Positron emission tomography (PET): Evaluation of chronic periaortitis

Authors


Abstract

Objective

To evaluate the presence and extent of large-vessel inflammation in patients with chronic periaortitis (CP) using 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET).

Methods

A consecutive case series consisting of 7 patients with CP seen over a 3-year period and a control group of 14 patients with malignancy were evaluated with FDG-PET. For every case we selected 2 age- and sex-matched controls who underwent PET imaging for malignancy. The diagnosis of CP was made by means of computed tomography. PET imaging was performed at diagnosis before therapy was started. Measurement of vascular uptake was graded using a 4-point semiquantitative scale.

Results

All patients had evidence of grade 2+ or 3+ vascular uptake in the abdominal aorta and/or iliac artery. No controls showed vascular uptake greater than 1+. Vascular uptake in the thoracic aorta and/or in its branches was seen in 3 (43%) of 7 patients. Vascular uptake in abdominal aorta and/or iliac artery was observed in patients with CP but not in controls (100% versus 0%). There was also a significantly more frequent FDG uptake in the large thoracic arteries in case-patients compared with controls (43% versus 0%; P = 0.03).

Conclusion

FDG-PET scan shows in patients with CP the presence of a large-vessel vasculitis involving abdominal aorta and common iliac arteries, which in some patients is also extended to thoracic aorta and/or its branches.

INTRODUCTION

Chronic periaortitis (CP) is a clinical pathologic entity characterized by a fibroinflammatory reaction that extends from the adventitia of the abdominal aorta and of the common iliac arteries into the retroperitoneum, and often leads to the encasement of adjacent structures (ureters, inferior vena cava). Three main entities are included: idiopathic retroperitoneal fibrosis (IRF), inflammatory abdominal aortic aneurysms (IAAAs), and perianeurysmal retroperitoneal fibrosis (PRF) (1–3). IRF is characterized by the periaortic deposition of collagen, which by extending into the retroperitoneum, often obstructs the ureters and other abdominal organs. A dilated aorta is usually not present in IRF. In IAAAs the mass develops around a dilated aorta and usually does not cause obstructions. PRF may represent a link between IAAAs and IRF. These conditions have common clinical and histopathologic findings, and thus probably represent different manifestations of the same disease.

Although CP has traditionally been considered an excessive local inflammatory response to atherosclerotic plaque antigens (2, 4, 5), its clinical presentation is characterized by constitutional symptoms and elevated inflammatory markers, which suggests a systemic process. Furthermore, an association between CP and systemic autoimmune disorders, in particular systemic vasculitis, has frequently been reported (6–13).

Recently, 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET) has been proposed as a potentially useful tool for detection of inflammatory changes in large-vessel vasculitis (14, 15). FDG is a deoxyglucose labeled with fluorine-18 (18F), a positron-emitting radionuclide that identifies areas of high glucose metabolic activity. Since inflammatory cells have an increased glucose uptake, high levels of glucose metabolism are seen in vessel walls if inflammation is present. FDG-PET may be a more sensitive measure of vascular inflammation than conventional methods. We have used FDG-PET to evaluate the presence and extent of large-vessel inflammation in a group of patients diagnosed with CP by computed tomography (CT).

SUBJECTS AND METHODS

Subjects.

We recruited 7 patients with CP seen consecutively in Parma and Reggio Emilia Hospitals over a 3-year period. The diagnosis of CP was made by CT (Figure 1). None had any clinical or laboratory evidence of chronic infections or tumors, none had undergone previous abdominal surgery, and none were taking any medications known to induce the development of CP. All the patients received a standardized laboratory evaluation, as previously described (7). PET imaging was performed at diagnosis before therapy was started.

Figure 1.

Computed tomographic appearance of idiopatic retroperitoneal fibrosis in patient 4. The scan (taken at the level of the lower abdominal aorta) shows a peri-aortic retroperitoneal mass (arrow).

We also included a control group consisting of 14 patients with malignancy. For every case we selected 2 age- and sex-matched controls who underwent PET imaging for malignancy.

Methods.

All PET studies were performed on a dedicated system (C-PET, Philips, Cleveland, OH). The patients underwent at least 4 hours of fasting and good oral hydration before the study. Euglycemia was documented before 18F-FDG administration by measuring blood glucose level. An intravenous line was positioned in an antecubital vein to make the intravenous administration of 18F-FDG (2 MBq/kg body weight) easier and to improve the hydration by administering 500 ml of saline solution in the following 40 minutes. During the distribution phase of FDG, the patients were sitting in a comfortable armchair in a quiet room. A whole-body scan was performed 60–90 minutes after the FDG administration, acquiring both transmission and emission images. Raw data were processed using an iterative reconstruction method (RAMLA algorithm) and both nonattenuation-corrected and attenuation-corrected images were reconstructed (144 x 144 matrix, 4 mm slice thickness).

Two nuclear medicine physicians with expertise in PET imaging reviewed and scored the images in an interactive computer system with use of a linear grey scale and varying degrees of background subtraction. They were blinded to the patients' diagnosis. Their results were compared and any disagreements were resolved by consensus.

The following arteries were evaluated: aortic arch, ascending and descending thoracic aorta, brachiocephalic artery, carotid arteries, subclavian arteries, axillary arteries, brachial arteries, abdominal aorta, iliac arteries and femoral arteries. Measurement of vascular uptake was graded using a 4-point semiquantitative scale where 0 = absence of uptake; 1+ = low grade uptake (uptake present but lower than liver uptake); 2+ = intermediate grade uptake (similar to liver uptake); and 3+ = high grade uptake (between liver and cerebral uptake or similar to the uptake in cerebral cortex). Scores 0 and 1+ were considered negative, while scores 2+ and 3+ were considered positive. The study was approved by the local Ethics Committee and all patients gave oral informed consent.

Statistical analysis.

Chi-square test or Fisher's exact test, as appropriate, were used to compare categorical data.

RESULTS

The demographic characteristics and clinical findings in the 7 case patients (5 patients with IRF and 2 with PRF) are shown in tables 1 and 2. The case patients consisted of 4 men and 3 women with a median age of 60 years (range 37–82 years) at diagnosis. The most common symptoms were abdominal and/or back pain, as well as constitutional symptoms such as fatigue, anorexia, weight loss, fever, diffuse myalgias, and arthralgias.

Table 1. Demographics of the 7 patients with chronic periaortitis at the time of diagnosis*
PatientSexAge, yearsType of CPDistribution of CPUreteral involvementARFESR (mm/hour)CRP (mg/liter)Autoantibodies
  • *

    CP = chronic periaortitis; ARF = acute renal failure; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein (normal value <5 mg/liter); PRF = perianeurysmal retroperitoneal fibrosis; ANA = antinuclear antibodies; IRF = idiopathic retroperitoneal fibrosis; RF = rheumatoid factor; pANCA = perinuclear antineutrophil cytoplasmic antibodies; anti-MPO = antimyeloperoxidase.

1M60PRFPeriaortic & periiliacNo9920
2M75PRFPeriaortic & periiliacNo11227ANA (1:2560, speckled pattern)
3M53IRFPeriaortic & periiliacUnilateralNo9126
4F60IRFPeriaortic & periiliacNo11443
5F82IRFPeriiliacUnilateralNo8227RF & pANCA (anti-MPO titer: 60 EU/ml)
6F63IRFPeriaortic & periiliacBilateralYes100182ANA (1/320, speckled pattern)
7M37IRFPeriaortic & periiliacBilateralYes97105
Table 2. Characteristics of the seven patients with chronic periaortitis at the time of diagnosis
PatientClinical symptomsConstitutional Symptoms*Surgical treatmentRetroperitoneal biopsy
  • *

    Constitutional symptoms = fatigue, anorexia, weight loss, fever, diffuse myalgias, and arthralgias.

1Abdominal pain, bilateral hydrocele, left varicoceleAbsentAneurysm repairYes
2AbsentAneurysm repairNo
3Abdominal, back and right testicular pain, right varicocelePresentExploratory laparoscopyYes
4Abdominal and back painPresentNo
5Back pain, inferior vena cava syndromePresentNo
6Abdominal and back pain, left femoral-popliteal thrombosisPresentLaparotomy for ureterolysisYes
7Abdominal and back pain, left iliac-femoral-popliteal vein thrombosisAbsentLaparotomy for ureterolysisYes

Ureteral obstructive disease was present in 4 patients, with unilateral hydronephrosis occurring in 2 patients and bilateral in 2. The 2 patients with bilateral hydronephrosis associated rapidly progressive renal failure.

Laboratory values observed in all patients at diagnosis were as follows: elevated erythrocyte sedimentation rate (median 99 mm/hour, range 82–114 mm/hour) and C-reactive protein level (median 27 mg/liter, range 20 mg/liter–182 mg/liter; normal 0–5 mg/liter). Two patients were antinuclear antibody positive (speckled pattern). None of the patients was antiextractable nuclear antigen antibody or anti-double stranded DNA antibody positive. One patient was serum antineutrophil cytoplasmic antibodies (ANCA) positive; indirect immunofluorescence revealed perinuclear ANCA, which proved to be myeloperoxidase-ANCA at enzyme-linked immunosorbent assay (60 EU/ml, normal range 0–5) (Medipan; Selchow-Berlin, Germany).

Two patients had surgical repair of the inflammatory aneurysm and 2 patients underwent laparotomy for ureterolysis. Periaortic retroperitoneum tissue samples were obtained from 4 patients. In agreement with previous findings, retroperitoneal specimens consisted of sclerotic tissue with a marked inflammatory infiltrate mainly consisting of mononuclear cells distributed within and around the adventitia of the aorta and of the small and medium sized retroperitoneal vessels (Figure 2) (7).

Figure 2.

Retroperitoneal soft tissue specimens in two patients with idiopathic retroperitoneal fibrosis. A,In patient 6, sclerotic tissue with diffuse and nodular perivascular (arrow) inflammatory infiltrate. (Hematoxylin and eosin stained; original magnification × 4.) B, In patient 7, high-magnification appearance of abundant collagen bundles infiltrated by an intense inflammatory tissue mainly consisting of mononuclear cells. (Hematoxylin and eosin stained, original magnification × 40.)

Table 3 shows the results of FDG-PET in patients with CP. None of the subjects showed vascular uptake greater than 1+ (data not shown). All patients had evidence of grade 2+ or 3+ vascular uptake in the abdominal aorta and/or iliac artery. Vascular uptake in the thoracic aorta and/or in its branches was seen in 3 of 7 patients (43%) (Figure 3). Vascular uptake in the abdominal aorta and/or iliac artery was observed in patients with CP but not in the controls (100% versus 0%). There was also a significantly more frequent FDG uptake in the large thoracic arteries in case patients compared with controls (43% versus 0%; P = 0.03).

Table 3. 18F-fluorodeoxyglucose uptake score in the 7 patients with chronic periaortitis*
PatientAscending aortaAortic archDescending aortaCarotid arterySubclavian arteryAxillary arteryBrachial arteryAbdominal aortaIliac arteryFemoral artery
  • *

    0 = absence of uptake; 1+ = low grade uptake (uptake present but lower than liver uptake); 2+ = intermediate grade uptake (similar to liver uptake); 3+ = high grade uptake (between liver and cerebral uptake or similar to the uptake in cerebral cortex).

11+0000002+2+1+
201+2+1+2+2+2+3+2+1+
32+2+1+1+1+1+1+3+1+1+
43+1+01+1+01+3+2+0
500000001+2+0
600000003+3+0
700000002+2+1+
Figure 3.

Positron emission tomography shows increased 18F-fluorodeoxyglucose uptake (arrows) in thoracic aorta and abdominal aorta in a patient with idiopathic retroperitoneal fibrosis. A,Sagittal view. B, Coronal view.

DISCUSSION

In all our patients, the diagnosis of CP was made by means of contrast-enhanced CT, which is the diagnostic test of choice for this condition. We have used FDG-PET to evaluate the presence and extent of large-vessel inflammation in patients with CP. FDG-PET suggests that CP is a vascular inflammatory process mainly localized to the abdominal aorta and/or to the common iliac arteries. However, involvement of the thoracic aorta and its branches may also occur, and in fact has been observed in 43% of our patients but none of the controls. These FDG-PET findings are similar to those observed in giant cell arteritis (GCA) and Takayasu's arteritis (TA) (14, 15). These data suggest that CP may represent a large-vessel vasculitis involving the abdominal aorta and common iliac arteries, which in some patients also extends to the thoracic aorta and/or its branches.

The pathogenesis of CP is obscure. Parums et al have suggested that it may be the result of an exaggerated local immune response to atherosclerotic plaque antigens such as oxidized low-density lipoproteins and ceroid (2, 4, 5). However, the role of atherosclerosis in CP remains controversial, mainly because no substantial difference in the incidence of advanced atherosclerotic disease has been clearly demonstrated between CP patients and healthy controls (16, 17).

The presence of constitutional symptoms, the markedly raised acute-phase reactants, the presence of autoantibodies, and the frequent association with other autoimmune conditions suggest an ongoing systemic process rather than a local reaction (1–3, 6–13). A recent case-control study that compared inflammatory and noninflammatory abdominal aortic aneurysms showed that IAAAs are more frequently associated with systemic autoimmune diseases (6). Likewise, in a recent study, we found that in 7 (44%) of 16 cases CP was associated with other autoimmune conditions, namely ANCA-positive renal disease and autoimmune tyroiditis (7). Finally, various case reports have described an association between CP and numerous autoimmune diseases including well-defined small- and medium-vessel vasculitis (Wegener's granulomatosis, polyarteritis nodosa, Henoch-Schoenlein purpura) (9–11), unclassifiable “systemic vasculitis” (8, 12), aortitis (13), or different types of immune-mediated glomerulonephritis (membranous, membrano-proliferative, and rapidly progressive glomerulonephritis) (18, 19). Ankylosing spondylitis, a condition sometimes linked to ascending aortitis, has also been reported to be associated with CP (12, 20, 21), whereas the association with rheumatoid arthritis and systemic lupus erythematosus is uncommon (7, 22). Finally, CP (particularly IRF) may frequently be associated with fibroinflammatory disorders affecting other organs (sclerosing cholangitis, mediastinal fibrosis, Riedel's and chronic autoimmune thyroiditis) (3, 8, 23), most of which are thought to have an autoimmune pathogenesis.

The concept of a systemic inflammatory nature of CP is also borne out by the histologic similarities of CP with large-vessel vasculitis, such as the prominent adventitial inflammation and the involvement of vasa vasorum (24). Moreover, autopsy studies have shown that moderate adventitial inflammation and fibrosis may not be limited to the abdominal aorta, but also involve its thoracic portion (25, 26). In fact, aortic arch syndrome has been reported in CP, confirming the possible involvement of the thoracic aorta (13). Our findings show that by using FDG-PET, vascular inflammation in the thoracic aorta and/or its branches may be detected in approximately half of the cases.

In light of these findings, it is tempting to speculate that CP may originate as a primary arteritis displaying a particular tropism for the abdominal aorta and common iliac arteries. The vascular inflammation may subsequently extend to the retroperitoneum, eliciting a fibroinflammatory reaction. Large-vessel vasculitis may have different disease expression depending on the territories of the arterial trees involved. For instance, large-vessel GCA produces a distinct spectrum of clinical manifestations related to inflammation of subclavian and axillary vessels and often occurs, differently from classic GCA, without involvement of the cranial arteries (27). However, GCA of the abdominal aorta and iliac arteries is rare.

TA primarily targets the subclavianaxillary and proximal brachial arteries in a pattern indistinguishable from that of large-vessel GCA (28). However, the abdominal aorta, renal, and mesenteric superior arteries are involved in 30–40% of patients with TA. In GCA, TA, and CP, intracranial arteries are essentially spared. The mechanisms underlying the different pattern of arterial involvement in large-vessel vasculitis are unclear.

Atherosclerosis might produce false-positive vascular FDG uptake. Yun et al have evaluated the presence of FDG vascular uptake in the abdominal aorta, iliac, and proximal femoral arteries in 137 consecutive patients undergoing PET scan, mainly for malignancy (29). Fifty percent of the patients showed vascular uptake in at least one vessel, with an increased prevalence in older patients. The abdominal aorta was visualized less frequently compared with the iliac and femoral arteries. Although not the primary focus of the study, the thoracic aorta and pulmonary arteries were also evaluated, but their walls could be identified only in a limited number of cases. This increased vascular uptake could be partially explained by the presence of macrophages within the atherosclerotic plaque. Thus, it could be argued that the presence of atherosclerotic plaques might produce false-positive vascular FDG uptake.

However, to confirm that vascular uptake in CP is specific (i.e., due to inflammation of the vessel wall), we compared our patients with an adequate number of age- and sex-matched controls. Using a semiquantitative (0–3) grading, where grade 0–1 was considered negative and 2–3 positive, we found that most patients (6 of 7) had vascular abdominal uptake of grade 2 or 3; however, in the controls vascular uptake was consistently ≤1, i.e., negative. Furthermore, positive vascular FDG uptake was also seen in some patients in the thoracic aorta/subclavian arteries, which are less prone to atherosclerosis than the abdominal aorta, which further confirms the specificity of high FDG uptake for true vasculitis. Therefore, we feel that the high FDG uptake (considered positive) observed in our patients in the abdominal and thoracic aorta is expression of true vasculitis.

In conclusion, FDG-PET demonstrates in patients with CP, large vessel involvement of the abdominal aorta and of the common iliac arteries, which in some patients also extends to the thoracic aorta and/or its branches. The high-grade FDG vascular uptake and the concomitant inflammatory features found in patients with CP suggest that this vascular involvement may represent vasculitis. This raises the question as to whether CP should be included in the group of the large-vessel vasculitides.

Acknowledgements

We are indebted to Dr. Domenico Corradi for providing histologic documentation

Ancillary