The New England Journal of Medicine
e-mail icon  FREE NEJM E-TOC    HOME   |   SUBSCRIBE   |   CURRENT ISSUE   |   PAST ISSUES   |   COLLECTIONS   |    Advanced Search
Sign in | Get NEJM's E-Mail Table of Contents — Free | Subscribe
 
Original Article
Volume 328:1-9 January 7, 1993 Number 1
NextNext

The Diagnosis of Thoracic Aortic Dissection by Noninvasive Imaging Procedures
Christoph A. Nienaber, Yskert von Kodolitsch, Volkmar Nicolas, Volker Siglow, Angela Piepho, Carsten Brockhoff, Dietmar H. Koschyk, and Rolf P. Spielmann

 

This Article
-Abstract

Commentary
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation
ABSTRACT

Background and Methods This study was designed to assess the safety and reliability of new noninvasive imaging methods as compared with aortography in the diagnosis of dissection of the thoracic aorta. One hundred ten patients with clinically suspected aortic dissection followed a diagnostic protocol that included transthoracic and transesophageal color-flow Doppler echocardiography (TTE and TEE), contrast-enhanced x-ray computed tomography (CT), and magnetic resonance imaging (MRI). Imaging results were compared in a blinded fashion and validated independently against intraoperative findings in 62 patients, autopsy findings in 7, and the results of contrast angiography in 64.

Results The sensitivities of MRI, TEE and x-ray CT for detecting dissection were similar, at 98.3, 97.7, and 93.8 percent, respectively; TTE had a sensitivity of only 59.3 percent (P<0.005). The specificities of both TTE (83.0 percent) and TEE (76.9 percent) were lower than those of x-ray CT (87.1 percent) and MRI (97.8 percent; P<0.05), mainly as a result of false positive findings in the ascending aorta. MRI and x-ray CT were more sensitive than TTE in detecting the formation of thrombus in the entire thoracic aorta (P<0.05), but were not superior to TEE in this regard. CT was not effective in detecting an entry site or aortic regurgitation, but MRI and TEE accurately identified both. Two patients died during or soon after CT and TEE, and three died between retrograde angiography and surgery.

Conclusions A noninvasive diagnostic strategy using MRI in all hemodynamically stable patients and TEE in patients who are too unstable to be moved should be considered the optimal approach to detecting dissection of the thoracic aorta. Comprehensive and detailed evaluation can thus be reduced to a single noninvasive diagnostic test in the investigation of suspected dissection of the thoracic aorta.

Patients with suspected dissection of the thoracic aorta require immediate diagnostic evaluation so that urgent therapeutic interventions can begin. Detailed information on associated findings, such as the extent of the dissection, the location of the entry site, the formation of thrombus in the false lumen, and evidence of pericardial effusion or aortic regurgitation may be helpful in selecting the best management plan and assessing the patient's prognosis1,2,3,4,5,6,7. The established diagnostic procedures, including aortography,8,9,10 contrast-enhanced x-ray computed tomography (CT),11,12,13,14,15,16 and two-dimensional transthoracic color-flow Doppler echocardiography (TTE),17,18,19,20,21 all have inherent procedural or diagnostic shortcomings. Transesophageal echocardiography (TEE)22,23,24,25 and magnetic resonance imaging (MRI)26,27,28,29,30 are recent advances in noninvasive imaging technology that have been useful in assessing the thoracic aorta31,32,33. This study was designed to explore our previous findings33 on this important clinical problem in greater depth by comparing the diagnostic accuracy of all noninvasive procedures, including x-ray CT, with that of invasive aortography in an attempt to identify the most reliable, comprehensive, and efficacious diagnostic strategy in the investigation of suspected aortic dissection.

Methods

Study Design

All 110 patients who presented to our institutions with clinically suspected dissection of the thoracic aorta34 underwent an initial screening with TTE, followed by TEE, x-ray CT, or MRI (every patient underwent at least two procedures); the sequence of imaging procedures was determined at random. All techniques, including MRI, were available on a 24-hour basis. Informed consent was obtained for each procedure. Forty-seven patients (42.7 percent) underwent all four noninvasive tests, 97 (88.2 percent) underwent three or more, and all 110 underwent two or more imaging procedures before validation of the results (Table 1). The results of each imaging procedure were analyzed for safety and reliability in detecting and classifying both the extent of dissection and the presence of associated findings before validation with contrast angiography (in 64 patients), by intraoperative inspection (62), or by autopsy (7). (Twenty-three patients underwent more than one validation procedure).

View this table:
[in this window]
[in a new window]
  		Table 1. Number of Imaging Procedures per Patient.

 
Patients

Over a period of five years, 110 consecutive patients referred to two medical centers with suspected dissection of the thoracic aorta were analyzed; there were 40 women and 70 men, with a mean (±SD) age of 54 ±14 years (range, 19 to 96). According to an accepted convention, 35 patients were considered to have acute dissections, with sudden onset of symptoms within the previous 48 hours, and 27 patients had subacute dissections, with symptoms presenting during the previous two weeks35,36. Patients with chronic dissections were not included. Echocardiographic studies were performed with hemodynamic monitoring in the emergency room or intensive care unit. X-ray CT and MRI studies were performed in a nearby unit under the close surveillance of a cardiologist or thoracic surgeon, with voice communication to comfort the patient; a telemetric electrocardiograph and the patient's blood pressure were monitored continuously. Antihypertensive and mild sedative treatments were administered if necessary. Five patients were studied while receiving mechanical ventilation.

Echocardiographic Evaluation

Screening TTE was performed in all patients, with standard and suprasternal cross-sectional echo projections obtained with 2.25-and 3.5-MHz transducers and sector scanners (V3400 R CV60, Diasonics, Palo Alto, Calif.; or HP 77065 or HP Sonos 1000, Hewlett-Packard, Andover, Mass.); the sector scanners were fitted to video copy processors (Sony, Cologne, Germany; or 77570 Mitsubishi, Mitsubishi, Kyoto, Japan). In 70 patients TEE with optional color-flow Doppler mapping was performed with a wide-angle transducer at 5.0 MHz (HP 21362A, Hewlett-Packard). Images were recorded on a 0.5-inch (1.25 cm) VHS video recorder (AG-7330 E, Panasonic, Osaka, Japan). Photoprints were obtained from a Mitsubishi black-and-white or Sony color video printer during playback of the recorded documentation.

In patients with subacute symptoms TEE was performed after a fasting period of one hour; all the other patients were studied as rapidly as possible. The upper pharyngeal region was locally anesthetized with 10 percent lidocaine spray. After the echoscope was introduced, gradual rotation and tilting provided an image of the descending aorta; withdrawal of the probe permitted the mapping of the ascending aorta and the arch. The image could be improved by upward angulation and close contact with the esophageal wall. TEE was performed by experienced operators, with the patients sedated and undergoing hemodynamic monitoring, and the procedure required 9 to 18 minutes (average, 13 ±6).

X-Ray CT

Third-generation CT scanners (Somatom II and Somatom Plus, Siemens, Erlangen, Germany) were used in 79 patients. After intravenous bolus injections of 80 to 100 ml of non-ionic contrast medium (Ultravist 370, Schering, Berlin; or Solutrast 370, Byk-Gulden, Konstanz, Germany), tranverse scans were obtained during shallow respiratory excursions from the arch to the aortic bifurcation in intervals of 2 cm11,12. Acquisition time ranged from 1 to 5 seconds per slice, or 5 to 15 minutes of total scanning time.

MRI

One hundred five patients underwent MRI with a whole-body magnet at 1.5 tesla (Gyroscan S15, Philips, Best, the Netherlands). Patients were monitored with continuous telemetric electrocardiography, sphygmomanometric blood-pressure readings, and voice communication. With long intravenous lines in place for antihypertensive medication, gated spin-echo pulse sequences with a trigger delay of 100 msec from the R wave were performed (echo time, 30 msec). With scout images as a guide, transverse scans were acquired encompassing the ascending aorta, the aortic arch, and the descending aorta; slice thickness was 8 to 10 mm, and the diameter of the field of view was 40 cm. The size of the acquisition matrix was 256 by 192 to 256 phase-encoding steps. Longitudinal scans were obtained in a similar fashion. Dissections involving the ascending aorta were routinely imaged in coronal slices, whereas dissections confined to the descending aorta were imaged in planes parallel to the aortic arch (oblique sagittal view). In some patients, however, the tomograms were perpendicular to the dissected flap on a transverse image. If dissections were not visible on either transverse images or coronal scout images, an oblique sagittal scan was also obtained. The slice thickness in longitudinal scans was 5 to 7 mm, and the field of view was 40 to 45 cm. In addition, cine MRI was performed in 85 of 105 patients to assess aortic-valve competence by acquiring transverse scans of 10-mm slices through the aortic valve and the left ventricular cavity with an echo time of 12 msec37,38. Customized tubing was used in five patients who were receiving mechanical ventilation. Scanning time for MRI was 23 ±3 minutes for all spin-echo sequences; with the addition of cine MRI, the average scanning time was 39 ±16 minutes (range, 24 to 60).

Reference Techniques

For validation we performed conventional or digital contrast angiography, using retrograde arterial catheterization from the femoral artery under fluoroscopic guidance. Injections of 20 to 40 ml of contrast material into the aortic root were used to assess aortic-valve regurgitation. The diagnosis of aortic dissection was established with the identification of an intimal flap or a double lumen; indirect signs were a compressed true lumen and a severely thickened aortic wall34,39. The entry site was identified by the shunting of contrast material between the true and false lumens, passage of the catheter or guide wire, or disruption of the intimal flap40,41. A diagnosis of dissection was excluded when conventional or digital angiography produced unequivocally negative results. The time required for angiography was 40 ±16 minutes. Arterial digital-subtraction angiography was used in 25 of 64 patients; the diagnostic accuracy of digital angiography of the aorta was considered similar to that of conventional cine angiography42,43.

Intraoperative (n = 62) and postmortem (n = 7) inspections of the aorta and adjacent tissues were performed by pairs of experienced surgeons or pathologists and were documented at the time of visual or digital examination.

Diagnostic Criteria

With TTE and TEE, a diagnosis of dissection was confirmed by the presence of two vascular lumens separated by an intimal flap; if there was complete thrombosis of the false lumen, a central displacement of intimal calcifications was considered to be diagnostic of aortic dissection17,18,19,23. According to the criteria of Daily et al., dissections involving the ascending aorta were classified as type A regardless of the site of the primary intimal tear (Figure 1); all other dissections were designated type B44. The site of entry was defined by a disruption of the dissected membrane or by a typical communication between the two lumens on color-flow Doppler echocardiography20,21,24,25. Mural echogenic structures in the false or true lumen were identified as thrombi33. Aortic regurgitation was diagnosed if there was diastolic fluttering of the mitral leaflets or backflow on pulsed or color-flow Doppler examination25,45. The separation of pericardial contrast material was considered to indicate pericardial effusion.


View larger version (96K):
[in this window]
[in a new window]
  		Figure 1. TEE Scans in a Patient with a Type A Aortic Dissection.

In Panel A and Panel B, a transverse two-dimensional TEE sector scan shows the ascending aorta 2 cm above the plane of the aortic valve and a flap in oblique orientation separating the true lumen (TL) from the false lumen (FL), with a crescent-shaped, partially formed mural thrombus (Th). A superimposed color-flow Doppler signal in Panel B identifies flow in the true lumen in red. Panel C and Panel D show the dissection continuing into the lower and upper descending thoracic aorta, respectively. The crescent-shaped false lumen reveals some swirling echo reflections indicative of low flow or near-stasis.

 
For x-ray CT, we used established criteria to identify aortic dissection and thrombus formation46,47; briefly, an abrupt transition to a larger lumen at the origin of a side branch, the visualization of a dissecting membrane, or the presence of a poorly or not at all opacified crescent portion of the aorta after contrast enhancement in a patient with a compressed or flattened true lumen was considered diagnostic. No attempt was made to assess aortic regurgitation or the entry site with x-ray CT48.

With MRI, an aortic-wall dissection was diagnosed if there were two separated lumens. Less specific criteria, such as aortic widening or the spiraling of a thrombosed false lumen, were not considered diagnostic of dissection. An entry was located at the site of disruption or communication between two lumens. Using previously described cine techniques, we analyzed MRI scans for thrombus formation in the false or true lumen49 and for evidence of aortic regurgitation and pericardial effusion30,37. Figure 2 shows a typical type A lesion with detailed mapping of the dissecting process and involvement of a side branch.


View larger version (166K):
[in this window]
[in a new window]
  		Figure 2. Spin-Echo MRI Scan of a Type A Aortic Dissection in the Coronal Orientation.

The flap (arrow) begins in the ascending aorta and extends through the aortic arch into the left carotid artery, clearly separating the true (TL) and false (FL) lumens.

 
Statistical Analysis

Before validation, each image was interpreted by two or three experienced readers who were unaware of other imaging results, using a consensus method29. The sensitivity and specificity of each imaging technique were determined from the percentage of true positive and true negative results. All images were jointly reevaluated by the same physicians and by one technologist at a second unblinded reading session, to analyze the effect of the disclosure of the other imaging results on interpretation. The significance of changes in the frequency ratio of two dependent distributions (imaging methods) was assessed by the McNemar test with continuity correction50. All P values are two-sided, and the level of significance was set at 5 percent.

Results

Data on the demographic characteristics of the patients with proved dissections, their underlying disease, the treatment strategy, the interval from hospitalization to surgery, and the outcomes are available elsewhere (*). There were 24 acute and 8 subacute type A aortic dissections, and 11 acute and 19 subacute type B dissections. The age distribution did not differ between subgroups, but the patients with type A lesions were on average 52 ±14 years old, and those with type B lesions were 59 ±8 years old (P = 0.4). Twenty-seven of the 32 patients with type A lesions underwent a surgical procedure; there were three preoperative and three postoperative deaths between days 11 and 58. The median interval between the times of initial hospitalization and surgical intervention was 13.5 hours for patients with acute type A dissections and eight days for those with subacute lesions (P<0.01). Surgery was performed less frequently for type B dissections (15 of 30 cases) and always electively, between five days and five months after hospitalization. Four of the patients with medically treated type B dissections died from 10 days to 3 months after hospital admission. The sensitivities of TEE, x-ray CT, and MRI for the detection of dissection were excellent and did not differ between acute and subacute lesions; the sensitivity of TTE was suboptimal for type A dissection (especially for acute lesions) and was not useful for either acute or subacute type B lesions (Table 2).

View this table:
[in this window]
[in a new window]
  		Table 2. Sensitivity of Imaging Procedures According to Type of Aortic Dissection.

 
The characteristics of the groups of patients and information on the validation procedures are shown in Table 3; in 62 of 110 patients the dissection of the thoracic aorta was confirmed by angiography or morphologic evaluation that revealed a spectrum of findings similar to those of previous imaging studies in patients with aortic disease19,21,24,25,45,46.

View this table:
[in this window]
[in a new window]
  		Table 3. Patient Characteristics, Diagnostic Procedures, and Validated Findings.

 
Table 4 summarizes the independent diagnostic potential of each imaging method, both for the detection and classification of a dissection and for associated findings. MRI was diagnostic of aortic dissection in 58 of 59 patients with positive findings; similarly, TEE was diagnostic in 43 of 44 patients, whereas x-ray CT failed to identify 3 of 48 patients whose diagnoses were validated by angiography, morphologic evaluation, or both. The sensitivities of MRI, TEE, and x-ray CT were 98.3, 97.7, and 93.8 percent, respectively, significantly higher than that of TTE (59.3 percent, P<0.005). There was one false positive finding on MRI, four on x-ray CT, and six on TEE, resulting in specificities of 97.8 percent for MRI, 87.1 percent for CT, and 76.9 percent for TEE (P<0.05 for the comparison with MRI).

View this table:
[in this window]
[in a new window]
  		Table 4. Detection and Classification of Thoracic Aortic Dissection, According to Imaging Procedure.

 
The sensitivities of MRI and TEE for type A dissections were 100 and 96.4 percent, respectively, and were thus significantly higher than those of CT and TTE (P<0.05). Moreover, with specificities of 98.6 and 100 percent, MRI and CT were the most specific methods for excluding a dissection involving the ascending aorta (P<0.05 for the comparison with TEE and TTE). For type B dissections, MRI, TEE, and x-ray CT had similar diagnostic potential, with sensitivities of 96.5, 100, and 96.0 percent, respectively, and were significantly better than TTE (P<0.005). This notion was substantiated by a subanalysis of 47 patients who underwent all four imaging procedures. The comparison in individual patients shown in Table 5 confirmed that the specificities of x-ray CT and MRI were better than those of TEE and TTE in the ascending aorta (P<0.05). Again, TTE was not useful in assessing the descending aorta.

View this table:
[in this window]
[in a new window]
  		Table 5. Diagnostic Potential of TTE, TEE, X-Ray CT, and MRI for the Detection of Thoracic Aortic Dissection in the 47 Patients Who Underwent All Four Procedures.

 
The reliability of each imaging method in identifying the spatial extent of dissections and yielding associated findings of potential prognostic importance is shown in Table 6. Both MRI and TEE were highly reliable in identifying an entry site; x-ray CT was not useful in providing this information, and conventional TTE was not helpful in detecting communications within the descending aorta. Thrombus formation was best detected by x-ray CT and MRI. Both echocardiographic techniques were less sensitive for thrombus formation in the ascending aorta and in the arch; four of seven instances were missed by TEE. Aortic regurgitation was best identified by Doppler echocardiographic techniques; however, in contrast to x-ray CT, MRI could also identify aortic regurgitation with a high degree of reliability at the expense of an additional 15 minutes for cine acquisition. Pericardial effusion was detected by all the methods except angiography.

View this table:
[in this window]
[in a new window]
  		Table 6. Identification of Thoracic Aortic Dissection and Associated Findings, According to Imaging Procedure.

 
Despite an overall low rate of procedural complications, severe adverse effects were noted in addition to fluctuating blood pressures. One patient with a type A dissection died from rupture of the aorta during CT scanning; in another with a type A dissection the aorta ruptured within 10 minutes of a completed TEE study. The deaths of three patients (one with a type A dissection and two with type B dissections) after retrograde angiography and before surgery may have been related to catheter propagation of the dissecting aneurysm. No adverse effects were noted in association with MRI, and all attempted MRI scans were completed.

The unblinded reexamination of echocardiographic tapes, CT scans, and MRI scans by the same readers revealed no differences between the blinded and unblinded evaluations.

Discussion

This prospective comparison between established and new imaging techniques provided evidence that MRI and TEE are preferable as noninvasive but definitive diagnostic approaches in the investigation of suspected thoracic aortic dissection. Neither x-ray CT nor aortography provided additional vital information, and these techniques may therefore be avoided or relegated to a complementary role.

Conventional TTE was of limited diagnostic value in the assessment of the thoracic aorta, especially the descending segment, because of anatomical and technical drawbacks such as a limited field of view and a lack of visualization of branch vessels, as well as the need for optimal technical performance. For aortic regurgitation and pericardial effusion, however, TTE proved to be highly valuable as an adjunct to spin-echo MRI sequences.

Both TEE and MRI had excellent sensitivity for a dissection at any level of the thoracic aorta. MRI was the most reliable method of avoiding false positive findings and provided the most comprehensive information on associated findings and side-branch involvement. As compared with MRI, the specificity of TEE was suboptimal in the ascending aorta. False positive findings in the ascending aorta have been reported anecdotally with TEE, and the explanation has been either extensive plaque formation or echo reverberations in an ectatic vessel24,33. Routine echocardiographic investigation for a flow signal in the false lumen could probably enhance the specificity of TEE25. The lower specificity of TEE in the ascending aorta may not be surprising, given the lack of true specificity data for TEE; previous studies that reported higher rates of specificity involved either selected cases of proved dissection or the unblinded and intraoperative use of TEE24,25,45.

False positive results with MRI have not been previously reported; the single such result in our study may have been caused by pulsating artifacts or may have been a true positive result in a patient with an intraluminal hematoma and a small tear that may not have been recognizable when the aortic valve was replaced. The quality of the MRI and x-ray CT images was not affected by air in the trachea, a methodologic problem that may explain the suboptimal specificity of ultrasound imaging of the ascending aorta. This inherent limitation may be overcome with biplanar or omniplanar TEE; there may still be false positive findings in the ascending aorta, but visualization of the aortic arch is likely to improve with multiple planes25,51.

In this study, four cases of ascending aortic dissection were missed with x-ray CT, as compared with one case with TEE and none with MRI. The sensitivity of x-ray CT in our study (82.6 percent) confirmed observations by Erbel et al.24 and Ballal et al25.; they reported sensitivities from 67 to 83 percent. With CT, dissections may be obscured by complete thrombosis of one lumen or similar opacification in both the true and false lumens, without temporal or densitometric differences; there may also be streak artifacts12,15. Moreover, x-ray CT was the only method we considered inappropriate for locating the entry site (point of origin of a dissection)14,48. In contrast, TEE proved to be highly efficacious in identifying communications between the true and false lumens, as Ballal et al. also found25. Our data extend this diagnostic potential to MRI, although the prognostic importance of such communications remains unclear, given the possibility of retrograde dissection and intramural hemorrhage.

The formation of thrombus in a false lumen was detected most reliably by MRI and x-ray CT. Conventional TTE in all segments of the aorta and TEE in the ascending aorta were more likely to miss this good prognostic sign41; Hashimoto et al. could identify thrombus formation in only 2 of 12 patients by TEE45.

As in previous studies, aortic regurgitation was an associated finding in 50 percent of type A lesions and 10 percent of type B lesions21,24,45. Both color Doppler echocardiography and cine MRI were highly reliable in detecting aortic-valve regurgitation, whereas CT was not48. The echocardiographic diagnosis of aortic regurgitation was usually obtained within 2 minutes, but cine MRI required 15 ±5 minutes; this additional scanning time did not result in more complications.

Pericardial effusion was diagnosed in 25 percent of the patients with type A dissections; this proportion is slightly higher than previously reported24. In no patient was this poor prognostic sign6,41 missed by any imaging technique.

Limitations

The traditional gold standard of contrast aortography carries a substantial risk of procedural complications and diagnostic pitfalls; in patients with thrombotic occlusion of the false lumen, dissection may be missed. Conversely, mural thrombi may be present in true aneurysms that are not dissecting10,34,40,52,53. In our study, the results of contrast angiography were discordant with intraoperative morphologic findings in two patients because the false lumen was not opacified; both MRI and TEE provided the correct diagnosis in these patients. Morphology and angiography were concordant in the remainder of this series, confirming the value of contrast angiography as an excellent reference method, with a reported accuracy of 99 and 100 percent34,39,46,48. Moreover, patients with negative findings on contrast angiography had an uneventful follow-up of 28 ±5 months and tested negative on follow-up evaluation.

The problems of coronary artery involvement and the severity of aortic regurgitation were not directly addressed in this study. Our results, however, indirectly support the notion that the need for coronary angiography may be overemphasized, particularly in patients with acute type A dissections, since coronary involvement in the dissecting process is relatively rare, ranging from 1.5 to 7.5 percent48,54. Furthermore, selective angiography is potentially dangerous and may injure the coronary ostia in acute ascending dissections. Finally, coronary angiography is not helpful in the emergency setting, because urgent surgical intervention is advised whenever a dissection encompasses the ascending aorta, regardless of coronary involvement. In any case, visual inspection is an excellent way for the surgeon to assess potential involvement of the coronary ostia and to decide on the reimplantation or grafting of coronary arteries3,55,56.

In principle, with improving techniques both the right and left main coronary arteries may be imaged by TEE and MRI25,57. But for the sake of time, one should not perform bypass grafting for coexisting coronary artery stenoses in the emergency situation of aortic dissection; coronary artery disease can instead be treated medically, with elective angioplasty or bypass grafting during follow-up. These options may also make selective coronary angiography obsolete in the emergency management of aortic dissection, although patients over the age of 40 with acquired cardiac disease conventionally undergo coronary angiography before open-heart surgery.

Feasibility and Safety

In this series of 105 patients, MRI was established as a safe procedure, even for emergency diagnostic evaluation in severely ill patients with acute aortic dissection. With continuous electrocardiography and blood-pressure monitoring and with a physician present, no patient was in fact isolated during MRI. There were no serious side effects, and the fact that access to the patient was limited during scanning was not associated with an increased risk. However, two patients did not undergo scanning because they had pacemakers, and one because the patient had claustrophobia. In contrast, one patient died during x-ray CT despite antihypertensive treatment, and another died 10 minutes after TEE from acute rupture of the ascending aorta; similar observations have recently been reported58,59. Although both TEE and x-ray CT are generally considered safe, they are semi-invasive. In addition to stress, TEE may cause fluctuations in blood pressure, and CT requires the infusion of contrast material, with the possibility of allergic reactions, hypotension, and nephrotoxicity. The deaths of three patients that were associated with retrograde angiography may have been related to propagation of the dissection60. Mechanical ventilation did not interfere with either imaging procedure.

Recommended Approach to Diagnosis

On the basis of our data, we believe that TTE should be followed by a single definitive diagnostic step before surgical treatment. Therefore, one safe and reliable imaging procedure might be used rather than a stepwise diagnostic approach including x-ray CT, digital subtraction, or conventional angiography as previously advocated48,61,62,63. MRI combines safety, superb accuracy, and unrestricted and comprehensive mapping of the entire aorta, side branches, and adjacent tissues. With conventional TTE as an adjunct, MRI may be limited to spin-echo sequences that require no more time than CT or TEE but offer the advantage of a free choice of imaging planes. Excellent sensitivity was also obtained with TEE, without the need to move the patient and therefore in less time. TEE had a lower specificity in the ascending aorta, however, and is a semi-invasive procedure with a small but definite risk. Although the information obtained with TEE appears to be less comprehensive than that obtained with MRI, both methods could provide all the diagnostic information required for surgery, with no need for x-ray CT or angiography. Unlike angiography, MRI and TEE may even identify precursors of dissection, such as intramural hemorrhage with no luminal component. The potential disadvantages of current forms of MRI technology, such as the need to move the patient, limited direct access, and longer examination time when cine MRI sequences are used, did not increase individual risk. Improved technology and more rapid acquisition may alleviate these shortcomings. At present, the use of MRI in all stable patients with acute or subacute lesions and the use of TEE in unstable patients for whom transportation does not appear to be safe should be considered the diagnostic strategies of choice in investigating suspected aortic dissection and guiding urgent surgical therapy.

We are indebted to Dr. David Carmichael, Scripps Memorial Hospital, La Jolla, California, for his encouragement; to the staff members of the emergency units and departments of radiology, anesthesiology, thoracic surgery, and pathology for their support; to the referring doctors; to Dr. George T. O'Byrne for his thoughtful suggestions; and to Miss Dorte Oestreich for assistance in the preparation of the manuscript.

* See NAPS document no. 04993 for two pages of supplementary material. To order, contact NAPS c/o Microfiche Publications, 248 Hempstead Tpk., West Hempstead, NY 11552.


Source Information

From the Division of Cardiology, Department of Internal Medicine II (C.A.N., Y.v.K., V.S., C.B., D.H.K.), the Department of Diagnostic Radiology (V.N.), and the Department of Thoracic Surgery (A.P.), Universitats-Krankenhaus Eppendorf, Hamburg; and the Department of Radiology, Christian-Albrechts-Universitat, Kiel (R.P.S.) -- both in Germany.

Address reprint requests to Dr. Nienaber at the Department of Internal Medicine II, Division of Cardiology, Universitats-Krankenhaus Eppendorf, Martinistrasse 52, D-2000 Hamburg 20, Germany.

References

  1. Miller DC, Stinson EB, Oyer PE, et al. Operative treatment of aortic dissections: experience with 125 patients over a sixteen-year period. J Thorac Cardiovasc Surg 1979;78:365-382. [Abstract]
  2. Wolfe WG, Oldham HN, Rankin JS, Moran J. Surgical treatment of acute ascending aortic dissection. Ann Surg 1983;197:738-742. [Medline]
  3. DeBakey ME, McCollum CH, Crawford ES, et al. Dissection and dissecting aneurysms of the aorta: twenty-year follow-up of five hundred twenty-seven patients treated surgically. Surgery 1982;92:1118-1134. [Medline]
  4. Doroghazi RM, Slater EE, DeSanctis RW, Buckley MJ, Austen WG, Rosenthal S. Long-term survival of patients with treated aortic dissection. J Am Coll Cardiol 1984;3:1026-1034. [Abstract]
  5. Miller DC, Stinson EB, Shumway NE. Realistic expectations of surgical treatment of aortic dissections: the Stanford experience. World J Surg 1980;4:571-578. [CrossRef][Medline]
  6. Haverich A, Miller DC, Scott WC, et al. Acute and chronic aortic dissections -- determinants of long-term outcome for operative survivors. Circulation 1985;72:Suppl:II-22. 
  7. Eagle KA, DeSanctis RW. Aortic dissection. Curr Probl Cardiol 1989;14:225-278. [CrossRef][Medline]
  8. Dinsmore RE, Willerson JT, Buckley MJ. Dissecting aneurysm of the aorta: aortographic features affecting prognosis. Radiology 1972;105:567-572. [Medline]
  9. Suchato C, Pekanan P, Singjaroen T, Sereerat P. Indication of dissecting aortic aneurysm on noncontrast computed tomography. J Comput Assist Tomogr 1980;4:115-116. [Medline]
  10. Eagle KA, Quertermous T, Kritzer GA, et al. Spectrum of conditions initially suggesting acute aortic dissection but with negative aortograms. Am J Cardiol 1986;57:322-326. [CrossRef][Medline]
  11. Godwin JD, Herfkens RL, Skioldebrand CG, Federle MP, Lipton MJ. Evaluation of dissections and aneurysms of the thoracic aorta by conventional and dynamic CT scanning. Radiology 1980;136:125-133. [Free Full Text]
  12. Vasile N, Mathieu D, Keita K, Lellouche D, Bloch G, Cachera JP. Computed tomography of thoracic aortic dissection: accuracy and pitfalls. J Comput Assist Tomogr 1986;10:211-215. [Medline]
  13. Iliceto S, Ettorre G, Francioso G, Antonelli G, Biasco G, Rizzon P. Diagnosis of aneurysm of the thoracic aorta: comparison between two non invasive techniques: two-dimensional echocardiography and computed tomography. Eur Heart J 1984;5:545-555. [Free Full Text]
  14. Godwin JD, Breiman RS, Speckman JM. Problems and pitfalls in the evaluation of thoracic aortic dissection by computed tomography. J Comput Assist Tomogr 1982;6:750-756. [Medline]
  15. Thorsen MK, San Dretto MA, Lawson TL, Foley WD, Smith DF, Berland LL. Dissecting aortic aneurysms: accuracy of computed tomographic diagnosis. Radiology 1983;148:773-777. [Free Full Text]
  16. Singh H, Fitzgerald E, Ruttley MST. Computed tomography: the investigation of choice for aortic dissection? Br Heart J 1986;56:171-175. [Free Full Text]
  17. Mintz GS, Kotler MN, Segal BL, Parry WR. Two dimensional echocardiographic recognition of the descending thoracic aorta. Am J Cardiol 1979;44:232-238. [CrossRef][Medline]
  18. Kasper W, Meinertz T, Kersting F, Lang K, Just H. Diagnosis of dissecting aortic aneurysm with suprasternal echocardiography. Am J Cardiol 1978;42:291-294. [CrossRef][Medline]
  19. Victor MF, Mintz GS, Kotler MN, Wilson AR, Segal BL. Two dimensional echocardiographic diagnosis of aortic dissection. Am J Cardiol 1981;48:1155-1159. [CrossRef][Medline]
  20. Mathew T, Nanda NC. Two-dimensional and Doppler echocardiographic evaluation of aortic aneurysm and dissection. Am J Cardiol 1984;54:379-385. [CrossRef][Medline]
  21. Iliceto S, Nanda NC, Rizzon P, et al. Color Doppler evaluation of aortic dissection. Circulation 1987;75:748-755. [Free Full Text]
  22. Hanrath P, Kremer P, Langenstein BA, Matsumoto M, Bleifeld W. Transosophageale Echokardiographie. Dtsch Med Wochenschr 1981;106:523-525. [Medline]
  23. Borner N, Erbel R, Braun B, Henkel B, Meyer J, Rumpelt J. Diagnosis of aortic dissection by transesophageal echocardiography. Am J Cardiol 1984;54:1157-1158. [CrossRef][Medline]
  24. Erbel R, Engberding R, Daniel W, et al. Echocardiography in diagnosis of aortic dissection. Lancet 1989;1:457-461. [Medline]
  25. Ballal RS, Nanda NC, Gatewood R, et al. Usefulness of transesophageal echocardiography in assessment of aortic dissection. Circulation 1991;84:1903-1914. [Free Full Text]
  26. Amparo EG, Higgins CB, Hricak H, Sollitto R. Aortic dissection: magnetic resonance imaging. Radiology 1985;155:399-406. [Free Full Text]
  27. Geisinger MA, Risius B, O'Donell JA, et al. Thoracic aortic dissections: magnetic resonance imaging. Radiology 1985;155:407-412. [Free Full Text]
  28. Dinsmore RE, Wedeen VJ, Miller SW, et al. MRI of dissection of the aorta: recognition of the intimal tear and differential flow velocities. AJR Am J Roentgenol 1986;146:1286-1288. [Free Full Text]
  29. Kersting-Sommerhoff BA, Higgins CB, White RD, Sommerhoff CP, Lipton MJ. Aortic dissection: sensitivity and specificity of MR imaging. Radiology 1988;166:651-655. [Free Full Text]
  30. Spielmann RP. MR bei akuter Aortendissektion. ROFO 1990;152:316-20.
  31. Goldman AP, Kotler MN, Scanlon MH, Ostrum B, Parameswaran R, Parry WR. The complementary role of magnetic resonance imaging, Doppler echocardiography and computed tomography in the diagnosis of dissecting thoracic aneurysms. Am Heart J 1986;111:970-981. [CrossRef][Medline]
  32. Dinsmore RE, Liberthson RR, Wismer GL, et al. Magnetic resonance imaging of thoracic aortic aneurysms: comparison with other imaging methods. AJR Am J Roentgenol 1986;146:309-314. [Free Full Text]
  33. Nienaber CA, Spielmann RP, von Kodolitsch Y, et al. Diagnosis of thoracic aortic dissection: magnetic resonance imaging versus transesophageal echocardiography. Circulation 1992;85:434-447. [Free Full Text]
  34. Slater EE, DeSanctis RW. The clinical recognition of dissecting aortic aneurysm. Am J Med 1976;60:625-633. [CrossRef][Medline]
  35. Levinson DC, Edmeades DT, Griffith GC. Dissecting aneurysm of the aorta: its clinical, electrocardiographic and laboratory features: a report of fifty-eight autopsied cases. Circulation 1950;1:360-387. [Medline]
  36. Anagnostopoulos CE, Prabhakar MJS, Kittle CF. Aortic dissections and dissecting aneurysms. Am J Cardiol 1972;30:263-273. [CrossRef][Medline]
  37. Schiebler M, Axel L, Reichek N, et al. Correlation of cine MR imaging with two-dimensional pulsed Doppler echocardiography in valvular insufficiency. J Comput Assist Tomogr 1987;11:627-632. [Medline]
  38. Sechtem U, Pflugfelder PW, Cassidy MM, et al. Mitral or aortic regurgitation: quantification of regurgitant volumes with cine MR imaging. Radiology 1988;167:425-430. [Free Full Text]
  39. Sutton MS, Oldershaw PJ, Miller GAH, Paneth M, Williams B, Braimbridge M. Dissection of the thoracic aorta: a comparison between medical and surgical treatment. J Cardiovasc Surg (Torino) 1981;22:195-202. [Medline]
  40. Shuford WH, Sybers RG, Weens HS. Problems in the aortographic diagnosis of dissecting aneurysm of the aorta. N Engl J Med 1969;280:225-231.
  41. Dinsmore RE, Willerson JT, Buckley MJ. Dissecting aneurysm of the aorta: aortographic features affecting prognosis. Radiology 1972;105:567-572.
  42. Guthaner DF, Miller DC. Digital subtraction angiography of aortic dissection. AJR Am J Roentgenol 1983;141:157-161. [Free Full Text]
  43. Brenot P, Pernes JM, Angel C, Seurot M, Hallali P, Gaux JC. Contribution of intravenous digital subtraction angiography to the diagnosis of aortic dissection. Ann Radiol 1984;27:5-14.
  44. Daily PO, Trueblood HW, Stinson EB, Wuerflein RD, Shumway NE. Management of acute aortic dissections. Ann Thorac Surg 1970;10:237-247. [Medline]
  45. Hashimoto S, Kumada T, Osakada G, et al. Assessment of transesophageal Doppler echography in dissecting aortic aneurysm. J Am Coll Cardiol 1989;14:1253-1262. [Abstract]
  46. Moncada R, Salinas M, Churchill R, et al. Diagnosis of dissecting aortic aneurysm by computed tomography. Lancet 1981;1:238-241. [Medline]
  47. White RD, Lipton MJ, Higgins CB, et al. Noninvasive evaluation of suspected thoracic aortic disease by contrast-enhanced computed tomography. Am J Cardiol 1986;57:282-290. [CrossRef][Medline]
  48. DeSanctis RW, Doroghazi RM, Austen WG, Buckley MJ. Aortic dissection. N Engl J Med 1987;317:1060-1067. [Medline]
  49. von Schulthess GK, Augustiny N. Calculation of T2 values versus phase imaging for the distinction between flow and thrombus in MR imaging. Radiology 1987;164:549-554. [Free Full Text]
  50. Models for measuring changes. In: Bishop YMM, Fienberg SE, Holland PW. Discrete multivariate analysis: theory and practice. Cambridge, Mass.: MIT Press, 1975:257-79.
  51. Nanda NC, Pinheiro L, Sanyal RS, Storey O. Transesophageal biplane echocardiographic imaging: technique, planes, and clinical usefulness. Echocardiography 1990;7:771-788. [Medline]
  52. Dinsmore RE, Rourke JA, DeSanctis RD, Harthorne JW, Austen WG. Angiographic findings in dissecting aortic aneurysm. N Engl J Med 1966;275:1152-1157.
  53. Mugge A, Daniel WG, Laas J, Grote R, Lichtlen PR. False-negative diagnosis of proximal aortic dissection by computed tomography or angiography and possible explanations based on transesophageal echocardiographic findings. Am J Cardiol 1990;65:527-529. [Medline]
  54. Hirst AE Jr, John VJ Jr, Kime SW Jr. Dissecting aneurysm of the aorta: a review of 505 cases. Medicine (Baltimore) 1958;37:217-279. [Medline]
  55. Jex RK, Schaff HV, Piehler JM, et al. Repair of ascending aortic dissection: influence of associated aortic valve insufficiency on early and late results. J Thorac Cardiovasc Surg 1987;93:375-384. [Abstract]
  56. Cheitlin MD. Commentary on: Anderson MW, Higgins CB. Should the patient with suspected acute dissection of the aorta have MRI, CAT scan, or aortography as the definitive study? In: Cheitlin MD, Brest AM, eds. Dilemmas in cardiology. Philadelphia: F.A. Davis, 1990:293-306.
  57. Paulin S, von Schulthess GK, Fossel E, Krayenbuehl HP. MR imaging of the aortic root and proximal coronary arteries. AJR Am J Roentgenol 1987;148:665-670. [Free Full Text]
  58. Hansmann A, Storck M, Kaltenecker A, Gockel M, Lauterjung L. Akute Aortenruptur bei vorbestehender Aortendissektion Typ B. Cor Vasa 1990;4:173-178.
  59. Silvey SV, Stoughton TL, Pearl W, Collazo WA, Belbel RJ. Rupture of the outer partition of aortic dissection during transesophageal echocardiography. Am J Cardiol 1991;68:286-287. [CrossRef][Medline]
  60. Hart WL, Berman EJ, LaCom RJ. Hazard of retrograde aortography in dissecting aneurysm. Circulation 1963;27:1140-1142. [Free Full Text]
  61. Danza FM, Fusco A, Falappa P. The role of computed tomography in the evaluation of dissecting aortic aneurysms. Radiology 1984;152:827-829. [Medline]
  62. Laas J, Schluter G, Daniel W, Hendrickx P, Haverich A. Acute type-A dissection of the aorta: which diagnostic modes remain for surgical indication? Eur J Cardiothorac Surg 1987;1:169-172. [Abstract]
  63. Schopp D, Wimmer B. Die heutige Diagnostik des disseziierenden Aortenaneurysmas. Radiologe 1989;29:237-244. [Medline]

 

This Article
-Abstract

Commentary
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation

Related Letters:

The Diagnosis of Thoracic Aortic Dissection by Noninvasive Imaging Procedures
Hartnell G., Costello P., Goldstein S. A., Lindsay J., Vasan R., Nienaber C. A., Spielmann R. P., Cigarroa J. E., Eagle K. A., Isselbacher E. M.
Extract | Full Text  
N Engl J Med 1993; 328:1637-1638, Jun 3, 1993. Correspondence

This article has been cited by other articles: