Article Text
PDF
Abstract
Objective: A substantial number of patients with acute aortic syndrome (AAS) require invasive therapy because of disease progression. Our study aimed to assess the impact of positron emission tomography (PET)/computed tomography (CT) and serological markers of inflammation to identify patients at high risk for disease progression.
Methods: 33 patients with AAS (thoracic aortic aneurysm 5, thoracic aortic dissection 14, penetrating aortic ulcer 8, intramural haematoma 6) were included. After intravenous administration of [18F] fluorodeoxyglucose a non-contrast enhanced PET/CT of the body trunk and CT angiography of the entire aorta was performed. Serological levels of D-dimers and C-reactive protein (CRP) were measured in all patients. Follow-up imaging was performed to detect disease progression.
Results: 11 (33%) of 33 patients showed elevated tracer uptake within the aortic pathology, whereas 22 patients were PET-negative. In 23 patients a CRP level exceeding 1.0 mg/dl or a D-dimer level larger than 250 μg/l was found. The follow-up time was 224 (195) days. Nine of 11 PET-positive patients (82%) showed progression of AAS. In contrast, 55% of PET-negative patients showed stable disease or regression during the follow-up period. Kaplan-Meier analysis showed a clear, but not yet significant trend to longer survival in PET-negative patients, whereas elevated CRP and D-dimers did not allow for distinguishing of high-risk patients.
Conclusions: Vessel wall inflammation was found in one-third of the patients with AAS and this patient group seems to have a high risk for disease progression. These initial results needs further investigation.
Statistics from Altmetric.com
The imaging procedure 18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) has shown excellent results in the diagnosis of various diseases, based on the detection of increased glucose metabolism. In oncology there are several clear indications for FDG-PET, including detection of malignant tumours, tumour staging and therapy control.1 In patients with inflammatory vascular diseases FDG-PET can be helpful for monitoring therapy response.2–4 For atherosclerotic abdominal aortic aneurysms, a strong correlation was found between increased FDG-PET uptake and the progression of aneurysm size.5 Furthermore, there is a correlation between the stage of plaque development and the level of glucose metabolism within atherosclerotic plaques.6 Advanced plaques are more likely to show increased tracer uptake, as a result of the accumulation of active macrophages.7
FDG-PET alone is limited by the low spatial resolution and a lack of morphological information. This can be overcome with the dual-modality PET/computed tomography (CT), which has emerged as a new diagnostic technique for vascular applications. Among those, acute aortic syndrome (AAS) belongs to a group of important clinical findings, because of high mortality rate.8 In addition, acute aortic syndromes are frequently associated with elevated markers of inflammation like D-dimers or C-reactive protein (CRP).9 10 These elevated serological markers can be caused by the underlying disease or by the acute aortic pathology itself.11 12 Whereas morphological imaging with contrast-enhanced CT represents the diagnostic modality of choice for patients with suspected AAS, the value of CT to identify the extent of inflammation in AAS is limited.
Our study aimed to assess the correlation of dual-modality PET/CT imaging and serological markers of inflammation with the clinical course in patients with AAS.
MATERIAL AND METHODS
Patients
All patients with suspicion of AAS or known aortic pathologies of unknown extension referred to our department between January 2003 and December 2005 were included in this prospective study. Written informed consent was obtained from all patients and the study was performed in full accordance with guidelines issued by the institutional review board.
PET/CT imaging
Dual-modality PET/CT was performed using the Biograph system (Siemens Medical Solutions, Hoffman Estates, IL, USA). The system acquires the CT first, followed by PET.13
CT images were acquired with 130 mA, 130 kV, section width of 5 mm and table feed of 8 mm per rotation. Whole-body spiral CT was performed starting at the upper thigh and scanning in the caudocranial direction up to the neck. This CT scan was used for the attenuation correction of the PET data. It was performed without intravenous contrast to avoid hardening artefacts and to identify hyperdense intramural haematomas. A limited breath-hold technique was used to avoid motion-induced artefacts.14
Patients had been instructed to fast for a minimum of 6 hours before tracer injection. In addition, blood samples were collected immediately before the injection of the radioactive tracer to ensure blood glucose levels to be within the normal range (up to 130 mg/dl).
The PET system has a field of view of 15.5 cm per bed position and an in-plane spatial resolution of 4.6 mm. PET images were acquired 60 minutes after administration of 350 MBq FDG. The acquisition time of PET was adapted according to the patients’ weight, using 3 minutes per bed position for patients up to 65 kg, 4 minutes per bed position for patients ranging between 65 and 85 kg and 5 minutes per bed position for patients above 85 kg.
After the PET scan an additional contrast enhanced CT scan covering the chest and abdomen was obtained to identify the vascular pathology within the thoracic aorta and its extension to either supraaortic vessels or infrarenal aorta and iliac arteries. A volume of 120 ml of iodinated contrast (Xenetix 300, Guerbet, Sulzbach, Germany) was administered using an automated injector (Liebel Flarsheim; Medical Supplies, Kerpen, Germany) for a special contrast protocol.15 The CT scan was performed with a fixed delay of 50 seconds following the start of the contrast injection.
Image evaluation
The CT datasets included non-contrast enhanced as well as contrast-enhanced CT scans and additional multiplanar reformats. These data were evaluated by two radiologists experienced in cardiovascular imaging in consensus to identify the true extend of a given aortic pathology.
Thereafter, two different readers (one radiologist, one nuclear physician) evaluated the fused PET/CT data in consensus. The datasets read by this second reader group included the non-contrast-enhanced CT data essentially used for the attenuation correction and the PET data only. The focus of this data evaluation was the diagnosis of a possible focal elevated FDG uptake in the aortic wall. The two reader groups had the same clinical information but they were unaware of the results of the other imaging modality.
Morphological definitions
On the CT datasets, the assessment of an aortic pathology was based on the detection of typical morphological features.16 17
The diagnostic criteria of an intramural haematoma (IMH) included semicircular or circular aortic wall thickening without intimal disruption exceeding 5 mm, and hyperdense areas in the aortic wall in the non-contrast-enhanced CT scan.
Penetrating aortic ulcer (PAU) was defined as a crater-like ulceration in the aortic wall with or without adjacent subintimal haematoma. The diagnosis of an aortic dissection (AD) was made in case of an intimal flap separating true and the false lumen that showed contrast enhancement. Thoracic aortic aneurysm (TAA) was defined as an aortic diameter exceeding 4 cm on multiplanar reconstructions perpendicular to the vessel.
Metabolic definitions
On fused PET/CT images an aortic wall pathology was considered PET-positive if its glucose uptake was found to be above levels of the surrounding tissue on qualitative analysis. For quantitative analysis the maximum standardised uptake value (SUVmax, correlated to body weight) was determined by placing a region of interest (ROI) over the complete vessel circumference and measuring the highest uptake inside (calculation of SUV as described by Antoch et al15). A SUVmax of 2.5 was used as threshold to distinguish physiological and pathological FDG uptake.3
Serological markers of inflammation
In all patients the serum levels of D-dimers and C-reactive protein CRP were measured within 24 hours before or after the PET/CT examination. The threshold for pathological D-dimer levels was set to 250 μg/l, the threshold for a pathological elevated CRP was defined as 1.0 mg/dl. In addition a higher CRP-threshold of 6.3 mg/dl was registered according to published data.18
Statistical analysis
Categorical variables are presented as frequencies. All analyses were performed using the SAS software package. For the comparison of different patient subgroups during follow-up a combined end point was defined by summarising:
A measurable progression of a given aortic pathology on imaging or escalation of initial therapy (for example, conversion from medical treatment to either endovascular or open aortic repair), or
Death related to the aortic pathology.
A measurable progression was defined by an enlargement of an ulcerated plaque or a thickening of IMH of more than 25% to the size measured in the PET/CT examination.
The decision to invasive treatment was based on the following findings:
Progression of a measured aortic pathology in follow-up imaging,
Clinical progression—for example, recurrent chest pain under antihypertensive medication,
Necessity to escalate the initial medical treatment,
Clinical findings necessitating immediate treatment—for example, malperfusion of lower limbs or mesenteric structures.
The Kaplan-Meier method was used to compare the incidence of the combined end point in patient subgroups with either PET-positive or PET-negative aortic pathologies as well as in the subgroups of patients with an elevated or normal CRP levels or D-dimer levels. The log-rank test was used to define differences in survival. For the comparison of the serological markers with either PET-positive or PET-negative aortic pathologies the Fisher’s exact test was used.
Follow-up
The follow-up was based on all available data including all clinical data, physical examinations, radiological procedures (CT, PET/CT, MRI) as well as laboratory tests.
RESULTS
Patients
Between January 2003 and December 2005, 124 patients (91 male and 33 female, age range 35–86 years, mean age 66 years) presented with acute aortic syndrome. Fifteen (12%) patients died within 24 hours after admission; 40 (32%) patients received immediate surgery; 16 (13%) patients underwent immediate endovascular repair procedures; 16 patients (13%) were clinically unstable and a PET/CT examination could not be performed.
FDG-PET/CT could be successfully performed in 37 patients (30%) within 24 hours after admission. Four of these 37 patients dropped out during follow-up, resulting in a study group of 33 patients (28 male and 5 female, age range 46–85 years, mean age 67 years). The mean time of follow-up was 224 (SD 195) days (range 14–821 days).
Imaging
CT-based diagnosis
The combined non-enhanced and intravenous contrast-enhanced CT imaging revealed type B AD in 14 (43%) of 33 patients, a penetrating aortic ulcer in eight (24%) of 33 patients and an intramural haematoma in six (18%) of 33 patients. Five (15%) of the 33 patients had a TAA. The patient characteristics, measured data, diagnosed types of AAS as well as risk factors are summarised in table 1.
PET/CT imaging
The FDG-PET identified an elevated glucose metabolism within the aortic wall lesion in 11 (33%) of the 33 patients. The FDG-uptake measured as SUVmax showed a mean value of 4.5 (SD 0.80, range 3.3–6.2).
Four (37%) of these 11 patients were diagnosed to have an AD-B, and three (27%) of 11 to have a TAA, whereas in each of the groups of IMH and PAU, two patients showed elevated glucose uptake in the aortic wall lesion (fig 1).
Serological markers of inflammation
Levels of CRP exceeding 1.0 mg/dl (mean value 6.72 mg/dl, maximum level 19.7 mg/dl) was found in 23 (70%) of the 33 patients. Also, 23 (70%) of the 33 patients showed a pathologically elevated D-dimer with a mean value of 621 μg/l and a maximum of 2212 μg/l. In 13 (39%) of the 33 patients the CRP level was larger than 6.3 mg/dl.
There were no significant differences in D-dimers or CRP levels between PET-positive or PET-negative patients. The mean value of D-dimers was 877 μg/l and 573 μg/l, respectively. The mean value of CRP was calculated with 10.9 mg/dl in the PET-positive group and 5.1 mg/dl in the PET-negative group.
Correlation with clinical outcome
During the clinical course, 19 (58%) of the 33 patients showed a progression of the initial aortic wall lesion or required an escalation of initial therapy. Eleven patients (33%) remained stable and showed no progression of the initial lesion. A regression of elevated glucose utilisation or thrombosis of a local type-B dissection under antihypertensive therapy was found in three patients.
Twenty-eight (88%) of the 33 patients survived acute aortic pathology during follow-up. Three of the five deceased patients had pathological glucose uptake in the initial aortic wall lesion. Two patients died from progression of the aortic pathology, in two patients death was related to surgical complications. The last patient died from cerebral haemorrhage awaiting stent-graft placement.
When comparing clinical outcome and FDG uptake, nine (82%) of the 11 patients with pathological glucose metabolism received definite treatments, or showed a progression of the wall pathology under conservative therapy. Three of these nine patients died. However, only 10 of the 22 patients with normal glucose metabolism in the aortic wall (45%) were treated with replacement surgery, interventional stent-graft implantation or developed a progression of the underlying aortic pathology under antihypertensive drug therapy. The correlation of pathological FDG-uptake and clinical outcome is summarised in table 2.
The comparison of both PET subgroups by using the combined end point shows a difference.
In the group of PET-positive patients, nine (82%) of 11 patients experienced progression, or received definite treatment or died. In contrast, this could be observed only in 10 (45%) of 22 patients in the PET-negative group. Nevertheless the cumulative survival numbers show no statistical difference using the log-rank test (p>0.121) (fig 2).
The comparisons of the patient subgroups with pathological or normal CRP levels as well as with pathological or normal D-dimer levels show no significant differences in survival either (p>0.87 and p>0.24).
The comparison of cumulative survival by combining the results of PET and serological markers showed the best results for the combination of PET and D-dimers with a p value of 0.077 in the log-rank test (fig 3). The combination of PET and elevated CRP as well the combination of PET, elevated D-dimers and CRP were less distinctive (p values 0.48 and 0.46, respectively).
DISCUSSION
The combination of morphological and metabolic information provided by PET/CT imaging seems to be helpful in identifying patients with AAS with an increased risk for disease progression. Whereas the majority of patients showed pathological elevated D-dimers and CRP, only a third of all patients had pathological glucose metabolism in the aortic wall lesion. In addition, the single parameter of FDG uptake presented the clearest trend when testing for survival benefit. Even if these preliminary results are not supported by statistical significances, this is to the authors’ knowledge the first paper examining the value of PET/CT in patients with acute aortic syndrome.
The acute vessel wall lesions summarised as acute aortic syndrome represent a group of clinical findings with high mortality, but the outcome differs within subgroups. Acute aortic dissection in the ascending aorta shows a mortality of 1–2% per hour after onset of symptoms and is a surgical emergency.19 Medical treatment alone results in a 30-day mortality rate of 50%. The uncomplicated acute type-B dissection has a 30-days-mortality rate of 10%–30% and the intramural haematoma can develop a mortality rate of 20% at 30 days.20 Given the more focal nature of vessel wall lesion in PAU und IMH, both subtypes of AAS were revealed to show a more malignant nature with a higher tendency to aortic rupture or dissection.21 22 Depending on the localisation PAU and IMH should be treated differently. Focal lesions in the ascending aorta require immediate surgical therapy, but a focal wall lesion in the aortic arch or descending aorta can initially be handled conservatively with an intensive medical treatment of hypertension. In these cases a thorough follow-up imaging is essential.23 The poor prognosis of these vessel pathologies makes the correct and timely decision for a case and stage-adapted therapy mandatory.
The rate of PET-positive findings in the aorta was somewhat higher in our patient population than in a normal population when examining consecutive cancer patients, at 7%, but the results vary remarkably in different studies.24 This could be explained by the mean age of the study population, since it is known that the likelihood of FDG uptake rises with patient age.25 26 Another explanation has to be seen in the selection bias in our study owing to the high dropout rate from the initially included patient group.
The overwhelming majority of acute aortic wall lesions are based on arteriosclerotic changes inside the vessel wall. Therefore the diagnosis of inflammatory changes in the manifestation of atherosclerotic plaques may add important information to the evaluation of aortic pathologies. The uptake of FDG in atherosclerotic lesions has been proved to correlate with the number of macrophages in the plaque and may be able to characterise vulnerable plaques. In addition, symptomatic unstable plaques show higher glucose metabolism than asymptomatic or stable lesions.6 7 Aortic wall lesions occur more often in patients with advanced atherosclerosis, and patients with penetrating ulcers have a poorer prognosis compared to patients with classic AD.27 Our finding of a poorer outcome of PET-positive patients compared to the PET-negative group supported our hypothesis.
Inflammation can be seen as an independent marker for a poor prognosis of aortic aneurysm and dissection and CRP levels higher than 6.3 mg/dl indicate a high risk for short-term mortality.18 This correlates with our finding that the subgroup of 13 patients with a CRP level >6.3 mg/dl had a poorer prognosis in survival analysis than the group of 25 patients with a CRP level >1.0 mg/dl.
The finding that D-dimers have a better correlation with outcome than CRP was established by Ohlmann and colleagues, but their patient population showed a remarkably higher D-dimer level with a median of 8610 ng/ml ( = μg/l) compared to the 621 μg/l in our group.10
In a paper by Sbarouni and colleagues the level of D-dimers was significantly higher in a group of patients with acute aortic dissection compared to chronic aortic aneurysms or controls.28 Also they reported a much higher average level of D-dimers in the AD group with 4630 (3010) ng/ml compared to our data. The conclusion in this paper is that D-dimer could be used as a “rule-out” test in patients with suspected acute aortic dissection (AAD) and seems useful in the discrimination between AAD and chronic uncomplicated aneurysms. All in all there is growing evidence that elevated D-dimers may have diagnostic and prognostic value in patients with acute aortic dissection. Therefore a possible combination of PET results and D-dimer levels seems promising, since this combination showed the best results when testing for differences in survival. Taking into account the time and effort needed to perform a PET/CT in patients with AAS, the simple laboratory measurement of D-dimers has a clear benefit. This again underlines the need for a (multicentre) study to compare the value of PET/CT and D-dimers .
Our study has clear limitations. The small number of patients is the most important limitation of our study. Predominantly this is due to the high rate of patients (70%) who dropped out of our initial study group because of death, immediate endovascular or surgical therapy or general instability.
Since AAS summarises several subgroups with very different prognoses, the limited number of patients in the different subgroups weakens the power of the chosen statistical tests. Therefore the focus of the study was set solely on the correlation between the detection of pathological glucose metabolism in the aortic wall and the outcome. Nevertheless, we are aware that this approach simplifies the very complex clinical situation of patients with AAS.
The technical features of the PET scan require an in-room time of 30–45 minutes for the patient undergoing the PET/CT examination. This must result in a preselection bias, since severely ill patients—that is, especially unstable patients or patients in an emergency setting requiring immediate therapy, had to be excluded. In future the use of more sensitive PET detectors will decrease the time per bed position down to 1 minute and will result in in-room times of 15 minutes or less.
The follow-up period of 7 months is rather short, but we think that even this short time frame is sufficient to detect the subacute changes in PET-positive aortic processes.
The last point to mention is the limited capability of our CT scanner. We used a dual-slice CT with a limited heat capacity in the x-ray tube compared to more advanced (16-detector, 40-detector or 64-detector rows) CT systems. This resulted in longer imaging sessions to scan the body trunk, upwards of more than 70 seconds. Studies of contrast dynamics in a dissection or gated-CT studies require more sophisticated systems. In future combined PET/CT imaging using CT-installations with 64 detector-rows will enable not only examination of patients with AAS using ECG or breath-gating, but will play an important part in combined non-invasive coronary and myocardial imaging of patients with acute coronary syndrome.29
CONCLUSION
Combined morphological and metabolic imaging with PET/CT is feasible in clinically stable patients with acute aortic syndrome. Thirty-three per cent of the examined patients showed elevated glucose metabolism in aortic wall lesions. These patients were at higher risk for disease progression or death during follow-up. The correlation between pathological FDG uptake and outcome was stronger than between any of the serological markers and outcome, but did not reach a significance level. Thus, these preliminary results need further evaluation in larger patient groups.
REFERENCES
Footnotes
Competing interests: None.