CT of the thorax has been shown to aid routine preoperative planning for redo cardiac surgery in evaluating retrosternal anatomic relations,6 and more recently in minimally invasive or catheter based cardiac procedures.7 There are no guidelines to suggest its use in routine conventional cardiac surgical procedures done via sternotomy. In the authors’ institution, the need for CT thorax as part of preoperative planning is left to the discretion of the performing surgeon.
There were some distinct differences in demographics between the groups that did and did not undergo routine preoperative screening CT thorax. Most of the patients who underwent surgeries after 2015 by the lead surgeon in this study routinely underwent CT thorax as part of preoperative planning especially if they were older (age >60 years), due to change in the surgeon’s clinical practice. This likely explains the higher mean age in the screening CT group. Patients with recent MI and poor EF were more likely to be part of the subset of patients who warranted urgent surgery, and in whom obtaining a preoperative screening CT was not essential.
Despite the time lag between the no CT and CT groups, there was no significant difference between CPB and crossclamp times, which may have been expected with increasing surgeon experience. EAUS was used more often in the no CT group if there were palpable plaques felt on the aorta, to aid in finding the most appropriate cannulation or crossclamp sites, whereas in the CT group, the preoperative CT images were assessed for this purpose. The cost of epiaortic ultrasound and a non-contrasted CT thorax is similar. The CT thorax is able to provide additional anatomic information though not in real-time.
Change in operative strategy in this study was lower than the 9.2% in a more recent study by de Hartog-Dikhoff et al3 and occurred due to palpable aortic calcified plaques intraoperatively rather than the presence of calcifications on the CT. In most patients who faced an altered operative strategy, the aorta was still cannulated for CPB, but some underwent off-pump or on-pump beating heart surgery. Majority of the cases were coronary artery bypass grafting (CABG) surgeries. The surgeon routinely performs left internal mammary artery (LIMA) to the left anterior descending (LAD), and aortocoronary grafts for the rest. While many cases were modified to single crossclamp technique, the modifications did not appear to affect routine graft configuration. All patients who had change of operative strategy within the screening CT group had aortic calcifications present on CT, suggesting that identifying aortic calcifications on a preoperative CT may help with operative planning for routine cardiac surgery.
A meta-analysis by den Harder et al showed 7 articles pertaining to primary cardiac operation of which 1194 subjects received a preoperative CT scan.8 The study suggested preoperative CT may be beneficial, albeit with weak evidence. However, only 2 studies reported outcomes separately for a control group. These studies showed a decrease in mortality and stroke rates in favour of the CT group, and the main cause of death was stroke.9,10 There was no significant difference in mortality or stroke between the two groups in this study, which included all-comers who underwent cardiac surgery. A recent propensity matched study by Sandner et al enrolling 2320 consecutive patients who underwent isolated CABG showed 11.2% of patients in the CT group received modifications of the conventional CABG procedure versus 3.3% in the non-CT group.11 There was a significant reduction in stroke incidence among patients who underwent CT aortic angiography in both the matched and unmatched cohorts. The significant difference in stroke rates may be related to the use of contrasted CT-angiography instead of plain CT, which may have detected soft atheromas in addition to calcifications.
A study by Linden et al showed that the presence of ascending aortic atheroma was associated with increased postoperative stroke, was detected in 26.2% of patients and most often involved the distal anterior segment.3 It showed an 8.7% incidence of stroke despite minor surgical modifications. This is in contrast to this study wherein none of the patients with a modified operative strategy developed stroke, as was the case in the study by Nishi et al.12 Here too, change of operative strategy itself appeared to be protective of stroke. This is unsurprising as manipulation of a diseased aorta is often associated with stroke. The review by Gaudino et al suggested that early postoperative stroke was likely related to operative technique, and was inversely associated with off-pump surgery.1 Also, the presence of aortic calcifications on CT alone was not associated with development of postoperative stroke when stratified within the CT group. The relatively higher stroke rate overall may be related to the number of higher risk surgeries undertaken, evidenced by the proportion of patients with poor ejection fractions and combined surgeries.
While all the variables in Table 4 showed significance in relation to postoperative stroke individually, the effects were mitigated by adjusting for age and EF. This is likely because atrial fibrillation and aortic calcifications are both associated with increased age,13 and the use of IABP is associated with poor EF. Analysis for an age cut-off for patients to undergo preoperative CT thorax showed a significant cut-off at 61 years for change in operative strategy. However, this did not translate to postoperative stroke. This suggests that performing preoperative CT thorax may be useful for preoperative planning in older patients, but not for reducing the incidence of stroke. The finding that palpable plaque was the most significant predictor for change in operative strategy when adjusted for age suggests that CT thorax may play a supportive role, rather than a definitive role in decision-making for altering surgical strategy.
This study also sought to reflect pulmonary infection in relation to CT features of emphysema. This has not been reported in other studies. In this study, the incidence of chronic pulmonary diseases of asthma and chronic obstructive pulmonary disease (COPD) were low in both groups. The presence of emphysema features on CT was also low despite more than half the study population in both groups comprising smokers. There was no significant association between emphysema features and pulmonary complications. Thus, preoperative screening CT thorax does not seem useful in determining risk of postoperative pneumonia for patients undergoing cardiac surgery.
On the whole, this study provides a fair comparison between the screening CT group and control group without preoperative CT thorax, and outcomes were comparable despite increased age and poorer EF in the no CT group. Most of the CTs were non-contrasted, with low radiation dose. Almost all patients underwent primary cardiac surgery. However, there are several limitations that must be noted. First, this is a retrospective observational study with potential for selection bias. Second, the difference in time period between the 2 groups coincides with increased operative experience of the surgeon for the CT group, potentially taking on more complex or higher risk cases. The surgical risk is not reported in this study. Third, there were a few patients who did not undergo surgery due to discovery of a porcelain aorta on CT thorax, and are not captured in this study. Lastly, other embolic complications such as mesenteric ischemia are not reflected in this study, though these were very rare.