this injury experimentally using acute deceleration mod-els. In fact, studies have shown that the tensile strength ofthe aorta exceeds the gravitational forces generated invehicular trauma [2]. Parmley and colleagues also ex-pressed doubt that the usual deceleration forces in caraccidents were sufficient to rupture the aorta [1]. Thewhiplash theory fails to explain the other sites of aorticrupture; aortic lacerations proximal and distal to theisthmus, as well as injuries to the roots of the greatvessels, can also occur [3]. In addition, investigatorsdisagree as to which part of the aorta, the aortic arch [4]or descending aorta [1], is mobile on impact.

The osseous pinch mechanism is based on normalphysiologic motion of the bony thorax, including therotation of the ribs about their vertebral articulations.Blunt trauma to the chest causes the manubrium, first rib,and the medial clavicles to rotate posteriorly and inferi-orly about the axes of the posterior rib articulations. Thismovement compresses the vascular structures betweenthe anterior osseous elements and the vertebral column.

Crass and coworkers [5], in an experimental model ofthoracic compression, showed that compression of thearticulated skeleton in an antero-posterior direction re-sulted in postero-inferior displacement of the manu-brium, first rib, and medial clavicle, which impinged onand caused near-transection of the aortic model. Thelevel of injury was just distal to the subclavian artery.Cohen and associates [6], using calculations made fromthe cross-sectional anatomy shown on computed tomo-graphic scans, demonstrated the point of aortic impinge-ment and provided further evidence in support of theosseous pinch mechanism. Our case report clearly sup-ports the osseous pinch theory. The transected aorta inour patient is unlikely to have resulted from any mech-anism other than the impinging bony structures. How-ever, it is not possible to conclude that “osseous pinch” isthe mechanism of injury in all cases of aortic transection.

References

1. Parmley LF, Mattingly TW, Manion WC, et al. Non-penetrating traumatic injury of the aorta. Circulation 1958;17:1086–101.

2. Greendyke RM. Traumatic rupture of aorta. JAMA 1966;195:119–22.

3. Fisher RG, Hadlock F, Ben-Menachem Y. Laceration of thethoracic aorta and brachiocephalic arteries by blunt trauma.Radiol Clin North Am 1981;19:91–110.

4. Zehnder MA. Delayed post-traumatic rupture of the aorta ina young healthy individual after closed injury: mechanical-etiological considerations. Angiology 1956;7:252–67.

5. Crass JR, Cohen AM, Motta AO. A proposed new mechanismof traumatic aortic rupture: the osseous pinch. Radiology1990;176:645–9.

6. Cohen AM, Crass JR, Thomas HA. CT evidence for the“osseous pinch” mechanism of traumatic aortic injury. AJRAm J Roentgenol 1992;159:271–4.

Successful Repair of IntraoperativeAortic Dissection Detected byTransesophageal EchocardiographyDavid Varghese, MB, ChB, Bernhard J. C. J. Riedel,FCA, S. Nicholas Fletcher, FRCA, Mohammed I. Al-Momatten, FRCS, and Asghar Khaghani, FRCS

Department of Cardiothoracic Surgery and Anesthetics,Harefield Hospital, Middlesex, United Kingdom

Aortic dissection is a rare but devastating complication ofcardiac surgery. Early intraoperative diagnosis and man-agement are essential for a favorable outcome. We de-scribe the case of a 69-year-old man with worseningdyspnea who was admitted for mitral valve replacementhaving previously had a mitral valve repair. Pre-cardiopulmonary bypass transesophageal echocardiogra-phy confirmed mitral regurgitation and showed mildatherosclerotic changes in the descending aorta. Follow-ing successful replacement of the mitral valve, an attemptto wean from cardiopulmonary bypass failed. This wascharacterized by acute onset hypovolemia. The trans-esophageal echocardiography showed the presence offeatures of acute aortic dissection involving only thedescending aorta without identifying the entry point.The tear was successfully repaired by direct suturewithin the lumen.

(Ann Thorac Surg 2002;73:953–5)© 2002 by The Society of Thoracic Surgeons

Aortic dissection is a rare complication of open car-diac surgery that needs early recognition and

repair. We report an unusual case of intraoperativeaortic dissection with an intimal tear in the descendingaorta and its diagnosis by transesophageal echocardiog-raphy.

A 69-year-old man who underwent mitral valve repair forposterior mitral valve leaflet prolapse 2 years previouslypresented with a short history of worsening dyspnea. Hewas admitted for mitral valve replacement. He had nohistory of peripheral vascular disease, hypertension, orhypercholesterolemia. On examination, he was noted tohave long-standing atrial fibrillation, ankle edema, andreduced breath sounds on the right base. The preopera-tive chest radiograph showed cardiomegaly and a locu-lated pleural effusion on the right side. A transthoracicechocardiograph demonstrated moderate mitral regurgi-tation and mild tricuspid regurgitation, with good left

Accepted for publication April 21, 2001.

Address reprint requests to Dr Varghese, 20 Lytham Close, Thamesmead,London, SE 28 8QH, England; e-mail: david@varghese.freeserve.co.uk.

953Ann Thorac Surg CASE REPORT VARGHESE ET AL2002;73:953–5 INTRAOPERATIVE AORTIC DISSECTION

© 2002 by The Society of Thoracic Surgeons 0003-4975/02/$22.00Published by Elsevier Science Inc PII S0003-4975(01)02867-326

ventricular ejection. The aortic root and valve were nor-mal. His warfarin therapy was stopped 6 days previouslyand anticoagulation was maintained with a heparin in-fusion. Other regular medication included digoxin,amiloride, and bumetanide.

Anesthesia consisted of a balanced low dose opioid-volatile (fentanyl-isoflurane-propofol) general anesthetictechnique. Monitoring was facilitated by a left radialartery cannula, a right internal jugular central venouscatheter, pulmonary artery catheter, and a transesopha-geal echocardiograph (TEE). The precardiopulmonarybypass TEE examination revealed mobile mitral valvecusps with a moderately severe central regurgitant jetsecondary to ruptured chordae. The left atrium wasdilated (8.5 cm) with systolic pulmonary artery pressuresestimated at 35 mm Hg. Left ventricular function wasnormal. The ascending aorta was normal but there weregrade I-II atheromatous changes in the descending aorta.

Surgery was commenced through a median sternot-omy. Cardiopulmonary bypass (CPB) was institutedthrough the ascending aorta and bicaval cannulation.The aortic line pressure was normal during the whole

procedure. Left atrial and mitral annular dilatation werenoted on inspection. The anterior mitral leaflet was foundto be partially prolapsed with ruptured chordae. Themitral valve was replaced with a CarboMedics 31 mmtype M (Sulzer Carbomedics Limited, UK). After theprocedure, while still on CPB, TEE examination showedanterior displacement of the left atrium by the descend-ing aorta (Figs 1, 2, 3) and increased shadowing aroundthe descending aorta from the level of the left subclavianartery to the level of the diaphragm. No intimal tearcould be detected on TEE. In addition to echocardio-graphic findings, there was acute volume loss requiringextra volume to be added to the CPB reservoir repeat-edly. There was also increasing abdominal distension,with anterior displacement of the mediastinum that co-incided with an attempt to wean the patient off CPB.There were immediate signs of hypovolemia. Inspectionof the mitral valve using color flow doppler indicatedgood valve function. There was no visible sign of ascend-

Fig 1. Horizontal plane of descending aorta showing dissection.

Fig 2. Horizontal plane of descending aorta showing compression ofleft atrium and appendage by hematoma.

Fig 3. Horizontal view of aorta showing normal relationship be-tween left atrium and descending aorta.

Fig 4. Computed tomographic scan (7 months postop) showing re-solving hematoma.

954 CASE REPORT VARGHESE ET AL Ann Thorac SurgINTRAOPERATIVE AORTIC DISSECTION 2002;73:953–5

ing aortic involvement. A presumptive diagnosis of dis-section of either the distal aortic arch or descending aortawas made. CPB was reinstituted by placing the arterialcannula in the left femoral artery. After cooling to 20°C,total circulatory arrest was established. Intermittent ret-rograde perfusion to the cerebral circulation was main-tained through the superior vena cava line. The leftpleura was opened, but no blood was found in thepleural cavity. A large hematoma was however palpatedaround the whole length of the descending aorta com-patible with TEE findings (Fig 4). Examination of theaortic cannulation site and ascending aorta up to theproximal arch revealed no abnormality or external signsof dissection. The ascending aorta was opened at the siteof cannulation. A 2.5-cm transverse tear was found in thedistal arch just opposite the origin of the left subclavianartery. The entry site was repaired from within the lumenwith Teflon (Impra Inc., Subsidiary of L.R. Bard, Tempe,AZ) pledgetted interrupted Prolene (Ethicon, Somerville,NJ) sutures. The patient was rewarmed and successfullyweaned from CPB. The cross-clamp time was 59 minutes,and the circulatory arrest time was 22 minutes. Residualhemodynamic instability precluded sternal closure atthis time. The chest was stented and a delayed sternalclosure was performed 3 days later.

Immediately postoperatively, the patient had a com-puted tomographic scan. This showed an aortic dissec-tion extending to the abdominal aorta involving bothrenal and mesenteric vessels with a large retroperitonealhematoma. The early postoperative period was charac-terized by over 2 liters of drainage from the chest tubeswithout a decrease in hematocrit. The drainage wasthought to be from the posterior mediastinum.

A computed tomographic scan prior to dischargeshowed no aortic luminal compromise, no intimal flap,and resolving mediastinal and retroperitoneal hema-toma. The patient was discharged from hospital 32 daysafter surgery. He had no shortness of breath or neuro-logic deficit at the 6-week follow-up.

Comment

Intraoperative aortic dissection is a rare but catastrophiccomplication of cardiac surgery with a well-describedpathophysiology. In one study, the frequency of iatro-genic aortic dissection was 0.16% [1]. The site of dissec-tion as seen in our patient (descending aorta) is very rare.

There are two types of clinical presentations regardingiatrogenic aortic dissection [2]. Type I is that seen in ourpatient where the aortic dissection presents intraopera-tively [2]. Type II are those patients who present withaortic dissections from any time during the postoperativeperiod up to 20 years after cardiac surgery [2, 3].

There was evidence of grade I to II atherosclerosis inthe descending aorta. Atherosclerosis is a known predis-posing factor for development of aortic dissection [1]. We

postulate that a jet of blood from the aortic cannula mayhave caused an intimal tear and subsequent dissection inthe already atherosclerotic aorta as suggested by thetransesophageal echocardiogram. In this case, the TEEfindings were very helpful in suspecting the presence ofacute dissection while still on CPB. Mortality increaseswith a delay in diagnosis. Still and colleagues, in aretrospective study of 14,877 operative procedures,showed mortality to be 20% for dissections discoveredintraoperatively and 50% for dissections discovered post-operatively [1]. This confirms the value of routine preop-erative TEE in cardiac surgery. Pericardiopulmonary by-pass TEE would also greatly aid in the placement of aorticcannula away from sites of aortic plaques and avoidanceof aortic cross-clamping at these sites. Once diagnosed,there are two types of repair: a conservative surgery withprimary suture [1, 4] as done in our patient, or a patch-tube graft interpositioning. The extent and the location ofthe intimal tear determines the type of repair. It isrelatively easy to repair ascending aortic dissections, butit becomes technically more challenging when the dis-section is in the arch or descending aorta. In our patient,the tear was distal and the only simple way was to repairthe tear from within the lumen by direct suture. The sizeof the periaortic hematoma was very extensive in thiscase, extending from the posterior mediastinum to theretroperitoneal area due to lack of reentry point. Thiscould result in compression of aortic branches such asthe mesenteric vessels, which may have been the cause ofthe acute pancreatitis in this case.

In conclusion, we suggest that in circumstances ofacute aortic dissection due to a limited intimal tear,simple primary repair is the procedure of choice partic-ularly when the access to the area is restricted. Also, TEEis a useful tool in the preoperative diagnosis of aorticdissection and guidance of cannula placement [5].

References

1. Still RJ, Hilgenburg AD, Atkins CW, Daggart WM, BuckleyMJ. Intraoperative aortic dissection. Ann Thorac Surg 1992;53:374–80.

2. Ruchat P, Hurni M, Stumpe F, Fischer AP, von Segesser LK.Acute ascending aortic dissection complication open heartsurgery: cerebral perfusion defines the outcome. Eur J Car-diothorac Surg 1998;14:449–52.

3. Orszulak TA, Pluth JR, Schaff HV, Piehler JM, Smith HC,McGoon DC. Results of surgical treatment of ascending aorticdissections occurring late after cardiac operation. J ThoracCardiovasc Surg 1982;83:538–45.

4. Leclercq F, Albat B, Messner-Pellenc P, et al. Successfulconservative surgery of acute ascending aortic dissectionoccurring during coronary angiography. J Cardiovasc Surg(Torino) 2000;41:61–3.

5. Anderson C, Joyce FS, Tingleff J, Arendrup H. Aortic dissec-tion after cardiopulmonary bypass detected by intraoperativetransesophageal echocardiography. Acta Anaesthesiol Scand1997;41:1227–8.

955Ann Thorac Surg CASE REPORT VARGHESE ET AL2002;73:953–5 INTRAOPERATIVE AORTIC DISSECTION