The value of computed tomography in the management of symptomatic abdominal aortic aneurysms Kara H . V. Kv'flekval, M D , I r w i n M. Best, M D , R o b e r t A. Mason, M D , G. Broadie N e w t o n , M D , and Fabio Giron, M D , PhD, Stony Brook, N.T. The use of computed tomographic (CT) scanning in the diagnosis of ruptured abdominal aortic aneurysm is controversial because the delay created by the procedure, it has been argued, may increase overall mortality. However, if emergency surgery can be avoided in the medically compromised patient, surgical results may improve. To assess the value of CT scanning, we studied the 1983 to 1988 records of 65 hemodynamicaUy stable patients with abdominal aortic aneurysms, who underwent diagnostic CT scanning for acute abdominal or back pain. Twenty-one patients had a history of severe cardiac, renal, or pulmonary disease. The average duration of the examination was 63 minutes; no episodes of hypotension occurred. Subsequently, 17 of 18 patients with ruptured aneurysms had emergency surgery, with 31% morbidity and 29% mortality. Of 44 patients found to have nonruptured aneurysms, 13 had other causes for their pain, nine were not considered surgical candidates, and 24 had elective aneurysmectomies, with 8% morbidity and 0% mortality. In three patients CT scanning excluded the diagnosis of aneurysm. Additional information provided by CT scanning enhanced the safety of the perioperative management of four patients with rupture and 14 without. In conclusion, the delay imposed by obtaining a preoperative CT scan in patients with possible ruptured aneurysm did not adversely affect patient outcome, and the information obtained from it aided significantly in both preoperative and intraoperative management. (J VAse St3RG 1990;12:28-33.)
Aneurysms become symptomatic by rupture or by some process in the aneurysm wall that leads to expansion or dissection. However, in clinical practice a patient with a known aneurysm or a pulsatile abdominal mass who has new onset abdominal or back pain is considered to have a possible ruptured aneurysm until proved otherwise, irrespective of the patient's hemodynamic status at the time of evaluation. These patients are considered in this report to have a symptomatic aneurysm. The use of computed tomography (CT) in the management of patients with symptomatic abdominal aortic aneurysms has generated controversy over its clinical value. Several studies have found CT to be useful in identifying rupture both in the patient with a known aneurysm and in the patient for whom the diagnosis of abdominal aortic aneurysm is in question, 1-s but some investigators have recently arFrom the Vasofiar Divisionof the Department of Surge~, UniversityHospital, StateUniversityof New Yorkat StonyBrook. Reprint requests: Fabio Giron, MD, PhD, Vascular Division, Department of Surgery, School of Medicine, Health Sciences Center T19, State University of New York at Stony Brook, Stony Brook. NY 11794-8191. 24/1/19859 28
gued that the delay incurred by obtaining the CT scan may increase the overall mortality. 9 The issue o f ~ time is most relevant in the care of the medically compromised patient. Any delay in surgery can further increase mortality. However, if an emergency operation can be avoided in this patient populatio~ survival may ultimately improve. The information that CT provides about intraabdominal anatomy also allows the surgeon to be aware before operation o f the nature and extent o f the aneurysm and any anomalies (e.g., venous, renal), thereby enhancing the safety of the operation. To better define the potential benefits o f CTassisted diagnosis, we conducted a retrospective review of patients who presented with symptoms consistent with a possible ruptured abdominal aortic aneurysm and were evaluated by CT scanning. METHODS From 1983 to 1988, 95 patients were evaluated for a possible ruptured abdominal aortic aneurysm. Patients who were hemodynamically unstable went directly to the operating room. Those patients who were either stable or easily stabilized with crystaUoid fluid underwent a CT scan. Patients were considere~
Volume 12 Number 1 Julv 1990
CT and symptomatic abdominal aortic aneurysms 29
Fig. 1. Patient with paraaortic fluid (black arrow) and streaking in the rnesentery (white arrow). CT findings were considered positive for rupture. At surgery he was found to have an intraabdominal abscess with carcinomatosis.
to be at increased operative risk if they had concurrent cardiac disease including congestive heart failure, arrhythmias, myocardial infarction; renal disease with renal insufficiency (creatinine >2) or failure requiring dialysis; or, a history of chronic obstructive pulmonary disease or forced expiratory volume in 1 second (FEV1) < 1. Intravenous bolus contrast was used at the discretion of the attending vascular surgeon. If the diagnosis of rupture was established by CT scanning, the patient was taken directly to the operating room from the radiology department. If a~e CT findings were negative for rupture, the patient was further evaluated for other causes of his or her pain. Subsequently, the decision for aneurysm repair ,was made. The patients were retrospectively divided into four categories. Category I consisted of those patients who were hemodynamically unstable and went directly to the operating room. Category II were those hemodynamically stable patients who were found on CT scanning to have evidence of rupture, Category III were those hemodynamically stable patients who had no evidence of rupture on CT scanning, and category IV were those found on CT scanning to have no aneurysm. RESULTS
Of the 95 patients reviewed, 30 were in category I (hemodynamically unstable). They went directly to the operating room. Six of these 30 patients had a ~..rdiac arrest in the emergency depamnent and were
resuscitated on the way to surgeD,. The mortality in this group was 60%. Sixty-five hemodynamically stable patients underwent CT scanning. Twenty-one of these 65 patients (32%) were at increased operative risk. The average duration of the examination was 63 minutes, and no episodes of hypotension occurred during the procedure. There were 18 patients (28%) in category II (stable, ruptured). One patient with terminal lung cancer, when told he had a leaking thoracoabdominal aneurysm, declined surgery and subsequently died. The other 17 patients were taken immediately to the operating room. At operation one patient had no evidence of rupture of his aneurysm. This patient had recently undergone, at another hospital, a Billroth II gastrectomy and colon resection for gastric cancer invading the transverse colon. His abdominal aortic aneurysm had been found during surgery. He developed abdominal and back pain after operation and was transferred to our institution for evaluation of a possible ruptured aneurysm. The CT scan showed paraaortic fluid and streaking in the mesentery. These findings were considered consistent with rupture (Fig. 1). At surgery he was found to have an intraabdominal abscess and carcinomatosis with a desmoplastic reaction in the mesentery. He therefore presented a false-positive CT scan. Thus of the 18 patients diagnosed by CT scanning to have a ruptured aneurysm, 17 underwent immediate surgery. A rupture was confirmed at surgery in i6 of these patients, resulting in a sensitivity of 94%.
Journal of VASCULAR SURLY
30 Kvilekval et al.
Fig. 2. A, This patient had a ruptured aneurysm involving the left renal artery (note the paraaortic fluid [black arrow] and the superior mesenteric artery [white arrow] ). B, He was found to have a left inferior vena cava (arrow).
The overall mortality for category II was 29%. However, four of the 16 patients found to have ruptured aneurysms died. Three of the four had significant concurrent medical problems. One died during operation of a cardiac arrhythmia (ventricular tachycardia); two died of postoperative pulmonary and renal complications. The patient with no additional risk factors died during surgery of hemorrhage. Five of these 16 (31%) had postoperative complications (cardiac, pulmonary, renal), and two of these five patients had lower extremity ischemia, one required a revascularization procedure and the other an amputation.
The CT scan gave additional information about four patients in category II. One patient had a leftsided inferior vena cava and an aneurysm involving the left renal artery (Fig. 2). His aneurysm was repaired by means of a retroperitoneal approach, allowing the surgeon to avoid the inferior vena cava, which crossed over the aorta, and to reimplant the left renal artery. Another patient had an inflammatory aneurysm. A third patient was found to have both a thoracic aneurysm and an infrarenal abdominal aortic aneurysm. Because the CT scan showed that only the infrarenal aneurysm was leaking, only this portion was repaired. In the patient with known lung canc,~,
olume 12 umber 1 2' 1_990
CT and symptomatic abdominal aortic aneurysms 31
Fig. 3. Patient with carcinoid invading the small bowel mesentery (arram).
the knowledge that he had a leaking thoracoabdominal aneurysm, as noted above, led him to refuse surgery. Twenty-six (59%) of 44 patients in category III (stable, nonruptured) had concurrent medical problems putting them at increased operative risk. In 30 patients no cause(s) other than their aneurysm could be found for their pain. Twenty-one of these 30 patients had their aneurysm repaired on a semielective basis, with no mortalities. One patient had a postoperative pneumonia and another developed thromb~,sis of his popliteal aneurysm, requiring a femoral peroneal bypass. Total morbidity in this group was 8%. Thirteen of the 44 patients (30%) in category III were found to have another cause for their symptoms (Table I), and two of these 13 required emergency general surgical operations, one for strangulated bowel and the other for diverticulitis. The patient with diverticulitis was on chronic anticoagulation therapy for atrial fibrillation, and the knowledge that his aneurysm was not leaking allowed preoperative normalization of his coagulation parameters. Three patients in this group subsequently had elective repair of their aneurysm without incident. One patient had no evidence of rupture on CT scanning; his pain was attributed to prostatitis. The next day, while in the hospital, he developed increasing abdominal pain and hypotension. He was taken immediately to the operating room, and his ruptured aneurysm was successfully repaired. His case is con~dered to be a false-negative finding on CT scanning,
Table I. Additional diagnoses of patients with symptoms found on CT scanning not to have a ruptured aneurysm (category II1) Diagnosis Anastomotic ulcer S/P Billroth II Bile gastritis S/P Billroth II Cholecystitis Colon cancer Compression fracture Dissection with lower extremity ischm~a Diverticulitis Ischemic cofitis Renal cyst Strangulated small bowel Villous adenoma
No.ofpatients 1 1 3 1 1 1 1 1 1 1 1
even though further review of his CT scan showed no evidence of a leak. Therefore, of the 24 patients undergoing aneurysmectomy after having a CT scan outcome negative for rupture, 23 were found to have no evidence of rupture at surgery, and one ruptured in hospital, resulting in a specificity of 95%. Nine patients in category III were not considered surgical candidates because of their poor medical condition or the nature of their aneurysm or both. Three of these nine patients died of their medical problems within 2 months of being evaluated by CT scanning; one of cardiac arrhythmias and two of respiratory failure. No patients with CT findings negative for rupture died of their anemysms.
Journalof VASCULAR SUP,G ~
32 Kvilekval et al.
Fig. 4. Patient with retroperitoneal fibrosis (arrow). Table II. Additional information obtained from the CT scan Category II (n)
Category III (n)
Category I V (n)
(stable,ruptured)
(stable, nonruptured)
(no aneurysm)
Left-sided inferior vena cava (1) Inflammatoryaneurysm (1) Thoracic & abdominal aneurysmwith only the infrarenalportion leaking (1) Thoracoabdominalaneurysmin patient with terminal lung cancer (1)
Retroaortic renal vein (1) Left-sided inferior vena cava (1) Suprarenal componet to aneursym (9)
Horseshoe kidney (1) Carcinoid invading small bowel mesentery (1) Retroperitoneal fibrosis (1)
Table III. Characteristics o f patients undergoing aneurysm surgery after CT scanning Category II Category III
.....
No. of patients
Increased risk ~
Morbidity
Mortality
16 24
63% 55%
31% 8%
25% 0%
Concurrent cardiac disease including congestiveheart failure, arrhythmias,myocardialinfraction; renal disease with renal insufficiency (creatinlne >2) or failure requiring dialysis; or, a history of chronic obstructive pulmonary disease or forced expiratory volume in 1 second (FEV1)
Aneurysms become symptomatic by rupture or by some process in the aneurysm wall that leads to expansion or dissection. However, in clinical practice a patient with a known aneurysm or a pulsatile abdominal mass who has new onset abdominal or back pain is considered to have a possible ruptured aneurysm until proved otherwise, irrespective of the patient's hemodynamic status at the time of evaluation. These patients are considered in this report to have a symptomatic aneurysm. The use of computed tomography (CT) in the management of patients with symptomatic abdominal aortic aneurysms has generated controversy over its clinical value. Several studies have found CT to be useful in identifying rupture both in the patient with a known aneurysm and in the patient for whom the diagnosis of abdominal aortic aneurysm is in question, 1-s but some investigators have recently arFrom the Vasofiar Divisionof the Department of Surge~, UniversityHospital, StateUniversityof New Yorkat StonyBrook. Reprint requests: Fabio Giron, MD, PhD, Vascular Division, Department of Surgery, School of Medicine, Health Sciences Center T19, State University of New York at Stony Brook, Stony Brook. NY 11794-8191. 24/1/19859 28
gued that the delay incurred by obtaining the CT scan may increase the overall mortality. 9 The issue o f ~ time is most relevant in the care of the medically compromised patient. Any delay in surgery can further increase mortality. However, if an emergency operation can be avoided in this patient populatio~ survival may ultimately improve. The information that CT provides about intraabdominal anatomy also allows the surgeon to be aware before operation o f the nature and extent o f the aneurysm and any anomalies (e.g., venous, renal), thereby enhancing the safety of the operation. To better define the potential benefits o f CTassisted diagnosis, we conducted a retrospective review of patients who presented with symptoms consistent with a possible ruptured abdominal aortic aneurysm and were evaluated by CT scanning. METHODS From 1983 to 1988, 95 patients were evaluated for a possible ruptured abdominal aortic aneurysm. Patients who were hemodynamically unstable went directly to the operating room. Those patients who were either stable or easily stabilized with crystaUoid fluid underwent a CT scan. Patients were considere~
Volume 12 Number 1 Julv 1990
CT and symptomatic abdominal aortic aneurysms 29
Fig. 1. Patient with paraaortic fluid (black arrow) and streaking in the rnesentery (white arrow). CT findings were considered positive for rupture. At surgery he was found to have an intraabdominal abscess with carcinomatosis.
to be at increased operative risk if they had concurrent cardiac disease including congestive heart failure, arrhythmias, myocardial infarction; renal disease with renal insufficiency (creatinine >2) or failure requiring dialysis; or, a history of chronic obstructive pulmonary disease or forced expiratory volume in 1 second (FEV1) < 1. Intravenous bolus contrast was used at the discretion of the attending vascular surgeon. If the diagnosis of rupture was established by CT scanning, the patient was taken directly to the operating room from the radiology department. If a~e CT findings were negative for rupture, the patient was further evaluated for other causes of his or her pain. Subsequently, the decision for aneurysm repair ,was made. The patients were retrospectively divided into four categories. Category I consisted of those patients who were hemodynamically unstable and went directly to the operating room. Category II were those hemodynamically stable patients who were found on CT scanning to have evidence of rupture, Category III were those hemodynamically stable patients who had no evidence of rupture on CT scanning, and category IV were those found on CT scanning to have no aneurysm. RESULTS
Of the 95 patients reviewed, 30 were in category I (hemodynamically unstable). They went directly to the operating room. Six of these 30 patients had a ~..rdiac arrest in the emergency depamnent and were
resuscitated on the way to surgeD,. The mortality in this group was 60%. Sixty-five hemodynamically stable patients underwent CT scanning. Twenty-one of these 65 patients (32%) were at increased operative risk. The average duration of the examination was 63 minutes, and no episodes of hypotension occurred during the procedure. There were 18 patients (28%) in category II (stable, ruptured). One patient with terminal lung cancer, when told he had a leaking thoracoabdominal aneurysm, declined surgery and subsequently died. The other 17 patients were taken immediately to the operating room. At operation one patient had no evidence of rupture of his aneurysm. This patient had recently undergone, at another hospital, a Billroth II gastrectomy and colon resection for gastric cancer invading the transverse colon. His abdominal aortic aneurysm had been found during surgery. He developed abdominal and back pain after operation and was transferred to our institution for evaluation of a possible ruptured aneurysm. The CT scan showed paraaortic fluid and streaking in the mesentery. These findings were considered consistent with rupture (Fig. 1). At surgery he was found to have an intraabdominal abscess and carcinomatosis with a desmoplastic reaction in the mesentery. He therefore presented a false-positive CT scan. Thus of the 18 patients diagnosed by CT scanning to have a ruptured aneurysm, 17 underwent immediate surgery. A rupture was confirmed at surgery in i6 of these patients, resulting in a sensitivity of 94%.
Journal of VASCULAR SURLY
30 Kvilekval et al.
Fig. 2. A, This patient had a ruptured aneurysm involving the left renal artery (note the paraaortic fluid [black arrow] and the superior mesenteric artery [white arrow] ). B, He was found to have a left inferior vena cava (arrow).
The overall mortality for category II was 29%. However, four of the 16 patients found to have ruptured aneurysms died. Three of the four had significant concurrent medical problems. One died during operation of a cardiac arrhythmia (ventricular tachycardia); two died of postoperative pulmonary and renal complications. The patient with no additional risk factors died during surgery of hemorrhage. Five of these 16 (31%) had postoperative complications (cardiac, pulmonary, renal), and two of these five patients had lower extremity ischemia, one required a revascularization procedure and the other an amputation.
The CT scan gave additional information about four patients in category II. One patient had a leftsided inferior vena cava and an aneurysm involving the left renal artery (Fig. 2). His aneurysm was repaired by means of a retroperitoneal approach, allowing the surgeon to avoid the inferior vena cava, which crossed over the aorta, and to reimplant the left renal artery. Another patient had an inflammatory aneurysm. A third patient was found to have both a thoracic aneurysm and an infrarenal abdominal aortic aneurysm. Because the CT scan showed that only the infrarenal aneurysm was leaking, only this portion was repaired. In the patient with known lung canc,~,
olume 12 umber 1 2' 1_990
CT and symptomatic abdominal aortic aneurysms 31
Fig. 3. Patient with carcinoid invading the small bowel mesentery (arram).
the knowledge that he had a leaking thoracoabdominal aneurysm, as noted above, led him to refuse surgery. Twenty-six (59%) of 44 patients in category III (stable, nonruptured) had concurrent medical problems putting them at increased operative risk. In 30 patients no cause(s) other than their aneurysm could be found for their pain. Twenty-one of these 30 patients had their aneurysm repaired on a semielective basis, with no mortalities. One patient had a postoperative pneumonia and another developed thromb~,sis of his popliteal aneurysm, requiring a femoral peroneal bypass. Total morbidity in this group was 8%. Thirteen of the 44 patients (30%) in category III were found to have another cause for their symptoms (Table I), and two of these 13 required emergency general surgical operations, one for strangulated bowel and the other for diverticulitis. The patient with diverticulitis was on chronic anticoagulation therapy for atrial fibrillation, and the knowledge that his aneurysm was not leaking allowed preoperative normalization of his coagulation parameters. Three patients in this group subsequently had elective repair of their aneurysm without incident. One patient had no evidence of rupture on CT scanning; his pain was attributed to prostatitis. The next day, while in the hospital, he developed increasing abdominal pain and hypotension. He was taken immediately to the operating room, and his ruptured aneurysm was successfully repaired. His case is con~dered to be a false-negative finding on CT scanning,
Table I. Additional diagnoses of patients with symptoms found on CT scanning not to have a ruptured aneurysm (category II1) Diagnosis Anastomotic ulcer S/P Billroth II Bile gastritis S/P Billroth II Cholecystitis Colon cancer Compression fracture Dissection with lower extremity ischm~a Diverticulitis Ischemic cofitis Renal cyst Strangulated small bowel Villous adenoma
No.ofpatients 1 1 3 1 1 1 1 1 1 1 1
even though further review of his CT scan showed no evidence of a leak. Therefore, of the 24 patients undergoing aneurysmectomy after having a CT scan outcome negative for rupture, 23 were found to have no evidence of rupture at surgery, and one ruptured in hospital, resulting in a specificity of 95%. Nine patients in category III were not considered surgical candidates because of their poor medical condition or the nature of their aneurysm or both. Three of these nine patients died of their medical problems within 2 months of being evaluated by CT scanning; one of cardiac arrhythmias and two of respiratory failure. No patients with CT findings negative for rupture died of their anemysms.
Journalof VASCULAR SUP,G ~
32 Kvilekval et al.
Fig. 4. Patient with retroperitoneal fibrosis (arrow). Table II. Additional information obtained from the CT scan Category II (n)
Category III (n)
Category I V (n)
(stable,ruptured)
(stable, nonruptured)
(no aneurysm)
Left-sided inferior vena cava (1) Inflammatoryaneurysm (1) Thoracic & abdominal aneurysmwith only the infrarenalportion leaking (1) Thoracoabdominalaneurysmin patient with terminal lung cancer (1)
Retroaortic renal vein (1) Left-sided inferior vena cava (1) Suprarenal componet to aneursym (9)
Horseshoe kidney (1) Carcinoid invading small bowel mesentery (1) Retroperitoneal fibrosis (1)
Table III. Characteristics o f patients undergoing aneurysm surgery after CT scanning Category II Category III
.....
No. of patients
Increased risk ~
Morbidity
Mortality
16 24
63% 55%
31% 8%
25% 0%
Concurrent cardiac disease including congestiveheart failure, arrhythmias,myocardialinfraction; renal disease with renal insufficiency (creatinlne >2) or failure requiring dialysis; or, a history of chronic obstructive pulmonary disease or forced expiratory volume in 1 second (FEV1)
