Eur J Cardiothorac Surg 1998;13:650-654
© 1998 Elsevier Science NL
Thoracic computed tomography prior to redo coronary surgery
Jochen Cremera,
Omke E. Teebkena,
Andrè Simona,
Alfred Hutzelmannb,
Martin Hellerb,
Axel Havericha
a Department of Cardiovascular Surgery, Christian-Albrechts-University, Arnold-Heller-Strasse 7, 24105 Kiel, Germany
b Department of Radiology, Christian-Albrechts-University, Arnold-Heller-Strasse 9, 24105 Kiel, Germany
Received 8 December 1997;
received in revised form 16 March 1998;
accepted 24 March 1998.
Corresponding author. Division of Thoracic and Cardiovascular Surgery, Hannover Medical School, Carl-Neuberg-Strasse 1, D-30625 Hannover, Germany; Tel.: +49 511 5326583; fax: +49 511 5325404.
 |
Abstract
|
|---|
Objective: Median resternotomy in coronary redo surgery represents a specific risk for injury of adjacent retrosternal structures. Aiming at improved preoperative evaluation of retrosternal structures, computed tomography (CT) techniques were routinely applied in redo cases. Methods: Of 99 patients undergoing coronary reoperations since April 93, thoracic CT scans were retrospectively analyzed for retrosternal vicinity of cardiovascular structures, condition of the ascending aorta and structural abnormalities of the sternum. Results: The minimal median distance between the posterior sternum surface and the anterior aortic wall was measured at 1.9±0.9 cm, whereas the mean closest distance to the anterior right ventricular wall was 0.4±0.5 cm. In 28 cases, the distance between sternum and aorta was smaller than or equal to 1 cm. No measurable distance between the sternum and the right ventricle was noted in 41 patients. Calcification of the ascending aorta became obvious 56 times. With respect to potential injury of the ascending aorta or the right ventricle, a safe reentry by resternotomy was facilitated in all cases. However, following complete sawing, the innominate vein became injured during retrosternal dissection in two cases. Preventive femoral vessel exposure was not performed and urgent femoral cannulation (n=1) was infrequent. Conclusions: Thoracic CT scanning prior to redo coronary surgery allows for detailed assessment of retrosternal relations and facilitates reopening of the sternum. Preventive femoral vessel exposure and lateral thoracotomies may be avoidable in many cases.
Key Words: Cardiac surgery • CT scanning • Sternotomy
|
Introduction
|
|---|
Along with the increasing number of coronary redo procedures a proper preoperative evaluation of retrosternal relations appears necessary to allow for a safe reopening of the chest, especially when applying a median resternotomy. Particular structures at risk for injury during reopening are the right ventricle, the ascending aorta, the innominate vein and retrosternal grafts. Although injuries of these structures in the setting of coronary reoperations have been numerously reported, additional diagnostic means beyond basic conventional chest X-ray and left heart catheterization have not been generally recommended. Computed tomography (CT) of the chest under these conditions has been mentioned by several groups
[1]
[2] but so far has not been postulated as a preoperative standard. To address the question to what extent preoperative CT scanning allows for a substantial evaluation of the relation of retrosternal structures prior to coronary redo surgery we reviewed CT findings with respect to major surgical complications occurring along with resternotomy in 99 elective consecutive cases.
|
Materials and methods
|
|---|
Along with major changes within the surgical team in March 1993 thoracic CT scanning was introduced into the preoperative diagnostic program prior to redo cardiac surgery. Subsequently, the original CT scans of 99 consecutive patients undergoing coronary redo surgery became available for a retrospective analysis by an independent radiologist. Criteria of special surgical interest were as follows: distance between the posterior sternal surface and the right ventricle and also to the ascending aorta, suspicion of innominate vein adhesion or patent graft adhesion to the posterior sternal surface, sclerosis and calcification of the ascending aorta, diameter of the ascending aorta and abnormalities of the sternal structure (e.g. pseudoarthrosis, dehiscence). Concomitantly, the surgical records considering the redo operation were reviewed, focusing on injuries of the previously mentioned retrosternal structures and mode of cannulation. As such, aortic cannulation, femoral cannulation prior to sawing, femoral cannulation immediately following sawing and preventive exposure of the femoral vessels without cannulation were noted.
Technique and evaluation of CT scanning
All CT scans were performed with a Somatom Plus (Siemens, Erlangen, Germany) by a spiral technique (120 kV, 210 mA). Routine 10-mm scans of the chest were obtained at 10-mm intervals. For distance measurements at individual scans, the Sienet Magic View 1000 System was applied. In the presence of anterior located patent grafts accompanied by a distance between aorta and sternum smaller than 1 cm, a second set of scans was produced by applying contrast medium imaging.
|
Results
|
|---|
CT findings
By means of thoracic CT scanning a mean distance of 1.9±0.9 cm was found between the posterior sternal surface and the closest segment of the anterior ascending aorta. In 28 cases the distance was smaller than 1 cm but a layer of soft tissue in between could always be identified (
Fig. 1
a). In seven of those patients, CT scans revealed distances even smaller than 0.5 cm. The corresponding lateral X-ray view of the chest (
Fig. 1b) usually did not allow for a reliable assessment of the distance and location of the ascending aorta to the posterior sternum. However, distances of more than 1 cm between aorta and sternum (
Fig. 2
) were regarded to be uncritical for reopening. Localized calcifications (
Fig. 3
) or sclerotic lesions were noted 56 times, predominantly extending into anterior segments of the ascending aorta or close to the origin of the brachiocephalic trunk. Mere sclerotic lesions of the ascending aorta, without, calcification, were detected in 26 patients. The maximal aortic diameter varied between 2.5 and 5 cm (mean 3.6±0.5 cm).
View larger version (64K):
[in this window]
[in a new window]
|
Fig. 1. (a) Anterior ascending aorta directly adjacent to the posterior sternum. (b) Corresponding conventional lateral X-ray of the chest.
|
|
For the right ventricle, however, a close relation to the sternum was frequently found as assessed by a minimal distance smaller than 1 cm in 92 patients. In 41 cases adhesions of both structures had to be suspected due to a missing separation layer in between (
Fig. 4
). In contrast, adhesions of the innominate vein to the sternum were assumed only in nine cases, but due to the horizontal scanning technique in 10-mm layers, the location of the innominate vein appeared less reliable.
Graft identification by use of contrast medium bolus infusion was attempted only when a patent anterior graft has been already assessed by coronary angiography and the distance between the sternum and the ascending aorta was smaller than 1 cm. Under these particular conditions patent grafts could be localized in five cases, fortunately being separated from the sternum (
Fig. 5
) in all cases. Visualization of a closed pericardium (
Fig. 6
) by means of CT scanning represented a rare finding.
Unexpected proximal sternum dehiscence was noted three times (
Fig. 7
). In all other cases the sternal structure appeared without pathologic findings.
Surgical access
As a result of CT findings, a median resternotomy using an oscillating saw appeared justified in all cases. Preventive exposure of the femoral vessels was not performed.
Applying this approach, catastrophic events following resternotomy could be prevented in all cases. However, lacerations of the innominate vein (n=2) while dissecting retrosternal structures after complete sawing required one urgent femoral cannulation. Even though superficial right ventricular epicardial lesions occasionally occurred, a deep injury or complete rupture of the right ventricular wall did not happen in our redo coronaries. In contrast the aorta or patent grafts had not been injured in any of these cases.
Although it has not consistently been reported, whether the pericardium had been closed or left open in primary operations, it is our impression that a pericardium previously left open does not result in an increased rate of epicardial lesions.
In presence of prediagnosed partial or complete sternal dehiscence (n=3), a partial bilateral longitudinal resection (slice <0.5 cm) was performed before reapproximation of the divided sternum.
|
Discussion
|
|---|
Despite increasing numbers, coronary reoperations are still associated with a higher perioperative morbidity and mortality rate comparing to primary procedures. Even in recent reports operative mortality rates differ between 3.4% and 11.4% for reoperative grafting
[3]
[4]
[5]. To what extent these mortality rates depend on surgical complications while reentering the mediastinum via repeat sternotomy cannot be exactly specified as these data are either not in particular described
[3]
[6], or the contribution of complications during reentry to the overall mortality and morbidity is not reported
[7]
[8]
[9]. Few details are available about injury rates and involvement of retrosternal structures when performing resternotomies.
From a questionnaire poll in 1983, a catastrophic hemorrhage may occur in about 2–6% of reoperative adult cases
[10]. Verkkala
[9] reported 2.8% injury of the right ventricle or the left anterior descending artery. More precise information about the injury rates due to resternotomy are given by Kaul
[8], who described 1.2% injury of the right ventricle and great arteries and 0.8% injury of patent anterior grafts or left anterior descending arteries. In contrast, Christenson and co-workers observed no injuries to the heart or bypasses among 597 redo patients
[7]. However, imaging techniques for a subtle preoperative evaluation of retrosternal relations appear essential for a safe reopening of the anterior chest. Whether CT scanning may allow for a more detailed assessment beyond lateral chest X-ray and aortograms has not been conclusively answered so far. Few publications on redo coronary surgery just mention the possibility of preoperative CT scanning for this purpose, without defining general or specific indications for this diagnostic technique. In 1983 only five of 131 surgeons applied CT scanning prior to reoperations under specific indications
[10].
As a conclusion of our retrospective analysis, CT scanning seems to allow for a clear evaluation of retrosternal distances to structures at risk for injury. Of these the ascending aorta and the right ventricle are of major importance. Thus, a risk-adjusted operative strategy is possible, avoiding unnecessary exposure of the groin vessels as preferred by the majority of surgeons responding to the questionnaire. Therein, a general exposure of the groin vessels has been reported by 81 of 131 surgeons. Also, indications for lateral thoracotomy
[2] in reoperative coronary surgery may be influenced by CT findings.
According to our current strategy based on routine preoperative CT evaluation in redo cases, additional anterior or parasternal thoracotomies would only be applied in case of aortic adhesions to the posterior sternal surface to allow for a preventive dissection of these structures before sawing. In addition, in cases with median sternal adhesions of patent vein or internal mammary artery grafts anterolateral minithoracotomies and femoral vessel exposure or modified ministernotomies are applied to allow for a safer graft identification and dissection before sawing. Conversely, dealing with right ventricular adhesions to the sternum just vertical traction to the sternum will be exerted by use of sharp clamps fixing the soft tissue around the xyphoid. Only if right ventricular adhesions are combined with pulmonary hypertension or sternum irregularities we consider preventive unloading of the heart by institution of extracorporal circulation via femoral vessel cannulation. Otherwise minimally invasive approaches with extrasternal incisions may obviate the risk for severely adherent retrosternal structures to become injured along with a median resternotomy.
The fact that sclerosis and calcification of the ascending aorta are better visualized by CT scans than by aortograms or conventional chest X-rays may have direct implications for the site of cannulation, location of proximal anastomoses and X-clamp modalities.
Besides various surgical options, no-touch techniques of the ascending aorta combined with deep hypothermic circulatory arrest or grafting to the fibrillating heart could become necessary in extreme cases. Femoral or subclavian arterial cannulation instead of aortic cannulation and multiple in situ arterial grafting or arterial T grafting would avoid direct access to the aorta. In cases with moderate or localized sclerotic lesions of the ascending aorta the cross clamp is posed within a short period of circulatory arrest under subtle palpation of the vascular wall. Similarly, the punch holes for the proximal anastomoses are created still under cross-clamping with an emptied aorta facilitating a better evaluation of the aortic wall thickness and localized sclerosis.
Singular attempts at detailed preoperative evaluation of retrosternal structures by other means such as transesophageal echocardiography or nuclear magnetic resonance imaging were not similarly successful. For instance, transesophageal echocardiography permits even a structural evaluation of the ascending aorta but fails in a reliable estimation of retrosternal relations and dimension. Nuclear magnetic resonance imaging, however, allows the differentiation of retrosternal structures but the resolution of structures is inferior to computerized tomography scanning.
As demonstrated with the selected illustrations we think, that CT scans give substantially more information, which aids a safe reopening of the sternum. As exemplified in
Fig. 1b, we found quite frequently that lateral X-rays of the chest did not allow for a reliable estimation of the distances of the aorta and right ventricle to the posterior sternum, especially not in patients with chronic obstructive pulmonary disease. This is, however, of significant importance as direct adhesion of the right ventricle to the sternum (more than 40% in our patient group) is probably more common than expected. Even if the estimation of a close distance between the posterior sternum and the anterior aorta would be sufficiently represented in lateral X-rays, the exact localization of critical adhesions would be difficult from lateral and also from transversal chest X-rays. Due to the summation effect of radiodense structures in lateral X-rays, particular information of small size structures as anterior grafts cannot be regularly visualized, whereas the superior resolution qualities of CT scanning allows for imaging of such structures (
Fig. 5). Moreover, sclerotic lesions especially when consisting of soft plaques and increased aortic wall thickness can rather be assessed by CT scanning (
Fig. 3) than by conventional X-rays.
On the basis of the available information a cost–benefit analysis could not be calculated exactly. Currently additional costs of about 260 DM (five-fold costs of a conventional chest roentgenogram) are linked with thoracic CT scanning. Despite lack of a comparative study we consider CT scanning to finally reduce overall costs, as detrimental complications associated with sternal reentry could be avoided in each of our patients. In addition the time required for the surgical team and operative facilities to preventively expose femoral vessels could be saved.
To what extent specific techniques during primary surgery anticipating the surgical risk of a sternal reentry as posterior routing of the left internal thoracic artery or application of polytetrafluoroethylene membranes or protective tubes may overcome the discussion on preoperative imaging of mediastinal structures in redo procedures remains open.
In consequence, based on a retrospective analysis of 99 consecutive patients undergoing coronary reoperation preoperative thoracic CT scanning allowed for detailed assessment of retrosternal relations and facilitated safe reopening of the sternum. Preventive femoral vessel exposure and lateral thoracotomies were be avoided in all cases.
|
References
|
|---|
- Galbut D.L., Traad E.A., Dorman M.J., DeWitt P.L., Larsen P.B., Kurlansky P.A., Button J.H., Ally J.M., Gentsch T.O. Bilateral internal mammary artery grafts in reoperative and primary coronary bypass surgery. Ann Thorac Surg 1991;52:20-28.[Abstract]
- Suma H., Kigawa I., Horii T., Tanaka J., Fukada S., Wanibuchi Y. Coronary artery reoperation through the left thoracotomy with hypothermic circulatory arrest. Ann Thorac Surg 1995;60:1063-1066.[Abstract/Free Full Text]
- Lytle B.W., Loop F.D., Cosgrove D.M., Taylor P.C., Goormastic M., Peper W., Gill C.C., Golding L.A.R., Stewart R.W. Fifteen hundred coronary reoperations. J Thorac Cardiovasc Surg 1987;93:847-859.[Abstract]
- He G.-W., Acuff T.E., Ryan W.H., He Y.-H., Mack M.J. Determinants of operative mortality in reoperative coronary artery bypass grafting. J Thorac Cardiovasc Surg 1995;110:971-978.[Abstract/Free Full Text]
- Salomon N.W., Page U.S., Bigelow J.C., Krause A.H., Okies J.E., Metzdorff M.T. Reoperative coronary surgery. J Thorac Cardiovasc Surg 1990;100:250-260.[Abstract]
- Loop F.D., Lytle B.W., Cosgrove D.M., Woods E.L., Stewart R.W., Golding L.A.R., Goormastic M.T.P. Reoperation for coronary atherosclerosis. Ann Surg 1990;212:378-386.[Medline]
- Christenson J.T., Schmuziger M., Simonet F. Reoperative coronary artery bypass procedures: risk factors for early mortality and late survival. Eur J Cardiothorac Surg 1997;11:129-133.[Abstract]
- Kaul T.K., Fields B.L., Wyatt D.A., Jones C.R., Kahn D.R. Reoperative coronary artery bypass surgery: early and late results and management in 1300 patients. J Cardiovasc Surg (Torino) 1995;36(4):303-312.[Medline]
- Verkkala K., Järvinen A., Virtanen K., Keto P., Pellinen T., Salminen U.-S., Ketonen P., Luosto R. Indications for and risks in reoperation for coronary artery disease. Scand J Thor Cardiovasc Surg 1990;24:1-6.[Medline]
- Dobell A.R.C., Jain A.K. Catastrophic hemorrhage during redo sternotomy. Ann Thorac Surg 1983;37:273-278.[Abstract]
Top
Abstract
Introduction
