Abstract
The rates of growth of medically treated abdominal aortic aneurysms (AAA) are difficult to determine, and relationships with parietal inflammation and with metabolic parameters from 18F-FDG PET remain unclear. This 18F-FDG PET sequential observational study was aimed at analyzing the metabolic changes accompanying the growth phases of medically treated AAA. Methods: Thirty-nine patients (37 men; age [mean ± SD], 71 ± 12 y) exhibiting small and medically treated AAA (maximal diameter, 46 ± 3 mm) underwent 18F-FDG PET and CT angiography at baseline and 9 mo later. Clinical and imaging parameter correlates of the 9-mo increase in maximal diameter were investigated; these included 18F-FDG maximal standardized uptake values (SUVmax) averaged for slices encompassing the AAA volume. Results: Of the 39 patients, 9 (23%) had a significant (≥2.5 mm) increase in maximal diameter at 9 mo, whereas the remaining 30 did not. The patients with an increase in maximal diameter at 9 mo exhibited lower SUVmax within the AAA at baseline than patients who did not have such an increase (1.80 ± 0.45 vs. 2.21 ± 0.52; P = 0.04); they also displayed a trend toward greater changes in SUVmax at 9 mo (difference between 9 mo and baseline: +0.40 ± 0.85 vs. −0.06 ± 0.57; P = 0.07). Similar levels were ultimately reached in both groups at 9 mo (2.20 ± 0.83 and 2.15 ± 0.66). SUVmax was a significant, yet modest, baseline predictor of the absolute change in maximal diameter during follow-up (P = 0.049). Conclusion: The enhancement in the maximal diameter of small AAA was preceded by a stage with a low level of 18F-FDG uptake, but this low level of uptake was no longer documented after the growth phases, suggesting a pattern of cyclic metabolic changes.
The prevalence of abdominal aortic aneurysms (AAA) remains high worldwide, affecting approximately 5% of elderly men (1). For most AAA, which are small and uncomplicated, only a periodic follow-up of the maximal diameter is required (1)—even if such screening is not sufficient to accurately determine the rupture risk at the individual level (2). Small AAA can rupture at nonnegligible rates of approximately 1% per year (3), whereas many AAA can grow to dramatic sizes without rupturing.
However, diameter-based screening is still important because large randomized trials have shown that survival is not improved by prophylactic surgery until a maximal diameter of 5.5 cm is reached (4). Additionally, enhanced risk is observed when increase rates are greater than or equal to 2 mm/y (5).
Several recent reports suggested that 18F-FDG PET imaging has the potential to aid in risk stratification for AAA. In particular, the increased 18F-FDG uptake in the AAA wall corresponds locally to macrophage infiltration and to enhanced proteolysis (6–9), 2 parameters having major roles in the growth and rupture of AAA (10). This increased uptake is also frequently documented when there is an obvious need for rapid surgical repair, such as for a symptomatic or rapidly expanding AAA (9,11).
In contrast, an abnormally low level of 18F-FDG uptake is commonly documented for uncomplicated AAA (12,13), presumably because of a decrease in cellular density (8,13). Furthermore, the level of uptake may be even lower in AAA that are more likely to exhibit a subsequent enhancement in maximal diameter (14). During the growth phases, however, how this 18F-FDG uptake evolves along with the metabolism and inflammation of AAA walls is unknown.
This 18F-FDG PET sequential observational study was aimed at analyzing the metabolic changes accompanying the growth phases of medically treated AAA.
MATERIALS AND METHODS
Patient Selection and Study Design
The study protocol was approved by the local ethics committee (Comité de Protection des Personnes agreement 06.11.02) and released on the ClinicalTrials.gov website under the identifier NCT02182908.
All study patients gave signed informed consent to participate and were prospectively included if they met the following criteria: presence of an infrarenal AAA of a common atherosclerotic origin and for which only medical therapy was required (a criterion that implies the exclusion of symptomatic AAA and AAA with a maximal diameter of ≥55 mm), AAA maximal diameter of greater than or equal to 40 mm at CT angiography, age of greater than 18 y, no contraindication for 18F-FDG PET or CT angiography, no antiinflammatory or immunosuppressant treatment, and no known disease that could lead to a marked inflammatory response.
The study patients were referred for baseline 18F-FDG PET, which needed to be performed no later than 15 d after the initial CT angiography, and for a control assessment that was scheduled to be performed 9 mo (±15 d) from the initial CT angiography and that involved additional 18F-FDG PET and CT angiography. On the days on which baseline 18F-FDG PET and control 18F-FDG PET were performed, the patients underwent a medical assessment with blood pressure measurement.
Hybrid PET/CT Recording
18F-FDG PET images were recorded on a Biograph hybrid system with 6-detector CT for attenuation correction and anatomic localization (Biograph 6 True Point; Siemens) (15,16). An 18F-FDG activity of 5.5 MBq/kg was injected intravenously after an overnight fast. The blood glucose level had been previously verified with a dedicated electronic device so as to ensure a level of less than 1.5 g⋅L−1 in all patients. CT imaging was initiated 90 min later and was immediately followed by a 3-dimensional PET recording with 3 or 4 bed positions of 7 min each; all patients were scanned from the apex of the lungs to the iliac bifurcation.
The main CT parameters were as follows: 130 kV, intensity adapted to noise index, 512 × 512 matrix, 2-mm slice thickness, and pitch of 1. 18F-FDG PET images were reconstructed by the ordered-subset expectation maximization method (3 iterations and 8 subsets) with corrections for scatter and attenuation and were subsequently displayed in a 128 × 128 matrix with 3.0 × 3.0 × 3.0 mm voxels. The standardized uptake value (SUV) was calculated by dividing the activity measured in each voxel by the total injected activity expressed per gram of body weight and corrected for radioactive decay.
CT Angiography Recording
Images of the AAA were recorded during breath-hold periods of about 10 s on a 64-detector CT scanner (LightSpeed VCT; GE Healthcare). A total iodinated contrast volume of 1.5–2 mL/kg, with an iodine concentration of 300–400 mg/mL, was injected in an upper-limb vein at a rate of 3–5 mL/s; next, 20–60 mL of saline physiologic solution was injected at the same rate. Acquisition was triggered by bolus-tracking software. The main CT parameters were as follows: craniocaudal recording direction; 100–140 kV, depending on the patient’s morphology; intensity adapted to noise index; rotation time of 0.6 s; pitch close to 1; 0.625-mm slice thickness; 512 × 512 matrix; and field of view adapted to the patient.
Analysis of Hybrid PET/CT Images
18F-FDG PET images were analyzed with a paired display of both CT and fused PET/CT images provided by an Esoft station (Siemens) (15,16). The averaged values of aortic activities were quantified on consecutive transaxial slices encompassing the abdominal aorta. On each slice, a region of interest was drawn around the aorta, including the aneurysm wall and content, to allow the maximal 18F-FDG uptake to be collected (maximal SUVs [SUVmax]). The values were averaged for slices covering the suprarenal abdominal aorta and the aneurysmal volume, with the exclusion of inferior and superior extremities (at a distance of 1 cm from each neck region). The averaged SUVmax was also divided by the mean blood voxel activity, estimated from the inferior vena cava, to provide maximal tissue-to-background ratios (17).
Analysis of CT Angiography
The maximal diameters of the AAA as well as diameter changes at 9 mo were determined with dedicated software, allowing a side-by-side display and reorientation of both baseline and 9-mo CT angiography studies (Volume Viewer; GE Healthcare). This analysis was performed separately and in a masked fashion by 2 experienced observers in accordance with current recommendations (delimitation of the external parietal limits after careful orientation, perpendicular to the long axis of the aneurysm long axis (18)). The maximal diameters measured by the 2 observers were subsequently averaged for further analyses, except when the difference was greater than 3 mm; in such cases, an additional and consensual determination was obtained from the 2 observers.
For assessment of the reproducibility of this process, maximal diameters were determined a second time from 36 CT angiography studies, and the mean ± SD of the differences between the 2 determinations was calculated. Thereafter, the corresponding limits of agreement were used to define the cutoff value for significant changes in maximal diameter between baseline and 9 mo.
In addition, the total volume of the AAA as well as the volumes of the lumen and thrombus portions of the AAA were evaluated with A3Dmax segmentation software (Object Research System) (19). This software enables the segmentation and 3-dimensional modeling of the lumen and thrombus volumes of the AAA after the manual placement of upper- and lower-limit planes. In the present study, these limit planes were placed at the levels of the inferior renal artery and the aortoiliac bifurcation.
Statistical Analysis
Discrete variables were expressed as percentages, and continuous variables were expressed as mean ± SD. Patients with a significant increase in the maximal diameter between baseline and 9 mo were defined as those for whom this difference exceeded the limits of agreement that were documented in the reproducibility analysis (upper limit of the 95% confidence interval). Mann–Whitney tests were used for 2-group comparisons of continuous variables, and Fisher exact tests were used for 2-group comparisons of discrete variables.
Wilcoxon signed rank tests were used for paired comparisons of continuous variables between baseline and 9 mo, the McNemar test was used for discrete variables, and the Spearman rank correlation coefficient method was used for determining the correlates of the differences in maximal diameter and volume of AAA between 9 mo and baseline. These correlates were investigated for baseline values only for CT angiography parameters and for both baseline and 9-mo values for the remaining parameters (Table 1).
Paired Comparisons of Main Recorded Variables at Baseline and 9 Months (Control) (n = 39)
For all tests, a P value of less than 0.05 was considered to indicate statistical significance.
RESULTS
A total of 51 patients were initially included and completed the baseline investigations. However, 12 patients were subsequently excluded, 1 because of a problem in the recording of 18F-FDG PET images and 11 because of the lack of 9-mo investigations (7 patient refusals, 2 early referrals to AAA surgery, 1 allergy to iodinated contrast material, and 1 referral to carotid surgery at the time of the 9-mo investigations). Thus, 39 patients completed the overall protocol and constituted the final study population.
Baseline Data
The mean age of the 39 patients was 71 y (SD, 12 y; range, 38–88 y). Only 2 patients (5%) were women, and 24 (62%) had a history of cardiovascular disease (stroke in 2, coronary artery disease in 15, and peripheral arterial obstructive disease of the lower limbs in 12). All but 2 (95%) had a history of smoking, 22 (56%) were treated for hypertension, and 5 (13%) were treated for diabetes.
Additional baseline characteristics are shown in Table 1. At CT angiography, the average maximal diameter of the AAA was 46 mm (SD, 3 mm; range, 40–54 mm), and the mean AAA volume was 101 mL (SD, 21 mL), including a thrombus volume of 40 mL (SD, 25 mL). At 18F-FDG PET, the mean SUVmax of the AAA was 2.1 (SD, 0.5), significantly lower than that of the suprarenal abdominal aorta (2.3 ± 0.6; P < 0.001). The mean SUVmax was unrelated to AAA diameter or volume. A high level of 18F-FDG uptake, which was easily detectable within the AAA, was a rare finding; a segment with an SUVmax of greater than or equal to 3 was documented in only 4 patients.
Evolution at 9 Months
As shown in Table 1, there was no significant difference between the baseline and the 9-mo control for all analyzed clinical and blood parameters.
There were significant enhancements at 9 mo for all CT parameters as opposed to none for 18F-FDG PET parameters (Table 1). In particular, the maximal diameter of the AAA was increased at 9 mo, with a mean difference from the baseline of 1.4 mm (SD, 1.9 mm). However, this difference was greater than or equal to 2.5 mm and considered to be significant in only 9 patients, who were placed in a group designated D+ (i.e., those with a significant [≥2.5 mm] increase in maximal diameter at 9 mo), whereas the remaining 30 patients were placed in a group designated D− (i.e., those without such an increase). This threshold of 2.5 mm corresponded to the upper limit of the 95% confidence interval of the reproducibility analysis.
Correlates of Growth Rates at Baseline and at 9 Months
The sole baseline imaging parameters showing significant differences between D+ and D− patients were SUVmax within the AAA, which was lower in D+ patients, and AAA volume, which was higher in D+ patients (Table 2).
Comparison of Imaging Parameters for Patients With (D+) and Patients Without (D−) Increase in Maximal Diameter at 9 Months
As shown in Table 2 and Figure 1, D+ patients exhibited lower SUVmax within the AAA at baseline than D− patients (1.80 ± 0.45 vs. 2.21 ± 0.52; P = 0.04); they also displayed a trend toward greater changes in SUVmax at 9 mo (difference between 9 mo and baseline: +0.40 ± 0.85 vs. −0.06 ± 0.57; P = 0.07). Similar levels were ultimately reached in both groups at 9 mo (2.20 ± 0.83 and 2.15 ± 0.66). The changes in SUVmax during the growth phases could be particularly striking, as shown in Figure 2.
Comparison of 9-mo evolution of 18F-FDG activity within AAA (SUVmax) between patients with (▪) and patients without (□) significant (≥2.5 mm) increase in maximal diameter during same period. *P < 0.05 for between-group comparisons.
Frontal and axial fused images from 18F-FDG PET and CT angiography showing marked enhancement in 18F-FDG uptake (arrows) between baseline (SUVmax, 1.9) and 9 mo (control; SUVmax, 4.2) after growth phases of AAA (47 mm for maximal diameter at baseline and 57 mm at 9 mo). Area with enhanced 18F-FDG uptake corresponds to thickening of presumably inflammatory tissue outside intimal calcifications on CT scan.
D+ patients also had a lower mean SUVmax within the suprarenal aorta than D− patients, but the difference was not significant (2.01 ± 0.51 vs. 2.41 ± 0.57; P = 0.11).
In addition, the difference in maximal diameter between 9 mo and baseline was slightly but significantly related to the baseline SUVmax of the AAA (P = 0.049) (Fig. 3). None of the other analyzed variables represented a significant predictor. However, trends toward greater differences in diameter were documented for patients who were still actively smoking at 9 mo (2.3 ± 2.7 mm for such patients vs. 1.2 ± 1.5 mm for other patients; P = 0.08) and for AAA with larger volumes at baseline (P = 0.07 for the analysis of the correlation between differences in diameter and AAA volume).
Relationship between difference (Δ) in maximal AAA diameter between baseline and 9 mo (control) and 18F-FDG activity within AAA (SUVmax) at baseline.
Finally, none of the baseline imaging parameters was predictive of the difference in AAA volume between baseline and 9 mo.
DISCUSSION
One of the key questions underlying the present study was whether 18F-FDG PET has the potential to aid in risk stratification for AAA. Indeed, on the one hand, increased 18F-FDG uptake within the AAA wall is correlated with macrophage infiltration and hyperproteolysis in situ (6–9). Furthermore, parietal 18F-FDG foci can be observed when AAA are prone to rupture and, more precisely, when they are large, rapidly expanding, or symptomatic (11). Such foci have yet to be documented at the actual site of rupture (9). However, on the other hand, focal or diffuse increases in 18F-FDG uptake are likely to be a rare finding in patients with asymptomatic AAA, even when the maximal diameter is close to surgical indications (12,20). Surprisingly, an inverse correlation between 18F-FDG uptake and the subsequent rates of growth of asymptomatic AAA was recently documented (14).
These findings are further confirmed by the results of the present 18F-FDG PET sequential observational study of patients with asymptomatic AAA considered to be below the threshold for surgical repair (46 ± 3 mm for baseline maximal diameter). First, the mean level of 18F-FDG uptake, as determined with the SUVmax, was significantly lower than that documented in the suprarenal nondilated portion of the aorta, in accordance with previous observations (12,20). Second, an even lower level of 18F-FDG uptake was documented at baseline in the AAA of the 9 patients for whom CT angiography yielded evidence of a subsequent increase in maximal diameter at 9 mo (control). Third, an inverse relationship was observed between baseline 18F-FDG uptake and the absolute change in the maximal diameter at 9 mo (Fig. 3), in accordance with a previous observation (14).
Recent histopathologic studies showed that the low level of 18F-FDG uptake of asymptomatic AAA reflects the metabolic activity of a small number of cells present within arterial walls (13), especially in fibrotic and necrotic regions (8). It is noteworthy that these cells include not only inflammatory cells but also smooth muscle cells exhibiting a sustained increase in glucose uptake, especially when stimulated and chronically exposed to inflammatory cytokines, a common situation within atherosclerotic walls (21,22).
Thus, in the atherosclerotic walls of most asymptomatic AAA, the overall number of cells able to entrap 18F-FDG is considerably reduced, leading to a global low level of 18F-FDG uptake (8,13,20). These data imply that, in such cases, chronic inflammation would not be sufficiently active to result in increased glucose metabolism detectable by PET cameras (20). From a technical imaging standpoint, such detection is also likely to be further impaired by partial-volume effects related to the thinness of arterial walls (7).
In the present study, as well as in the study by Kotze et al. (14), the level of 18F-FDG uptake was found to be even lower in AAA that were more likely to exhibit a subsequent enhancement in maximal diameter. This inverse relationship was lost when 18F-FDG uptake was assessed through the tissue-to-background ratio, similar to what was observed by Kotze et al. (14), presumably because this method of correcting for background activity is not well adapted to this setting. High and variable proportions of AAA walls are indeed covered by thrombus, in which activity is dramatically lower than in circulating blood. The activity documented in such thrombus-covered walls is likely to be minimally influenced by the level of 18F-FDG in blood.
Figure 3 shows that a clear increase in maximal diameter was a rare finding in our study population, even for AAA exhibiting a low level of 18F-FDG uptake at baseline. However, it must be kept in mind that AAA growth is likely to involve not only parietal but also important blood determinants. In particular, the thrombus of AAA is metabolically active, with variable production of cytokines and proteases that may influence the metabolism and structure of arterial walls (23).
On average, growth phases were found to be preceded by a stage of low metabolism, with a low level of 18F-FDG uptake, but the main original observation of the present sequential study was the reversible nature of this stage. A trend toward greater changes in 18F-FDG uptake was indeed documented in patients with an increased maximal diameter at 9 mo (Fig. 1); the uptake levels seen in the other patients at 9 mo (control) were ultimately reached.
These cyclic changes in 18F-FDG uptake are in keeping with the cyclic inflammatory process of atherosclerosis development and with the transient nature of 18F-FDG uptake within atherosclerotic lesions (24,25). These changes are also in accordance with the current model of AAA progression, which involves repetitive sequences of enlargement and inflammation (26,27). These cyclic changes are best illustrated in Figure 2, which shows a dramatic enhancement in 18F-FDG uptake and in wall thickness after a growth phase, an observation similar to that previously documented in a case report (9). Such an enhancement in 18F-FDG uptake may relate to the recruitment of additional inflammatory cells but also to the inflammation-related activation of smooth muscle cells (21,22).
An additional observation was that patients with an increased diameter at 9 mo showed a trend toward a lower level of 18F-FDG uptake within the suprarenal aorta at baseline. Even if the difference between these patients and the other patients was not significant, this observation leads us to wonder to what extent these metabolic changes are diffuse rather than strictly located within the AAA walls.
Taken together, these observations led us to consider that enhanced 18F-FDG uptake, in relation to an enhanced inflammatory reaction, does not precede the growth phases of uncomplicated AAA, and we wonder whether enhanced 18F-FDG uptake may accompany the growth phases. Only a trend toward enhanced 18F-FDG uptake was documented here, together with an increase in the maximal diameter. However, our 9-mo delay was likely too long to image all inflammatory phases. On serial PET imaging from cancer patients, arterial 18F-FDG foci were indeed found to normalize at a shorter and highly variable delay (6.8 ± 4.3 mo (25)). However, further studies with larger populations and higher imaging rates during follow-up are needed to clarify the relationship between growth phases and cyclic changes in 18F-FDG uptake.
CONCLUSION
The present study showed that the enhancement in the maximal diameter of small AAA was preceded by a stage with a low level of 18F-FDG uptake, but this low level of uptake was no longer documented after the growth phases, suggesting a pattern of cyclic metabolic changes. At the level of individual subjects, 18F-FDG PET was not an effective tool for predicting the growth phases of small and asymptomatic AAA.
DISCLOSURE
The costs of publication of this article were defrayed in part by the payment of page charges. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734. This work was funded by a grant from the French Health Ministry (Programme Hospitalier de Recherche Clinique Inter-régional). The Nancyclotep Experimental Imaging Platform provided financial and organizational support. No other potential conflict of interest relevant to this article was reported.
Acknowledgments
We thank Pierre Pothier for critical review of the article.
Footnotes
Published online Mar. 19, 2015.
- © 2015 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication July 30, 2014.
- Accepted for publication January 5, 2015.
Jump to section
Related Articles
Cited By...
- Aortic sodium [18F]fluoride uptake following endovascular aneurysm repair
- Aortic sodium [18F]fluoride uptake following endovascular aneurysm repair
- Systematic Review of Circulating, Biomechanical, and Genetic Markers for the Prediction of Abdominal Aortic Aneurysm Growth and Rupture
- Novel Molecular Imaging Approaches to Abdominal Aortic Aneurysm Risk Stratification