[18F] FDG uptake in patients with spondyloarthritis: correlation with serum inflammatory biomarker levels – EJNMMI Research

SpA is a group characterized by chronic inflammatory arthritis, particularly sacroiliitis. Chronic inflammation causes progression of structural damages that are largely irreversible. These irreversible damages cause arthralgia, stiffness, and limited flexibility in SpA patients. These symptoms relate to the severity of SpA and can interfere with a patient’s quality of life (QOL). The aims of treatment are to alleviate symptoms, maintain and improve QOL, and decrease disease complications. It is difficult to achieve complete remission, and symptoms of relapse are often seen. Assessment of disease activity is essential for therapy because treatment may have to be switched or strengthened if disease activity is maintained at a high level or increases.

Assessment tools, including BASDAI or ASDAS, for determining disease activity in patients with SpA, particularly AS are based on the visual analogue scale and CRP or ESR value. Additional evaluation indices assess enthesitis, joint structure and function, and patient dysfunction in its entirety. Notably, these indices are only applicable to AS, and no suitable index is available for nr-axSpA, another type of axSpA. This has led to the utilization of the above mentioned indices for nr-axSpA, which has resulted in limited appropriate assessment of disease activity in SpA.

Although these assessment tools mainly comprise subjective patient-reported analogue scales, imaging modalities can objectively assess lesions. MRI of the spine and sacroiliac joints (SIJ) is used in clinical assessment, and various scoring systems using MRI for spine and SIJ have been developed. Zhang et al. demonstrated that MR DAS of SIJ, the Spondyloarthritis Research Consortium of Canada (SPARCC) MRI index, and apparent diffusion coefficient values in SIJ correlated significantly with the BASDAI score in AS patients [23]. However, another study reported that in axSpA patients there were no significant correlations between the SPARCC score and BASDAI and ASDAS, and no significant difference in the SPARCC score with high and low clinical DAS [24]. Furthermore, MR DAS is currently applied for axial skeletal lesions, but not for peripheral lesions. In peripheral lesions, US has been used for detecting enthesitis or synovitis in SpA, but to the best of our knowledge, there has been no study that has shown a relation between US and clinical DAS.

FDG PET can provide us whole body examination, and this is an advantage over MRI and US. Particularly, in pSpA patients with not only peripheral lesions but also axial lesions, it is advantageous that FDG PET can detect both axial and peripheral lesions in a single examination because the presence of axial lesions may affect treatment options. Furthermore, because FDG uptake has been observed in inflammatory tissues, FDG PET may be useful for early detection, similar to MRI. In fact, FDG PET has been used for the detection and evaluation of enthesitis or to distinguish it from similarly presenting rheumatic diseases, such as RA and polymyalgia rheumatica in SpA patients [7, 9, 11, 25]. Takata et al. demonstrated that SUVmax in lesions correlated with PsA DAS, including the Psoriasis Area and Severity Index and the DAS28 [26]. Abdelhafez et al. also reported that in PsA patients, FDG PET metrics showed a significant correlation with DAPSA [13]. These findings highlight the possible utility of FDG PET for the assessment of disease activity in PsA patients. In view of the fact that no reports are available on FDG PET in the context of disease activity in patients with other types of SpA, we evaluated the correlation between FDG uptake and disease activity assessed by serum inflammatory biomarker levels in SpA patients. Our findings demonstrated that FDG uptake in patients with untreated pSpA correlated with serum inflammatory biomarker levels.

While we conclusively demonstrated significant correlation between PET indices and serum inflammatory biomarker levels in untreated SpA patients, similar observations were not evident in treated SpA patients. Kaijasilta et al. showed that while sulfasalazine, a DMARD, reduces FDG PET-detectable inflammation in sacroiliac joints in axSpA, a concomitant decrease in inflammatory serum biomarker levels, such as CRP and ESR, did not occur [27]. This may add to the difficulty of establishing a relationship between FDG uptake and serum inflammatory biomarker levels during treatment. Moreover, in the present study, patients in the treated group were included based on their treatment contents at the time of PET/CT scanning only, and treatment periods and clinical effects were not considered. Therefore, the treated group included patients with different disease activities. Additional PET/CT scanning to assess treatment effect was not performed in the present study. Further studies are needed to validate whether FDG PET/CT can help assess treatment effect in SpA patients.

Significant correlation between CRP and all PET indices was evident in untreated pSpA but not in untreated axSpA. Similarly, MMP-3 expression correlated significantly with the number of positive regions and the total score in untreated pSpA, but not in untreated axSpA. Conventionally, SpA has been associated with entheses and RA with the synovium; however, recent reports suggest that both these conditions involve the synovium [28, 29]. Appel et al. also reported that synovitis is a hallmark of peripheral arthritis, and inflammation within the axial skeleton, chest wall, and hip is mostly associated with inflammation in the entheses or subchondral bone marrow [29]. Synoviocytes are potent sources of the cytokine interleukin 6 (IL-6), which is commonly induced by inflammation and known to stimulate production of acute-phase proteins, including CRP, serum amyloid A, and fibrinogen [30, 31]. Consequently, the predominant involvement of peripheral articular tissue in pSpA may result in high IL-6 secretion, thus reflecting synovitis. Further, predominant lesions of axSpA have little synovium. This may have affected the results that there were significant correlations between PET indices and CRP in untreated pSpA but not in untreated axSpA. The elevation of IL-6 secretion may be much higher in pSpA than that in axSpA, which may subsequently make the correlations between CRP and all PET indices more significant in untreated pSpA patients. Synovial inflammatory cytokines also induce expression of MMP-3, a proteinase, whose levels have been noted to correlate with disease activity in RA patients [32]. The correlation between MMP-3 and the number of positive regions and the total score in untreated pSpA may be reflective of synovitis in peripheral articular.

ESR, an inflammatory biomarker, often finds usage in the evaluation of disease activity scoring systems for RA, along with CRP. Beckers et al. reported correlation of ESR and CRP levels with FDG PET findings, including the number of FDG positive joints and the sum of SUVs in FDG positive joints in RA patients [33]. We did not observe any significant correlation between ESR and PET indices in untreated pSpA and axSpA patients. ESR levels are determined by various factors, and analysis is further complicated by the fact that pSpA patients often suffer from other diseases, including IBD, chlamydia infection, and psoriasis. Further studies on ESR are thus warranted, especially since we were unable to obtain sufficient ESR data.

Upon classification of SpA, all patients were classified as axSpA or pSpA based on ASAS classification criteria in the present study. However, the axial skeletal lesion preferentially occurring in axSpA may be involved in pSpA as well, and conversely, the peripheral lesion preferentially occurring in pSpA may also be involved in axSpA [2, 34]. The axial skeletal lesion in pSpA is often seen in PsA and arthritis related to IBD [35]. In the present study, there were also some patients with both axial and peripheral manifestations. Accordingly, if peripheral lesions included synovitis and our results were affected by synovitis, consideration of such a mixed type could alter the results.

This study had some further limitations, such as its retrospective design and its small sample size. These limitations may led to statistically non-significant correlations between any serum inflammatory biomarkers and PET indices in treated SpA and untreated axSpA, and between ESR and PET indices. While CRP data were available from all patients, ESR values were obtained from a limited number of patients. Additionally, visual PET indices may have underestimated axial skeletal lesions and disease activity in patients with axSpA, on account of the spine being considered one region. Finally, there was no follow-up on treatment effects or changes in disease activity after the PET/CT scan. In RA patients, FDG PET can predict treatment effects potentially earlier than serum biomarkers [19, 36]. We could not investigate treatment effects on SpA in FDG PET; however, its utility is expected. To overcome these limitations and validate our results, additional large-scale prospective studies are essential.