Abdominal aortic aneurysm, diagnosis, biomarkers
List of abbreviations:
AA: aortic aneurysm
AD: aortic dissection
AAA: abdominal aortic aneurysm
APC: activated protein C
HsCr: high-sensitivity C-reactive protein
IL-1: interleukin -1 (-4, -5, -6, -10)
IFN-gamma: interferon gamma
MMP: matrix metalloproteinase
MIF: macrophage migrating inhibitory factor
PAI-1: plasminogen activator inhibitor type 1
PIIInP: aminoterminal propeptide of type III procollagen
PICP: carboxyterminal propeptide of type I procollagen
PF-1 -2: prothrombin fragment-1 -2
SEP: serum elastin peptides
TIMP: tissue inhibitor of metalloproteinase
TPA: tissue plasminogen activator
TNF: tumour necrosis factor
TGF-1-beta: transforming growth factor-beta
Rationale and drawbacks of the use of biochemical markers in AAA
In order to improve the assessment of patients with suspected AAA, a number of predictive models and new diagnostic screening tools have been developed. The predictive models include utilisation of various clinical variables, while the new diagnostic screening tools utilise several serologic biochemical markers (defined as measurable proteins, peptides, genes, or metabolic products that represent biologic processes in an organism) with a view to assessing the biological activity of the aortic wall for prediction of AAA progression and risk of rupture.
Although rupture risk increases with increasing aortic diameter, the natural history of AAA has been shown to differ substantially between individual patients. Thus, identification of patients at high risk of rapid AAA expansion and rupture is essential to provide targeted surveillance. Moreover, after surgical and/or percutaneous exclusion, the reappearance of specific biochemical markers (such as continuous aortic dilatation or the presence of endoleaks following AAA exclusion), suggesting a biological reactivation of aortic disease, may allow follow-up of the magnitude of the progression.
Biochemical diagnosis may be inexpensive, readily available, and non-invasive, making it ideal for use in rapid diagnosis. However, elevation of non-specific inflammatory markers such as high-sensitivity C-reactive protein, homocysteine, lipoprotein-a, and different inflammatory cytokines such as interleukin-2, - 3 and -6, may reflect the unspecific atherosclerotic burden characteristic of patients affected by AAA but do not specifically tackle the problem of identification of abdominal aortic conditions.
It is worth noting that identification of AAA biomarkers is complicated by the fact that AAA is a multifactorial disease with a complex pathophysiology. Whenever biomarkers are found to be associated with a given disease, the question usually arises whether these markers are just innocent bystanders or guilty offenders. Thus, we believe it is unlikely that a single biomarker will be found to predict AAA progression; on the other hand, a multi-marker or pattern approach should be able to do so . These biomarkers can perhaps be incorporated into a clinical scoring system taking into account other major cardiovascular risk factors such as old age, male sex, smoking status, blood pressure and dyslipidaemia, etc., in the case of AAA.
In this review, for practical purposes, we will focus only on circulating biomarkers for AAA development, progression, and rupture in addition to markers which may predict the stabilisation of AAA after surgical or endovascular repair (Figure 1). Unspecific inflammatory markers, suggestive more of the patient’s atherosclerotic burden, are not discussed but are reported and referenced in Table 1.
Figure 1. Schematic drawing of different serologic biomarkers related to AAA development, progression, rupture and stabilisation after surgical or endovascular graft exclusion.
Table 1. AAA status and corresponding biomarker elevation* that can be potentially utilised for diagnosis and follow-up.
|Biomarker||Development||Progression||Symptomatic ruptured AAA||Symptomatic non-ruptured AAA||After repair||Correlation|
|MMPs/TIMPs [2,3,4]||Yes ↑||Yes ↑||No||No||Yes ↑||
|Cathepsin/Cystatin C ||Yes ↓||Yes ↓||No||No||No||
|PAI activity ||No||Yes ↑||No||No||No||
|Fibrinogen ||Yes ↑||No||No||Yes ↑||No||
|Tissue plasminogen activator ||Yes ↑||No||No||Yes ↑||No||
|PIIInP PICP ||No||Yes ↑||No||No||No||
|D-Dimer [9)||Yes ↑||Yes ↑||Yes ↑||Yes ↓||Yes ↑||
|Bystander atherosclerotic biomarkers unspecific for AAA|
|High-sensitivity C-reactive protein ||Yes ↑||No||No||No||Yes ↑||
|TNF ||Yes ↑||No||No||No||Yes ↑||
|IFN-gamma ||Yes ↑↓||Yes ↑||
|TGF-1 beta ||Yes ↓||No||No||No||No||
|IL-1 ||Yes ↑||Yes ↑||No||No||No||
|IL-4 ||Yes ↑||Yes ↑||No||No||No||
|IL-5 ||Yes ↑||Yes ↑||No||No||No||
|IL-6 ||Yes ↑||Yes ↑ ↓||No||No||No||
|IL-10 ||Yes ↑||Yes ↑||No||No||No||
|Homocysteine ||Yes ↑||Yes ↑||No||No||No||
|Lipoprotein(a) ||No||No||No||No||Yes ↑||
|NT-proBNP ||Yes ↑||Yes ↑||No||No||No||
|MIF ||Yes ↑||Yes ↑||No||No||No||
|Osteoprotegerin ||Yes ↑||Yes ↑||No||No||No||
|Osteopontin ||Yes ↑||No ↑ ↓||No||No||No||
|Cotinine ||Yes ↑||No||No||No||No||
*Serum levels of biomarkers are considered elevated/decreased when compared with normal values of age and sex-matched healthy subjects. For abbreviations see the beginning of the article.
Markers of AAA development/progression/rupture
Matrix metalloproteinases (MMPs)/ tissue inhibitors of metalloproteinase (TIMPs)
It is now accepted that inappropriate remodelling or damage to the extracellular matrix is thought to contribute to the initiation and progression of AAA disease. Direct functions of MMPs include extracellular matrix degeneration, control of inflammation, and induction of apoptosis. Loss of elastin and collagen degradation lead to weakening, dilatation, and ultimately rupture of the aortic wall. Proteases active within the AAA wall include members of the MMP, cysteine, and serine protease families. The endogenous tissue inhibitors of MMPs (TIMP-1 to TIMP-4) regulate the activity of MMPs, the latter depending on the rate of synthesis, activation, and balance between active enzymes and inhibitors.
Patients with AAAs exhibit arterial dilation and altered matrix composition throughout the vasculature . Degradation occurs as a consequence of complex interactions between genetic factors, inflammatory cytokines, MMPs, and others. Phenotypically, there is dissolution and fragmentation of collagen and elastin, which leads to expansion of the vessel wall that can no longer withhold the repetitive expansible forces of systolic contraction. More specifically, with the study of MMPs, it has been shown that collagen neosynthesis may decrease with relatively increased collagen degradation in patients with AAA. As such, expression of MMP-9 and MMP-12 is increased in patients with AAA. The ability of some MMPs (MMP-2, -3, -9, and -12) to hydrolyse elastin is of great importance in terms of their effects on the vascular system and especially for the aortic wall. During the last few years, several studies have demonstrated an increase in the activities of various MMPs in atherosclerotic AAA. These observations have strongly correlated with immunohistochemical findings of increased immunoreactivity to these components. Furthermore, the use of MMPs seems to help to differentiate aneurysmal from stenotic aortic lesions, since the former show a different pattern of expression of MMP, as demonstrated by a marked selective increase in MMP-3. The genetic implications of finding increased MMP-3 in AAA are of particular interest. It has been shown in a Finnish population that a specific and more active allele (5A MMP-3) might be a genetic risk factor for AAA . Matrix metalloproteinase-2 (MMP-2) is the dominant elastase in small AAAs, and overexpression of MMP-2 in vascular smooth muscle cells may be a primary aetiological event in aneurysm genesis.
Elastase is the major catabolic enzyme for the breakdown of mature form elastin and its activity is inhibited by trypsin, which, in turn, is inhibited by α1-proteinase inhibitor (α1-antitrypsin [A1AT]). Patients with AAA have higher serum levels of A1AT than patients with aortic occlusive disease . Elastase stimulates degradation of extracellular matrix proteins, and subsequent inflammation of the aortic wall, suggesting a direct link with aortic dilation in AAA. However, the role of elastase–A1AT complexes in the development of AAA is ambiguous.
Elastin is a major component of the human abdominal aorta. It has been shown that, in AAA, the structure and the amount of elastin is changed, and increased levels of elastase and other elastolytic proteases in aneurysmal walls have been demonstrated. It is of note that increased elastin degradation also takes place in normal parts of the aorta in cases of AAA. With the study of the serum elastin peptides (SEP), plasma elastin alpha1 antitrypsin complex, and protocollagen III-N-terminal propeptide (all measured by enzyme-linked immunoassay and indicative of increased elastolysis), it has also been suggested that increased elastolysis is associated with increased AAA wall distensibility, whereas increased collagen turnover is associated with reduced distensibility . Previous studies have proposed that AAA distensibility may be an independent predictor of growth and rupture, possibly because it reflects changes in the structure and composition of the aortic wall. The Chichester Aneurysm Screening Group showed that serological elastin peptide levels were significantly higher in patients with large AAAs (>60 mm) that ultimately ruptured, as compared with those whose aneurysms did not rupture (p=0.041) . This finding indicates that loss of elastin, in addition to collagen degradation, predisposes an individual to AAA rupture. In addition, SEP, which are the end-products of elastolytic degradation, have been prospectively described as a strong predictor of expansion in small vessel AAA (sized 3-5 cm). Recently, it has been suggested that one sampling of SEP combined with AAA size in patients referred for AAA surgery may be a useful indicator of high risk of rupture, with good correlations (r=0.51) between SEP and AAA progression.
Fibrillar collagens give mechanical stability to tissue when axial forces are transmitted to the arterial wall. MMPs and cysteine protease families can degrade the highly protease-resistant structures of type I and III collagens. This increased collagen degradation associated with reduced collagen deposition has been associated with progression and rupture of AAAs by measuring collagen metabolism products.
In this context, aminoterminal propeptide of type III procollagen (PIIInP) reflects the synthesis of collagen type III, while carboxyterminal propeptide of type I procollagen (PICP) reflects the synthesis of type I collagen. The serum PIIInP levels in patients with AAA have been shown to be significantly higher (p>0.0001) than in patients with aortic occlusive disease . In addition, serological levels of PIIInP were correlated with aneurysm diameter and maximum thickness of intraluminal thrombus. PIIInP concentrations and AAA diameter correlation was found in a long-term study of 55 patients followed for a minimum of 2 years (r=0.55, p=0.002) . Conversely, since AAA is a disease of the medial layer, no studies have found a correlation between PICP and AAA development/progression or rupture.
The cysteine protease family consists of cathepsins, caspases, and calpains. Cathepsins K, l, and S have potent elastolytic activity, and similarly to MMPs release adhesion molecules from the endothelial cell surface. The presence of cathepsins S, K, and l has been demonstrated in AAA . Cystatin C has been found to correlate with AAA progression with a negative correlation with the expansion rate (r=0.24) in a single study .
Serine proteases: plasmin, tissue plasminogen activator, activated protein C
Plasmin directly degrades some components of the extracellular matrix and might also indirectly affect other extracellular matrix constituents through the activation of MMPs and latent pro-MMPs. Recently, it has been suggested that plasmin is a common activator of the known proteolytic systems involved in aneurysmal degradation and this marker has been reported to be associated with the expansion of AAA. It has been shown that aortic matrix degradation in AAA may be partly caused by activation of plasminogen by a tissue plasminogen activator .
Fibrinolytic activity and gene expression have been shown within the aortic wall of patients with asymptomatic AAA. A recent study has provided evidence that proteolytic pathways are involved in the aortic wall degradation in AAA by showing that acutely symptomatic non-ruptured AAAs are associated with increased systemic fibrinolysis and reduced thrombin generation as compared with ruptured AAAs. Furthermore, preoperative haemostatic markers, particularly plasma prothrombin fragment and plasminogen activator inhibitor (PAI) activity, may distinguish acutely ruptured from non-ruptured AAAs . Despite the importance of the above findings, one must take into account that the majority of the above-mentioned studies included only a small number of patients. Moreover, a rapid assay of PAI is not yet available.
Inhibition of systemic fibrinolysis, in patients with high levels of PAI-1, was significantly associated with AAA  in 438 patients versus 438 healthy subjects. In the same study, a significant association between larger abdominal aortic diameters and the number of thrombophilic parameters was also found (r=0.13, p=0.005).
Protein C is a vitamin-K-dependent serine protease that, when activated by thrombin bound to its endothelial cell cofactor (thrombomodulin), becomes activated protein C (APC), which is a potent physiological anticoagulant. Protein C inhibitor (PCI) is the major inhibitor of APC and it forms a complex (APC–PCI), which is a sensitive indicator of thrombin formation. Kölbel et al found that patients with AAA had a threefold higher median APC–PCI level than control individuals and, interestingly, patients with low levels of APC–PCI did not have a thrombus-lined aneurysm . Moreover, APC–PCI level was significantly higher in patients with AAA compared with controls, and APC-PCI was reported to correlate significantly with AAA diameter (r=0.22; p=0.001) but not with AAA growth rate (r=0.11, p=0.14) .
Endothelin is an endothelium-derived contracting factor which regulates vascular tone. Its release is stimulated by a variety of factors such as angiotensin II, vasopressin, adrenaline, hypoxia, and vascular injury. Some of these processes may well be implicated in the pathogenesis of aortic aneurysms. Endothelin 1 and 2 levels have been studied in patients with large or symptomatic AAAs. They were found to be significantly increased in patients with large as compared to small aneurysms, thus offering the potential of being an endogenous marker of aneurysm diameter . In addition, levels of plasma hepatocyte growth factor concentration, the member of the endothelium-specific growth factor family with the greatest mitogenic activity, have been proposed to predict the presence of atherosclerotic lesions from the abdominal aorta to the femoral arteries. The above findings could be especially helpful in evaluating small AAA, which currently are only detected with elaborate imaging. Further studies of alterations in endothelin levels in aneurysms being followed are needed to test these hypotheses.
Markers for AAA monitoring after exclusion
MMPs and TIMPs
Our group has shown that plasma MMP-9 and MMP-3 levels are significantly elevated in patients with AAA compared with healthy control subjects and that, after endovascular graft exclusion, these proteins decreased to a level similar to that of patients undergoing open surgical repair . Additionally, we showed that a lack of decrease in MMP levels, after endovascular graft exclusion, may help to identify patients who have an endoleak and consequent aneurysm expansion caused by ongoing sac pressurisation during follow-up. Conversely, a decrease in the amount of circulating MMPs could represent a simple marker of successful aneurysm exclusion.
Georgiadis and co-authors  investigated whether plasma and connective tissue MMPs and their inhibitors (TIMPs) may predict late high-pressure endoleaks after endovascular aortic aneurysm repair. In relation to this aim, baseline levels of MMP-2 and -9 and TIMP-1 and -2 were measured from blood and inguinal fascia specimens in 72 consecutive patients who were treated by EVAR (EndoVascular Aneurysm Repair) patients. Baseline plasma levels of MMP-2, MMP-9, TIMP-1, and TIMP-2 and baseline MMP-2 and MMP-9 activity estimated using gelatin zymography were compared between patients who developed a late endoleak during follow-up and those who did not. The authors also investigated whether MMPs and TIMPs and also MMP activity differed between patients with moderate (50-59 mm) and large (≥60 mm) diameter AAA at primary EVAR. The mean follow-up period was 63.1 months. ProMMP-9 concentrations were found to be higher in patients with high-pressure late endoleaks (n=13) than in patients without an endoleak (n=59) (p=0.03, SE 4.46 [95% CI: -19.653-1.087]). Patients with larger diameter AAAs had higher mean tissue homogenate levels of total MMP-9/total MMP-2, and higher proMMP-9/total MMP-9 levels, compared to those with moderate diameter AAAs (p=0.025, p=0.05, p=0.018, and p=0.021, respectively). Regression analysis revealed significant association between proMMP-9 and late endoleak (p=0.025, OR 1.055, 95% CI: 1.007-1.106). Furthermore, proMMP-9 and active MMP-9 were strong predictors of late endoleak in patients with large AAAs (p=0.018 and p=0.041, respectively).
The authors concluded that proMMP-9/active MMP-9 were strongly associated with late endoleak in patients with large AAAs, suggesting that continuous exposure to systemic circulation of AAA wall connective tissue environment, reflected by extracellular matrix dysregulation and altered MMP activity, may be coupled to AAA biology changes leading to late endoleak formation.
D-dimer is a typical degradation product of cross-linked fibrin. Increased levels of D-dimers have been found in blood samples of 74 AAA endograft patients and controls . These authors tested the hypothesis that levels of D-dimer could be a marker for incomplete aneurysm exclusion after AAA repair. Elevated levels of D-dimer served as a useful marker for fixation problems after endovascular AAA repair and helped to rule out type I endoleak, thus allowing some patients to avoid unnecessary invasive tests.
Protein profiling of cardiovascular associated proteins
Recently, in a case control study of 25,589 patients, Memon and co-authors  identified 415 patients with AAA. Of these, 134 patients underwent evaluation of plasma levels of 21 proteins associated with proteolysis, oxidative stress, lipid metabolism and inflammation. Patients were matched for comorbidities with 139 subjects with an aortic diameter <30 mm. Of all proteins examined, a combination of growth/differentiation factor-15 and cystatin-B had the best ability to discriminate AAA from non-AAA (ROC 0.76; sensitivity 80%, and specificity 52%). Myeloperoxidase showed the best prognostic value (area under the curve 0.71; sensitivity 80%, and specificity 59%), and higher baseline levels of myeloperoxidase were significantly associated with faster AAA growth compared with lower levels, independently of baseline diameter.
The increase in the burden of abdominal aortic disease makes future advances in serologic biomarkers and molecular imaging coupled with biological activity of the aortic wall extremely important  in order to detect AAA development, progression and risk of rupture. The decision-making process of earlier interventions in case of high biological activity of AAA in the setting of smaller aneurysm size, facilitating prompt and timely initiation of appropriate therapy, can be utilised in the near future. Furthermore, patients presenting for long-term assessment of surgically or endovascularly repaired AAA may benefit from detection of changes in serologic levels of biomarkers. The reappearance of a biomarker in the circulation may be coupled with an imaging modality to evaluate the status of the aorta better.
Although several cytokines, D-dimer and hepatocyte growth factor offer potential to be predictive of clinical disease, the laboratory tests to assess AAA are still limited to those that are employable in clinical settings, have commercially available assays that can be standardised, and have adequate precision. On the basis of these considerations, at this time, it is most reasonable to limit current assays to D-dimers, serine proteases and MMPs until more data from ongoing large prospective studies are reported.