Carotid arterial trauma: assessment with the Glasgow Coma Scale (GCS) as a guide to surgical management1

Cardiovascular Surgery, Vol. 5, No. 2, pp.196–200, 1997 (0 1997 The International Society for Cardiovascular Surgery Published hyElsevier Science Ltd. Printed in Great Britain 0967–2109/97$17.00 +0.00

E SCIENCEg

Carotidarterialtrauma:assessmentwiththe GlasgowComaScale(GCS)as a guideto surgical management E. I? Teehan,F.T. PadbergJr, 1?N. Thompson, B. C. Lee, M. SilvaJr, Z. Jamil, K. G. Swan and R. W. Hobson II University of Medicine and Dentistry of New Jersey New Jersey Medical School, and University Hospital, Depatiment of.%rgery Section of VascularSurgery Newark, NJ07103, USA

Managementof carotid arterial injuriesassociated with focalneurologicaldeficitor altered state of consciousness (SCON)remains unresolved. Experience with these injuries in one particular hospital was reviewed and the Glasgow Coma Scale (GCS) utilized to assist with clinical stratificationof these patients. A literature reviewwas also conductedto better defineindications for repair or ligation of carotid injuries. From 1978 to 1990, 34 patients with carotid arterial injuries were reviewed with reference to the GCS,focal deficit, hypotension, anatomic site and mechanism of injury.The literature from 19S2 to 1993 was surveyedfor carotid artery injuries (13 16 patients). Outcomeof treatment with or without repair was compared with pre-operative necrologicstatus. Thirty-fourpatients with injuriesof the common (24) or internal (1O) carotid arteries were managed with repair (68%), ligation (24%) or observation (9Yo). The SCONwas normal in 18 patients; 16 patients (88Yo)underwent repair and allremained normal.Allpatients with GCS9–14 regained a normalSCONafter surgicalrepair,while 10 patients with GCSe 8 had repair (S), ligation(3), and non-operativemanagement (2); fivereturned to normal,four died and one remained comatose. However, outcomes correlated poorly with management. Of 1316 patients cited in the surgical literature, patients with no deficitand patients with pre-operative deficits did significantlybetter after repair as compared with ligation (Pc 0.001). In comatose patients, managementdidnot affectoutcome. It is concludedthat carotidarterialinjuriesshouldbe repairedinpatients with normalnecrologicevaluation,focalpre-operativenecrologicdeficitsand in patients with GCS>9. Comatosepatients with GCS<8 do poorlyregardless of management. The GCSprovidesan objectivefor stratificationofpatients with altered SCONwho benefitfrom repairof carotid arterial injuries.01997 The InternationalSocietyfor CardiovascularSurgery. Keywords: carotid artery trauma, GalsgowComaScale

The appropriate management of traumatic carotid arterial injuries continues to present a complex surgical challenge. Associated injuries outside the neck are uncommon; morbidity and mortality are frequently Presented at the Seventh Annual Meeting of the Eastern Vascular Society,held at the Sheraton Society Hill Hotel, Philadelphia,PA, USA on 30 April 1993 Correspondence to: Dr R. W. Hobson II, Section of Vascular Surgery, Medical Science Building G-5.32, 185 S. Orange Ave., University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark, NJ 07103, USA No corrected proof was received from the author for this pape~

196

related to associated hemorrhage, shock and stroke. Since outcome has been correlated with pre-operative necrologic status, improved results from treatment of these injuries is likely to occur with this group. In particular, managementof the arterial injury in patients with altered state of consciousness remains controversial. Traumatic carotid arterial injury may result in focal or global necrologic deficits. Revascularization of patients with preoperative focal deficits has been reported to give better results than simple Iigationl’2. Since repair of the arterial injury in the comatose patient has been associated with limited benefit, some authors

CARDIOVASCULAR SURGERY APRIL 1997 VOL S NO r?

GlasgowComaScale:E. P Teehanet al.

have advocated ligation of the injury rather than ~rimary repairs-~. Although some patients maydoWell, it is difficult to determine which patients with altered state of consciousness (SCON) will improve with revascularization. Previous reports evaluating outcome in patients with deficits have not clearly distinguished between those with Iateralizing deficits, altered mental status and coma. Additionally, coma has not been uniformly defined in these series. The Glasgow Coma Scale (GCS) was developedto provide a standardizedmeasureof the severityof the patient’saltered levelof consciousness. It allows an objective and reproducible assessment of the depth of impaired consciousness and coma. This report analyses the clinical experience with carotid arterial injuries in 34 patients at one particular hospital. The GCS was used to stratify patients for developmentof specific clinical managementguidelines. In addition, a review of the surgical literature was conducted to complement this analysis.

Table 1 The GlasgowComaScale(GCS) Eye opening Verbal response

Motor response

Total score

Never To pain To verbal stimuli Spontaneously No response Incomprehensiblesounds Inappropriate words Disorientedand converses Oriented and converses No response Extension(decerebraterigidity) Flexion (decorticaterigidity) Flexionwithdrawal Localizesto pain Obeys

1 2 3 4 5 1 2 3 4 5 6 3-15

The best response from each category is determined by its corresponding numericalscore.The resultsfrom the three categoriesare then added together to give the overall GlasgowComa Score (normal, 15: moderatelydiminished SCON,9–14: severelydiminished SCON,s8)

Patients and methods For the period from 1978 to 1990, clinical records of 34 consecutive patients with penetrating injuries to the common and internal carotid arteries were reviewed. Hospital records were reviewed retrospectively for the following information: mechanism and anatomical distribution of injuries, haemodynamic stability, necrologic status, associated injuries, surgical management, morbidity, and mortality associated with the carotid arterial injury. Immediately following resuscitation, patients were stratified as normal, exhibiting the presence of a focal deficit or possessing an altered SCON. The GCS was applied to objectively define the magnitude of coma. The GCS employs a 15-point scale to assess necrologic responsiveness in three categories: eye opening, motor response and verbal responseG.The results from each category were addedtogether to give a total score (T’uble 1). A GCS of 15 was considered normal; a GCS of 9-14

was a moderately diminishedstate of consciousness;and a GCS of less than 8 was a severelydiminished state of consciousness. Hypotension was defined as a systolic blood pressure <90 mmHg on admission. At the authors’ institution, cervical wounds thought to penetrate the platysma muscle layer are routinely explored. Operative management involvedeither vascular repair or ligation. Injuries that occurred above the angle of the mandible (Zone III) were defined as high injuries’. The data were analysed with the Statistical Analysis Systemsoftware. Significancewas definedas 1’<0.05. In the authors’ series of patients the following factors were assessed: influence of hypotension on admission, anatomic location of injury and neurological status on admission (normal, neurological deficit or coma). All comparisons were tested using the two-tailed Fisher’s

CARDIOVASCUIJ4R SURGERY APRIL 1997 VOL S NO 2

Exact test. In the reviewof the literature all comparisons were tested using the Wilcoxon rank sum test.

Results Penetrating injuries of the common or internal carotid arteries were observed in 34 patients. Thirty patients were male (88 YO) and four were female; the mean age was 29 years. The distribution of injuries was 19 right and 15 left. The common carotid artery (CCA) was injured in 24 cases and the internal carotid artery (ICA) in 10. A majority of these injuries were from gunshot wounds (23), followed by stab wounds (8) and shotgun wounds (3). Compared to our previous reportss, the proportion of injuries due to firearms has increased from 53 to 76Y0. Likewise, as assailants become more remote, the proportion of left-sided injuries has decreased from 67 to 44~0. Six deaths resultedin a mortality rate of 17Y0. Four of these patients presented with coma and hypotension and two patients with focal deficits which progressedto coma. There was a significant correlation betweencoma and death (P c 0.05). Hypotension was present in 17 patients (50’%0).There were four deaths in this group. Three patients were comatose, in two the injuries were repaired, one was treated with ligation; post-operatively all remained comatose and expired. The fourth patient expired in the emergency room prior to exploration of the neck. Hypotension on admission did not correlate with postoperative necrologic deficit or death (F’>0.05). Associated injuries were common due to the density of important structures in the neck. Venous injuries occurred in six patients and were managed by lateral venorrhaphy(3) and ligation (3). Five patients sustained

197

GlasgowComaScale:E. /? Teehanet al.

mandibular fractures. There were three oesophageal injuries. Three patients sustained fractures of cervical vertebrae; one patient who sustained a fracture of C3 presented with coma, underwent carotid arterial ligation, and subsequentlydied. A second patient presented with a fracture of the seventhcervical vertebra, resulting in quadriplegia.The third patient also had a fracture of C7 which resulted in a sensory deficit without paralysis. Other neural injuries included brachial plexus (2), cranial nerves (5), peripheral nerve (1) and phrenic nerve (1). One tracheal injury required tracheotomy. One patient presented with blindness secondary to bilateral ruptured globes from shotgun pellets. Surgical management is summarized in Table 2. Injuries were managed by vascular repair in 23 cases (67%), carotid arterial ligation in eight (25%) and observation in three (8Y0). Vascular repair included saphenous vein interprimary repair in 16 (70~o), position graft in five (22%) and vein patch angioplasty in two (80/0). Carotid ligation was performed in eight patients. Four Iigations were performed because distal ICA injuries were inaccessible and because of expansion of an inaccessible pseudoaneurysmin one other patient. Two patients underwent ligation because of an associated necrologic deficit. One artery was ligated because of a technically difficult repair in a patient who had two intra-operative cardiac arrests. Two patients were managed non-operatively.Both had high injuries and underwent arteriography. One patient had a focal necrologic deficit and was found to have a carotid-}cavernous sinus fistula which was subsequently treated by balloon occlusion and the other refused intervention for a common carotid pseudoaneurysm. A normal state of consciousness (GCS 15) was present on admission in 17 patients. They were managed with repair (82%), ligation (12%) and observation (6Y0). All remained neurologically stable and recovered. A diminished state of consciousness was present in 14 patients, who were separated into two groups based on the GCS score. Ten patients had a severely diminished SCON (GCS < 8) and four patients had a moderately diminished SCON (GCS 9–14). Management of the 10 patients with severe reductions in SCON consisted of repair (5), ligation (3) and observation (l).

Table 2

Surgical managementof carotid artery injuries

Treatment

CCA

ICA

Total

Vascularrepair Primary repair Vein interposition Patchangioplasty Ligation Observation

19 12 5 2 2 3

4 4

23

0 0 6

0

CCA,common carotid artery; [CA,internal carotid artery

198

8 3

Five of the ten regained a normal mental state, four died and one remained in coma. In the subgroup of patients with a moderately diminished SCON, two underwentrepair and two underwent ligation. All four patients regained a normal mental state following surgical treatment. One patient who underwentligation developed a hemiparesis post-operatively. In three patients the SCON by GCS could not be determined because they required incubation. Focal neurological deficits were present in nine patients. Of these nine, four patients underwent repair, four ligation and one endovascular balloon occlusion of a carotid-cavernous fistula. Following repair two patients improved, one worsened and one progressedto coma and died. After ligation two patients improved and two progressed to coma and died.

Reviewof the literature The English languageliterature from 1952 to 1993 was surveyed for papers concerned with vascular trauma, penetrating wounds of the neck or carotid injuries. An earlier seriesl summarized the literature from 1952 to 1979. Articles from 1980 to 19939-22 were acceoted for , inclusion, provided information on patients’ demographic data, necrologic status and hospital course were included. Those cases associated with head injury or other significant injury which may have influenced patient outcome were not included. These results, including the authors’ 34 cases, were combined in an effort to relate operative management to the preoperative necrologic state. The collected series totalled 1316 cases of carotid arterial injury — 695 patients had a normal necrologic examination prior to operation, of which 625 underwent arterial repair. Of these patients, 581 remained normal while 44 became worse. Of seventypatients who were not repaired, 55 remainednormal while 15 became worse. A favorable outcome was associated with carotid artery repair (1’< 0.001) (Figure 1). Three hundred and fifty patien~s presented with a severeneurological deficit. This was defined as aphasia, cortical motor paralysis, profound central sensory deficit or coma which was present after adequate resuscitation. Two hundredand forty-sevenpatients had vascular repair and 45°A improved, which was significantly better (P c 0.001 ) than the 200/0that improved without repair. only 30~0 of patients who were revascularized became worse, while those undergoing ligation or expectant management deteriorated in-55% of cases (P< 0.001) (Figure 2). Ninety-eight of the patients had a pre-operative necrologic status of coma. Arterial repair was performed in 57 patients, of which 36% improved while 46Y0 died. Forty-one patients did not undergo repair; 21% improved and 66% died. Although there seemed to be a trend favouring repair, no significant difference (F’= 0.13) existed between the two forms of treatment.

CARDIOVASCULAR SURGERY APRIL 1997 VOL 5 NO 2

GlasgowComaScale:E. L?Teehanet al.

attempted to correlate the incidence of shock with stroke. However, neither reports by Thai et U1.3 nor Liekweg and Greenfield2were able to confirm a direct relationship. Unger et al.l, while unable to show a difference in their own patients, report a significant differencein their reviewof the literature up to 1979. In the authors’ series, a correlation between shock and post-operative necrologic deficit or death could not be established. Conversely,Ledgerwood et al.13 and Fabian et al 10have reported that shock increased the incidence

NORMAL NECROLOGIC EXAM / 10WO

80’%0 60Y0

40% 20% o%

/ Repair

No Repair

HNormal(n=625)

■ Worse(n=

70)

p
ABNORMAL

NECROLOGIC EXAM *

60%

I

50’YO 40Y0 30% 20Y0 lo’% o% Improved

❑ Repair(n=

Unchanged 242)

■ No

Worse

Repair(n=98)

●p
Figure 2

Discussion Management of carotid arterial injuries remains complex and continues to generate controversy. Reduction in cerebral blood flow frequently leads to major necrologic morbidity and/ordeath, with several recent series reporting mortality rates of 10–22Y08-*2’23. Shock, location of arterial injury and ischaemic time have been implicated as adverse influences on necrologic outcomez,10,24,25. Several prior reviews have

CARDIOVASCULAR SURGERY APRIL 1997 VOLS NO 2

of adverse necrologic outcome. Although one might expect that the anatomic site of injury might influence outcome, this could not be demonstrated in the present series. It has been theorized that patients with common carotid injuries may do better than those with internal carotid injuries because of the potential for collateral circulation from the external to the internal carotid artery13. Demetraides et al.24 reported a significantly greater mortality rate in common carotid injuries as compared to internal carotid injuries. However, no fatal injuries resulted from internal carotid injuries in the present series. For patients with no pre-operative necrologic deficit, most surgeons would agree that, if technically feasible, the carotid artery should be repaired. This is supported by the authors’ report, in which over 80?’owere repaired and all remained normal. The literature review also confirmed this observation. In 695 patients, surgical repair was associated with a significantly improved outcome. For patients with necrologic deficits resulting from carotid arterial injuries, optimal management has been controversial. Bradley5 reported a 50Y0 mortality following revascularization in 10 patients with pre-operative necrologic deficits. He described hemorrhagic cerebral infarction following revascularization and recommended ligation for patients with pre-operative necrologic deficits. In the authors’ series, two patients deteriorated following repair; however, cerebral oedema, not hemorrhagic infarction, was noted. Liekweg and Greenfield2 recommended revascularization for patients with pre-operative necrologic deficits but not for patients with coma. Based upon their review of the literature Unger et aL1 also noted a significant difference favouring revascularization in patients with pre-operativenecrologic deficits. In the present authors’ literature review, 252 patients with pre-operative focal necrologic deficits were identified. A highly significant difference favouring repair was demonstrated in this group of patients. For the comatose patient, data from the authors’ seriesas well as the literature reviewconfirm that results remain poor regardlessof the surgical treatment. Several authors have advocated ligation for comatose others have recommended rePair patients3–5 . However, in these patients9’16’~7’24. A review by Liekweg and Greenfield2 recommended repair in the comatose

199

GlasgowComaScale:E. P Teehanet al. patient if prograde flow could be demonstrated. Unger et al.l concluded that the available data were too limited to make a definitive recommendation for patients with coma. The current review showed a trend which appeared to favour repair, with S7’%0 improved versus

21% with no repair, and with 46?4. mortality versus 66Y0 with no repair. However, these apparent differences were not statistically significant. In previousreports of carotid arterial injury,coma has been a subjective description of the patient’s condition which has not been defined uniformlyl’2’8-11. The authors employed the Glasgow Coma Scale in order to assess objectively the patients’ necrologic responsiveness and relate this to treatment options. The authors recommend that future reports include an estimation of the level of consciousness in their assessmentof patients with this injury. A subgroup of patients with a moderately diminished level of consciousness appeared to do better with revascularization. For the clinician faced with a carotid artery injury and a depressed level of consciousness, a more aggressive approach to revascularization may be warranted. The present authors agree with Blaisde1127that a multicentre prospective randomized trial is appropriate to determine the optimal treatment for the comatose patient. Patients with no necrologic deficit have had uniformly good results following repair of the injured carotid artery. Patients with preoperative necrologic deficits in the absence of coma have done significantly better with repair than with carotid ligation. Patients with severely impaired consciousness (GCS < 8) do poorly regardless of surgical management. Although limited by small numbers, an aggressive approach to repair of carotid arterial injuries in patients with GCS29 seemswarranted. The authors recommend patient stratification for repair of carotid arterial injuries using GCS and standard necrologic examination.

References 1. Unger S, Tucker W, Mrdeza M, Wellons H, Chandler J. Carotid artery trauma. Surgery 1980; 87: 477–87. 2. Liekweg W, Greenfield L. Management of penetrating carotid arterial injury. Am Surg 1978; 188: 587–92. 3. Thai E, Snyder W, Hays R, Perry M. Management of carotid artery injuries. Surgery 1974; 76: 955–60. 4. Cohen A, Brief D, Mathewson C. Carotid artery injuries. AWI J Swg 1970; 120: 210-14.

200

5. Bradley E. Management of penetrating carotid injuries: an alternative approach. J Trauma 1973; 13: 248–55. 6. Teasdale G, Jennett B. Assessment of coma and impaired consciousness, a practical scale. Lancet 1974; 2: 81–3. 7. Thai E. Injury to the neck. In: Moore E, Mattox K, Feliciano D, eds. Trauma. Norwalk, CT: Appleton & Lange, 1991: 306. 8. Padberg F, Hobson R, Yeager R, Lynch T. Penetrating carotid arterial trauma. Am Surg 1984; 50: 277–82. 9. Brown M, Graham J, Feliciano D, Mattox K, Bean A, DeBakey M. Carotid artery injuries. Am] Szwg1982; 144: 748-53. 10. Fabian T, George S, Croce M, Mangiante E, Voeller G, Kudsk K. Carotid artery trauma: Management based on mechanism of injury.J Trauma 1990; 30: 953–63. 11. Jebara V, Tabet G, AshoushR et al. Penetratingcarotid injuries— a wartime experience.J Vase .Mrg 1991; 14: 117–20. 12. Meyer J, Walsh J, Barrett J et al. Analysis of 18 recent cases of penetratinginjuries to the common and internal carotid arteries. Am J Surg 1988; 156: 96-9. 13. LedgerwoodA, Mullins R, Lucas C. Primaryrepair vs ligation for carotid artery injuries. Arch Swg 1980; 115: 488–92. 14. Richardson R, Obeid F, Richardson D, Hoyt D, Wisner D, Gomez G, Johanson K, McSwain N, Weigek J, Blaisdell F. Necrologic consequences of cerebrovascular injury. J Trauma 1992; 32: 755-60. 15. Moses B, Tolson M, Boyd C. Fifteen year review of extracranial carotid injury.J Med Assoc Georgia 1987; 76: 129–32. 16. Weaver F, Yellin A, Wagner W, Brooks S, Weaver A, Milford M. The role of arterial reconstruction in penetratingcarotid injuries. Arch Surg 1988; 123: 1106-11. 17. Robbs J, Rajaruthnam P, Duncan H, Vawda I, Baker L. Neurological deficit and injuries involvingthe neck arteries. BrJ Swg 1983; 70(4): 220-2. 18. Timberlake G, Rice J, Kerstein M, Rush D, McSwain N. Penetratinginjury to the carotid artery. A reappraisalof management. Am Srwg1989; 55: 154–7. 19. Khoury G, Haji H, Khoury S, Basil A, Speir R. Penetratingtrauma to the carotid vessels.J Vase Swg 1990; 4: 607–10. 20. RichardsonJ, Simpson C, Miller F. Management of carotid artery trauma. Surgery 1988; 104: 673–80. 21. Maravasti M, Parker F, Bredenberg C. Injuries to arterial branches of the aortic arch. Tborac Cardiouasc .%rg 1984; 32: 293-8. 22. Burruss G, Mangiante E, Fabian T. Penetrating carotid artery injuries. J Ten Med Assoc 1986; 79: 492–3. 23. Fry R, Fry W. Extracranial carotid artery injuries. Surgery 1980; 80: 581–6. 24. Demetraides D, Skalkides J, Sofianos C, Melissas J, Franklin J. Carotid artery injuries: Experience with 124 cases. J Trauma 1989; 29: 91-4. 25. Rubio P, Reul G, Bean A, Jordan G, DeBakey M. Acute carotid artery injury: 25 years’ experience. J Trauma 1974; 14: 967-73. 26. Karlin R, Marks C. Extracranial carotid artery injury. Current surgical management. Am J Stwg 1983; 146: 225–7. 27. Blaisdell F. Cerebrovascular injury — necrologic consequences. J Trauma 1991; 31(2): 1024. Paper accepted 25 May 1995

CARDIOVASCULAR SURGERY APRIL 1997 VOL 5 NO 2