Abdominal Aortic Aneurysms: Diagnostic Review and New Technique

Transcription

1 CURRENT REVIEW Abdominal Aortic Aneurysms: Diagnostic Review and New Technique Mario N. Gomes, M.D., Dieter Schellinger, M.D., and Charles A. Hufnagel, M.D. ABSTRACT A new technique of computed tomography (CT) applied to the diagnosis of abdominal aortic aneurysms is described, and the experience in 58 patients is reviewed. In all instances the abdominal aorta was clearly demonstrated, even when wall calcification was absent. A series of 37 patients with abdominal aortic aneurysm underwent evaluation by physical examination, abdominal roentgenograms, and ultrasonic and CT scanning. Measurements of the transverse diameter at the point of maximal dilatation were compared with the measurements made at operation. CT not only confirmed the diagnosis in all patients but the measurements obtained by this technique were the most accurate, correlating extremely well with the true dimensions of the aneurysm. The addition of contrast enhancement to CT scanning allowed clear delineation of the aortic lumen and intraaneurysmal thrombus, not possible with any other method, including ultrasonography. This technique appears useful as a screening procedure and in the differential diagnosis of a tortuous abdominal aorta. Patients with a small aortic dilatation can be followed accurately by scanning. The incidence of abdominal aortic aneurysm has been estimated to be between ten and twenty persons per thousand [31. The increase in the average age associated with degenerative processes is probably the major factor in the apparent rise in incidence of these aneurysms. The report on the natural history of abdominal aortic aneurysms by Estes [61 brought attention to the high incidence of rupture and the low survival rate. The efficacy of operation in the prevention of rupture has been demonstrated but the indications for operative inter- From the Departments of Surgery and Radiology, Georgetown University Medical Center, Washington, DC. Address reprint requests to Dr. Gomes, Division of Cardiovascular Surgery, Georgetown University Hospital, 3800 Reservoir Rd, NW, Washington, DC vention range from resection in all patients with abdominal aneurysms [7, 201 to resection in only symptomatic patients [29]. Other recommendations fall between these two extremes 124, 26, 281. Some of the controversy regarding the surgical indications may derive from the diagnostic methodology; most of the data were published in the 1960s when diagnosis was obtained only by physical examination or roentgenograms of the abdomen. Because other means of assessing the size of the aneurysm in untreated patients were not available, the validity of some of the conclusions regarding the natural history of small abdominal aortic aneurysms is subject to question. A very low mortality rate for elective procedures in unselected patients, regardless of age, can be expected now in institutions with expertise in vascular operations. Our reported overall operative mortality of 4% in 1965 to 1966 [91 has been lowered slightly since then and compares favorably with figures from other centers [2, 231. For selected patients with small aneurysms, the risk seems to be even less, ranging from 2.5 to 3% [2, 7, 271. The operative mortality in ruptured aneurysms is still high. A study reported in 1977 shows a mortality of 55% in a group of 66 patients with rupture [201. Large aneurysms rupture two and one-half to three times more frequently than small ones [7, 271. Even in small aneurysms the risk of rupture is still substantial; it was 16 and 19.5%, respectively, in the series of Foster [71, Szilagyi [271, and their colleagues. Accurate evidence of progression in aneurysm size may be difficult to obtain by the traditional means of diagnosis. The positive diagnosis of aneurysm can also be difficult in obese patients, in patients in whom there is abdominal guarding, and when calcification of the aneurysmal wall is absent or minimal. A technique combining simplicity, safety, reliability, by Mario N. Gomes

2 480 The Annals of Thoracic Surgery Vol 27 No 5 May 1979 and accuracy of measurement of size should be available to make a definitive diagnosis in all prospective patients, particularly in the follow-up of patients with small aneurysms for which conservative therapy was elected. Diagnosis Most abdominal aneurysms are asymptomatic until the time of rupture or enlargement. Our experience [9] has been that, with careful questioning, symptoms can be related to nonruptured aneurysms in only 34% of patients. In 11% awareness of an abdominal pulsation was a complaint. Identification and measurement of the pulsatile mass can often be done by physical examination alone. The success of diagnosis by palpation is determined by aneurysm size and thickness of the abdominal wall. Detection has ranged from 22 to 60% [121. Higher figures are found in surgical series. In ours [9], 90% of the aneurysms were readily palpable. When the clinical examination is unsatisfactory or when the aneurysm is small, the diagnosis of aneurysm can easily be missed. It is important to remember that hyperlordosis of the lumbar spine, a tortuous aorta, or a retroperitoneal tumor may resemble an aneurysm on physical examination. Roentgenographic examination of the abdomen with anteroposterior views as well as overexposed lateral or oblique views or both has been most helpful. Calcification of the aneurysm wall is usually necessary for radiological diagnosis of abdominal aneurysm. In general, the lateral views are more definitive since a tortuous aorta of normal caliber may mimic an aneurysm in the anteroposterior view. An estimation of aneurysm size can be made when calcification is present at the same level in both walls of the aneurysm. This occurs in only 55% of patients [25]. In approximately 20 to 45% of patients [22] there are no roentgenographic manifestations. Contrast aortography visualizes only the lumen, not the laminated thrombus; this may make the assessment of true aneurysm size impossible. Even in large aneurysms, a normalsized lumen may produce a false diagnosis. Indirect demonstration can be found when aortic calcifications are seen outside the opacified area or when there is displacement of the neighboring vessels. A certain degree of risk is attached to this procedure, depending on the experience of the angiographer. In the series reported by Robicsek and co-workers [21], the aneurysm was not visualized in 48 of 218 patients; an incidence of 8% of false-negative aortograms has been reported also [lo]. The value of aortography derives from the additional information it may provide regarding associated involvement of the renal, mesenteric, and peripheral arteries. The place of radionuclide aortography with technetium 99m in the diagnosis of aneurysm is still not well defined. Like contrast angiography, it reveals the size of the aortic lumen but generally not the overall aortic diameter. Ultrasonography applied to the abdominal aorta has been a useful tool in patients in whom calcification is absent. At present, a twodimensional (B scan) ultrasonic picture is preferred. Usually, the aorta and adjacent organized inner thrombus can be recognized. Both sagittal and transverse ultrasonic studies are made in order to increase accuracy and eliminate possible confusion with other paraaortic structures. With ultrasonic techniques, diagnostic accuracy rates of 98.8 and 100% have been reported [5, 191. With conventional B-mode scanning, exact measurements of wall thickness are difficult to obtain [5]. The use of a gray-scale unit that improves definition and allows better differentiation of the echo-reflecting structures would probably provide higher accuracy [5, 81. The presence of abdominal gas decreases the reliability of results obtained with ultrasonic methods, and in acute situations, use of ultrasonography is limited. Computed tomography (CT) of the head has been recognized as one of the most revolutionary diagnostic tools of recent decades. It was developed by Hounsfield [131 and then was introduced by the EM1 Company in With the advent of newer technology, the usefulness of CT scanning has been extended to the entire body [16], thus exposing almost all organ systems to this new and exciting diagnostic method. Details and technical data have been

3 481 Current Review: Gomes, Schellinger, and Hufnagel: Abdominal Aortic Aneurysms published elsewhere [13, 141. Briefly, two basic principles constitute the fundamental difference between conventional radiography and CT. (1) In computer technology the roentgenographic film is replaced by multiple detectors, which feed absorption data into an attached computer. The computer composes an image of significantly higher resolution that can be displayed on an oscilloscope screen and photographed as a permanent record. Diseased areas can be shown to better advantage. (2) Unlike conventional radiography in which body structures are looked at in an anteroposterior, lateral, or oblique geometry, CT scans the body in multiple cross-sectional slices 3 to 15 mm thick. With 90 to 95% of abdominal aortic aneurysms arising below the renal arteries and frequently extending to the aortic bifurcation, it seemed natural to extend the CT procedure to this field, given the anatomical constancy of appearance in a relatively circumscribed area (Fig 1). At our institution where CT body scanning was originally developed 1161, the usefulness of this new diagnostic method in the evaluation of aortic aneurysms has been recognized [l, 111. Our three-year experience is the basis of this report, which assesses the value of CT in establishing the diagnosis of abdominal aneurysm and the accuracy of its measurements compared with other established modalities, with the diameter as measured at operation as the standard. Materials and Methods Fifty-eight patients are included in this study of CT scanning of the abdominal aorta. There were 43 men and 15 women ranging in age from 47 to 85 years. The majority were in the sixth and seventh decades. Throughout this experience, the CT images were obtained on ACTA scanners (the 100 and 200 FS). The initial study was directed to establish the validity and reliability of CT scanning in a group of 8 patients in whom the diagnosis of aortic aneurysm already had been definitively made by other means. The clinical or radiological findings or both were compared with the image obtained by CT scanning and later were confirmed at the operating table. Once we had established an image of consistently good quality and diagnostic accuracy, we investigated the determination of aortic size in a group of 37 patients. These patients were admitted to Georgetown University Hospital for elective resection of arteriosclerotic abdominal aortic aneurysms and were evaluated in a prospective, blind study, which included physical examination, abdominal roentgenograms, ultrasonic determination, CT scanning, and final comparison with surgical measurements. The transverse diameter of the aneurysm at its widest level was estimated independently by two experienced vascular surgeons. A radiopaque marker was taped to the skin at that level, and anteroposterior and lateral roentgenograms of the abdomen were made using a higher penetration. Measurements of the visualized aneurysm, when sufficient wall calcification was present, were obtained by a different examiner and recorded separately. The same protocol was carried out for the ultrasonic and CT scanning. In 25 of these patients, ultrasonography was done using a standard B- mode unit in longitudinal and transverse plans. At operation just before cross-clamping, measurements of the transverse diameter of the aorta were obtained with a caliper and were considered to be the true size of the aneurysm. The third group of patients studied included 11 in whom an aortic aneurysm was suspected. An average of five CT scans were made at intervals of 2 to 3 cm, with the xiphoid or the umbilicus as baseline. We now consider the umbilical area as baseline and obtain three scans above and one or two below at 2-cm intervals; the whole of the infrarenal aorta and the common iliac arteries are generally encompassed by these settings. Direct identification of intraaneurysmal thrombus still cannot be made with the newer generation of scanners probably because of radiographic absorption of similar levels of thrombus and intraaortic blood. To overcome this problem, contrast enhancement also was used in 14 of the 58 patients. In this subgroup, CT scans were made before and after a drip infusion of 25% sodium diatrizoate. Allergic reactions to the contrast medium were tested, and then about 150 ml was infused rapidly over

4 482 The Annals of Thoracic Surgery Vol 27 No 5 May 1979

5 483 Current Review: Gomes, Schellinger, and Hufnagel: Abdominal Aortic Aneurysms a five-minute period; then scanning was initiated, with a slower drip of another 150 ml during the remainder of the procedure. Two other patients with the diagnosis of false aneurysm of the abdominal aorta, not included in this group, were similarly studied. Results In all 58 patients the abdominal aorta was clearly visualized on CT scans, regardless of its size and location. Characteristically, it appears as a round, fairly homogenous structure (Fig 2) Of diameter and is located anterior to the vertebral body and generally to the left. Its degree of radiodensity is moderate and similar in these scans to that of the kidneys, spleen, or liver. Calcification of the aortic wall is made strikingly apparent by the sharp increase in radiodensity (Fig 3). The degree of calcification encountered ranged greatly from none to almost circumferential involvement. Lack of calcification does not impair the clear delineation of the aorta. Occasionally, erosions of the anterior aspect of the vertebral body are visible. In the first group of 8 patients who received a clinical diagnosis of aortic aneurysm, scanning also was used for diagnosis, and the results were confirmed subsequently at operation. The maximal transverse diameters of the aortas were evaluated and compared by the different methods in the group of 37 patients with known aneurysms. Surgical measurements were not obtained for 4 of them, and they are excluded from the results. One of these 4 patients refused operation, and the other 3 were such poor surgical candidates that operation was not performed. Analysis of the data (Figs 4, 5) in the remaining 33 patients shows that not only did CT scanning demonstrate the aneurysm in all instances but also that the estimated measurements of size correlated extremely well with the true diameter obtained at the time of operation. In all patients the measurements were either identical or within a 5% variation. The CT values were far more accurate Fig I. Multiple computed tomography scans of the abdominal aorta showing the size variations of the aneurysm encountered at different levels. The accompanying aortogram indicates the corresponding position of the cuts. Fig 2. Typical computed tomography scan of a noncalcified aneurysm (A) of the abdominal aorta taken at the level of the umbilicus. The vena cava (v) also is well visualized posterior to and to the right of the dilatated aorta. Fig 3. Smaller abdominal aneurysm with partial calcification of the anterior and posterior walls, as seen in a standard computed tomography scan. JG AL DA BD UI CP EC FD KN NG HC JK caw. Fig 4. Comparative measurements of the maximal transverse diameter of abdominal aortic aneurysms in 12 patients by computed tomography (CT) scanning and at operation. The interrupted line indicates a 5% variation from the exact diameter and the continuous line, a 10% variation.

6 484 The Annals of Thoracic Surgery Vol 27 No 5 May 1979 MS RW JN WD LV WT MR JV ES NS JA JE JW JD X AP RP LS TA A( Fig 5. Comparison of the maximal transverse diameter of the abdominal aortic aneurysm in a group of21 patients. Measurements were obtained by ultrasonography and computed tomography scanning, and the true diameter was measured at operation. than those obtained by ultrasonic examination in 21 of these patients. As we had anticipated, there was a discrepancy between the clinical estimates and the true diameter despite the large experience of the observers. The discrepancy was greater in heavier patients. In 1 patient, the estimated size nearly doubled the actual size of the aorta; at operation, a group of lymph nodes anterior to a small abdominal aneurysm was found. In 2 extremely heavy and tense patients, the large aneurysms (7.5 cm and 8.2 cm) could not be felt. In only 4 patients were identical figures obtained; in the great majority (24 patients) the aneurysm size was overestimated in the physical examination. Review of the conventional abdominal radiography demonstrated varying degrees of aortic calcification ranging from none (4 patients) to very heavy (5 patients). When calcification was heavy, not only the aneurysm wall but also the adjacent aorta and iliac arteries were involved. Sufficient calcification was present in 23 of the 33 studied patients (70%) to allow the diagnosis of aneurysm. In only 12 of those patients could measurements of the transverse diameter be reliably obtained; in the other 11 patients, the typical curvilinear calcification was present in only one wall. The lateral roentgenograms of the lumbar spine were consistently more helpful for the diagnosis because they showed the characteristic wall calcification, ordinarily in the anterior wall. In the 12 patients the transverse diameter was overestimated in 7 and identical to the true dimensions in only 1. Radiographic diagnosis of abdominal aneurysm was not made in 4 patients and was suggestive but not definitive in 6 others. B-mode ultrasonograms of the abdominal aorta were made in 25 of the 37 patients in this group. Because 4 of these patients had no surgical confirmation of the diagnosis, only 21 are considered. This technique successfully identified the aneurysm in all 21, but it did not provide accurate measurements of its size. We found almost consistent underestimation of the true diameter, by as much as 20% in 5 of the 21 patients. In only 4 patients were the figures within a 5% variation (Fig 6). No specific attempts were made to visualize the upper branches of the abdominal aorta by CT scanning. Occasionally, superior mesenteric, splenic, and renal arteries were demonstrated. Now we routinely obtain lower cuts (-6 cm) that include the pelvis and usually the iliac arteries (Fig 7). Even when not calcified they are easily apparent. Associated aneurysms of the iliac arteries were seen by CT scanning in 5 of our patients, and all were confirmed subsequently at operation. We can anticipate the value of this method in establishing suprarenal or thoracoabdominal involvement of the aorta. Contrast enhancement was used in the final 14 patients seen in this group and adequate definition of the aortic lumen and mural thrombus was obtained in all of them. The circulating contrast medium fills the vessel lumen and is seen in the scan as an area of higher radiodensity than the peripheral area of the organized clot (Fig 8). Different degrees of wall thickness, circumferentially or unilaterally located, are evidenced easily in this fashion, varying from patient to patient or with the level of the CT scan within the aneurysmal sac (Fig 9). Surgical confirmation of the findings was obtained by sectioning transversally the resected specimens. Intraluminal thrombus was also

7 485 Current Review: Gomes, Schellinger, and Hufnagel: Abdominal Aortic Aneurysms JN CP ED LM WT FD JC UP NG JK TA WH Filp 6. Comparison of maximal aneurysm size estimated by physical examination with size obtained in roentgenograms of the abdomen in relation to the transverse diameter obtained directly at the time of resection. Fig 7. Normal-sized common iliac arteries (arrows) with some degree of wall calcification as seen in a computed tomography scan made 3 cm below the umbilical area. Fig 8. Contrast enhancement added to a computed scan of an aortic aneu ysm. A large thrombus (T), almost circumferential but markedly heavier in the anterior aneurysmal wall, is clearly demonstrated. Findings were confirmed at resection. (L = lumen.) Fig 9. Computed tomography scans taken at different levels with contrast enhancement show well the unequal distribution and thickness of the intraaneurysmal thrombus and corresponding variations in the aortic lumen.

8 486 The Annals of Thoracic Surgery Vol 27 No 5 May 1979 Fig 10. An abdominal aorta with a transverse diameter of 3 cm, as measured electronically by computed tomography scan. Aortic wall calcification seen in a plain abdominal roentgenogram and well demonstrated here raised the possibility of the presence of an aneurysm. identified in 5 of these patients by ultrasonography; in one instance of a large saccular aneurysm (7.6 cm), the ultrasonograms were considered negative for a thrombus that was seen in an enhanced CT scan and in the surgical specimen. Contrast was not used in the CT scanning of 2 other patients because of known allergy to iodine. In both, intraluminal clot was identified by ultrasonography. In the last group of patients the presumptive diagnosis of abdominal aneurysm was raised because of clinical symptoms and suspicious physical and radiological findings. Of the 11 patients studied, CT scanning showed a normal aorta in 7 (Fig 10) and in 2, a tortuous, elongated aorta deviated to the right. In 3 of these 9 patients, ultrasonograms also were obtained and were found to be normal. The 2 remaining patients were seen initially with unexplained back pain of short duration and were both obese, hypertensive, and in the sixth decade of life. Palpation of the abdomen was unrevealing, and abdominal roentgenograms showed no aortic calcification. In both an aortic aneurysm was seen in the CT scan and confirmed at operation. Comment In an evaluation of the presently available methodology for the diagnosis of abdominal aortic aneurysms, it becomes clear that in certain cases the diagnosis may not be made when the aneurysm is present, or it may be erroneous due to extraneous circumstances, such as a tortuous aorta, or retroperitoneal masses. Accurate evaluation of aneurysm size is a very important factor, especially for those patients for whom the decision concerning elective surgical intervention depends on the diameter of the aneurysm. It can be even more valuable in the follow-up of patients with small aneurysms for which conservative therapy is elected. With the present generation of CT scanners, the abdominal aorta and its size can be demonstrated in all instances. This fact alone establishes its merits for the detection, differential diagnosis, and follow-up of aortic aneurysms. Scanners of the rotary type are now available. They reduce the exposure time to 20 and even 4.8 seconds [17] and the radiation dose to the skin from 2 to as low as 0.1 rads [30]. Although in a great many patients physical examination alone may suffice for a diagnosis of abdominal aneurysm, evaluation of its size by this method is less reliable and makes the method inexact in the follow-up of aneurysm size. In about 75% of our patients thus evaluated, we overestimated the transverse diameter. Wall calcification is a prerequisite for the diagnosis of abdominal aneurysm when using abdominal roentgenograms. No calcification was seen in 30% of our studied patients regardless of aneurysm size. Assessment of size by measuring the transverse diameter was disappointing. Other researchers [5] found this technique relatively accurate, but they measured the anteroposterior diameter of the aneurysm from the calcified anterior wall to the calcified posterior wall or, if not visible, to the presumed location of the anterior spinal ligament. The possible overlay of the spine and aneurysm wall in the frontal projection is probably the main reason for our lower diagnostic and accuracy yield. It was chosen because of the difficulty in assessing the anteroposterior diameter of the aorta by physical examination. Aortography was not used in this study as a method of routine evaluation of abdominal aneurysms. Of particular interest was the comparison of CT scanning with ultrasonography. In all pa-

9 487 Current Review: Gomes, Schellinger, and Hufnagel: Abdominal Aortic Aneurysms tients, the diagnosis of aneurysm was made by both techniques. However, the applicability of ultrasonography in the follow-up of small aneurysms, with the standard mode, seems less accurate based on our findings. Its reliability is related to the experience of the observer since standardization of the technique is more difficult. Like other authors [5, 8, 181, we found underestimation of the reported diameters compared with the ones obtained at operation. The problem arises from the acoustical difficulty in separating aortic wall echoes from those of adjacent tissues [18]. Brewster and coworkers [5] believe that the measurements reflect the diameter of the aneurysmal lumen and do not include the wall thickness. The discrepancy seems to be directly proportional to the thickness of the aneurysmal clot. We found the highest accuracy in aneurysms with minimal or absent thrombus and thin aortic wall. Intraluminal clot was identified frequently but not in all of the patients in whom it was present at operation; even when reported it was not necessarily in the same anatomical location within the aneurysmal sac. The accuracy of CT scanning in the determination of exact aneurysm size was demonstrated in our study. The capability to obtain a sharp image of the aorta, with excellent contrast from the surrounding tissues, combined with the incorporation in our present scanner of a programmed mechanism that automatically computes the desired diameter will almost certainly provide a high degree of accuracy in measurements. In 1977, it was demonstrated that the aorta is one of the few structures whose CT image can be directly measured with accurate results [151. Longitudinal ultrasonography for the evaluation of aneurysm length is an asset compared with CT scanning although this capability probably will be incorporated in the next generation of scanners. An advantage of CT scanning over ultrasonography is that sources of artifact giving rise to acoustical interference can be overcome. The introduction of contrast enhancement to CT scanning adds a new dimension: the recognition of lumen and mural thrombus. This addition may be even more valuable when Fig 11. Abdominal computed tomography scan showing not only on aortic aneurysm (A) but also large, bilateral, polycystic kidneys (K). applied to the diagnosis of dissecting aortic aneurysms when true and false lumens must be identified. We have been able to do this recently. The preoperative identification of intraluminal thrombus, accomplished in the 14 studied patients, can be helpful in the prevention of distal embolization during aneurysmectomy. Visualization of the iliac arteries by CT scanning is also helpful since it identifies before operation the degree of dilatation of these vessels in the aneurysmal process. Accessory information regarding other organs such as the liver, spleen, pancreas, or kidneys can be obtained also by CT scanning. For example, polycystic kidneys (Fig 11) were demonstrated during a recent study of a patient with an abdominal aortic aneurysm. CT scanning is more accurate than clinical examination, roentgenography, or aortography in the diagnosis of abdominal aortic aneurysm. Like ultrasonography, it is a very good screening method when the diagnosis is questionable and is good for differentiating a tortuous aorta from an aortic aneurysm or evaluating paraaortic masses that may simulate aneurysms. Accuracy in the evaluation of aortic size with CT scanning appears to be higher, which makes it a better procedure in the follow-up of small aneurysms. Being noninvasive it can be used easily and rapidly in all patients and in emergency situations in which ultrasonography may not be possible. A limited view of the aorta is obtained with a single scan. Four to five scans are necessary to

10 488 The Annals of Thoracic Surgery Vol 27 No 5 May 1979 estimate aneurysm length. It is likely that longitudinal scans will be available soon. There already is reduction in both the exposure time and the radiation dose to the skin. Better resolution may occur in the future; perhaps this would make the use of contrast enhancement unnecessary. Such modifications probably will reduce the cost of the examination and further widen the applications of CT scanning. References 1. Axelbaum SP, Schellinger D, Gomes MN, et al: Computerized tomographic evaluation of aortic aneurysms. Am J Roentgenol 127:75, Baker AG, Roberts B, Berkowitz HD, et al: Risk of excision of abdominal aortic aneurysms. Surgery 68:1129, Bergan JJ, Yao JST: Modern management of abdominal aortic aneurysms. Surg Clin North Am 54:175, Bernstein EF, Fisher JC, Varco RL: Is excision the optimum treatment for all abdominal aortic aneurysms? Surgery 6133, Brewster DC, Darling RC, Raines JK, et al: Assessment of abdominal aortic aneurysm size. Circulation 56:Suppl 2:164, Estes JE: Abdominal aortic aneurysm: a study of 102 cases. Circulation 2:258, Foster JE, Bolasny BL, Gobbel WG Jr, et al: Comparative study of elective resection and expectant treatment of abdominal aortic aneurysms. Surg Gynecol Obstet 129:1, Freimanis AK: Echographic diagnosis of lesions of the abdominal aorta and lymph nodes. Radio1 Clin North Am 13:557, Friedman SA, Hufnagel CA, Conrad PW, et al: Abdominal aortic aneurysm: clinical status and results of surgery in 100 consecutive cases. JAMA 200:1147, Gardner RJ, Lancaster JR, Tarnay TJ, et al: Fiveyear history of surgically treated abdominal aortic aneurysms. Surg Gynecol Obstet 130:981, Gomes MN: ACTA scanning in the diagnosis of abdominal aortic aneurysms. Comput Tomog 1:51, Gore I, Hirst AE: Arteriosclerotic aneurysms of the abdominal aorta: a review. Prog Cardiovasc Dis 17:113, Hounsfield GN: Computerized transverse axial scanning (tomography): I. Description of system. Br J Radio1 46:1016, Hounsfield GN: Picture quality of computed tomography. Am J Roentgenol 127:3, Kricheff 11, Saulnier J, Zalesnek M, et al: Area and volume measurements of CT images: errors and their correction (abstract 32), CT Symposium, Heidelberg, West Germany, Sept 30, Ledley RS, DiChiro G, Luessenhop AJ, et al: Computerized transaxial X-ray tomography of the human body. Science 186:207, Margulis AR, Boyd DP, Korobkin MT: Advantages and disadvantages of rotary body CT scanners (abstract 2), CT Symposium, Heidelberg, West Germany, Sept 29, McGregor JC, Pollock JG, Anton HC: The value of ultrasonography in the diagnosis of abdominal aortic aneurysm. Scott Med J 20:133, Mulder DS, Winsberg F, Cole, CM, et al: U1- trasonic B scanning of abdominal aneurysms. Ann Thorac Surg 16:361, Robicsek F: Aneurysms of the abdominal aorta: whether or not, when, and whom? Ann Thorac Surg 24:241, Robicsek F, Daugherty HK, Mullen DC, et al: The value of angiography in the diagnosis of unruptured aneurysm of the abdominal aorta. Ann Thorac Surg 11:538, Ryan EA, Spittell JA Jr, Kincaid OW: Roentgenographic manifestations of abdominal aortic aneurysms. Postgrad Med 36:A77, Dec Sanger PW, Robicsek F, Daugherty HK, et al: Operative mortality rate of aortic aneurysms. Surgery 64:359, Schatz IJ, Fairbaim JF 11, Juergens JL: Abdominal aortic aneurysms, a reappraisal. Circulation 26:200, Smith BF, Maloney JD, Pairolero PC, et al: U1- trasound evaluation of abdominal aortic aneurysms (abstract). Circulation 53, 54:Suppl2:11-12, Stokes J, Butcher HR Jr: Abdominal aortic aneurysms: factors influencing operative mortality and criteria of operability. Arch Surg 107:297, Szilagyi DE, Smith RF, DeRusso FJ, et al: Contribution of abdominal aneurysmectomy to prolongation of life. Ann Surg 164:678, Voorhees AB Jr, McAllister FF: Long-term results following resection of arteriosclerotic abdominal aortic aneurysms. Surg Gynecol Obstet 117:355, Wolffe JB, Colcher RE: Diagnosis and conservative management of atherosclerotic aneurysms of the abdominal aorta. Vasc Dis 3:49, Zaklad H: Low dose in computerized tomography (abstract 13), CT Symposium, Heidelberg, West Germany, Sept 29, 1977