double-blind randomised controlled trial, Freiermuth et al. tested whether SPECT-CT could identify patients with facet (zygapophyseal) pain by comparing SPECT results to diagnostic medial branch blocks (the gold-standard test for facet pain) (Freiermuth et al., 2015). They found SPECT-CT had only moderate sensitivity (∼57%) and specificity (∼77%) in predicting block-confirmed facet pain (Freiermuth et al., 2015). Many patients with true facet pain did not show SPECT uptake at the implicated joint (false negatives), and some without facet pain had SPECT “hot” facets (false positives). The authors concluded SPECT-CT could not reliably replace diagnostic blocks as a screening tool. This result suggests that the presence of osteoblastic activity on imaging does not perfectly align with the presence of nociceptive pain arising from that joint. Some facets may hurt without much bone turnover, and some may have bone turnover without pain, possibly because the remodelling is an adaptive or quiescent process. Without a biological marker of pain, SPECT uptake alone is an imperfect surrogate.

Similarly, the correlation of SPECT uptake with discogenic pain or inflammation is unclear. Degenerative disc disease often induces MRI-visible changes in adjacent vertebral endplates (Modic changes). Type I Modic changes (oedema) correlate with active inflammation, whereas Type II (fatty) indicate more chronic degeneration. One might expect SPECT to light up in Modic I (inflammatory) changes due to active bone turnover. However, clinical observations are mixed. Some practitioners note that SPECT can show increased uptake at vertebral endplates corresponding to Modic changes or high-intensity zones in discs, potentially indicating painful disc pathology. A small prospective series by Harisankar et al. reported that SPECT-CT uptake in anterior vertebral bodies correlated with degenerative changes on MRI at those levels (Harisankar et al., 2012). Yet, how that uptake relates to the disc's cellular processes (such as pro-inflammatory cytokines, neoinnervation, or marrow oedema) is not firmly established. We lack histological proof linking a SPECT-positive disc to specific pathological findings (e.g. granulation tissue or nerve in-growth in annular tears). Moreover, degenerative changes can be present without uptake. For instance, chronic endplate sclerosis (Modic Type III) might involve less active remodelling and hence be cold on SPECT, despite representing severe degeneration. Conversely, an actively healing endplate microfracture could be hot on SPECT but not distinguishable from typical degeneration on routine MRI. Ultimately, the mismatch between imaging signal and underlying biology is a well-recognized issue. Without insight into the cellular or molecular processes driving tracer accumulation, the clinical significance of a given uptake pattern remains uncertain.

It should also be noted that increased SPECT uptake is not specific to any single pathology - it is a relatively nonspecific indicator of bone turnover. Degenerative arthritis, stress reactions, occult fractures, infection, or even tumour could all cause focal uptake. In older patients, it is not uncommon to see multiple hot spots on a SPECT-CT of the spine, reflecting multilevel spondylosis or old injuries (Kato et al., 2019). Distinguishing which (if any) of those corresponds to the patient's pain requires careful clinical correlation as a patient's most symptomatic level may not always be the “hottest” on SPECT, and an intense uptake at one level could be incidental, contributing to diagnostic uncertainty. In summary, the correlation between SPECT signal intensity and the underlying pathological or pain-generating process is imperfect and remains an area of active inquiry. Improved understanding, perhaps via studies combining imaging with biochemical markers or tissue analysis, is needed to interpret SPECT-CT with more confidence in lumbar spine disease.

Collectively, the discordant findings presented so far indicate that SPECT-CT positivity is neither necessary nor sufficient to establish a symptomatic pain source. Increased uptake may occur at clinically silent sites, whereas symptomatic structures may not demonstrate increased uptake; positive and negative findings therefore require correlation with the clinical phenotype and other diagnostic information.

4.5. Unclear impact on surgical decision-making

While SPECT-CT can reveal areas of high metabolic activity potentially contributing to symptomatic pain, how a surgeon uses that information is variable. In theory, a positive SPECT signal at a certain level could sway a surgeon to fuse that level (or perform a targeted facet rhizotomy), whereas a negative SPECT might dissuade operating at a level that looks degenerated on MRI (assuming it's a bystander and not causing pain). Furthermore, small case series have shown value in reducing the invasiveness of surgery. Kato et al. described 5 elderly patients initially planned for multi-level fusion based on conventional imaging, in whom SPECT-CT pinpointed the pain source to one or two levels or even to a nonsurgical lesion (like an insufficiency fracture); as a result, three patients had a shorter fusion (1–2 levels) and two avoided fusion in favour of conservative treatment, all with good outcomes (Kato et al., 2019). Similarly, in a retrospective study by Brusko, 23 patients underwent spinal fusion at levels only indicated by positive SPECT-CT findings, regardless of varying MRI appearances (Brusko et al., 2019). Over 80% of patients in this study reported symptomatic pain relief and over a half achieved full resolution of symptoms (Brusko et al., 2019). This further suggests that SPECT-CT can refine treatment by preventing unnecessary extensive fusions. However, thesand the absence of SPECT-negative or conventionally selected controls limits the strength of this evidence. Indirectly, a retrospective study of the effect of SPECT-CT on surgical decision making in the management of OA of the foot and ankle found that its inclusion changed the diagnosis in 53% of the 107 patients, altering the treatment plan in 27% of cases (van Hasselt et al., 2023). Similar retrospective studies on the impact of SPECT-CT on surgical decision making in evaluation of lower back pain are yet to be carried out. There is a lack of randomized data demonstrating that SPECT-CT–guided surgery yields superior outcomes to standard care with respect to conventional diagnostic pathways. Without that, many surgeons are hesitant to rely heavily on an imaging test that might add cost and radiation without proven benefit.

4.6. Multi-level disease complexity

Many patients present with multi-level degeneration, and SPECT-CT often reveals tracer uptake at several levels. Integrating these findings into clinical decision-making remains complex. When adjacent discs demonstrate increased uptake, it can be difficult to determine which level is the primary pain generator and whether one or both levels require surgical intervention. Some practitioners might perform diagnostic injections at the levels indicated by SPECT to further confirm the pain source while others might treat the most symptomatic level first. In any case, the presence of multiple positive findings complicates surgical planning rather than simplifying it. Likewise, if SPECT is completely negative in a patient with severe back pain and multiple degenerative changes on MRI, it raises a diagnostic conundrum. Absence of tracer uptake may suggest that pain is not discogenic or facetogenic in origin but rather reflects a different mechanism – non-structural pain, muscular pain or central sensitisation processes. Some clinicians would interpret it that way and avoid surgical intervention, opting for conservative or non-fusion treatments. However, solid evidence to support that approach (i.e., that a negative SPECT predicts poor outcome from fusion) is lacking. Thus, integration into decision algorithms is still based on low-level evidence and surgeon preference.

4.7. Timing, logistics, communication and interpretation issues

From a practical standpoint, adding a SPECT-CT means another diagnostic step in the workup, which includes extra costs and radiation exposure. For some healthcare systems or patients, this is not readily accessible or may cause delays. If its contribution to planning is unclear, some surgeons prefer to proceed with established diagnostic measures (e.g., MRI + confirmatory blocks) rather than obtain SPECT-CT routinely. Instead, its use is assessed on a case-by-case basis, often reserved for complex or unclear cases. Several challenges hinder seamless integration. A proposed clinical pathway is outlined in Fig. 3.

Fig. 3.

Lower back pain management and the role of SPECT-CT in the diagnostic pathway.

A prudent approach advocated in the literature is to treat SPECT-CT as an adjunct to traditional workflows. For example, Hirsch et al. conclude that for facetogenic pain, SPECT-CT findings should be correlated with clinical exam, and any facet identified as a pain suspect by SPECT should be confirmed with a diagnostic block before definitive treatment (Hirsch et al., 2024). This layered approach (clinical exam → SPECT → confirmatory block → treatment) illustrates the cautious integration needed given the uncertainties. Until more robust evidence or standardization is achieved, SPECT-CT will remain a helpful but not standalone tool in surgical decision-making for lumbar spine disease.

Appropriate training, standardized reading protocols, and possibly quantitative metrics will be needed to improve consistency in interpreting SPECT-CT for lumbar spine disease. For now, clinicians should be aware of this variability and, when possible, seek expert interpretation in a multi-disciplinary consensus setting if the SPECT-CT result is pivotal in a patient's management.

5. Emerging MRI techniques as complementary and potential alternative approaches

While SPECT-CT currently serves as a valuable adjunct for identifying metabolically active sites within the degenerative lumbar spine, recent advances in magnetic resonance imaging (MRI) are closing the functional imaging gap. These novel MRI techniques aim to extract physiological, compositional, and even biomechanical information traditionally available only through nuclear imaging. As these MRI-based modalities mature, they may significantly reduce the clinical reliance on SPECT-CT or possibly render it unnecessary in certain diagnostic scenarios.

One of the most promising areas is quantitative MRI, which extends beyond structural imaging to assess tissue composition and microarchitecture. Techniques such as T1ρ and T2 mapping can quantify proteoglycan and collagen content within intervertebral discs, offering insight into disc health and matrix degeneration (Johannessen et al., 2006; Gullbrand et al., 2016). Unlike standard T2-weighted images that infer hydration qualitatively, these quantitative maps provide reproducible metrics that can detect early biochemical changes before overt morphological degeneration is visible. These imaging biomarkers have shown strong correlations with Pfirrmann grading and may help identify pain generators, challenging SPECT-CT's current niche in evaluating discogenic back pain.

A particularly exciting frontier is Chemical Exchange Saturation Transfer (CEST) MRI, especially gagCEST, which targets glycosaminoglycan content - a critical marker of disc integrity. gagCEST MRI allows for the direct, non-invasive assessment of proteoglycan levels within the nucleus pulposus (Latz et al., 2019). Degenerated discs with significant GAG depletion can be differentiated from age-related or asymptomatic degeneration. Thus, gagCEST offers a direct view into the disc's biochemical state unlike SPECT-CT, which infers bone remodelling as a proxy for adjacent disc pathology.

Parallel advances in ultrashort echo time (UTE) and zero echo time (ZTE) MRI are overcoming traditional limitations in imaging cortical bone (Larson et al., 2016). These sequences allow for the visualization of bone structures with contrast approaching that of CT, enabling the assessment of facet joints, pars defects, and endplate sclerosis - all common SPECT targets, within the same MRI exam. These sequences produce what is colloquially referred to as “synthetic CT” or “MR-based CT,” and are particularly beneficial in evaluating postoperative patients where metal artifact hampers standard imaging. When combined with standard soft tissue and nerve imaging, these MRI techniques provide a comprehensive, radiation-free evaluation of the lumbar spine. However, clinical adoption is in its infancy and synthetic CT remains primarily research based.

In addition, 3D MRI bone sequences such as volumetric interpolated breath-hold examination (VIBE) and susceptibility-weighted imaging (SWI) can render high-resolution, multiplanar images of the osseous spine, enabling assessment of facet arthropathy and subchondral changes with increasing fidelity. These methods are evolving rapidly and, in combination with AI-based post-processing, can simulate CT-like renderings of spinal anatomy, narrowing the diagnostic gap between MRI and SPECT-CT. Perhaps most significantly, these advanced MRI sequences offer functional information without ionizing radiation, a critical consideration in younger or recurrently imaged patients. With the integration of deep learning algorithms, MRI data can now be post-processed to derive structural, chemical, and even biomechanical metrics previously exclusive to nuclear medicine or CT-based studies.

Taken together, these developments suggest that a single, extended MRI protocol could soon replace the need for separate SPECT-CT exams in many cases. As novel MRI sequences become more validated and clinically accessible, the relative invasiveness, cost, and radiation exposure of SPECT-CT may limit its use to only highly specific cases, such as equivocal postoperative pain or suspected pseudarthrosis. However, these techniques should currently be regarded as emerging complementary or potential alternative approaches rather than established replacements for SPECT-CT. T1rho and T2 mapping and gagCEST primarily provide information regarding tissue composition, whereas UTE, ZTE and synthetic CT improve depiction of osseous anatomy. AI-based reconstruction and post-processing remain developmental. Although these techniques may improve structural or compositional characterisation, their ability to identify symptomatic spinal levels, alter treatment selection or improve outcomes has not been established in prospective comparative clinical pathways.

6. Limitations

This study has several limitations. Firstly, the quality of evidence presented is limited by the study design, which is a descriptive narrative review by nature. The search was restricted to a single database, PubMed, and English-language publications. Relevant evidence indexed exclusively in other databases or published in other languages may therefore have been missed, introducing potential retrieval and language bias. Reference-list screening improved coverage but could not eliminate this limitation. Absence of quantitative synthesis further limits the clinical impact of this study.

7. Conclusion and future directions

SPECT-CT is helpful in the evaluation of lumbar spine disease by offering a window into the metabolic activity of spinal structures that traditional imaging can miss. It holds particular promise in cases of chronic back pain by detecting areas of active bone remodelling and precisely localizing them, potentially improving diagnostic specificity and, in select cases, guiding targeted interventions. Reports of improved patient outcomes when SPECT-CT is used to guide treatment such as higher fusion success rates or avoidance of unnecessary surgery are encouraging. However, as this review has highlighted, significant limitations and unknowns temper the enthusiasm for widespread use.

First, the field lacks standardized quantitative thresholds for interpreting SPECT-CT, making results somewhat subjective and difficult to generalize. Second, the biological significance of SPECT uptake in degenerative spine conditions remains elusive - a hot spot is not a clear synonym for pain, and the disconnect between imaging findings and actual painful pathology (as evidenced by poor SPECT-MRI correlation and only moderate SPECT-block correlation) reminds us that we are still making indirect inferences about pain generators. Bridging that gap will require multidisciplinary research combining imaging with molecular and histological analysis of tissues, perhaps through animal models or special circumstances where tissue can be obtained.

Third, without a gold standard validation, it is hard to know how accurate SPECT-CT truly is. Future studies that use rigorous clinical endpoints (like sustained pain relief after treating a SPECT-positive target vs. a SPECT-negative one) can provide higher-level evidence. Randomized trials or at least prospective controlled studies would greatly aid in defining SPECT-CT's role. For example, a trial could randomize patients with multi-level degeneration to fusion at the level indicated by SPECT vs. the level indicated by conventional imaging - and then compare outcomes.

Fourth, to integrate SPECT-CT into clinical algorithms, consensus needs to be built among specialists. This might take the form of appropriateness indications and decision pathways for positive and negative results. Spine societies and working groups could use emerging evidence to formulate guidelines, which in turn could standardize practice and facilitate multicentre research.

Finally, the evolving landscape of imaging technology will influence SPECT-CT's future role. If advanced MRI sequences or PET tracers eventually provide more direct or reliable markers of pain-generating pathology, the reliance on SPECT-CT could decrease. Conversely, improvements in SPECT technology (e.g., newer SPECT cameras with higher resolution and quantitation) could enhance its value. For example, one could imagine a SPECT tracer that targets nerve inflammation or osteoclast activity specifically - that might yield a more pain-specific signal. Molecular imaging research is ongoing in that realm.

In conclusion, SPECT-CT may assist in anatomically localising sites of increased osseous metabolic activity in selected diagnostically challenging cases; however, its ability to identify the true symptomatic source, improve treatment selection or improve patient outcomes remains incompletely established. Therefore, SPECT-CT in the lumbar spine should be considered an adjunct rather than a standalone test or routine preoperative investigation. Its findings should be interpreted alongside the clinical phenotype, conventional imaging and, where appropriate, confirmatory procedures. Prospective controlled studies and standardised acquisition and interpretation criteria are required to determine whether SPECT-CT improves treatment selection or patient outcomes.

For spine surgeons, radiologists, and imaging scientists, ongoing collaboration will be key to uncover the full potential of SPECT-CT. By designing studies that address its unknowns and by leveraging advances in complementary modalities (like novel MRI), the goal is to improve the precision of diagnosing the source of back pain.

Consent to participate

Not applicable.

Author contributions

Katarzyna Pleśniar, Jake McDonnell and Rohan Tewari: writing and editing.

Stacey Darwish and Joseph S. Butler: conceptualisation, review and editing, supervision.

Ethical considerations

Not applicable.

Consent for publication

Not applicable.

Data availability statement

Not applicable.

Funding statement

Not applicable.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.