325

Arteriography after Carotid Endarterectomy

John Holder 1 Of 55 patients undergoing carotid endarterectomy, 16 had abnormal postoperative Eugene F. Binet2 angiograms by accepted literature criteria. Five of the 16 were symptomatic. The other Stevenson Flanigan3 11 were neurologically stable or improved from their preoperative condition. None of the 16 patients underwent reoperation. Of those 11 who had abnormal postoperative Ernest J. Ferris 1 angiograms but a good clinical result, four had a second postoperative angiogram some months later that demonstrated marked improvement in the appearance of the endarterectomy site. Patients undergoing carotid endarterectomy should not be sub­ jected to routine postoperative angiography without clinical indications nor should they undergo reoperation on the basis of angiographic findings alone without consideration of their clinical status.

Cerebral angiography remains th e most precise method to evaluate the path­ ologic changes that occur in extracrani al cerebrovascular disease involving th e internal carotid artery in th e neck. Several authors have strongly endorsed its use in the immediate postoperative period to evalu ate the patency of th e end­ arterectomy site. In some institutions reoperation is performed if the angiographic findings appear unsatisfactory without consideration of the patient's neurologic status. This report will examine the preoperative and postoperative angiograms of patients who were not subjected to reoperation despite having abnormal appearing postoperative carotid angiograms.

Materials and Methods

Postoperative arteriography has been routinely perform ed on all pati ents subiected to carotid endarterectomy by the Neurosurgery Service at the Little Rock Veterans Adminis­ tration Medical Center. The angiographic studies of 55 such patients operated on during a Received August 28, 1980; accepted after re­ 6 year period were evaluated. These male patients were 42- 81 years old (average, 60 '12 vision March 4 , 1981 years). Seventy percent of the patients were seen with transient ischemi c epi sodes and Presented at the annual meeting of the Ameri­ 30 % were evaluated for completed stroke. In addition to th eir extracran ial cerebrovascular can Roentgen Ray Society , Las Vegas, NV, April disease, two patients had diabetes mellitus, two had hyperlipidemi a, and 1 2 showed 1980. intracranial cerebrovascular disease at angiograph y. , Department of Radiology, University of Arkan­ The preoperative angiographic evaluation in these cases consisted of selecti ve cathe­ sas for Medical Sciences, 4301 W. Markham, terizati on of both common carotid arteries with anteroposteri or and lateral angiography of Little Rock, AR 72205. Address reprint requests both the head and neck vessels performed in all patients. In certain cases selecti ve to J. Holder. vertebral angiograms were also obtained . Oblique projecti ons of th e carotid bifurcati on 2 Department of Radiology, Veterans Ad minis­ were also included when necessary. Aortic arch examinati on preceded the selecti ve studies tration Medical Center, Little Rock , AR 72206. in 75% of the cases. Th e postoperative angiographic evalu at ion consisted of selective 3 Department of Neurosurgery, University of catheterization of the " operated artery " with an teroposteri or and lateral angiograms of th e Arkansas for Medical Sciences, Little Rock , AR 72205. head and neck plus selective study of the opposite carotid artery if significant disease was noted involving that vessel on th e preoperative study. This article appears in July / August 198 1 AJNR and September 198 1 AJR. All patients were operated under general anesthesia, most with electroencephalographic mon itoring. Intraarterial shunts were not used unless electroencephalog raphic changes AJNR 2:325-329, July / AU9ust 1981 0 195-6108/ 8 1/ 0204 - 0325 $00.00 occurred or disease on th e contralateral side was present. © Am eri can Roentg en Ray Society Th e hospital charts on 48 (87%) of the 55 pati en ts were avail able for review. These 48 326 HOLDER ET AL. AJNR:2, July/ August 198 1

pati ents were foll owed an average of 15'/2 months; th e longest carotid artery on postoperative arteriography, three with foll ow-up was 5 years, 3 months and th e shortest was 6 weeks. significant stenosis in the distal limb of their endarterectomy, Two patients died in th e immediate postoperati ve peri od. and three with stenosis of th e proximal limb. Also included were three patients with large persistent filling defects or webs on the postoperative angiographic study. Four of the Results 11 patients in this group had a second follow-up angiogram Of th e 48 patients avail able for c linical follow-up, 77% at 19, 19.5, 20, and 33 months, respectively. Their case were stated to have had a good c linical result. That is, their presentations foll ow. transient ischemi c episodes ceased or their cerebrovascular disease did not progress in the postoperative period. On the o th er hand , 23% of th e pati ents available for foll ow-up had a poor result; th at is, they suffered a stroke in the postop­ Case Reports erative period or th eir transient ischemi c episodes contin­ ued. Case 1 Of the 55 immediate postoperative angiograms obtained A 55-year-old man had acute onset of ri ght-sided weakness and in this seri es, 9% were given an excellent rating. An excel­ expressive aphasia. His preoperative angiogram demonstrated 95% lent postoperative angiogram was interpreted as entirely stenosis of the left intern al carotid artery (fig. 1 A) . A follow-up normal or showing very minimal postoperative changes. angiog ram 9 days aft er a successful endarterectomy showed a Another 57.5% of the angiograms were given a good rating. thick cuff of ti ssue at th e proximal margin of th e endarterectomy They showed only irregularity of th e vessel wall , occlu sion site with 50% stenosis of the common carotid artery and 99% stenosis of the external carotid artery (fig . 1 B). The paiient was of the external carotid artery, or a stenosis of the internal clinicall y stabl e and was not returned to th e operating room . Foll ow­ caroti d artery th at did not exceed 30% . The other 33.5% of up angiogram 19 months later showed very minimal postoperative angiograms were given a poor appearance rating as they changes. Th e cuff had disappeared and the extern al carotid artery showed stenosis of the internal carotid artery greater than had returned to normal (figs. 1 C and 1 D) . Th e pati ent remained 30% . Almost two-thirds of th e patients had th eir follow-up well. angiogram within 30 days of th eir surgical procedure, most of th em while still in the hospital. In an attempt to compare th e results of th e surgical procedure and th e radiographic Case 2 appearance of the carotid artery on the postoperative an­ A 60-year-old man was seen after several attacks of aphasia. A giogram, the patient population was divided into four groups. 95% stenosis of th e carotid artery was seen on preoperati ve angio­ Group 1. Thi s group compri sed the 26 patients having a gram (fig. 2A). Repeat angiogram 5 days after su rg ery showed a good c linical result and a good or excell ent appearance of large intimal flap within the lumen of the common caroti d artery (fig. the carotid artery on their postoperative angiogram. The 2B). The patient was asymptomati c so surgery was deferred. Fol­ average c linical follow-up for this group was 12.7 months low-up angiogram at 95 days again identified th e intimal flap (figs. with the longest pati ent being foll owed 3 '12 years and th e 2C and 2D) . Th e patient remained asymptomatic. A subsequent shortest, 2 months. angiogram at 19.5 months after endarterectomy revealed th e fl ap Group 2. These six patients had a poor c linical result but to still be present but less prominent (figs. 2E and 2F). The patient's a good or excell ent appearance on th eir postoperati ve an­ c linical statu s remained stable during thi s period . giogram. Their average follow-up time was 3'12 months. In cluded in thi s group were several pati ents whose transient Case 3 ischemic episodes either persisted into or returned in the early postoperative period despite a satisfactory appearing A 59-year-old known hypertensive white man had sudden onset internal carotid artery on postoperative arteriography. of left hemiparesis and left facial numbness. Hi s preoperati ve ca­ Group 3 . These fi ve patients had a poor c linical result and rotid angiogram demonstrated a 40% stenosis of the internal carotid a poor radiographic appearance on their postoperative an­ artery with ulcerati on (fig . 3A) . A foll ow-up study 6 days aft er giograms. They had an average clinical follow-up of 21 surgery showed a thick cuff of ti ssue at the proximal margin of the months with th e longest being 5 years, 3 months and the endarterectomy causing a 50% stenosis of th e common carotid artery (fig. 3 B). The patient was asymptomati c so no surgery was shortest being 1 day. In this group were several patients performed. A second angiogram at 20 months after surgery showed who complained of persistent transient ischemic attacks or only very minimal postoperative c hanges (figs. 3C and 3D) . The vertebrobasilar insuffic iency. One patient died of a massive patient has been well since surgery. cerebral infarc t. Several postoperati ve angiograms in this group showed internal carotid artery occlusion or intralu­ minal c lots or webs. Case 4 Group 4. These 11 patients were said to have a good A 60-year-old man had transient diplopia. Hi s preoperative an­ c linical result, but a poor radiographic appearance on their giogram showed an ulcerated plaque and a 50% stenosis of the follow-up angiogram. They were foll owed for an average of internal carotid artery (fig. 4A). Repeat angiogram 26 days after sli ghtly over 11 months, with the longest follow-up being 4 surgery showed 40% stenosis of the carotid artery (fig. 4B). Re­ years and the shortest, 15 days. In c luded in this group were operation was not performed. A second foll ow-up angiogram at 33 two patients with asymptomati c occlusion of the internal months showed minimal postoperative changes (figs. 4C and 4D). AJ NR :2, July/ August 1981 ARTERIOGRAPHY AFTER CAROTID ENDARTERECTOMY 327

Fig. 1.-Case 1, 55-year-old man with acute onset of right hemiparesis and expressive aphasia. A , Preoperative ca­ rotid angiogram shows 95% stenosis of intern al carotid artery. B, Postoperative angiogram 9 days later. Thick cuff of tissue at proximal margin of end­ arterectomy si te with 50% stenosis of common carotid artery and 99% steno­ sis of external carotid art ery. C and D, Follow-up angiogram 19 months later. Very minimal postoperative changes. Cuff has disappeared and extern al ca­ rotid artery has returned to normal.

A B c o

A B c o E F Fig. 2. - Case 2, 60-year-old man with multiple attacks of aphasia. A, angiogram at 95 days. Flap is unc hanged. E and F, Final angiogram al 19 .5 Preoperative carotid angiogram shows 95% stenosis of carotid artery. B, months. Fl ap is still present but less prominent (arrowheads). Repeat angiogram 5 days after surgery. Large intimal flap. C and D, Repeat

Fig. 3. -Case 3, 55-year-old man with acute onset of left hemiparesis and facial numbness. A , Preoperative carotid angiogram shows 40% stenosis of inter­ nal carotid artery with ulceration. B, Fol­ low-up study 6 days aft er surgery. Thick cuff of tissue at proximal margin of end­ arterectomy causes 50% stenosis of common caro tid artery. C and D, Repeat angiogram 20 months after surgery. Only very minimal postoperative changes. A B c o 328 HOLDER ET AL. AJNR: 2 , July/ August 1981

Fig. 4. -Case 4, 60-year-old man with transient diplopia. A , Preoperative carotid angiogram show s ulcerati ve plaque and 50% stenosis o f internal ca­ rotid artery. B, Repeat study 26 days after surgery shows 40% stenosis of ca­ rotid artery. C and D, Foll ow-up stu dy at 33 months after operation. Only minimal postoperative c hanges.

A B c o

properly tacked down. They also felt that an angiogram Discussion should be obtained to evaluate the hemodynamic signifi­ Angiography after carotid endarterectomy may be per­ cance of externall y apparent narrowings or to evaluate the formed either within the operative suite while the patient is final result of a technically difficult operation. still under anesthesia (intraoperative angiography) or in the Shultz et al. [3] indicated that angiography in the imme­ radiology department at varying times after surgery (post­ diate postoperative period should be performed if a transient operative angiography). There are advantages and disad­ or progressive neurologic deficit occurred. His group also vantages to both techniques. Intraoperative angiograms are performed postoperative studies to determine the " smooth­ usuall y of poor technical quality for several reasons, in c lud­ ness" of a vessel wall before operating on the contralateral ing insufficient contrast materi al concentration within the carotid artery. blood vessel. Sterile fi eld constrai nts often preclude proper There is a diversity of opinion as to what constitutes an positioning of the patient. The physical li mitations of the x­ acceptable angiogram after carotid endarterectomy. Au­ ray equipment do not allow for high milliamperage, short thors differ as to what angiographic findings would require exposure, or rapid , seri al radiographs. The resultant product them to reoperate on a carotid artery. Plecha and Pories [4] is frequently overexposed, mistimed, and poorly positioned. revised their vascular reconstructions when a significant Routine intraoperative angiography has enjoyed its greatest defect was noted on angiography. They did not comment success at those medical centers where the operating room on the meaning of the word " significant. " Gurdjian et al. [5] is equipped with x-ray apparatus capable of producing reoperated if the foll ow-up angiogram showed an obvious radiographs of high diagnostic quality. stenosis but did not elaborate further on " obvious. " On th e other hand, angiography in the immediate post­ The presence of a suture lin e stricture, platelet thrombi, operative peri od is accomplished under more ideal c ircum­ atheroscleroti c debris, or an intimal fl ap on angiography stances. It is usuall y done in a speciall y equipped angiog­ were all c ited by Dardik et al. [6] as reasons for reoperation. raphy suite within the radiology department by trained an­ Of the seven reoperated patients in th e series of 13 1 con­ giographers using catheter techniques. High quality, secutive cases reported by Anderson et al. [2], one intra­ properly positioned, rapid sequence serial fi lms are obtain­ operative angiogram showed a significant stenosis, two able. showed exceptionally thick c uffs at the proximal ends of the In the experience of several investigators the technical endarterectomies, and three showed occlusion of th e exter­ difficulties associated with intraoperative studies are more nal carotid arteries. In a series of 100 patients reported by than offset by the desire to immediately assess their oper­ Blaisdell et al. [1], only patients demonstrating a stenosis of ative result. Blaisdell et al. [1] performed routine intraoper­ 30% or greater on intraoperative angiography were reop­ ative carotid angiography before c losing the in c ision in the erated. belief that a normal operative study assures them that th e Usi ng the most stringent criterion reported in th e litera­ immediate and late technical results will be excell ent. On ture, that is, a restenosis of 30% or more of an operated the other hand, Anderson, Collins and Rich [2] found that carotid artery, 16 of th e patients in our series had an routine intraoperative angiography did not change their in­ unacceptable post endarterectomy angiogram. Of that num­ c idence of postoperative neurologic complications. Their ber, 11 (group 4) were asymptomati c. These 11 patients experience with 13 1 consecutive operative angiograms would have been reoperated on immediately had their an­ failed to support th e contention that its use should be giograms been obtained while they were still in the operating routine. They indicated that operative angiography should room under anesthesia. Their stable neurologic examination be performed in only certain c ircumstances such as when would not have pl ayed a role in determing the need for the operator had difficulty passing a shunt tube or when it reexploration of their endarterectomy sites. could not be ascertai ned that the distal intima had been Of the five patients (group 3) who had both an abnormal AJNR:2, July / August 198 1 ARTERIOGRAPHY AFTER CAROTID ENDARTERECTOMY 3 29 postoperative angiogram and a poor clinical result, two had should not be subjected to reoperation on the basis of total occlusion of the operated carotid artery at postopera­ abnormal postendarterectomy angiograms alone. The pa­ tive angiography. Immediate reoperation of an occluded ti ent's clinical statu s should also be considered when mak­ carotid artery has met with mixed results. In the other three ing that decision. patients in group 3, intraoperative angiography may have Intraoperative angiography will continue to be performed been of value in delineating a remedial lesion but was not in those instances when the operator is concern ed about done for technical reasons. the integrity of the surgical procedure. Postoperative studies In four of the 11 patients in group 4 , a second angiogram should be performed when th e patient's clinical conditi on was obtained at 19, 19.5, 20, and 33 months postend­ demands it. Routine postoperative studies after caroti d end­ arterectomy, respectively (figs. 1-4). In three of the four arterectomy should not be done on asymptomati c or neu­ patients with this second postoperative angiogram, the sig­ rologically improving pati ents. nificant abnormalities identified on the immediate postop­ erative study had resolved . In case 4, while the intimal flap remained visible, it had become less prominent. By literature REFERENCES criteria, all four of these patients should have been returned 1 . Blaisdell FW , Lim R, Hall AD. Technical result of caroti d end­ to the operating room for reoperation in the immediate arterectomy. Am J Surg 1967;11 4 :239- 24 6 postoperative period. Instead, because of the lack of clinical 2. Andersen CA, Collins GJ , Ri ch NM. Routine operati ve art eri­ symptoms they were not reoperated and their repeat follow­ ography during caroti d endarterectomy: a reassessment. Sur­ up angiograms demonstrated that the changes on arteriog­ gery 1978;83: 6 7 -73 raphy seen in the immediate postoperative period were 3 . Schutz H, Fl eming JFR, Awerbuck B. Arteri ographic assess­ temporary. Speculation as to why the appearance of these ment of carotid endarterectomy. Ann Surg 1970; 171 : 509- vessels would so drastically change in periods of less than 521 3 years is conjecture but may be related to vessel wall mediate assessmen t of arteri al reconstructi on. Arch Surg 1972; 105: 902- 90 7 edema and tissue fragmentation caused by the trauma of 5. Gurdjian ES, Hardy WG, Lind ner DW, Th omas LM . Results of the surgical procedure [7]. Perhaps reendothelialization is endarterectomy in th e treatm ent of cerebrovascular disease. principally responsible for the improved appearance of the Angiology 1964; 15: 88-1 0 2 vessel wall on the follow-up study. 6. Dardik II , Ibrahim 1M , Sprayregen S, Veith F, Dard ik H. Routi ne While the group of patients studied here is of only modest intraoperati ve angiography. Arch Surg 1975; 110: 184-190 size, their clinical and radiographic findings strongly suggest 7 . Dirrenberger RA, Sundt TM Jr. Caroti d endarterectorn y. J that patients who are neurologically stable or improved Neurosurg 1978;48 ; 201 - 2 1 9