Diagnostic challenges in aortic disease
During cardiac surgery, manipulation of the thoracic aorta can cause the release of emboli into the systemic circulation. Aortic atherosclerosis has indeed been associated with post-operative stroke and other embolic complications. Atherosclerosis of the distal ascending aorta has the largest risk for post-operative stroke, because surgical procedures including aortic cannulation and placement of a cross-clamp are usually performed on this part of the aorta. Accurate identification of aortic plaques can guide the surgical management to avoid dislodgement of plaques.
Multiple diagnostic tests are available to visualize atherosclerosis of the distal ascending aorta; these modalities were reviewed in Chapter 2. Computed tomography (CT) has a low sensitivity and negative predictive value, and is thus not very useful for the focused screening for aortic atherosclerosis. Magnetic resonance imaging (MRI) is an accurate test with anatomical and functional information, but is infrequently used. More importantly, both tests lack the possibility of real-time intraoperative visualization. Epiaortic ultrasound (EAU) is considered the gold standard for the intraoperative visualization of atherosclerosis of the distal ascending aorta. Its use has been found to lead to changes in the surgical management, but an improvement of patient outcomes has not yet been established. Manual palpation is frequently applied, but is known to underestimate the atherosclerotic burden and may itself cause the release of emboli. Transesophageal echocardiography (TEE) is used most often for the focused screening for aortic atherosclerosis, as it has the advantages of visualizing the aorta before sternotomy and to provide real-time imaging during surgery. A major limitation however is the poor sensitivity for atherosclerosis of the distal part of the ascending aorta due to the air-filled trachea. A possible solution for this limitation is offered with modified TEE, in which a fluid-filled balloon is positioned in the trachea and left main bronchus thereby allowing for visualization of the distal ascending aorta and its branches is possible. Compared to EAU, in previous studies modified TEE was found to have a good overall diagnostic accuracy (area under curve: 0.89), with a high sensitivity and negative predictive value (95% and 97% respectively) but limited specificity and positive predictive value (79% and 67% respectively). However, good diagnostic accuracy by itself is not going to improve patient outcomes. This thesis consists of studies investigating next steps, by putting modified TEE to use in practice, to improve patient outcomes. The aim of modified TEE screening is to improve patient outcomes through a more accurate diagnosis of atherosclerosis of the distal ascending aorta. Therefore, modified TEE should provide information that is not already captured by other tests. In Chapter 3 the added value of modified TEE was studied, compared to patient characteristics and conventional TEE imaging. Modified TEE indeed improved the diagnosis of atherosclerosis of the distal ascending aorta beyond conventional TEE, as indicated by the “net reclassification improvement” and “integrated discrimination improvement”.
In Chapter 4 we studied the prognostic implications of atherosclerosis of the distal ascending aorta (DAA) diagnosed with modified TEE. For this aim we studied the follow-up of patients included in a diagnostic accuracy study, in which the results of modified TEE were blinded during surgery. Atherosclerosis of the DAA was associated with increased one-year and long-term all-cause mortality (hazard ratio [HR]1.82, 95% CI: 1.10 – 3.24). The predictive performance was superior to the results of conventional TEE, but inferior to the prognosis of one-year mortality with the original EuroSCORE. Also, there was no association with short-term mortality or stroke, possibly related to the limited number of patients with these end-points. These results indicate that the findings of modified TEE are of relevance for the prognosis of patients, and that this information is not already available using conventional TEE.
The ultimate goal of modified TEE is however not to provide a new invasive tool to predict mortality in the coming years, but to prevent intra-operative complications during surgery. The aim of Chapter 5 was to study if the use of modified TEE was indeed associated with improved patient outcomes. This non-randomized observational study compared 1,391 patients who underwent perioperative modified TEE In this study, modified TEE was used on indication, i.e. in patients with a higher probability of atherosclerosis based on clinical characteristics and conventional TEE imaging. The EuroSCORE predicted operative mortality was indeed higher in the intervention- compared to the control group (5.9% vs. 4.0%, p<0.001); the observed 30-day mortality was similar in both groups (2.2% vs. 2.5%, p = 0.55). The use of modified TEE was associated with a lower mortality after multivariable- (relative risk [RR] 0.70, 95% CI: 0.50 – 1.00) and propensity score adjustment (RR 0.67, 95% CI: 0.45 – 0.98). In contrast to our expectations, modified TEE was not associated with a lower incidence of post-operative stroke (adjusted RR: 1.01, 95% CI: 0.71 – 1.43). Limitations of this study, besides the observational design, were the absence of a protocol to guide the diagnosis and management of aortic atherosclerosis, and an insufficient registration of the steps between test application and patient outcomes. To overcome these limitations would require randomization to a diagnostic protocol with- or without modified TEE. In Chapter 6 the results of the pilot-phase of such a diagnostic trial were presented. This study included 32 patients who were randomly allocated to perioperative screening with conventional TEE (and optional use of epiaortic ultrasound; control), or the addition of modified TEE (intervention). The primary end-point was the incidence of new diffusion-weighted lesions on a post-operative MRI. The aims of this pilot study were to asses the feasibility of the study design and to estimate the incidence of the primary end-point. A pre-operative MRI turned out not to be feasible and was therefore removed from the study protocol after inclusion of four patients. Modified TEE showed severe atherosclerosis of the DAA in 24% of patients, the surgical management was changed in 12% of all patients based on imaging results. A post-operative MRI was completed in 97% of patients, which was considerably more than expected from previous studies; one patient refused because of claustrophobia and dyspnea. The incidence of post-operative lesions was 29%, which was considerably lower than expected. We recommended changes in the design of a future full-cohort trial, i.e. to remove the preoperative MRI from the study protocol, to include patients with a higher prevalence of aortic atherosclerosis, and to perform a new sample-size calculation based on the current incidence of post-operative ischemic lesions.
In Chapter 7 we studied the outcomes after conventional AVR in patients aged >80 years (octogenarians). In the past decade transcatheter aortic valve replacement (TAVR) has become available as an alternative treatment for symptomatic aortic stenosis. The indication for this newer, less-invasive treatment is a very high surgical risk, or presence of contra-indications for conventional surgery. The latter may also constitute a severely calcified aorta, in which case a transapical or direct aortic approach may be preferable to prevent (cerebral) embolization. We therefore perform modified TEE imaging in all these patients before deciding on the best treatment. Registries have reported the regular off-label use of TAVR in elderly patients also. We compared the outcomes after conventional surgery in octogenarians. In a pre-TAVR cohort of 762 patients, we compared the outcomes of octogenarians 163 (21.4%) and patients aged 80 years. Octogenarians more often had a post-operative delirium compared (11.0% vs 6.2%, P = 0.03), the quality of life 30-days after surgery had not improved, but the operative mortality was similar (1.9% vs. 2.9%). The mortality at one-year follow-up was low, and comparable to younger patients (6.5% vs. 6.3%); more importantly, the quality of life was at or above the expected level for age-matched Dutch inhabitants. Our results indicate that, in octogenarians deemed suitable, surgical AVR was associated with a good long-term prognosis and quality of life.
PART II – AORTIC DISSECTION
Acute aortic dissection (AD) is a rare but highly lethal condition. A timely diagnosis is thus of the utmost importance. This first requires an adequate interpretation of the symptoms and signs and second requires accurate diagnostic tests.
In Chapter 8 we studied a cohort of 200 consecutive patients with an acute aortic dissection as the final diagnosis. As shown in previous studies, the symptoms on admission were diverse. The study focused on the initial differential diagnosis, which did not include an aortic dissection in 31% of the patients. Women and patients without back pain were more likely to be missed initially. Additional analyses indicated that “painless” AD was more likely to be missed, possibly because of a different clinical presentation with hemodynamic instability. Patients without an AD in the initial differential diagnosis had a longer time from admission to surgery, but similar short- and long-term outcomes. In hemodynamically stable patients with suspected acute aortic dissection, CT (or MRI) should be used as a primary test. Recent guidelines no longer recommend TEE imaging in these patients, because of limited visualization of the upper thoracic aorta. In hemodynamically unstable patients, TEE is recommended however, since this allows continuous hemodynamic monitoring. In Chapter 9, we describe how modified TEE can improve the diagnosis and management of patients with a dissected aorta. In the acute setting, modified TEE can be used to establish the diagnosis of a dissection, assess progression into the cerebral vessels and differentiate Stanford type A and type B dissections. Second, we describe the potential of intraoperative monitoring, which includes the guidance of cannulation of the true lumen and monitoring of the flow in the proximal cerebral vessels. Finally, modified TEE can be used for the diagnosis of intraoperative (iatrogenic) aortic dissection which, as we present, can be missed using conventional TEE.
Starting time 14.30 Building Waaier - Prof.dr.G. Berkhoff-zaal
