Parameters related to a positive test result for FDG PET(/CT) for large vessel vasculitis: a multicenter retrospective study

Clin Rheumatol (2012) 31:861–871 DOI 10.1007/s10067-012-1945-0

ORIGINAL ARTICLE

Parameters related to a positive test result for FDG PET(/CT) for large vessel vasculitis: a multicenter retrospective study G. A. Hooisma & H. Balink & P. M. Houtman & R. H. J. A. Slart & K. D. F. Lensen

Received: 28 February 2011 / Revised: 4 December 2011 / Accepted: 19 January 2012 / Published online: 10 February 2012 # Clinical Rheumatology 2012

Abstract The purpose of this study was to identify clinical and laboratory parameters that may improve the effectiveness of the use of fluorodeoxyglucose positron emission tomography (18 F-FDG PET)(/CT) for diagnosing large vessel vasculitis (LVV), and secondarily to assess the contribution of 18 F-FDG PET/CT in finding other diagnoses for patients without signs of LVV on the scan. A multicenter retrospective study of 18 F-FDG PET(/CT) scans performed between January 2000 and December 2009 for clinical suspicion of LVV was conducted. A total of 304 18 F-FDG PET(/CT) scans were included, of which 62 (20%) were positive and 242 (80%) were negative for LVV. Univariate analysis showed that patients with a positive scan were older (65.9±13.4 versus 58.6±16.5 years, p00.002), were more frequently female (76% versus 55%, p00.002), more often had a history of temporal arteritis (10% versus 3%, p0 0.044), less frequently had artralgia (31% versus 67%, p0 0.000), and had higher thrombocyte counts (434±161 versus 373±168×109/l, p00.049) and a higher erythrocyte G. A. Hooisma : H. Balink (*) Department of Nuclear Medicine, Medical Center Leeuwarden, Leeuwarden, The Netherlands e-mail: [email protected] K. D. F. Lensen Department of Internal Medicine, Institute for Cardiovascular Research, VU University Medical Center, Amsterdam, The Netherlands P. M. Houtman Department of Rheumatology, Medical Center Leeuwarden, Leeuwarden, The Netherlands R. H. J. A. Slart Department of Nuclear Medicine, University Medical Center Groningen, Groningen, The Netherlands

sedimentation rate (ESR) (72.6 ± 31.0 versus 51.4 ± 30.5 mm/h, p00.001) than patients with a negative scan. In the multivariate analysis, only artralgia (OR 0.091; 95% CI 0.023–0.366) and ESR (OR 1.024; 95% CI 1.002–1.046) remained statistically significant predictors. The presence of artralgia is a statistically significant negative predictor and an elevated ESR a statistically significant positive predictor of LVV showing up on 18 F-FDG PET(/CT). A reliable prediction of the outcome of the scan, based on these two parameters, is not possible however. 18 F-FDG PET(/CT) allows early diagnosis of LVV and may discover occult inflammatory or neoplastic disorders. Keywords 18F-FDG PET . Artralgia . CT . Erythrocyte sedimentation rate . Large vessel vasculitis

Introduction During the Chapel Hill Consensus Conference in 1994, two forms of primary large vessel vasculitis (LVV) were distinguished: giant cell arteritis (GCA) and Takayasu arteritis (TA). [1] GCA is the most common form and has an incidence of 20 per 100,000 in the population aged over 50 years. Women are affected twice as often as men [2–5]. Characteristically, it affects the temporal artery, resulting in temporal arteritis as a synonym for GCA. Temporal arteritis does not cover the whole clinical spectrum however, since the entire aorta and all its branches can be affected [6]. The clinical presentation is variable and includes typical signs of temporal arteritis (e.g., headache, jaw claudication, and scalp tenderness) but also several nonspecific symptoms, such as fever, malaise, fatigue, and myalgia [2–4, 7]. GCA is also associated with polymyalgia rheumatica (PMR) in about 40% of the cases [8]. The diagnostic

862

reference method for GCA is a temporal artery biopsy, but the test results can be false negative. A falsenegative rate of 15–70% is reported; therefore, the incidence of GCA might be underestimated [9–14]. TA is the second form of primary LVV and primarily affects the aorta and its main branches as well as the coronary and pulmonary arteries. It has an incidence of only 2 per 1,000,000. The mean age of onset is 35 years and the prevalence in women is 2 to 25 times higher than in men, depending on the series considered [2, 4]. The clinical picture is somewhat similar to that of giant cell arteritis [15], but many of the typical signs of temporal arteritis are not commonly found in TA [16, 17]. TA can have a quite devastating course, reflected in a mortality rate as high as 35% at 5 years, similar to that seen in malignancies [18]. Several findings however, suggest that GCA and TA may be two processes within the spectrum of the same single disease [15]. Despite that in the course of the disease stenoses, occlusions, and aneurysms may occur, it is reported that early diagnosis and monitoring of TA are both hampered as serological and inflammation parameters are unreliable [19, 20]. Both fluorodeoxyglucose positron emission tomography (18 F-FDG PET) and pathology studies revealed that the inflammation of TA affects mainly the aorta and its branches, but in a more focal and localized, inhomogeneous pattern than seen in GCA and with more intense FDG uptake than GCA. TA may also present as isolated involvement of either renal arteries, pulmonary arteries, vertebral arteries, or coronary arteries [21–25]. Imaging studies play an important role in diagnosing and monitoring LVV. Angiography, ultrasonography, CT, and MRI used to be the most commonly employed imaging techniques. The differential diagnosis with accompanied atherosclerotic development can be very difficult however. Even if vasculitis can be diagnosed by angiography, differentiation between active vasculitic lesions and plaque formation is difficult. In more recent years 18 F-FDG PET has become an increasingly important diagnostic tool for diagnosing LVV [26–31]. Because the morphological changes in the vessel wall are preceded by inflammatory activity, LVV can be diagnosed in an earlier stage than with the conventional imaging techniques [27]. 18 F-FDG PET also identifies more regions of the aorta and its branches involved in the inflammatory process than MRI [26, 28]. Hybrid 18 F-FDG PET/ CT seems to be superior to 18 F-FDG PET alone in this; by fusing the functional images of the 18 F-FDG PET with the anatomical images of the CT, it is easier to localize regional 18 F-FDG uptake [32, 33] (Figs. 1, 2, and 3). Hybrid 18 F-FDG PET/CT is able to identify focal vascular inflammation and focal vascular calcification as different phases of atherosclerosis in order to exclude this as the cause of the increased 18 F-FDG uptake [34, 35]. As a result, the number of false-

Clin Rheumatol (2012) 31:861–871

positive test results will be reduced, thereby increasing the specificity of the technique. Because of the aforementioned considerations, 18 F-FDG PET(/CT) is increasingly used in patients with elevated inflammatory parameters and clinical characteristics compatible with LVV. In a certain amount of these cases, 18 FFDG PET(/CT) will directly point to LVV, thereby avoiding several time-consuming and sometimes unpleasant tests. On the other hand, due to the low incidence, a high proportion of negative test results of 18 F-FDG PET(/CT) for LVV can be expected. This is undesirable from a cost-effectiveness point of view, although a negative test result for LVV can still be of use, if it provides clues to other diagnoses (e.g., lymphoma, infection). The aim of this study was to possibly identify clinical and laboratory parameters that may improve the efficient use of 18 F-FDG PET(/CT) scans in patients who are suspected of LVV, and secondarily to assess the contribution of 18 F-FDG PET/CT in finding other diagnoses for the patients without signs of LVV.

Materials and methods Patients This multicenter retrospective study was performed at the Medical Center Leeuwarden, University Medical Center Groningen, VU University Medical Center, and the Isala Klinieken with its affiliated hospitals, in The Netherlands. Patients who had undergone a 18 F-FDG PET(/CT) scan between January 2000 and December 2010 with LVV (GCA or TA) as clinical question were included. Patients who had undergone a 18 F-FDG PET(/CT) scan for, e.g., oncological reasons, but were diagnosed with LVV as an incidental finding, were also included. Patients had to be 18 years or older at the time of the performance of the scan. When a patient had undergone multiple 18 F-FDG PET(/CT) scans for LVV, only the first scan was included. Patients whose files explicitly stated that they used immunosuppressive drugs at the time of the scan were excluded. Patients with LVV in their medical history were also excluded. The time point for requesting a 18 F-FDG PET(/CT) in the diagnostic work-up was chosen by the referring physician. Data The clinical and laboratory data were recorded from the electronic medical records. Besides sex and age at the time of the performance of the 18 F-FDG PET(/CT) scan, the following data were recorded, based on findings in the literature [2]: &

Systemic: malaise, weight loss, nausea, fatigue, fever, and night sweats

Clin Rheumatol (2012) 31:861–871 Fig. 1 Left image is a maximum intensity projection of the PET investigation; pathological activity in aortic arch and abdominal aorta until the bifurcation, and in the subclavian, brachial, and carotic arteries. SUV max 5.4. Right image is a transverse CT slice; no wall thickening in thoracic and abdominal aorta

Fig. 2 Left image is a maximum intensity projection of the PET investigation showing pathological uptake in the carotids and bilateral subclavia. SUV max 2.4. Discrete increased uptake in thoracic aorta. The patchy uptake, with low intensity, in the abdominal aorta is mainly due to atherosclerosis seen on the coronal CT slice. Right image is a coronal CT slice of the abdominal aorta

863

864

Clin Rheumatol (2012) 31:861–871

Fig. 3 Left image is a maximum intensity projection of the PET investigation showing pathological activity in aortic arch and abdominal aorta until the bifurcation, and in both subclavian arteries and art. brachialis on both sides. SUV max 4.3. Right image is a transverse CT slice showing wall thickening of the aorta at the level of the crus of the diaphragm

& & &

& &

&

Temporal arteritis: jaw claudication, hyperesthesia of the scalp, thickening of the temporal artery or pain on palpation, temporal arteritis in the past, and visual disorders Musculoskeletal: myalgia, artralgia, and PMR in the past Cardiovascular: claudication of the arms, claudication of the legs, weak peripheral pulse, Raynaud’s phenomenon, orthostatic hypotension, and echocardiographically confirmed aortic regurgitation Pulmonary: radiographically confirmed pulmonary infiltrate Neurological: dizziness, headache, cerebrovascular accident or transient ischemic attack in the past year, symptoms of polyneuropathy (weakness, numbness, tingling, pain), and electromyographically proven polyneuropathy Laboratory tests: hemoglobin, lymphocytes, thrombocytes, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), albumin, creatinine, and alkaline phosphatase

The symptoms ought not to be explained by another medical condition the patient had already been diagnosed with. The clinical parameters were considered positive or negative only when the results were explicitly stated in the

patient’s history and as missing data when no data were recorded. Temporal arteritis in the past, PMR in the past, and cerebrovascular accident or transient ischemic attack in the past were considered negative when no data were recorded. The 18 F-FDG PET(/CT) scans were considered positive if a smooth linear pattern of 18 F-FDG uptake was found in the aorta or its main branches with an intensity higher than the liver [34]. 18

F-FDG PET(/CT) scanners

Each participating medical center used another equipment and protocol: & & &

Medical Center Leeuwarden: Siemens Biograph 6 LSO HI-REZ (hybrid PET/CT), 4 MBq/kg, 90 min between injection and start acquisition, 3 min per bed position University Medical Center Groningen: Siemens ECAT HR+(stand-alone PET), 5 MBq/kg, 60 min between injection and start acquisition, 7 min per bed position VU University Medical Center: Siemens HR+(standalone PET), 3.8 MBq/kg, 60 min between injection and start acquisition, 5 min per bed position; Philips Gemini TF (hybrid PET/CT), 3.8 MBq/kg, 60 min between injection and start acquisition, 2 min per bed position

Clin Rheumatol (2012) 31:861–871

&

Isala Klinieken: GE (hybrid PET/CT), 3.8 MBq/kg, 60 min between injection and start acquisition, 4 min per bed position

Statistical analysis The statistical analysis was performed with SPSS (version 17.0) for Windows. For the analysis of the categorical parameters, the chi-square test was used. Where applicable the Fisher’s exact test was used. The continuous parameters were analyzed using the independent t test or, in case of a non-normal distribution, the Mann–Whitney U test. The multivariate analysis was performed using logistic regression; for the inclusion a p value of 0.05 was chosen, for the exclusion a p value of 0.10. All tests were two sided, and results with a p value of ≤0.05 were considered statistically significant.

Results Patient characteristics Between January 2000 and December 2009, a total of 304 18 F-FDG PET(/CT) scans were performed for LVV in the participating centers. Of these 162 were 18F-FDG PET scans and 142 18F-FDG PET/CT scans. The proportion of patients with a positive scan was similar in the groups that underwent a 18F-FDG PET scan or a 18F-FDG PET/CT scan. The patients from the different hospitals were equally distributed between the group with large vessel 18F-FDG uptake and the group without large vessel 18F-FDG uptake. A total of 62 scans (20%) were positive for LVV. In 24 cases (39%), a temporal artery biopsy was performed, of which 12 biopsies (50%) were positive and 12 biopsies (50%) were negative. In the remaining 38 cases (61%), no temporal artery biopsy was performed. Even in the cases that the biopsy was negative or not performed, all patients were treated for LVV, based on the positive 18 F-FDG PET/ CT scan. These and other characteristics are shown in Table 1. Results of

18

F-FDG PET(/CT) scans

In the group with a scan positive for LVV, all but one patient demonstrated homogeneous smooth linear and long segmental uptake, with higher intensity compared to the liver and in the thoracic aorta and its main branches, which is considered to be a characteristic pattern of GCA [36]. Only one patient showed a different pattern: pathological uptake was found only in the wall of the aortic arch and particularly intense uptake in the lateral wall and perivascular space adjacent to

865 Table 1 Baseline characteristics Positive PET(/CT) Negative PET(/CT) (N062) (N0242) Age

65.9±13.4

58.6±16.5

Female sex

47 (76%)

132 (55%)

Stand-alone FDG-PET Hybrid FDG-PET/CT

33 (53%) 29 (47%)

129 (53%) 113 (47%)

Positive result temporal artery biopsy Negative result temporal artery biopsy

12 (19%) 12 (19%)

the truncus pulmonalis. “Atypical” Cogan’s syndrome was diagnosed on the basis of sensorineural deafness with improvement on steroids and circumscript large vessel vasculitis in the aortic arch [37]. Seventeen patients with doubtful increased large vessel 18 F-FDG uptake (inferior to the liver) on the 18F-FDG PET(/CT) scan, were subsequently treated for LVV. In these cases the diagnosis was based on the clinical picture, and it usually involved patients with typical signs of temporal arteritis. Other frequent diagnoses in this group were PMR, inflammatory disorders of the joints, other systemic diseases, and malignancies. These diagnoses were based on the results of the 18F-FDG PET(/CT) scan in about 50% of the cases. In 86 patients (36%), no diagnosis could eventually be made. The most frequent diagnoses and their respective numbers are shown in Table 2. Comparing both groups The clinical parameters of both groups are shown in Table 3. Because of a large amount of missing data, the number of recorded patients varied widely between parameters. In the Table 2 Most frequent diagnoses in the group with a negative test result for large vessel vasculitis Frequency (N0242) Large vessel vasculitis Polymyalgia rheumatica Inflammatory disorders of the jointsa Other systemic disordersb Malignancies Other diagnosesc No classifying diagnosisd a

17 (7%) 22 (9%) 25 (10%) 31 (13%) 13 (5%) 48 (20%) 86 (36%)

Among others: rheumatoid arthritis and spondylarthropathies

b

Among others: small vessel vasculitis and M. Sjögren

c

Among others: vascular disease and infections

d No classifying diagnosis made or no diagnosis recorded in medical record

866

Clin Rheumatol (2012) 31:861–871

group with a positive 18F-FDG PET(/CT) scan, the most frequent symptoms were fatigue, headache, and weight loss. In the group with a negative scan, artralgia, fatigue, and myalgia were the most frequent symptoms. Both groups differed significantly regarding age (65.9 ± 13.4 versus 58.6±16.5 years, p00.002), female sex (76% versus 55%, p00.002), temporal arteritis in the past (10% versus 3%, p0 0.044), and artralgia (31% versus 67%, p00.000). In the group with a positive test result, 90% of the patients were aged over 50, compared to 69% in the group with a negative test result. This difference was significant with a p value of 0.001.

The results of the laboratory tests in both groups are shown in Table 4. Both groups differed significantly regarding thrombocytes (434±161 versus 373±168×109/l, p0 0.049) and ESR (72.6±31.0 versus 51.4±30.5 mm/h, p0 0.001). For the CRP, when a cutoff point of 10 mg/l was chosen, none of the patients in the group with a positive test result had a normal value, resulting in a sensitivity of 100% and a specificity of 26%. For the ESR, when a cutoff point of 20 mm/h was chosen, a sensitivity of 94% and a specificity of 16% were found. On the parameters that differed significantly between both groups in the univariate analysis, univariate and

Table 3 Clinical variables in both groups

Age Female General Malaise Weight loss Nausea Fatigue Fever Night sweats Temporal arteritis Jaw claudication

p

Positive PET(/CT) (N062)

Negative PET(/CT) (N0242)

65.9±13.4 47/62 (76%)

58.6±16.5 132/242 (55%)

17/19 (89%) 27/36 (75%) 6/11 (55%)

60/69 (87%) 62/106 (58%) 13/29 (45%)

NS NS NS

29/31 (94%) 18/35 (51%) 9/14 (64%)

77/81 (95%) 61/137 (45%) 22/47 (47%)

NS NS NS NS

0.002 0.002

5/18 (28%)

7/34 (21%)

Hyperesthesia of the scalp

7/16 (44%)

6/24 (25%)

Thickening temporal artery or pain on palpation Temporal arteritis in the past

1/23 (4%) 6/62 (10%)

5/28 (18%) 8/242 (3%)

NS 0.044

10/25 (40%)

16/52 (31%)

NS

18/24 (75%)

69/81 (85%)

NS

9/29 (31%) 8/62 (13%)

78/116 (67%) 14/242 (6%)

6/10 (60%) 3/4 (75%)

5/9 (56%) 6/14 (43%)

NS NS

11/56 (20%) 19/40 (48%) 2/3 (67%) 5/9 (56%)

NS NS NS NS

21/175 (12%)

NS

15/20 (75%) 43/78 (55%) 11/242 (5%) 30/45 (67%) 9/16 (56%)

NS NS NS NS NS

Visual disorders Musculoskeletal Myalgia Arthralgia Polymyalgia rheumatica in the past Cardiovascular Claudication of the arms Claudication of the legs Weak peripheral pulse Raynaud phenomenon Orthostatic hypotension Aortic regurgitation Pulmonary Pulmonary infiltrate Neurological Dizziness Headache CVA or TIA in the past Polyneuropathy (symptoms) Polyneuropathy (EMG) NS non-significant (p