An Experience with Upper-Extremity Vascular Trauma Dolores F. Cikrit, MD, Michael C. Dalsing, MD, Benjamin J. Bryant, ~D, Stephen G. Lalka, MD, Alan P. Sawchuk, ~D, Jeffrey E. Schulz, M~, Indianapolis,Indiana In this 6-year study of 101 limbs requiring surgical intervention for upper-extremity vascular trauma, most patients were male, young, and injured by penetrating objects. Injured vessels included 13 axillary/subclavian, 23 brachial, 40 radial, and 25 ulnar arteries. Concomitant injuries included nerve injury in 50 cases, tendon laceration in 29, and bony fracture in 1 |. Arterial repair was accomplished by primary repair in 54 limbs, vein graft in 26 limbs, and vein patch in 3 limbs. Seventeen arterial injuries were ligated. Ancillary procedures included 30 nerve or 27 tendon repairs. The limb salvage rate was 99%. No functional deficits were noted in those cases with only a vascular injury. In 64% and 25% of patients with nerve or musculoskeletal injury, respectively, the arm was functionally impaired. Prompt diagnosis and surgical intervention eliminate vascular injury as a factor in upper-extremity limb loss or disability. Functional deficits are the result of nerve or orthopedic injuries.
" pper-extremity vascular trauma constitutes nearly half of all extremity vascular trauma in the United U States [1]. The majority of injuries are of a violent penetrating nature. Since these injuries generally occur in young males, the social and economic implications can be significant, resulting in marked morbidity from poor functional outcomes or amputation. Although the collateral circulation to the arm is extensive, vascular insufficiency can still occur at the time of injury or at a later date from thrombosis, embolization, or rupture of an unrecognized vascular injury. This study was undertaken in an effort to better define these injuries and to help decrease the disability that can be associated with upperextremity vascular trauma. From the Department of Surgery, Indiana University Medical Center, Indianapolis, Indiana. Requests for reprints should be addressed to Dolores F. Cikrit, MD, Department of Surgery, Indiana University Medical Center, 1001 West 10th Street, Indianapolis, Indiana 46202. Presented at the 18th Annual Meeting of the Society for Clinical Vascular Surgery, Palm Desert, California, March 7-11, 1990.
PATIENTS AND M E T H O D S In this retrospective study, 329 consecutive patients were evaluated for upper-extremity vascular trauma at Indiana University Medical Center Trauma Unit at Wishard Memorial Hospital, a county hospital which is one of two major trauma centers in the city of Indianapolis. Iatrogenic injuries were excluded from this study. From January 1983 through August 1989, vascular injury was excluded angiographically in 229 patients. The remaining 100 patients (101 limbs) required surgical intervention for their vascular injury and form the basis of this report. They were evaluated for presentation, physical findings, method of diagnosis, associated injuries, surgical intervention, postoperative course, functional assessment, and long-term follow-up. Most patients were male (79%) and young, with a mean age of 28 years (range: 5 to 89 years). Arterial injuries involved the axillary/subclavian (13 patients), brachial (22), radial (38), ulnar (25), brachial/radial (1), and radial/ulnar (2) arteries (Figure 1). This included one patient who had bilateral brachial artery injuries. Penetrating wounds accounted for 94% of the injuries while blunt trauma comprised the remainder. There were 59 lacerations, including 15 that were caused by glass and 7 that were self-inflicted in suicide attempts. Ninety percent of the lacerations involved the radial and/or ulnar arteries. The remainder of the penetrating injuries included 15 stab wounds, 12 gunshot wounds, and 7 shotgun blasts (Table I). Proximal vessel (subclavian/axillary/ braehial artery) injuries were usually (67% of the time) the result of stab, gunshot, or shotgun blast wounds, whereas distal arterial (radial/ulnar artery) injuries usually resulted from lacerations. The blunt trauma included two motor vehicle accidents, two elbow dislocations, one conveyor belt injury, one fall, and one supracondylar humeral fracture. All but one of the cases of blunt trauma involved proximal vessels. Twenty-nine patients presented with absent pulses, while 36 patients were found to have a palpable pulse despite having a major vascular injury. Distal pulses were absent in 31% of the limbs with an axillary/subclavian artery injury, 43% with a brachial artery injury, 24% with an ulnar artery injury, and 23% with a radial artery injury. Palpable radial or ulnar pulses were noted in 32% of patients with an axinary/subclavian injury, 35% of limbs with a brachial artery injury, 52% of patients with an ulnar artery injury, and 28% of limbs with a radial artery injury. An abnormal Allen test suggestive of vascular injury was noted in only six patients. The remaining 30 patients had diminished pulses. Arterial bleeding was noted in the emergency room in 45 patients (Table I). Bleeding was associated with 55% and 40%, respectively, of radial and ulnar artery injuries. Nearly half (46%) of
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160 AUGUST 1990
229
CIKRIT ET AL
, . ~ L N A R
RADIAL
/=/ BRACHIAL
AXILLARY SUBCLAVIAN Figure 1. Distal (radial and ulnar arteries) injuries comprised the majority of vessels injured with only 36% of injuries involving proximal upper-extremity vessels (subclavian, axillary, brachial arteries).
TABLE I
Types and Presentation of Injury In Upper-Extremity Vascular Trauma* Axillary/ Subclavian Brachial t Radialt Wounding force Laceration Stab Gunshot Shotgun blast Blunt Presentation Palpable pulses
Absent pulses Arterial bleeding Concomitant injury Total Limbs
Ulnar
Total
1 (8) 4 (31) 5 (38) 1 (8) 2 (15)
5 (22) 4 (17) 4 (17) 6 (26) 4 (17)
32 (80) 21 (84) 5 (13) 2 (8) 2 (5) 1 (4) 1 (3) 0 0 1 (4)
59 15 12 8 7
4 (31) 4 (31) 6 (48)
8 (35) 10 (43) 7 (30)
11 (28) 13 (52) g (23) 6 (24) 22 (55) 10 (40)
36 29 45
13
21
26
21
81
13
23
40
25
101
* Values in parenthesesare percentages. "t Includes one limb with both a brachial and radial artery injury. ~tIncludestwo limbs with both a radial and ulnar artery injury.
subclavian/axillary artery injuries presented with arterial bleeding. Only 20 patients had an isolated arterial injury. These isolated arterial injuries involved the radial or ulnar arteries in 18 patients and the brachial artery in the remaining 2. There were seven major venous injuries, which were all associated with proximal arterial injuries. Concomitant nonvascular injuries included nerve damage in 50 limbs, at least 1 tendon laceration or transection in 29 limbs, bony fracture in 11 limbs, and remote injuries in 17 patients (Table II). Nerve injuries included 41 transections or lacerations and 9 contusions, including 7 brachial plexus injuries. The remaining nerve injuries were divided nearly equally between the radial, ulnar, and median nerves. Of ulnar and radial artery injuries, 230
THE AMERICAN JOURNAL OF SURGERY
60% and 35%, respectively, had associated nerve injury. This was also the ease in 65% and 46%, respectively, of brachial and axiUary/subclavian artery injuries. The majority of patients with tendon lacerations had more than one tendon damaged, and these obviously all occurred in association with radial or ulnar artery injury. Sixty-three patients went directly from the shock room to the surgical suite because of arterial bleeding (31 patients), severe ischemia (20), or localized injury that required simultaneous tendon or nerve repair (12). Biplanar angiography in the radiology department was utilized in 14 patients for injury in close proximity to a major vessel and in 23 patients for localization of a clinically apparent injury. In patients who underwent angiography, there were 17 occlusions, 12 intimal flaps, 8 pseudoaneurysms, and 3 artedovenous fistulas noted (Table III). In all three patients with an arteriovenous fistula, a pseudoaneurysm was also noted. Three false-positive angiograms resulted in negative explorations. Only nine venograms were obtained, revealing extravasation in two patients, compression in three, and intimal flap in one; two were negative, although on exploration for arterial repair one of these had a vein wall defect noted and corrected. Arterial repair was accomplished by primary anastomosis in 54 vessels, interposition vein grafting in 26 vessels, and vein patch angioplasty in only 3 vessels. Two thirds of the distal vessels (ulnar or radial artery) were repaired primarily, which differed markedly from the more proximal vessel (subelavian/axillary/brachial artery) repairs, where less than one third of the vessels could be repaired primarily. Seventeen arterial injuries were simply ligated for hemorrhage control (Table IV). Vessels ligated included 10 radial and 3 ulnar arteries. Four ligations occurred in the axillary/subelavian area but did not actually involve ligation of either the subclavian or axillary artery. These ligations involved major branches of the axillary vessel that were believed by the surgeon responsible to be hemodynamicallysignificant, requiring surgical intervention. This included an arteriovenous fistula involving a branch of the axillary/subelavian artery in two patients and significant extravasation in two other patients. Lateral venography or vein grafting corrected the venous injuries. There were four negative explorations. Three of these were the result of false-positive arteriograms. The last negative exploration occurred after a wrist pulse was lost during reduction of a humeral fracture. The majority of vascular repairs were performed by vascular surgeons, but some patients were treated by plastic surgeons for distal vascular injury in association with their associated treatment of tendon or nerve injuries. Ancillary procedures included primary nerve repair in 30 limbs, tendon repair in 27 limbs, fasciotomy in 7 limbs, and bone fixation in 7 limbs. Of the 17 patients with remote injuries, only 1 required vascular repair in another portion of the body, while 4 underwent exploratory laparotomies. There were three pneumo-hemothoraxes that required chest tubes. These occurred in two patients with an axillary/subelavian artery injury and in one patient with a proximal brachial artery injury.
VOLUME 160 AUGUST 1990
UPPER-EXTREMITY VASCULAR TRAUMA
RESULTS There were no deaths in our 100 patients. Complications that occurred in these patients included one wound infection, three vessel repair thromboses, two laryngeal spasms, and one amputation for a Volkmann's ischemic contracture; tendon lacerations were missed in one patient and subsequently repaired at a later date. In the three arteries that thrombosed postoperatively, none had undergone an intraoperative completion angiogram. In two of these, there were no associated injuries and neither had an adverse effect on the patient's outcome. The last one also had a missed concurrent ulnar injury and resuited in our only amputation. There were three missed vascular injuries; two resulted in pseudoaneurysm formation that was subsequently recognized and repaired at a later date, while the third resulted in the only amputation as already mentioned. One of these missed injuries was the result of failing to obtain an angiogram in a wound examined for proximity only. Two of the patients with missed injuries had concurrent arterial injuries (brachial injury in two locations, radial and ulnar injury) that may have been detected on intraoperative completion angiography, In reviewing the charts and arteriograms of the four patients who underwent negative explorations, it was found that one was performed for a pulse deficit with no preoperative angiogram. The remaining three negative explorations occurred after angiography suggested an arterial injury in two stab wounds to the forearm and one gunshot wound to the upper arm. Review of the three false-positive angiograms suggested an overly cautious interpretation of the films. Two of the angiograms had a long-segment stenosis, while the last angiogram revealed a questionable intimal flap. In the 20 patients with no associated injuries, no functional defects were present at discharge from the hospital or in follow-up. In the remaining 80 patients, functional defects were evident and in some cases, very severe disability was noted. In 64% of the patients with associated nerve damage and in 25% of those with musculoskeletal injury, the arm was functionally impaired. A third of the patients with a nerve injury had a moderate to severe deficit that resulted in significant long-term disability, while 15% of the patients with a tendon injury had a moderate to severe disability. A third of proximal nerve injuries in association with axiUary/subclavian artery injuries resulted in a severe deficit while only 17% of nerve injuries in association with distal arterial injuries resulted in a severe deficit. Their associated vascular injuries did not contribute to the functional deficits. Since our trauma unit is an inner-city hospital that cares primarily for the indigent of Indianapolis, our longterm follow-up was less than optimal. Sixteen patients failed to keep any clinic appointments and were lost to all follow-up. Eighty-four patients returned for follow-up but not necessarily at their scheduled appointment. The average length of follow-up in the remaining 84 patients at their most recent clinic visit was 3.6 months, with follow-up ranging from 2 weeks to 28 months. Thirty-two patients were seen only once following injury, which was
T A B L E II
Associated Injuries In Upper-Extremity Vascular Trauma* Associated Injuries t Nerve Tendon Bony fracture Venous Remote None
Axillary/ Subclavian 6 0 3 6 4 0
(46) (23) (46) (31)
Brachial~
Radialw
Ulnar
Total
15 (65) 0 4 (17) 1 (4) 9 (39) 2 (9)
14 (35) 15 (38) 2 (5) 0 2 (5) 14 (35)
15 (60) 14 (56) 2 (8) 0 2 (8) 4 (16)
50 29 11 7 17 20
* Values in parentheses are percentages. t Several limbs had more than one associated injury. t Includes one limb with both a brachial and radial artery injury. wIncludes two limbs with both a radial and ulnar artery injury.
TABLE Ill
Arlerlogram Results In Upper-Extremity Vascular Trauma Axillary/ Subclavian
Brachial
Radial
Ulnar
Total
Occlusions Intimal flaps Pseudoaneurysms Arteriovenous fistula
1 3 4 2~
7 8t 2 0
6 1 2 1~
3* O 0 0
17 12 8 3t
Total
8~
9r
3
37~
17
* Includes one false-positive arteriogram, T Includes two false-positive arteriograms. t All arteriovenous fistulas had an associated pseudoaneurysm,
T A B L E IV
Method of Repair In Upper-Extremity Vascular Trauma* Type of Repair
Axillary/ Subclavian
Brachial t
Radialt
Ulnar
Total
Primary repair Vein graft Vein patch Ligation Negative exploration
2 (15) 6 (46) 1 (8) 4 (31) 0
9 (38) 10 (42) 2 (8) 0 3 (13)
23 (55) 8 (19) 0 11 (26) 0
20 (80) 2 (8) 0 2 (8) 1 (4)
54 26 3 17 4
24
42
25
Total vessels repaired
13
104w
* Values in parentheses are percentages. t Includes one limb with both a brachial and radial artery injury. Includes two limbs with both a radial and ulnar artery injury. wIncludes a total of 101 limbs plus 3 limbs with 2 vessels injured.
usually 2 weeks to 1 month after their discharge from the hospital. They failed to keep any subsequent clinic appointments. The few patients with long-term follow-up were generally those with a poor functional outcome secondary to nerve or skeletal/tendon injuries. COMMENTS Upper-extremity vascular trauma is nearly as common as lower-extremity vascular trauma, comprising
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160
AUGUST 1990
231
CIKRIT ET AL
47% of such noniatrogenic injuries treated at our medical is only a 6% incidence of arterial injury requiring surgical center. This is similar to reports from other institutions intervention, the social and economic implications of an [1,2]. Historically, the management of these injuries was upper-extremity amputation in this less-than-compliant developed during a war experience. During World War population may justify an angiogram for all proximity II, the preferred means of treatment was ligation with injuries. The indigent nature of our patients, as evidenced relatively high rates of amputation (40%) [3]. With sub- by their poor follow-up clinic attendance, does not allow sequent military encounters, the management of these us to observe these wounds for evidence of impending injuries evolved to expeditious repair, with decreased am- ischemia as Dennis and co-workers have suggested [15]. Patients who are noted to have a significant arterial putation rates ranging from 7% to 13% [4,5]. The civilian management of upper-extremity vascular trauma has uti- injury should undergo surgical repair. Most distal injuries lized many of the principles developed during these mili- can be repaired primarily, while proximal injuries may tary experiences, resulting in amputation rates ranging require autogenous vein grafting or patch angioplasty. Prior to completion of the surgical procedure, an intraopfrom 1% in our 6-year experience to 28% [1,6,7]. Penetrating trauma is the most common cause of up- erative arteriogram should be obtained to assess the adeper-extremity vascular trauma in the United States, as in quacy of the repair and also to rule out missed injuries. the present study [1,6,8-10]. Lacerations comprised 59% Postoperatively, the patients' pulses should be monitored of our injuries and may have contributed to the low am- and an upper-extremity Doppler examination in the vasputation rate and lack of mortality in our series. In the cular laboratory should be performed prior to discharge, Dallas experience and the New Orleans experience, the every 3 months for the first year, every 6 months for the most common wounding force was a bullet wound (42%) next 2 years, and once a year thereafter [16]. In patients with an equivocal arteriographic vascular and over 60% of the vessels injured were proximal vessels (subclavian, axillary, brachial artery) [6,10]. Stab injury and no clinical indications for surgical intervenwounds were the most common (48%) wounding force in tion, a period of close observation and a repeat angiogram San Antonio, with the most commonly injured vessels could be justified to decrease the number of negative lying again in a proximal position (60%) [1]. The Balti- explorations. The definition of an equivocal angiographic more experience had a significant number of distal vessel vascular injury includes segmental narrowing, especially lacerations (67%), which was similar to our experience of a long segment, with contrast extending past the lesion [8]. In contradistinction, less violent societies, such as in suggesting arterial spasm, or intimal flaps that are small Denmark, show a predominance of blunt trauma (92%) and do not infringe on more than 30% of the lumen. The incidence of nerve injury in association with upwith proximal vessel injuries predominating (86%) [7]. Clinical hard signs suggestive of an arterial injury per-extremity vascular trauma was 50% in this series, include pulse deficit, distal ischemia, local bruit, arterial which is similar to other reports [1,9,17,18]. In contrast, bleeding, or an expanding or pulsatile hematoma, while lower-extremity vascular trauma has a much lower incisoft signs include a small or moderately stable hematoma, dence of associated nerve injury [19,20]. The long-term proximity injury, and adjacent neurologic injury. The disability associated with upper-extremity neurologic inpresence of a pulse, as noted in 36% of our patients and jury is 27% to 44% [10,17,19]. Proximal extremity inju25% of the patients of Borman et al [6], does not ensure ries with nerve involvement have a poorer functional rethe absence of a significant arterial injury. In some cases, sult than injuries with distal nerve injuries [19]. Axillary/ the failure of a limb to demonstrate an absent pulse is a subclavian artery injuries were associated with the highreflection of the rich collateral circulation to the arm and est incidence of severe impairment (33%) in our series. also the fact that total occlusions are present in less than Nerve injuries in association with radial artery injuries half of the patients sustaining a vascular injury, The had the best functional outcome in our series and in the remaining arterial injuries include intimal defects, partial series of Hardin and co-workers [10]. Nichols and Lillelacerations, pseudoaneurysms, or arteriovenous fistula. hei [19] recommend primary nerve repair when feasible Although these nonocclusive injuries are not an immedi- for penetrating trauma (lacerations and stab wounds), ate threat to the viability of the limb, they all have the while with gunshot wounds, because of the degree of potential to cause ischemia by thrombosis, embolization, contusion, acute nerve repair is rarely indicated. Upper-extremity vascular trauma can have major soor rupture. Such events may suddenly change a viable limb to a limb in ischemic jeopardy. Consequently, these cial and economic ramifications since it generally occurs injuries need to be detected early and dealt with prompt- in a young population. Although the amputation rate is low, the functional disability that results from associated ly. The reported incidence of positive angiograms per- injuries can be devastating. To avoid the contribution of formed in patients for proximity injuries ranges from 2% vascular injury to the disability or amputation rate, we to 20% [11-15]. Dennis and associates [15] recently sug- recommend the liberal use of angiography preoperatively gested that angiography was not necessary in proximity both for proximity injuries and to localize clinically susinjuries and that these patients could be treated by close pected injuries, followed by prompt surgical correction of follow-up. Assuming the 229 negative angiograms in the obvious defects including defects in either the radial or present series were all done for proximity only (which was ulnar arteries. In all repairs, intraoperative completion not the case), then 14 angiograms in 243 studies were angiography should be used to assess the quality of the positive for a significant proximity injury. Although this repair and to rule out other vascular injuries. Intraopera232
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160
AUGUST 1990
UPPER-EXTREMITYVASCULARTRAUMA
tivc angiography should also be used in negative explorations to confirm the surgical impression. In conclusion, through this aggressive therapeutic and diagnostic approach, vascular injuries can be eliminated as the source of amputation or disability. Our experience in this series demonstrates that such care can result in superb results for limb salvage. Functional results, however, are dependent on nerve or musculoskeletal deficits. REFERENCES 1. Orcutt MB, Levin BA, Gaskill HV, Sirinek KR. Civilian vascular trauma of the upper extremity. J Trauma 1986; 26: 63-7. 2. Perry MO, Thai ER, Shires GT. Management of arterial injuries. Ann Surg 1971; 173: 403-8. 3. Debakey MD, Simeone FA. Battle injuries of the arteries in World War II: an analysis of 2,471 cases. Ann Surg 1946; 123: 534-79. 4. Rich NM, Bauch JH, Hughes CW. Acute arterial injuries in Vietnam: 1000 cases. J Trauma 1970; 10: 359-69. 5. Adar R, Schramek A, Khodadadi J, Zwieg A, Goleman L, Romanoff H. Arterial combat injuries of the upper extremity. J Trauma 1980; 20: 297-302. 6. Borman KR, Snyder WH, Weigeit JA. Civilian arterial trauma of the upper extremity: an 11 year experiencein 267 patients. Am J Surg 1984; 148: 796-9. 7. Kruse-Anderson S, Lorentzen JE, Rohr N. Arterial injuries of the upper extremities. Acta Chir Scand 1983; 149: 473-7. 8. Sitzmann JV, Ernst CB. Management of arm arterial injuries. Surgery 1984; 96: 895-901. 9. McCready RA. Upper-extremity vascular injuries. Surg Clin North Am 1988; 68: 725-40. 10. Hardin WD, O'Connell RC, Adinolfi MF, Kerstein MD. Traumatic arterial injuries of the upper extremity: determinants of disability. Am J Surg 1985; 150: 266-70. 11. Gomez GA, Kreis DJ, Ratner L, et al. Suspected vascular trauma of the extremities: the role of arteriography in proximity injuries. J Trauma 1986; 26: 1005-8. 12. Menzoian JO, Doyle JE, Cantelmo NL, eta/. A comprehensive approach to extremity vascular trauma. Arch Surg 1985; 120:8015. 13. Frykberg ER, Crump JM, Vines FS, et al. A reassessment of the role of arteriography in penetrating proximity extremity trauma: a prospective study. J Trauma 1989; 29: 1041-52. 14. Howard CA, Thai ER, Redman HC, et al. Intra-arterial digital substraetion angiography in the evaluation of peripheral vascular trauma. Ann Surg 1989; 210: 108-11. 15. Dennis JW, Frykberg ER, Crump JM, et al. New perspectives on the management of penetrating trauma in proximity to major limb arteries. J Vase Surg 1990; 11: 84-93. 16. Rutherford RB. Diagnostic evaluation of extremity vascular trauma. Surg Clin North Am 1988; 68: 683-91. 17. Peacock JB, Proctor JH. Factors limiting extremity function following vascular injury. J Trauma 1977; 17: 532-4. 18. Smith RF, Elliott JP, Hagerman JH, Szilagyi DE, et al. Acute penetrating arterial injuries of the neck and limbs. Arch Surg 1974; 109: 198-205.
19. Nichols JS, Lillehei KO. Nerve injury associated with acute vascular trauma. Surg Clin North Am 1988; 68: 837-52. 20. Visser PA, Hermreck AS, Pierce GE, et al. Prognosis of nerve injuries incurred during acute trauma to peripheral arteries. Am J Surg 1980; 140: 596-9.
DISCUSSION Stanley R. Klein (Los Angeles, CA): Dr. Cikrit, how long were patients with only arteriograms in the hospital, assuming that the injuries were isolated extremity trauma, and were there any complications in that group? What were the findings on the positive arteriograms? Were the patients re-examined prior to the arteriogram, i.e., was the patient who had an arteriovenous fistula reexamined? Did anyone hear a bruit? Would you modify your recommendation for angiography for proximity to colleagues in rural locations who may not see as many cases but who have better patient follow-up? You have said that not all instances of arterial injury seen on arteriograms warrant surgery. These injuries mainly occurred in nonbrachial vessel segments that were treated by ligation. In the absence of clinical evidence of transmural vessel disruption or concomitant injury requiring exploration, do you feel that all these lesions require exploration? Finally, would you remark on the correlation of vascular and neurologic injury and the appropriate time in follow-up in order to give a definite prognosis? Dolores F. Cikrit (closing): Most patients went to arteriography within 6 hours of arrival at the emergency room. The patients were initially examined by the emergency room physician and then by the chief resident on call prior to angiography. There were no complications arising from angiography. The 14 injuries detected on angiography based on proximity only were examined before and after the procedure and prior to surgery without any clinical indication of the injury. Eight of these patients had an intimal flap and six had an occlusion or near-occlusion. We can't follow patients in our population because of their noncompliance. Many of them fail to keep their clinic appointments. In other localities, they may be able to be followed by periodic clinical examinations. With regard to patients with serious neurologic injury and deficit, they had a long follow-up because they came in periodically for physiotherapy; by a severe deficit, we mean probably lifelong.
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 AUGUST 1990 9.33
" pper-extremity vascular trauma constitutes nearly half of all extremity vascular trauma in the United U States [1]. The majority of injuries are of a violent penetrating nature. Since these injuries generally occur in young males, the social and economic implications can be significant, resulting in marked morbidity from poor functional outcomes or amputation. Although the collateral circulation to the arm is extensive, vascular insufficiency can still occur at the time of injury or at a later date from thrombosis, embolization, or rupture of an unrecognized vascular injury. This study was undertaken in an effort to better define these injuries and to help decrease the disability that can be associated with upperextremity vascular trauma. From the Department of Surgery, Indiana University Medical Center, Indianapolis, Indiana. Requests for reprints should be addressed to Dolores F. Cikrit, MD, Department of Surgery, Indiana University Medical Center, 1001 West 10th Street, Indianapolis, Indiana 46202. Presented at the 18th Annual Meeting of the Society for Clinical Vascular Surgery, Palm Desert, California, March 7-11, 1990.
PATIENTS AND M E T H O D S In this retrospective study, 329 consecutive patients were evaluated for upper-extremity vascular trauma at Indiana University Medical Center Trauma Unit at Wishard Memorial Hospital, a county hospital which is one of two major trauma centers in the city of Indianapolis. Iatrogenic injuries were excluded from this study. From January 1983 through August 1989, vascular injury was excluded angiographically in 229 patients. The remaining 100 patients (101 limbs) required surgical intervention for their vascular injury and form the basis of this report. They were evaluated for presentation, physical findings, method of diagnosis, associated injuries, surgical intervention, postoperative course, functional assessment, and long-term follow-up. Most patients were male (79%) and young, with a mean age of 28 years (range: 5 to 89 years). Arterial injuries involved the axillary/subclavian (13 patients), brachial (22), radial (38), ulnar (25), brachial/radial (1), and radial/ulnar (2) arteries (Figure 1). This included one patient who had bilateral brachial artery injuries. Penetrating wounds accounted for 94% of the injuries while blunt trauma comprised the remainder. There were 59 lacerations, including 15 that were caused by glass and 7 that were self-inflicted in suicide attempts. Ninety percent of the lacerations involved the radial and/or ulnar arteries. The remainder of the penetrating injuries included 15 stab wounds, 12 gunshot wounds, and 7 shotgun blasts (Table I). Proximal vessel (subclavian/axillary/ braehial artery) injuries were usually (67% of the time) the result of stab, gunshot, or shotgun blast wounds, whereas distal arterial (radial/ulnar artery) injuries usually resulted from lacerations. The blunt trauma included two motor vehicle accidents, two elbow dislocations, one conveyor belt injury, one fall, and one supracondylar humeral fracture. All but one of the cases of blunt trauma involved proximal vessels. Twenty-nine patients presented with absent pulses, while 36 patients were found to have a palpable pulse despite having a major vascular injury. Distal pulses were absent in 31% of the limbs with an axillary/subclavian artery injury, 43% with a brachial artery injury, 24% with an ulnar artery injury, and 23% with a radial artery injury. Palpable radial or ulnar pulses were noted in 32% of patients with an axinary/subclavian injury, 35% of limbs with a brachial artery injury, 52% of patients with an ulnar artery injury, and 28% of limbs with a radial artery injury. An abnormal Allen test suggestive of vascular injury was noted in only six patients. The remaining 30 patients had diminished pulses. Arterial bleeding was noted in the emergency room in 45 patients (Table I). Bleeding was associated with 55% and 40%, respectively, of radial and ulnar artery injuries. Nearly half (46%) of
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160 AUGUST 1990
229
CIKRIT ET AL
, . ~ L N A R
RADIAL
/=/ BRACHIAL
AXILLARY SUBCLAVIAN Figure 1. Distal (radial and ulnar arteries) injuries comprised the majority of vessels injured with only 36% of injuries involving proximal upper-extremity vessels (subclavian, axillary, brachial arteries).
TABLE I
Types and Presentation of Injury In Upper-Extremity Vascular Trauma* Axillary/ Subclavian Brachial t Radialt Wounding force Laceration Stab Gunshot Shotgun blast Blunt Presentation Palpable pulses
Absent pulses Arterial bleeding Concomitant injury Total Limbs
Ulnar
Total
1 (8) 4 (31) 5 (38) 1 (8) 2 (15)
5 (22) 4 (17) 4 (17) 6 (26) 4 (17)
32 (80) 21 (84) 5 (13) 2 (8) 2 (5) 1 (4) 1 (3) 0 0 1 (4)
59 15 12 8 7
4 (31) 4 (31) 6 (48)
8 (35) 10 (43) 7 (30)
11 (28) 13 (52) g (23) 6 (24) 22 (55) 10 (40)
36 29 45
13
21
26
21
81
13
23
40
25
101
* Values in parenthesesare percentages. "t Includes one limb with both a brachial and radial artery injury. ~tIncludestwo limbs with both a radial and ulnar artery injury.
subclavian/axillary artery injuries presented with arterial bleeding. Only 20 patients had an isolated arterial injury. These isolated arterial injuries involved the radial or ulnar arteries in 18 patients and the brachial artery in the remaining 2. There were seven major venous injuries, which were all associated with proximal arterial injuries. Concomitant nonvascular injuries included nerve damage in 50 limbs, at least 1 tendon laceration or transection in 29 limbs, bony fracture in 11 limbs, and remote injuries in 17 patients (Table II). Nerve injuries included 41 transections or lacerations and 9 contusions, including 7 brachial plexus injuries. The remaining nerve injuries were divided nearly equally between the radial, ulnar, and median nerves. Of ulnar and radial artery injuries, 230
THE AMERICAN JOURNAL OF SURGERY
60% and 35%, respectively, had associated nerve injury. This was also the ease in 65% and 46%, respectively, of brachial and axiUary/subclavian artery injuries. The majority of patients with tendon lacerations had more than one tendon damaged, and these obviously all occurred in association with radial or ulnar artery injury. Sixty-three patients went directly from the shock room to the surgical suite because of arterial bleeding (31 patients), severe ischemia (20), or localized injury that required simultaneous tendon or nerve repair (12). Biplanar angiography in the radiology department was utilized in 14 patients for injury in close proximity to a major vessel and in 23 patients for localization of a clinically apparent injury. In patients who underwent angiography, there were 17 occlusions, 12 intimal flaps, 8 pseudoaneurysms, and 3 artedovenous fistulas noted (Table III). In all three patients with an arteriovenous fistula, a pseudoaneurysm was also noted. Three false-positive angiograms resulted in negative explorations. Only nine venograms were obtained, revealing extravasation in two patients, compression in three, and intimal flap in one; two were negative, although on exploration for arterial repair one of these had a vein wall defect noted and corrected. Arterial repair was accomplished by primary anastomosis in 54 vessels, interposition vein grafting in 26 vessels, and vein patch angioplasty in only 3 vessels. Two thirds of the distal vessels (ulnar or radial artery) were repaired primarily, which differed markedly from the more proximal vessel (subelavian/axillary/brachial artery) repairs, where less than one third of the vessels could be repaired primarily. Seventeen arterial injuries were simply ligated for hemorrhage control (Table IV). Vessels ligated included 10 radial and 3 ulnar arteries. Four ligations occurred in the axillary/subelavian area but did not actually involve ligation of either the subclavian or axillary artery. These ligations involved major branches of the axillary vessel that were believed by the surgeon responsible to be hemodynamicallysignificant, requiring surgical intervention. This included an arteriovenous fistula involving a branch of the axillary/subelavian artery in two patients and significant extravasation in two other patients. Lateral venography or vein grafting corrected the venous injuries. There were four negative explorations. Three of these were the result of false-positive arteriograms. The last negative exploration occurred after a wrist pulse was lost during reduction of a humeral fracture. The majority of vascular repairs were performed by vascular surgeons, but some patients were treated by plastic surgeons for distal vascular injury in association with their associated treatment of tendon or nerve injuries. Ancillary procedures included primary nerve repair in 30 limbs, tendon repair in 27 limbs, fasciotomy in 7 limbs, and bone fixation in 7 limbs. Of the 17 patients with remote injuries, only 1 required vascular repair in another portion of the body, while 4 underwent exploratory laparotomies. There were three pneumo-hemothoraxes that required chest tubes. These occurred in two patients with an axillary/subelavian artery injury and in one patient with a proximal brachial artery injury.
VOLUME 160 AUGUST 1990
UPPER-EXTREMITY VASCULAR TRAUMA
RESULTS There were no deaths in our 100 patients. Complications that occurred in these patients included one wound infection, three vessel repair thromboses, two laryngeal spasms, and one amputation for a Volkmann's ischemic contracture; tendon lacerations were missed in one patient and subsequently repaired at a later date. In the three arteries that thrombosed postoperatively, none had undergone an intraoperative completion angiogram. In two of these, there were no associated injuries and neither had an adverse effect on the patient's outcome. The last one also had a missed concurrent ulnar injury and resuited in our only amputation. There were three missed vascular injuries; two resulted in pseudoaneurysm formation that was subsequently recognized and repaired at a later date, while the third resulted in the only amputation as already mentioned. One of these missed injuries was the result of failing to obtain an angiogram in a wound examined for proximity only. Two of the patients with missed injuries had concurrent arterial injuries (brachial injury in two locations, radial and ulnar injury) that may have been detected on intraoperative completion angiography, In reviewing the charts and arteriograms of the four patients who underwent negative explorations, it was found that one was performed for a pulse deficit with no preoperative angiogram. The remaining three negative explorations occurred after angiography suggested an arterial injury in two stab wounds to the forearm and one gunshot wound to the upper arm. Review of the three false-positive angiograms suggested an overly cautious interpretation of the films. Two of the angiograms had a long-segment stenosis, while the last angiogram revealed a questionable intimal flap. In the 20 patients with no associated injuries, no functional defects were present at discharge from the hospital or in follow-up. In the remaining 80 patients, functional defects were evident and in some cases, very severe disability was noted. In 64% of the patients with associated nerve damage and in 25% of those with musculoskeletal injury, the arm was functionally impaired. A third of the patients with a nerve injury had a moderate to severe deficit that resulted in significant long-term disability, while 15% of the patients with a tendon injury had a moderate to severe disability. A third of proximal nerve injuries in association with axiUary/subclavian artery injuries resulted in a severe deficit while only 17% of nerve injuries in association with distal arterial injuries resulted in a severe deficit. Their associated vascular injuries did not contribute to the functional deficits. Since our trauma unit is an inner-city hospital that cares primarily for the indigent of Indianapolis, our longterm follow-up was less than optimal. Sixteen patients failed to keep any clinic appointments and were lost to all follow-up. Eighty-four patients returned for follow-up but not necessarily at their scheduled appointment. The average length of follow-up in the remaining 84 patients at their most recent clinic visit was 3.6 months, with follow-up ranging from 2 weeks to 28 months. Thirty-two patients were seen only once following injury, which was
T A B L E II
Associated Injuries In Upper-Extremity Vascular Trauma* Associated Injuries t Nerve Tendon Bony fracture Venous Remote None
Axillary/ Subclavian 6 0 3 6 4 0
(46) (23) (46) (31)
Brachial~
Radialw
Ulnar
Total
15 (65) 0 4 (17) 1 (4) 9 (39) 2 (9)
14 (35) 15 (38) 2 (5) 0 2 (5) 14 (35)
15 (60) 14 (56) 2 (8) 0 2 (8) 4 (16)
50 29 11 7 17 20
* Values in parentheses are percentages. t Several limbs had more than one associated injury. t Includes one limb with both a brachial and radial artery injury. wIncludes two limbs with both a radial and ulnar artery injury.
TABLE Ill
Arlerlogram Results In Upper-Extremity Vascular Trauma Axillary/ Subclavian
Brachial
Radial
Ulnar
Total
Occlusions Intimal flaps Pseudoaneurysms Arteriovenous fistula
1 3 4 2~
7 8t 2 0
6 1 2 1~
3* O 0 0
17 12 8 3t
Total
8~
9r
3
37~
17
* Includes one false-positive arteriogram, T Includes two false-positive arteriograms. t All arteriovenous fistulas had an associated pseudoaneurysm,
T A B L E IV
Method of Repair In Upper-Extremity Vascular Trauma* Type of Repair
Axillary/ Subclavian
Brachial t
Radialt
Ulnar
Total
Primary repair Vein graft Vein patch Ligation Negative exploration
2 (15) 6 (46) 1 (8) 4 (31) 0
9 (38) 10 (42) 2 (8) 0 3 (13)
23 (55) 8 (19) 0 11 (26) 0
20 (80) 2 (8) 0 2 (8) 1 (4)
54 26 3 17 4
24
42
25
Total vessels repaired
13
104w
* Values in parentheses are percentages. t Includes one limb with both a brachial and radial artery injury. Includes two limbs with both a radial and ulnar artery injury. wIncludes a total of 101 limbs plus 3 limbs with 2 vessels injured.
usually 2 weeks to 1 month after their discharge from the hospital. They failed to keep any subsequent clinic appointments. The few patients with long-term follow-up were generally those with a poor functional outcome secondary to nerve or skeletal/tendon injuries. COMMENTS Upper-extremity vascular trauma is nearly as common as lower-extremity vascular trauma, comprising
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160
AUGUST 1990
231
CIKRIT ET AL
47% of such noniatrogenic injuries treated at our medical is only a 6% incidence of arterial injury requiring surgical center. This is similar to reports from other institutions intervention, the social and economic implications of an [1,2]. Historically, the management of these injuries was upper-extremity amputation in this less-than-compliant developed during a war experience. During World War population may justify an angiogram for all proximity II, the preferred means of treatment was ligation with injuries. The indigent nature of our patients, as evidenced relatively high rates of amputation (40%) [3]. With sub- by their poor follow-up clinic attendance, does not allow sequent military encounters, the management of these us to observe these wounds for evidence of impending injuries evolved to expeditious repair, with decreased am- ischemia as Dennis and co-workers have suggested [15]. Patients who are noted to have a significant arterial putation rates ranging from 7% to 13% [4,5]. The civilian management of upper-extremity vascular trauma has uti- injury should undergo surgical repair. Most distal injuries lized many of the principles developed during these mili- can be repaired primarily, while proximal injuries may tary experiences, resulting in amputation rates ranging require autogenous vein grafting or patch angioplasty. Prior to completion of the surgical procedure, an intraopfrom 1% in our 6-year experience to 28% [1,6,7]. Penetrating trauma is the most common cause of up- erative arteriogram should be obtained to assess the adeper-extremity vascular trauma in the United States, as in quacy of the repair and also to rule out missed injuries. the present study [1,6,8-10]. Lacerations comprised 59% Postoperatively, the patients' pulses should be monitored of our injuries and may have contributed to the low am- and an upper-extremity Doppler examination in the vasputation rate and lack of mortality in our series. In the cular laboratory should be performed prior to discharge, Dallas experience and the New Orleans experience, the every 3 months for the first year, every 6 months for the most common wounding force was a bullet wound (42%) next 2 years, and once a year thereafter [16]. In patients with an equivocal arteriographic vascular and over 60% of the vessels injured were proximal vessels (subclavian, axillary, brachial artery) [6,10]. Stab injury and no clinical indications for surgical intervenwounds were the most common (48%) wounding force in tion, a period of close observation and a repeat angiogram San Antonio, with the most commonly injured vessels could be justified to decrease the number of negative lying again in a proximal position (60%) [1]. The Balti- explorations. The definition of an equivocal angiographic more experience had a significant number of distal vessel vascular injury includes segmental narrowing, especially lacerations (67%), which was similar to our experience of a long segment, with contrast extending past the lesion [8]. In contradistinction, less violent societies, such as in suggesting arterial spasm, or intimal flaps that are small Denmark, show a predominance of blunt trauma (92%) and do not infringe on more than 30% of the lumen. The incidence of nerve injury in association with upwith proximal vessel injuries predominating (86%) [7]. Clinical hard signs suggestive of an arterial injury per-extremity vascular trauma was 50% in this series, include pulse deficit, distal ischemia, local bruit, arterial which is similar to other reports [1,9,17,18]. In contrast, bleeding, or an expanding or pulsatile hematoma, while lower-extremity vascular trauma has a much lower incisoft signs include a small or moderately stable hematoma, dence of associated nerve injury [19,20]. The long-term proximity injury, and adjacent neurologic injury. The disability associated with upper-extremity neurologic inpresence of a pulse, as noted in 36% of our patients and jury is 27% to 44% [10,17,19]. Proximal extremity inju25% of the patients of Borman et al [6], does not ensure ries with nerve involvement have a poorer functional rethe absence of a significant arterial injury. In some cases, sult than injuries with distal nerve injuries [19]. Axillary/ the failure of a limb to demonstrate an absent pulse is a subclavian artery injuries were associated with the highreflection of the rich collateral circulation to the arm and est incidence of severe impairment (33%) in our series. also the fact that total occlusions are present in less than Nerve injuries in association with radial artery injuries half of the patients sustaining a vascular injury, The had the best functional outcome in our series and in the remaining arterial injuries include intimal defects, partial series of Hardin and co-workers [10]. Nichols and Lillelacerations, pseudoaneurysms, or arteriovenous fistula. hei [19] recommend primary nerve repair when feasible Although these nonocclusive injuries are not an immedi- for penetrating trauma (lacerations and stab wounds), ate threat to the viability of the limb, they all have the while with gunshot wounds, because of the degree of potential to cause ischemia by thrombosis, embolization, contusion, acute nerve repair is rarely indicated. Upper-extremity vascular trauma can have major soor rupture. Such events may suddenly change a viable limb to a limb in ischemic jeopardy. Consequently, these cial and economic ramifications since it generally occurs injuries need to be detected early and dealt with prompt- in a young population. Although the amputation rate is low, the functional disability that results from associated ly. The reported incidence of positive angiograms per- injuries can be devastating. To avoid the contribution of formed in patients for proximity injuries ranges from 2% vascular injury to the disability or amputation rate, we to 20% [11-15]. Dennis and associates [15] recently sug- recommend the liberal use of angiography preoperatively gested that angiography was not necessary in proximity both for proximity injuries and to localize clinically susinjuries and that these patients could be treated by close pected injuries, followed by prompt surgical correction of follow-up. Assuming the 229 negative angiograms in the obvious defects including defects in either the radial or present series were all done for proximity only (which was ulnar arteries. In all repairs, intraoperative completion not the case), then 14 angiograms in 243 studies were angiography should be used to assess the quality of the positive for a significant proximity injury. Although this repair and to rule out other vascular injuries. Intraopera232
THE AMERICAN JOURNAL OF SURGERY
VOLUME 160
AUGUST 1990
UPPER-EXTREMITYVASCULARTRAUMA
tivc angiography should also be used in negative explorations to confirm the surgical impression. In conclusion, through this aggressive therapeutic and diagnostic approach, vascular injuries can be eliminated as the source of amputation or disability. Our experience in this series demonstrates that such care can result in superb results for limb salvage. Functional results, however, are dependent on nerve or musculoskeletal deficits. REFERENCES 1. Orcutt MB, Levin BA, Gaskill HV, Sirinek KR. Civilian vascular trauma of the upper extremity. J Trauma 1986; 26: 63-7. 2. Perry MO, Thai ER, Shires GT. Management of arterial injuries. Ann Surg 1971; 173: 403-8. 3. Debakey MD, Simeone FA. Battle injuries of the arteries in World War II: an analysis of 2,471 cases. Ann Surg 1946; 123: 534-79. 4. Rich NM, Bauch JH, Hughes CW. Acute arterial injuries in Vietnam: 1000 cases. J Trauma 1970; 10: 359-69. 5. Adar R, Schramek A, Khodadadi J, Zwieg A, Goleman L, Romanoff H. Arterial combat injuries of the upper extremity. J Trauma 1980; 20: 297-302. 6. Borman KR, Snyder WH, Weigeit JA. Civilian arterial trauma of the upper extremity: an 11 year experiencein 267 patients. Am J Surg 1984; 148: 796-9. 7. Kruse-Anderson S, Lorentzen JE, Rohr N. Arterial injuries of the upper extremities. Acta Chir Scand 1983; 149: 473-7. 8. Sitzmann JV, Ernst CB. Management of arm arterial injuries. Surgery 1984; 96: 895-901. 9. McCready RA. Upper-extremity vascular injuries. Surg Clin North Am 1988; 68: 725-40. 10. Hardin WD, O'Connell RC, Adinolfi MF, Kerstein MD. Traumatic arterial injuries of the upper extremity: determinants of disability. Am J Surg 1985; 150: 266-70. 11. Gomez GA, Kreis DJ, Ratner L, et al. Suspected vascular trauma of the extremities: the role of arteriography in proximity injuries. J Trauma 1986; 26: 1005-8. 12. Menzoian JO, Doyle JE, Cantelmo NL, eta/. A comprehensive approach to extremity vascular trauma. Arch Surg 1985; 120:8015. 13. Frykberg ER, Crump JM, Vines FS, et al. A reassessment of the role of arteriography in penetrating proximity extremity trauma: a prospective study. J Trauma 1989; 29: 1041-52. 14. Howard CA, Thai ER, Redman HC, et al. Intra-arterial digital substraetion angiography in the evaluation of peripheral vascular trauma. Ann Surg 1989; 210: 108-11. 15. Dennis JW, Frykberg ER, Crump JM, et al. New perspectives on the management of penetrating trauma in proximity to major limb arteries. J Vase Surg 1990; 11: 84-93. 16. Rutherford RB. Diagnostic evaluation of extremity vascular trauma. Surg Clin North Am 1988; 68: 683-91. 17. Peacock JB, Proctor JH. Factors limiting extremity function following vascular injury. J Trauma 1977; 17: 532-4. 18. Smith RF, Elliott JP, Hagerman JH, Szilagyi DE, et al. Acute penetrating arterial injuries of the neck and limbs. Arch Surg 1974; 109: 198-205.
19. Nichols JS, Lillehei KO. Nerve injury associated with acute vascular trauma. Surg Clin North Am 1988; 68: 837-52. 20. Visser PA, Hermreck AS, Pierce GE, et al. Prognosis of nerve injuries incurred during acute trauma to peripheral arteries. Am J Surg 1980; 140: 596-9.
DISCUSSION Stanley R. Klein (Los Angeles, CA): Dr. Cikrit, how long were patients with only arteriograms in the hospital, assuming that the injuries were isolated extremity trauma, and were there any complications in that group? What were the findings on the positive arteriograms? Were the patients re-examined prior to the arteriogram, i.e., was the patient who had an arteriovenous fistula reexamined? Did anyone hear a bruit? Would you modify your recommendation for angiography for proximity to colleagues in rural locations who may not see as many cases but who have better patient follow-up? You have said that not all instances of arterial injury seen on arteriograms warrant surgery. These injuries mainly occurred in nonbrachial vessel segments that were treated by ligation. In the absence of clinical evidence of transmural vessel disruption or concomitant injury requiring exploration, do you feel that all these lesions require exploration? Finally, would you remark on the correlation of vascular and neurologic injury and the appropriate time in follow-up in order to give a definite prognosis? Dolores F. Cikrit (closing): Most patients went to arteriography within 6 hours of arrival at the emergency room. The patients were initially examined by the emergency room physician and then by the chief resident on call prior to angiography. There were no complications arising from angiography. The 14 injuries detected on angiography based on proximity only were examined before and after the procedure and prior to surgery without any clinical indication of the injury. Eight of these patients had an intimal flap and six had an occlusion or near-occlusion. We can't follow patients in our population because of their noncompliance. Many of them fail to keep their clinic appointments. In other localities, they may be able to be followed by periodic clinical examinations. With regard to patients with serious neurologic injury and deficit, they had a long follow-up because they came in periodically for physiotherapy; by a severe deficit, we mean probably lifelong.
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 AUGUST 1990 9.33
