a x Reprinted from ANNALS OF SuRGERY, Vol. 144, No. 4, October 1956 Copyright © 1956 by J. B. Lippincott Company Printed in U. S. A. Aneurysm of Thoracoabdominal Aorta Involving the Celiac, Superior Mesenteric, and Renal Arteries. Report of Four Cases Treated by Resection and Homograft Replacement * MicHaEL EK. DeBaxey, M.D., Oscar Creecu, Jr., M.D., Gzorcr C. Morais, Jr., M.D.** Housten, Texas Most aneurysms of the abdominal aorta fortunately arise below the origin of the renal arteries, so that resection is not as- sociated with serious ischemic damage to vital structures.?° In the small proportion of cases, however, in which the aneurysm is located in the upper segment of the ab- dominal aorta including its major visceral branches, this problem assumes grave sig- nificance. In such cases there is consider- able danger of producing fatal ischemic damage to such vital structures as the liver, kidneys, and gastro-intestinal tract, as a consequence of temporary arrest of blood flow to these organs during the period re- quired to excise the aneurysm and replace it with an aortic homograft. This report is concerned with our experi- ences with resection of aneurysms of this latter type and replacement by homografts in four cases. The aneurysms in all of these cases extended from the lower descending thoracic aorta to the lower abdominal aorta and involved the celiac, superior mesen- teric, and one or both renal arteries. So far as we have been able to determine from a * Presented before the American Surgical As- sociation, White Sulphur Springs, West Virginia, April 11-13, 1956, Supported in part by the Association. ** From the Cora and Webb Mading Depart- ment of Surgery, Baylor University College of Medicine, and the surgical services of the Veterans Administration, Methodist, and Jefferson Davis Hospitals, Houston, Texas. Houston Heart 549 review of the literature, there have been no records of similar cases in which all these vessels were involved, although two cases with involvement of some of these vessels were recently reported. In one of these the aneurysm arose below the origin of the nor- mally placed left renal artery but involved the renal artery of an ectopic right kidney, and it was successfully resected and arterial continuity to the right kidney restored after replacement with a homograft.”! In the other case the aneurysm, involving the celiac, superior mesenteric, and left renal arteries, was successfully resected and the segment replaced with a homograft so that con- tinuity to the celiac and superior mesen- teric arteries was restored but the left kid- ney was excised.?? ; In the four cases to Ke reported the op- erative procedure consisted in excision of the aneurysm and replacement with an aortic homograft with restoration of con- tinuity to the celiac axis and superior mes- enteric arteries in all, as well as to both renal arteries in two and to one of the renal arteries in the other two cases. Because of the extensive nature of this operative pro- cedure and the problems it poses in terms of technical management as well as func- tional disturbances of vital organs, it seems desirable to record these cases and to con- sider certain observations derived from this experience, relating particularly to surgical management and to significant studies on renal and hepatic function. 550 DEBAKEY, CREECH AND MORRIS Annals of Surgery October 1956 Fic. 1. Composite roentgenogram of the chest and abdomen with diagram superimposed to show location and extent of the thoracoabdominal aneurysm in Case 1. Replacement of the excised segment of aorta and aneurysm by homograft with restoration of continuity of the aorta, celiac, superior mesenteric, and renal arteries is shown in the diagram on the right. CASE REPORTS Case 1. J. H., a Negro farmer 66-years-old, was admitted to the Houston Veterans Adminis- tration Hospital Aug. 10, 1955, because of pain in the abdomen, left flank and back, of 2 years’ dura- tion. The pain initially was felt only in the upper portion of the abdomen but gradually extended to the back and left flank. The patient had lost about 30 pounds in weight during his illness and had had to discontinue farming because of failing strength. The patient had had gonorrhea and probably a penile lesion, the nature of which could not be accurately ascertained. On admission the blood pressure was 158 mm. Hg systolic and 100 mm. Hg diastolic. The heart was not enlarged and no murmurs were detected. The abdomen was somewhat protuberant, particu- larly in the epigastrium, where vigorous pulsations were visible. Palpation disclosed a firm mass in the upper part of the abdomen extending from the xiphoid process almost to the umbilicus. It was somewhat irregular in outline, with an ex- pansile pulsation. A systolic murmur was audible over the entire mass. Roentgenographic examination of the chest and abdomen revealed a large soft-tissue shadow extending from the level of the eleventh thoracic to the third lumbar vertebra (Fig. 1). The bodies of the twelfth thoracic and first and second lum- bar vertebrae were extensively eroded (Fig. 2). An aortogram disclosed a large dumbbell-shaped aneurysm extending from the ninth thoracic verte- bra to the second lumbar vertebra. The waist of the dumbbell appeared to be at the level of the aortic hiatus of the diaphragm. Intravenous pye- Volume 144 Number 4 Fic. 2. Lateral roentgenogram of the abdomen in Case 1 showing erosion of the bodies of the twelfth thoracic, first and second lumbar vertebrae. lography outlined the collecting systems of both kidneys but failed to disclose their relationship to the aneurysm. An electrocardiogram showed normal sinus rhythm with slight evidence of left ventricular hypertrophy. Laboratory studies dis- closed slight increase in the globulin fraction of the serum proteins and moderate reduction in platelet count. On Aug. 31, 1955, operation was performed under general body hypothermia. Surface cooling was accomplished with a Thermo-O-Rite® blanket, the temperature having been reduced to 32.8° C. (91° F.). A left thoracolumbar approach was em- ployed with entrance into the pleural cavity through the bed of the eighth rib. The pleural space was obliterated by dense fibrous adhesions, and posteriorly in the paravertebral gutter and adjacent mediastinum there was evidence of old and recent hemorrhage. The lower lobe of the lung was mobilized and retracted upward. The descend- ing thoracic aorta was isolated about 6 cm. above the diaphragm and encircled with tape. The aneu- rysm was found to arise about 4 cm. above the diaphragmatic hiatus and extended well below the ANEURISM OF THORACOABDOMINAL AORTA 551 renal arteries. The posterior parietal peritoneum over the abdominal aorta and the ligament of Treitz were divided. The aorta below the aneurysm was encircled with umbilical tape. In order to ex- pose the anterior surface of the aneurysm, it was necessary to remove the spleen and divide the attachments of the tail of the pancreas. The splenic flexure of the colon was mobilized, and retraction of the stomach and pancreas to the right and the colon downward exposed the aneurysm. The left kidney lay on the left anterior lateral wall of the sac, its vein passing laterally across the aneurysm to the inferior vena cava. The left kidney was dis- sected from Gerota’s fascia and turned medially to expose the artery at the hilum. This vessel was then dissected medially and found to arise from the left lateral wall of the aneurysm about 5 cm. from its distal end. Mobilization of the left renal vein exposed the right renal artery which also arose from the aneurysm. Superior mesenteric and celiac arterial trunks were next isolated and en- circled with tapes. Occluding clamps were applied to the aorta above and below the aneurysm and to the renal, superior mesenteric, and celiac arteries. The sac was then opened widely and excised except for the portion adherent to the eroded vertebral bod- ies. The major branches were diveded at their origin from the aneurysm to preserve maximum length. A segment of abdominal. aortic homograft containing stumps of the renal, superior mesenteric, and celiac arteries was then used to bridge the de- fect and to restore continuity to all of these major vessels (Fig. 1). Following completion of these anastomoses the occluding clamps were released and normal circulation through the aorta was re- established. The respective periods of arrest of circulation are shown in Chart III. The remnant of aneurysmal sac at the site of vertebral erosion was sutured along its margin for hemostasis. The omentum was then brought posteriorly and placed around the graft and into the vertebral defect. Penrose drains were placed into the vertebral bed of the aneurysm and brought out through a stab wound in the left flank. The diaphragm was re- paired, an intercostal catheter was inserted into the left pleural space for underwater drainage, and the thoracic and abdominal wounds were closed. During operation the rectal temperature drifted to 32.2° C. (90° F.). Rewarming was begun about one hour prior to completion of the operation, and the temperature returned to normal 414 hours later. During the operation raw surfaces bled ex- cessively. At that time the prothrombin time was 15 per cent, clotting time 45 seconds, and plate- let count 100,000. Intravenous administration of 50 mg. of vitamin K-1 oxide caused prompt re- 552 DEBAKEY, CREECH AND MORRIS Annals of Surgery October 1956 Cuart I. Data Relating to Renal Function in Case 2 CuHartT I After Operation (days) Before Determination Operation 2 7 11 23 Urine Volume 1000 162 1175 1150 1140 ml./24 hrs. Urine Specific Gravity 1.023 1.024 1.014 1.013 Blood Urea Nitrogen 14 78 432. 20 22 mg. % Urea Clearance (1) 27%—normal max. clearance (2) 29% —normal standard clearance Renal Blood Flow ml./min. Renal Plasma Flow ml./min. Glomerular Filtration Rate ml./min. Plasma Sodium meq./l. Plasma Potassium meq./l. Sodium Excretion meq./l. Potassium Excretion meq./l. 29 644 16 389 1 69 141 133 138 144 5.9 4.3 4.0 2.6 8 8 12 30 duction in oozing and improvement in prothrombin and clotting times. On the first postoperative day the patient was fully conscious, moved the extremities well, and appeared to be in good condition. The urinary out- put averaged about 7 ml. an hour. On the second postoperative day urinary out- put remained about 7 ml. an hour and the blood urea nitrogen rose to 50 mg./100 ml. During the evening the patient developed severe acute pul- monary edema and was treated by venesection with removal of 500 ml. of blood and administra- tion of aminophyline, norepinephrine, and oxygen under positive pressure. The patient’s condition gradually improved, but it was necessary to con- tinue intravenous administration of norepinephrine. At 3 a.m. on the third postoperative day, severe pulmonary edema again developed and the patient became unconscious. Tourniquets were ap- plied to the extremities, tracheostomy was _per- formed, and oxygen was administered under pres- sure with a Halliburton positive pressure, demand- assist valve. The blood urea nitrogen had risen to 70 mg./100 ml. The urinary output was increased, however, to a total of 440 ml. in 24 hours. On the fourth postoperative day the patient’s general conditon remained about the same. There was, however, a significant rise in the total urine volume for the 24 hour period, the amount being 1,183 ml. Urinary steroid excretion studies showed a significant reduction in adrenal cortical activity; therefore, an infusion of 100 mg. of cortisone was given. Bronchopulmonary secretions remained a prob- lem throughout the next day although there ap- peared to be slight improvement. The blood urea nitrogen had risen to 96 mg./100 ml. Throughout the day the urinary output remained at a high level, 1,066 ml. having been excreted during the 24 hour period. However, renal blood flow, renal plasma flow, and glomerular filtration rate were reduced to negligible levels, the actual figures be- ing 22 ml. per minute, 12 ml. per minute, and 4 ml. per minute, respectively. Serum bilirubin had increased to 4.4 mg. and cephalin flocculation fe 2 +. Six days after operation the patient was still stuporous although he responded somewhat to painful stimuli. The blood urea nitrogen had risen to 124 mg./100 ml. and the hourly urinary volume gradually declined. A total of 500 mg. of cortisone was given intravenously. Although the total urinary volume for the 24 hour period was 522 ml., during the last 2 or 3 hours the urinary output was zero. The patient’s condition rapidly deteriorated on the seventh postoperative day, manifested by pro- Volume 144 Number 4 ANEURISM OF THORACOABDOMINAL AORTA 553 Cuart II. Studies of Hepatic Function Before and After Operation in the Cases Reported Herein Cuart II. Studies of Liver Function Case 1 Case 2 Case 3 Case 4 (105 min.)* (36 min.)* (47 min.)* (102 min.)* Con- Con- Con- Con- Determination tok: F071 P.075 trol. P.O. 19 tol FD. S tor. 0.7 P.O.9 Bilirubin 0.5 el 4.4 0.7 0.3 0.7 1.3 0.4 aon 2.0 Ceph. Flocc. Neg 1+ 2+ Neg. 2+ Neg 2+ Neg 2+ 2+ Thymol Turbidity 2 1 1 2 1 ae 1 1 1 Alk. Phosphatase es 4.3 4 a 30 a 11.0 Serum Protein 8.5 6.5 6.0 6.6 5.3 5.5 4.5 4.2 Prothrombin 85% 80% 80% 100% 75% 55% 100% 2. 4%. 3% * Period of simultaneous occlusion of aorta, celiac and superior mesenteric arteries. gressive fall in blood pressure despite vasopressor agents and by complete anuria. Pulmonary edema became overwhelming and the patient died at 3:50 a.m. on Sept. 7, 1955. At necropsy there were bilateral pulmonary edema and pneumonitis. With the exception of multiple small hemangiomas the liver was normal. The kidneys were swollen and pale, and on micro- scopic examination there was hyalinization of glomeruli and necrosis of tubules. The homograft was intact and patent. Case 2. C. B. M., a man 65-years-old, was ad- mitted to the Houston Veterans Administration Hospital on Oct. 13, 1955, for treatment of an aneurysm involving the lower thoracic and upper abdominal aorta. About 18 months previously, severe pain had developed in the left subcostal area, which seemed to be precipitated by sitting or standing. At first the pain was relieved by lying on the right side, but in recent months it had been unaffected by change in position and had increased in severity, involving also the left flank and upper lumbar region. About 6 weeks previously, the pa- tient had been admitted to another hospital, where roentgenograms revealed an aortic aneurysm. The patient had no knowledge of previous syphilitic infection. The blood pressure on admission was 120 mm. Hg systolic and 80 mm. Hg diastolic. His chest was emphysematous and slightly asymmetric owing to prominence of the left anterior costal margin. However, expansion was equal bilaterally. In the epigastrium, and particularly beneath the left cos- tal margin, vigorous aortic pulsations were evident, although a definite mass was not felt. ? Laboratory studies revealed slight albuminuria and hematuria. The erythrocyte count was 4,000,- 000, hemoglobin 13.7 grams, and leukocyte count 5,250, with a normal differential count. Blood urea nitrogen was 14 mg./100 ml., and urea clearance was normal (Chart I). Hepatic function studies were also normal (Chart II). Results of the sero- logic test for syphilis were positive. Postero-ante- rior and lateral roentgenograms of the chest and abdomen revealed a large fusiform aneurysm of the lower thoracic and upper abdominal aorta, the walls of which were outlined by calcium (Fig. 3). At its origin the thoracic aorta was sharply angulated to the left, and distally the aneurysm appeared to extend as far as the third lumbar intervertebral space, displacing the esophagus and stomach anteriorly (Fig. 3). The vertebral bodies were not eroded. Intravenous pyelograms showed normal excretory renal function and suggested that the renal arteries arose from the lower portion of the aneurysm. On Oct. 19, 1955, operation was performed under general anesthesia. A left thoracolumbar ap- proach was used with the pleural cavity being entered through the bed of the resected seventh rib. Preliminary exploration disclosed a large fusi- form aneurysm arising in the lower third of the thoracic aorta and extending into the left hemi- thorax. There was sharp angulation of the thoracic aorta just proximal to the origin of the aneurysm. _ The greatest diameter of the aneurysm was at the level of the diaphragm, where it measured ap- proximately 16 cm. The mediastinal pleura over the lower thoracic aorta was incised, the aorta was mobilized above the aneurysm, and a tape was passed about it. The aneurysm extended below the diaphragm for a distance of about 15 cm., its total length approximating 24 cm. The posterior peritoneum overlying the lower abdominal aorta was next incised together with the ligament of Treitz, the aorta in the region of the inferior mes- 554 DEBAKEY, CREECH AND MORRIS Annals of Surgery October 1956 Fic. 3. Posteroanterior and lateral roentgenograms in Case 2 reveal a large tusiform aneu- rysm involving the lower thoracic and upper abdominal portions of the aorta. The walls of the aneurysm are outlined by calcium. The thoracic aorta is sharply angulated to the left at the upper end of the aneurysm. An esophogram with barium shows the esophagus to be displaced anteriorly by the aneurysm. enteric artery was exposed, and a tape was passed about it. The parietal peritoneum was _ incised from the aortic hiatus, beneath the diaphragm to the left, behind the spleen, and downward in the left colic gutter. The spleen, splenic flexure of the colon, pancreas and stomach were then mobilized and retracted to the right, thus exposing the peri- renal fat about the left kidney and the left retro- peritoneal portion of the aneurysm. The left kidney was dissected from Gerota’s fascia, its vein and artery were identified, and tapes were passed about them. From above the diaphragm the mediastinal pleural attachments to the aneurysm were incised and reflected, and the dissection was continued inferiorly to free the diaphragm and its crura from the aneurysm. The celiac, superior mesenteric, and right renal arteries were then isolated. With the exception of the inferior mesenteric artery, all major visceral branches of the aorta arose from the aneurysm (Fig. 4). In order to minimize the period of arrest of visceral blood flow during resection of the aneu- rysm, a temporary shunt made of polyvinyl sponge, with an internal diameter of 14 mm. and approx- imately 40 cm. in length, was utilized. The shunt was attached to the descending thoracic aorta above the origin of the aneurysm by an end-to-side anastomosis and then similarly implanted into the abdominal aorta just above the level of the inferior mesenteric artery (Fig. llc). The aorta proximal to the aneurysm was then doubly clamped and divided. Blood flow to the. aneurysm and _ the branches arising from it now occurred by way of the shunt. Removal of the aneurysm was begun at its proximal end, but because of the thin necrotic posterior wall and the danger of hemorrhage from perforation, an occluding clamp was applied just proximal to the superior mesenteric artery and above the origin of the renal arteries and another occluding clamp applied to the celiac axis. Blood flow to the kidneys and superior mesenteric artery was thus maintained by the shunt, but blood flow to the celiac artery was interrupted. The proximal portion of the aneurysm between the occluding clamps was then removed, and the intercostal ves- sels arising from the posterior wall of its thoracic portion were secured by suture ligature. A portion of the posterior wall could not be completely ex- cised, but the intimal layer with the attached laminated thrombi ws removed. Occluding clamps were placed on the left renal artery, and it was divided. The left renal branch of the homograft was then anastomosed to this vessel (Fig. 122). Occluding clamps were then placed on the ab- dominal aorta distal to the aneurysm and on the superior mesenteric and right renal arteries, and the remainder of the aneurysm was removed. The distal end of the homograft was anastomosed to the abdominal aorta, following which a noncrush- ing clamp was placed across the graft just above the left renal artery and just below its right ren:l branch and the occluding clamp on the abdominal aorta released to restore blood flow to the left kid- ney by way of the shunt (Fig. 12c). The graft was then anastomosed to the right renal artery, and the occluding clamp was moved above that vessel on the homograft to permit blood flow to the right Volume 144 Number 4 ANEURISM OF THORACOABDOMINAL AORTA 555. Fic. 4. Composite roentgenogram of the chest and abdomen with diagram superimposed to show location and extent of the thoracoabdominal aneurysm in Case 2. Replacement of the excised segment of aorta and aneurysm by homograft with restoration of continuity of the aorta, celiac, superior mesenteric, and renal arteries is shown in the diagram on the right. kidney. This maneuver was repeated as the anas- tomoses to the superior mesenteric and celiac ar- teries were performed. The respective periods of arrest of circulation through these vessels are shown in Chart III. Anastomosis of the upper end of the graft to the thoracic aorta completed insertion of CuartT III. Periods of Occlusion of the Celiac, Superior Mesenteric, and Renal Arteries in the Cases Reported Herein in Minutes Cuart III. Period of Occlusion in Minutes Case Celiac Superior Right Left No. Artery Mesenteric Renal Renal 1 120 105 105 105 2 44 36 ar 23 3 114 47 15 43 4 116 102 46 0 the graft. Occluding clamps were then removed to allow blood to flow in a normal fashion through the thoracic and abdominal portions of the aorta and into its major branches (Fig. 4). Exclusion clamps were reapplied, the shunt was removed, and the openings were closed with interrupted silk sutures (Fig. 5). Incisions in the posterior peri- toneum over the abdominal aorta and beneath the diaphragm and in the left colic gutter were closed with continuous catgut sutures. The spleen was removed because of laceration. The aortic hiatus was then reconstructed, and the diaphragm was repaired. A catheter was inserted into the left pleural space through the eighth intercostal space for underwater drainage, and the thoracic and ab- dominal wounds were closed. A tracheostomy was then performed. Operation lasted 8% hours. During the proce- dure 4,500 ml. of whole blood was administered, 556 DEBAKEY, CREECH AND MORRIS Fic. 5. Photograph made at operation in Case 2 showing the homograft in place. The celiac, superior mesenteric, and left renal arteries are vis- ible, but the right renal artery lies behind the superior mesenteric artery. The proximal site of implantation of the shunt is just above the proximal aortic anastomosis. and at the end of operation the patient’s general condition appeared satisfactory. During the first postoperative day, the urinary output was 15 ml. By the end of the second post- operative day, however, urinary output had in- creased to 162 ml. and renal function studies showed severe depression (Chart I). There was progressive increase in urinary volume during the next few days and by the fifth day the 24 hour urinary volume was 1,240 ml. The blood urea nitrogen reached a maximal level of 79 mg./100 ml. on the 3rd postoperative day and subsequently declined to a normal level by the 11th postopera- tive day. Renal function studies showed a return to normal on the 23rd day (Chart I) and liver function studies were essentially normal on the 19th day (Chart II). Good peristalsis was evident on about the 4th postoperative day, and the patient was eating nor- Annals of Surgery October 1956 Fic. 6. Intravenous pyelogram in Case 2 four months after operation reveals normal function of both kidneys. mally by the end of the lst week. A necrotic area in the left groin resulting from extravasation of norepinephrine on the first postoperative day was debrided on the 26th postoperative day, and 15 days later the area was covered with a split thick- ness skin graft. The patient was discharged on Jan. 4, 1956, fully recovered and in good condition. He returned for examination on Feb. 19, 1956. Intravenous pyelograms disclosed normal function of both kidneys (Fig. 6). Discrete renal function studies at this time were found to be normal, the actual figures for renal blood flow, renal plasma flow and glomerular filtration rate being respec- tively 792, 475 and 68 ml./minute. An aortogram Volume 144 Number 4 ANEURISM OF THORACOABDOMINAL AORTA 557 Fic. 7, Aortogram made five months after operation in Case 2 reveals normal patency of the aorta, celiac, superior mesenteric and renal arteries. on March 23, 1956 demonstrated patency and nor- mal filling of the renal, superior mesenteric, and celiac arteries (Fig. 7). Case 3. P. T., a man 42-years-old, was ad- mitted to the Methodist Hospital on Nov. 5, 1955, complaining of a sense of fulness and increasing pain in the upper part of the abdomen of one years duration. During the preceding 4 months, the pain had become so severe that large doses of narcotics were necessary for relief. The pain, which was most severe in the epigastrium and left upper abdominal quadrant, was not related to eating and was not relieved by antacids. An aortogram made at another hospital 3 months previously had dis- closed an aneurysm of the lower thoracic and up- per abdominal portions of the aorta. Operation was performed at that time, but because the aneurysm involved the renal arteries, it was not resected. The patient had had brucellosis many years previously and was known to have had hyperten- sion for one year prior to admission. There was no history of venereal disease. On admission, the blood pressure was 180 mm. Hg systolic and 130 mm. Hg diastolic. An opera- tive scar extended in an oblique direction from the umbilicus to the left costal margin. Examination of the abdomen was difficult because of extreme tenderness in the epigastrium, although a pulsating mass was palpable. The only significant laboratory findings were a few granular casts and a trace of albumin in the urine. Roentgenograms of the chest showed a soft tissue mass in the posterior mediastinum protrud- ing to both the right and left of the spine (Fig. 8). Abdominal films showed the mass to protrude to the left of the lumbar spine at the level of the third and fourth lumbar vertebrae. Calcification was present in the peripheral portions of the mass. Intravenous pyelograms revealed a normal kidney and ureter on the right but poor concentration on the left. An electrocardiogram disclosed normal sinus rhythm with slight left ventricular hyper- trophy and slightly prolonged P-R interval. On Nov. 10, 1955, operation was performed through a left throacoabdominal approach with excision of the previous scar. Upon retracting the lung anteriorly, the mid-descending thoracic aorta was found to be sharply angulated in a horseshoe fashion. At the lower end of this angulation was a large aneurysm, which extended distally through the diaphragm. The descending thoracic aorta above the aneu- rysm was mobilized, and a tape was placed about it. The spleen was resected, and an incision was made in the left lateral posterior peritoneum; this detached the splenic flexure of the colon and per- mitted retraction of the stomach, colon, and pan- creas to the right. Gerota’s fascia was incised, and the left kidney was mobilized from its bed and retracted to the right and inferiorly. The left renal artery was isolated and was found to arise from the lower portion of the aneurysm. The ligament of Treitz and the posterior peritoneum overlying the abdominal aorta in the region of the inferior 558 DEBAKEY, CREECH AND MORRIS Annals of Surgery October 1956 Fic. 8. Composite roentgenogram of the chest and abdomen with diagram superimposed to show location and extent of the thoracoabdominal aneurysm in Case 3. Replacement of the excised segment of aorta and aneurysm by the homograft with restoration of continuity of the aorta, celiac, superior mesenteric and left renal arteries is shown in the diagram on the right. mesenteric artery were incised. The lower abdom- inal aorta was mobilized, and a tape was placed about it distal to the aneurysm. The aneurysm was then mobilized by removal of the diaphragmatic attachments to its surface, and the superior mes- enteric and celiac arteries were isolated. The aneu- rysm extended distally to a point just beyond the origin of the left renal artery but above the origin of the right renal artery (Fig. 8). A shunt of compressed polyvinyl sponge was then anastomosed to the side of the descending thoracic aorta above the aneurysm and to the anterior aspect of the abdominal aorta below the aneurysm. The left renal artery was then divided at its origin from the aneurysm, and the left renal branch of a lyophil- ized homograft was anastomosed to it. Occluding clamps were placed on the descending thoracic aorta above the aneurysm, on the abdominal aorta just proximal to the right renal artery and on the celiac and superior mesenteric arteries, and the aneurysm was excised. The distal portion of the abdominal aortic homograft was anastomosed to the abdominal aorta. The homograft was then oc- cluded above the renal arteries. Upon removal of the abdominal aortic clamp, blood flow to the kidneys was restored through the shunt. The superior mesenteric and then the celiac arteries were next anastomosed to the graft, fol- lowing which the homograft was anastomosed to the lower end of the descending thoracic aorta, and the occluding clamp on the aorta removed with resultant restoration of normal blood flow (Fig. 8). The periods of occlusion for the left renal, superior mesenteric, and celiac arteries are shown in Chart III. The shunt was removed and the aortotomy wounds were closed. The posterior remnant of the wall of the aneurysm was over- sewn with a continuous suture for hemostasis, a pedicled omental graft was wrapped about the homograft, and the posterior peritoneum over the abdominal aorta was closed with a continuous suture. The left kidney was replaced in its bed and the peritoneum in the left gutter was reap- proximated. Appendectomy was then performed. Label 144 ANEURISM OF THORACOABDOMINAL AORTA 559 umber 4 Cuart IV. Data Relating to Renal Function in Case 3 CHART IV After Operation (days) Before Determination Operation 2 7 11 22 Urine Volume 1000 306 3595 5560 1800 ml./24 hrs. Urine Specific Gravity 1.026 1.013 1.010 1.014 Blood Urea Nitrogen 20 60 80 27 mg. % Renal Blood Flow 461 212 396 596 ml./min. Renal Plasma Flow Zit 121 226 358 ml./min. Glomerular Filtration Rate 67 19 44 59 ml./min. Plasma Sodium 22 128 132 132 128 meq./l. Plasma Potassium 5.2 3.8 4.3 4.3 ao meq./l. Sodium Excretion 63 61 280 101 meq./l. Potassium Excretion 40 62 A 38 meq./1. The diaphragm was repaired, and an intercostal drainage tube was inserted through the ninth inter- costal space and connected to water sealed drain- age. The wound was then closed. A catheter was placed in the right side of the chest through the fifth intercostal space, since the right hemithorax had been entered through the mediastinum during the procedure. The patient received 6,500 ml. of whole blood during the procedure, which lasted about 7 hours. At the conclusion of the operation the blood pressure was 150 mm. Hg systolic and 90 mm. Hg. diastolic. Within the first 90 minutes after operation, 250 ml. of urine was excreted, and urinary output con- tinued at a rate of 5 to 20 ml. per hour during the first 48 hours (Chart IV). By the 4th day, the urinary volume had reached 1,820 ml. with max- imal diuresis on the ninth day, when 5,950 ml. of urine was excreted. Subsequently, the output de- clined to normal levels. The blood urea nitrogen level had risen to 100 mg. on the 5th day but returned to normal by the tenth day. Glomerular filtration rate was severely reduced on the day after operation but returned to normal by the 22nd day. Peristaltic activity became normal by the 4th post- operative day, when the nasogastric tube was re- moved. Hepatic function studies revealed a slight increase in serum bilirubin and a cephalin floccula- tion of 2+ (€ Chart’ If). After the first postoperative week the patient required only occasional small doses of codeine and by the 10th postoperative day he was com- pletely ambulatory, asymptomatic and taking a normal diet. He was discharged on Dec. 4, 1955, in good condition. On Mar. 31, 1956, excretory pyelograms were normal. Case 4. A. G., a white man 65-years-old, was admitted to the Houston Veterans Administration Hospital Oct. 25, 1955, for treatment of a thoraco- abdominal aneurysm, diagnosed about 2 months previously at the time of subtotal gastric resection for peptic ulcer. On admission, the blood pressure was 130 mm. Hg systolic and 90 mm. Hg diastolic. Examination revealed a well healed upper abdominal midline incision and a palpable pulsating mass in the left upper abdominal quadrant. Laboratory findings were essentially normal. Roentgenograms of the chest revealed no abnormalities, but an antero- posterior roentgenogram of the abdomen disclosed calcification of the wall of the aneurysm (Fig. 9). Aortography, performed through a catheter passed through the left brachial artery into the thoracic aorta, revealed aneurysmal involvement of» both the lower thoracic and upper abdominal portions of the aorta. Intravenous pyelograms were normal, although discrete renal function studies disclosed slight reduction in glomerular filtration rate. No ab- normalities were detected in the electrocardiogram. On Nov. 12, 1955, operation was performed through a left thoracoabdominal approach. Ex- ploration revealed a fusiform aneurysm involving 560 DEBAKEY, CREECH AND MORRIS Annals of Surgery October 1956 Fic. 9. Composite roentgenogram of the chest and abdomen with diagram superimposed to show location and extent of the thoracoabdominal aneurysm in Case 4. Replacement of the excised segment of aorta and aneurysm by the homograft with restoration of continuity to the aorta, celiac, superior mesenteric and right renal arteries is shown in the diagram on the right. the distal third of the descending thoracic aorta and extending through the diaphragm to include the upper abdominal aorta. The thoracic aorta above the aneurysm was mobilized, and a tape was passed about it. The diaphragm was then incised to the surface of the aneurysm. The posterior peri- toneum beneath the left leaf of the diaphragm was incised laterally, the incision continuing be- hind the spleen and down the paracolic gutter as far as the sigmoid colon. The stomach, spleen, pancreas and colon were then mobilized and re- tracted to the right to expose the left anterolateral surface~.of the aneurysm and distal abdominal aorta below. The posterior peritoneum over the lower abdominal aoxta was incised just above the inferior mesenteric artery, the aorta below the aneurysm was mobilized, and a tape was passed about it. The left renal vein was then isolated and mobilized. The left renal artery was exposed and found to arise just distal to the aneurvsm. The right renal artery, however, had its origin from the aneurysmal sac about 4 cm. from its distal end. The superior mesenteric and celiac arteries arising from the aneurysm were likewise exposed and mobilized, and tapes were passed about them. With exclusion clamps, a temporary by-pass shunt of compressed polyvinyl sponge was then im- planted into the aorta above and below the aneu- rysm (Fig. 10). With blood flow established through the shunt, occluding clamps were applied to the aorta above the aneurysm just distal to the shunt and below to the aneurysm at a level above the origin of the left renal but below the origin of the right renal artery. After occlusion of the supe- rior mesenteric and celiac arteries, the aneurysm was opened widely and excised. A lyophilized ab- dominal aortic homograft containing the right renal, celiac, and superior mesenteric arteries was utilized to replace the excised segment of aorta. The right renal branch of the homograft was first Volume 144 Number 4 ANEURISM OF THORACOABDOMINAL AORTA 561 Fic. 10. Photograph made at operation in Case 4. The shunt (a) has been implanted into the thoracic aorta above (b) and into the abdominal aorta below (g). The celiac (c) and superior mesenteric (d) arteries have been isolated and tapes passed about them. The left renal vein identified and a tape (f) passed about it. anastomosed to the right renal artery, after which the distal end of the aortic homograft was anas- tomosed to the distal segment of aorta. By removal of the occluding clamp above the origin of the right renal vessel, blood flow was restored to the right kidney, flow to the left kidney not having been interrupted during the procedure. Anastomo- sis of the superior mesenteric artery to the graft was completed, and following this the celiac artery was anastomosed to the graft. The proximal anas- tomosis between the homograft and the thoracic aorta was next completed (Fig. 9). The shunt was removed, the abdominal viscera were replaced, and the posterior peritoneum was approximated. Sev- eral rubber drains were placed into the bed of the aneurysm along the eroded surface of the verte- bral bodies and brought out in the left flank. The diaphragm was then repaired, an intercostal cath- eter was inserted into the left pleural space for underwater drainage, and the wound was closed. The periods of occlusion for the celiac, superior mesenteric, and right renal arteries are shown in Chart III. Following operation, the blood pressure re- mained at preoperative levels. The urinary volume (e) crosses the lower portion of the aneurysm. The left renal artery has been for the first postoperative day was 199 ml. and the blood urea nitrogen was 33 mg./100 ml. Subse- quently, the daily urinary volume progressively increased to 2,020 ml. by the 8th postoperative day. Blood urea nitrogen increased to 136 mg. per 100 ml. on the 8th day and began to decline there- after. Hepatic function studies showed some devia- tions from normal, but his general condition was satisfactory (Chart II). By the 8th postoperative day the patient was ambulatory, taking a soft diet, and asymptomatic. During the following week his condition seemed to be daily improving, and ar- rangements for his discharge were being made. On the 13th postoperative day, the patient be- gan to have bloody diarrheal stools and on the following day vomited a large quantity of bright red blood and became hypotensive. At abdominal exploration, the site of bleeding was found to be a marginal ulcer at the gastrojejunostomy. The ulcer was resected, but the patient became hypoxic and died after cardiac arrest developed. At necropsy, all the anastomoses were patent, and both kidneys appeared relatively normal. 562 DEBAKEY, CREECH AND MORRIS DISCUSSION There are two important considerations in extirpation of aneurysms of this type, stemming primarily from the location and extent of the lesion. The first of these is concerned with the technical performance of the procedure and the second with the potential ischemic damage to such vital abdominal organs as the kidneys, liver, and gastrointestinal tract as a result of tempo- rary arrest of circulation to them during performance of the procedure. The former has an important bearing upon the latter, since ischemic damage to tissues is largely dependent upon the period of circulatory arrest. To be sure, there is considerable variation in the tolerance of different tis- sues to ischemia, but in all of them limits exist as to the duration of circulatory arrest that will permit subsequent survival. Al- though these limits may be affected by a number of factors, including particularly development of collateral circulation as a result of the lesion itself, this cannot always be determined with sufficient reliability prior to operation. Accordingly, this aspect of the problem, namely, prevention of fatal ischemic damage to tissues during per- formance of the procedure, assumes major significance. Two methods are available to overcome this problem. The first consists in the use of hypothermia to reduce oxygen demand by the tissues and the second in the use of a temporary shunt to conduct blood around the occluded segment with performance of the procedure in a manner to minimize the period of circulatory arrest. The former method was employed in the first case, but because it did not prove successful, the lat- ter method was used in the subsequent three cases. The gratifying results obtained in these cases suggest that it is the prefer- able method, and for this reason its more detailed consideration seems desirable. TECHNICAL CONSIDERATIONS Because of the extensive nature of these aneurysms, involving both the lower tho- Annals of Surgery October 1956 racic and upper abdominal aorta, adequate exposure is essential. This may be satisfac- torily obtained by a left thoracoabdominal approach. The patient is placed in the su- pine position with the left side of the chest slightly elevated and the left arm sus- pended from an overhead support (Fig. lla). The incision is made over the left seventh or eighth rib, extending from the midaxillary line anteriorly and obliquely across the costal margin to the midabdom- inal line and then curving inferiorly as a midabdominal incision to a point well be- low the umbilicus (Fig. lla). The left pleural and pcritoneal cavities are entered, and after division of the costal cartilages, the diaphragm is incised radially from its peripheral attachment to the aortic hiatus, and the rib-spreader is inserted. This pro- vides satisfactory exposure of the entire extent of the aneurysm as well as the aorta immediately above and below the lesion. After adequate exploration to determine more precisely the extent and nature of the lesion, it is desirable to expose, by careful sharp and blunt dissection, the aorta im- mediately above and below the aneurysm and to encircle these segments of the aorta with umbilical tape as a safety measure for control of hemorrhage. The major visceral branches, such as the celiac, superior mes- enteric and renal arteries arising from the aneurysm, are then similarly treated (Fig. 11b). In order to expose these vessels satis- factorily, as well as to permit subsequent excision of the aneurysm, a left retroperi- toneal approach is employed. This is done by dividing the posterolateral parietal peri- toneal attachment along its left border with mobilization of the visceral organs to the right side. The shunt, which has previously been prepared, is then sutured into place by end- to-side anastomosis, with the use of partial tangential occlusion clamps, to the descend- ing thoracic aorta immediately above the aneurysm and to the abdominal aorta just ahove the bifurcation (Fig. llc). For this Volume 144 Number 4 ANEURISM OF THORACOABDOMINAL AORTA : 563 Fic. 11. Drawings made at operation in Case 2. (a) The patient is placed in the supine position with left shoulder slightly elevated and left arm supported overhead. Left thoraco- lumbar incision is employed. (b) The aorta above and below the aneurysm has been exposed and the celiac, superior mesenteric, and left renal arteries and the left renal vein have been isolated and tapes passed about them. ‘fhe right renal artery lies behind the superior mesenteric artery and is not seen. (c) A shunt made of compressed polyvinyl sponge has been attached as an end-to-side anastomosis to the descending thoracic aorta above the aneurysm and to the abdominal aorta below the aneurysm. purpose, it has been found desirable to use a shunt made of compressed polyvinyl sponge (Ivalon®) with a lumen 14 mm. in diameter 7 *® (Fig. 10). Once these steps have been completed and the shunt is functioning satisfactorily and conducting blood into the abdominal aorta below the aneurysm, attention may be directed toward mobilization and resec- tion of the aneurysm. By careful sharp and blunt dissection, the aneurysm is freed from surrounding structures as well as possible, 564 DeBAKEY, CREECH AND MORRIS Annals of Surgery October 1956 Fic. 12. (a) Anastomosis of the left renal artery to its comparable segment on the aortic homograft. The technic of anastomosis here as elsewhere consists in a continuous through-and- through suture of arterial silk. (b) Anastomosis of the distal end of the aortic homograft to the abdominal aorta. (c) With the anastomosis of the homograft to the left renal artery and to the abdominal aorta completed, an occluding clamp is applied to the graft in an oblique manner so that it lies below the origin of the right renal artery and above the origin of the left renal artery. The previously applied occluding or on the abdominal aorta has now been e released, permitting restoration of blood flow to the left kidney through the shunt. The anas- tomosis to the right renal artery is then performed. (d) Anastomosis of the graft to the superior mesenteric artery. It may now be observed that the occluding clamp lies between the origin of the superior mesenteric and both renal arteries, thus permitting blood flow to be restored to both kidneys during the anastomosis of the superior mesenteric artery. (e) Anastomosis of the celiac artery. It may be noted again that the previously applied occluding clamp below the superior mesenteric artery has been moved up so as to lie immediately above it in order to permit restoration of blood flow through this vessel during the anastomosis of the celiac artery. (f) With the celiac artery anastomosis completed and the occluding clamp reapplied above it, blood flow has been restored to all the major visceral branches of the homograft during the completion of the final anastomosis to the descending thoracic aorta above. following which special noncrushing arte- rial occlusion clamps are applied to the aorta immediately below the upper attach- ment of the shunt and immediately above its lower attachment, as well as to the celiac, superior mesenteric, and renal ar- teries. By this means circulation through the aneurysm is completely .arrested, and it mav be excised with minimal blood loss. In some instances it has not been feasible to free the aneurysm satisfactorily because of its intimate adherence to surrounding structures. Under these circumstances, as tor example in Cases 1, 3, and 4, it may be preferable to apply the occluding clamps as already described and, after complete arrest of circulation through the aneurysm, it may be safely entered and removed by intramural dissection of the aneurysmal sac. When the aneurysm has thus been ex- cised, aortic continuity may be restored by means of a properly fitting abdominal aortic homograft, including the branches of the celiac, superior mesenteric, and renal Volume 144 Number 4 ANEURISM OF THORACOABDOMINAL AORTA 565 Fic. 13. Drawing showing completion of the homograft replacement in Case 2. (a) All anastomoses have been completed and blood flow through the aorta has been restored although the shunt is still in place. (b) The shunt has been removed and the openings in the aorta repaired. arteries. In order to minimize the period of circulatory arrest to the abdominal organs, particularly the kidneys, which are con- sidered more susceptible to ischemia, the anastomoses are performed in the following order. The left renal artery is attached first to its counterpart on the graft by end-to- end anastomosis, with the use of continuous through-and-through suture of 00000 arte- rial silk (Fig. 12a). Similar anastomosis is performed to unite the graft to the opening in the abdominal aorta (Fig. 12b). An oc- clusion clamp is applied to the graft some- what obliquely so that it lies above the origin of the left renal artery but below the origin of the right renal artery, and the occluding clamp on the abdominal aorta immediately above the lower attachment of the shunt is released (Fig. 12c). This per- mits restoration of circulation into the left renal artery through the shunt. The anasto- mosis to the right renal artery is performed as previously described (Fig. 12c). Upon completion of this anastomosis, an occlud- 566 | DEBAKEY, CREECH AND MORRIS ing clamp is applied to the graft above the origin of both renal arteries but below the origin of the superior mesenteric artery, and the previously applied occluding clamp is released to permit blood flow into both renal arteries (Fig. 12d). In a similar man- ner, after the anastomosis to the superior mesenteric artery has been performed, the occluding clamp is moved above it while the anastomosis is performed to the celiac axs (Fig. 12e). After completing the latter anastomosis this maneuver with the occlud- ing clamp is repeated, following which the final anastomosis is made between the graft and the opening in the lower thoracic aorta (Fig. 12f). In this manner circulation by way of the shunt is restored successively into each major visceral artery as rapidly as its anastomosis is performed. Following completion of the final anastomosis of the graft to the thoracic aorta, the occluding clamps at this level are removed and blood flow is restored in a normal manner through the graft (Fig. 13a). The shunt is then re- moved and the openings in the aorta are closed by a continuous suture of arterial silk (Fig. 13b). In cases in which the aneu- rysm has produced erosion of the vertebral bodies and a large excavated area, it is desirable to wrap the graft with a pedicled flap of omentum in order to provide it with a protecting bed of vascular tissue. The abdominal viscera are then replaced in their normal positions, and the postero- lateral peritoneal attachments on the left side are approximated. The opening in the diaphragm is closed, and after a catheter has been placed in the left pleural space for underwater drainage, the wound is closed in layers. PHYSIOLOGIC CONSIDERATIONS The ischemic effects of temporary inter- ruption of circulation to the major abdom- inal organs in man have not been clearly defined owing to the paucity of such ob- servations. Moreover, studies directed to- Annals of Surgery October 1956 ward this problem in animals have not yielded entirely satisfactory information. Thus, although the effects of ischemia on the kidney have been extensively investi- gated experimentally, variable results have been obtained depending to a large extent on the methods employed to produce renal ischemia as well as the type of animal used for this purpose. It has been demonstrated, for example, that occlusion of the renal ar- tery for two hours in unilaterally nephrec- tomized rabbits uniformly produced death from renal damage. On the other hand, in rats this procedure is apparently well tolerated, since most of the animals will survive this period of occlusion, but exten- sion of this period beyond three hours re- duces survival by about 50 per cent.27 28 In dogs longer periods of renal artery occlu- sion may also be tolerated with progressive increase in mortality after two hours.1® 17 *°,49, 44 Although renal function is tempo- rarily reduced during the early period, it gradually returns to normal over a period of two to three weeks. Similarly, autotrans- plantation of the kidney into the groin or neck, with complete interruption of its blood supply for periods up to two and one-half hours, results in severe temporary depression of renal function.*+ Renal homo- transplants apparently behave in a similar fashion.** ':°> Dempster described a type of anuria that developed in 5 to 16 per cent of animals following autotransplantation and homotransplantation of kidneys. This anuria, which differs from the early depres- sion of function customarily encountered, consists in complete failure to excrete urine after re-establishment of blood flow. It probably results from ischemia and _arte- riolar spasm incident to transplantation, since increased speed in removing and transplanting the kidney apparently de- creases the incidence of this type of anuria. The significance of the method of pro- ducing renal ischemia is well illustrated by experiments carried out in our laboratory. Volume 144 Number 4 Thus, two-hour occlusion of the aorta im- mediately above the renal arteries has a negligible effect on renal function in the dog.*! 8? Similarly, unilateral renal artery occlusion for two hours is well tolerated, although it results in moderate depression of renal function in the occluded kidney but not in the contralateral kidney.**: #1 **: 3° On the other hand, simultaneous occlusion of the aorta and one renal artery for two hours results in severe depression of renal blood flow and glomerular filtration rate in the occluded kidney.**? Particularly sig- nificant is the fact that in this latter group of animals renal function in the kidney whose renal artery was not occluded re- mains unaffected, despite the fact that the aorta was occluded immediately above this point. It is also significant that the pressure in the aorta distal to the level of occlusion was found to be approximately 25 mm. Hg. These observations would suggest that this degree of subfiltration arterial pressure re- sulting from collateral blood flow provides a high degree of protection from ischemic damage to the kidney. These experiments were repeated utilizing hypothermia, and it was found that this afforded moderate pro- tection to the kidney subjected to complete ischemia.** Much less precise information is avail- able on the ischemic effects of temporary arrest of circulation to the kidney in man.° Semb,*! in performing segmental resection of the kidneys for tuberculosis, clamped the entire renal pedicle for one and one- half hours without significant alteration of renal function. Similarly, Bahnson** re- ported survival of a patient following oc- clusion of both renal arteries for 37 min- utes, although another of his patients died from renal failure after occlusion of the aorta above the renal arteries for 110 min- utes. In the case recently reported by Ellis and associates,” occlusion of the aorta and one renal artery for 95 minutes produced no significant reduction in renal function. ANEURISM OF THORACOABDOMINAL AORTA 567 In describing the results of renal homo- transplantation in nine patients, Hume and co-workers 7° reported resumption of func- tion in kidneys that had been totally ische- mic for 200 minutes. Similarly, Merrill and associates,®° in performing homotransplan- tation of a healthy kidney from one iden- tical twin to another, observed that total anoxia of the kidney for a period of 90 minutes does not mitigate against resump- tion of adequate function. Thus, it is apparent from these experi- mental and clinical observations that con- siderable variations exist in the safe period of tolerance of the kidney to anoxia, rang- ing from less than one hour to over three hours. Obviously, there are a number of variables that influence this problem, in- cluding age, pre-existing renal disease, col- lateral blood supply, individual variations, and the like. In this connection it may be significant that three of the four patients reported herein were in the seventh decade. In these cases arrest of circulation to the kidneys for periods of less than one hour was well tolerated, but prolongation of this period for more than one and one-half hours, despite the use of hypothermia, re- sulted in fatal ischemic damage to the kidneys. In the first case in which operation was performed under hypothermia complete renal ischemia bilaterally for 105 minutes resulted in acute, severe oliguria of 48 hours’ duration, followed by gradual in- crease in urinary volume. However, in spite of the fact that urinary output had_ in- creased to 1,183 ml. on the fourth day after operation, glomerular filtration rate and renal blood flow were minimal and death from uremia occurred on the seventh postoperative day. Microscopically, there were extensive hyalin changes in the glo- meruli and tubular necrosis. These events suggest that the fatal renal failure was a direct result of prolonged renal ischemia. If the patient could have survived the first 568 ! DEBAKEY, CREECH AND MORRIS week, perhaps. renal recovery would have taken place, a possibility that is suggested by the experience with renal autotrans- plants.13> 14 34 In the second case blood flow to the right and left kidneys was interrupted for 27 and 23 minutes, respectively (Chart III). The pattern of recovery was similar to that ob- served in acute renal failure from other causes and consisted of early oliguria fol- lowed by diuresis with return to normal function about the third week. Discrete renal function studies in this case point to ischemic involvement of the entire nephron initially with glomerular recovery preced- ing tubular recovery (Chart I). _ Renal response in the third case was strikingly similar to that in the second case although the right kidney was ischemic for only 15 minutes, while the left was ische- mic for 43 minutes. In this instance the oc- currence of pronounced diuresis suggested more extensive tubular involvement. None- theless, recovery was practically complete by the end of the third week. The response in the fourth case was somewhat similar, although as might be anticipated from the unilateral ischemia, depression of renal function was only moderate and recovery was more rapid than in the other cases. These limited observations do not permit definite conclusions to be drawn regarding the maximum safe period of complete renal ischemia. They do, however, point to the importance of minimizing this period. It seems reasonable to state that in cases of this kind periods up to 45 minutes may be safely tolerated, but extension beyond this time is increasingly hazardous. It would also appear that hypothermia of moderate degree, ie., with reduction of body tem- perature to 32.2° C. (90° F.), cannot be relied upon to provide adequate protection against fatal ischemic damage to the kid- neys when the period of ischemia extends over 100 minutes. Whereas observations on the effects of renal ischemia in man are relatively few, Annals of Surgery October 1956 there are even fewer data pertaining to the effects of ischemia on the liver and gas- tro-intestinal tract. There are a number of older reports of ligation of the thoracic aorta for aneurysm and considerable recent experience with temporary occlusion for resection of aneurysms,'*!2 but these are not pertinent since only partial ischemia of these structures is produced in this way. Nor are the observations following celiac or hepatic artery ligation for bleeding esopha- geal varices relevant, since an extensive col- lateral network is still functioning. The tolerance of the liver to ischemia has been extensively investigated in experimen- tal animals, but the results of these studies have been quite variable.*® 1. 18-20, 24, 29, 35, “6 88, 48 According to these experiments, for example, the maximum safe period of arrest of circulation to the liver varies from some- what less than one half hour to a little more than one hour. These variations in results are apparently due to a number of factors, including particularly the different species of experimental animals used, the different methods of producing arrest of hepatic cir- culation with consequent differences in the degree of completeness of occluding the afferent hepatic circulation, the presence of anomalous or accessory arteries to the liver, and the administration of antimicrobial agents. For these reasons it is difficult to evaluate these experimental observations in determining more precisely the critical pe- riod of hepatic anoxia. The weight of evi- dence would suggest, however, that this period closely approaches one half hour. There is also some experimental evidence to suggest that this period may be pro- longed to one hour by the use of hypo- thermia.® 37: 42 This problem in man is even less well defined owing to the paucity of such ob- servations. Wangensteen *> reported per- formance of left hepatic lobectomy in three cases with simultaneous occlusion of the hepatic, gastroduodenal, and superior mes- enteric arteries and the portal vein. In the Volume 144 Number 4 two patients who survived operation the periods of ischemia were 33 and 24 minutes respectively, whereas in the patient who died there were two periods of ischemia of 12 and 15 minutes each interrupted by a 20-minute period of release of ischemia. Two somewhat similar cases of left hepatic lobectomy for metastatic carcinoma have also been reported by Burch, Traphagen, and Folkman.‘ In their first patient who re- covered, the aorta above the celiac axis and the portal vein were occluded for 10 min- utes with no apparent disturbance in liver function. Similar occlusion for the same period of time was done in the second case, but the patient developed extensive mes- enteric thrombosis which subsequently caused his death. More recently, Shumway and Lewis #2 reported four cases of right hepatic lobectomy performed under hypo- thermia in which temporary arrest of he- patic circulation was produced by occlu- sion of the thoracic aorta at the level of the tenth intercostal space, the inferior vena cava above and below the liver, the hepatic artery, and the portal vein. In the two pa- tients who survived operation the period of ischemia was 45 minutes, while in the other two patients who died it was 33 and 40 minutes, respectively. The four cases reported herein afforded a unique opportunity to observe the effects of relatively complete interruption of arte- rial blood supply to the liver and gastro- intestinal tract. Thus, with the aorta oc- cluded above the diaphragm and below the renal arteries, and with celiac and superior mesenteric arteries interrupted, the liver was nourished only by the portal venous flow. The effectiveness of this afferent nu- trient vessel to the liver was also greatly diminished through reduction of arterial blood flow to the portal bed since this was derived primarily from the inferior mes- enteric artery. In the three successful cases the periods of hepatic ischemia ranged from 44 to 116 minutes. Yet in no instance ANEURISM OF THORACOABDOMINAL AOKTA 569 was hepatic function seriously deranged (Chart II). It cannot be stated with cer- tainty that hypothermia was responsible for the apparent tolerance of the liver to ische- mia of two hours in the first case, since the patient survived only one week. There was no evidence, however, during this time nor at autopsy of significant liver derangement. The most consistent changes were noted in serum bilirubin levels and in the cephalin flocculation. In view of the large amount of whole blood administered during opera- tion, however, the slight increase in serum bilirubin does not appear to be significant. Accordingly, these observations would sug- gest that in man the liver can tolerate oc- clusion of its major arterial blood supply for relatively long periods, almost two hours, without significant alteration in function. In the first case arterial blood flow to the entire gastro-intestinal tract was interrupted for 105 minutes. Postoperatively, there were moderately severe intestinal distention and diarrhea suggesting some disturbances but otherwise no evidence of serious ischemic damage to these organs. In the remaining cases the superior mesenteric artery was occluded for periods ranging from 36 to 102 minutes, but a temporary aortic shunt maintained blood flow through the inferior mesenteric artery so that the gastro-intes- tinal tract was not completely ischemic. In none of these cases was intestinal distention or diarrhea a problem, and resumption of gastro-intestinal function took place in a normal manner following operation. SUMMARY 1. Four cases of extensive thoracoabdom- inal aneurysm of the aorta treated by re- section and homograft replacement are re- ported. In all of these cases the aneurysm extended from the lower descending tho- racic aorta to the lower abdominal aorta and involved the celiac, superior mesen- teric, and one or both renal arteries. The 570 DEBAKEY, CREECH AND MORRIS operative procedure consisted in excision of the aneurysm and replacement with an aortic homograft with restoration of con- tinuity to the celiac axis and superior mes- enteric arteries in all as well as to both renal arteries in two and to one of the renal arteries in the other two cases. 2. The most important consideration in extirpation of aneurysms of this type arises from the potential ischemic damage to such vital organs as the kidneys, liver, and gas- tro-intestinal tract as a consequence of temporary arrest of circulation to them dur- ing performance of the procedure. Two methods are available to overcome this problem, namely hypothermia and the use of a temporary shunt to conduct blood around the occluded segment with the per- formance of the procedure in a manner to minimize the period of circulatory arrest. 3. Hypothermia was used in the first case, but the patient died of renal failure one week after operation. The superior mesenteric and both renal arteries were occluded for 105 minutes and the celiac artery for 120 minutes. 4. In the other three cases temporary shunts were employed permitting signif- icant reduction in the period of temporary arrest of circulation to the abdominal vis- cera. The successful results obtained in these cases emphasize the importance of performing the procedure in such a man- ner as to minimize the period of ischemia to these vital structures. 5. In the three successful cases the pe- riod of occlusion of the renal arteries ranged from 15 to 46 minutes. Serial renal function studies in these cases revealed a characteristic pattern of depression of func- tion, as reflected by increase in the blood urea nitrogen level and significant reduc- tion in renal blood flow and glomerular filtration rate, during the first four or five days after operation with progressive re- turn to normal during the subsequent ten days to two weeks. Annals of Surgery October 1956 6. The period of occlusion of the celiac artery in these cases ranged from 44 to 116 minutes. Studies of hepatic function re- vealed no significant alterations. 7. The period of arrest of circulation through the superior mesenteric artery varied from 36 to 102 minutes and no sig- nificant disturbances in gastro-intestinal function were observed. BIBLIOGRAPHY 1. Badenoch, A. W. and E. M. Darmady: The Effects of Temporary Occlusion of the Renal Artery in Rabbits and its Relationship to Traumatic Uraemia. J. Path. and Bact., 59: 79, 1947. 2. Bahnson, H. T.: Definitive Treatment of Sac- cular Aneurysms of the Aorta with Excision of Sac and Aortic Suture. Surg., Gynec. and Obst., 96: 383, 1953. 3. Bahnson, H. T.: Treatment of Abdominal Aortic Aneurysm by Excision and Replace- ment by Homograft. Circulation, 9: 494, 1954. 4. Burch, B. H., D. W. Traphagen and M. J. Folkman: The Use of Aortic Occlusion in Abdominal Surgery with a Report of Two Human Cases. Surgery, 34: 672, Oct. 1953. 5. Burch, B. H., D. W. Traphagen, M. J. Folk- man, D. A. Rosenbaum and E. C. Mueller: Temporary Aortic Occlusion in Abodminal Surgery. Surgery, 35: 684, May 1954. 6. Child, C. G., Ill, R. D. McClure, Jr. and D. M. Hays: Studies on Hepatic Circulation in Macaca Mulatta Monkey and in Man. Surg. Forum, American College of Surgeons, p. 140, 1952. 7. Cooley, D. A., D. E. Mahaffey and M. E. DeBakey: Total Excision of the Aortic Arch for Aneurysm. Surg., Gynec., & Obst., 101: 667, Dec. 1955. 8. Creech, O., Jr.. M. E. DeBakey and D. E. Mahaffey: Total Resection of the Aortic Arch. Surgery (In press). 9. Creech, O., Jr., M. E. DeBakey, G. C. Morris, Jr. and J. H. Moyer: Experimental and Clin- ical Observations on the Effects of Renal Ischemia. Surgery, 40: 129, July 1956. 10, DeBakey, M. E., D. A. Cooley and O. Creech, Jr.: Resection of the Aorta for Aneurysms and Occlusive Disease with Particular Refer- ence to the Use of Hypothermia. Analysis of 240 Cases. Trans. Amer. Coll. Cardiology, S: 153, 1055: Volume 144 Number 4 Li: 12. 13. 14. 15. 16. ki, 18. 19. 20. 22. 24. 25. DeBakey, M. E., D. A. Cooley and O. Creech, Jr.: Resection of Aneurysms of the Thoracic Aorta. Surg. Clin. N. America, W. B. Saunders Co., Philadelphia. Aug. 1956, pp. 969-982. DeBakey, M. E., G. C. Morris, Jr. and J. H. Moyer: Unpublished data. Dempster, W. J.: The Anurias Following Kid- ney Transplantation. Acta med. scandinav., 148: 91, 1954. Dempster, W. J. and A. M. Joekes: Functional Studies of the Kidney Autotransplanted to the Neck of Dogs. Acta med. scandinav., 147: 99, 1953. DeWeese, M. S. and C. Lewis: Partial Hepat- ectomy in the Dog. Surgery, 30: 642, 1951. Donnet, V., S. Maffre and A. Pruneyre: Evolu- tion de la Diurese Aprés Occlusion Tem- poraire ou Prolongée d’une Artére Rénal chez le Chien non Narcosé. J. Physiol., 43: 110. 195%. Donnet, V. and A. Pruneyre: Rétablissement des Fonctions du Vein Soumis 4 une Anémie Expérimentale Prolongée. Compt. rend. Soc. biol., 147: 442, 1953. Drapanas, T., D. R. Becker, G. S. Alfano, W. H. Potter and J. D. Stewart: Some Effects of Interrupting Hepatic Blood Flow. Ann. Surg., 142: 831, Nov. 1955. Duchinova, S. L.: Ueber temporaire Abklem- mung des Lig. Hepato-Duodenale fiir blut- lose Operatione an der Leber. Zentralorg. f. d. ges. Chir., 35: 581-582, 1925. Edwards, W. S., O. K. Tidwell and C. R. Lombardo: The Mechanism of Death from Thoracic Aortic Occlusion. Surg. Forum, American College of Surgeons, p. 90, 1954. “Ellis, Fo A, 32) 3 elercen and E. A. Hines, Jr.: Aneurysm of the Abdominal Aorta Involving the Right Renal Artery. Ann. Surg., 142: 992, 1955. Etheredge, S. N., J. Yee, J...V.. Smith, -S. Schonberger and M. J. Goldman: Successful Resection of a Large Aneurysm of the Upper Abdominal Aorta and Replacement with Homograft. Surgery, 38: 1071, Dec. 1955. . Friedman, S. M., R. L. Johnson and C. L. Friedman: The Pattern of Recovery of Renal Function Following Renal Artery Occlusion in the Dog. Circulation Res., 2: 231, 1954. Goodrich, E. O., Jr., H. F. Welch, J. A. Nel- son, T. S. Beecher and C. S. Welch: Homo- transplantation of the Canine Liver. Surgery, 39: 244, Feb. 1956. Hamilton, P. B., R. A. Phillips and A. Hiller: Duration of Renal Ischemia Required to Produce Uremia. Am. J. Physiol., 152: 517, 1948. ANEURISM OF THORACOABDOMINAL AORTA 26. 27. 28. 29. 30. ol; 32. 33. 34. 35. 36. 37, 38. 39. 571 Hume, D. M., J. P. Merrill, B. F. Miller and G. W. Thorn: Experiences with Renal Homo- transplantation in the Human: Report of Nine Cases. J. Clin. Invest., 34: 327, Feb. 1955. Koletsky, S.: Effects of Temporary Interrup- tion of Renal Circulation in Rats. Arch. Path., 58: 592, 1954. Koletsky, S. and B. J. Dillan: Survival of Rats after Temporary Complete Renal Ischemia. Proc. Soc. Exper. Biol. & Med., 70: 14, 1949. Markowitz, J., A. Rappaport and A. C. Scott: Prevention of Liver Necrosis Following Liga- tion of the Hepatic Artery. Proc. Soc. Exper. Biol. and Med., 70: 305, 1949. Merrill, J. P., J. E. Murray, J. H. Harrison and W. R. Guild: Successful Homotrans- plantation of the Human Kidney Between Identical Twins. J. A. M. A., 160: 277, Jan. 28, 1956. Morris, .G.. C.,::Jr,; Gc Fi: Heer and J.-H. Moyer: The Protective Effect of Subfiltra- tion Arterial Pressure on the Kidney. Surg. Forum, American College of Surgeons, p. 623, 1956. Morris, G. C., Jr., C. F. Heider, J. H. Moyer and C. Handley: The Effect of Various Types of Renal Ischemia on Renal Function. Clin. Res. Proceedings, 4: 42, 1956. Moyer, J. H., G. C. Morris, Jr. and M. E. De- Bakey: Renal Functional Response to Hypo- thermia and Ischemia in Man and Dog. National Research Council, Conference on the Physiology of Induced Hypothermia, Oct. 28-29, 1955, Washington, D. C. Murray, G. and R. Holden: Transplantation of Kidneys, Experimentally and in Human Cases. Am. J. Surg., 87: 508, April 1954. Popper, H. L., .N. (C, Jefferson and. H. Necheles: Survival of Dogs After Partial or Total Devascularization of the Liver. Ann. Surg., 140: 93, 1954. Raffucci, F. L.: The Effects of Temporary Oc- clusion of the Afferent Hepatic Circulation in Dogs. Surgery, 33: 342, March 1953. Raffucci, F. L., F. J. Lewis and O. H. Wan- gensteen: Hypothermia in Experimental He- patic Surgery. Proc. Soc. Exper. Biol. & Med., 83: 639, 1953. Rappaport, A. M., W. N. Lotto and W. M. Lougheed: Experimental Hepatic Ischemia; Collateral Circulation of the Liver. Ann. Surg., 140: 695, 1954. 2 Roof, B. S,, H. D.,; Lauson,: §, 'T.. Bella and H. A. Eder: Recovery of Glomerular and Tubular Function, Including P-aminohip- 572 DEBAKEY, CREECH AND MORRIS purate Extraction, Following Two Hours of Renal Artery Occlusion in the Dog. Am. J. Physiol., 166: 666, 1951. 40. Scheibe, J. R., E. Giraldi and C. W. Ver- meulen: The Effect of Temporary Renal Vascular Occlusion on Kidney Function. Surgery, 25: 724, 1949. 41. Semb, C.: Renal Tuberculosis and its Treat- ment by Partial Resection of the Kidney. Acta chir. scandinav., 98: 457, 1949. 42. Shumway, N. E. and F. J. Lewis: Hypo- thermia and Temporary Occlusion of the Hepatic Circulation. National Research Council, Conference on the Physiology of Annals of Surgery October 1956 Induced Hypothermia, Oct. 28-29, 1955, Washington, D. C. 43. Tinker, M. B. and M. B. Tinker, Jr.: Resec- tion of the Liver. J. A. M. A., 112: 2006, 1939. 44, Van Slyke, D. D., R. A. Phillips, P. B. Hamil- ton, R. M. Archibald, V. P. Dole and K. Emerson, Jr.: Effect of Shock on the Kidney. Tr. A. Am. Physicians, 58: 119, 1944. 45. Wangensteen, O. H.: Discussion of paper by Ravdin and Vars: Further Studies on Factors Influencing Liver Injury and Liver Repair. Ann. Surg., 132: 362, 1950. Discussion.—Dr. JoHNn H. Gipson, Jr., Phila- delphia, Pennsylvania: I make no apology for getting up a third time this afternoon to pay tribute to one of America’s and one of the world’s greatest surgeons. We have not only heard a description of one of the most brilliant technical achievements that to my knowledge has been accomplished in the last few years in the field of vascular surgery, but we have also heard a presentation by a man who is extraordinarily modest about his achieve- ments, and who not only in the technical field but in the field of devotion to the interests of surgeons and surgery at large, and surgical education, has had an extraordinary career. I think this Associa- tion should take note of the tremendous achieve- ments of this man. I believe we have a right to be proud of him as a member of this Association. (Applause ) Dr. HeNry T. BAHNson, Baltimore, Maryland: To hear a monumental and simply thrilling feat in surgical treatment emanating from Houston is cer- tainly not unusual and hardly remarkable. This afternoon is no exception. I am sure Dr. Dean Warren and I are delighted to serve as appetizers for this entree. (Laughter) For those of you who, like myself, perhaps are not quite so skillful, I think it should be men- tioned, and certainly not to detract from this work, that many of these aneurysms that are located in this area anse posteriorly. In many instances they do not involve the mouths of the visceral vessels. In two cases that we have seen in a much smaller group of patients, one was able to clamp the aorta above and below and to perform a posterolateral aortorraphy. It requires a much shorter time, and I believe is less risky, although admittedly the end result may not be as good as the procedure Dr. DeBakey has described. PRESIDENT Biatockx: I would like to say a word about a great man. One is now approaching the rostrum. Dr. ARTHUR H. BLAKEMORE, New York, New York: I was very embarrassed in New York just a short time before I left. A 79-year-old gentleman arrived at 5:30 a.m. from Binghamton, with a sup- posedly ruptured aneurysm of the abdominal aorta. Fortunately, the patient was in good condition upon arrival and at operation the aneurysm was found not to be ruptured; apparently the rapid development of a secondary sacculation of the aneurysm had been the cause of severe pain im- mediately following the act of shoveling water in a Hooded cellar. Whereas there was no blood to be seen surrounding the aneurysm, there was marked inflammatory edema of the surrounding tissues and aneurysm wall. The latter was necrotic in places. A finding of great concern and considerable em- barrassment at the time in this patient was in- volvement of the entire abdominal aorta by aneurysm. The secondary sacculation, starting im- mediately distal to the renal arteries, expanded rapidly to a maximum diameter of 15 centimeters, but fortunately contracted nearing the bifurcation level finally to a diameter approximating 6 centi- menters. Whereas proximal to the renal arteries, upward to the diaphragm, the aneurysm ranged from 7 to 9 centimeters in diameter. The aneurysm had the characteristic appear- ance of the arteriosclerotic variety. The patient’s systolic blood pressure ranged from 170-200 mm. Hg, and diastolic from 95-105 mm. Hg. It was realized at operation that rupture of the aneurysm sacculation below the renal arteries was imminent. Fortunately, a line of section relatively free of sclerotic plaques could be made distal to the renal arteries. The difficulty was the apparent diameter at this level approximated 9 centimeters, compared to approximately a diameter of 6 centimeters at the bifurcation. Encouragement to try resection of the aneurysm at the above levels was forthcoming when Doctor Arthur Voorhees produced a braided tube of orlon measuring 4 centimeters in diam- eter, the largest diameter that was leakproof to blood that the present braiding machine will fab- ricate. As it worked out, the proximal and distal diameters shrank a centimeter or two following resection of the aneurysm, and the ends of the braided tube dilated to accommodate the suture Volume 144 Number 4 line. Thus, with this special feature of the braid, the prosthesis became evenly streamlined toward the ends compensating to form even junctions at the sites of anastomosis. Following removal of the occluding clamps, the “run-off” was good, with all distal arteries patent. Fortunately, there were no calcific plaques encountered in the suture line and the overall quality of the tissue was, if anything, better than in some patients in the past in whom successful anastomoses were achieved. Recovery in this patient was uneventful, and I don't believe the aneurysm above the prosthesis is going to rup- ture in this old gentleman. In my entire experience I have seen only one ruptured arteriosclerotic aneurysm of the aorta oc- curring proximal to the renal arteries, and that one occurred in the ascending arch. This statement in no sense detracts from the fact that aneurysms do rupture, but rather to emphasize the extreme pre- dilection to rupture of arteriosclerotic aneurysms of the terminal aorta. Dr. DeBakey and his able associates have pur- sued relentlessly the old demon, aneurysm, up and down the aorta. Today it may be safely stated they have staged the “last round up.” This final conquest, embracing the ideal management for aneurysms of the abdominal aorta that involve the visceral arteries, will go down in the Annals of Surgery as one of the very great advances in vascular surgery. I take great pleasure in con- gratulating Dr. DeBakey on a perfectly magnificent achievement. Dr. MicuHaeL E. DeBaxkey, Houston, Texas: It would be impossible for me to express ade- quately my grateful appreciation for the very kind and extremely generous remarks that have been ANEURISM OF THORACOABDOMINAL AORTA 573 made. I don’t want to appear overmodest about it, but I do feel very sincerely that this was motivated in many ways by Dr. Gibbon’s sincere friendship. As far as the feasibility of using other means of attacking this problem is concerned, the prob- lem of aneurysms involving this segment of the aorta, I am sure there are instances in which the dilatation in the abdominal aorta below the renals, which tends to extend up and involve probably the upper segment of the abdominal aorta, can be treated, as Dr. Blakemore said. However, I would doubt that this would be regarded as an effective means of treatment of aneurysms of the kind to which I refer, since all of these aneurysms began actually in the ascending thoracic aorta, in the terminal portion and involved the abdominal portion. Secondly, the pathological features of these aneurysms after they were excised demonstrated the nature of the disease process with the progres- sive necrosis in the terminal portion, one, for ex- ample, during the dissection and immobilization; the necrosis was so extensive that we inadvertently entered the aneurysm. Fortunately we were already prepared to occlude the aneurysm, and did not have any trouble. I might say that technically it is often easier to be prepared to enter the aneurysm by arresting or occluding circulation to the vessels which arise from the aneurysm. Then one can enter the aneurysm quite safely and view the dissection and do the surgery end-to-side. It is a simple way to attack aneurysms of this extensive nature. Again I want to express my gratitude for the very generous remarks made by the discussers. Thank you.