Evaluating the Patterns of FAPI Uptake in the Shoulder Joint: a Preliminary Study Comparing with FDG Uptake in Oncological Studies

Persistent shoulder pain, including frozen shoulder, is one of the most frequent complaints across all generations, with a prevalence of 26% and a lifetime estimate of 67% [1]. Among others, chronic inflammation along with increased fibroblast proliferation and extracellular matrix deposition have been advocated to contribute to disease on-set and progression, as recently demonstrated in patients with idiopathic frozen shoulder undergoing surgery [2]. Moreover, also classified as chronic inflammatory diseases, shoulder-associated rheumatoid (RA) and osteoarthritis (OA) are also characterized by increased fibroblast activation protein (FAP) expression in chondrocytes, in particular in the superficial zone of cartilage tissues, thereby rendering FAP as an attractive, pro-inflammatory target [3, 4]. Pathologically, fibrosis by fibroblasts causes extracellular matrix stiffness, which can lead to contracture, ultimately limiting motion of joints [5]. Contractures, however, reduce the patient's quality of life (QOL). If the degree of joint fibrosis can be determined by non-invasive imaging modalities, it may be possible to delay the respective disease progression, preferably for initiating treatment at the right time for the right patient. Computed tomography (CT), magnetic resonance image (MRI), and ultrasonography are clinically used as indices of joint morphologic evaluation. While they can visualize lesions on a millimeter scale [6‐8], it is difficult to assess the activity at the cellular and molecular level within the lesion.

In recent years, fibroblast activation protein inhibitor (FAPI) positron emission tomography/computed tomography (PET/CT) has entered the clinical arena as a novel, non-invasive imaging biomarker for malignant neoplasms [9]. As an underlying rationale on a cellular level, cancer-associated fibroblasts are crucially involved in tumorigenesis and thus, this radiotracer may even serve as a potential substitute for F-18 fluorodeoxyglucose (F-18 FDG) PET/CT in varying oncological scenarios [10, 11]. Of note, in those cancer cohorts, incidental joint uptake on FAPI-directed PET/CT in the shoulders and in patients with rheumatoid arthritis has already been described [12‐15]. Expression of FAP in fibroblasts is required for the transient regenerative phase of normal tissue reconstruction, but is overexpressed in chronic pathological processes, causing fibrosis in the tissue [16]. In IgG₄-related diseases, F-18 FDG, which indicates the degree of glycolysis, can be used as an indicator of inflammation, and fibroblast-targeting radiotracers as a potential biomarker of fibrosis [17]. Thus, we aimed to provide a comprehensive characterization of FAPI upregulation in shoulder joints, thereby determining its role as a non-invasive image biomarker of chronic fibrosis and provide a head-to-head comparison with the inflammation-targeting reference F-18 FDG.

In a large cohort of cancer patients imaged with Ga-68 FAPI-04 PET/CT, we observed visual increased FAPI expression in more than 43% of patients, while distribution was relatively balanced among investigated AC or GH + SA joints (or both). In addition, frequency and in-vivo FAP expression was also markedly elevated when compared to the reference radiotracer F-18 FDG. There are various types of PET and SPECT tracers that visualize inflammation, e.g. targeting C-X-C motif chemokine receptor 4 or somatostatin receptor [26, 27]. FAPI-directed radiotracers, however, are directed towards fibroblasts, which are responsible for joint stiffness and contractures [5] and thus, in-vivo assessment of fibroblast activity could be related not only to pain, but also to restricted joint mobility function. Thus, Ga-68 FAPI-04 PET may provide a novel image biomarker to assess chronic remodeling in shoulder joints in-vivo. The herein presented findings may then enable for rigorous assessments in patients affected with RA or OA, e.g., to initiate FAP-tPDT for synovial depletion based on FAPI PET signal strength. Such an image-guided photodynamic treatment would then allow to adjust light dose for providing improved therapeutic efficacy while identifying peri-arthritic unaffected tissue prior to treatment on-set, ultimately avoiding off-target effects. Further indications include longitudinal response monitoring, e.g., to determine response based on declining FAPI PET signal after FAP-tPDT.

In our study, Ga-68 FAPI-04 uptake in the AC joints was often visible as focal hot spots on MIP images, thereby allowing a reliable visual read-out of affected joints. Although smaller than GH + SA joints, AC joints were also more often FAP + . AC joint arthritis is common and it often accompanied by other shoulder conditions such as rotator cuff disease or glenohumeral arthritis [28]. Of note, AC joint-related pathology is also frequently observed in relatively young patients, including over 30 years of age [29]. Based on our findings, however, shoulder uptake in AC joints is relatively common and thus, the interpreting nuclear medicine expert should be aware of such a potential pitfall in scan interpretation, e.g., when investigating a FAP + (metastatic) lymph node in close proximity to the herein investigated anatomical localizations. In this regard, a careful assessment of MIP images can already provide a useful hint whether respective joints are affected, while transaxial PET/CT reads can then rule out (or corroborate) uptake in AC joints (Fig. 2). In addition, recently introduced standardized reporting systems for theranostic radiotracers may also be useful to segregate between benign and malignant findings [30], once such frameworks become available in the context of FAPI-directed molecular imaging.

Clinical implication of Ga-68 FAPI-04 uptake in shoulder joints have not been fully elucidated yet. Since shoulder joint diseases include frozen shoulder, OA, RA and rotator cuff tear [31], FAPI uptake may be most likely associated with or caused by such pathological conditions. In the present study, the difference of age among the patients with none, unilateral, or bilateral shoulder(s) with increased FAPI expression was not significant. However, the present study included patients with malignancies in particular in elder subjects (88% of patients were aged 50 years or older). We cannot conclude that age does not correlate with quantified Ga-68 FAPI-04 joint uptake, because a relevant portion of patients is affected with shoulder diseases already in their 40 s and 50 s [29]. Thus, further studies including a broader age spectrum along with a rigorous assessment of shoulder pain prior to scanning are needed to link the herein derived uptake pattern and signal strength to age-related (chronic) shoulder pain. Nonetheless, frozen shoulder, RA and OA are all associated with fibroblast-mediated, chronic inflammation [32, 33]. For instance, FAP deficiency in dedicated mouse models led to less cartilage degradation [34]. Of note, a recent ex-vivo study investigating RA synovial explants demonstrated that anti-FAP targeted photodynamic therapy (FAP-tPDT) caused cell death on ex-vivo activated synovial fibroblasts with increasing light dose [35]. This FAP-tPDT may enable for rigorous assessments in patients affected with RA or OA, e.g., to initiate FAP-tPDT for synovial depletion based on FAPI PET signal strength. Such an image-guided photodynamic treatment would then allow to adjust light dose for providing improved therapeutic efficacy while identifying peri-arthritic unaffected tissue prior to treatment on-set, ultimately avoiding off-target effects. Moreover, Wandler and coworkers used dedicated questionnaires and investigated whether F-18 FDG may be a useful tool to segregate between various joint inflammation or injuries (with OA and bursitis being associated with diffuse uptake pattern) [36]. While F-18 FDG PET/CT may emerge as a relevant radiotracer in an active inflammatory setting right after disease on-set or shoulder damage [36], FAPI-targeted PET has been advocated to play a role in chronic destructive processes and thus, future studies may then also conduct such pain-related, patient-centered registries in the context of FAPI-PET/CT [37]. Of note, those considerations are further fueled by the fact that a recent trial investigating uptake on FAPI PET/CT in benign diseases also reported on the spine, followed by the shoulders as the most frequent locations of relevant in-vivo FAP upregulation in patients with arthritis or inflammatory-related abnormalities [13].

This study has several limitations. As one of the most complex joints in the human body including the rotator cuff, bursa, and ligaments, reference CT images may not allow for such a detailed anatomical assessment to reliably segregate PET uptake in all of those localizations. PET combined with MRI may overcome this issue [13]. Since it was a retrospective analysis, information on the patient's symptoms, formal OA/RA diagnosis and treatments affecting shoulder uptake, such as anti-inflammatory drugs, were not available. The data of the body height for body mass index, smoking history, blood pressure, and blood tests were also not available for this study. Future studies may also conduct matched-pair comparisons, e.g., in patients with and without symptoms and more rigorous use of joint-focused X-rays and/or MRI [38, 39]. Long-term evaluation is also needed to clarify the relationship between Ga-68 FAPI-04 uptake signal and disease progression, preferably by including a larger number of patients in a prospective study.