Penetrating atherosclerotic ulcers of the thoracic aorta

Penetrating atherosclerotic ulcers of the thoracic aorta Sajjad Hussain, M D , J o h n L. Glover, M D , R o b e r t Bree, M D , and Phillip J. Bendick, Phi), Royal Oak, Mich. Penetrating ulcer of the thoracic aorta is defined as an atherosclerotic lesion of the descending thoracic aorta with ulceration that penetrates the internal elastic lamina, allowing hematoma formation in the media. There is controversy whether this lesion differs from classic acute type III aortic dissection, based on its location, radiographic findings, natural history, and recommended therapeutic approach. Of 47 patients with a diagnosis of aortic dissection seen at our hospital during a 2-year period, five patients had clinical and radiographic findings of penetrating ulcer. Each of the five patients had characteristic computerized tomographic (CT) findings and two patients had angiographic confirmation. In all patients CT showed subintimal hemorrhage, aortic wall enhancement, absence of a double lumen, and contrast extravasation through the ulceration. In both patients who underwent angiography, ulceration, subintimal hematoma, and absence o f a false lumen were demonstrated. The clinical presentation in four patients simulated acute aortic dissection or expanding thoracic aneurysm. The other patient, w h o was normotensive, did not have symptoms referable to the thoracic aorta but was studied because o f an abnormal chest x-ray film. None o f these five patients required surgical intervention. All five patients were alive and free of symptoms at 6 months, 8 months, 14 months (two patients), and 30 months after the original diagnosis. Follow-up CT scans in four patients showed resolution of subintimal hematoma and some dilatation of the lumen but no progression to rupture or aneurysm. Other authors stress the importance of differentiating symptomatic penetrating atherosclerotic ulcers from acute type III aortic dissection because of the higher incidence of rupture of penetrating ulcers and therefore recommend early surgical intervention. Our series suggests that rupture of symptomatic atherosclerotic ulcers of the thoracic aorta may be less common than previously thought and that nonoperative management is appropriate initially, followed by serial CT evaluation and clinical foUow-up. (J VAsc SURG 1989;9:710-7.) Intimal defects resulting from atheroscleroric ulcers occur in patients with advanced atherosclerosis, and the most common location is probably the carotid bifurcation. These lesions have also been described in association with aortic dissection, but their significance is not known. 13 Recently Stanson et al. 4 and Cooke et al. s reported 16 patients who had atheromatous ulcerations in the descending thoracic aorta that penetrated the internal elastic lamina and resulted in the formation o f hematomas in the media. 4,5 Seven of their patients who underwent surgery because of symptoms that recurred after initial medical management had contained rupture; the authors From the Departments of Surgery (Drs. Hussain, Glover, and Bendick) and Radiology (Dr. Bree), William Beaumont Hospital. Presentedat the TwelfthAnnualMeetingof the MidwestemVascular SurgicalSociety,Rochester,Minn., Sept. 23-24, 1988. Reprint requests: John L. Glover,MD, Department of Surgery, William Beaumont Hospital, 3601 W. Thirteen Mile Rd., Royal Oak, MI 48072. 710 advised that this entity be regarded as more aggressive than dissecting aneurysms, recommending surgery as the standard treatment. Because that recommendarion is based on only one series it is imperative that the natural history of penetrating ulcers be verified by others. The purpose of this report is to review our experience with five patients encountered during a 2-year period and evaluated prospectively since then. METHODS Patients who demonstrated characteristic findings of penetrating atherosclerotic ulcers of the descending thoracic aorta on computerized tomography (CT) or angiography between April 1985 and April 1987 were selected for this study. Their records were reviewed, and the patients themselves and their physicians were interviewed to obtain follow-up informarion. When possible follow-up radiographs were obtained. Volttme 9 Number 5 May 1989 Penetrating atheroscleroticulcers 711 Fig. 1. CT scan of the chest of patient 1. The descending thoracic aorta appears dilated with an atheromatous ulcer (black arrow) and intramural hematoma (white arrow). Fig. 2. CT scan of the chest of patient 1 taken 4 weeks after discharge. There is dilatation of the aortic lumen and diminution in the size of the intramural hematoma (arrow). CASE HISTORIES Patient 1. A 74-year-old white woman was admitted to the hospital with an acute onset of dyspnea and chest pain radiating to her back. No prior cardiac history was noted; she had a 60 pack-year history of smoking. On examination her blood pressure was 220/130 m m H g in the right arm and 210/130 turn H g in the left arm. Peripheral pulses were equal and not diminished. A systolic ejection murmur was audible. An admission chest radiograph showed athero- sclerotic changes of t h e aorta and a left pleural effusion. An ECG showed flat T waves in leads 1, AVL, V5, and V6 and Q waves in leads V1 and V2. No prior ECG was available for comparison. There was no evolution of the ECG changes or elevation of the creatine phosphokinase (CPK) isozymes during this hospitalization. Pulmonary embolism was suspected, but a ventilation-perfusion scan was negative. The persistence ofmidthoracic pain made it necessary to obtain a radiograph of the spine, which showed a soft tissue 712 Hussain et aI. Journal of VASCULAR SURGERY Fig. 4. Thoracic aortogram of patient 2 shows an aneurysm of the descending thoracic aorta with a focal atheromatous ulcer (arrow). Fig. 3. Admission chest radiograph of patient 2. Note the soft tissue density adjacent to the descending thoracic aorta (arrow). mass in the left paraspinal area. A C T scan of the chest showed a dilated descending thoracic aorta with an atheromatous ulcer and an intramural hematoma (Fig. 1). The patient was transferred to the coronary care unit and treated with propranolol; later, digoxin was added because of the onset of atrial fibrillation. The blood pressure was maintained below 140/70 mm H g and her symptoms gradually subsided. The patient was discharged on the thirteenth day continuing with propranolol and digoxin therapy. A follow-up outpatient CT scan performed 4 weeks after discharge showed some dilatation of the lumen and diminution in the size of the intramural hematoma (Fig. 2). The patient has remained free of significant symptoms for 21/2 years. Patient 2. A 67-year-old black man with a history of chronic azoternia and insulin-dependent diabetes was admitted to the hospital because of suspected venous thrombosis in the left leg. Blood pressure was normal. Admission chest radiograph showed a soft tissue density adjacent to the descending thoracic aorta (Fig. 3). After lower extremity venous ultrasonography confirmed the admitting diagnosis, the patient was treated with heparin and then sodium warfarin (Coumadin). However, anticoagulants were discontinued because of upper gastrointestinal hemorrhage. Upper endoscopy showed a Mallory-Weiss tear. A C T scan of the chest showed an area of focal subintimal hemorrhage without demonstration of an ulcer or plaque. A thoracic aortogram showed an aneurysm of the descending thoracic aorta with a focal atheromatous ulcer (Fig. 4). After placement of a Greenfield filter the patient was managed nonoperatively. He remained normotensive and was discharged on the fifteenth hospital day. He was symptom free when evaluated 6 months after diagnosis. An outpatient CT scan done at that time showed a residual focal periaortic bulge compatible with a chronic subintimal hematoma and saccular aneurysm. The patient remains well and free of symptoms. Patient 3. A 72-year-old white woman with a long history of hypertension was seen with an acute onset of epigastric pain radiating to both flanks and the back. Her blood pressure was 212/132 mm Hg. There was no jugular venous distension, and no pulse deficit or heart murmur was noted. Her leukocyte count was 16,500 with normal serum amylase, liver enzyme, and CPK isozyme values. The ECG did not show acute changes. Hepatobiliary scan done be- Volume 9 Number 5 May 1989 Penetrating atherosclerotic ulcers 713 Fig. 5. CT scan of the chest of patient 3. There is an atheromatous ulcer in the descending thoracic aorta (black arrow) with an intramural hematoma (white arrow). cause o f the possibility of acute cholecystitis showed no obstruction of the cystic duct. An abdominal ultrasound scan showed cholelithiasis. On repeat chest radiograph a new left pleural effusion was noted. A ventilation-perfusion scan for suspected pulmonary embolism was suggestive, but a pulmonary angiogram showed no embolism. The pain persisted, and left pleurocentesis was performed, with aspiration of 550 ml of serous fluid. A CT scan showed an atheromatous ulcer in the descending thoracic aorta with an intramural hematoma; these findings were confirmed by aortography (Figs. 5 and 6). The patient was transferred to the intensive care unit where she required intravenous labetalol hydrochloride for a short period to control her blood pressure. She was later given oral propranolol, and her pain subsided in 48 hours. A GT scan was repeated in 1 week and showed diminution in the size of the hematoma with persistent wall thickening. However, the patient improved clinically and was discharged taking 13-blockers on the fourteenth hospital day. She had no recurrence of chest pain, and GT scan performed approximately 11 weeks after discharge from the hospital showed slight dilatation of the aortic lumen with resolution o f the hematoma (Fig. 7). Patient 4. A 70-year-old white woman was seen in the emergency room with epigastric, lower thoracic, and back pain for 24 hours' duration. She had a history o f hypertension and cigarette smoking. Her blood pressure on admission was 170/100 mm H g Fig. 6. Thoracic aortogram of patient 3 confirms the presence of a penetrating atheromatous ulcer (arrow). in the right arm and 180/100 mm H g in the left arm. The patient's peripheral pulses were equal bilaterally. Cholecystitis was suspected. An abdominal ultrasound scan showed cholelithiasis but a hepatobiliary scan showed no cystic duct obstruction. Chest x-ray Journal of VASCULAR SURGERY 714 Hussain et al. Fig. 7. CT scan of the chest of patient 3 taken approximately 11 weeks after discharge. The aortic lumen remains slightly dilated with resolution of the hematoma (arrow). film showed tortuosity of the aorta and bilateral pleural effusions. Because of increasing back pain a CT scan was done, which showed atherosclerotic ulceration in the descending thoracic aorta with a subintimal hematoma but no contrast extravasation. The patient was given sodium nitroprusside and propranolol in the intensive care unit to control her blood pressure. After clinical improvement and discontinuation of parenteral antihypertensive drugs the patient was discharged taking oral propranolol. There was no recurrence of chest pain during an 8-month follow-up and outpatient CT scans at 8 weeks and 8 months showed a slight increase in luminal diameter of the aorta with decrease in size of the hematoma. Patient 5. An 81-year-old white man with a past history of chronic dementia, hypertension, and myocardial infarction was admitted to the hospital after falling downstairs. Temporary loss of consciousness was suspected. On examination he was awake and his blood pressure was 120/76 mm Hg. Gross hematuria was noted. Injuries included fractures of the right acetabulum, pubic rami, and posterior iliac wing and a compression deformity of the twelfth thoracic vertebra. Intravenous pyelography showed a mass displac- ing the fight ureter and ultrasonography confirmed a pelvic hematoma. The ultrasound scan also showed a dilated distal thoracic aorta with subintimal hemorrhage. A C T scan was performed and this showed atherosclerotic ulceration of the thoracic aorta with a subintimal hematoma. During an 18-day hospital stay the patient was treated for the fractures. He required blood transfusions (2 units); a cystoscopy for continuing hematuria revealed no gross abnormalities, and this eventually resolved. A single episode of chest pain and dyspnea resolved with sublingual nitroglycerine. The patient was able to walk and was finally discharged to a nursing home. No chest symptoms were noted during a 14-month follow-up. RESULTS Five patients had clinical and radiographic findings of penetrating ulcer of the thoracic aorta during this 2-year period. Forty-seven patients with a diagnosis of aortic dissection were seen at our hospital during the same period. Four of these five patients were hypertensive and had symptoms similar to those of acute aortic dissection or expanding thoracic aneurysm. Other clinical problems, such as gallbladder disease and pulmonary embolus, were also considered by the attending physicians, as is frequently the case in Volume 9 Number 5 May 1989 Penetrating atherosclerotic ulcers 715 Fig. 8. Thoracic aortograms compare characteristic findings of penetrating ulcer (left) and type III dissection (right). Penetrating ulcers (arrow) are found in the descending thoracic aorta, distant from the origin of the left subclavian artery. Type III dissections generally begin as intirnal lacerations beginning just distal to the origin of the subclavian artery (black arrow) and show a "false" or double lumen (white arrow). Fig. 9. CT scans compare characteristic findings of penetrating ulcer (left) and type III dissection (right). The penetrating ulcer appears as a local outpouching of contrast within the lumen (solid arrow) with surrounding intramural hematoma (arrowhead). The typical double lumen created by the dissected intima (arrow) is seen on the r/ght. dissecting aneurysms. The patients underwent further investigation including CT and angiography because o f persistent symptoms. However, one patient did not have symptoms referable to the thoracic aorta and was normotensive. He was studied because o f an abnormal chest x-ray film. In aU these cases the clinical diagnosis was type III aortic dissection. In all patients CT showed subintimal hemorrhage, aortic wall enhancement, and absence of a double lumen of the thoracic aorta. The penetrating ulcer was seen clearly in three patients. In both patients who underwent angiography, ulceration, subintimal hematoma, and absence of a false lumen were demonst_rated. All patients were monitored in the intensive care unit. Aggressive antihypertensive therapy was used in all patients who had elevated blood pressure. In some cases this included intravenous labetalol hydrochloride and sodium nitroprusside. None of these five patients needed surgical intervention. All five patients were alive and free of symptoms at 6 months, 8 months, 14 months (two patients), and 30 months after the original diagnosis was made. Four patients were able to assume an active life-style, and one was in a nursing home. Follow-up GT scans in four patients showed resolution of subintimal hematoma and some dilatation of the lumen, but no progression to rupture. Journal of VASCULAR SURGERY 716 Hussain et al. DISCUSSION The usual concern about arterial intimal ulceration is embolism, from either atheromatous debris or thrombi trapped in the ulcer. There are very few case reports of ulcers that extend into the media. 6,7 Stanson et al.~ reported 16 cases found in a review of 684 aortograms, about 2.3%. Eyler and Clark2 found only one case in their series of 46 patients with aortic dissection. Our apparently higher incidence may be happenstance or may be related to the dedicated search for such ulcers. The reason for such a search was to help determine whether this lesion is a clinical entity separate from type III aortic dissection and whether it should be treated differently, as suggested by Stanson et al. It seems reasonable to assume that ulceration that penetrates the media would be followed by "dissection" of blood into the layers of the vessel; in fact in the past this mechanism was considered to be the cause of dissecting aneurysms by Virchow and other pathologists, s'9 Subsequent studies of pathologic specimens, however, showed that atherosclerotic plaques were rarely present at the site of initial dissection~'~°; it is generally accepted that an intimal laceration is the precipitating event. TM In addition in type III dissections these lacerations are located just distal to the origin of the left subclavian artery. All of the penetrating ulcers in Stanson et al's. 4 series, as well as in ours, were in the descending thoracic aorta, well beyond the origin of the subclavian artery. Furthermore none of the patients in either series developed a "false lumen" or double lumen. In fact studies of some of our patients showed healing, manifest by resolution of the hematoma and some dilatation of the aorta. However, a false lumen is characteristic of dissection; it undergoes thrombosis and organization in only about 10% of cases. These differences between these entities are illustrated in Figs. 8 and 9. Finally one of our patients was normtensive, had no history of trauma, and had no clinical manifestations of Marfan's syndrome or other connective tissue disorders. Dissecting aneurysm would be most unusual in such a patient. Based on these data we believe that penetrating atheromatous ulcers are a distinct clinical entity that should be differentiated from aortic dissection. The question of treatment is another matter. Stanson et al.4 recommended operation because of the propensity to rupture. Their recommendation was based on findings in the 14 (of 16) patients who underwent operation: 10 had intramural hematoma, with four of these having contained rupture at the site of the penetrating plaque. Three additional pa- tients had no extension of the intramural hematoma "beyond the confines of the penetrating ulcer" but also had contained rupture at the site of an atherosclerotic plaque, presumably a different, nonulcerated plaque. In the remaining patient who underwent surgery there was a false aneurysm, but no specific comment about the intramural hematoma. Two patients, aged 79 and 83 years, did not undergo surgery. The former refused surgery and was lost to follow-up; the latter had no surgery because she was elderly and symptom free. Based on these findings one can certainly understand the recommendation for surgery, especially because most (and perhaps all) of their patients were referred from other institutions. However, these same findings may support a recommendation for observation just as well. After all no patient died of frank rupture, and the long-term result of contained rupture is not known, although it is apparent that some patients harbor contained. rupture of abdominal aneurysms for months at least. 12 Even less is known about the fate of intramural hematoma, although it is well established that the conditions of some patients with true aortic dissection remain stable for years) 3 One might argue that a confined intramural hematoma is even more likely to remain stable because it is not subject to the hemodynamic changes associated with aortic flow in a double lumen. It is likely that false aneurysms will develop in some patients, but longer follow-up is needed to determine whether they will be a majority or a minority of the patients. In summary, we believe that penetrating aortic ulcer is a different pathologic entity from type III dissection, and we encourage others to search for it to help clarify its natural history. When such a patient is seen we advise appropriate treatment of the hypertension if it is present. If the diagnosis has been made by angiography, CT scan may be useful to look for contained rupture, because these patients are probably at greatest risk of dying of exsanguination. In any case observation may be indicated as the initial treatment, depending on the age of the patient and the severity of the associated medical problems. REFERENCES 1. Gore I, Hisrt AE Jr. Dissecting aneurysm of the aorta. Cardiovasc Clin 1973;2:239-60. 2. Eyler WR, Clark MD. Dissecting aneurysms of the aorta: roentgen manifestations including a comparison with other types of aneurysms. Radiology 1965;85:1047-57. 3. Stein HL, Steinberg I. Selective aortography, the definitive technique for diagnosis of dissecting aneurysm of the aorta; AJR 1968;102:333-48. Volume 9 Number 5 May 1989 4. Stanson AW, Kazmier FJ, HoUier LH, et al. Penetrating atherosderotic ulcers of the thoracic aorta: natural history and clinicopathologic correlations. Ann Vasc Surg 1986;1:15-23. 5. Cooke JP, Kazmier FJ, Orszulak TA. The penetrating aortic ulcer: pathologic manifestations, diagnosis, and management. Mayo Clin Proc 1988;63:718-25. 6. Lord RSA, Berry NA. Atherosclerotic ulceration of the brachiocephalic artery. Aust NZ J Surg 1974;44:370-4. 7. Nora JD, HoUier LH. Contained rupture of the suprarenal aorta. J VASCSURG1987;5:651-4. 8. Slater EE, DeSanctis RW. Theclinical recognition of dissecting aortic aneurysm. Am J Med 1976;60;625-33. 9. Shennan T. Dissecting aneurysms. Special report No. 193, Med Res Council (Gr Britain), 1933. DISCUSSION Dr. Tony Stanson (Rochester, Minn.). This is an entity that must be appreciated a little bit differently from a classic dissection. I cannot determine whether these patients need surgery, because I am a radiologist, but it has occurred to me since I first recognized this lesion in 1974 that it is potentially dangerous. We already had seen two of these ulcers rupture fatally, and yet a third patient still is alive 10 years after surgical exploration during which the source of the periaortic hematoma never was really identiffed. Perhaps an operation is not necessary for every patient; maybe the presence o f pain is the most important consideration in that regard. It is critically important to have a noncontrast scan to determine a diagnosis of hematoma. One just does not appreciate the high density o f the hematoma after contrast material is administered. Dr. John Joyee (Rochester, Minn.). The natural history o f this curious lesion obviously has been understood better since we have had the advantage of computerized tomographic (CT) scanning, magnetic resonance imaging, and more careful interpretation o f three-view angiograms. The observations presented this morning were well documented and carefully presented, so they also will be o f great help in the future. I believe that these ulcers have a spectrum of events. We have seen four of these radiographic lesions in the ascending aorta with associated chest pain. Two patients died o f rupture into the pericardium, a third underwent correction in the presence of hemopericardium, and the fourth underwent a successful operation before hemopericardium had occurred. We have seen four additional lesions in the abdomen. One ruptured fatally before any surgery could be done, and two patients survived operations after contained ruptured. In addition to our apparently severe cases, however, Penetrating atherosderotic ulcers 717 10. Larson EW, Edwards WD. Risk factors for aortic dissection: a necropsy study of 161 cases. Am J Cardiol 1984;53:84955. 11. Roberts WC. Aortic dissection: anatomy, consequences and causes. Am Heart J 1981;101:195-214. 12. Jones CS, ReiUyMK, Dalsing MC, Glover JL. Chronic contained rupture of abdominal aortic aneurysms. Arda Surg 1986;121:542-6. 13. Wheat/vlW, Palmer RF, BartleyTD, Seelman RC. Treatment of dissecting aneurysms of the aorta without surgery. J Thorac Cardiovasc Surg 1965;50:364-9. we have heard today about several examples that have behaved in a benign fashion. All of this indicates that these ulcers have a spectrum of severity and, like conventional aortic dissections, the controversy concerning whether they require surgical treatment probably will continue for some time. Considering the improvements in the safety of operations on the descending thoracic or thoracicoabdominal aorta, I tend to favor a surgical approach once these lesions have been discovered. Dr. Peter C. Pairolero (Rochester, Minn.). You have to remember that your study is retrospective and may necessarily include only those patients who have done well. What about the patient who has intense chest pain, dies before an angiogram is obtained, has no postmortem examination, and therefore is excluded from your series? As you indicated in your introductory remarks the treatment of this lesion is different from that of a standard dissection in the sense that it requires resection and graft replacement of the penetrating ulcer, rather than control of the proximal aorta. As Dr. Stanson mentioned we have missed at least one of these ulcers early in our experience because we directed OUt attention to the aortic segment near the left subclavian artery where dissections usually occur. Although the outcome of these lesions may represent a spectrum, I do not think we want to be in the position of telling our colleagues that they are benign. Dr. Hussain (closing). We agree that CT scanning is more useful than an aortogram becaiase it can identify a contained rupture. We also agree that prospective studies are necessary to define the best treatment for this entity. However, our attending physicians were concerned about the high risk for paraplegia associated with surgical management in the acute situation, because four of the 14 patients in the original article describing these lesions apparently had that complication. Therefore they elected to treat these patients medically in the first place.