Volumt: 6 Number 6 December 1987
Pulmonary angiography was performed by advancing a percutaneous venous catheter into the main pulmonary artery and injecting contrast material under pressure. Studies were performed selectively if the initiai injection did not clearly indicate embolism. Peripheral phlebography was performed after contrast material was injected into the dorsal vein of the foot. All noninvasive studies were interpreted by one of the authors. Studies were performed within 4 to 48 hours of the onset of symptoms or presentation to the hospital. Of the 16 patients with positive noninvasive studies, 10 had angiograms that indicated pulmonary embolism and six had normal angiograms. Of 34 patients with normal noninvasive studies, 16 had abnormal pulmonary angiograms whereas 18 had normal angiograms. The positive predictive value for the noninvasive studies in comparison to pulmonary angiography was 63% and the negative predictive value was 53%. By contrast, in 125 patients with suspected deep venous thrombosis, positive and negative predictive values for noninvasive studies in comparison to phlebography were 94% and 92%, respectively. In this retrospective study, noninvasive venous studies correlated well with phlebography for the diagnosis of deep venous thrombosis but were not shown to be clinically useful in the diagnosis of pulmonary embolism. D. Preston Flanigan, MD. University
of Illinois
College
ofMediciw
~ipyr~~~le~~a~~ scanning iu patients undergoing vascular surgery: optimizing preoperative evaluation of cardiac risk Eagle KA, Singer DE, Brewster DC, et al. JAMA 1987;257:2185-9. In this study, 111 patients underwent dipyridamole*hallium scanning before aortic surgery. Patients were also evaluated for the presence of one or more of the following cardiac risk factors: Q wave on preoperative electrocardiography (KG), history of congestive heart failure, diabetes, angina, or myocardial infarction. In the initial phase of the study, the scan results of 61 consecutive patients were compared with the incidence of postoperative cardiac events. There were eight patients with postoperative cardiac complications including two fatal myocardial infarctions among 18 patients with abnormal scans. Stepwise logistic regression analysis showed that a reversible defect on the thallium scan, indicating an area of potentially ischemit myocardium, was the best predictor of postoperative cardiac events (p < 0.001). An ECG Q wave or history of congestive heart failure was also of predictive value. Although diabetes or a history of angina or myocardial infarction did not bear a statistically significant relationship to postoperative cardiac complications, these variables were combined with the risk factors of ECG Q wave and congestive heart failure to subdivide patients into high- and lowrisk subgroups. Of 29 patients having none of the factors mentioned above, only one had a postoperative cardiac
event compared with seven complications among 32 patients with at least one risk factor (p = 0.04). In the second phase of the study, scan results of 50 patients were used to validate the predictive usef&xss of clinical risk factors. There were no postoperative cardiac complications among 23 patients in the low-risk group compared with 10 in the high-risk subset. Corn~in~~g the results from both groups, the likelihood of ~~rio~er~~~~~e myocardial complications in patients with ~all~~rn redistribution in the high-risk group was 45% compared with only 7% for those without redistribution. The authors conclude that patients lacking a clinical cardiac risk factor are unlikely to have a myocardial complication atier surgery and need no further evaluation. Of 52 patients who did not have symptoms ofcoronary artery disease and who had a normal EGG, 14% had an abnormal scan but only 2% had a postoperative cardiac complication. Both figures are surprisingly low. Pin other studies, routine preoperative coronary angiography has shown that 22% to 30% of asymptomatic patients have significant coronary occlusive disease. The incidence ofcardisc complications among such patients has been reported to be several times that observed in this study. Another concern is that multivariant analysis failed to show statistical significance for three of the five risk factors used to define the low-risk groups, possibly because of the small sample sizes involved. Nonetheless, this study reinforces the value of preoperative dipyridamole-thallium scanning as a screening test for coronary artery disease. The selective use of thallium scanning to stratify the cardiac risk ofsymptomatic patients is a potentially cost-effective approach, since only those patients with clinical risk factors and an abnormal scan need further evaluation. Bwce S. Cutler, M.D. University ofMmsaclmetts
Medical
Center
Percutaneous ~a~s~~~~ angioplas artery. A therapeutic alternative to stmction of proximal vertebral artery stenoses Briickmann H, Ringelstein EB, Buchner II, Zeumer J Neurol 1986;233:336-9.
H.
The authors report their use of percutaneous transluminal angioplasty (PTA) for treatment of proximal vertebral artery occlusive disease in 13 patients during 3 years. All patients had symptoms of vertebrobasilar insufficiency. PTA was performed only if there were bilateral hemodynamically significant vertebral artery lesions; the dilatation was done on the side with the more serious stenosis. During the procedure, acoustically evoked brain stem potentials and continuous-wave Doppler ultrasound flow signals from the mastoid loop of the homolaterdl vertebral artery were monitored. Immediately after PTA, angiography was performed to document results. During an obsewatlon period of 2 to 25 months (average 15 months) vertebral artery occlusion at the angioplasty site occurred in one patient and recurrent stenosis occurred in another, without addi
630
~ournd of VASCULAR SURGERY
Abstracts
tional neurologic sequelae in either instance. Of the 13 patients, eight showed improvement in both subjective and objective clinical symptoms. The procedure was clinically unsuccessful in the remaining five patients. In two of these five cases, CT scanning displayed signs of cerebral small vessel disease. Despite an improvement in neurologic status, one patient died of myocardial infarction 14 days after PTA. The authors interpret these results as encouraging for the use of PTA for the treatment of selected patients with proximal vertebral artery stenosis. Charles O’Mara, M.D. Universi~ of Mksiss+pi
School of Medicine
Metastatic cancer-a relative contraindication to vena cava filter replacement Walsh DB, Downing S, Nauta R, Gomes MN. Cancer 1987;59:161-3. Insertion of a Greenfield vena cava filter for the treatment of pulmonary embolism is effective treatment with low morbidity and mortality from the procedure. The authors retrospectively reviewed records of 41 patients who had Greenfield filters inserted between October 1979 and November 1984. Indications included contraindication to anticoagulation and deep vein thrombosis in 27 patients (66%)) recurrent pulmonary embolism on anricoagulation in 10 patients (24%), and prophylaxis in four patients (10%). No pulmonary emboli occurred after filter placement. Six complications were noted. The most serious of these was a filter ejected in the right atrium. Although no patients died because of filter placement, 20 of the 41 patients were dead at the time of the review. Eighteen deaths resulted from cancer, 10 within 2 months of filter placement and five before hospital discharge. All of these patients had widely metastatic cancer. The authors conclude that filter placement should be performed in patients with aggressive cancer and proven metastasis only after analysis of predicted survival. This decision is based on short survival, patient discomfort, and expense. The calculated cost of filter placement was $3235. In addition, recommendation is made for detailed discussion of the procedure and expected benefit with the patient and referring physician before insertion. Mar& Wayne
F. McNamava, State University
M.D.
Secondary aortoenteric fistulas-an analysis of 42 cases Bergqvist D, Alm A, Claes G, et al. Eur J Vast Surg 1987;1:11-8. Aortoenteric fistulas resulting from previous aortoiliac reconstructive operations remain a rare but grave complication of reconstructive vascular surgery. Their frequency
may increase because of the continued expansion of reconstructive abdominal aortic surgery. This study reviews retrospectively 42 cases of secondary aortoenteric fistula treated during a 12-year period (from 1972 to 1984) in the 13 Swedish hospitals that perform most of the major vascular surgical procedures in Sweden. The purpose was to identify factors that may result in such fistulas and to determine the results of surgical management. Twenty-five secondary aortoenteric fistulas were seen after operation for aortic aneurysms, 15 for aortoiliac occlusive disease, one after renal artery ligation for renovascular hypertension, and one after operation for an iliac pseudoaneurysm. The frequency of aortoenteric fistulas was 1.1% after aneurysm repair and 0.5% after aortoiliac reconstruction done to treat occlusive disease. In 48% of these cases, some type of technical problem complicated the original operation. The most important symptom of aortoenteric fistula was gastrointestinal hemorrhage, seen in 35 patients (84%). Most patients had several small prodromal hemorrhages. The second most common symptom was septicemia (55%). The median duration from the primary operation to the start of symptoms was 32 months with a wide range from 14 days to 120 months. The median delay from the start of symptoms to diagnosis was 30 days. Not surprisingly, the duodenum was the most common site for the fistula. The most reliable method of diagnosis remains exploratory laparotomy. Gastroduodenoscopy and angiography lead to the diagnosis in 29% and 38% of patients, respectively. Various combinations of extra-anatomic bypass and complete or partial graft removal were performed. The operative mortality rate was 58%. Repeat operations were common (56%) and were most often done because of a new aortoenteric fistula. In the remaining patients recurrence of the aortoenteric fistula or aortic stump bleeding was also common (47%). The late survival rate was a discouraging 17% at 5 years. In their discussion, the authors emphasize that parients with aortoiliac reconstructions are always at risk for this complication, even at late follow-up. All patients with aortoiliac reconstructions and gastrointestinal hemorrhage should be suspected of having a fistula until otherwise shown. Exploratory laparotomy remains the diagnostic method of choice. Although controversy continues over the optimal surgical treatment, the current recommendation is construction of an extra-anatomic bypass to avoid problems with leg ischemia and then complete removal of the involved graft. Since the major late problem was recurrence of the fistula or a blow-out of the aortic stump, the unsolved problems with secondary aortoenteric fistulas remain secure closure of the aortic stump and better protection of the sutured aorta with some types of vascularized tissue. John W. Hallett, Mayo Clinic
JY., M.D.