Fever of unknown origin (FUO) and inflammation of unknown origin (IUO) have always been challenging in clinical practice [1,2,3]. Connective tissue diseases (CTDs) account for a large proportion of FUO/IUO patients, and their occurrences tend to increase in years [4, 5]. Patients with CTDs who initially present as FUO/IUO often lack characteristic clinical manifestations. Malignancies or infections often need to be excluded during diagnostic process. Even in those patients with a confirmed history of CTDs, it is also necessary to determine whether the primary disease is activated or other conditions are concomitant when fever reemerged, since the long-term use of steroids and immunosuppressive agents increase the risk of concurrent tumor and infection [6,7,8,9,10,11,12,13,14,15,16]. The role of Fluorine-18 labeled fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) in diagnosis of FUO has been confirmed [4, 17,18,19]. However, its use in differential diagnosis of CTDs in FUO/IUO patients is less reported. To further explore the significance of FDG PET/CT in identifying CTDs in FUO/IUO patients, we conducted a retrospective analysis of the clinical data of a group of FUO/IUO patients who underwent FDG PET/CT examination.

Patients and methods

Study subjects

From January 2013 to December 2019, data of 503 consecutive patients with FUO/IUO who underwent FDG PET/CT examination in Peking University People’s Hospital were reviewed. Among them, 242 patients (83 males and 159 females, age 52.8 ± 19.4 years) who were eventually diagnosed as CTDs were included in the study. All patients met the FUO standard [20] (205 cases) or IUO standard [21] (37 cases). Final clinical diagnosis was established by rheumatology specialists based on the clinical manifestation, laboratory, imaging, histopathological examinations, response to treatment, as well as ≥ 6 months’ clinical follow-up. The relevant data of patients were obtained through the digital information system of our hospital. Laboratory examination data within 1 week before and after the PET/CT examination could be obtained for all patients.

PET/CT imaging

Patients were fasted at least 6 h before examination, and blood glucose was controlled below 11.2 mmol/L. 18F-FDG (provided by Atom high-tech Co., Ltd., Beijing, China) was injected intravenously at 5.55 MBq/kg (0.15 mCi/kg). Fifty minutes later, image acquisition was performed using Discovery VCT PET/CT Scanner (GE Healthcare, Milwaukee, Wisconsin, USA). During acquisition, the patients lied in a supine position in the examination bed with both upper limbs placed at the sides of the body. The acquisition field included at least ranging from the base of the skull to the middle of tibia, or extended to the toe, if necessary. Images were reconstructed as three-dimensional PET, CT, and fusion images with a slice thickness of 3.3 mm. More than three experienced nuclear medicine physicians read the images together by visual judgment, observing abnormal uptake of FDG and structural changes in the collection field. Location, number, and systemic distribution of the lesions were recorded, and the nature of the lesions was determined regarding clinical manifestations and laboratory tests. In addition, one fixed nuclear medicine physician measured and recorded the lesion SUVmax by using three-dimensional region of interest (3D ROI) technique.

Data analysis

In this group of CTD patients with FUO/IUO, data was analyzed including the distribution of diseases, main clinical manifestations, and their imaging characteristics on FDG PET/CT. In addition, the role of FDG PET/CT in the differential diagnosis of CTDs was evaluated through the course of clinical managements as well as a clinical questionnaire survey, which was obtained from the attending physicians at the end of follow-up. In the questionnaire survey, clinical significance of FDG PET/CT was graded into four levels: G0, FDG PET/CT findings did not help or had an adverse effect on the final diagnosis; G1, FDG PET/CT findings played an important part in ruling out malignancy, focal infection, or typical rheumatologic diseases; G2, FDG PET/CT findings provided additional diagnostic clues (e.g., detecting lesions and suggesting appropriate sites for pathology examinations); and G3, FDG PET/CT revealed the etiological diagnosis directly. It was considered that G0 did not contribute to the differential diagnosis of CTD, while G1–G3 did. The impact of FDG PET/CT on diagnosis of CTD patients was analyzed and compared with that of infection and malignancy patients.

Statistical analysis

Statistical analysis in this study was carried out by SPSS 16.0 software (SPSS Inc. Released 2007). Among the different groups, Student t test was used to analyze the difference of mean SUVmax. The chi-square test was used to compare the composition ratio between groups. Difference was considered statistically significant when P < 0.05.

Results

The 242 patients with CTDs accounted for 48.1% of FUO/IUO patients who received FDG PET/CT examination. Clinical diagnosis of these patients was listed in Table 1. Adult-onset Still’s disease was the most common disease. Other diseases included systemic vasculitis, undifferentiated connective tissue disease, rheumatoid arthritis, idiopathic inflammatory myopathy, systemic lupus erythematosus, and polymyalgia rheumatica.

Table 1 Final clinical diagnosis for the 242 FUO/IUO patients
Full size table

Main clinical manifestations of CTD patients who featured as FUO/IUO, other than fever, were arthralgia, rash, and muscle pain or weakness. Laboratory results generally showed elevated serum C-reactive protein (CRP) and/or erythrocyte sedimentation rate (ESR). However, clinical manifestations of different CTDs showed various patterns (Table 2).

Table 2 Main clinical manifestations of different connective tissue diseases
Full size table

On FDG PET/CT examination, 237/242 (97.9%) of the patients had positive findings. In these CTD patients, tissues and organs with active inflammation showed high uptake of FDG, while some of them did not show significant structural changes. Lesion mean SUVmax of CTDs was significantly lower than malignancies, but no significant difference was found between that of CTDs and infections or miscellaneous diseases (Table 3).

Table 3 FDG uptake of different disease types in FUO/IUO patients
Full size table

Inflammatory lesions were detected in 161 (66.5%) patients, and their distribution was consistent with corresponding susceptible organs and tissues in various diseases (Table 4 and Fig. 1). Non-specific abnormal uptake was found in other 76 (31.4%) patients, manifested as increased diffuse FDG uptake in spleen and bone marrow, accompanied by multiple, symmetrical distributed reactive proliferative lymph nodes. Non-specific abnormal uptake was also seen in most of the patients with detected inflammatory lesions. Negative PET/CT was found in 5 cases, including UCTD (2), systemic vasculitis (1), AOSD (1), and RS3PE (1).

Table 4 Inflammatory lesions detected by FDG PET/CT in patients with connective tissue diseases
Full size table
Fig. 1
figure 1

FDG PET/CT images in patients diagnosed with different CTDs. A. adult-onset Still’s disease, B. giant cell arteritis, C. polyarteritis nodosa, D. idiopathic inflammatory myopathy, E. rheumatoid arthritis, and F. polymyalgia rheumatica. Lesion distribution was consistent with the corresponding susceptible organs and tissues for each disease

Full size image

Before PET/CT examination, 178 out of 242 (73.6%) patients received experimental treatments including antibiotics (126), glucocorticoid (90), and/or non-steroidal anti-inflammatory drugs (42). Histopathologic biopsies were performed in 102 (42.1%) patients, located in bone marrow 71, skin and soft tissues 25, lymph nodes 14, and other organs 8. None of the biopsy results done before PET/CT had diagnostic significance. However, after PET/CT examination, 40 patients underwent 46 times histopathologic examinations (skin and soft tissues 13, lymph nodes 11, bone marrow 7, and other organs 15), and 10 of them (21.7%) directed to determined diagnosis. In addition, 210 (86.8%) patients changed their therapeutic regimens after the PET/CT scan.

Clinical questionnaire about the role of FDG PET/CT in differential diagnosis of CTDs showed that all but two patients benefited from PET/CT examination in clinical process (Table 5). In 110 (45.5%) patients, FDG PET/CT excluded malignancies, focal infections, or other typical CTDs, and thus helped to establish a certain diagnosis. This was commonly seen in AOSD, UCTD, SLE, and Sjögren’s syndrome. In 50 (20.6%) patients, PET/CT findings suggested important diagnostic clues or appropriate biopsy sites that contribute to diagnosis. In 80 (33.1%) patients, PET/CT findings could directly prompt a specific CTD, which was commonly seen in systemic vasculitis, RA, IIM, panniculitis, relapsing polychondritis, etc.

Table 5 Role of FDG PET/CT in differential diagnosis of different types of connective tissue diseases
Full size table

In this study, the contribution of FDG PET/CT for all of 503 FUO/IUO patients was listed in Table 6. The contribution of PET/CT with a role of G1-G3 for CTDs was higher than that for infections (χ2 = 7.507, P = 0.006) and similar to that for malignancies (χ2 = 3.439, P = 0.064).

Table 6 Role of FDG PET/CT in differential diagnosis of FUO/IUO
Full size table

Discussion

In recent years, incidence of CTDs in population is rising. Differential diagnosis and treatment are related to prognosis of patients [22,23,24]. Studies have shown that FDG PET/CT has advantages over conventional imaging techniques in detecting inflammatory lesions [18, 19, 25,26,27,28,29,30,31], but its significance in diagnosis of various CTDs has rarely been reported. Meanwhile, FUO/IUO is a clinically challenging diagnostic problem. Etiological diseases can be classified into infections, CTDs, malignancies, and other miscellaneous diseases. Current studies demonstrated that CTDs were main etiological reasons for FUO/IUO. Revealing the role of FDG PET/CT in CTD patients manifest as FUO/IUO is not only helpful for differential diagnosis but also of importance in treatment monitoring and prognosis assessment.

The study of FDG PET/CT imaging features in CTD patients manifested as FUO/IUO was valuable for differential diagnosis. In the current study, patients with CTDs underwent PET/CT examination mostly because lack of diagnostic clues or need to exclude malignancy, infection, and other CTDs. Usually the incidence and disease type distribution in FUO/IUO patients are related to age, sex, region, environment, and other factors [4, 17]. In our study, CTDs accounted for 48.1% of FUO patients, which is higher than previously reported [5], and more rare diseases are included. Considering this group of data were obtained from large general hospitals and had a large sample size, we believed that our results provide credible reference for those CTD patients with fever as the main clinical manifestation. In addition, as the diagnostic value of FDG PET/CT in CTDs being gradually recognized by clinicians, the number of FDG PET/CT examinations used for the differential diagnosis for FUO/IUO patient has been increasing year by year. This may also be a reason for the higher proportion of CTDs.

For now, FDG PET/CT is mainly used in diagnosis of malignant tumors. Its diagnostic value for various CTDs remains to be understood. Our study showed that in CTD patients who initially presented as FUO/IUO and underwent PET/CT examination, AOSD was the most common disease, accounting for 31.82% of the patients, followed by systemic vasculitis, UCTD, RA, IIM, SLE, PMR, etc. In FDG PET/CT images, the inflammatory lesions could be observed in the forms of high uptake of FDG, before significant structural changes could be seen. The overall distribution and pattern of the lesions usually corresponded with the clinical characteristics of respective diseases, which were also obviously different from the image manifestations of malignant tumors or infectious diseases [25, 27, 28, 30]. The SUV might contribute to differentiate CTDs from malignancies. However, within inflammatory lesions, such as infectious diseases, miscellaneous diseases, and various kinds of CTDs, the SUV of lesions overlaps greatly. Although the diagnostic value of FDG PET/CT in specific CTDs remains to be confirmed, it can be seen from this study that different image patterns could be found in PET/CT for different CTDs. In some patients, active inflammatory lesions detected by FDG PET/CT were asymptomatic and hard to detect by routine imaging; thus FDG PET/CT provided more evidence for diagnosis and clinical classification of CTDs. With the help of FDG PET/CT combined with clinical data, many difficult cases could be diagnosed correctly. Previous studies have also shown that FDG PET/CT had multi-level significance in diagnosis of inflammatory diseases, including detecting or excluding malignant tumors, revealing lesions that indicated diagnostic clues and appropriate biopsy sites, and providing evidence for experimental treatment [4, 17, 32,33,34]. And this multi-level diagnostic significance highlights its advantages in the diagnosis of CTDs.

In the study, the value of FDG PET/CT in differential diagnosis of CTDs was confirmed by the course of management and clinical questionnaire survey. Although histopathologic biopsies yielded relatively low diagnostic capacity in CTDs [4], PET/CT suggested more non-routine biopsy sites and improved the diagnostic significance of biopsies. Meanwhile, it played an important role on tailoring the treatment strategies. By the clinical questionnaire survey, FDG PET/CT detected inflammatory lesions that indicated important diagnostic clues or appropriate biopsy sites in 20.6% of patients and directly suggested a specific diagnose in 33.1% of patients. Even non-specific abnormal uptake on FDG PET/CT which was once considered to be of no diagnostic significance [33,34,35,36,37] actually played a diagnostic role in excluding other diseases. In our study, 45.5% of patients benefited this way, therefore receiving proper diagnosis and treatment. The contribution of FDG PET/CT in clinical diagnosis of CTD patients was significantly better than infection patients and was not inferior to malignancy patients. This implied that FDG PET/CT could be as helpful for rheumatologists as oncologists. In fact, as physicians started to realize the important role of FDG PET/CT in clinical practice, its use in diagnosis of CTDs had been increasing and is no longer limited to the differential diagnosis of FUO/IUO.

In our study, it was demonstrated that FDG PET/CT could reveal the inflammatory lesions in organs and tissues that reflect the clinical characteristics of different CTDs, thus providing an objective evidence for differential diagnosis, classification, and treatment decision in CTDs. We believe that the combination of clinical and imaging features may provide diagnostic criteria with higher performance. However, this work is difficult to fully explain in this study because of the inclusion of multiple types of CTDs, and this is also the inadequacy of this research. Future studies on different types of CTD need to be carried out.