Retroperitoneal fibrosis is characterized by the presence of a fibroinflammatory retroperitoneal mass that often entraps the ureters or other abdominal organs (1). In most cases it is idiopathic, but it may also be secondary to infections, malignancies, or drugs or it may be associated with systemic autoimmune diseases (2, 3).
Idiopathic retroperitoneal fibrosis (IRF) usually presents with constitutional symptoms and back or abdominal pain. Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level are frequently high, and in some cases patients are positive for autoantibodies (4). The diagnosis is made by means of computed tomography (CT) or magnetic resonance imaging (MRI), but a histologic examination of the retroperitoneal tissue is sometimes required. IRF frequently has a chronic-relapsing course but, except for acute-phase reactants, no parameters are used to monitor disease activity.
Positron emission tomography (PET) with 18F-fluorodeoxyglucose (18F-FDG) enables the in vivo measurement of metabolic processes. It is an established functional imaging modality in oncology and neurology (5), but has recently been introduced in the evaluation of various inflammatory diseases, particularly large-vessel vasculitis (6).
We describe the case of a patient with IRF who was studied by means of 18F-FDG–PET and discuss the potential role of this technique in assessing disease activity and clinically managing IRF patients.
A 60-year-old woman was admitted to the hospital because of flank pain, fatigue, anorexia, and weight loss. Mild hypertension had been diagnosed 6 years earlier and she was taking enalapril (10 mg/day). Upon admission, a physical examination revealed bilateral costovertebral angle and abdominal tenderness. Her serum creatinine was 7.5 mg/dl, ESR 105 mm/hour, and CRP 76 mg/liter (normal <5 mg/liter). She tested positive for antinuclear antibodies (titer 1/320, speckled pattern) but negative for antineutrophil cytoplasmic, antiextractable nuclear antigen, and anti–double-stranded DNA antibodies.
Sonography showed a hypoechoic mass surrounding the abdominal aorta and the inferior vena cava, extending laterally to involve the ureters with resultant bilateral hydronephrosis. Cytoscopically positioned ureteral stents resolved the hydronephrosis and normalized the serum creatinine level. An MRI revealed a periaortic and periiliac retroperitoneal mass (Figure 1A) that entrapped both ureters, which suggested a diagnosis of IRF. The patient underwent a CT-guided mass biopsy, and histology showed sclerotic tissue with an inflammatory infiltrate consisting of lymphocytes, plasma cells, and macrophages, thus confirming the diagnosis of IRF.
18F-FDG–PET was performed to evaluate the activity of IRF and to screen for other sites of fibrosis. A whole-body scan was acquired using a C-PET ADAC scanner (ADAC Laboratories, Milpitas, CA) 90 minutes after the intravenous administration of 18F-FDG (2 MBq/kg body weight). The images showed strikingly higher 18F-FDG metabolism in the aortoiliac region, with no other sites of pathologic uptake (Figure 1B, 1C).
Oral prednisone (1 mg/kg/day) was started: the patient's symptoms dramatically improved and 2 weeks later ESR and CRP normalized (12 mm/hour and 4.6 mg/liter, respectively). Steroid therapy was tapered and, when its dosage was 10 mg/day 4 months after the start of treatment, the patient was still asymptomatic with normal ESR (6 mm/hour) and CRP (3.8 mg/liter). MRI was repeated and showed a considerable reduction in the size of the retroperitoneal mass (Figure 2A), particularly in the periiliac region. 18F-FDG–PET was repeated and showed that the 18F-FDG uptake was markedly reduced, although it could still be observed in a small aortic region (Figure 2B, 2C).
The ureteral stents were removed and repeated sonographies during the following months showed only minimal left hydronephrosis; furthermore, ESR, serum creatinine, and CRP levels remained normal. Prednisone was stopped after 7 months of treatment, and a third MRI was recorded 1 month later: surprisingly, the retroperitoneal mass had considerably increased in size (Figure 3A), and bilateral hydronephrosis also recurred. At that time, the patient was still asymptomatic, ESR was 10 mm/hour, and CRP was 4.4 mg/liter. 18F-FDG–PET was therefore repeated and showed marked 18F-FDG accumulation in the periaortic and right periiliac regions; moreover, bilateral hydronephrosis also caused high 18F-FDG uptake in both kidney pelvic regions (Figure 3B, 3C). Ureteral stents were again positioned, and we started more aggressive treatment using prednisone and azathioprine.
PET is an imaging method based on the differential uptake of the 18F-FDG glucose analog by actively metabolizing cells. 18F-FDG is transported into cells on the basis of their rate of glycolysis; this characteristic has been exploited in cancer imaging: given the increased glycolytic rate of malignant cells, high 18F-FDG concentrations are achieved relative to surrounding healthy tissue (7). However, lymphocytes, macrophages, neutrophils, and fibroblasts also avidly take up 18F-FDG, particularly under activation conditions (8). Thus, primary inflammatory diseases (such as sarcoidosis) as well as infectious diseases (such as tuberculosis or aspergillosis) can be imaged by means of 18F-FDG–PET (9).
Several studies have recently highlighted the potential role of 18F-FDG–PET in the assessment of inflammatory vascular diseases, particularly giant cell arteritis and Takayasu's arteritis. In these conditions, it can detect occult areas of vascular inflammation and thus demonstrate the full extent and distribution of vasculitis (6, 10, 11); furthermore, in Takayasu's arteritis, 18F-FDG–PET proved to be more reliable than MRI in monitoring disease activity (10).
IRF usually presents as a fibroinflammatory retroperitoneal mass. In its early stages, histology shows active inflammation with a mixture of macrophages, lymphocytes, scattered plasma cells, and eosinophils in a framework of fibroblasts and collagen bundles. In the late stages, fibrous scarring occurs and the tissue becomes relatively avascular and acellular (1). The increased aortoiliac 18F-FDG accumulation observed in our patient at the times of disease onset and relapse was therefore probably due to the presence of inflammatory cells and actively metabolizing fibroblasts.
One of the crucial problems in IRF is assessing disease activity, because this may drive therapeutic decisions. IRF often has a chronic clinical course and may relapse in more aggressive cases, but its clinical manifestations are often vague. Therefore, disease activity cannot always be evaluated on the basis of a patient's signs or symptoms.
ESR and CRP are frequently used to monitor the clinical course of the disease; however in our patient, 18F-FDG–PET proved to be more reliable than the levels of acute-phase reactants in the assessment of IRF activity. At the time of diagnosis, ESR and CRP were high and PET revealed marked aortoiliac 18F-FDG uptake. After 4 months of therapy, however, ESR and CRP were normal (thus apparently indicating complete disease remission), whereas PET showed a residual aortic area of pathologic 18F-FDG uptake. The most striking discrepancy between the PET results and the acute-phase reactant levels was found at the time of disease relapse: PET revealed marked 18F-FDG accumulation but ESR and CRP were still normal.
A high ESR and high CRP level hallmark many inflammatory and autoimmune diseases, and are often thought to mirror disease activity (12); however, a number of exceptions have been reported. Giant cell arteritis and polymyalgia rheumatica are typically characterized by a high ESR, but up to 22% of patients have a low ESR at diagnosis, and their disease course is not different from that observed in patients with a high ESR. In Takayasu's arteritis, ESR and CRP often do not reflect the state of vascular inflammation: histopathologic studies showed that >40% of patients with normal acute-phase reactant levels actually still have active arteritis at the time of vascular surgery (12).
Because of its high cost, 18F-FDG–PET cannot be routinely recommended for IRF patients, but its ability to assess disease activity may be relevant in some clinical settings. For example if medical treatment leads to a reduction in the retroperitoneal mass, it can help to detect the presence of active inflammatory foci inside the residual tissue. Furthermore, if a patient with a previous diagnosis of IRF reports symptoms that may suggest a disease relapse, PET could show whether metabolic activity has reappeared within the mass and therefore whether the patient may benefit from additional treatment. Finally, it may help identify the most appropriate sites for retroperitoneal biopsy.
18F-FDG–PET allows whole-body imaging. It can reveal other diseased sites (e.g., thyroid, mediastinum), such as those found in multifocal fibrosclerosis (13), or it may detect infectious or neoplastic processes to which retroperitoneal fibrosis may be secondary or associated (14).
In conclusion, we believe that 18F-FDG–PET can be considered a useful functional imaging modality in IRF patients, and that it provides a highly reliable means of assessing disease activity.
The authors gratefully acknowledge Dr. P. Schianchi and Dr. R. Cobelli for their help in preparing the images.