- Email: [email protected]
Resection of the internal carotid artery and replacement with greater saphenous vein: A safe procedure for en bloc cancer resections with carotid involvement
Resection of the internal carotid artery and replacement with greater saphenous vein: A safe procedure for en bloc cancer resections with carotid involvement J. G o r d o n Wright, M D , R o y N i c h o l s o n , M D , D a v i d E. Schuller, M D , and William L. Smead, M D , Columbus, Ohio
Purpose:Many patients who have advanced cancer of the neck will have involvement of the internal carotid artery. The management of this condition remains controversial, and a wide range of therapeutic options have been suggested including ligation, "shaving" the tumor off the carotid, or en bloc resection and replacement of the internal carotid artery by polytetrafluoroethylene, vein, or superficial femoral artery. We reviewed our experience with en bloc resections of the internal carotid artery in a consecutive series of patients who had malignancies involving the internal carotid artery at a single institution from 1989 to 1995. Methods:We used a retrospective chart review based on a list of 20 patients generated by the Hospital Cancer Registry and our Vascular Surgery clinical database. Results: All patients had their internal carotid artery removed and replaced with a greater saphenous vein while they were under general anesthesia. A resection of their cervical malignancy was also performed. Concomitant myocutaneous flaps were rotated over the carotid bypass in six (30%) patients. Eight (40%) of the bypass grafts were nonreversed, and 12 (60%) were reversed, with a clear trend towards using nonreversed veins more recently. Shunts were used in 18 (90%). Eighteen of the 20 patients had some form ofintraoperative contamination including tracheostomies, pharyngostomies, or fistulas. Half of the patients had intraoperative radiation therapy, and 16 (80%) patients underwent operation for recurrent cancer. During the follow-up period two (10%) patients had strokes (one minor and one major), and one patient had a graft blowout, which was treated by ligation without stroke. One patient had an asymptomatic occlusion of his graft. Conclusions: From these results we conclude that the use of the greater saphenous vein to replace the internal carotid artery after en bloc resection is not attended by a high rate of infectious complications or graft blowout even in the presence of intraoperative tracheopharyngeal contamination and that the greater saphenous vein is the conduit of choice for replacing an internal carotid artery after cancer resections. (J Vasc Surg 1996;23:775-82.)
When attempting a cure for patients with advanced head and neck cancer, the surgeon will remove all neck metastases along with the primary tumor From the Departments of Surgeryand Otorhinolaryngology,The Ohio State UniversityCollegeof Medicine and the Arthur James Cancer Hospital. Presented at the Nineteenth Annual Meeting of the Midwestern VascularSurgicalSociety,Chicago,Ill., September 23-25, 1995. Reprint requests: J. Gordon Wright, MD, Division of Vascular Surgery, The Ohio State University, Department of Surgery, RoomN-716 Doan Hall, 410 WestTenth Ave., Columbus, OH 43210. Copyright © 1996 by The Society for Vascular Surgery and International Societyfor CardiovascularSurgery,North American Chapter. 0741-5214/96/$5.00 + 0 24/6/72502
while trying to preserve as much function as possible. Occasionally the malignancy will involve the carotid artery. This poses a special problem, because resection of the carotid artery (with or without replacement) puts the patient at additional risk for the potential complications o f stroke, graft infection, or graft blowout. In the past, these complications occurred frequently enough to dissuade many surgeons from undertaldng cn-bloc resections that involved the carotid artery. However, we have always believed that the best way to manage advanced head and neck tumors invading the carotid artery is to use a team approach that includes vascular surgeons and oto775
776
JOURNALOFVASCULARSURGERY May 1996
Wright et al.
laryngologists. We have not used the balloon occlusion test routinely to try and predict which patients will be better treated with carotid ligation alone, because in our experience too many false-negative test results occur, and the complication rate from carotid replacement is so low that we simply try and replace the carotid artery in all patients who need carotid artery resection. Autogenous saphenous vein is our graft conduit of choice, but other materials may be used. Graft patency and healing require appropriate concomitant soft-tissue coverage of any overlying mucosal or cutaneous defects. This study reviews the outcomes in 20 consecutive patients treated between 1989 and 1995 at The Ohio State University with en bloc resection and replacement of the carotid artery for head and neck cancer. Some controversy exists regarding the value of carotid artery resection and replacement with respect to its impact on the long-term survival of patients with advanced head and neck malignancies; however, most authors have suggested that carotid resection does not increase the long-term survival? In a recent meta-analysis and review of the pertinent literature, Snyderman and D'Amico 2 concluded that the impact of carotid resection on long-term survival remains unproven, but it does improve local/regional control of the disease. We would agree with their conclusions. Our minimum goal in resecting and replacing the carotid artery has always been to achieve better local/regional control without subjecting the patient to undue risk of postoperative stroke regardless of any survival benefit it may or may not provide. We believe that ours has been a rational approach, because most series reviewing the results ofunselected carotid ligations (i.e., without revascularization) report a 15% to 30% incidence of postoperative stroke and a 3% to 5% mortality specifically resulting from strokes. Therefore reduction of stroke risk should be an important goal for any palliative procedure that might require carotid artery resection, even though the putative survival benefit of carotid resection remains unproven. Many investigators have hoped that patients can be selected for a "safe" ligation of the internal carotid artery (ICA), if some type of temporary balloon occlusion test (BOT) could reliably predict that permanent occlusion of the internal carotid artery would not result in a stroke. Unfortunately, no one has demonstrated that any form of BOT (either by clinical examination alone or in conjunction with electroencephalography, transcranial Doppler scanning, or single photon emission computed tomography scanning) can reliably predict which
patients can undergo a "safe" ligation of the ICA. In most published series that have investigated the prcdictivc value of the BOT on the basis of clinical response alone, the stroke rates after a negative BOT result are approximately 30%, which is not significantly different from the stroke rate that would be expccted after unselected ligation alone. The predictive value of a negative BOT result can be increased slightly by adding some adjunctive test of cerebral blood flow (e.g., transcranial Doppler scanning, electroencephalography, or single photon emission computed tomography). In most published series a negative BOT result with some adjunctive test followed by carotid ligation carries a 10% stroke risk. We believe that even a 10% to 30% stroke rate is too high among patients who supposedly can have a "safe" ICA ligation and essentially renders the test useless, because we are able to achieve a major stroke rate of less than 5% by replacing the internal carotid arteries in all of these patients without any selection criteria. We believe that the reason that BOTs have been unsuccessful in predicting postoperative strokes is because most postoperative strokes do not occur immediately; this is because most postoperative strokes are due to the embolization of fresh clot that has recently formed in the ligated distal ICA stump. This opinion is based on our own unpublished experience with emergency ligations of the carotid artery for tumor erosion or penetrating trauma and is supported by the recent publication of Meleca and Marks, 3 who reported that six of eight of the strokes occurring after carotid resection had a delayed onset of this complication (typically on postoperative day 2 or 3). In addition, careful review of the literature demonstrates that at least half the reported strokes occurring after acute carotid ligation for tumor or trauma do not occur immediately but typically occur a few days later. It seems to us that the temporary occlusion of an internal carotid artery for 15 or 30 minutes is a poor way to predict what might happen on the second or third postoperative day. METHODS Patients were identified by manual comparison of a computer-generated list of all patients in the Ohio State University Cancer Registry with stage 3 or greater head and neck cancers who underwent a resection with another computer-generated list of all patients in the Vascular Surgery Registry who underwent a carotid artery procedure. The list was also compared with a list of all patients during this time period who were billed for a carotid artery procedure
JOURNAL OF VASCULAR SURGERY Volume 23, Number 5
and was found to be complete and comprehensive. All information was obtained from existing patient medical records or from information obtained during patient interviews. The information was abstracted onto data sheets and entered into a computer for analysis. Postoperative death and stroke were defined as occurring during the hospitalization for which the patient underwent the carotid resection and replacement. Average results are presented as the mean + SEM. RESULTS Patient population. From 1989 to 1995, 20 patients had resection of advanced head and neck tumors for which the internal or common carotid artery was resected and replaced electively. A total of 16 men and four women underwent operation. All the patients were white. The average age was 62.4 + 4.4 years (range, 20 to 78 years). Eight of the 20 patients were diabetic, 6 had significant coronary artery disease, and 5 were hypertensive. The patients reported in this series come from a larger cohort of 80 patients, 60 of whom did not undergo a carotid resection but had tumor close enough to their carotid artery that they were evaluated with a carotid angiogram. The mean survival for the 60 patients who did not undergo carotid resection and replacement was 18.1 months. The mean survival rate for the 20 patients who did undergo carotid resection and replacement was 19.2 months. This finding is not dissimilar from what has been reported in other series. We excluded from this analysis three patients who required carotid artery ligation to control exigent hemorrhage caused by tumor invasion. Also excluded from this analysis were three patients (early in the series) who had carotid artery ligation without an attempt to replace the carotid artery. We excluded these patients because the focus of this review was to analyze the septic, thrombotic, neurologic, and hemorrhagic complications of carotid artery resection and replacement with autogenous greater saphenous vein, and we were malting no attempt to compare resection and ligation alone with resection and replacement. Malignancies. The most common tumor type was squamous cell carcinoma (16, 80%), three patients had advanced thyroid cancer, and one patient had an advanced basal cell carcinoma. In 16 (80%) patients the cancer was recurrent. The primary tumors originated from the larynx (six), pyriform sinus (five), base of the tongue (two), nasopharynx (two), tonsillar fossae (one), and oropharynx (one). In three patients the primary tumor site was unlmown and could not be determined. All 16 patients with squa-
Wright et al.
777
mous cell carcinoma had a significant history of alcohol and cigarette consumption, and two of the remaining four patients also consumed significant amounts of alcohol and cigarettes. Six (30%) patients had contaminated necks on gross examination including two patients with necrotic tumor, two with active wound infections from previous biopsies, and two with draining fistulas. Preoperative evaluation. All patients had biopsyproven malignancies. Computed axial tomography (CAT) scans (18), magnetic resonance imaging studies (6), cerebral angiography (19), cervical ultrasonography4 studies (6), and BOTs (4) were routinely obtained when clinically indicated, but none of these tests was obtained in all patients. CAT scanning was the diagnostic modality that identified tumor invasion most often, but because it was also the test that was ordered most often, we cannot comment on the relative accuracy of CAT scanning versus magnetic resonance imaging ultrasonography or angiography in determining whether the carotid artery is involved with tumor. Indeed, it is our experience that when we suspect that the carotid might be involved (excluding those cases in which it is clearly involved or not), a clean tissue plane can be identified between the tumor and the carotid artery in at least half the cases. Because the Ohio State University serves as the primary referral center for central Ohio, many of these patients arrived with their preoperative evaluations complete or nearly complete. Preoperative treatment. In all 16 patients with recurrence a previous neck dissection was performed. Seven of these 16 patients had previously undergone some type of myocutaneous flap procedure. Preoperative irradiation was performed in 16 of the 20 patients, ~ and one patient had previously undergone brachytherapy with 1125 seed implantation. Operative treatment. The decision to resect the carotid artery en bloc with the tumor mass was made in the operating room by the otorhinolaryngologic surgeon. In all cases the tumor either encircled the carotid artery, or it was so densely adherent that it could not be safely removed. Before the carotid artery was clamped, the entire tumor mass was mobilized as much as possible. Heparin was used in all cases except one. The vein harvest site was kept isolated from the cervical operative site by using separate instruments and by changing gowns and gloves. Once a suitable length of greater saphenous vein was harvested, heparin was administered, clamps were applied, and the tumor mass and the involved segment of carotid artery were removed en bloc. The internal carotid artery was replaced by a greater saphenous vein in the
JOURNAL OF VASCULAR SURGERY
778
May 1996
Wright et al.
reversed (12, 60%) or nonreversed (8, 40%) orientation. When the vein was placed in the nonreversed orientation, a modified Mills valvulotome was used to cut the valves. Shunts were used in 18 (90%) of the cases. Typically, the entire extracranial and extrathoracic portion of the carotid artery was removed from a point 3 or 4 cm above the clavicle to a point 1 or 2 cm below the base of the skull. No special attempts were made to revascularize the external carotid artery, unless the carotid bifurcation and the external carotid were clearly free of any tumor involvement and the bypass could be safely terminated proximal to or distal from the bifurcation. In 18 (90%) of the cases the patient had some form of intraoperative contamination including tracheostomies, pharyngostomies, or fistulas. Half of the patients had intraoperative radiation therapy. New concomitant myocutaneous flaps were rotated over the carotid bypass in six (30%) patients. These included four deltopectoral flaps6 and two pectoralis major flaps. 7 In seven other patients flaps had already been fashioned during previous procedures. In one patient the carotid artery was covered with a skin graft, and in the remaining six patients the carotid artery was covered with local flaps. Postoperative management. Patients were given intravenous antibiotics for at least 24 hours or until all wounds were healing without signs of infection. Drains were left in and removed by the otorhinolaryngologic service, typically on postoperative day 2. In four of the patients a carotid duplex study was performed before discharge to determine graft patency. In the remaining 16 patients a carotid duplex scan was obtained sometime after discharge. In all cases except one the initial duple x study demonstrated graft patency. Patients were then monitored by both the otorhinolaryngologic and Vascular Surgery services with periodic carotid duplex studies. Early postoperative complications. None of the patients died during the early postoperative period. One patient had a major stroke that manifest on his emergence from anesthesia. This patient had undergone replacement of his carotid artery without the use ofheparin. On removal of the shunt, clot was noted at the end of the shunt, and poor back bleeding was seen. Therefore a no. 2 Fogarty catheter was passed distally until all retrievable clot was removed. Three patients had wound infections in the early postoperative period. Two of these were simple wound infections that resolved with prolonged antibiotics. One of these three was a phayngocutaneous fistula, which closed spontaneously while the patient was an outpatient. No other patients had any central
neurologic deficit during their hospitalization, nor did any graft occlusions, graft infections, or graft blowouts occur during the early postoperative phase. Late postoperative follow-up. One patient had a minor stroke i2 months after the operation that was manifest by contralateral hand and arm weakness that completely resolved during the subsequent 6 weeks. At the time he had his stroke, this patient underwent a carotid duplex scan, which was normal. Another patient had arterial hemorrhage from his contralateral neck dissection and myocutaneous muscle flap. This condition was later diagnosed as being a pharyngeocutaneous fistula with erosion into the common carotid artery but was on the side opposite our earlier carotid replacement. However, during the evaluation of this patient's hemorrhage, he underwent cerebral angiography, which demonstrated an asymptomatic occlusion of the carotid graft. A third patient had hemorrhage from his carotid graft approximately 1 month after discharge. This was the only carotid graft in this series that was covered with a sldn graft. The skin graft was used because the local flaps left approximately 1 square cm of the carotid graft uncovered. When this patient had exigent hemorrhage 1 month after the operation, he was treated with ligation of his carotid graft. He did not have any neurologic deficit. All other patients remained free of complications and had patent grafts after a mean follow-up of 15.9 + 7.8 months. In summary, no graft blowouts or strokes occurred as the result of septic complications. The only graft blowout that occurred did so after a skin graft coverage, and none of the remaining 19 patients who had their grafts covered with local flaps or myocutaneous rotation flaps had a graft blowout. Two (10%) patients had a stroke, one (5%) major stroke in the immediate postoperative period and one (5%) minor stroke without residual 1 year after the operation. DISCUSSION Elective carotid artery resection in the treatment of paticnts with advanced malignancies of the head and neck remains a controversial issue. Tumor invasion or adhcrencc to thc carotid artcry may require carotid resection to achieve control of disease. However, the significant risks of stroke and death and the low cure rate deter many surgeons from electively resecting the carotid artery, s This can result in some patients being denicd an appropriate operation and left exposed to the possibility of carotid rupture or other local tumor problems. Furthermore when thcse patients have a recurrent tumor involving the carotid artery, surgical treatment becomes more treacherous
JOURNAL OF VASCULAR SURGERY Volume 23, Number 5
because o f the previous neck dissection, irradiation therapy, myocutaneous flap rotation, and possible fistulae in the field. Unfortunately, few alternatives exist. Failure to fully resect the segment of the carotid involved with tumor may result in tumor recurrence or death o f carotid rupture .9 Several methods have been described to predict the likelihood o f stroke after carotid artery resection and ligation without replacement. These include intraoperative electroencephalography, carotid stump pressure monitoring, ocular plethysmography, somatosensory-evoked cortical potentials, and temporary balloon occlusion with or without xenon-enhanced computed tomography o f cerebral blood flow. Ehrenfeld et al.~0 concluded that a carotid stump pressure o f greater than 70 mm H g indicates adequacy o f collateral circulation sufficient to avoid a stroke after ligation. They also noted that intermediate carotid stump pressures (in the range of 55 to 69 m m Hg) represent an intermediate risk of stroke and that pressures less than 55 m m H g result in an unacceptable stroke rate. Others, however, have shown that carotid stump pressure is an inaccurate predictor o f subsequent neurologic compromise. 11 Ocular plethysmography as described by Mar tinez et al. 12 allows accurate and reliable prediction o f the adequacy o f collateral flow to compensate for carotid artery ligation. Atkinson et al.~3 used somatosensoryevoked cortical potentials in conjunction with carotid stump pressure and achieved a 17% stroke rate with ligation. Although this is one o f the lowest stroke rates after carotid ligation in the literature, it still seems unacceptably high to us. One o f the most recent advances in this area is the use by de Vries et al. ~4 of temporary balloon occlusion with xenon-enhanced computed tomography o f cerebral blood flow. This m e t h o d seems promising, although further studies are necessary to obtain statistically significant data. In our opinion none o f these methods has produced sufficiently accurate results to justify the conversion o f a patent carotid artery to an occluded carotid artery. The probable reason these methods are unreliable is simply that most strokes after carotid ligation do not occur immediately but usually occur from 1 to 5 days after the ligation. Therefore we should not expect that the effects o f acute temporary carotid artery occlusion can predict what will happen several days after permanent ligation. Furthermore the best series in the literature has a 17% stroke rate after simple ligation. This result does not compare very well with the current series and others that have also achieved significantly lower stroke rates with very acceptable septic and hemorrhagic complications.
Wright et al.
779
Therapeutic options for management o f head and neck tumors with invasion o f the carotid artery include irradiation t h e r a p y , ls-17 palliative carotid peeling, 18'19and resection with 2° 24 or without 2s 30 bypass grafting. A review o f the literature demonstrates a reported stroke rate o f 17% to 45% and a death rate o f 8% to 58% when treated with ligation alone, 2 and these rates seem to be significantly higher than when patients are treated with resection and replacement. Radiation therapy carries the additional risk o f radiation arteritis, 31 accelerated atherosclerosis, 3~ and delayed healing, 32 but might help achieve better local control o f the tumor. We have used both reversed and nonreversed orientations for the greater saphenous vein and have found advantages to both. Ifa significant difference is found in the diameter of the proximal and distal ends of the harvested veins (distal vein being larger), then we prefer to use the vein in the nonreversed orientation so we can match up the large end of the vein with the c o m m o n carotid artery and the small end o f the vein with the internal carotid artery. Because many o f these patients have their entire carotid artery resectcd from just above the clavicle to a point within 1 or 2 cm from the base o f the skull, the length o f greater saphenous vein is often 20 cm or more, and this often results in a considerable size differential between the proximal and distal ends of the vein. On the other hand, if no significant difference is found between the proximal and distal ends of the vein, then we prefer to use the vein in the reversed orientation to avoid the need for valve lysis. In our experience artcriography has not been especially useful in identifying tumor invasion o f the carotid artery. This has also been the experience o f Reilly et al.,1 who found that arteriography correctly identified tumor invasion o f the carotid artery in only 1 of 12 patients. In our uncontrolled experience CAT scans seem to be the most c o m m o n way that the surgeon is alerted to the possibility o f carotid involvement. Recent reports have suggested that ultrasonography might be the most accurate method for determining tumor invasion o f the carotid artery before operation. 33 We are not overly concerned with the preoperative diagnosis o f tumor invasion of the carotid artery, because ultimately this will always be a clinical judgment based on the findings o f the operating surgeon. CONCLUSIONS From these results we conclude that the use o f the greater saphenous vein to replace the internal carotid
780
Wright et al.
artery after e n bloc resection is n o t a t t e n d e d by a high rate o f infectious complications or graft b l o w o u t even in the presence o f intraoperative tracheopharyngeal c o n t a m i n a t i o n a n d that the greater saphenous vein is the c o n d u i t o f choice for replacing an internal carotid artery after cancer resections. W h e n c o m p a r i n g o u r results with the r e p o r t e d results in the literature after carotid resection a n d ligation w i t h o u t replacement, it is clear that carotid r e p l a c e m e n t has a m u c h lower risk o f postoperative stroke a n d death. I t is unlikely that the resection a n d r e p l a c e m e n t o f the carotid artery will achieve significantly better t u m o r c o n t r o l in these patients with already advanced malignancies, b u t further studies m i g h t prove this to be the case. REFERENCES
1. ReillyMK, Perry MO, Netterville JL, Meacham PW. Carotid artery replacement in conjunction with resection ofsquamous cell carcinoma of the neck: preliminary results. J Vasc Surg 1992;15:324-30. 2. Snyderman CH, D'Amico F. Outcome of carotid artery resection for neoplastic disease: a meta-analysis.Am J Otolaryngol 1992;13:373-80. 3. MelecaRJ, MarksSC. Carotidarteryresection for cancerofthe head and neck. Arch Otolaryngol 1994;120:974-8. 4. Zaragoza L, Sendra F, Soloano J, et al. Ultrasonography is more effective than computed tomography in excluding invasion of the carotid wall by cervical lympadenopathies. Eur J Radiol 1993;17:191-4. 5. Jones TR, Frusha JD. Carotid revascularizationafter cervical irradiation. South Med J 1986;79:1517-20. 6. BakamjianVY, Long M, Rigg B. Experiencewith the medially based deltopectoral flapsin reconstructive surgery of the head and neck. Br J Plast Surg 1971;24:174-83. 7. AriyanS. The pectoralis major myocutaneous flaps: a versatile flap for reconstruction in the head and neck. 1971;63:73-81. 8. Kenedy JT, Krause CJ, Loevy S. The importance of tumor attachment to the carotid artery.Arch Otolaryngo11977;103: 70-3. 9. Mares MD, Bruns MD, Keenan MJ. Carotid artery resection for head and neck cancer. Ann Otol Rhinol Laryngol 1992; 101:778-81. 10. EhrenfeldWK, Stoney RJ, WylieEJ. Relation of carotid stump pressure to safety of carotid artery ligation. Surgery 1983;93: 299-305. 11. KellyJJ, Callow AD, O'Donnell TF, et al. Failure of carotid stump pressures. Arch Surg 1979;114:1361-6. 12. Martinez SA, Oiler DW, Gee W, deFries HO. Electivecarotid artery resection. Arch Otolaryugol 1975;101:744-7. 13. Atkinson DP, Jacobs LA, WeaverAW. Elective carotid resection for squamous cell carcinomasof the head and neck. Am J Surg 1984;148:483-8.
JOURNALOF VASCULARSURGERY May 1996
14. De Vries EJ, Sekhar LN, Horton JA, et al. A new method to predict safe resection of the internal carotid artery. Laryngoscope 1990;100:85-8. 15. Fee W, Goff J, Dinat D, et al. Intraoperative iodine 125 implants. Their use in large tumors in the neck attached to the carotid artery. Arch Otolaryngol 1983;109:727-30. 16. Lee DJ, Liberman FZ, Park RI, Zinreigh ES. Intraoperative I-125 seed implantation of extensiverecurrent head and neck carcinomas, Radiology 1991;178:879-82. 17. Toita T, Nakano M, TakizawaY, et aL Intraoperative radiation therapy (IORT) for head and neck cancer. Int J Rad One Biol Plays 1994;30:i219-24. 18. Ketchum AS, Haye RC. Spontaneous carotid artery hemorrhage after head and neck surgery. Am J Surg 1965;110:64955. 19. Dibble DG, Gowen GF, Sheed DR. Observation on postoperative carotid hemorrhage. Am J Surg 1965;109:765-70. 20. Conley JJ. Free autogenous vein graft to the internal and common carotid arteries in the treatment of tumors of the neck. Ann Surg 1953;137:205-14. 21. ConleyJJ. Carotidarterysurgeryinthetreatmentoftumorsof the neck. Arch Otolaryngol 1957;65:437-46. 22. Reha AJ, Rongetti JR, Bisi R, Replacement of carotid arteries with prosthetic grafts. Arch Otolaryngol 1962;76:76-80. 23. Lore JM, Boulous JE. Resection and reconstruction of the carotid artery in metastatic squamous cell carcinoma. Am J Surg 1981;142:437-42. 24. Karam F, Schaefer S, Cherryholmes D, Dagher FJ. Carotid artery resection and replacement in patients with head and neck malignant tumors. J Cardiovas Surg 1990;31:697-701. 25. Watson WL, SilverstoneSM. Ligature of the common carotid artery in cancer of the head and neck. Ann Surg 1939;109: 127-32. 26. Moore O, Baker HW. Carotid artery ligation in surgery of the head and neck. Cancer 1955;8:712-26. 27. Gandhi K, Oppenheimer P. Emergency carotid ligation. Arch Otolaryngol 1962;75:451-6. 28. Moore OS, IGarlanM, Sigler L. Factors influencing the safety of carotid ligation. Am J Surg 1969;118:666-8. 29. Shumrick DA. Carotid artery rupture. Laryngoscope 1973; 83:1051-61. 30. Hefler KS, Strong EW. Carotid arterial hemorrhage after radical head and neck surgery. Am J Surg 1979;138:607-10. 31. Fonkalsrud EW, SanchezR, Zerubavi R, et al. Serialchangesin arterial structure following irradiation therapy. Surg Gynecol Obstet 1977;134:395-40. 32. Marcial-Rojas RA, Castro JK Irradiation injury to elastic arteries in the course of treatment of neoplastic disease. Ann Otolaryngol 1962;71:945-58. 33. Mann WJ, Beck A, Schreiber J, Manrer J, Amedee RG, Gluckmann JL. Ultrasonography of evaluation of the carotid artery in head and neck cancer. Laryngoscope 1994;104: 885-8.
Submitted Oct. 11, 1995; accepted Feb. 6, 1996.
JOURNAL OF VASCULARSURGERY Volume 23, Number 5
Wright et al.
781
DISCUSSION
Dr. R a m o n Berguer (Detroit, Mich.). The mortality and morbidity of head and neck tumors that invade the carotid artery has changed dramatically since the introduction o f en-bloc carotid resection and replacement with autogenous tissue. The mortality has dropped from 27% in the Sloan-Kettering series to 0% or 5% in the two series involved in this discussion. Among their 20 patients, the authors experienced two graft occlusions, two strokes, and one blowout requiring excision and ligation. In one of the strokes the patient had not received heparin, and the blowout occurred in a patient who had a skin graft coverage of a vein graft. Both problems probably could have been avoided. Before 1991 we replaced carotid arteries with saphenous vein in the manner described by the authors. Sometimes we placed saphenous veins that were small and posed a problem with size mismatch. We had two saphenous grafts occlude shortly after their placement. In 1991 we adopted a protocol for carotid replacement using superficial femoral artery as an autograft while replacing the femoral segment with a prosthesis of polytetrafluoroethylene (PTFE). Our experience with this protocol was presented to this society last year. The authors' and our series are comparable in numbers, age, gender distribution, length of follow-up, and incidence of previous radical neck dissection. O f the 24 patients in whom we did carotid arterial autograft replacement, the only neurologic complication noted was in one patient who had a stroke when his carotid graft was avulsed 6 hours after operation as the surgeons responsible for the myocutaneous flap were attempting to control bleeding from a muscular vessel. The carotid artery was ligated, and 24 hours later the patient had a massive stroke and eventually died. The 23 surviving patients had patent grafts and no strokes at a mean follow-up o f l 1 months. The other complications were at the site of the donor artery: a silent thrombosis of the PTFE in the femoral location of one patient and an infected false aneurysm of the PTFE proximal anastomosis requiring ligation in another. In both cases the legs remained viable. These numbers are too small to statisticallyconclude the superiority o f one method over the other. It can be said, however, that because arterial autografts do as well or better than saphenous vein grafts, that one should consider them a good choice if the preoperative duplex shows a saphenous vein of inadequate size or quality. We have used twice a saphenous vein graft in patients that showed atheroma of the superficial artery in the preoperative duplex. We will continue to use autogenous artery as our first choice until we accrue numbers that will support or negate this policy, and we will follow with interest the series of Dr. Wright and associates. J. G o r d o n Wright. I would agree with your comments entirely. I think that the superficial femoral artery is a good alternate to use in this situation, and clearly the number of cases in our series precludes a valid a statistical comparison between superior femoral artery and greater saphenous vein
at this point. I think in your hands, you have had good results with the superficial femoral artery, whereas we have had good results with the greater saphenous vein. I think the important take-home message from this study is that if you are faced with a malignant invasion of the carotid, you should not simply ligate and resect; you should make every reasonable attempt to revascularize because, as you mentioned, the mortality and morbidity has dropped significantly by adopting a policy of trying to revascularize all of these. I cannot think of a reason a superficial femoral artery would be superior except as you have mentioned the theoretic possibilities o f less infectious complications, although we did not see any infectious complications in this particular series. Dr. Joseph J. Piotrowski (Cleveland, Ohio). I did a number of these resections and replacements with saphenous vein when I was at the University of Arizona. Similar to your experience, most of our patients had had previous neck resections and postoperative irradiation. Because of this, we tunneled our vein grafts into the posterior triangle. I wonder if the single case of blowout that you experienced could have been prevented by an alternate routing of your vein graft, because you already were considering covering this with a somewhat scary skin graft. Dr. Wright. There is no doubt about that. This blowout was clearly due to an inadequate soft-tissue coverage, and it could have been prevented by the means you had suggested or by simply rotating some kind ofmyocutaneous flap. Dr. William D. Turnipseed (Madison, Wis.). I think your clinical series would suggest that some, but not all, of these patients require carotid reconstruction. This is born out by the fact that you have lost graft function in a couple of patients with no clinical consequence. If you could document before surgery an adequate intracranial collateral system, do you think it might simplify your surgical approach and eliminate the need for this algorithm in selected patients? Dr. Wright. If there were such a test, it would simplify things. However, we do not like the balloon occlusion test because we have had some patients who have had a negative balloon occlusion test who were ligated and subsequently stroked. Furthermore, our involvement with these cases is usually approximately 45 minutes, and many of these procedures take 6 or 8 hours. I am not sure that replacing the carotid in these large head and neck resections is really adding significantly more complexity to an already complex operation. Dr. Turnipseed. You might want to look into the use of magnetic resonance angiography and Diamox chemical stress testing. Magnetic resonance angiography can clearly define the anatomy of the circle of Willis. Magnetic resonance angiography can also be used to evaluate blood flow
JOURNAL OF VASCULARSURGERY
782
Wright et al.
May 1996
through watershed areas within the cerebral circulation before and after vasodilation with Diamox. This may prove to be an accurate preoperative screening test for these patients. Dr. Wright. One other point that is emphasized in the article is that these tests are all acute tests, and in this series (and others) most patients who have stroke after ligation do not have stroke immediately but usually stroke on the second or third day after ligation. Dr. Donald P. Spadone (Columbia, Mo.). I would have you entertain another hypothesis for the cause of stroke rather than inadequate intracerebral collaterals. It is possible that with ligation there is propagation ofthrombus within the ligated internal carotid and propagation or
embolization ofthrombus produces stroke 2 or 3 days after ligation. Is there evidence to support propagation ofthrombus or embolization in patients who did have stroke? Dr. Wright. Yes, that is exactly what I think happens with most of these patients. There is no objective evidence to prove or disprove that pathophysiology except as suggested by their clinical course. The fact that they usually do not have stroke immediately, that they usually have stroke on the second, third, or fourth postoperative day suggests that it is the propagation of thrombus into the circle of Willis, which embolizes and causes the stroke. Once again, this is why we do not like the balloon occlusion test or any other test of acute temporary occlusion.
T_HE MO ve? 0enaus, D o n ' t m i s s a s i n g l e i s s u e of t h e j o u r n a l ! To e n s u r e p r o m p t s e r v i c e w h e n y o u c h a n g e y o u r a d d r e s s , p l e a s e p h o t o c o p y a n d c o m p l e t e t h e f o r m below.
Please send your change of address notification at least six weeks before your move to ensure continued service. We regret we cannot guarantee replacement of issues missed due to late notification. J O U R N A L TITLE: Fill in the title of the journal here.
O L D ADDRESS:
N E W ADDRESS:
Affix the address label from a recent issue of the journal here.
Clearly print your new address here.
Name Address City/State/ZIP
COPY AND MAIL THIS FORM TO: Journal S u b s c r i p t i o n Services M o s b y - Y e a r Book, Inc. 11830 Westline I n d u s t r i a l Dr. St. Louis, M O 63146-3318
O R FAX T O : 314-432-1158
Mosby
O R PHONE: 1-800-453-4351 OutsidetheU.S.,call 314-453-4351
Purpose:Many patients who have advanced cancer of the neck will have involvement of the internal carotid artery. The management of this condition remains controversial, and a wide range of therapeutic options have been suggested including ligation, "shaving" the tumor off the carotid, or en bloc resection and replacement of the internal carotid artery by polytetrafluoroethylene, vein, or superficial femoral artery. We reviewed our experience with en bloc resections of the internal carotid artery in a consecutive series of patients who had malignancies involving the internal carotid artery at a single institution from 1989 to 1995. Methods:We used a retrospective chart review based on a list of 20 patients generated by the Hospital Cancer Registry and our Vascular Surgery clinical database. Results: All patients had their internal carotid artery removed and replaced with a greater saphenous vein while they were under general anesthesia. A resection of their cervical malignancy was also performed. Concomitant myocutaneous flaps were rotated over the carotid bypass in six (30%) patients. Eight (40%) of the bypass grafts were nonreversed, and 12 (60%) were reversed, with a clear trend towards using nonreversed veins more recently. Shunts were used in 18 (90%). Eighteen of the 20 patients had some form ofintraoperative contamination including tracheostomies, pharyngostomies, or fistulas. Half of the patients had intraoperative radiation therapy, and 16 (80%) patients underwent operation for recurrent cancer. During the follow-up period two (10%) patients had strokes (one minor and one major), and one patient had a graft blowout, which was treated by ligation without stroke. One patient had an asymptomatic occlusion of his graft. Conclusions: From these results we conclude that the use of the greater saphenous vein to replace the internal carotid artery after en bloc resection is not attended by a high rate of infectious complications or graft blowout even in the presence of intraoperative tracheopharyngeal contamination and that the greater saphenous vein is the conduit of choice for replacing an internal carotid artery after cancer resections. (J Vasc Surg 1996;23:775-82.)
When attempting a cure for patients with advanced head and neck cancer, the surgeon will remove all neck metastases along with the primary tumor From the Departments of Surgeryand Otorhinolaryngology,The Ohio State UniversityCollegeof Medicine and the Arthur James Cancer Hospital. Presented at the Nineteenth Annual Meeting of the Midwestern VascularSurgicalSociety,Chicago,Ill., September 23-25, 1995. Reprint requests: J. Gordon Wright, MD, Division of Vascular Surgery, The Ohio State University, Department of Surgery, RoomN-716 Doan Hall, 410 WestTenth Ave., Columbus, OH 43210. Copyright © 1996 by The Society for Vascular Surgery and International Societyfor CardiovascularSurgery,North American Chapter. 0741-5214/96/$5.00 + 0 24/6/72502
while trying to preserve as much function as possible. Occasionally the malignancy will involve the carotid artery. This poses a special problem, because resection of the carotid artery (with or without replacement) puts the patient at additional risk for the potential complications o f stroke, graft infection, or graft blowout. In the past, these complications occurred frequently enough to dissuade many surgeons from undertaldng cn-bloc resections that involved the carotid artery. However, we have always believed that the best way to manage advanced head and neck tumors invading the carotid artery is to use a team approach that includes vascular surgeons and oto775
776
JOURNALOFVASCULARSURGERY May 1996
Wright et al.
laryngologists. We have not used the balloon occlusion test routinely to try and predict which patients will be better treated with carotid ligation alone, because in our experience too many false-negative test results occur, and the complication rate from carotid replacement is so low that we simply try and replace the carotid artery in all patients who need carotid artery resection. Autogenous saphenous vein is our graft conduit of choice, but other materials may be used. Graft patency and healing require appropriate concomitant soft-tissue coverage of any overlying mucosal or cutaneous defects. This study reviews the outcomes in 20 consecutive patients treated between 1989 and 1995 at The Ohio State University with en bloc resection and replacement of the carotid artery for head and neck cancer. Some controversy exists regarding the value of carotid artery resection and replacement with respect to its impact on the long-term survival of patients with advanced head and neck malignancies; however, most authors have suggested that carotid resection does not increase the long-term survival? In a recent meta-analysis and review of the pertinent literature, Snyderman and D'Amico 2 concluded that the impact of carotid resection on long-term survival remains unproven, but it does improve local/regional control of the disease. We would agree with their conclusions. Our minimum goal in resecting and replacing the carotid artery has always been to achieve better local/regional control without subjecting the patient to undue risk of postoperative stroke regardless of any survival benefit it may or may not provide. We believe that ours has been a rational approach, because most series reviewing the results ofunselected carotid ligations (i.e., without revascularization) report a 15% to 30% incidence of postoperative stroke and a 3% to 5% mortality specifically resulting from strokes. Therefore reduction of stroke risk should be an important goal for any palliative procedure that might require carotid artery resection, even though the putative survival benefit of carotid resection remains unproven. Many investigators have hoped that patients can be selected for a "safe" ligation of the internal carotid artery (ICA), if some type of temporary balloon occlusion test (BOT) could reliably predict that permanent occlusion of the internal carotid artery would not result in a stroke. Unfortunately, no one has demonstrated that any form of BOT (either by clinical examination alone or in conjunction with electroencephalography, transcranial Doppler scanning, or single photon emission computed tomography scanning) can reliably predict which
patients can undergo a "safe" ligation of the ICA. In most published series that have investigated the prcdictivc value of the BOT on the basis of clinical response alone, the stroke rates after a negative BOT result are approximately 30%, which is not significantly different from the stroke rate that would be expccted after unselected ligation alone. The predictive value of a negative BOT result can be increased slightly by adding some adjunctive test of cerebral blood flow (e.g., transcranial Doppler scanning, electroencephalography, or single photon emission computed tomography). In most published series a negative BOT result with some adjunctive test followed by carotid ligation carries a 10% stroke risk. We believe that even a 10% to 30% stroke rate is too high among patients who supposedly can have a "safe" ICA ligation and essentially renders the test useless, because we are able to achieve a major stroke rate of less than 5% by replacing the internal carotid arteries in all of these patients without any selection criteria. We believe that the reason that BOTs have been unsuccessful in predicting postoperative strokes is because most postoperative strokes do not occur immediately; this is because most postoperative strokes are due to the embolization of fresh clot that has recently formed in the ligated distal ICA stump. This opinion is based on our own unpublished experience with emergency ligations of the carotid artery for tumor erosion or penetrating trauma and is supported by the recent publication of Meleca and Marks, 3 who reported that six of eight of the strokes occurring after carotid resection had a delayed onset of this complication (typically on postoperative day 2 or 3). In addition, careful review of the literature demonstrates that at least half the reported strokes occurring after acute carotid ligation for tumor or trauma do not occur immediately but typically occur a few days later. It seems to us that the temporary occlusion of an internal carotid artery for 15 or 30 minutes is a poor way to predict what might happen on the second or third postoperative day. METHODS Patients were identified by manual comparison of a computer-generated list of all patients in the Ohio State University Cancer Registry with stage 3 or greater head and neck cancers who underwent a resection with another computer-generated list of all patients in the Vascular Surgery Registry who underwent a carotid artery procedure. The list was also compared with a list of all patients during this time period who were billed for a carotid artery procedure
JOURNAL OF VASCULAR SURGERY Volume 23, Number 5
and was found to be complete and comprehensive. All information was obtained from existing patient medical records or from information obtained during patient interviews. The information was abstracted onto data sheets and entered into a computer for analysis. Postoperative death and stroke were defined as occurring during the hospitalization for which the patient underwent the carotid resection and replacement. Average results are presented as the mean + SEM. RESULTS Patient population. From 1989 to 1995, 20 patients had resection of advanced head and neck tumors for which the internal or common carotid artery was resected and replaced electively. A total of 16 men and four women underwent operation. All the patients were white. The average age was 62.4 + 4.4 years (range, 20 to 78 years). Eight of the 20 patients were diabetic, 6 had significant coronary artery disease, and 5 were hypertensive. The patients reported in this series come from a larger cohort of 80 patients, 60 of whom did not undergo a carotid resection but had tumor close enough to their carotid artery that they were evaluated with a carotid angiogram. The mean survival for the 60 patients who did not undergo carotid resection and replacement was 18.1 months. The mean survival rate for the 20 patients who did undergo carotid resection and replacement was 19.2 months. This finding is not dissimilar from what has been reported in other series. We excluded from this analysis three patients who required carotid artery ligation to control exigent hemorrhage caused by tumor invasion. Also excluded from this analysis were three patients (early in the series) who had carotid artery ligation without an attempt to replace the carotid artery. We excluded these patients because the focus of this review was to analyze the septic, thrombotic, neurologic, and hemorrhagic complications of carotid artery resection and replacement with autogenous greater saphenous vein, and we were malting no attempt to compare resection and ligation alone with resection and replacement. Malignancies. The most common tumor type was squamous cell carcinoma (16, 80%), three patients had advanced thyroid cancer, and one patient had an advanced basal cell carcinoma. In 16 (80%) patients the cancer was recurrent. The primary tumors originated from the larynx (six), pyriform sinus (five), base of the tongue (two), nasopharynx (two), tonsillar fossae (one), and oropharynx (one). In three patients the primary tumor site was unlmown and could not be determined. All 16 patients with squa-
Wright et al.
777
mous cell carcinoma had a significant history of alcohol and cigarette consumption, and two of the remaining four patients also consumed significant amounts of alcohol and cigarettes. Six (30%) patients had contaminated necks on gross examination including two patients with necrotic tumor, two with active wound infections from previous biopsies, and two with draining fistulas. Preoperative evaluation. All patients had biopsyproven malignancies. Computed axial tomography (CAT) scans (18), magnetic resonance imaging studies (6), cerebral angiography (19), cervical ultrasonography4 studies (6), and BOTs (4) were routinely obtained when clinically indicated, but none of these tests was obtained in all patients. CAT scanning was the diagnostic modality that identified tumor invasion most often, but because it was also the test that was ordered most often, we cannot comment on the relative accuracy of CAT scanning versus magnetic resonance imaging ultrasonography or angiography in determining whether the carotid artery is involved with tumor. Indeed, it is our experience that when we suspect that the carotid might be involved (excluding those cases in which it is clearly involved or not), a clean tissue plane can be identified between the tumor and the carotid artery in at least half the cases. Because the Ohio State University serves as the primary referral center for central Ohio, many of these patients arrived with their preoperative evaluations complete or nearly complete. Preoperative treatment. In all 16 patients with recurrence a previous neck dissection was performed. Seven of these 16 patients had previously undergone some type of myocutaneous flap procedure. Preoperative irradiation was performed in 16 of the 20 patients, ~ and one patient had previously undergone brachytherapy with 1125 seed implantation. Operative treatment. The decision to resect the carotid artery en bloc with the tumor mass was made in the operating room by the otorhinolaryngologic surgeon. In all cases the tumor either encircled the carotid artery, or it was so densely adherent that it could not be safely removed. Before the carotid artery was clamped, the entire tumor mass was mobilized as much as possible. Heparin was used in all cases except one. The vein harvest site was kept isolated from the cervical operative site by using separate instruments and by changing gowns and gloves. Once a suitable length of greater saphenous vein was harvested, heparin was administered, clamps were applied, and the tumor mass and the involved segment of carotid artery were removed en bloc. The internal carotid artery was replaced by a greater saphenous vein in the
JOURNAL OF VASCULAR SURGERY
778
May 1996
Wright et al.
reversed (12, 60%) or nonreversed (8, 40%) orientation. When the vein was placed in the nonreversed orientation, a modified Mills valvulotome was used to cut the valves. Shunts were used in 18 (90%) of the cases. Typically, the entire extracranial and extrathoracic portion of the carotid artery was removed from a point 3 or 4 cm above the clavicle to a point 1 or 2 cm below the base of the skull. No special attempts were made to revascularize the external carotid artery, unless the carotid bifurcation and the external carotid were clearly free of any tumor involvement and the bypass could be safely terminated proximal to or distal from the bifurcation. In 18 (90%) of the cases the patient had some form of intraoperative contamination including tracheostomies, pharyngostomies, or fistulas. Half of the patients had intraoperative radiation therapy. New concomitant myocutaneous flaps were rotated over the carotid bypass in six (30%) patients. These included four deltopectoral flaps6 and two pectoralis major flaps. 7 In seven other patients flaps had already been fashioned during previous procedures. In one patient the carotid artery was covered with a skin graft, and in the remaining six patients the carotid artery was covered with local flaps. Postoperative management. Patients were given intravenous antibiotics for at least 24 hours or until all wounds were healing without signs of infection. Drains were left in and removed by the otorhinolaryngologic service, typically on postoperative day 2. In four of the patients a carotid duplex study was performed before discharge to determine graft patency. In the remaining 16 patients a carotid duplex scan was obtained sometime after discharge. In all cases except one the initial duple x study demonstrated graft patency. Patients were then monitored by both the otorhinolaryngologic and Vascular Surgery services with periodic carotid duplex studies. Early postoperative complications. None of the patients died during the early postoperative period. One patient had a major stroke that manifest on his emergence from anesthesia. This patient had undergone replacement of his carotid artery without the use ofheparin. On removal of the shunt, clot was noted at the end of the shunt, and poor back bleeding was seen. Therefore a no. 2 Fogarty catheter was passed distally until all retrievable clot was removed. Three patients had wound infections in the early postoperative period. Two of these were simple wound infections that resolved with prolonged antibiotics. One of these three was a phayngocutaneous fistula, which closed spontaneously while the patient was an outpatient. No other patients had any central
neurologic deficit during their hospitalization, nor did any graft occlusions, graft infections, or graft blowouts occur during the early postoperative phase. Late postoperative follow-up. One patient had a minor stroke i2 months after the operation that was manifest by contralateral hand and arm weakness that completely resolved during the subsequent 6 weeks. At the time he had his stroke, this patient underwent a carotid duplex scan, which was normal. Another patient had arterial hemorrhage from his contralateral neck dissection and myocutaneous muscle flap. This condition was later diagnosed as being a pharyngeocutaneous fistula with erosion into the common carotid artery but was on the side opposite our earlier carotid replacement. However, during the evaluation of this patient's hemorrhage, he underwent cerebral angiography, which demonstrated an asymptomatic occlusion of the carotid graft. A third patient had hemorrhage from his carotid graft approximately 1 month after discharge. This was the only carotid graft in this series that was covered with a sldn graft. The skin graft was used because the local flaps left approximately 1 square cm of the carotid graft uncovered. When this patient had exigent hemorrhage 1 month after the operation, he was treated with ligation of his carotid graft. He did not have any neurologic deficit. All other patients remained free of complications and had patent grafts after a mean follow-up of 15.9 + 7.8 months. In summary, no graft blowouts or strokes occurred as the result of septic complications. The only graft blowout that occurred did so after a skin graft coverage, and none of the remaining 19 patients who had their grafts covered with local flaps or myocutaneous rotation flaps had a graft blowout. Two (10%) patients had a stroke, one (5%) major stroke in the immediate postoperative period and one (5%) minor stroke without residual 1 year after the operation. DISCUSSION Elective carotid artery resection in the treatment of paticnts with advanced malignancies of the head and neck remains a controversial issue. Tumor invasion or adhcrencc to thc carotid artcry may require carotid resection to achieve control of disease. However, the significant risks of stroke and death and the low cure rate deter many surgeons from electively resecting the carotid artery, s This can result in some patients being denicd an appropriate operation and left exposed to the possibility of carotid rupture or other local tumor problems. Furthermore when thcse patients have a recurrent tumor involving the carotid artery, surgical treatment becomes more treacherous
JOURNAL OF VASCULAR SURGERY Volume 23, Number 5
because o f the previous neck dissection, irradiation therapy, myocutaneous flap rotation, and possible fistulae in the field. Unfortunately, few alternatives exist. Failure to fully resect the segment of the carotid involved with tumor may result in tumor recurrence or death o f carotid rupture .9 Several methods have been described to predict the likelihood o f stroke after carotid artery resection and ligation without replacement. These include intraoperative electroencephalography, carotid stump pressure monitoring, ocular plethysmography, somatosensory-evoked cortical potentials, and temporary balloon occlusion with or without xenon-enhanced computed tomography o f cerebral blood flow. Ehrenfeld et al.~0 concluded that a carotid stump pressure o f greater than 70 mm H g indicates adequacy o f collateral circulation sufficient to avoid a stroke after ligation. They also noted that intermediate carotid stump pressures (in the range of 55 to 69 m m Hg) represent an intermediate risk of stroke and that pressures less than 55 m m H g result in an unacceptable stroke rate. Others, however, have shown that carotid stump pressure is an inaccurate predictor o f subsequent neurologic compromise. 11 Ocular plethysmography as described by Mar tinez et al. 12 allows accurate and reliable prediction o f the adequacy o f collateral flow to compensate for carotid artery ligation. Atkinson et al.~3 used somatosensoryevoked cortical potentials in conjunction with carotid stump pressure and achieved a 17% stroke rate with ligation. Although this is one o f the lowest stroke rates after carotid ligation in the literature, it still seems unacceptably high to us. One o f the most recent advances in this area is the use by de Vries et al. ~4 of temporary balloon occlusion with xenon-enhanced computed tomography o f cerebral blood flow. This m e t h o d seems promising, although further studies are necessary to obtain statistically significant data. In our opinion none o f these methods has produced sufficiently accurate results to justify the conversion o f a patent carotid artery to an occluded carotid artery. The probable reason these methods are unreliable is simply that most strokes after carotid ligation do not occur immediately but usually occur from 1 to 5 days after the ligation. Therefore we should not expect that the effects o f acute temporary carotid artery occlusion can predict what will happen several days after permanent ligation. Furthermore the best series in the literature has a 17% stroke rate after simple ligation. This result does not compare very well with the current series and others that have also achieved significantly lower stroke rates with very acceptable septic and hemorrhagic complications.
Wright et al.
779
Therapeutic options for management o f head and neck tumors with invasion o f the carotid artery include irradiation t h e r a p y , ls-17 palliative carotid peeling, 18'19and resection with 2° 24 or without 2s 30 bypass grafting. A review o f the literature demonstrates a reported stroke rate o f 17% to 45% and a death rate o f 8% to 58% when treated with ligation alone, 2 and these rates seem to be significantly higher than when patients are treated with resection and replacement. Radiation therapy carries the additional risk o f radiation arteritis, 31 accelerated atherosclerosis, 3~ and delayed healing, 32 but might help achieve better local control o f the tumor. We have used both reversed and nonreversed orientations for the greater saphenous vein and have found advantages to both. Ifa significant difference is found in the diameter of the proximal and distal ends of the harvested veins (distal vein being larger), then we prefer to use the vein in the nonreversed orientation so we can match up the large end of the vein with the c o m m o n carotid artery and the small end o f the vein with the internal carotid artery. Because many o f these patients have their entire carotid artery resectcd from just above the clavicle to a point within 1 or 2 cm from the base o f the skull, the length o f greater saphenous vein is often 20 cm or more, and this often results in a considerable size differential between the proximal and distal ends of the vein. On the other hand, if no significant difference is found between the proximal and distal ends of the vein, then we prefer to use the vein in the reversed orientation to avoid the need for valve lysis. In our experience artcriography has not been especially useful in identifying tumor invasion o f the carotid artery. This has also been the experience o f Reilly et al.,1 who found that arteriography correctly identified tumor invasion o f the carotid artery in only 1 of 12 patients. In our uncontrolled experience CAT scans seem to be the most c o m m o n way that the surgeon is alerted to the possibility o f carotid involvement. Recent reports have suggested that ultrasonography might be the most accurate method for determining tumor invasion o f the carotid artery before operation. 33 We are not overly concerned with the preoperative diagnosis o f tumor invasion of the carotid artery, because ultimately this will always be a clinical judgment based on the findings o f the operating surgeon. CONCLUSIONS From these results we conclude that the use o f the greater saphenous vein to replace the internal carotid
780
Wright et al.
artery after e n bloc resection is n o t a t t e n d e d by a high rate o f infectious complications or graft b l o w o u t even in the presence o f intraoperative tracheopharyngeal c o n t a m i n a t i o n a n d that the greater saphenous vein is the c o n d u i t o f choice for replacing an internal carotid artery after cancer resections. W h e n c o m p a r i n g o u r results with the r e p o r t e d results in the literature after carotid resection a n d ligation w i t h o u t replacement, it is clear that carotid r e p l a c e m e n t has a m u c h lower risk o f postoperative stroke a n d death. I t is unlikely that the resection a n d r e p l a c e m e n t o f the carotid artery will achieve significantly better t u m o r c o n t r o l in these patients with already advanced malignancies, b u t further studies m i g h t prove this to be the case. REFERENCES
1. ReillyMK, Perry MO, Netterville JL, Meacham PW. Carotid artery replacement in conjunction with resection ofsquamous cell carcinoma of the neck: preliminary results. J Vasc Surg 1992;15:324-30. 2. Snyderman CH, D'Amico F. Outcome of carotid artery resection for neoplastic disease: a meta-analysis.Am J Otolaryngol 1992;13:373-80. 3. MelecaRJ, MarksSC. Carotidarteryresection for cancerofthe head and neck. Arch Otolaryngol 1994;120:974-8. 4. Zaragoza L, Sendra F, Soloano J, et al. Ultrasonography is more effective than computed tomography in excluding invasion of the carotid wall by cervical lympadenopathies. Eur J Radiol 1993;17:191-4. 5. Jones TR, Frusha JD. Carotid revascularizationafter cervical irradiation. South Med J 1986;79:1517-20. 6. BakamjianVY, Long M, Rigg B. Experiencewith the medially based deltopectoral flapsin reconstructive surgery of the head and neck. Br J Plast Surg 1971;24:174-83. 7. AriyanS. The pectoralis major myocutaneous flaps: a versatile flap for reconstruction in the head and neck. 1971;63:73-81. 8. Kenedy JT, Krause CJ, Loevy S. The importance of tumor attachment to the carotid artery.Arch Otolaryngo11977;103: 70-3. 9. Mares MD, Bruns MD, Keenan MJ. Carotid artery resection for head and neck cancer. Ann Otol Rhinol Laryngol 1992; 101:778-81. 10. EhrenfeldWK, Stoney RJ, WylieEJ. Relation of carotid stump pressure to safety of carotid artery ligation. Surgery 1983;93: 299-305. 11. KellyJJ, Callow AD, O'Donnell TF, et al. Failure of carotid stump pressures. Arch Surg 1979;114:1361-6. 12. Martinez SA, Oiler DW, Gee W, deFries HO. Electivecarotid artery resection. Arch Otolaryugol 1975;101:744-7. 13. Atkinson DP, Jacobs LA, WeaverAW. Elective carotid resection for squamous cell carcinomasof the head and neck. Am J Surg 1984;148:483-8.
JOURNALOF VASCULARSURGERY May 1996
14. De Vries EJ, Sekhar LN, Horton JA, et al. A new method to predict safe resection of the internal carotid artery. Laryngoscope 1990;100:85-8. 15. Fee W, Goff J, Dinat D, et al. Intraoperative iodine 125 implants. Their use in large tumors in the neck attached to the carotid artery. Arch Otolaryngol 1983;109:727-30. 16. Lee DJ, Liberman FZ, Park RI, Zinreigh ES. Intraoperative I-125 seed implantation of extensiverecurrent head and neck carcinomas, Radiology 1991;178:879-82. 17. Toita T, Nakano M, TakizawaY, et aL Intraoperative radiation therapy (IORT) for head and neck cancer. Int J Rad One Biol Plays 1994;30:i219-24. 18. Ketchum AS, Haye RC. Spontaneous carotid artery hemorrhage after head and neck surgery. Am J Surg 1965;110:64955. 19. Dibble DG, Gowen GF, Sheed DR. Observation on postoperative carotid hemorrhage. Am J Surg 1965;109:765-70. 20. Conley JJ. Free autogenous vein graft to the internal and common carotid arteries in the treatment of tumors of the neck. Ann Surg 1953;137:205-14. 21. ConleyJJ. Carotidarterysurgeryinthetreatmentoftumorsof the neck. Arch Otolaryngol 1957;65:437-46. 22. Reha AJ, Rongetti JR, Bisi R, Replacement of carotid arteries with prosthetic grafts. Arch Otolaryngol 1962;76:76-80. 23. Lore JM, Boulous JE. Resection and reconstruction of the carotid artery in metastatic squamous cell carcinoma. Am J Surg 1981;142:437-42. 24. Karam F, Schaefer S, Cherryholmes D, Dagher FJ. Carotid artery resection and replacement in patients with head and neck malignant tumors. J Cardiovas Surg 1990;31:697-701. 25. Watson WL, SilverstoneSM. Ligature of the common carotid artery in cancer of the head and neck. Ann Surg 1939;109: 127-32. 26. Moore O, Baker HW. Carotid artery ligation in surgery of the head and neck. Cancer 1955;8:712-26. 27. Gandhi K, Oppenheimer P. Emergency carotid ligation. Arch Otolaryngol 1962;75:451-6. 28. Moore OS, IGarlanM, Sigler L. Factors influencing the safety of carotid ligation. Am J Surg 1969;118:666-8. 29. Shumrick DA. Carotid artery rupture. Laryngoscope 1973; 83:1051-61. 30. Hefler KS, Strong EW. Carotid arterial hemorrhage after radical head and neck surgery. Am J Surg 1979;138:607-10. 31. Fonkalsrud EW, SanchezR, Zerubavi R, et al. Serialchangesin arterial structure following irradiation therapy. Surg Gynecol Obstet 1977;134:395-40. 32. Marcial-Rojas RA, Castro JK Irradiation injury to elastic arteries in the course of treatment of neoplastic disease. Ann Otolaryngol 1962;71:945-58. 33. Mann WJ, Beck A, Schreiber J, Manrer J, Amedee RG, Gluckmann JL. Ultrasonography of evaluation of the carotid artery in head and neck cancer. Laryngoscope 1994;104: 885-8.
Submitted Oct. 11, 1995; accepted Feb. 6, 1996.
JOURNAL OF VASCULARSURGERY Volume 23, Number 5
Wright et al.
781
DISCUSSION
Dr. R a m o n Berguer (Detroit, Mich.). The mortality and morbidity of head and neck tumors that invade the carotid artery has changed dramatically since the introduction o f en-bloc carotid resection and replacement with autogenous tissue. The mortality has dropped from 27% in the Sloan-Kettering series to 0% or 5% in the two series involved in this discussion. Among their 20 patients, the authors experienced two graft occlusions, two strokes, and one blowout requiring excision and ligation. In one of the strokes the patient had not received heparin, and the blowout occurred in a patient who had a skin graft coverage of a vein graft. Both problems probably could have been avoided. Before 1991 we replaced carotid arteries with saphenous vein in the manner described by the authors. Sometimes we placed saphenous veins that were small and posed a problem with size mismatch. We had two saphenous grafts occlude shortly after their placement. In 1991 we adopted a protocol for carotid replacement using superficial femoral artery as an autograft while replacing the femoral segment with a prosthesis of polytetrafluoroethylene (PTFE). Our experience with this protocol was presented to this society last year. The authors' and our series are comparable in numbers, age, gender distribution, length of follow-up, and incidence of previous radical neck dissection. O f the 24 patients in whom we did carotid arterial autograft replacement, the only neurologic complication noted was in one patient who had a stroke when his carotid graft was avulsed 6 hours after operation as the surgeons responsible for the myocutaneous flap were attempting to control bleeding from a muscular vessel. The carotid artery was ligated, and 24 hours later the patient had a massive stroke and eventually died. The 23 surviving patients had patent grafts and no strokes at a mean follow-up o f l 1 months. The other complications were at the site of the donor artery: a silent thrombosis of the PTFE in the femoral location of one patient and an infected false aneurysm of the PTFE proximal anastomosis requiring ligation in another. In both cases the legs remained viable. These numbers are too small to statisticallyconclude the superiority o f one method over the other. It can be said, however, that because arterial autografts do as well or better than saphenous vein grafts, that one should consider them a good choice if the preoperative duplex shows a saphenous vein of inadequate size or quality. We have used twice a saphenous vein graft in patients that showed atheroma of the superficial artery in the preoperative duplex. We will continue to use autogenous artery as our first choice until we accrue numbers that will support or negate this policy, and we will follow with interest the series of Dr. Wright and associates. J. G o r d o n Wright. I would agree with your comments entirely. I think that the superficial femoral artery is a good alternate to use in this situation, and clearly the number of cases in our series precludes a valid a statistical comparison between superior femoral artery and greater saphenous vein
at this point. I think in your hands, you have had good results with the superficial femoral artery, whereas we have had good results with the greater saphenous vein. I think the important take-home message from this study is that if you are faced with a malignant invasion of the carotid, you should not simply ligate and resect; you should make every reasonable attempt to revascularize because, as you mentioned, the mortality and morbidity has dropped significantly by adopting a policy of trying to revascularize all of these. I cannot think of a reason a superficial femoral artery would be superior except as you have mentioned the theoretic possibilities o f less infectious complications, although we did not see any infectious complications in this particular series. Dr. Joseph J. Piotrowski (Cleveland, Ohio). I did a number of these resections and replacements with saphenous vein when I was at the University of Arizona. Similar to your experience, most of our patients had had previous neck resections and postoperative irradiation. Because of this, we tunneled our vein grafts into the posterior triangle. I wonder if the single case of blowout that you experienced could have been prevented by an alternate routing of your vein graft, because you already were considering covering this with a somewhat scary skin graft. Dr. Wright. There is no doubt about that. This blowout was clearly due to an inadequate soft-tissue coverage, and it could have been prevented by the means you had suggested or by simply rotating some kind ofmyocutaneous flap. Dr. William D. Turnipseed (Madison, Wis.). I think your clinical series would suggest that some, but not all, of these patients require carotid reconstruction. This is born out by the fact that you have lost graft function in a couple of patients with no clinical consequence. If you could document before surgery an adequate intracranial collateral system, do you think it might simplify your surgical approach and eliminate the need for this algorithm in selected patients? Dr. Wright. If there were such a test, it would simplify things. However, we do not like the balloon occlusion test because we have had some patients who have had a negative balloon occlusion test who were ligated and subsequently stroked. Furthermore, our involvement with these cases is usually approximately 45 minutes, and many of these procedures take 6 or 8 hours. I am not sure that replacing the carotid in these large head and neck resections is really adding significantly more complexity to an already complex operation. Dr. Turnipseed. You might want to look into the use of magnetic resonance angiography and Diamox chemical stress testing. Magnetic resonance angiography can clearly define the anatomy of the circle of Willis. Magnetic resonance angiography can also be used to evaluate blood flow
JOURNAL OF VASCULARSURGERY
782
Wright et al.
May 1996
through watershed areas within the cerebral circulation before and after vasodilation with Diamox. This may prove to be an accurate preoperative screening test for these patients. Dr. Wright. One other point that is emphasized in the article is that these tests are all acute tests, and in this series (and others) most patients who have stroke after ligation do not have stroke immediately but usually stroke on the second or third day after ligation. Dr. Donald P. Spadone (Columbia, Mo.). I would have you entertain another hypothesis for the cause of stroke rather than inadequate intracerebral collaterals. It is possible that with ligation there is propagation ofthrombus within the ligated internal carotid and propagation or
embolization ofthrombus produces stroke 2 or 3 days after ligation. Is there evidence to support propagation ofthrombus or embolization in patients who did have stroke? Dr. Wright. Yes, that is exactly what I think happens with most of these patients. There is no objective evidence to prove or disprove that pathophysiology except as suggested by their clinical course. The fact that they usually do not have stroke immediately, that they usually have stroke on the second, third, or fourth postoperative day suggests that it is the propagation of thrombus into the circle of Willis, which embolizes and causes the stroke. Once again, this is why we do not like the balloon occlusion test or any other test of acute temporary occlusion.
T_HE MO ve? 0enaus, D o n ' t m i s s a s i n g l e i s s u e of t h e j o u r n a l ! To e n s u r e p r o m p t s e r v i c e w h e n y o u c h a n g e y o u r a d d r e s s , p l e a s e p h o t o c o p y a n d c o m p l e t e t h e f o r m below.
Please send your change of address notification at least six weeks before your move to ensure continued service. We regret we cannot guarantee replacement of issues missed due to late notification. J O U R N A L TITLE: Fill in the title of the journal here.
O L D ADDRESS:
N E W ADDRESS:
Affix the address label from a recent issue of the journal here.
Clearly print your new address here.
Name Address City/State/ZIP
COPY AND MAIL THIS FORM TO: Journal S u b s c r i p t i o n Services M o s b y - Y e a r Book, Inc. 11830 Westline I n d u s t r i a l Dr. St. Louis, M O 63146-3318
O R FAX T O : 314-432-1158
Mosby
O R PHONE: 1-800-453-4351 OutsidetheU.S.,call 314-453-4351