- Email: [email protected]
Renal endarterectomy through the transected aorta: A new technique for combined aortorenal atherosclerosis—a preliminary report
Renal endarterectomy through the transected aorta: A new technique for combined aortorenal atherosclerosis--A preliminary report R o n a l d J. Stoney, M D , Louis M. Messina, M D , Jerry G o l d s t o n e , M D , and L i n d a M. ReiUy, M D , San Francisco, Calif. The optimal surgical management of combined aortic and renal atherosclerosis has not been defined. A modified technique of renal endarterectomy performed through the transected aorta before anastomosis of the aortic prosthesis is presented. The surgical course and early follow-up results of the first 44 consecutive patients treated by this technique are reported. Twenty-three patients (25%) had symptomatic coronary artery disease, 36 patients (82%) had arterial hypertension, and 28 patients (64%) had elevated serum creatinine levels (mean 2.10 mg/dl). Primary indications for operation related to aortic disease in 39 patients (89%) and to poorly controlled hypertension in five patients (11%). Seventy-five renal arteries were revascularized; 11% of these arteries were occluded. Revascularizations were bilateral in 27 patients (61%), involved a solitary kidney in three patients (6%), and were unilateral in 14 patients (32%). Aortobffemoral bypass grafting was performed in 16 patients (36%), aortoiliac bypass was performed in 15 patients (34%), and infrarenal aortic replacement was performed in 11 patients (25%). Postoperative complications developed in 14 patients (32%). There were two postoperative deaths (4%). Nineteen patients had transient increases in serum creatinine levels >0.5 mg/dl. Two patients required postoperative dialysis, neither related to failure of their renal:revascularization. There were five late deaths, all as a result of myocardial infarction. Renal function improved in four patients and remained stable in 22 patients. Blood pressure was improved in 64% of patients previously hypertensive. These preliminary results indicate that renal endarterectomy through the transected aorta along with aortic reconstruction is a safe, efficient, and effective means of treating these complex lesions. (l VASe SURG 1989;9:224-33.)
The first successful renal revascularization for renal artery stenosis was performed by Edwin J. Wylie on Nov. 10, 1952, although not reported at that time? T h e patient was a 46-year-old woman who had severe hypertension, a left renal artery stenosis, and an infrarenal aortic occlusion. This combination o f renal artery and aortic disease was cured by an aortoiliac and renal endarterectomy. Although this first renal revascularization was performed as part o f From the Section of Vascular Surgery, Department of Surgery, University of California. Supported in part by the PacificVascular Research Foundation, San Francisco, Calif. Presented at the Thirty-sixth Scientific Meeting of the Nort_h American Chapter, International Society for Cardiovascular Surgery, Chicago, Ill., June 14-15, 1988. Reprint requests: Louis M. Messina, MD, Section of Vascular Surgery,Universityof Michigan Medical Center, 2210 L Tanbman Health Care Center, 1500 E. Medical Center Dr., Ann Arbor, MI 48109-0329. 224
a combined aortic and renal reconstruction, over the ensuing 3 decades the management o f focal or isq~ lated renal artery disease has been the most su& ~ cessfu[. 2 However, at our institution and at other centers 3,4 that perform renal artery reconstructions regularly, a distinct change in the profile o f patients requiting renal revascularization has been noted. Elderly patients who h a v e generalized atherosclerosis now have combined symptomatic aortic and renal disease, often complicated by severe hypertension and azotemia. The renal occlusive lesions are typically bilateral (Fig. 1). They exhibit clinical features that have been repeatedly associated with increased mottalky compared to the management o f isolated renal or aortic disease. These risk factors include the presence o f bilateral renal artery stenoses, 5 impaired renal fianction, 6 and the presence o f aortic aneurysmal disease. 7-9Untreated, these lesions o f the aorta and renal arteries carry an ominous clinical prognosis? ° The
Volume 9 Number 2 February 1989
Modified transaortic endarterectomy 225
Fig. 2. Lateral view of aorta after dissection shows celiac artery (ca), superior mesenteric artery (sma), right renal artery (rra), and left renal artery (Ira). Fig. 1. Incision of crura of the diaphragm to facilitate posterolateral mobilization of the aorta. optimal surgical management of these complex, highrisk patients has not been defined. Transaortic renal endarterectomy through a longitudinal aortotomy has proved to be effective and durable in the treatment of patients with renal artery occlusive disease. 1 However, when concurrent aortic grafting is necessary, transaortic endarterectomy through a longitudinal aortotomy necessitates intersection of the suture lines of the aortotomy and graft anastomosis. Under these circumstances, reimplantation of the renal ~arteries or multiple renal artery grafting has been advocated. The increasing number of patients requiring combined aortic and renal reconstruction and our continuing satisfactory experience with transaortic renal endarterectomy stimulated the development of a modified technique for ~,ansaortic renal endarterectomy through the transected infrarcnal aorta preceding proximal aortic graft anastomosis. The purpose of this article is to describe this modified technique of transaortic endarterectomy and to report the clinical course and early follow-up of the first 44 consecutive patients treated by this method.
OPERATIVE TECHNIQUE Transaortic renal endarterectomy through the transected abdominal aorta permits one to successfully manage the typical features of symptomatic combined aortic and renal artery atherosclerotic disease. The renal artery lesions are frequently proximal extensions of aortic atherosclerosis typically involving both or multiple renal arteries. This technique permits satisfactory disobliteration of single or multiple renal arteries and concomitant aortic grafting,
requiring only a single suture line and a short, usually well-tolerated period of obligatory renal ischemia. Either a long midline incision or a generous transverse incision is made depending on the patient's body habitus. The aorta may be exposed by a transperitoneal incision. Alternatively, the upper abdominal aorta may be approached by incision of the left lateral peritoneal reflection and medial rotation of the abdominal viscera. Careful and precise dissection of the upper abdominal aorta and its visceral branches is critical to the success of this operation. Dissection of the pararenal aorta commences with complete mobilization of the left renal vein from the confluence of its renal branches to the vena cava. The posterior tributary of the renal vein to the ascending lumbar vein is routinely transected and ligated. Similarly the adrenal and gonadal veins are sacrificed as necessary to permit tension-free manipulation of the renal vein. The aorta at and above the level of the superior mesenteric artery (SMA) is encased in a dense layer of neural tissue. This neural tissue should be incised to expose the anterior and lateral surfaces of the aorta. The lateral and posterior surfaces of the aorta at this same level are encircled to a variable extent by the crura of the diaphragm. Sharp division of the crura greatly facilitates further mobilization of the posterior aorta (Figs. 1 and 2), thus permitting occlusion of the aorta by the vertical placement of an aortic clamp. MobiJ,ization of the renal arteries should be extended beyond the palpable extent of the renal atherosclerosis. After systemic heparinization, clamps are first applied individually to the superior mesenteric and renal arteries, and finally to the aorta proximal and distal to the site of proposed aortic reconstruction and to any intervening intercostal or lumbar arteries.
226
Journal of VASCULAR SURGERY
Stoney et al.
Fig. 3. Left, Vertical placement of an aortic clamp proximal to SMA. Aorta is transected just below renal arteries. Right, Aortic endarterectomy commences by developing the plane within the media on anterior ao~ic wall.
Fig. 4. Left, After dissection of the aortic plaque circttmferentially. Right, LRA endarterectomy performed by maintaining gentle traction on aortic atheroma; this permits prolapse of renal artery into the aorta. This prolapsing of the renal artery into the aorta is facilitated fmTJaer by pushing the damp occluding the renal artery toward the aorta.
The proximal aortic clamp should be placed above any significant aortic atherosclerotic lesion and at a level that will permit an uncompromised aortic and renal endarterectomy (Fig. 3). Most frequently this is above the level of the SMA. The aorta is then transected immediately below the level of the renal arteries. The distal 3 cm of the aorta down to the level of the inferior mesenteric artery is resected to provide space for an unrestricted end-to-end anastomoses of the prosthetic graft to the aorta after endarterectomy. Transaortic endarterectomy of the proximal segment is facilitated by gently rotating the aortic clamp to evert the aortic stump anteriorly and thus permit a direct view of the aortic lumen (Fig. 3). An endarterectomy dissector is used to develop and circumferentially extend the plane between the aortic plaque and the media. This dissection is continued proxi-
Fig. 5. Left, Endarterectomy plane is extended by using dissector to push renal artery wall away from the lesion. Right, With continued traction on the aortic lesion a well-defined end point is feathered and the lesion breaks away.
mally to a level immediately distal to the S/vIA orifice. It is occasionally helpful to longitudinally split r3~ aortic atherosclerotic lesion. This maneuver allows separate transaortic renalendarterectomy without the still-attached contralateral renal atheroma tethering the spccimcn and limiting mobility. This endarterectomy plane is extended in continuity into each of the renal arteries. Thc renal cndartercctomy is performed by maintaining gentle traction on the aortic lesion; this prolapses the renal artery into the aortic lumen (Figs. 4 and 5). Separation of the prolapsing renal artery media from the emerging core of atherosclerotic intima of the renal artery is achieved by use of the dissector blade to gently push the residual arterial wall away from the lesion. With continued traction on the specimen, a well-defined end point is usually obtained and the specimen separates easily. After each individual renal lesion has been removed, the plaque ~. transected just distal to the SMA orifice (Figs. 6 an8 7). The aorta and renal arteries are irrigated with a heparinized saline solution. Residual intimal fragments are removed. Back bleeding and flushing of the renal arteries are performed in the standard fashion. The graft is then anastomosed to the aorta in an end-to-end fashion with a continuous synthetic suture (Fig. 8). After this anastomosis and flushing are completed, a clamp is placed distal to the aortic anastomoses and renal perfusion is thereby restored. Careful assessment of our renal revascularizations was undertaken with the Hoffrel 540A (Hoffrel Instruments Inc., South Norwalk, Conn.) intraoperative probe and duplex scanner. This assessment is undertaken by viewing the aorta and the renal arteries in transverse sections from their origin to beyond the end point of the endarterectomy. This permits de-
Volume 9 Number 2 Februais~ 1989
Modified transaortic endarterectomy
227
Fig. 7. Transection of the lesion just distal to the SMA in sagittal view.
Fig. 6. After the aortorenal endarterectomy, the specimen is transected just distal to the SMA orifice,
tection of vessel wall irregularities, flaps, or persistent lesions. After this, each vessel is viewed in longitudinal section and the audible Doppler signal is noted as are the peak systolic and diastolic flow velocities. If a residual lesion is palpated at the completion of the endarterectomy or is identified by duplex scanning, an occluding clamp is applied to the involved renal artery at its origin. A transverse incision is made just distal to the site of the lesion. A tapered (beveled) end point is created by incising the intima and media beyond the residual lesion with a Beaver blade. The remaining lesion is then removed, working from distal to proximal segment. METHODS OF POSTOPERATIVE F O L L O W - U P We reviewed the records of all patients who underwent combined aortorenal reconstruction by this technique since its introduction at the University of California, San Francisco. Pertinent preoperative data obtained from hospital records were (1) immediate signs and symptoms, (2) pertinent past medical history, (S) electrocardiograms, (4) chest radiographs, and (5) results of angiography. In addition, the extent of coronary artery disease, arterial hypertension (blood pressure > 1 5 0 / 9 0 mm Hg), and renal insufficiency (creatinine level >1.5 mg/dl) were carefully documented. Pcrioperative assessment of surgical results was based on review of the surgical record and postoperative course of these patients. The indications for surgery, the operative procedure, the level of aortic cross-clamping, the number of kidneys revascular-
Fig. 8. After completion of the endarterectomy, the aortic graft is anastomosed to the aorta in a end-to-end fashion by means of continuous synthetic suture. ized, and the renal ischemia time were documented in each patient. The results of intraoperative duplex scanning and postoperative angiography were noted. Defects seen on duplex scanning were categorized as small (luminal defect or < 4 0 % narrowing), moderate (40% to 60% narrowing), or large (>60% narrowing). The postoperative course of the patients was carefully analyzed. All significant intraoperative and postoperative complications were recorded. In addition, the maximal elevation in creatinine level and the creatinine level at the time of discharge were noted in each patient. A change of the creatinine level by more than 0.5 mg/dl was considered significant. Late follow-up results were obtained on all patients if 6 months had elapsed from the time of their operation. Both renal function and the course of the patient's hypertension were noted. Improved management of arterial hypertension was defined as a normal blood pressure while on the same medications or a reduction in the number of medications. When patients showed evidence of decreasing renal function or greater difficulty in control of hypertension, angiography was performed.
Journal of VASCULAR SURGERY
228 Stoney et al.
Table I. Indications for operation Disease
Aortic Aneurysm Atheroscleroticocclusivedisease Aneurysm and atheroscleroticocclusive
Table II. Methods of reconstruction No. of
patients
9 (20%) 18 (41%) 9 (20%)
disease
Infected graft Renal Uncontrolled hypertension Uncontrolled hypertension and elevated creatinine
3 (7%) 2 (5%) 3 (7%)
RESULTS Perioperative assessment o f surgical results Between September 1984 and April 1988, 44 patients underwent treatment of their combined aortic and renal disease by this technique. The 26 men and 18 women had a mean age of 64.8 years (range 42 to 86 years). There was a high incidence of significant coronary artery disease. Twenty-three patients had symptomatic coronary artery disease. Eleven patients were known to have had a myocardial infarction preoperatively. Two of these patients had their infarctions no more than 6 months before undergoing this operation. Three patients had congestive heart failure. Ten patients had prior coronary artery bypass grafting. Thirty-six patients (82%) had arterial hypertension. Thirteen patients (36%) required multidrug therapy (more than three drugs), 15 (42%) patients had poorly controlled arterial hypertension (blood pressure higher than 150/90 rnm H g on multiple medications), and 33 (92%) had hypertension of long-standing duration (more than 5 years). Twentyeight patients (64%) had evidence of decreased renal function, mean creatinine level of 2.10 mg/dl (range 1.3 to 5.0 mg/dl). The primary indication for operation was aortic disease in 39 patients (89%) and poorly controlled hypertension in five patients (11%), three of whom had an elevated serum creatmine level (Table I). Thirty (68%) of the patients referred for the management of aortic disease had hypertension or an elevated creatinine level or both. Of the nine other patients referred for the management of aortic disease with normal blood pressure or untreated hypertension and normal serum creatinine level, five had bilateral renal artery stenoses and four had unilateral renal artery stenosis of a moderate to severe degree (>50% diameter reduction). Three of these patients were less than 50 years old.
Operation
No. of patients
Renal revascularization Bilateral
27 (61%)
Unilateral Solitary (Renal ischemia = mean 30 min; range, 17-81 min) Aortic reconstruction Aortic-aortic Aortoiliac Aortobifemoral Aortic resection Clamp level Supraceliac Supra-SMA Suprarenal
14 (32%) 3 (7%)
11 (25%) 15 (34%) 16 (36%) 2 (5%) 3 (7%) 20 (45%) 21 (48%)
The indications for reconstruction of thc aorta represented the full spectrum of diseases familiar k~. vascular surgeons (Table I). Three patients required aortic reconstruction for the management of graft infection. Two patients required removal of an infected aortic prosthesis. Both had significant renal artery stenoses. Endarterectomy of the pararenal aortic stump was necessary to provide an adequate aortic wall for closure and to avoid thrombosis of the stenotic renal arteries after the aortic stump was oversewn. The third patient had an infection of an axillofemoral graft that was placed initially for the treatment of an infected aortic graft. Thus the aortic stump, which had been previously oversewn, required aortorenal endarterectomy to permit attachment of a new prosthesis to the reopened aorta. Each operation was designed to treat all significant aortic and renal disease present in a given patient (Table II). A total of 75 renal arteries were reva~ cularized in 44 patients. Eleven percent of the arterids revascularized were occluded preoperatively. The mean time of renal ischemia that was required to perform the transaortic renal endarterectomy and anastomosis of an aortic prosthesis was 30 minutes (range 17 to 81 minutes). To effectively treat the combined aortic and renal disease of these patients, supraceliac aortic cross-clamping was necessary in three patients, supramesenteric clamping in 20 patients, and suprarenal clamping in 21 patients. Assessment of the renal artery revascularization was made intraoperatively or immediately postoperatively in 91% of the patients. Intraoperative duplex scanning was undertaken in 37 patients (84%). Thirty-seven defects were identified in 25 patients. Twenty-eight of the 37 defects were small, caused by small pieces of residual intima or platelet aggregates
Volume 9 Number 2 February 1989
along the endarterectomized surface of the renal artery. These defects were of no clinical or hemodynamic significance. Seven patients had moderate defects and two patients had large defects. All moderate and large defects were explored and corrected when technically feasible. Angiography was performed before discharge of 25 patients (57%). These angiograms revealed two small defects, two moderate defects, and one large defect. One postoperative renal occlusion was identified by angiography on the seventh postoperative day. The results of the intraoperative duplex scan were normal, for this patient, but the patient had an occluded renal artery preoperatively. Intraoperative and perioperative complications. Intraoperative technical complications occurred in four patients. In two patients a distal renal artery flap was identified and removed through a trans:rse renal arteriotomy that was dosed primarily. In one patient previously unsuspected occlusive disease extended into small branches of the renal artery. Because both endarterectomy and bypass grafting were not feasible in this one instance, this vessel was ligated. In one patient the aortic anastomosis had to be redone at a more proximal level because the aortic, wall was too thin at the original site of anastomosis. Perioperative mortality. There were two deaths in the postoperative period. The first was that of an 80-year-old man 23 days after he underwent resection of an 8 cm abdominal aortic aneurysm and transaortic bilateral renal endarterectomy- He had a preoperative creatinine level of 3.0 mg dl and poor pulmonary function. After prolonged postoperative ventilatory support pneumonia, sepsis, and multisystem organ failure including anuric renal failure ~quiring dialysis developed in the patient. He died and his autopsy revealed patent renal arteries bilateral. The second death was that of a 76-year-old woman with poorly controlled hypertension and an elevated creatinine level. She died 2 months after undergoing an uncomplicated bilateral renal endarterectomy and aortobifemoral bypass grafting. Symptoms of persistent gastric outlet obstruction led to her hospitalization in her own community where her gastric outlet obstruction was operated on. However, the gastroenterostomy failed. No autopsy was performed after she died. Serious postoperative complications developed in 14 patients (32%) (Table III). Three patients required repeat operation. Two patients were successfially operated on for limb ischemia and one patient with thrombocytopenia underwent repeat operation on the groin for bleeding after removal of an infected
Modified transaortic endarterectomy 229
Table III. Postoperative complications Complication No. ofpatients Repeat operation Pulmonary Respiratory insufficiency Pneumonia Cardiac Myocardial infarction Congestive heart failure Gastointestinal Hyperamylasemia Elevated fiver enzymes Other
14 3 8 3 4 1 3 4 4
axillofemoral graft. No patient required abdominal reexploration. Pulmonary complications occurred in 11 patients. Eight patients had respiratory insufficiency, arbitrarily defined as requiring a ventilator beyond the initial 24 hours postoperatively. Pneumonia developed in three of these patients. Thus it should also be noted that 82% of the patients were extubated successfully within 24 hours of operation. Four patients had myocardial infarction. Two of these patients had negative serial creamine phosphokinase isoenzymes. One patient had congestive heart failure. Four patients had elevated liver enzymes in the immediate postoperative period usually related to supraceliac or supra-SMA aortic cross-clamping. One patient each had phlebitis, atheroembolism to the feet not resulting in tissue loss, peroneal palsy, and a decubitus ulcer. Transient renal dysfunction was common postoperatively. Nineteen patients had an increase in the creatinine level of greater than 0.5 mg/dl. Sixteen of these 19 patients (84%) had preoperative renal insufficiency. At the time of discharge eight of the 19 patients had return of the creatinine level to or below the preoperative baseline tevcl. Two patients required postoperative dialysis. One is the patient previously mentioned to have died. The other patient had an initially uncomplicated course with normal renal function. However, an infection requiring treatment with nephrotoxic antibiotics developed and renal failure ensued. This patient had normal results on a postoperative renal angiogram. No patient required transient dialysis. LATE F O L L O W - U P RESULTS t
Follow-up was achieved in 36 patients for a mean period of 1.7years after operation. Five patients died, all as a result of complications of coronary artery occlusive disease. Data were available to obtain follow-up on renal function in 33 patients. Twenty-
230 Stoney et al.
six patients (79%) had improved (4 patients) or stable (22 patients) renal function as determined by serum creatinine levels. In the subgroup (79%) who had preoperative renal insufficiency, an identical percent had improved or stable sertma creatinine levels. One patient required dialysis just before her death. She had had a normal result on her postoperative renal angiogram and improvement in her renal function postoperatively. Thus the cause of her preterminal dialysis-dependent renal failure is unknown. No patient who had normal creatinine and blood pressure levels preoperatively had renal failure or hypertension after operation. Arterial hypertension was cured or its management improved in 64% of the patients who were hypertensive preoperatively. Late postoperative angiography was performed because of worsening renal function or increased difficulty in controlling arterial hypertension in five patients and as a matter of routine follow-up in two patients. These studies revealed two late renal artery occlusions. One of the two occlusions occurred in a patient who had an occluded renal artery preoperatively. A second patient with recurrent renal artery stenosis was identified and underwent successful percutaneous renal transluminal angioplasty. DISCUSSION We believe that these preliminary results in the management of combined aortorenal atherosclerosis in 44 consecutive patients indicate that renal cndarterectomy through a transected aorta and concomitant aortic grafting provides a safe, efficient, and effective means to manage these complex arterial lesions. This technique permitted successful surgical management of a wide spectrum of aortic and renal occlusive disease. Although this technique allows for the successful management of a wide spectrum of combined aortic and renal lesions, these operations clearly represent a major surgical stress and are associated with significant postoperative morbidity. Forty-three percent of the patients had elevation of their creatininc levels of more than 0.5 mg/dl. This elevation in creatinine level occurred almost exclusively (84%) in patients who had an elevated creatinine level preoperatively. Fortunately approximately half of these elevations returned to or below baseline by the time the patient was discharged. Interestingly the two patients who required dialysis postoperatively had objectively documented patent renal arteries. Their renal dysfunction was largely a result of postoperative complications rather than intraoperative renal ischemia. Another area of major postoperative morbidity
Journal of VASCULAR SURGERY
was a result of coronary artery occlusive disease. In four patients an electrocardiogram showed evidence of postoperative myocardial infarction. The high incidence of preoperative symptomatic coronary artery disease, the 9% incidence ofperioperative myocardial infarction, and the high late-mortality rate as a result of coronary artery disease sharply underline the very advanced state of generalized atherosclerosis experienced by these patients. Intraoperative technical problems associated with the use of this technique were uncommon. The transected endarterectornized aortic stump provided a secure site for proximal anastomosis of the aortic prostheses. One patient required revision of the primary aortic anastomosis. Technical factors that facilitate the safe conduct of this procedure are complete mobilization of the left renal vein, which permits its tension-free manipulation; incision of the dense neural tissue anterolateraUy and of the cru~;~ posterolateraUy around the paramcsenteric aorta; identification and development of the proper endarterectomy plane within the media of the aorta, and careful eversion endarterectomy of the renal arteries. The latter is accomplished by maintaining tension on the aortic plaque, which causes the renal artery to prolapse into the aorta, while gently pushing the renal artery wall away from the lesion with the endarterectomy dissector. Two of the three renal artery occlusions that developed postoperatively (one early and one late) occurred in patients who had occluded renal arteries preoperatively. Thus the use of transaortic renal endarterectomy to open occluded renal arteries should be undertaken with extra care to assure the adequacy of the distal end point. We will continue to follow carefully this important subgroup of patients. hatraoperativc duplex scanning has provided a simple and sensitive technique with which to evaluate the status of the renal arteries after endarterectomy. Most lesions revealed by this technique are minor defects that are the result of residual intima or platelet aggregates. Experience has shown that these rarely require intervention. This is supported by their absence on the postoperative arteriograms. All larger defects are considered for direct examination to allow correction and prevent late stenosis or occlusion of the artery. The major contraindications to the use of this technique of renal endarterectomy are aneurysmal degeneration of the aorta at the level of the renal arteries, or extensive distal renal artery atherosclerosis. Our preliminary follow-up results indicate that
Volume 9 Number2 February 1989
Modified transaortic endarterectomy 231
79% o f the patients had stable or improved renal function compared to their preoperative status. In addition, o f those patients w h o had hypertension preoperatively, 64% o f the patients had improved management o f their arterial hypertension. N o patient w h o underwent prophylactic renal revascularization, suffered loss o f renal function or the development o f hypertension postoperatively. These nine patients were younger (three of nine patients were less than 50 years o f age) and were at less risk than the remainder o f the patients w h o had multiple clinical manifestations o f advanced atherosclerosis. Included in this group o f nine are the three patients who had graft infections. We believe aggressive surgical treatment o f the combined aortic and renal disease in these patients is justified in view o f the very poor natural history o f these patients 1H4 if untreated. In one series, once ~_ 2enal artery stenosis became significant, there was a 44% rate o f stenosis progression and a 15% rate o f occlusion over an observation period o f 3 to 72 months. 1~ This progression was accompanied by the development o f azotemia in 50% o f these patients. Similar findings have been found in other studies. 1°-t4 A consistent finding o f these studies ~°,n was that the progression o f the renal artery stenosis and the elevation o f serum creatinine level were not heraided by a loss o f control o f the patient's arterial hypertension by drug therapy. Based on this experience in the treatment o f these 44 consecutive patients the following recommendations appear justified: renal endarterectomy through the transected aorta and concomitant aortic grafting provides a safe, efficient, and effective means for the management o f combined aortic and renal athero~'qerosis. Because o f the p o o r prognosis o f these patients when treated nonoperatively, aortorenal revascularization should be undertaken promptly after careful preoperative evaluation and optimization o f coexisting myocardial or other critical organ function. The precise role o f this technique in the treatment o f patients with combined aortic and renal ath-
erosclerosis will ultimately be determined by its long-term success in correcting the complex aortic lesions, controlling hypertension, and preserving renal function.
DISCUSSION
endarterectomized renal arteries is imperative. In this area there is really no margin for error. Dr. Stoney says that lifting the aorta gives enough exposure with this technique so that the plaque end point can be seen. Another concern is the potential time of damping and embolization. We have approached this differently in that we clamp off the aorta below the renal arteries and corn-
Dr. J. Harold Harrison (Atlanta, Ga.). We have been doing transaortic endarterectomy of the renal arteries through the anterior approach for 30 years. We have transected the aorta for complete occlusion extending to the level of the renal arteries, but we have not used this tech,nique. Visualization of the endpoint of the plaque in the
REFERENCES
1. Wylie EJ, Perloff DL, Stoney RJ. Autogenous tissue revascularization techniques in surgery for renovascularhypertension. Ann Surg 1969;170:416-28. 2. Ernst CB, Stanley JC, Marshall FF~ Fry WJ. Renal revascularization for arteriosclerotic renovascular hypertension: prognostic implications for focal renal arterial vs. overt generalized arteriosclerosis. Surgery 1973;73:859-67. 3. Libertino JA, Thierry AF, Zinman LN, et al. Changing concepts in surgical management of renovascular hypertension. Arch Intern Med 1988;I48:357-9. 4. NovickAC, Ziegelbaum M, Vidt DG, Gifford RW Jr, Pohl MA, GoormaasticM. Trends in surgical revascularizationfor renal artery disease--Ten year's experience. JAMA 1987; 257:498-501. 5, Tarazi RY, Hertzer NR, Beven EG, O'Hara PJ, Anton GE, Krajewski LP. Simultaneous aortic reconstruction and renal revascularizafion: risk factors and late results in 89 patients. 1 VASCSURG1987;5:707-14. 6, StewartMT, Smith RB Ill, Fulenwider IT, Perdue GD, Wells JO. Concomitant renal revascularization in patients undergoing aortic surgery. J VASCSUinG1985;2:400-5. 7. Brewster DC, Buth 1, Darling RC, Austen WG. Combined aortic and renal artery reconstruction. Am J Surg 1976; 131:457-63. 8. Shahian DM, Najafi H, Iavid H, Hunter JA, Goldin MD, Monson DO. Simultaneous aortic and renal artery reconstruction. Arch Surg 1980;115:1491-7. 9. Sterpetti AV, Schultz RD, Fddhaus RJ, Peetz DJ. Aortic and renal atherosderotic disease. Surg Gynecol Obstet 1986; 163:54-9. 10. Dean RH, Kieffer RW, Smith BM, et al. Renovascular hypertension: anatomic and renal function changes during a drug therapy. Arch Surg 1981;116:1408-15. 11. Pohl MA, Novick AC. Natural history-of atherosclerotic and fibrous renal artery disease: clinicalimplications. Am J Kidney Dis 1985:A120-30. 12. Schreiber MJ, Pohl MA, Novick AC. The natural history of atherosderotic and fibrous renal artery disease. Urol Clin North Am 1984;11:383-92. 13. WollenweberJ, Sheps SG, Davis GD. Clinical course of ath~ erosderotic renovascular disease. Am I Cardiol 1968;21:6071. 14. Meaney TF, Dustan HP, McCormack LI- Natural history of renal arterial disease. Radiology 1968;91:881-7.
232
Stoney et al.
plete any distal surgery. We then expose the proximal aorta and renal and mesenteric vessels and use an anterior approach to the renal arteries. If at this point we find a widened or sickly looking aorta, and particularly if we find "cheese" when we cut across the aorta, we will do bypass grafting. Dr. Messina, what do you do when you have completed an extensive dissection of the vessels, you have placed an occluding clamp above the renal and/or ruesenteric arteries, you have not done your surgery below, and you run into this extensive disease? In the article the exposure is described as midline or a high transverse. To me a high bilateral subcostal incision offers the best exposure of vessels in this area. Dr. William J, Fry (Dallas, Texas). I express my appreciation to the authors for allowing me to review the manuscript and compliment them on the low mortality in this group of patients at very high risk. We have had a great deal of difficulty treating our group of patients who have atherosclerosis of the renal arteries. In a series ofrnore than 125 cases that we studied recently, all deaths (a 5% mortality) have occurred when we have tried to do combined operations. Therefore we have tried to avoid this whenever possible, and our mortality rate with staged operation so far has been zero. We think that the severe cardiomyopathy in these patients makes them a poor risk for a prolonged procedure. Nevertheless we are impressed with the outstanding results that Dr. Messina and his group have brought to us today. I have several questions. It is not quite clear in the article exactly what size defects revealed by ultrasound scanning postoperatively will make you open the renal artery and operate and which ones you will let go. This is important to us, since the operative ultrasound is a very sensitive instrument and sometimes it is a bit confusing as to exactly what is significant and what is not. Your indications for operations are not entirely clear to me. Which ones require endarterectomy and which ones do not? I am not aware of any data that have shown progression of atherosclerotic lesion in the renal artery in the absence of hypertension; therefore I ask you why you did approximately 20% of these operations prophylactically. I echo what Dr. Harrison said; in the patient with an aneurysm it is difficult to do endarterectomies. We tend to avoid trying to do endarterectomies in aneurysmal disease and would rather do a bypass. Is there a particular plane to get in that would give more Strength and a better division at the end of the specimen? Dr. Richard H. Dean (Winston-Salem, N.C.). I thank the authors for giving me a preliminary copy of their manuscript and compliment Dr. Messina on his excellent presentation. I would like to make a couple of points and ask a couple of questions. Although I disagree with the use of transaortic endarterectomy, I recognize that under the able hands of the San Francisco group, the results with the transaortic endarterectomy have been quite good and are comparable to results with other techniques. However, I continue to be concerned about the adaptability of this technique to the range of expertise of those dealing with
lournal of VASCULAR SURGERY
renal vascular surgery in the country as a whole, and I am also concerned that the visualization of the end point by the transaortic route lasts for only about half a second. I have difficulty in believing that visualization is adequate during that very brief moment. One can only infer the condition of the transection point by examination of the endarterectomy specimen. The reasonI bring this up is the 28 small defects the authors identified during intraoperative post_reconstruction duplex scanning. With the transverse scanning technique we have commonly seen loose pieces of intima at the distal transection point, and these have been relatively simple to pull away. I continue to be concerned about what would happen if I left those alone. If I correctly interpret the 28 small defects as probably being similar to those that we have seen and removed, perhaps there is no difficulty if they are left in place. I question the indications for using the techniques reported. Specifically, as best as I can tell from reading the manuscript, about 20% of the patients had no clear-ct~ indication for revascularization or transaortic endarterec~ tomy other than the fact that they had aortic disease and renal artery stenoses. My question to Dr. Messina in that regard is whether the San Francisco group have changed their indications for renal revascularization so that it is now done prophylactically, or do those patients have peculiar stenoses or are they patients in whom the context of therapeutic intervention had to be modified for some specific reason? Clearly, if their suggestion is that there are now data showing that all nonhemodynamic lesions seen angiographically should be treated with endarterectomy or bypass, I am not familiar with those data, and I therefore continue to believe, alternatively, that we should not be operating on such patients prophylactically. In that regard, I have one further question that relates to the results. Eighty percent of the patients' renal function was either improved or stable. Does that imply that 20% of the patients had worsening of their renal function as consequence of the procedure? Similarly, as relates to h~~ pertension where 13% of the patients had a worsening of their hypertension, do these match up with the technical issues identified by the duplex scanner or in early followup studies showing residual or recurrent lesions? Again, I would compliment the authors and their surgical colleagues on the technique reported and I appreciate the opportunity to discuss it. Dr. Messlna (closing). I thank all the discussants for their comments. I will address Dr. Harrison's concern about the applicability of this technique to various forms of combined aortorenal atherosclerosis and then describe some of the techniques that can be used to avoid complications associated with aortorenal endarterectomy. We believe that this technique is well suited to most forms of combined aortorenal atherosclerosis. Typically these renal lesions are bilateral and proximal extensions of aortic atheromas. The major contraindications to the application of this procedure are the presence of a pararenal aneurysm or
Volume 9 Number 2 February 1989
a degenerated or patulous aorta, or the occurrence of distal renal artery disease. I would emphasize a number of points to avoid technical complications. The completeness of the initial dissection is critical to the successful conduct of this operation. It is particularly important to obtain proper exposure of the aorta in the area of superior mesenteric and renal arteries. This usually entails incision o f the netlral tissue around the aorta at the level of the S/vIA as well as incision of the crura of the diaphragm bilaterally. The left renal vein should be fully mobilized, which usually requires transection and ligation o f its nonrenal branches. Transection of the aorta just beyond the level of the renal arteries has provided more than an adequate view of the aorta and allows one to see directly down into the orifices of the renal artery during the renal endarterectomy. To avoid embolization we emphasize careful dissection, avoiding unnecessary manipulation of the aorta, and clamping distally before applying a proximal clamp. In addition, we emphasize thorough irrigation of the aorta a~lcl renal arteries and back bleeding and flushing of the vessels before restoring flow, all to help minimize the occurrence of atheroembolization. Dr. Fry raised the concern about doing endarterectomies on patients with aneurysmal disease. If the endarterectomy plane is too deep into the media or the aorta is abnormal, a "blow out" of the aorta can occur. As previously mentioned, the main contraindication to the application of this procedure is the presence of a pararenal aneurysm or a degenerated, patulous aorta. Both Dr. Fry and Dr. Dean raised questions about the intraoperative duplex scan results. This is an evolving technique. Twenty-eight small defects were seen in 37 patients. However, most patients had postoperative angiography, and only two small defects were seen on these studies. I believe the differences between these two studies indicate how sensitive intraoperative duplex scanning is. Scanning identifies tiny intimal fragments or platelet aggregates that ar~enot seen on angiography. All moderate and large defects al~ opened when appropriate. The most common indication for opening the artery is an inadequate end point; the artery is transected distally and a good end point obtained.
Modified transaortic endarterectomy 233
Both Dr. Dean and Dr. Fry wanted more information concerning the 20% of patients who had a "prophylactic operation." That represented nine of the 44 patients. Three of the nine were less than 50 years old, arid the indication for renal revascularization was the presence of either bilateral renal artery stenoses or a high-grade unilateral stenosis. Another group of three patients required aortic and renal endarterectomy during the course of the management of infected prosthetic bypass grafts. In two, the aortic stump required endarterectomy to ensure against late occlusion of the renal arteries because of the obligatory infrarenal aortic occlusion necessary to manage their graft infection. In the third patient the endarterectomy was performed to reopen the pararenal aorta to provide adequate cuff for anastomosis o f a new prosthetic graft. An infection of an axillofemoral graft had developed in this patient. Dr. Fry is correct; there are not good natural history data on the management of renal artery stenoses in the absence of clinically significant hypertension. The only studies published, which include an excellent one by Dr. Dean, refer to patients who are being evaluated for hypertension. However, in these studies it is clear that when a renal artery stenosis becomes high-grade these lesions will show progression in 60% of patients within 2 years. In these patients there was up to a 54% incidence of deterioration of renal function not heralded by loss of blood pressure control. Most of our patients had serious, clinically identifiable complications of their renal artery disease. Sixty-four percent had elevated creatinine levels and 82% had hypertension. These patients had good responses of their hypertension management postoperatively. The patients who were operated on in the "prophylactic group" represented patients with unusual, atypical features that required selective judgment. We believe that young patients who are good operative risks and have hemodynamically significant stenosis should undergo prophylactic endarterectomy to prevent loss of renal function. However, this is predicated on close monitoring of one's own results so that the benefits outweigh the risk of this aggressive management policy. None o f the patients operated on prophylactically had deterioration of renal fimction or hypertension postoperatively.
The first successful renal revascularization for renal artery stenosis was performed by Edwin J. Wylie on Nov. 10, 1952, although not reported at that time? T h e patient was a 46-year-old woman who had severe hypertension, a left renal artery stenosis, and an infrarenal aortic occlusion. This combination o f renal artery and aortic disease was cured by an aortoiliac and renal endarterectomy. Although this first renal revascularization was performed as part o f From the Section of Vascular Surgery, Department of Surgery, University of California. Supported in part by the PacificVascular Research Foundation, San Francisco, Calif. Presented at the Thirty-sixth Scientific Meeting of the Nort_h American Chapter, International Society for Cardiovascular Surgery, Chicago, Ill., June 14-15, 1988. Reprint requests: Louis M. Messina, MD, Section of Vascular Surgery,Universityof Michigan Medical Center, 2210 L Tanbman Health Care Center, 1500 E. Medical Center Dr., Ann Arbor, MI 48109-0329. 224
a combined aortic and renal reconstruction, over the ensuing 3 decades the management o f focal or isq~ lated renal artery disease has been the most su& ~ cessfu[. 2 However, at our institution and at other centers 3,4 that perform renal artery reconstructions regularly, a distinct change in the profile o f patients requiting renal revascularization has been noted. Elderly patients who h a v e generalized atherosclerosis now have combined symptomatic aortic and renal disease, often complicated by severe hypertension and azotemia. The renal occlusive lesions are typically bilateral (Fig. 1). They exhibit clinical features that have been repeatedly associated with increased mottalky compared to the management o f isolated renal or aortic disease. These risk factors include the presence o f bilateral renal artery stenoses, 5 impaired renal fianction, 6 and the presence o f aortic aneurysmal disease. 7-9Untreated, these lesions o f the aorta and renal arteries carry an ominous clinical prognosis? ° The
Volume 9 Number 2 February 1989
Modified transaortic endarterectomy 225
Fig. 2. Lateral view of aorta after dissection shows celiac artery (ca), superior mesenteric artery (sma), right renal artery (rra), and left renal artery (Ira). Fig. 1. Incision of crura of the diaphragm to facilitate posterolateral mobilization of the aorta. optimal surgical management of these complex, highrisk patients has not been defined. Transaortic renal endarterectomy through a longitudinal aortotomy has proved to be effective and durable in the treatment of patients with renal artery occlusive disease. 1 However, when concurrent aortic grafting is necessary, transaortic endarterectomy through a longitudinal aortotomy necessitates intersection of the suture lines of the aortotomy and graft anastomosis. Under these circumstances, reimplantation of the renal ~arteries or multiple renal artery grafting has been advocated. The increasing number of patients requiring combined aortic and renal reconstruction and our continuing satisfactory experience with transaortic renal endarterectomy stimulated the development of a modified technique for ~,ansaortic renal endarterectomy through the transected infrarcnal aorta preceding proximal aortic graft anastomosis. The purpose of this article is to describe this modified technique of transaortic endarterectomy and to report the clinical course and early follow-up of the first 44 consecutive patients treated by this method.
OPERATIVE TECHNIQUE Transaortic renal endarterectomy through the transected abdominal aorta permits one to successfully manage the typical features of symptomatic combined aortic and renal artery atherosclerotic disease. The renal artery lesions are frequently proximal extensions of aortic atherosclerosis typically involving both or multiple renal arteries. This technique permits satisfactory disobliteration of single or multiple renal arteries and concomitant aortic grafting,
requiring only a single suture line and a short, usually well-tolerated period of obligatory renal ischemia. Either a long midline incision or a generous transverse incision is made depending on the patient's body habitus. The aorta may be exposed by a transperitoneal incision. Alternatively, the upper abdominal aorta may be approached by incision of the left lateral peritoneal reflection and medial rotation of the abdominal viscera. Careful and precise dissection of the upper abdominal aorta and its visceral branches is critical to the success of this operation. Dissection of the pararenal aorta commences with complete mobilization of the left renal vein from the confluence of its renal branches to the vena cava. The posterior tributary of the renal vein to the ascending lumbar vein is routinely transected and ligated. Similarly the adrenal and gonadal veins are sacrificed as necessary to permit tension-free manipulation of the renal vein. The aorta at and above the level of the superior mesenteric artery (SMA) is encased in a dense layer of neural tissue. This neural tissue should be incised to expose the anterior and lateral surfaces of the aorta. The lateral and posterior surfaces of the aorta at this same level are encircled to a variable extent by the crura of the diaphragm. Sharp division of the crura greatly facilitates further mobilization of the posterior aorta (Figs. 1 and 2), thus permitting occlusion of the aorta by the vertical placement of an aortic clamp. MobiJ,ization of the renal arteries should be extended beyond the palpable extent of the renal atherosclerosis. After systemic heparinization, clamps are first applied individually to the superior mesenteric and renal arteries, and finally to the aorta proximal and distal to the site of proposed aortic reconstruction and to any intervening intercostal or lumbar arteries.
226
Journal of VASCULAR SURGERY
Stoney et al.
Fig. 3. Left, Vertical placement of an aortic clamp proximal to SMA. Aorta is transected just below renal arteries. Right, Aortic endarterectomy commences by developing the plane within the media on anterior ao~ic wall.
Fig. 4. Left, After dissection of the aortic plaque circttmferentially. Right, LRA endarterectomy performed by maintaining gentle traction on aortic atheroma; this permits prolapse of renal artery into the aorta. This prolapsing of the renal artery into the aorta is facilitated fmTJaer by pushing the damp occluding the renal artery toward the aorta.
The proximal aortic clamp should be placed above any significant aortic atherosclerotic lesion and at a level that will permit an uncompromised aortic and renal endarterectomy (Fig. 3). Most frequently this is above the level of the SMA. The aorta is then transected immediately below the level of the renal arteries. The distal 3 cm of the aorta down to the level of the inferior mesenteric artery is resected to provide space for an unrestricted end-to-end anastomoses of the prosthetic graft to the aorta after endarterectomy. Transaortic endarterectomy of the proximal segment is facilitated by gently rotating the aortic clamp to evert the aortic stump anteriorly and thus permit a direct view of the aortic lumen (Fig. 3). An endarterectomy dissector is used to develop and circumferentially extend the plane between the aortic plaque and the media. This dissection is continued proxi-
Fig. 5. Left, Endarterectomy plane is extended by using dissector to push renal artery wall away from the lesion. Right, With continued traction on the aortic lesion a well-defined end point is feathered and the lesion breaks away.
mally to a level immediately distal to the S/vIA orifice. It is occasionally helpful to longitudinally split r3~ aortic atherosclerotic lesion. This maneuver allows separate transaortic renalendarterectomy without the still-attached contralateral renal atheroma tethering the spccimcn and limiting mobility. This endarterectomy plane is extended in continuity into each of the renal arteries. Thc renal cndartercctomy is performed by maintaining gentle traction on the aortic lesion; this prolapses the renal artery into the aortic lumen (Figs. 4 and 5). Separation of the prolapsing renal artery media from the emerging core of atherosclerotic intima of the renal artery is achieved by use of the dissector blade to gently push the residual arterial wall away from the lesion. With continued traction on the specimen, a well-defined end point is usually obtained and the specimen separates easily. After each individual renal lesion has been removed, the plaque ~. transected just distal to the SMA orifice (Figs. 6 an8 7). The aorta and renal arteries are irrigated with a heparinized saline solution. Residual intimal fragments are removed. Back bleeding and flushing of the renal arteries are performed in the standard fashion. The graft is then anastomosed to the aorta in an end-to-end fashion with a continuous synthetic suture (Fig. 8). After this anastomosis and flushing are completed, a clamp is placed distal to the aortic anastomoses and renal perfusion is thereby restored. Careful assessment of our renal revascularizations was undertaken with the Hoffrel 540A (Hoffrel Instruments Inc., South Norwalk, Conn.) intraoperative probe and duplex scanner. This assessment is undertaken by viewing the aorta and the renal arteries in transverse sections from their origin to beyond the end point of the endarterectomy. This permits de-
Volume 9 Number 2 Februais~ 1989
Modified transaortic endarterectomy
227
Fig. 7. Transection of the lesion just distal to the SMA in sagittal view.
Fig. 6. After the aortorenal endarterectomy, the specimen is transected just distal to the SMA orifice,
tection of vessel wall irregularities, flaps, or persistent lesions. After this, each vessel is viewed in longitudinal section and the audible Doppler signal is noted as are the peak systolic and diastolic flow velocities. If a residual lesion is palpated at the completion of the endarterectomy or is identified by duplex scanning, an occluding clamp is applied to the involved renal artery at its origin. A transverse incision is made just distal to the site of the lesion. A tapered (beveled) end point is created by incising the intima and media beyond the residual lesion with a Beaver blade. The remaining lesion is then removed, working from distal to proximal segment. METHODS OF POSTOPERATIVE F O L L O W - U P We reviewed the records of all patients who underwent combined aortorenal reconstruction by this technique since its introduction at the University of California, San Francisco. Pertinent preoperative data obtained from hospital records were (1) immediate signs and symptoms, (2) pertinent past medical history, (S) electrocardiograms, (4) chest radiographs, and (5) results of angiography. In addition, the extent of coronary artery disease, arterial hypertension (blood pressure > 1 5 0 / 9 0 mm Hg), and renal insufficiency (creatinine level >1.5 mg/dl) were carefully documented. Pcrioperative assessment of surgical results was based on review of the surgical record and postoperative course of these patients. The indications for surgery, the operative procedure, the level of aortic cross-clamping, the number of kidneys revascular-
Fig. 8. After completion of the endarterectomy, the aortic graft is anastomosed to the aorta in a end-to-end fashion by means of continuous synthetic suture. ized, and the renal ischemia time were documented in each patient. The results of intraoperative duplex scanning and postoperative angiography were noted. Defects seen on duplex scanning were categorized as small (luminal defect or < 4 0 % narrowing), moderate (40% to 60% narrowing), or large (>60% narrowing). The postoperative course of the patients was carefully analyzed. All significant intraoperative and postoperative complications were recorded. In addition, the maximal elevation in creatinine level and the creatinine level at the time of discharge were noted in each patient. A change of the creatinine level by more than 0.5 mg/dl was considered significant. Late follow-up results were obtained on all patients if 6 months had elapsed from the time of their operation. Both renal function and the course of the patient's hypertension were noted. Improved management of arterial hypertension was defined as a normal blood pressure while on the same medications or a reduction in the number of medications. When patients showed evidence of decreasing renal function or greater difficulty in control of hypertension, angiography was performed.
Journal of VASCULAR SURGERY
228 Stoney et al.
Table I. Indications for operation Disease
Aortic Aneurysm Atheroscleroticocclusivedisease Aneurysm and atheroscleroticocclusive
Table II. Methods of reconstruction No. of
patients
9 (20%) 18 (41%) 9 (20%)
disease
Infected graft Renal Uncontrolled hypertension Uncontrolled hypertension and elevated creatinine
3 (7%) 2 (5%) 3 (7%)
RESULTS Perioperative assessment o f surgical results Between September 1984 and April 1988, 44 patients underwent treatment of their combined aortic and renal disease by this technique. The 26 men and 18 women had a mean age of 64.8 years (range 42 to 86 years). There was a high incidence of significant coronary artery disease. Twenty-three patients had symptomatic coronary artery disease. Eleven patients were known to have had a myocardial infarction preoperatively. Two of these patients had their infarctions no more than 6 months before undergoing this operation. Three patients had congestive heart failure. Ten patients had prior coronary artery bypass grafting. Thirty-six patients (82%) had arterial hypertension. Thirteen patients (36%) required multidrug therapy (more than three drugs), 15 (42%) patients had poorly controlled arterial hypertension (blood pressure higher than 150/90 rnm H g on multiple medications), and 33 (92%) had hypertension of long-standing duration (more than 5 years). Twentyeight patients (64%) had evidence of decreased renal function, mean creatinine level of 2.10 mg/dl (range 1.3 to 5.0 mg/dl). The primary indication for operation was aortic disease in 39 patients (89%) and poorly controlled hypertension in five patients (11%), three of whom had an elevated serum creatmine level (Table I). Thirty (68%) of the patients referred for the management of aortic disease had hypertension or an elevated creatinine level or both. Of the nine other patients referred for the management of aortic disease with normal blood pressure or untreated hypertension and normal serum creatinine level, five had bilateral renal artery stenoses and four had unilateral renal artery stenosis of a moderate to severe degree (>50% diameter reduction). Three of these patients were less than 50 years old.
Operation
No. of patients
Renal revascularization Bilateral
27 (61%)
Unilateral Solitary (Renal ischemia = mean 30 min; range, 17-81 min) Aortic reconstruction Aortic-aortic Aortoiliac Aortobifemoral Aortic resection Clamp level Supraceliac Supra-SMA Suprarenal
14 (32%) 3 (7%)
11 (25%) 15 (34%) 16 (36%) 2 (5%) 3 (7%) 20 (45%) 21 (48%)
The indications for reconstruction of thc aorta represented the full spectrum of diseases familiar k~. vascular surgeons (Table I). Three patients required aortic reconstruction for the management of graft infection. Two patients required removal of an infected aortic prosthesis. Both had significant renal artery stenoses. Endarterectomy of the pararenal aortic stump was necessary to provide an adequate aortic wall for closure and to avoid thrombosis of the stenotic renal arteries after the aortic stump was oversewn. The third patient had an infection of an axillofemoral graft that was placed initially for the treatment of an infected aortic graft. Thus the aortic stump, which had been previously oversewn, required aortorenal endarterectomy to permit attachment of a new prosthesis to the reopened aorta. Each operation was designed to treat all significant aortic and renal disease present in a given patient (Table II). A total of 75 renal arteries were reva~ cularized in 44 patients. Eleven percent of the arterids revascularized were occluded preoperatively. The mean time of renal ischemia that was required to perform the transaortic renal endarterectomy and anastomosis of an aortic prosthesis was 30 minutes (range 17 to 81 minutes). To effectively treat the combined aortic and renal disease of these patients, supraceliac aortic cross-clamping was necessary in three patients, supramesenteric clamping in 20 patients, and suprarenal clamping in 21 patients. Assessment of the renal artery revascularization was made intraoperatively or immediately postoperatively in 91% of the patients. Intraoperative duplex scanning was undertaken in 37 patients (84%). Thirty-seven defects were identified in 25 patients. Twenty-eight of the 37 defects were small, caused by small pieces of residual intima or platelet aggregates
Volume 9 Number 2 February 1989
along the endarterectomized surface of the renal artery. These defects were of no clinical or hemodynamic significance. Seven patients had moderate defects and two patients had large defects. All moderate and large defects were explored and corrected when technically feasible. Angiography was performed before discharge of 25 patients (57%). These angiograms revealed two small defects, two moderate defects, and one large defect. One postoperative renal occlusion was identified by angiography on the seventh postoperative day. The results of the intraoperative duplex scan were normal, for this patient, but the patient had an occluded renal artery preoperatively. Intraoperative and perioperative complications. Intraoperative technical complications occurred in four patients. In two patients a distal renal artery flap was identified and removed through a trans:rse renal arteriotomy that was dosed primarily. In one patient previously unsuspected occlusive disease extended into small branches of the renal artery. Because both endarterectomy and bypass grafting were not feasible in this one instance, this vessel was ligated. In one patient the aortic anastomosis had to be redone at a more proximal level because the aortic, wall was too thin at the original site of anastomosis. Perioperative mortality. There were two deaths in the postoperative period. The first was that of an 80-year-old man 23 days after he underwent resection of an 8 cm abdominal aortic aneurysm and transaortic bilateral renal endarterectomy- He had a preoperative creatinine level of 3.0 mg dl and poor pulmonary function. After prolonged postoperative ventilatory support pneumonia, sepsis, and multisystem organ failure including anuric renal failure ~quiring dialysis developed in the patient. He died and his autopsy revealed patent renal arteries bilateral. The second death was that of a 76-year-old woman with poorly controlled hypertension and an elevated creatinine level. She died 2 months after undergoing an uncomplicated bilateral renal endarterectomy and aortobifemoral bypass grafting. Symptoms of persistent gastric outlet obstruction led to her hospitalization in her own community where her gastric outlet obstruction was operated on. However, the gastroenterostomy failed. No autopsy was performed after she died. Serious postoperative complications developed in 14 patients (32%) (Table III). Three patients required repeat operation. Two patients were successfially operated on for limb ischemia and one patient with thrombocytopenia underwent repeat operation on the groin for bleeding after removal of an infected
Modified transaortic endarterectomy 229
Table III. Postoperative complications Complication No. ofpatients Repeat operation Pulmonary Respiratory insufficiency Pneumonia Cardiac Myocardial infarction Congestive heart failure Gastointestinal Hyperamylasemia Elevated fiver enzymes Other
14 3 8 3 4 1 3 4 4
axillofemoral graft. No patient required abdominal reexploration. Pulmonary complications occurred in 11 patients. Eight patients had respiratory insufficiency, arbitrarily defined as requiring a ventilator beyond the initial 24 hours postoperatively. Pneumonia developed in three of these patients. Thus it should also be noted that 82% of the patients were extubated successfully within 24 hours of operation. Four patients had myocardial infarction. Two of these patients had negative serial creamine phosphokinase isoenzymes. One patient had congestive heart failure. Four patients had elevated liver enzymes in the immediate postoperative period usually related to supraceliac or supra-SMA aortic cross-clamping. One patient each had phlebitis, atheroembolism to the feet not resulting in tissue loss, peroneal palsy, and a decubitus ulcer. Transient renal dysfunction was common postoperatively. Nineteen patients had an increase in the creatinine level of greater than 0.5 mg/dl. Sixteen of these 19 patients (84%) had preoperative renal insufficiency. At the time of discharge eight of the 19 patients had return of the creatinine level to or below the preoperative baseline tevcl. Two patients required postoperative dialysis. One is the patient previously mentioned to have died. The other patient had an initially uncomplicated course with normal renal function. However, an infection requiring treatment with nephrotoxic antibiotics developed and renal failure ensued. This patient had normal results on a postoperative renal angiogram. No patient required transient dialysis. LATE F O L L O W - U P RESULTS t
Follow-up was achieved in 36 patients for a mean period of 1.7years after operation. Five patients died, all as a result of complications of coronary artery occlusive disease. Data were available to obtain follow-up on renal function in 33 patients. Twenty-
230 Stoney et al.
six patients (79%) had improved (4 patients) or stable (22 patients) renal function as determined by serum creatinine levels. In the subgroup (79%) who had preoperative renal insufficiency, an identical percent had improved or stable sertma creatinine levels. One patient required dialysis just before her death. She had had a normal result on her postoperative renal angiogram and improvement in her renal function postoperatively. Thus the cause of her preterminal dialysis-dependent renal failure is unknown. No patient who had normal creatinine and blood pressure levels preoperatively had renal failure or hypertension after operation. Arterial hypertension was cured or its management improved in 64% of the patients who were hypertensive preoperatively. Late postoperative angiography was performed because of worsening renal function or increased difficulty in controlling arterial hypertension in five patients and as a matter of routine follow-up in two patients. These studies revealed two late renal artery occlusions. One of the two occlusions occurred in a patient who had an occluded renal artery preoperatively. A second patient with recurrent renal artery stenosis was identified and underwent successful percutaneous renal transluminal angioplasty. DISCUSSION We believe that these preliminary results in the management of combined aortorenal atherosclerosis in 44 consecutive patients indicate that renal cndarterectomy through a transected aorta and concomitant aortic grafting provides a safe, efficient, and effective means to manage these complex arterial lesions. This technique permitted successful surgical management of a wide spectrum of aortic and renal occlusive disease. Although this technique allows for the successful management of a wide spectrum of combined aortic and renal lesions, these operations clearly represent a major surgical stress and are associated with significant postoperative morbidity. Forty-three percent of the patients had elevation of their creatininc levels of more than 0.5 mg/dl. This elevation in creatinine level occurred almost exclusively (84%) in patients who had an elevated creatinine level preoperatively. Fortunately approximately half of these elevations returned to or below baseline by the time the patient was discharged. Interestingly the two patients who required dialysis postoperatively had objectively documented patent renal arteries. Their renal dysfunction was largely a result of postoperative complications rather than intraoperative renal ischemia. Another area of major postoperative morbidity
Journal of VASCULAR SURGERY
was a result of coronary artery occlusive disease. In four patients an electrocardiogram showed evidence of postoperative myocardial infarction. The high incidence of preoperative symptomatic coronary artery disease, the 9% incidence ofperioperative myocardial infarction, and the high late-mortality rate as a result of coronary artery disease sharply underline the very advanced state of generalized atherosclerosis experienced by these patients. Intraoperative technical problems associated with the use of this technique were uncommon. The transected endarterectornized aortic stump provided a secure site for proximal anastomosis of the aortic prostheses. One patient required revision of the primary aortic anastomosis. Technical factors that facilitate the safe conduct of this procedure are complete mobilization of the left renal vein, which permits its tension-free manipulation; incision of the dense neural tissue anterolateraUy and of the cru~;~ posterolateraUy around the paramcsenteric aorta; identification and development of the proper endarterectomy plane within the media of the aorta, and careful eversion endarterectomy of the renal arteries. The latter is accomplished by maintaining tension on the aortic plaque, which causes the renal artery to prolapse into the aorta, while gently pushing the renal artery wall away from the lesion with the endarterectomy dissector. Two of the three renal artery occlusions that developed postoperatively (one early and one late) occurred in patients who had occluded renal arteries preoperatively. Thus the use of transaortic renal endarterectomy to open occluded renal arteries should be undertaken with extra care to assure the adequacy of the distal end point. We will continue to follow carefully this important subgroup of patients. hatraoperativc duplex scanning has provided a simple and sensitive technique with which to evaluate the status of the renal arteries after endarterectomy. Most lesions revealed by this technique are minor defects that are the result of residual intima or platelet aggregates. Experience has shown that these rarely require intervention. This is supported by their absence on the postoperative arteriograms. All larger defects are considered for direct examination to allow correction and prevent late stenosis or occlusion of the artery. The major contraindications to the use of this technique of renal endarterectomy are aneurysmal degeneration of the aorta at the level of the renal arteries, or extensive distal renal artery atherosclerosis. Our preliminary follow-up results indicate that
Volume 9 Number2 February 1989
Modified transaortic endarterectomy 231
79% o f the patients had stable or improved renal function compared to their preoperative status. In addition, o f those patients w h o had hypertension preoperatively, 64% o f the patients had improved management o f their arterial hypertension. N o patient w h o underwent prophylactic renal revascularization, suffered loss o f renal function or the development o f hypertension postoperatively. These nine patients were younger (three of nine patients were less than 50 years o f age) and were at less risk than the remainder o f the patients w h o had multiple clinical manifestations o f advanced atherosclerosis. Included in this group o f nine are the three patients who had graft infections. We believe aggressive surgical treatment o f the combined aortic and renal disease in these patients is justified in view o f the very poor natural history o f these patients 1H4 if untreated. In one series, once ~_ 2enal artery stenosis became significant, there was a 44% rate o f stenosis progression and a 15% rate o f occlusion over an observation period o f 3 to 72 months. 1~ This progression was accompanied by the development o f azotemia in 50% o f these patients. Similar findings have been found in other studies. 1°-t4 A consistent finding o f these studies ~°,n was that the progression o f the renal artery stenosis and the elevation o f serum creatinine level were not heraided by a loss o f control o f the patient's arterial hypertension by drug therapy. Based on this experience in the treatment o f these 44 consecutive patients the following recommendations appear justified: renal endarterectomy through the transected aorta and concomitant aortic grafting provides a safe, efficient, and effective means for the management o f combined aortic and renal athero~'qerosis. Because o f the p o o r prognosis o f these patients when treated nonoperatively, aortorenal revascularization should be undertaken promptly after careful preoperative evaluation and optimization o f coexisting myocardial or other critical organ function. The precise role o f this technique in the treatment o f patients with combined aortic and renal ath-
erosclerosis will ultimately be determined by its long-term success in correcting the complex aortic lesions, controlling hypertension, and preserving renal function.
DISCUSSION
endarterectomized renal arteries is imperative. In this area there is really no margin for error. Dr. Stoney says that lifting the aorta gives enough exposure with this technique so that the plaque end point can be seen. Another concern is the potential time of damping and embolization. We have approached this differently in that we clamp off the aorta below the renal arteries and corn-
Dr. J. Harold Harrison (Atlanta, Ga.). We have been doing transaortic endarterectomy of the renal arteries through the anterior approach for 30 years. We have transected the aorta for complete occlusion extending to the level of the renal arteries, but we have not used this tech,nique. Visualization of the endpoint of the plaque in the
REFERENCES
1. Wylie EJ, Perloff DL, Stoney RJ. Autogenous tissue revascularization techniques in surgery for renovascularhypertension. Ann Surg 1969;170:416-28. 2. Ernst CB, Stanley JC, Marshall FF~ Fry WJ. Renal revascularization for arteriosclerotic renovascular hypertension: prognostic implications for focal renal arterial vs. overt generalized arteriosclerosis. Surgery 1973;73:859-67. 3. Libertino JA, Thierry AF, Zinman LN, et al. Changing concepts in surgical management of renovascular hypertension. Arch Intern Med 1988;I48:357-9. 4. NovickAC, Ziegelbaum M, Vidt DG, Gifford RW Jr, Pohl MA, GoormaasticM. Trends in surgical revascularizationfor renal artery disease--Ten year's experience. JAMA 1987; 257:498-501. 5, Tarazi RY, Hertzer NR, Beven EG, O'Hara PJ, Anton GE, Krajewski LP. Simultaneous aortic reconstruction and renal revascularizafion: risk factors and late results in 89 patients. 1 VASCSURG1987;5:707-14. 6, StewartMT, Smith RB Ill, Fulenwider IT, Perdue GD, Wells JO. Concomitant renal revascularization in patients undergoing aortic surgery. J VASCSUinG1985;2:400-5. 7. Brewster DC, Buth 1, Darling RC, Austen WG. Combined aortic and renal artery reconstruction. Am J Surg 1976; 131:457-63. 8. Shahian DM, Najafi H, Iavid H, Hunter JA, Goldin MD, Monson DO. Simultaneous aortic and renal artery reconstruction. Arch Surg 1980;115:1491-7. 9. Sterpetti AV, Schultz RD, Fddhaus RJ, Peetz DJ. Aortic and renal atherosderotic disease. Surg Gynecol Obstet 1986; 163:54-9. 10. Dean RH, Kieffer RW, Smith BM, et al. Renovascular hypertension: anatomic and renal function changes during a drug therapy. Arch Surg 1981;116:1408-15. 11. Pohl MA, Novick AC. Natural history-of atherosclerotic and fibrous renal artery disease: clinicalimplications. Am J Kidney Dis 1985:A120-30. 12. Schreiber MJ, Pohl MA, Novick AC. The natural history of atherosderotic and fibrous renal artery disease. Urol Clin North Am 1984;11:383-92. 13. WollenweberJ, Sheps SG, Davis GD. Clinical course of ath~ erosderotic renovascular disease. Am I Cardiol 1968;21:6071. 14. Meaney TF, Dustan HP, McCormack LI- Natural history of renal arterial disease. Radiology 1968;91:881-7.
232
Stoney et al.
plete any distal surgery. We then expose the proximal aorta and renal and mesenteric vessels and use an anterior approach to the renal arteries. If at this point we find a widened or sickly looking aorta, and particularly if we find "cheese" when we cut across the aorta, we will do bypass grafting. Dr. Messina, what do you do when you have completed an extensive dissection of the vessels, you have placed an occluding clamp above the renal and/or ruesenteric arteries, you have not done your surgery below, and you run into this extensive disease? In the article the exposure is described as midline or a high transverse. To me a high bilateral subcostal incision offers the best exposure of vessels in this area. Dr. William J, Fry (Dallas, Texas). I express my appreciation to the authors for allowing me to review the manuscript and compliment them on the low mortality in this group of patients at very high risk. We have had a great deal of difficulty treating our group of patients who have atherosclerosis of the renal arteries. In a series ofrnore than 125 cases that we studied recently, all deaths (a 5% mortality) have occurred when we have tried to do combined operations. Therefore we have tried to avoid this whenever possible, and our mortality rate with staged operation so far has been zero. We think that the severe cardiomyopathy in these patients makes them a poor risk for a prolonged procedure. Nevertheless we are impressed with the outstanding results that Dr. Messina and his group have brought to us today. I have several questions. It is not quite clear in the article exactly what size defects revealed by ultrasound scanning postoperatively will make you open the renal artery and operate and which ones you will let go. This is important to us, since the operative ultrasound is a very sensitive instrument and sometimes it is a bit confusing as to exactly what is significant and what is not. Your indications for operations are not entirely clear to me. Which ones require endarterectomy and which ones do not? I am not aware of any data that have shown progression of atherosclerotic lesion in the renal artery in the absence of hypertension; therefore I ask you why you did approximately 20% of these operations prophylactically. I echo what Dr. Harrison said; in the patient with an aneurysm it is difficult to do endarterectomies. We tend to avoid trying to do endarterectomies in aneurysmal disease and would rather do a bypass. Is there a particular plane to get in that would give more Strength and a better division at the end of the specimen? Dr. Richard H. Dean (Winston-Salem, N.C.). I thank the authors for giving me a preliminary copy of their manuscript and compliment Dr. Messina on his excellent presentation. I would like to make a couple of points and ask a couple of questions. Although I disagree with the use of transaortic endarterectomy, I recognize that under the able hands of the San Francisco group, the results with the transaortic endarterectomy have been quite good and are comparable to results with other techniques. However, I continue to be concerned about the adaptability of this technique to the range of expertise of those dealing with
lournal of VASCULAR SURGERY
renal vascular surgery in the country as a whole, and I am also concerned that the visualization of the end point by the transaortic route lasts for only about half a second. I have difficulty in believing that visualization is adequate during that very brief moment. One can only infer the condition of the transection point by examination of the endarterectomy specimen. The reasonI bring this up is the 28 small defects the authors identified during intraoperative post_reconstruction duplex scanning. With the transverse scanning technique we have commonly seen loose pieces of intima at the distal transection point, and these have been relatively simple to pull away. I continue to be concerned about what would happen if I left those alone. If I correctly interpret the 28 small defects as probably being similar to those that we have seen and removed, perhaps there is no difficulty if they are left in place. I question the indications for using the techniques reported. Specifically, as best as I can tell from reading the manuscript, about 20% of the patients had no clear-ct~ indication for revascularization or transaortic endarterec~ tomy other than the fact that they had aortic disease and renal artery stenoses. My question to Dr. Messina in that regard is whether the San Francisco group have changed their indications for renal revascularization so that it is now done prophylactically, or do those patients have peculiar stenoses or are they patients in whom the context of therapeutic intervention had to be modified for some specific reason? Clearly, if their suggestion is that there are now data showing that all nonhemodynamic lesions seen angiographically should be treated with endarterectomy or bypass, I am not familiar with those data, and I therefore continue to believe, alternatively, that we should not be operating on such patients prophylactically. In that regard, I have one further question that relates to the results. Eighty percent of the patients' renal function was either improved or stable. Does that imply that 20% of the patients had worsening of their renal function as consequence of the procedure? Similarly, as relates to h~~ pertension where 13% of the patients had a worsening of their hypertension, do these match up with the technical issues identified by the duplex scanner or in early followup studies showing residual or recurrent lesions? Again, I would compliment the authors and their surgical colleagues on the technique reported and I appreciate the opportunity to discuss it. Dr. Messlna (closing). I thank all the discussants for their comments. I will address Dr. Harrison's concern about the applicability of this technique to various forms of combined aortorenal atherosclerosis and then describe some of the techniques that can be used to avoid complications associated with aortorenal endarterectomy. We believe that this technique is well suited to most forms of combined aortorenal atherosclerosis. Typically these renal lesions are bilateral and proximal extensions of aortic atheromas. The major contraindications to the application of this procedure are the presence of a pararenal aneurysm or
Volume 9 Number 2 February 1989
a degenerated or patulous aorta, or the occurrence of distal renal artery disease. I would emphasize a number of points to avoid technical complications. The completeness of the initial dissection is critical to the successful conduct of this operation. It is particularly important to obtain proper exposure of the aorta in the area of superior mesenteric and renal arteries. This usually entails incision o f the netlral tissue around the aorta at the level of the S/vIA as well as incision of the crura of the diaphragm bilaterally. The left renal vein should be fully mobilized, which usually requires transection and ligation o f its nonrenal branches. Transection of the aorta just beyond the level of the renal arteries has provided more than an adequate view of the aorta and allows one to see directly down into the orifices of the renal artery during the renal endarterectomy. To avoid embolization we emphasize careful dissection, avoiding unnecessary manipulation of the aorta, and clamping distally before applying a proximal clamp. In addition, we emphasize thorough irrigation of the aorta a~lcl renal arteries and back bleeding and flushing of the vessels before restoring flow, all to help minimize the occurrence of atheroembolization. Dr. Fry raised the concern about doing endarterectomies on patients with aneurysmal disease. If the endarterectomy plane is too deep into the media or the aorta is abnormal, a "blow out" of the aorta can occur. As previously mentioned, the main contraindication to the application of this procedure is the presence of a pararenal aneurysm or a degenerated, patulous aorta. Both Dr. Fry and Dr. Dean raised questions about the intraoperative duplex scan results. This is an evolving technique. Twenty-eight small defects were seen in 37 patients. However, most patients had postoperative angiography, and only two small defects were seen on these studies. I believe the differences between these two studies indicate how sensitive intraoperative duplex scanning is. Scanning identifies tiny intimal fragments or platelet aggregates that ar~enot seen on angiography. All moderate and large defects al~ opened when appropriate. The most common indication for opening the artery is an inadequate end point; the artery is transected distally and a good end point obtained.
Modified transaortic endarterectomy 233
Both Dr. Dean and Dr. Fry wanted more information concerning the 20% of patients who had a "prophylactic operation." That represented nine of the 44 patients. Three of the nine were less than 50 years old, arid the indication for renal revascularization was the presence of either bilateral renal artery stenoses or a high-grade unilateral stenosis. Another group of three patients required aortic and renal endarterectomy during the course of the management of infected prosthetic bypass grafts. In two, the aortic stump required endarterectomy to ensure against late occlusion of the renal arteries because of the obligatory infrarenal aortic occlusion necessary to manage their graft infection. In the third patient the endarterectomy was performed to reopen the pararenal aorta to provide adequate cuff for anastomosis o f a new prosthetic graft. An infection of an axillofemoral graft had developed in this patient. Dr. Fry is correct; there are not good natural history data on the management of renal artery stenoses in the absence of clinically significant hypertension. The only studies published, which include an excellent one by Dr. Dean, refer to patients who are being evaluated for hypertension. However, in these studies it is clear that when a renal artery stenosis becomes high-grade these lesions will show progression in 60% of patients within 2 years. In these patients there was up to a 54% incidence of deterioration of renal function not heralded by loss of blood pressure control. Most of our patients had serious, clinically identifiable complications of their renal artery disease. Sixty-four percent had elevated creatinine levels and 82% had hypertension. These patients had good responses of their hypertension management postoperatively. The patients who were operated on in the "prophylactic group" represented patients with unusual, atypical features that required selective judgment. We believe that young patients who are good operative risks and have hemodynamically significant stenosis should undergo prophylactic endarterectomy to prevent loss of renal function. However, this is predicated on close monitoring of one's own results so that the benefits outweigh the risk of this aggressive management policy. None o f the patients operated on prophylactically had deterioration of renal fimction or hypertension postoperatively.