A re-evaluation of 18F-FDG PET/CT for diagnosing relapsing polychondritis and monitoring the therapeutic response to treatment

Background: Relapsing polychondritis (RP) is a rare degenerative disease and its diagnosis is often delayed and difficult. Although the use of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) for diagnosing RP has been reported previously, its true diagnostic value is uncertain due to the limited number of published case reports. Moreover, its value for biopsy guidance and therapeutic response monitoring remain controversial. Materials and Methods: Data on 30 patients with a clinical diagnosis of RP who underwent PET/CT scans were retrospectively reviewed. Follow-up scans were performed in 10 patients, and visual scores (VS) and maximal standard uptake (SUVmax) values were analysed. Results: With the initial scan, 83.3% of patients showed lesions in more than one cartilage. Mean VS and SUVmax values in the cartilages were 2.92 ± 0.38 and 4.06 ± 0.18, respectively. In 18 patients with no history of corticosteroid treatment, more cartilages were affected than in patients who previously received corticosteroids (3.89 ± 0.29 vs 2.75 ± 0.63, respectively; p = 0.07). Positive rates for PET/CT-guided biopsy in nasal, auricular and tracheal/bronchial cartilages were 100%, 88.9% and 10.5%, respectively. In comparison, the positive rate for non-PET/CT-guided biopsy of auricular cartilage was 92.3%. Compared with the initial scan, the second scan had significantly lower mean VS (1.41 ± 0.20 vs 2.92 ± 0.38, respectively; p < 0.0001) and SUVmax values (2.76 ± 0.14 vs 4.06 ± 0.18, respectively; p < 0.001). Conclusions: FDG uptake in multiple cartilages suggests a diagnosis of RP, but repeated administration of corticosteroids may undermine the diagnostic value of 18F-FDG PET/CT. Further investigation of its use in locating biopsy sites and in monitoring the response to treatment is required.


INTRODUCTION
Relapsing polychondritis (RP), first described by Jaksch-Wartenhorst in 1923, [1] is a rare autoimmune inflammatory disease characterised by episodic inflammation of cartilaginous tissue and proteoglycanrich structures throughout the body [2,3]. Currently, the diagnosis of RP is mainly based on characteristic clinical findings and is generally made on the basis of a set of clinical criteria established by McAdam et al. in 1976 [4]. However, the initial symptoms are often atypical, and the prognosis is unfavourable when the respiratory tract is involved. Airway involvement is a major cause of morbidity and mortality in affected patients, [5] and this occurs in up to 50% of RP patients during the course of the illness [4][5][6][7].
In 2007, Nishiyama et al. [8] first described a case of RP diagnosed using fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT). Since then, a series of case reports have demonstrated that PET/ CT is capable of determining the distribution of all lesions and facilitating their early recognition, locating biopsy sites, and evaluating disease activity and the therapeutic response to treatment [9][10][11][12][13][14][15][16]. However, these were mostly single case reports, and there has been no comprehensive study focusing on the application of PET/CT in RP. In 2014, we retrospectively investigated 6 patients with RP, and found that PET/CT was a valuable tool for diagnosing RP and monitoring the response to treatment [17]. Almost at the same time, Yamashita et al. [18] and Lei et al. [19] analysed a series of RP cases who had undergone PET/CT scans (including their own and others from the literature), and also concluded that PET/CT is a reliable procedure for early diagnosis of RP. Other studies have also suggested that PET/CT is a useful radiological tool for the assisting in the selection of biopsy sites in RP [9,16,19]. However, the value of these investigations was limited by relatively heterogeneous populations and small sample sizes. In our previous study, [17] some biopsies of tracheal/bronchial cartilages were negative for RP even though FDG uptake was increased. In this regard, our data contrast with the conclusion of Lei et al. [19] who reported that combining PET/CT with transbronchial needle aspiration (TBNA) was an effective diagnostic approach. There is also controversy as to whether PET/CT is of value for monitoring the therapeutic response to treatment in view of its uncertain accuracy, high cost and radiation exposure.
In this study, we retrospectively investigated the available data on 30 patients with RP, which is largest sample analysed to date, and summarized our experience. We aimed to re-evaluate the value of 18 F-FDG PET/CT for diagnosing RP, guiding biopsies, and monitoring the therapeutic response to treatment.

Patient characteristics
The study population comprised 30 patients (17 males, 13 females) of mean age of 44.8 years (range, 25 to 66 years). Their clinical characteristics are shown in Table 1. The main clinical symptom was cough, which was noted in 24 patients (80%); other symptoms included shortness of breath in 15 patients (50%), wheezing in 6 (20%), excessive sputum in 6 (20%), fever in 4 (13.3%), hoarseness in 3 (10%), chest pain in 3 (10%) and pharyngalgia in 1 (3.3%). The duration of these complaints ranged from 6 months to more than 10 years. Eight patients were misdiagnosed as asthma, 3 as chronic obstructive pulmonary disease (COPD), and 1 as tracheal amyloidosis. Corticosteroids had been administered irregularly to these patients before a definitive diagnosis was made. All the patients received comprehensive treatment with corticosteroids after the definitive diagnosis of RP. Most of them showed improved clinical symptoms, including a complete or partial resolution of cough, easier breathe less sputum and fever, etc. Immune inhibitors (Azathioprine) were plus with in two patients (Nos. 8 and 26) when symptoms were not obviously improved.

F-FDG PET/CT findings in RP (Table 2)
The initial 18 F-FDG PET/CT scan was performed in all patients. Twenty-five patients (83.3%) showed lesions in more than one cartilage, and 20 (66.6%) had lesions in three or more cartilages. The most commonly involved cartilages were tracheal/bronchial cartilage (26 patients; 86.7%), nasal cartilage (18 patients; 60%) and auricular cartilage (15 patients; 50%). The mean VS in the cartilages was 2.92 ± 0.38. The RP lesions demonstrated an increased SUV max in all patients except Patient No. 7, and the mean SUV max value was 4.06 ± 0.18. Figure 1 shows the affected cartilages or organs in the 30 patients. When patients were classified into 2 groups according to their history of corticosteroid treatment, all 18 patients with no history of corticosteroid therapy were found to have lesions in 2 or more cartilages [ Figure 2 shows the PET/CT images for Patient No.12]. In the other 12 patients who had received corticosteroids, fewer cartilages tended to be affected (mean number, 2.75 ± 0.63 vs 3.89±0.29 in those with no history of corticosteroid therapy), although there was no statistical difference between the 2 groups (p = 0.07). Among the latter 12 patients, 4 (Nos. 4, 13, 21 and 29) had only 1 cartilage site with intense FDG uptake, and 1 patient (No. 7) did not have intense FDG uptake in any cartilage.
Follow-up PET/CT examinations were performed in 10 patients. In comparison with the initial scan, the second scan had a significantly lower mean VS (1.41 ± 0.20 vs 2.92 ± 0.38, respectively; p < 0.0001), and the mean SUV max value with the second scan was also significantly lower than that with the initial scan (2.76 ± 0.14 vs 4.06 ± 0.18, respectively; p < 0.001). Most of the intense uptake decreased or disappeared during corticosteroid therapy [ Figure 3 shows the PET/CT images for Patient No.12 after treatment]. FDG uptakes diminished in most of the affected sites, except in the auricular cartilage of Patient No. 24 who showed symptomatic and inflammatory improvement. However, nasal cartilage FDG uptake was obviously increased in Patient No. 20.

Cartilage biopsy in the RP patients
The results of cartilage biopsies, which were performed at different sites in all patients, are shown in Table 3. Nasal cartilage biopsies were all positive for patients with increased FDG uptake detected by PET/ CT-guided nasal cartilage biopsy (5/5, 100%). Fiberoptic bronchoscopy was performed in most of the patients and two-sample paired-proportions test showed significant difference in RP diagnostic performance between 18 F-FDG PET/CT and biopsy in tracheal/bronchial cartilage. Biopsies of the tracheal/bronchial cartilage by bronchoscopy were performed with the aid of PET/ CT in 19 patients, but only 2 had positive results (2/19, 10.53%). Biopsies of the tracheal/bronchial cartilage by bronchoscopy were also applied for 4 RP patients negative for PET/CT scan, and none of them showed positive outcomes (0/4, 0%). PET/CT-guided auricular cartilage biopsy was performed in nine patients and eight of them had positive results (8/9, 88.9%) [ Figure 4 shows the characteristic RP pathology in Patient No.12]. There were 15 patients in which FDG uptakes at auricular cartilage were negative. Auricular cartilage biopsy was conducted in 13 patients among them. Surprisingly, the biopsy positive rate was as high as 92.31% (12/13). Therefore, no correlation was found between PET/CT and biopsy in RP diagnosis with respect to auricular cartilage.

DISCUSSION
RP is a rare inflammatory chondropathy in multiple cartilages that is not readily diagnosed in the absence of typical clinical findings [2,17,[25][26][27][28][29]. The diagnosis of RP diagnosis is based on clinical criteria, such as those described by McAdam et al. [4] and on a biopsy of affected cartilage. According to Damiani and Levine, [21] the diagnosis also can be established when one or more of clinical features are present in conjunction with biopsy confirmation. Our previous study revealed that PET/ CT is a powerful tool for diagnosing RP as it can detect multisystemic cartilaginous abnormalities via increased FDG uptake. [17] In the study of Yamashita et al. [18]  several specific patterns of PET/CT findings in RP were found, such as FDG uptake in bilateral auricular cartilage, and FDG uptake was noted in various combinations of affected lesions in single patients (for example, auricular and nasal, trachea and unilateral bronchial tree, costal cartilage and joints, etc.). These authors recognised that the specific PET/CT findings may be of high diagnostic value. However, the numbers of patients evaluated in these previous studies were so small that a re-evaluation is necessary. To our knowledge, the current study is the largest series of patients in which the value of PET/CT has been investigated in RP. Significantly, PET/CT scans in the 30 patients whose data were evaluated showed that 83.3% (25/30) had disease involvement in more than one cartilage. We also found that all 18 patients without a history of prior corticosteroid treatment had involvement in more than one cartilage. Therefore, FDG uptake in multiple cartilages, such as tracheal/bronchial cartilage, nasal cartilage and auricular cartilage, in a single patient suggests the diagnosis of RP and a biopsy might not be necessary. However, the other 12 patients who received irregular corticosteroid treatment previously had fewer affected cartilages detected by PET/CT, and 4 of them (Nos. 4, 13, 21 and 29) had only one site of intense FDG In the present study, the positive rate for PET/CT-guided biopsy was relatively high in nasal cartilage (100%) and auricular cartilage (88.9%), but the positive rate for non-PET/CT-guided biopsy of auricular cartilage in was 92.3%. Thus, while PET/CT was effective for guiding biopsies in nasal and auricular cartilage, a negative PET/CT scan result will not completely exclude RP involvement, especially when auricular cartilage is affected. We consider that there could be a high risk of false-negative results with PET/CT scans of auricular cartilage. A possible explanation is that FDG uptake is consistent with an adequate blood supply which is usually poor in auricular cartilage. The positive rate for PET/CT-guided biopsies was low in tracheal/bronchial cartilage (10.5%) in the present study, even though FDG uptakes were high. Therefore, a positive result in tracheal/bronchial walls with a PET/ CT scan is not always consistent with the biopsy result for tracheal/bronchial walls sampled by bronchoscopy. A possible reason for this inconsistency is technical difficulty/ challenge in reaching the affected site of tracheal/bronchial cartilage when there is tracheal collapse, fibrosis or stenosis. Bronchoscopy is invasive and consequently may exacerbate mucosal swelling or cartilage inflammation via mechanical stimulation. Serious complications arising from cartilage biopsies of the trachea or bronchus in RP patients have been reported [30]. Thus, the value of PET/CT-guided biopsy may be dependent on the cartilage site affected.
Corticosteroid therapy is the mainstay of treatment for RP, and some researchers have reported that PET/CT is valuable for monitoring the response to treatment. Lei et al. [19] noted that post-treatment PET/CT examinations showed obvious decreases or complete disappearance of high 18 F-FDG-uptake lesions in 10 of 22 cases, which was highly consistent with symptom improvement, and Yamashita et al. [18] reported similar findings. In our research, a follow-up scan performed in 10 of the 30 patients (33%) showed that mean VS and SUV max values were significantly lower with this scan than with the initial scan. These changes were also correlated with clinical symptom improvements and related laboratory data.
These findings suggest that PET/CT is a promising radiological tool for monitoring the treatment response and disease progress. However, because there is no consensus about the duration of corticosteroid treatment, there is no consensus regarding the use of PET/CT to optimise corticosteroid strategies, although some reports have pointed to its potential value [9][10][11][12][13][14][15]. The time interval for PET/CT scans after corticosteroid therapy remains unclear, as it has been reported to range from 1 month to 13 months [9-11, 13, 18]. To our knowledge, there have been no studies that have sought to determine how many and what percentage of patients have had their clinical management altered on the basis of PET/ CT results. Therefore, there is no evidence that PET/ CT can provide prognostic information in RP patients and could be considered a guide for making treatment decisions. In 2 of our patients, in whom follow-up data indicated symptomatic and inflammatory improvement, FDG uptakes diminished in most of the affected sites but increased in 2 other sites (nasal and auricular cartilage). Thus, in some patients, PET/CT results may differ depending ono their symptoms. In some cartilage sites, FDG accumulation after corticosteroid treatment of patients with RP may be correlated with other factors (except inflammatory activity). In the present study, only 33.3% patients (10/30) underwent follow-up PET/ CT scans, mainly because of their high cost, and further studies involving larger numbers of patients are necessary to evaluate the role of follow-up PET/CT scans during treatment. In doing so, the cost-effectiveness of repeated PET/CT scans and the resultant radiation exposure must also be taken into account. PET/CT-guided biopsy was performed in affected cartilage when PET/CT showed increased FDG uptake. PET/CT, positron emission tomography/computed tomography; NaN, not a number; RP, relapsing polychondritis.
Our study has some limitations. Firstly, it was a retrospective study and therefore, the results may not completely translate to clinical practice. Secondly, because all patients included in the study presented with respiratory symptoms, it is possible that we underestimated the value of PET/CT for RP patients without respiratory involvement. Thirdly, although this was the largest dataset studied thus far, the number of patients evaluated was still small reflecting the low incidence of this disease and welldefined patient populations. Thus, a larger, multicentre, prospective, randomised study is needed for further validation of our results.
In conclusion, FDG uptake in multiple involved cartilages suggests the diagnosis of RP, but repeated application of corticosteroids may undermine the diagnostic value of 18 F-FDG PET/CT. Further investigation of 18 F-FDG PET/CT is needed confirm its value in guiding biopsies and in monitoring the response to treatment.

Study design and setting
The clinical characteristics of 30 patients with a diagnosis of RP who were treated at the First Affiliated Hospital of Guangzhou Medical University, Guangzhou between January 2010 and January 2016, were reviewed retrospectively. The patients were admitted to hospital  due to respiratory symptoms and underwent PET/CT examinations in the Department of Radiology. Highresolution computed tomography (HRCT) revealed diffuse thickening and narrowing of the trachea and/or major bronchi, tracheal calcification, or enlargement of the mediastinal lymph node in all patients except one (Patient No. 4). Patient Nos. 1 and 3 have previously been described in separate case reports, [12,20] and data on 6 patients (Nos. 1-6) have been retrospectively reviewed in a previous study [17]. PET/CT was performed as a wholebody imaging technique to determine the underlying cause of the disease or to exclude malignant conditions when the diagnosis was difficult to establish on the basis of the available clinical evidence.
All patients were finally diagnosed with RP according to clinical criteria. [4,21] None had serious underlying diseases such as diabetes or malignant tumours. All patients were treated with corticosteroids following establishment of the diagnosis, initially with prednisolone doses 1 mg/kg/day with subsequent reductions of 5-10 mg per month.
Follow-up 18 F-FDG PET/CT scans were performed in 10 patients to assess the therapeutic response and optimise further therapeutic strategies. Six patients had their follow-up PET/CT scans 3 months after beginning corticosteroid therapy, and the other 4 patients had their follow-up PET/CT scans after 2.5 or 4 months. In patient No. 3, follow-up scans were performed at 3, 9 and 15 months during prednisolone treatment. No follow-up scans were performed in the other 20 patients.
Written informed consent was obtained from all patients whose data were analysed and for publication of any accompanying images.

F-FDG PET/CT scans
As described in a previous report, [17] patient preparation for 18 F-FDG PET/CT scans included the following steps: (1) avoidance of strenuous work or exercise in the preceding 24 hours, fasting for more than 6 hours before the scan, and measurement of fasting blood glucose concentrations prior to the scan (which were required to be under 130 mg/dL [7.2 mmol/L] before 18 F-FDG injection); (2) an intravenous injection of 18 F-FDG in a dosage of 3.70-5.55 MBq/kg together with 10 mg furosemide to accelerate renal 18 F-FDG elimination; (3) a CT scan from the brain to the pelvis using a multidetector spiral CT scanner (3.75 mm slice thickness; pitch 0.875; rotation speed 140 keV; 120 mAs) prior to the PET scan; and (4) acquisition of whole-body PET images using a PET/CT system (GE Discovery ST; GE Medical Systems Inc., WI, USA) 60 min after tracer administration (acquisition time, 2.5 min per bed position for 6 to 7 bed positions).

Data analysis
The PET/CT images were reconstructed using the ordered subset expectation maximisation method (OSEM), with and without attenuation correction. The visual analysis of PET/CT characteristics and pattern was interpreted by at least 2 experienced nuclear medicine physicians who had the patients' clinical information available on a dedicated workstation. When disputes regarding visual interpretation occurred, consensus was reached by discussion. The 18 F-FDG PET scans were analysed visually and semi-quantitatively by calculating the SUV max . The intensity of 18 F-FDG uptake by cartilages relative to the background and incompatible with normal anatomy/physiology was assessed, and the intensity was scored using a 4-point visual score (VS) scale [22,23]. A score of 0 on this scale = absence, not visible on the image display; 1 = faint or less intense than mediastinal blood pool activity; 2 = moderate or equal in intensity to mediastinal blood pool activity; and 3 = more intense than mediastinal blood pool activity. A region of interest (ROI) was placed over the entire area of any abnormal uptake (scale 2 or 3) site. The SUV max was calculated (maximum ROI activity/injected dose per kg body weight), and the highest SUV max was used to evaluate the multiple abnormal uptake sites observed in a given lesion or tissue.
As the blood glucose level may affect SUV max results, the SUV max values needed to be corrected for blood glucose (SUV glu ) as follows: SUV glu = SUV max × blood glucose level/130 when the blood glucose level was >130 mg/dL (7.2 mmol/L) [24]. A positive PET/CT result was judged when the VS was ≥ 1 or SUV max was ≥ 2.0.

Statistical analysis
Continuous variables were presented as the median and range, and qualitative variables as numbers and percentages. Intergroup differences were analysed statistically using SPSS ® 17.0 (SPSS Inc, Chicago, IL, USA). Differences in continuous variables between groups were compared using Student's t-test. McNemar's test was used to determine the difference between 18 F-FDG PET/ CT and cartilage biopsy in RP diagnosis. Significance for statistical analyses was set at p < 0.05.

Author contributions
JW, ML, JH and YZ were involved in the study concept and design. JW, SC and SL were involved in data acquisition, and ML, SC and SL were involved in the statistical analysis and interpretation of the data. JW, JH and YZ drafted the manuscript. All authors critically reviewed/revised the manuscript and approved the final version. JH and YZ takes responsibility for the integrity of the data and the accuracy of the data analysis.