Molecular PET/CT in Systemic Sclerosis-Associated Interstitial Lung Disease (SSc-ILD)

Abstract

Systemic sclerosis-associated interstitial lung disease (SSc-ILD) is a leading cause of morbidity and mortality in patients with systemic sclerosis. High-resolution computed tomography (HRCT) provides anatomical detail but cannot directly assess disease activity, inflammation, or fibroblast activation. Molecular positron emission tomography/computed tomography (PET/CT) offers functional imaging that may complement structural assessment. This narrative review examines the role of molecular PET/CT in SSc-ILD, including fluorodeoxyglucose (FDG) PET/CT for metabolic activity assessment, fibroblast activation protein inhibitor (FAPI) tracers for fibrosis imaging, and other molecular probes targeting inflammation and tissue remodelling. We synthesize evidence on the diagnostic feasibility, prognostic value, and clinical applications of molecular PET/CT in SSc-ILD and related fibrotic interstitial lung diseases. Quantitative PET metrics, radiomics approaches, and artificial intelligence integration are also discussed. Although molecular PET/CT shows potential for phenotyping disease activity and predicting outcomes, current evidence is limited by small sample sizes and heterogeneous study designs. Standardization of imaging protocols, validation in multicenter cohorts, and integration with clinical and molecular biomarkers are needed before the clinical utility of molecular PET/CT in SSc-ILD can be established.

1. Search Strategy and Methods

1.1. Review Design

This narrative review was conducted in accordance with the Scale for the Assessment of Narrative Review Articles (SANRA) guidelines. The review aims to provide a comprehensive, clinically oriented synthesis of current evidence on molecular PET/CT in SSc-ILD, without the formal systematic screening and meta-analytic procedures of a systematic review.

1.2. Literature Search

A structured literature search was performed across four electronic databases: PubMed/MEDLINE, Embase, Scopus, and Web of Science. The search was conducted in January 2026 and covered publications from inception to December 2025. A total of 283 records were identified across all databases; after deduplication, 177 unique records were screened. Full search strings, database-specific record counts, and additional search notes are provided in Supplementary File S1.

Briefly, search terms combined controlled vocabulary (MeSH terms and Embase subject headings) with free-text keywords for systemic sclerosis, interstitial lung disease, and molecular PET imaging (including FDG, FAPI, and fibroblast activation protein). Equivalent keyword combinations with Boolean operators were applied in Scopus and Web of Science.

1.3. Inclusion and Exclusion Criteria

Inclusion criteria:

• Studies reporting PET/CT imaging (any tracer) in SSc-ILD or related fibrotic ILDs.
• Original research articles, case reports, case series, and conference abstracts.
• Review articles specifically addressing molecular imaging in ILD or SSc.
• Publications in English.

Exclusion criteria:

• Studies exclusively reporting non-PET imaging modalities without PET comparison.
• Animal studies without human data.
• Case reports unrelated to SSc-ILD or ILD.

1.4. Quality Appraisal

No formal quality appraisal tool (e.g., Newcastle–Ottawa Scale or QUADAS-2) was applied to the included studies. This is acknowledged as a limitation of the narrative review design (see Section 9.3). Evidence quality is discussed qualitatively throughout the review, and study design is indicated in all summary tables.

2. Introduction

Systemic sclerosis (SSc) is a chronic autoimmune connective tissue disease characterized by vasculopathy, immune dysregulation, and progressive fibrosis affecting the skin and internal organs. Interstitial lung disease (ILD) develops in a substantial proportion of SSc patients and represents a major determinant of prognosis [1]. SSc-ILD encompasses a spectrum of histopathologic patterns, most commonly nonspecific interstitial pneumonia (NSIP) and usual interstitial pneumonia (UIP), with variable inflammatory and fibrotic components [2]. The clinical course is heterogeneous; some patients experience stable or slowly progressive disease, whereas others develop rapid functional decline, requiring aggressive immunosuppression or lung transplantation.

Current diagnostic and monitoring strategies primarily rely on pulmonary function tests (PFTs), particularly forced vital capacity (FVC) and diffusing capacity for carbon monoxide (DLCO), and HRCT to assess the extent and pattern of lung involvement [3]. Although HRCT depicts structural abnormalities, such as ground-glass opacities, reticulation, and honeycombing, it cannot reliably distinguish between active inflammation and established fibrosis, or predict disease progression [4]. Biomarkers, such as Krebs von den Lungen-6 (KL-6), correlate with disease extent but lack specificity for disease activity [5].

Molecular imaging with PET/CT visualizes and quantifies the biological processes underlying SSc-ILD pathogenesis, including glucose metabolism, fibroblast activation, and immune cell trafficking. FDG PET/CT has been explored as a marker of inflammatory activity in ILD [6]. Tracers targeting fibroblast activation protein (FAP), a cell-surface protease expressed by activated fibroblasts, have emerged as tools for imaging fibrotic processes [7]. Other molecular probes targeting integrins, chemokine receptors, and macrophage activity are under investigation [8].

This narrative review synthesizes current evidence on molecular PET/CT in SSc-ILD and related fibrotic ILDs, examining the biological rationale, technical considerations, clinical applications, and limitations of FDG PET/CT, FAPI PET/CT, and other molecular tracers. Quantitative image analysis approaches, including radiomics and artificial intelligence, which may enhance disease phenotyping, prognostication, and treatment monitoring, are also discussed.

Figure 1 summarizes the pathophysiologic basis of SSc-ILD and the corresponding molecular PET imaging targets across inflammatory, fibrotic, epithelial/endothelial injury, and immune-cell trafficking pathways.

3. Clinical Background: SSc-ILD

3.1. Pathophysiology of SSc-ILD

Systemic sclerosis is characterized by three hallmark pathological processes: autoimmune activation, vasculopathy, and fibrosis. In the lung, these processes converge to produce SSc-ILD, with activated fibroblasts and myofibroblasts depositing excessive extracellular matrix, leading to progressive parenchymal distortion and functional impairment. Inflammatory cells, including macrophages, T lymphocytes, and mast cells, contribute to the early inflammatory phase, which may precede or accompany fibrotic remodelling.

3.2. Current Role of HRCT in SSc-ILD

HRCT remains the reference standard for diagnosing and monitoring SSc-ILD. The predominant pattern is NSIP, characterized by bilateral, symmetric ground-glass opacity and reticulation with basal predominance. UIP pattern, with honeycombing and traction bronchiectasis, portends a worse prognosis. HRCT’s extent correlates with PFT impairment and predicts outcomes [10], but is insensitive to early or active disease.

3.3. Pulmonary Function Tests and Biomarkers

PFTs, particularly FVC and DLCO, are the primary monitoring tools for SSc-ILD progression. A decline of ≥10% in FVC or ≥15% in DLCO is considered clinically significant. Serum biomarkers, including KL-6, SP-D, and CCL18, correlate with disease extent and activity but are not specific enough for clinical decision-making. Circulating fibroblast activation protein (FAP) has been explored as a serum marker of fibrotic activity in SSc-ILD [11].

3.4. Unmet Needs in SSc-ILD Assessment

Despite advances in HRCT and PFT monitoring, major unmet needs persist: (1) the inability to distinguish active inflammation from established fibrosis, (2) poor prediction of disease progression, (3) lack of validated tools for treatment response monitoring; and (4) the absence of molecular biomarkers for fibroblast activation in vivo. Molecular PET/CT has the potential to address these gaps, though prospective validation is required before clinical translation.

3.5. Technical Parameters for HRCT and PET/CT: Key Differences

A clinically important consideration for studies combining PET/CT and HRCT is whether the CT component of PET/CT provides sufficient image quality for ILD pattern classification (ground-glass opacity, reticulation, and honeycombing) or whether standalone diagnostic HRCT is required. Recent advancements in CT assessment of fibrotic ILDs have been reviewed elsewhere [12].

The key technical differences between the CT component of PET/CT and diagnostic HRCT are summarized in

Table 1. PET/CT vs. diagnostic HRCT: A technical comparison.

Parameter	PET/CT (CT Component)	Diagnostic HRCT
Tube voltage	120 kVp	120–140 kVp
Tube current	40–80 mAs (low-dose)	200–400 mAs
Slice thickness	3–5 mm	0.5–1.5 mm
Reconstruction kernel	Standard (soft tissue)	High-resolution (sharp/bone)
Primary purpose	Attenuation correction + anatomical localization	Detailed parenchymal assessment

.

These differences result in (1) higher image noise in PET/CT, reducing the conspicuity of subtle ground-glass opacities and fine reticulation; (2) lower spatial resolution in PET/CT, limiting the detection of small cysts, traction bronchiectasis, and early honeycombing; and (3) thicker slices in PET/CT, causing volume-averaging artefacts that obscure fine parenchymal detail.

While PET/CT can identify major ILD patterns (ground-glass opacity, consolidation, coarse reticulation, and honeycombing), it is not a substitute for diagnostic HRCT for detailed ILD pattern classification. This is reflected in the studies reviewed: Bellando-Randone et al. [6] performed FDG PET/CT and separate thin-section HRCT within one month; Peelen et al. [13] co-registered PET/CT with prior diagnostic HRCT; and Schniering et al. [14] used dedicated HRCT for radiomics analysis rather than the PET/CT component.

• Recommended Imaging Protocol for SSc-ILD

Optimal imaging protocols should include: (1) baseline assessment—diagnostic HRCT (0.5–1.5 mm slices, high-resolution kernel) for detailed pattern classification, plus molecular PET/CT (low-dose CT component) for metabolic activity assessment; (2) image co-registration—software-based registration of PET and diagnostic HRCT for spatial correlation of metabolic activity with structural patterns; (3) follow-up imaging—serial diagnostic HRCT for structural progression (standard of care), with selective PET/CT for disease activity in patients with unclear treatment response; and (4) radiation optimization—low-dose CT for PET/CT to minimize cumulative radiation from dual imaging.

Figure 2 summarizes the clinical PET/CT workflow and quantitative methods used for lung imaging in SSc-ILD.

4. FDG PET/CT in SSc-ILD

4.1. Biological Rationale for FDG PET/CT in SSc-ILD

¹⁸F-fluorodeoxyglucose (FDG) is a glucose analogue that accumulates in metabolically active cells, reflecting glucose uptake via GLUT transporters. In SSc-ILD, elevated FDG uptake in affected lung regions reflects increased metabolic activity of inflammatory cells (macrophages, neutrophils, and lymphocytes) and activated fibroblasts. FDG PET/CT thus provides a functional correlate of disease activity that is distinct from the structural information provided by HRCT. The hypothesis that metabolic activity may distinguish active, potentially reversible inflammatory disease from quiescent fibrosis provides the key rationale for FDG PET/CT in SSc-ILD [15].

4.2. Clinical Studies of FDG PET/CT in SSc-ILD

Multiple clinical studies have evaluated FDG PET/CT in SSc-ILD. Bellando-Randone et al. [6] demonstrated that normalized mean SUV (nmSUV)—a normalized metric comparing patient lung uptake against sex-, age-, height-, and weight-matched controls from a database of negative PET/CT scans—was elevated in SSc-ILD lung regions, with ground-glass opacity segments showing approximately 24% higher nmSUV than morphologically normal segments. This study used nmSUV rather than raw SUVmax or SUVmean; the normalization methodology is specific to this study and distinguishes it from other SSc-ILD FDG PET/CT studies. Peelen et al. [13] showed that FDG PET/CT detected metabolic activity in SSc-ILD patients with stable or improving PFTs, suggesting potential utility for disease activity monitoring beyond functional metrics. Ledoult et al. [16] reported that FDG uptake (SUVmax and the target-to-background ratio [TBR]) was elevated in SSc-ILD and correlated with PFT impairment. Broens et al. [17] evaluated FDG PET/CT in SSc patients eligible for autologous stem cell transplantation, demonstrating metabolic activity in ILD regions.

4.3. Prognostic Value

FDG PET/CT demonstrates emerging, though not yet validated, prognostic value in SSc-ILD. Lilburn et al. [18] conducted a prospective cohort study enrolling 45 patients with SSc-ILD (mean follow-up 44.8 months; 15 deaths recorded), demonstrating that the background lung minimum standardized uptake value (SUVmin)—reflecting diffuse metabolic activity in non-focal lung parenchyma—independently predicted patient survival. Background lung SUVmin enhanced the prognostic performance of the ILD-GAP index (a composite score incorporating ILD diagnosis, age, gender, and PFT parameters), improving the C-statistic for survival prediction. This study represents an important advance in quantitative FDG PET/CT prognostication for SSc-ILD; however, replication in independent multicenter cohorts is required before clinical adoption.

4.4. Comparison with HRCT and PFTs

Quantification methods vary across studies, with some employing maximum SUV (SUVmax), others using mean SUV (SUVmean) or TBR, and others calculating volumetric metrics. The optimal metric for SSc-ILD assessment remains unknown. Respiratory motion can degrade image quality and quantification accuracy, particularly in the lung bases; deep-inspiration breath-hold acquisitions may mitigate this limitation [19].

4.5. Radiation Dosimetry for Serial Imaging

Radiation exposure is an important consideration for serial PET/CT imaging in SSc-ILD patients, who may require multiple scans over years for disease monitoring and treatment response assessment.

Approximate effective dose estimates for FDG PET/CT, FAPI PET/CT, diagnostic HRCT, and annual background radiation are summarized in

Table 2. Effective radiation dose estimates for PET/CT and HRCT imaging.

Modality	Effective Dose per Exam	Source
18F-FDG PET/CT (standard protocol)	~7–14 mSv	FDG ~5–7 mSv + CT ~2–7 mSv
68Ga-FAPI PET/CT (standard protocol)	~4–6 mSv	FAPI ~2–3 mSv + CT ~2–3 mSv
Diagnostic HRCT	~1–3 mSv	High-resolution protocol
Annual background radiation	~3 mSv/year	ICRP reference

.

• Cumulative Dose Estimates for Serial Imaging (illustrative 2-year protocol):

• FDG PET/CT every 6 months (4 scans): ~28–56 mSv cumulative.
• FAPI PET/CT every 6 months (4 scans): ~16–24 mSv cumulative.
• ICRP occupational recommendation: <20 mSv/year averaged over 5 years.

• Clinical Implications:

The cumulative radiation dose from serial FDG PET/CT may approach or exceed ICRP occupational recommendations in intensive monitoring protocols. FAPI PET/CT offers a dosimetric advantage (approximately a 40–60% lower dose than FDG PET/CT) due to lower injected activity and shorter biological half-life. Dose-reduction strategies include: (1) low-dose CT protocols for attenuation correction (replacing diagnostic CT); (2) iterative reconstruction algorithms; (3) weight-based activity dosing; and (4) longer interscan intervals where clinically appropriate. Radiation risk must be weighed against the clinical benefit of serial functional imaging in individual patients, particularly in younger patients or those requiring frequent monitoring.

4.6. Limitations of FDG PET/CT Evidence in SSc-ILD

The FDG PET/CT evidence base in SSc-ILD has several important limitations that must be acknowledged when interpreting the findings:

• Small sample sizes: Most published studies enrolled 10–30 patients, with only a few exceeding 40 participants [6,13,17,18]. The largest prospective cohort to date (Lilburn 2025 [18]) enrolled 45 patients with a mean 44.8-month follow-up, representing a significant advance but still insufficient for definitive clinical validation.
• Heterogeneous study designs: Studies vary in patient selection criteria, disease duration, immunosuppressive treatment exposure, PET acquisition protocols, and quantification methods, limiting cross-study comparisons.
• Single-centre studies: All published SSc-ILD FDG PET/CT studies are single-centre, introducing potential selection bias (see Section 9.5).
• Lack of standardized quantification: No consensus exists on the optimal PET metric (SUVmax, SUVmean, TBR, or volumetric measures) for SSc-ILD assessment (see Section 7.2).
• Absence of prospective treatment response data: No randomized controlled trial has used FDG PET/CT as a primary endpoint or response biomarker in SSc-ILD.

The FDG PET/CT evidence base in SSc-ILD is summarized in

Table 3. FDG PET/CT studies in SSc-ILD.

Study	Year	Study Design	n	Tracer	Key Finding
Bellando-Randone [6]	2019	Letter—Retrospective case series	7	18F-FDG	nmSUV (normalized mean SUV) elevated in GGO segments; GGO segments ~24% higher uptake than morphologically normal segments
Ledoult [16]	2021	Full article—Pilot study	19	18F-FDG	SUVmax and TBR elevated; correlates with PFT impairment
Peelen [13]	2020	Full article—Cross-sectional	30	18F-FDG	PET detects metabolic activity in stable PFT patients
Uehara [19]	2016	Full article—Retrospective	18	18F-FDG	Breath-hold improves image quality; mixed CTD population
Bastos [20]	2022	Full article—Cross-sectional	23	18F-FDG	SULmean elevated across all HRCT patterns; FDG cannot distinguish fibrotic patterns in chronic SSc; CCL2 correlated with FVC
Broens [17]	2022	Full article—Retrospective	15	18F-FDG	Metabolic activity in SSc-ILD pre-ASCT
Broens [21]	2023	Conference abstract (EULAR 2023) *	13	18F-FDG	Early SSc (≤2 years): FDG uptake higher in ILD vs. no-ILD (p = 0.03); preliminary data only
Peelen [22]	2017	Conference abstract *	12	18F-FDG	Pilot feasibility
Ivorra [23]	2018	Conference abstract (EULAR 2018) *	14	18F-FDG	PET predicts PFT decline
Nishiyama [24]	2010	Conference abstract *	10	18F-FDG	FDG in connective tissue disease
Lilburn [18]	2025	Full article—Prospective cohort	45	18F-FDG	Background lung SUVmin independently predicts survival; enhances ILD-GAP index prognostic performance

* Conference abstracts represent preliminary data without full peer review and should be interpreted with caution. Abbreviations: n = sample size; SUVmax = maximum standardized uptake value; TBR = target-to-background ratio; CTD = connective tissue disease; ASCT = autologous stem cell transplantation; ILD-GAP = ILD–Gender–Age–Physiology index; GGO = ground-glass opacity; nmSUV = normalized mean SUV; SULmean = mean standardized uptake value lean body mass-corrected.

.

5. FAPI PET/CT in SSc-ILD

5.1. Biological Rationale

Fibroblast activation protein (FAP) is a type II transmembrane serine protease expressed by activated fibroblasts and myofibroblasts in fibrotic and inflammatory conditions. FAP expression is minimal in normal adult tissues but markedly upregulated in activated fibroblasts in fibrotic lung disease, cancer-associated stroma, and wound healing. Small-molecule FAP inhibitors (FAPIs) radiolabeled with ⁶⁸Ga or ¹⁸F enable PET imaging of FAP-expressing cells, providing a direct molecular signal for fibroblast activation—a key pathological process in SSc-ILD.

5.2. Clinical Studies in SSc-ILD

FAPI PET/CT evidence in SSc-ILD is currently limited to a small number of studies, and the evidence base must be interpreted with caution given the small sample sizes and heterogeneous populations.

Schmidkonz et al. [7] conducted a single-centre pilot study of ⁶⁸Ga-FAPI-04 PET/CT in SSc-ILD, demonstrating elevated FAP expression in fibrotic lung regions and correlation with HRCT extent and PFT impairment. Lam et al. [25] reported ⁶⁸Ga-FAPI-04 PET/CT findings in five patients with fibrotic ILD (including SSc-ILD), showing heterogeneous FAP uptake patterns. It is important to note that the Lam et al. [25] cohort included patients with fibrotic ILD broadly defined, not exclusively SSc-ILD patients; therefore, the findings may not be fully generalizable to SSc-ILD specifically. Khadka et al. [26] described a single case report of FAPI PET/CT in ILD, demonstrating the feasibility of FAP imaging.

A recently published exploratory study by Tai et al. [27] evaluated early (10 min) and late (60 min) ⁶⁸Ga-FAPI-46 PET/CT volume-based parameters in SSc-ILD patients, comparing progressive (n = 8) versus non-progressive disease (n = 14), defined by the INBUILD trial criteria. Key findings included: (1) early total lesion FAPI uptake (TL-FAPI at 0 min, median 1245 vs. 714, p = 0.020) and affected lung volume (AV at 0 min, median 514 mL vs. 368 mL, p = 0.042) were significantly higher in progressors than non-progressors; (2) early PET parameters showed stronger, more consistent correlations with PFT and quantitative HRCT parameters than late PET parameters; (3) volume-based parameters more consistently correlated with PFT and HRCT than intensity-based parameters; and (4) TL-FAPI at the 0 min threshold of 928 had the best diagnostic performance (AUC 0.80). The authors concluded that early ⁶⁸Ga-FAPI-46 PET/CT imaging, specifically early volume-based parameters, was associated with recent disease progression in SSc-ILD and that prospective validation in treatment response studies is needed [27].

5.3. Comparison with FDG PET/CT

FAPI PET/CT offers several theoretical advantages over FDG PET/CT in SSc-ILD: (1) direct molecular targeting of activated fibroblasts (the primary effector cell in fibrosis), rather than nonspecific metabolic activity; (2) lower background uptake in normal lung, potentially improving the signal-to-noise ratio; and (3) lower radiation dose (~4–6 mSv vs. ~7–14 mSv for FDG PET/CT). However, head-to-head comparison data in SSc-ILD are currently absent, and these theoretical advantages require prospective validation.

Figure 3 presents a proposed conceptual framework for molecular PET/CT-guided management of SSc-ILD.

5.4. Limitations of FAPI PET/CT Evidence in SSc-ILD

The FAPI PET/CT evidence base in SSc-ILD is at an early stage and subject to important limitations.

• Very limited SSc-ILD-specific data: As of the search date, FAPI PET/CT evidence in SSc-ILD consists of one single case report (Khadka 2025 [26], n = 1), one small case series in fibrotic ILD (Lam 2023 [25], n = 5, not exclusively SSc-ILD), one single-centre pilot study in SSc-ILD (Schmidkonz 2021 [7], n = 12), and one exploratory study in SSc-ILD (Tai 2026 [27], n = 22).
• Tracer heterogeneity: Published studies have used different FAPI tracers (⁶⁸Ga-FAPI-04 and ⁶⁸Ga-FAPI-46), different imaging timepoints (10 min and 60 min), and different quantification methods, limiting comparability.
• Absence of longitudinal data: No published study has evaluated FAPI PET/CT as a longitudinal monitoring tool in SSc-ILD with serial imaging over time.
• No treatment response data: No study has evaluated FAPI PET/CT changes in response to immunosuppressive or antifibrotic therapy in SSc-ILD.

The FAPI PET/CT evidence base in SSc-ILD and related fibrotic ILDs is summarized in

Table 4. FAPI PET/CT studies in SSc-ILD and related fibrotic ILDs.

Study	Year	Study Design	n	Tracer	Population	Key Finding
Schmidkonz [7]	2021	Full article—Single-centre pilot	12	68Ga-FAPI-04	SSc-ILD	FAP uptake correlates with HRCT extent and PFTs
Lam [25]	2023	Full article—Case series	5	68Ga-FAPI-04	Fibrotic ILD (not exclusively SSc-ILD) *	Heterogeneous FAP uptake; feasibility demonstrated
Khadka [26]	2025	Full article—Case report **	1	68Ga-FAPI	ILD	Single case; feasibility only
Tai [27]	2026	Full article—Exploratory prospective	22	68Ga-FAPI-46	SSc-ILD	Early volume-based parameters (TL-FAPI at 0 min) distinguish progressors from non-progressors (AUC 0.80); early timepoint outperforms late timepoint

* Lam et al. [25] included patients with fibrotic ILD broadly defined, not exclusively SSc-ILD. Findings may not be fully generalizable to SSc-ILD specifically. ** Case report (n = 1) evidence should be interpreted as feasibility only. Note: FAPI tracer variants differ between studies (⁶⁸Ga-FAPI-04 vs. ⁶⁸Ga-FAPI-46); these are distinct compounds with potentially different pharmacokinetics, and direct cross-study comparisons should be made with caution. Abbreviations: FAP = fibroblast activation protein; TL-FAPI = total lesion FAPI uptake; AUC = area under the curve; HRCT = high-resolution computed tomography; PFTs = pulmonary function tests.

.

6. Other Molecular Tracers in SSc-ILD and Related Fibrotic ILDs

6.1. Overview of Emerging Tracers

Beyond FDG and FAPI, several other molecular tracers have been investigated in interstitial lung disease, as reviewed by Broens et al. (2022) [8]. These include tracers targeting chemokine receptors (CXCR4), integrins (αvβ3), macrophage activity, and other fibrotic pathways.

Table 5. Other molecular tracer studies in SSc-ILD and related fibrotic ILDs.

Study	Year	Study Design	Tracer	Population	Key Finding
Schniering [9]	2019	Full article—Preclinical + human pilot	Integrin αvβ3 tracer	SSc-ILD/ILD	Feasibility of integrin αvβ3-targeted imaging in ILD demonstrated
Kopp [28]	2025	Full article	CXCR4 ligand	SSc-ILD	CXCR4 uptake correlates with peripheral CXCR4+ immune cell counts
Kopp [29]	2025	Conference abstract (EULAR 2025)	CXCR4 ligand	SSc-ILD	Peripheral CXCR4+ immune cells correlate with CXCR4 PET uptake in lung regions; preliminary data only
Branley [30]	2008	Full article	Macrophage tracer (11C-PK11195)	Scleroderma alveolitis (FASSc); n = 15 FASSc + n = 7 controls	Trend of reduced PK11195 uptake in FASSc (p = 0.09, not statistically significant); lung density significantly elevated (p < 0.005)
Broens [8]	2022	Review article (not a primary clinical study)	Various (MMPs, collagen markers, apoptosis markers)	Preclinical/multiple	Comprehensive narrative review of novel and preclinical PET tracers for ILD molecular imaging

Conference abstract (EULAR 2025): preliminary data without full peer review; interpret with caution. Notes: Broens et al. [8] (Autoimmunity Reviews, 2022, 21(12):103202) is a narrative review article and is not a clinical study. It is cited as the primary source for preclinical tracer content in Section 6.5. It should not be confused with Broens et al. [17] (Frontiers in Immunology, 2022), which is a separate clinical study of FDG PET/CT in SSc patients. All studies in this table are single-centre and represent early stage or feasibility-level evidence. None have been replicated in independent cohorts. Abbreviations: CXCR4 = C-X-C chemokine receptor type 4; MMPs = matrix metalloproteinases; FASSc = fibrosing alveolitis in systemic sclerosis; αvβ3 = alpha-v beta-3 integrin.

summarizes clinical studies that have applied these tracers in SSc-ILD or related fibrotic ILDs. Additional preclinical tracers targeting matrix metalloproteinases (MMPs), collagen synthesis, and apoptosis pathways are under development but have not yet been evaluated in SSc-ILD patients.

The evidence base for other molecular tracers in SSc-ILD and related fibrotic ILDs is summarized in Table 5.

6.2. CXCR4-Targeted Imaging

Kopp et al. [28] evaluated CXCR4-based PET imaging in SSc-ILD using a radiolabeled CXCR4 ligand. CXCR4 is a chemokine receptor expressed by immune cells and fibroblasts in inflammatory and fibrotic conditions. The study demonstrated elevated CXCR4 uptake in SSc-ILD lung regions, correlating with peripheral CXCR4+ immune cell counts. These findings are exploratory and require replication in larger prospective cohorts.

6.3. Integrin-Targeted Imaging

Schniering et al. [9] evaluated the molecular imaging of integrin αvβ3 and somatostatin receptor 2 in ILD. Integrin αvβ3 is expressed by activated fibroblasts and endothelial cells in fibrotic tissue. The study demonstrated the feasibility of integrin-targeted imaging in SSc-ILD and related fibrotic ILDs. These findings are preliminary and limited to a single study.

6.4. Macrophage-Targeted Imaging

Branley et al. [30] performed PET scanning of macrophages in patients with scleroderma fibrosing alveolitis (SSc-ILD) using ¹¹C-PK11195. The study identified a trend toward reduced PK11195 uptake in patients with fibrosing alveolitis compared to controls (p = 0.09), which did not reach statistical significance. Lung density was significantly elevated (p < 0.005). Macrophage-targeted imaging may provide insights into the inflammatory phase of SSc-ILD, but this evidence is limited to a single study using older PET technology.

6.5. Preclinical and Emerging Tracers

Broens et al. (2022) provide a comprehensive review of emerging and preclinical PET tracers for the molecular imaging of interstitial lung disease [8]. Additional PET imaging targets under exploration include:

• Matrix metalloproteinases (MMPs), which mediate extracellular matrix remodelling. Radiolabeled MMP inhibitors have shown uptake in fibrotic lung tissue in preclinical models, but no human SSc-ILD data are available [8].
• Collagen synthesis markers, including radiolabeled collagen-binding peptides targeting newly synthesized type I collagen. These tracers aim to directly quantify active fibrogenesis but remain in preclinical stages [8].
• Apoptosis markers, including radiolabeled annexin V and caspase-targeted probes. These may provide insights into cell death pathways in SSc-ILD but have not been evaluated in human ILD cohorts [8].
• Other emerging targets include lysyl oxidase (involved in collagen crosslinking), transglutaminase, and cytokine receptors (e.g., IL-13Rα2 and TGF-β receptors) [8].

Most of these approaches are currently in the preclinical or early clinical stages, with no human SSc-ILD data available. Translation to clinical SSc-ILD studies will require demonstration of target expression in SSc-ILD lung tissue, development of GMP-grade radiolabeled compounds, and phase I/II safety and feasibility studies [8].

It should be noted that Broens et al. [8] (Autoimmunity Reviews, 2022, 21(12):103202) is a narrative review article summarizing preclinical and early clinical tracers for ILD molecular imaging. It is not a clinical study and should not be confused with Broens et al. [17] (Frontiers in Immunology, 2022), which is a separate clinical study of FDG PET/CT in SSc patients eligible for autologous stem cell transplantation.

6.6. Limitations of Other Tracer Evidence

Evidence for non-FDG, non-FAPI tracers in SSc-ILD is extremely limited.

• CXCR4 imaging: Evidence is limited to one full article (Kopp 2025 [28]) and one conference abstract (Kopp 2025 [29], preliminary conference data). These represent two separate publications reporting related data; the conference abstract [29] should be interpreted as preliminary evidence pending full peer-reviewed publication.
• Integrin imaging: Evidence consists of one study in SSc-ILD/ILD (Schniering 2019 [9]). No replication data exist.
• Macrophage imaging: Evidence consists of one study in scleroderma alveolitis (Branley 2008 [30], n = 15 FASSc + n = 7 controls), with a non-significant trend for PK11195 uptake reduction (p = 0.09). This is the oldest study in the review and uses older PET technology.
• Preclinical tracers: No human SSc-ILD data exist for MMPs, collagen synthesis markers, or apoptosis markers. All evidence is from animal models or other disease contexts [8].

7. Quantitative PET Metrics and Artificial Intelligence

7.1. Overview of Quantitative PET Metrics

Quantitative PET metrics in SSc-ILD are in early investigational stages and have not been validated for routine clinical use. The following metrics have been reported in published studies, but standardization and prospective validation are needed before clinical implementation.

Quantitative PET metrics in SSc-ILD include SUVmax, SUVmean, TBR, and volumetric measures, such as metabolically active lung volume. SUVmax reflects the peak tracer uptake in the most active region, whereas SUVmean represents the average uptake over a defined volume. TBR normalizes lung uptake to the background (typically the blood pool or reference tissue), potentially reducing variability.

7.2. Limitations of SUVmax and Alternative Quantification Metrics

The standardized uptake value (SUV) is the most widely used quantification metric in PET imaging, normalizing tracer uptake to injected activity and body weight. However, different SUV metrics have distinct advantages and limitations that influence their suitability for SSc-ILD assessment.

7.2.1. SUVmax (Maximum SUV)

SUVmax represents the highest voxel value within a region of interest (ROI) and is the most commonly reported metric in the studies reviewed.

Advantages: Simple to measure (single-voxel and no ROI drawing required); highly reproducible across readers; and sensitive to focal areas of high metabolic activity.

Limitations: Sensitive to image noise—a single noisy voxel can artificially elevate SUVmax, making it susceptible to statistical outliers; susceptible to partial volume effects (underestimation in small lesions); does not reflect overall disease burden or spatial extent; and may not capture heterogeneous disease activity across multiple lung regions in diffuse ILD.

7.2.2. SUVmean (Mean SUV)

SUVmean represents the average SUV within a defined ROI. Bellando-Randone et al. [6] employed a normalized mean SUV (nmSUV)—a normalized metric referenced to matched controls—in SSc-ILD, demonstrating that ground-glass opacity regions show approximately 24% higher nmSUV than morphologically normal segments. This normalized approach is the only SSc-ILD study using PET metrics referenced to matched controls, and the methodology is specific to that study.

Advantages: More stable than SUVmax (less sensitive to noise and outliers); reflects average metabolic activity across the ROI; and better suited for heterogeneous, diffuse diseases such as SSc-ILD.

Limitations: Requires manual or semi-automated ROI definition (inter-reader variability); sensitive to ROI size and placement; and may dilute focal areas of high activity if ROI includes normal tissue.

7.2.3. SUVpeak (Peak SUV)

SUVpeak represents the mean SUV within a small fixed-volume ROI (typically 1 cm³ sphere) centred on the hottest region—a compromise between SUVmax’s simplicity and SUVmean’s stability.

Advantages: Less sensitive to noise than SUVmax; recommended by PERCIST (PET Response Criteria in Solid Tumors) for oncology.

Limitations: Limited validation in ILD populations; may not be optimal for diffuse lung disease (designed for focal lesions).

7.2.4. Volumetric Metrics

Volumetric metrics integrate both tracer uptake intensity and spatial extent. Metabolically Active Volume (MAV) and Total Lesion Glycolysis (TLG) have been used in SSc-ILD studies [18]. These metrics may better reflect overall disease burden but are highly sensitive to threshold selection, and no consensus threshold exists for ILD.

7.2.5. Target-to-Background Ratio (TBR)

The TBR normalizes lesion uptake to background reference tissue and has been used in SSc-ILD FDG PET/CT studies [13,16]. The TBR may reduce variability from technical factors but requires the definition of “normal” reference tissue—a challenge in diffuse ILD—where the background lung may itself be affected.

7.2.6. Recommendation

Until consensus guidelines are established, researchers should report multiple complementary metrics (e.g., SUVmax, SUVmean, and MAV) to enable comprehensive assessment and facilitate future meta-analyses. Standardization of the PET quantification methodology is essential for multicenter studies and clinical implementation.

7.3. Harmonization with HRCT and PFTs

The integration of PET metrics with HRCT and PFT data may improve the diagnostic and prognostic value of molecular imaging in SSc-ILD. Several studies have demonstrated correlations between FDG uptake and HRCT extent scores, as well as PFT impairment [6,13,16]. However, the added value of PET over HRCT alone for clinical decision-making remains to be established in prospective studies.

7.4. Radiomics and Texture Analysis

Radiomics approaches extract quantitative features from medical images, including PET images, to characterize tissue heterogeneity and predict outcomes. In SSc-ILD, CT-based radiomics has demonstrated prognostic value [14,31,32]. Radiomics approaches have also been applied to other fibrotic ILDs, including idiopathic pulmonary fibrosis [33], providing methodological context for SSc-ILD radiomics development. Schniering et al. [14] integrated CT radiomics with molecular data to decode prognostic and molecular differences in SSc-ILD. It is important to note that the majority of AI/radiomics studies in

Table 6. AI and radiomics studies in SSc-ILD.

Study	Year	Study Design	Imaging	Key Finding
Schniering [14]	2021	Full article	CT + PET (only study integrating PET/HRCT radiomics)	CT-based radiomics decodes prognostic and molecular differences in SSc-ILD
Chassagnon [34]	2021	Full article	CT only	Elastic registration-driven deep learning for longitudinal SSc-ILD assessment
Walsh [35]	2022	Full article	CT only	Deep learning outcome prediction in progressive fibrotic ILD
Stock [36]	2024	Full article	CT only	Deep learning algorithm to detect UIP pattern in SSc-ILD
Jia [37]	2024	Full article	CT only	Explainable AI CT scoring of ILD
Lee [38]	2025	Full article	CT only	Risk factors for progressive fibrosing ILD using deep learning CT quantification
Ammeter [39]	2024	Full article	CT only	Spatial heterogeneity of ILD by integrative radiomics and single-nucleus transcriptomics
Hoang-Thi [40]	2023	Full article	CT only	Prognostic value of automated CT assessment of ILD in systemic sclerosis
Kim [41]	2010	Full article	CT only	Computer-aided diagnosis for quantitative scoring of lung fibrosis extent in scleroderma
Chassagnon [31]	2020	Full article	CT only	Deep learning-based automated assessment of SSc-ILD on CT images
Martini [32]	2021	Full article	CT only	Applicability of radiomics in SSc-ILD: proof of concept

Note: All studies except Schniering et al. [14] used HRCT alone for AI/radiomics analysis. The AI/radiomics evidence base in SSc-ILD is therefore predominantly CT-based rather than PET-based. Claims about AI integration with PET imaging in SSc-ILD should be interpreted with this limitation in mind. Note: Fröhner et al. [12] (J Thorac Imaging, 2023) is a general CT review of fibrotic ILDs and is cited as background context for the CT assessment methodology, not as a primary SSc-ILD AI/radiomics study; it is therefore not included as a row in this table. Refaee et al. [33] (Front Med, 2022) addresses radiomics diagnosis of idiopathic pulmonary fibrosis (IPF), not SSc-ILD, and is cited as a related fibrotic ILD radiomics reference; it is not included as a row in this table. Abbreviations: UIP = usual interstitial pneumonia; CT = computed tomography; HRCT = high-resolution computed tomography; PET = positron emission tomography; AI = artificial intelligence.

used HRCT alone for analysis; only Schniering et al. [14] integrated PET/HRCT radiomics. The AI/radiomics evidence base in SSc-ILD is therefore predominantly CT-based rather than PET-based.

7.5. Artificial Intelligence Integration

AI and deep learning approaches for SSc-ILD assessment are in early investigational stages and may potentially enable automated pattern recognition, disease quantification, and outcome prediction. However, clinical implementation requires prospective validation in independent multicenter cohorts before these tools can be recommended for routine use.

Deep learning models applied to CT images have demonstrated utility for ILD pattern classification, disease extent quantification, and outcome prediction in SSc-ILD [14,31,32,34,35,36,37,38,39,40,41]. The integration of PET metabolic data with CT structural data may potentially enhance AI model performance, but this approach has been explored in only one published study (Schniering 2021 [14]).

7.6. Limitations of AI/Radiomics Evidence in SSc-ILD

The AI/radiomics evidence base in SSc-ILD has important limitations.

• CT-based predominance: The majority of AI studies in Table 6 used HRCT alone. Only Schniering 2021 [14] integrated PET/HRCT radiomics.
• Single-centre studies: Most AI studies are single-centre, limiting generalizability.
• Retrospective design: Most studies are retrospective, introducing selection bias.
• No prospective validation: No AI model has been prospectively validated in an independent multicenter SSc-ILD cohort.

8. Clinical Applications and Future Perspectives

8.1. Current Clinical Applications

Molecular PET/CT in SSc-ILD currently has no established role in routine clinical practice. Its applications remain investigational, including: (1) research tool for disease activity phenotyping; (2) exploratory biomarker for treatment response in clinical trials; and (3) prognostic assessment in selected patients with uncertain disease trajectory. Clinical adoption will require standardized protocols, prospective multicenter validation, and regulatory approval of novel tracers.

8.2. Future Perspectives

Future research priorities in molecular PET/CT for SSc-ILD include: (1) prospective multicenter studies with standardized imaging protocols and quantification methods; (2) head-to-head comparison of FDG vs. FAPI PET/CT in SSc-ILD; (3) longitudinal studies evaluating PET changes with treatment (immunosuppression and antifibrotics); (4) integration of PET biomarkers with serum biomarkers and genetic data; (5) the development of AI tools integrating PET and HRCT data; and (6) the validation of early FAPI volume-based parameters (as suggested by Tai et al. 2026 [27]) in larger prospective cohorts.

9. Limitations of This Review

9.1. Evidence Quality and Study Design Limitations

The evidence base for molecular PET/CT in SSc-ILD has several important limitations.

• Small sample sizes: Most studies enrolled 10–30 patients. Only Lilburn in 2025 [18] exceeded 40 participants (n = 45).
• Heterogeneous study designs: Studies vary in patient selection, disease duration, treatment exposure, PET protocols, and quantification methods.
• Single-centre studies: All published SSc-ILD PET studies are single-centre, limiting generalizability.
• Predominantly retrospective design: Most studies are retrospective, introducing potential selection bias.
• Conference abstract data: Several included studies are conference abstracts, representing preliminary data without full peer review.
• Tracer heterogeneity: Different FAPI tracer variants (FAPI-04, FAPI-46) and imaging timepoints are used across studies.

9.2. Narrative Review Methodology

This review is a narrative review and does not employ a systematic review methodology with formal screening by two independent reviewers, meta-analytic synthesis, or GRADE evidence quality assessment. The study selection reflects expert judgement and may be subject to selection bias. The transparent reporting of search strategy and database results (Section 1) is intended to enhance reproducibility, while acknowledging that this approach does not fulfil the requirements of a systematic review.

9.3. Lack of Formal Quality Appraisal

No formal quality appraisal tool (e.g., Newcastle–Ottawa Scale for observational studies or QUADAS-2 for diagnostic accuracy studies) was applied to the included studies. This is a recognized limitation of narrative reviews. The following types of bias may affect the included studies:

• Selection bias: Patients in single-centre PET studies are typically recruited from tertiary referral centres, enriching the cohort for more severe or complex disease (see Section 9.5).
• Information bias: Retrospective studies rely on medical records, which may have incomplete or inconsistently recorded data.
• Measurement bias: Variability in PET acquisition protocols, scanner hardware, and reconstruction algorithms across studies may introduce systematic measurement error.
• Attrition bias: Longitudinal studies may have differential loss to follow-up related to disease severity.
• Publication bias: Positive or feasibility studies are more likely to be published than negative studies, potentially overestimating the utility of molecular PET/CT in SSc-ILD.

Readers are advised to interpret the findings in the context of these limitations and to await prospective multicenter validation before clinical implementation.

9.4. Accessibility and Generalizability of PET/CT

PET/CT technology requires specialized infrastructure (cyclotron or generator for radiopharmaceutical production, dedicated PET/CT scanner, and radiation safety facilities) and is not universally available. The evidence reviewed here is derived from specialized academic centres in Europe and Asia, and the findings may not be generalizable to healthcare systems with limited PET/CT access.

9.5. Selection Bias and Generalizability

The reviewed studies are subject to several sources of selection bias that limit generalizability:

• Tertiary referral centre bias: All published SSc-ILD PET studies were conducted at tertiary academic medical centres. Patients referred to these centres typically have more severe, complex, or treatment-refractory disease compared to the broader SSc-ILD population managed in community settings.
• Disease severity enrichment: PET studies tend to enrol patients with established ILD and measurable PFT impairment, underrepresenting patients with early or subclinical disease.
• Treatment exposure confounding: Patients enrolled in PET studies are often on immunosuppressive therapy (mycophenolate mofetil, cyclophosphamide, and nintedanib), which may suppress the metabolic and fibrotic activity detectable by PET, potentially underestimating true PET signal in untreated disease.
• Socioeconomic and geographic barriers: PET/CT access is limited by cost, infrastructure, and geographic availability. Studies from high-resource academic centres may not reflect real-world practice in lower-resource settings.
• Implications for clinical translation: The generalizability of current PET findings to the broader SSc-ILD population, including patients with early disease, mild ILD, or those managed in non-academic settings, remains uncertain. Prospective multicenter studies recruiting diverse patient populations are needed.

10. Conclusions

Molecular PET/CT imaging represents a promising but early stage approach for SSc-ILD assessment, and its role in clinical practice currently remains investigational. FDG PET/CT provides a functional correlate of metabolic activity in SSc-ILD, with emerging prognostic data from a single prospective cohort study (Lilburn 2025 [18]); however, this finding requires replication in multicenter cohorts before clinical adoption. FAPI PET/CT offers direct molecular targeting of fibroblast activation, with early evidence suggesting discriminatory value for disease progression (Tai 2026 [27]), but is currently supported by small, early stage studies only. Other molecular tracers—including CXCR4, integrin αvβ3, and macrophage-targeted agents—are at even earlier stages of investigation and should be considered exploratory. The evidence base is currently limited by small sample sizes, single-centre designs, heterogeneous protocols, and the absence of prospective treatment response data. Standardization of imaging protocols, quantification methods, multicenter prospective validation, and integration with clinical biomarkers are essential prerequisites for clinical translation. Molecular PET/CT should currently be regarded as a research tool with significant potential, rather than a validated clinical imaging modality for SSc-ILD management.