- Email: [email protected]
Protecting the brachial plexus during median sternotomy
J
THORAC CARDIOVASC SURG
1987;94:297-301
Protecting the brachial plexus during median sternotomy Injuryto the brachial plexuswas prospectively assessedin 335 patients Wldergoing mediansternotomyfor cardiac operation. All patients were placed in the hands-up position(elbows elevated, arms abducted 90 degrees, and elbows flexed) after right internal jugular vein cannulation (23 cannulation-attempts were bilateral). Twenty-eight patients bad new upper extremity complaints after the operation, of whom 16 (4.8 % ) bad symptoms considered related to injury of the brachial plexus: one with generalized weakness of the left arm, six with localized weakness, pain, or paresthesia plus objective hypesthesia or weakness, and ninewith paresthesiasbut no objective signs. Four injurieswere right sided,four left sided, and eight bilateral. At the time of discbarge,15 of 16 weresymptomfree and the patient with generalized weakness wasrapidly improving. Postoperativeplexopathy was not related to degreeof sternal retraction, dissection of the internal mammary artery, or cannulationof the internal jugular vein. We believe the lowincidence and benign course of brachial plexus problems in these patients resulted from careful sternal retraction and use of the-bands-up position. Finally,our data do not support internal jugular cannulationas a major cause of plexus injuries after median sternotomy.
David L. Tomlinson, M.D., Irving A. Hirsch, M.D., S. V. Kodali, M.D., and Stephen Slogoff, M.D., Houston, Texas
Complaints of pain, weakness, numbness, and tingling of shoulders, arms, and hands are not unusual after operations requiring median sternotomy. These are commonly attributed to brachial plexus injury from stretch or compression by malposition of the arms or excessive sternal retraction."! Direct injury to the plexus byoccultfractures of the first rib because of improperly applied sternal retractors" S and by internal jugular vein cannulation'-" have been described. A recent prospective study of neurologic injury after median sternotomy suggested that internal jugular vein cannulation was the primary cause of brachial plexus injury after median sternotomy. Since we adopted the hands-up position of Jackson and Keats? in 1965, clinically obvious brachial plexus injury has been rare in our patients after median sternotomy. In addition, we have routinely catheterized theright internal jugular vein since 1975 without noting From the Division of Cardiovascular Anesthesiology, Texas Heart Institute, Houston, Texas. Received for publication July 24, 1986. Accepted for publication Aug. 24, 1986. Address for reprints: Stephen Slogoff, M.D., Cardiovascular Anesthesia, Texas Heart Institute, P.O. Box 20269, Houston, Texas 77025.
an increase in brachial plexus injury or other postoperative complaints referable to the right arm. To document our low incidence of nerve injury with the hands-up position, we performed a prospective study of this complication, which also failed to support any relationship between internal jugular vein cannulation and brachial plexus injury. Methods Three hundred sixty-five adult patients undergoing elective cardiac operations performed through a median sternotomy and necessitating central venouscannulation were prospectively evaluated before and after operation for symptoms and signs possibly related to brachial plexusdysfunction. Specificpreoperativeassessment (by LA.H. and D.L.T.) consisted of eliciting a detailed history of upper extremity pain, paresthesia, loss of sensationor weakness, assessmentof sensory responseto pin prick, and examination of motor function of muscle groups innervated by the nerves derived from the brachial plexus. These included adduction of the arm (C S_7) , forward thrust of the shoulder (C S_7) , elevation of the scapula (C), abduction, lateral and medial rotation of the arm (C s-s), flexion and supination of the forearm (C S-6) , ulnar flexion of the hand (C 7•S' T 1) , thumb apposition and flexion (C 7•s,T t ) , extension of the fore297
The Journal of
2 9 8 Tomlinson et al.
Fig. 1. A, A patient in the hands-up position on a foamwedge pad. The shoulders are elevated, elbows are 20 em above the horizontal, arms are abducted, and elbows flexed. Position is maintained by masking tape placed to avoid compression of the ulnar nerve at the elbow. B, The same patient is shown from above. Note that the arms are abducted to no more than 90% and the wedge and arm board are placed sufficiently cephalad to extend no further caudad than the arm. Thus the hands-up position provides ready access to arterial and venous cannulas without interfering with the surgeon's proximity to the patient.
arm and phalanges (C6-s), and radial extension of the hand and extension of the thumb (C 7•S) ' Eight patients were excluded because of symptoms or objective findings possibly related to central or peripheral nerve dysfunction. The first 50 patients were interviewed and examined by both observers to develop a uniform technique. After induction of anesthesia, the right internal jugular vein was cannulated with a 16 gauge, 5V
Thoracic and Cardiovascular Surgery
elevate elbows 20 em above the level of the operating room table (Fig. 1). Arms were abducted to no more than 90 degrees and the elbows were flexed. This position was maintained with masking tape placed to avoid compression of the ulnar nerve at the elbow. A Cooley sternal retractor with the crossbar caudad and retractor blades at the middle or lower third of the sternum was used in all patients. Patients were reexamined as described 5 to 7 days after operation. Any symptoms or objective signs not present on preoperative examination required confirmation by the second observer for inclusion. Patients with abnormal findings were examined at 2 day intervals until discharge. All abnormal observations were reviewe'd with a consulting neurologist to judge their relationship to injury of the brachial plexus. These relationships were defined for classification purposes as definite (extensive motor paralysis of the arm), probable (both subjective and objective findings attributable to injury of an anatomical portion of the plexus), possible (paresthesias without objective signs localized to established nerve distributions), and not related (symptoms or signs not related to an identifiable nerve distribution). Twenty-two patients examined preoperatively were excluded from the study because postoperative examination could not be performed: Four had central neurologic dysfunction, six were undergoing continued mechanical ventilation, 10 were discharged before the seventh day, and two died before the fifth day. The final group of 335 patients consisted of 244 who underwent coronary artery bypass, 58 who had valve replacement or repair, and 33 who had coronary artery bypass plus valve or other intracardiac procedure. All operations lasted less than 4 hours. Maximum perfusion time was 100 minutes and the average was 63 minutes. Statistical analysis of relative frequencies was by Fisher's exact test.
Results New symptoms or signs involving one or both upper extremities were reported by 28 (8.4%) patients (Table I). In 16 patients (4.8%) these new findings were considered related to brachial plexus injury: Of these, seven (2.1 %) were considered to be definitely (Patient 1, Table I) or probably related (Patients 2 to 7, Table I) and nine (2.7%) (Patients 8 to 16, Table I) were considered only possibly related. New symptoms of the remaining 12 patients (Patients 17 to 28, Table I) were not considered related to brachial plexus injury. Patient 17 complained of bilateral glove distribution paresthesias believed to be of hysterical origin. Patient 18, who had bilateral hand paresthesias on raising his arms, was
Volume 94 Number 2 August 1987
Protection of brachial plexus during median sternotomy
299
Table I. New symptoms, signs, or both, present 5 to 7 days after median sternotomy in 28 patients Patient
Condition at discharge
Involved side
Symptoms
Signs
L
Tingling, numbness, and weakness entire arm Generalized decrease in sensation and coordination Generalized pain
Profound weakness of all muscle groups; normal pin prick response
Significantly improving
Weakness of all muscle groups, hypesthesia C6-8,T, (L only)
Normal
Weakness on flexion and extension of forearms Weakness on abduction and adduction of fingers Hypesthesia C6-8,T, Hypesthesia C, Hypesthesia C,
Normal
2
LandR
3
LandR
4
LandR
5 6 7
LandR L R
8-12 13,14 15 16
Land R L R R
17
Land R
18
Land R
19-23
L
24,25
R
26 27 28
L R R
Decreased sensation and tingling C6-8,T, Tingling Co-8,T, Shoulder pain Shoulder pain on abduction Tingling C 6..,T, Tingling C 7_8,T, Tingling C'_8,T, Hypesthesia and tingling of shoulder, arm, and hand Tingling in glove distribution Tingling of hands on raising arms Generalized arm pain on movement Shoulder pain on abduction Generalized hand pain Tingling of thumb Tingling of fifth finger
Normal Normal Normal Normal
None None None None
Normal Normal Normal Normal
None
Normal
None
No change
None
Normal
None
Normal
None None None
Normal Normal Normal
Legend: L, Left. R, Right.
found to have markedly diminished radial pulses on abduction. Eight had generalized pain in the shoulder, arm, or hand not related to any nerve distribution. Patients 27 and 28 had paresthesias limited to a single finger on the right hand, which was used for intravenous infusion with a large-bore cannula. In all but one patient with plexopathy, symptoms and objective signs disappeared by the time of discharge 8 to 14 days after operation. Patient 1, with the greatest disability, left the hospital with residual weakness but was rapidly improving and had a functional arm. The extent of sternal retraction did not influence frequency of plexus injuries. Since the same retractor opening results in more chest distortion in short persons, this relationship was also examined in terms of patient height (Table II). Greater sternal retraction was usually required for operations that involved harvesting of the internal mammary artery. Of 206 patients who underwent dissection of this vessel, 11 (5.3%) had possible
plexus injury, but of these only four had symptoms referrable only to the side used. The right internal jugular vein was cannulated in 312 patients and the left in 23 patients after failure on the right side. Although all plexus injuries occurred in the patients with right-sided cannulation, the injury was equally likely to occur on the left side as on the right (Table III). Plexus injury occurred in 13 of 240 (5.4%) patients in whom the internal jugular vein was cannulated on the first attempt and after three of 95 (3.2%) cannulations necessitating multiple attempts. Trauma of internal jugular cannulation could not be related to plexus injury.
Discussion In its most severe form, postoperative brachial plexus palsy is characterized by flaccidity and numbness of the entire arm. 10 Less severe forms involve one or more roots or cords, with paralysis of the innervated muscle
The Journal of Thoracic and Cardiovascular Surgery
3 0 0 Tomlinson et al.
Table II. Relationship between patient height, degree of sternal retraction, and incidence of brachial plexus injury (differences by columns and rows were not statistically significant) Sternal retractor opening Patient height
7.5 em
8-/0.5 em
11-18 em
Tota/*
Less than 173 em More than 173 em
0/3t (0%) 1/7 (14%)
5/67 (7%) 3/55 (5%)
5/99 (5%) 2/95 (2%)
10/169 (6%) 6/157 (4%)
'Degree of sternal retractor opening was not recorded in nine patients, none of whom had injury. tFrequency of injury/number of patients.
Table m. Laterality of 16 brachial plexus injuries in 312 patients with cannulation of the right internal jugular vein Injuries Side of injury Right Left Both
No.
%
4 4 8
1.3 1.3 2.6
groups.'? Sensory loss is less prominent because of overlapping innervation, but paresthesias are common. Injury to the plexus during noncardiac operations is rare, with an estimated incidence of 0.03%,1l and tends to involve the more cephalad plexus roots." The mechanism is primarily stretching, less often compression, of the plexus because of malposition of the arm in relation to the shoulder and head.'? Plexus injury after median sternotomy differs in that it is common, with an estimated incidence of 5% to 38% despite care to avoid malposition of the arms and typically involves the lower plexus roots.t- s,12, 13 Because the hands-up position was associated with no compression and less tension on the plexus in the study reported by Jackson and Keats," we adopted this position 20 years ago for all supine patients but never documented its value in preventing plexus injury. In this prospective study of 335 patients, the incidence of all neurologic abnormalities was 4.8% (definite, probable, and possible). Severe plexus palsy occurred in only one patient (0.3%), and in only six others (1.8%) were objective signs present (probable). Most significantly, all patients were free of symptoms at the time of leaving the hospital except for one patient whose muscle weakness was improving rapidly. Comparison of these rates with those reported is handicapped by incomplete information, primarily failure to grade severity of injury and to report the position of the arms during operation. After median sternotomy with arms at the sides, Seyfer and associates' found motor and sensory neuropathies,
all involving the ulnar nerve, in 20 of 53 patients (38%) with disability persisting for an average of 2.3 months. After 312 coronary bypass operations reported by Shaw and colleagues," brachial plexopathies, all affecting the C g- T 1 roots, were present in 7% of patients. Hanson and co-workers" reported that 26 of 531 patients (5.0%) had brachial plexus injury after median sternotomy, with 22 injuries involving the C g- T 1 portion of the plexus. Sixteen of the 26 injuries were classified as moderate or severe, defined as disabling dysesthesia and pain or marked weakness lasting more than 3 months. Only Vander Salm, Cereda, and Cutler' studied the role of arm position on the incidence of this injury. After median sternotomy, plexus injury occurred in 14.5% of 90 patients with arms abducted to 90 degrees and in 23.5% of 90 patients with arms at their sides. The difference was not significant and severity of injury was not recorded. In a subsequent study to explore the role of rib fracture as the cause of plexus injury, 19 of 120 (15%) patients had plexus injury." Comparison of our data with these reports clearly indicates a low incidence of plexus injury in the hands-up position and a striking contrast in the short duration of symptoms and disability when injury occurred in our patients. . Although the mechanism of plexus injuries after median sternotomy is not clear, its high incidence typically related to the C g- T 1 distribution strongly suggests a role for sternal retraction in its cause. Direct trauma to the plexus from occult first rib fracture has been demonstrated.v' Wide sternal retraction in cadavers was found to push the clavicle into the retroclavicular space and rotate the first rib superiorly, which stretches the plexus.' By placing sternal retractor blades caudad and opening the retractor less widely, Vander Salm, Cutler, and Okike" reduced the severity but not the incidence of plexus injury in 120 patients. They concluded that minor plexus injuries probably result from stretching and compression but more serious injuries from rib fracture. Twelve of our patients had upper extremity symptoms unrelated to any nerve distribution. We believedysesthe-
Volume 94 Number 2 August 1987
Protection of brachial plexus during median sternotomy
sias of atypical distribution are unrelated to plexus injury and can be accounted for by operative trauma to bony structures of the shoulder, pressure neuropathy distal to the plexus, vascular insufficiency of the arms, trauma or hematoma at the radial artery cannulation site, and ischemia or phlebitis from vasoactive infusions. Our data convincingly fail to support any role for internal jugular cannulation in producing this injury. We found equal incidences of contralateral, ipsilateral, and bilateral injuries in 312 patients, all of whom had cannulation of the right internal jugular vein. Further, if trauma of cannulation does cause injury, then injury should be more frequent after multiple attempts. This was not the case even though the vast majority of cannulations were performed by relatively inexperienced physicians. In summary, we documented our impression of a low incidence and 'benign course of brachial plexus injury after median sternotomy when the hands-up position was used. We were unable to ascribe any role to internal jugular cannulation as a cause of this injury. We were also unable to identify the degree of retractor opening as an etiologic factor. This failure was probably related to the use of the. Cooley retractor with caudad placement of the blades and care to use the least retraction that provided an adequate surgical field. The hands-up position not only appears to prevent plexus injury from posterior displacement of the shoulder, but may minimize compression injury by first rib rotation during retraction of median sternotomy. The hands-up position is also of great advantage to the anesthesiologist in providing direct immediate access to intravenous and intra-arterial cannulas. We gratefully acknowledgethe advice, counsel, and support provided by Jack N. Alpert, M.D., Chief, Section of Neurology, St. Luke's Episcopal Hospital REFERENCES I. Kirsh MM, Magee KR, Gago 0, Kahn DR, Sloan H. Brachial plexus injury following median sternotomy incision. Ann Thorac Surg 1971;11:315-9.
30 1
2. Graham JG, Pye IF, McQueen INF. Brachial plexus injury after median sternotomy. 1 Neurol Neurosurg Psychol 1981;44:621-5. 3. Seyfer AE, Grammer NY, Bogumill GP, Provost 1M, Chandry U. Upper extremity neuropathies after cardiac surgery. 1 Hand Surg 1985;1OA:16-9. 4. Vander Salm TJ, Cereda JM, Cutler BS. Brachial plexus injury following median sternotomy. 1 THORAC CARDIOVASC SURG 1980;80:447-52. 5. Baisden CE, Greenwald LV, Symbas PN. Occult rib fractures and brachial plexus injury following median sternotomy for open-heart operations. Ann Thorac Surg 1984;38: 192-4.
6. Briscoe CE, Bushman lA, McDonald WI. Extensive neurologic damage after cannulation of the internal jugular vein. Br Med J 1974;1:314. 7. Paschall RM, Mandel S. Brachial plexus injury from percutaneous cannulation of the internal jugular vein. Ann Emerg Med 1983;12:58-60. 8. Hanson MR, Breuer AC, Furlan AJ, et al. Mechanism and frequency of brachial plexus injury in open heart surgery: a prospective analysis. Ann Thorac Surg 1983;36:675-9.
9. Jackson L, Keats AS. Mechanism of brachial plexus palsy following anesthesia. Anesthesiology 1965;26: 190-4. 10. Britt BA, Joy N, Mackay MB. Positioning trauma. In: Orkin FK, Cooperman LH, eds. Complications in anesthesiology. Philadelphia: lB Lippincott, 1983:648-54. 11. Dhuner KG. Nerve injuries following operations: survey of cases during a six-year period. Anesthesiology 1950; 11:289-93. 12. Vander Salm TJ, Cutler BS, Okike ON. Brachial plexus
injury following median sternotomy: part II. 1 THORAC CARDIOVASC SURG 1982;83:914-7. 13. Shaw Pl, Bates B, Cartlidge NEF, Heaviside D, lulian DG, Shaw DA. Early neurological complications of coronary artery bypass surgery. Br Med J 1985;291: 1384-7.
THORAC CARDIOVASC SURG
1987;94:297-301
Protecting the brachial plexus during median sternotomy Injuryto the brachial plexuswas prospectively assessedin 335 patients Wldergoing mediansternotomyfor cardiac operation. All patients were placed in the hands-up position(elbows elevated, arms abducted 90 degrees, and elbows flexed) after right internal jugular vein cannulation (23 cannulation-attempts were bilateral). Twenty-eight patients bad new upper extremity complaints after the operation, of whom 16 (4.8 % ) bad symptoms considered related to injury of the brachial plexus: one with generalized weakness of the left arm, six with localized weakness, pain, or paresthesia plus objective hypesthesia or weakness, and ninewith paresthesiasbut no objective signs. Four injurieswere right sided,four left sided, and eight bilateral. At the time of discbarge,15 of 16 weresymptomfree and the patient with generalized weakness wasrapidly improving. Postoperativeplexopathy was not related to degreeof sternal retraction, dissection of the internal mammary artery, or cannulationof the internal jugular vein. We believe the lowincidence and benign course of brachial plexus problems in these patients resulted from careful sternal retraction and use of the-bands-up position. Finally,our data do not support internal jugular cannulationas a major cause of plexus injuries after median sternotomy.
David L. Tomlinson, M.D., Irving A. Hirsch, M.D., S. V. Kodali, M.D., and Stephen Slogoff, M.D., Houston, Texas
Complaints of pain, weakness, numbness, and tingling of shoulders, arms, and hands are not unusual after operations requiring median sternotomy. These are commonly attributed to brachial plexus injury from stretch or compression by malposition of the arms or excessive sternal retraction."! Direct injury to the plexus byoccultfractures of the first rib because of improperly applied sternal retractors" S and by internal jugular vein cannulation'-" have been described. A recent prospective study of neurologic injury after median sternotomy suggested that internal jugular vein cannulation was the primary cause of brachial plexus injury after median sternotomy. Since we adopted the hands-up position of Jackson and Keats? in 1965, clinically obvious brachial plexus injury has been rare in our patients after median sternotomy. In addition, we have routinely catheterized theright internal jugular vein since 1975 without noting From the Division of Cardiovascular Anesthesiology, Texas Heart Institute, Houston, Texas. Received for publication July 24, 1986. Accepted for publication Aug. 24, 1986. Address for reprints: Stephen Slogoff, M.D., Cardiovascular Anesthesia, Texas Heart Institute, P.O. Box 20269, Houston, Texas 77025.
an increase in brachial plexus injury or other postoperative complaints referable to the right arm. To document our low incidence of nerve injury with the hands-up position, we performed a prospective study of this complication, which also failed to support any relationship between internal jugular vein cannulation and brachial plexus injury. Methods Three hundred sixty-five adult patients undergoing elective cardiac operations performed through a median sternotomy and necessitating central venouscannulation were prospectively evaluated before and after operation for symptoms and signs possibly related to brachial plexusdysfunction. Specificpreoperativeassessment (by LA.H. and D.L.T.) consisted of eliciting a detailed history of upper extremity pain, paresthesia, loss of sensationor weakness, assessmentof sensory responseto pin prick, and examination of motor function of muscle groups innervated by the nerves derived from the brachial plexus. These included adduction of the arm (C S_7) , forward thrust of the shoulder (C S_7) , elevation of the scapula (C), abduction, lateral and medial rotation of the arm (C s-s), flexion and supination of the forearm (C S-6) , ulnar flexion of the hand (C 7•S' T 1) , thumb apposition and flexion (C 7•s,T t ) , extension of the fore297
The Journal of
2 9 8 Tomlinson et al.
Fig. 1. A, A patient in the hands-up position on a foamwedge pad. The shoulders are elevated, elbows are 20 em above the horizontal, arms are abducted, and elbows flexed. Position is maintained by masking tape placed to avoid compression of the ulnar nerve at the elbow. B, The same patient is shown from above. Note that the arms are abducted to no more than 90% and the wedge and arm board are placed sufficiently cephalad to extend no further caudad than the arm. Thus the hands-up position provides ready access to arterial and venous cannulas without interfering with the surgeon's proximity to the patient.
arm and phalanges (C6-s), and radial extension of the hand and extension of the thumb (C 7•S) ' Eight patients were excluded because of symptoms or objective findings possibly related to central or peripheral nerve dysfunction. The first 50 patients were interviewed and examined by both observers to develop a uniform technique. After induction of anesthesia, the right internal jugular vein was cannulated with a 16 gauge, 5V
Thoracic and Cardiovascular Surgery
elevate elbows 20 em above the level of the operating room table (Fig. 1). Arms were abducted to no more than 90 degrees and the elbows were flexed. This position was maintained with masking tape placed to avoid compression of the ulnar nerve at the elbow. A Cooley sternal retractor with the crossbar caudad and retractor blades at the middle or lower third of the sternum was used in all patients. Patients were reexamined as described 5 to 7 days after operation. Any symptoms or objective signs not present on preoperative examination required confirmation by the second observer for inclusion. Patients with abnormal findings were examined at 2 day intervals until discharge. All abnormal observations were reviewe'd with a consulting neurologist to judge their relationship to injury of the brachial plexus. These relationships were defined for classification purposes as definite (extensive motor paralysis of the arm), probable (both subjective and objective findings attributable to injury of an anatomical portion of the plexus), possible (paresthesias without objective signs localized to established nerve distributions), and not related (symptoms or signs not related to an identifiable nerve distribution). Twenty-two patients examined preoperatively were excluded from the study because postoperative examination could not be performed: Four had central neurologic dysfunction, six were undergoing continued mechanical ventilation, 10 were discharged before the seventh day, and two died before the fifth day. The final group of 335 patients consisted of 244 who underwent coronary artery bypass, 58 who had valve replacement or repair, and 33 who had coronary artery bypass plus valve or other intracardiac procedure. All operations lasted less than 4 hours. Maximum perfusion time was 100 minutes and the average was 63 minutes. Statistical analysis of relative frequencies was by Fisher's exact test.
Results New symptoms or signs involving one or both upper extremities were reported by 28 (8.4%) patients (Table I). In 16 patients (4.8%) these new findings were considered related to brachial plexus injury: Of these, seven (2.1 %) were considered to be definitely (Patient 1, Table I) or probably related (Patients 2 to 7, Table I) and nine (2.7%) (Patients 8 to 16, Table I) were considered only possibly related. New symptoms of the remaining 12 patients (Patients 17 to 28, Table I) were not considered related to brachial plexus injury. Patient 17 complained of bilateral glove distribution paresthesias believed to be of hysterical origin. Patient 18, who had bilateral hand paresthesias on raising his arms, was
Volume 94 Number 2 August 1987
Protection of brachial plexus during median sternotomy
299
Table I. New symptoms, signs, or both, present 5 to 7 days after median sternotomy in 28 patients Patient
Condition at discharge
Involved side
Symptoms
Signs
L
Tingling, numbness, and weakness entire arm Generalized decrease in sensation and coordination Generalized pain
Profound weakness of all muscle groups; normal pin prick response
Significantly improving
Weakness of all muscle groups, hypesthesia C6-8,T, (L only)
Normal
Weakness on flexion and extension of forearms Weakness on abduction and adduction of fingers Hypesthesia C6-8,T, Hypesthesia C, Hypesthesia C,
Normal
2
LandR
3
LandR
4
LandR
5 6 7
LandR L R
8-12 13,14 15 16
Land R L R R
17
Land R
18
Land R
19-23
L
24,25
R
26 27 28
L R R
Decreased sensation and tingling C6-8,T, Tingling Co-8,T, Shoulder pain Shoulder pain on abduction Tingling C 6..,T, Tingling C 7_8,T, Tingling C'_8,T, Hypesthesia and tingling of shoulder, arm, and hand Tingling in glove distribution Tingling of hands on raising arms Generalized arm pain on movement Shoulder pain on abduction Generalized hand pain Tingling of thumb Tingling of fifth finger
Normal Normal Normal Normal
None None None None
Normal Normal Normal Normal
None
Normal
None
No change
None
Normal
None
Normal
None None None
Normal Normal Normal
Legend: L, Left. R, Right.
found to have markedly diminished radial pulses on abduction. Eight had generalized pain in the shoulder, arm, or hand not related to any nerve distribution. Patients 27 and 28 had paresthesias limited to a single finger on the right hand, which was used for intravenous infusion with a large-bore cannula. In all but one patient with plexopathy, symptoms and objective signs disappeared by the time of discharge 8 to 14 days after operation. Patient 1, with the greatest disability, left the hospital with residual weakness but was rapidly improving and had a functional arm. The extent of sternal retraction did not influence frequency of plexus injuries. Since the same retractor opening results in more chest distortion in short persons, this relationship was also examined in terms of patient height (Table II). Greater sternal retraction was usually required for operations that involved harvesting of the internal mammary artery. Of 206 patients who underwent dissection of this vessel, 11 (5.3%) had possible
plexus injury, but of these only four had symptoms referrable only to the side used. The right internal jugular vein was cannulated in 312 patients and the left in 23 patients after failure on the right side. Although all plexus injuries occurred in the patients with right-sided cannulation, the injury was equally likely to occur on the left side as on the right (Table III). Plexus injury occurred in 13 of 240 (5.4%) patients in whom the internal jugular vein was cannulated on the first attempt and after three of 95 (3.2%) cannulations necessitating multiple attempts. Trauma of internal jugular cannulation could not be related to plexus injury.
Discussion In its most severe form, postoperative brachial plexus palsy is characterized by flaccidity and numbness of the entire arm. 10 Less severe forms involve one or more roots or cords, with paralysis of the innervated muscle
The Journal of Thoracic and Cardiovascular Surgery
3 0 0 Tomlinson et al.
Table II. Relationship between patient height, degree of sternal retraction, and incidence of brachial plexus injury (differences by columns and rows were not statistically significant) Sternal retractor opening Patient height
7.5 em
8-/0.5 em
11-18 em
Tota/*
Less than 173 em More than 173 em
0/3t (0%) 1/7 (14%)
5/67 (7%) 3/55 (5%)
5/99 (5%) 2/95 (2%)
10/169 (6%) 6/157 (4%)
'Degree of sternal retractor opening was not recorded in nine patients, none of whom had injury. tFrequency of injury/number of patients.
Table m. Laterality of 16 brachial plexus injuries in 312 patients with cannulation of the right internal jugular vein Injuries Side of injury Right Left Both
No.
%
4 4 8
1.3 1.3 2.6
groups.'? Sensory loss is less prominent because of overlapping innervation, but paresthesias are common. Injury to the plexus during noncardiac operations is rare, with an estimated incidence of 0.03%,1l and tends to involve the more cephalad plexus roots." The mechanism is primarily stretching, less often compression, of the plexus because of malposition of the arm in relation to the shoulder and head.'? Plexus injury after median sternotomy differs in that it is common, with an estimated incidence of 5% to 38% despite care to avoid malposition of the arms and typically involves the lower plexus roots.t- s,12, 13 Because the hands-up position was associated with no compression and less tension on the plexus in the study reported by Jackson and Keats," we adopted this position 20 years ago for all supine patients but never documented its value in preventing plexus injury. In this prospective study of 335 patients, the incidence of all neurologic abnormalities was 4.8% (definite, probable, and possible). Severe plexus palsy occurred in only one patient (0.3%), and in only six others (1.8%) were objective signs present (probable). Most significantly, all patients were free of symptoms at the time of leaving the hospital except for one patient whose muscle weakness was improving rapidly. Comparison of these rates with those reported is handicapped by incomplete information, primarily failure to grade severity of injury and to report the position of the arms during operation. After median sternotomy with arms at the sides, Seyfer and associates' found motor and sensory neuropathies,
all involving the ulnar nerve, in 20 of 53 patients (38%) with disability persisting for an average of 2.3 months. After 312 coronary bypass operations reported by Shaw and colleagues," brachial plexopathies, all affecting the C g- T 1 roots, were present in 7% of patients. Hanson and co-workers" reported that 26 of 531 patients (5.0%) had brachial plexus injury after median sternotomy, with 22 injuries involving the C g- T 1 portion of the plexus. Sixteen of the 26 injuries were classified as moderate or severe, defined as disabling dysesthesia and pain or marked weakness lasting more than 3 months. Only Vander Salm, Cereda, and Cutler' studied the role of arm position on the incidence of this injury. After median sternotomy, plexus injury occurred in 14.5% of 90 patients with arms abducted to 90 degrees and in 23.5% of 90 patients with arms at their sides. The difference was not significant and severity of injury was not recorded. In a subsequent study to explore the role of rib fracture as the cause of plexus injury, 19 of 120 (15%) patients had plexus injury." Comparison of our data with these reports clearly indicates a low incidence of plexus injury in the hands-up position and a striking contrast in the short duration of symptoms and disability when injury occurred in our patients. . Although the mechanism of plexus injuries after median sternotomy is not clear, its high incidence typically related to the C g- T 1 distribution strongly suggests a role for sternal retraction in its cause. Direct trauma to the plexus from occult first rib fracture has been demonstrated.v' Wide sternal retraction in cadavers was found to push the clavicle into the retroclavicular space and rotate the first rib superiorly, which stretches the plexus.' By placing sternal retractor blades caudad and opening the retractor less widely, Vander Salm, Cutler, and Okike" reduced the severity but not the incidence of plexus injury in 120 patients. They concluded that minor plexus injuries probably result from stretching and compression but more serious injuries from rib fracture. Twelve of our patients had upper extremity symptoms unrelated to any nerve distribution. We believedysesthe-
Volume 94 Number 2 August 1987
Protection of brachial plexus during median sternotomy
sias of atypical distribution are unrelated to plexus injury and can be accounted for by operative trauma to bony structures of the shoulder, pressure neuropathy distal to the plexus, vascular insufficiency of the arms, trauma or hematoma at the radial artery cannulation site, and ischemia or phlebitis from vasoactive infusions. Our data convincingly fail to support any role for internal jugular cannulation in producing this injury. We found equal incidences of contralateral, ipsilateral, and bilateral injuries in 312 patients, all of whom had cannulation of the right internal jugular vein. Further, if trauma of cannulation does cause injury, then injury should be more frequent after multiple attempts. This was not the case even though the vast majority of cannulations were performed by relatively inexperienced physicians. In summary, we documented our impression of a low incidence and 'benign course of brachial plexus injury after median sternotomy when the hands-up position was used. We were unable to ascribe any role to internal jugular cannulation as a cause of this injury. We were also unable to identify the degree of retractor opening as an etiologic factor. This failure was probably related to the use of the. Cooley retractor with caudad placement of the blades and care to use the least retraction that provided an adequate surgical field. The hands-up position not only appears to prevent plexus injury from posterior displacement of the shoulder, but may minimize compression injury by first rib rotation during retraction of median sternotomy. The hands-up position is also of great advantage to the anesthesiologist in providing direct immediate access to intravenous and intra-arterial cannulas. We gratefully acknowledgethe advice, counsel, and support provided by Jack N. Alpert, M.D., Chief, Section of Neurology, St. Luke's Episcopal Hospital REFERENCES I. Kirsh MM, Magee KR, Gago 0, Kahn DR, Sloan H. Brachial plexus injury following median sternotomy incision. Ann Thorac Surg 1971;11:315-9.
30 1
2. Graham JG, Pye IF, McQueen INF. Brachial plexus injury after median sternotomy. 1 Neurol Neurosurg Psychol 1981;44:621-5. 3. Seyfer AE, Grammer NY, Bogumill GP, Provost 1M, Chandry U. Upper extremity neuropathies after cardiac surgery. 1 Hand Surg 1985;1OA:16-9. 4. Vander Salm TJ, Cereda JM, Cutler BS. Brachial plexus injury following median sternotomy. 1 THORAC CARDIOVASC SURG 1980;80:447-52. 5. Baisden CE, Greenwald LV, Symbas PN. Occult rib fractures and brachial plexus injury following median sternotomy for open-heart operations. Ann Thorac Surg 1984;38: 192-4.
6. Briscoe CE, Bushman lA, McDonald WI. Extensive neurologic damage after cannulation of the internal jugular vein. Br Med J 1974;1:314. 7. Paschall RM, Mandel S. Brachial plexus injury from percutaneous cannulation of the internal jugular vein. Ann Emerg Med 1983;12:58-60. 8. Hanson MR, Breuer AC, Furlan AJ, et al. Mechanism and frequency of brachial plexus injury in open heart surgery: a prospective analysis. Ann Thorac Surg 1983;36:675-9.
9. Jackson L, Keats AS. Mechanism of brachial plexus palsy following anesthesia. Anesthesiology 1965;26: 190-4. 10. Britt BA, Joy N, Mackay MB. Positioning trauma. In: Orkin FK, Cooperman LH, eds. Complications in anesthesiology. Philadelphia: lB Lippincott, 1983:648-54. 11. Dhuner KG. Nerve injuries following operations: survey of cases during a six-year period. Anesthesiology 1950; 11:289-93. 12. Vander Salm TJ, Cutler BS, Okike ON. Brachial plexus
injury following median sternotomy: part II. 1 THORAC CARDIOVASC SURG 1982;83:914-7. 13. Shaw Pl, Bates B, Cartlidge NEF, Heaviside D, lulian DG, Shaw DA. Early neurological complications of coronary artery bypass surgery. Br Med J 1985;291: 1384-7.