- Email: [email protected]
The microcirculatory effects of peripheral sympathectomy
The Microcirculatory Effects of Peripheral Sympathectomy L. Andrew Koman, MD, Beth P. Smith, PhD, F. Edward Pollock, Jr, MD, Thomas L. Smith, PhD, David Pollock, MD, Gregory B. Russell, MS, Winston-Salem, NC Microvascular physiology following peripheral artery sympathectomy was evaluated in sevon hands with refracto R pain (n = 7) and ulceration (n = 7) by serial isolated cold stress testing, which measures digital temperature and lcutaneous perfusion (laser Doppler fluxmetry). All patients (n = 6) had vasospasm (secondary Raynaud's phenomenon) and arteriographically proven digital and palmar occlusive disease. Microcirculatory flow responses were correlated with symptoms and signs (including ulcer healing) before and after (2-8 weeks, 12-15 weeks, and 24 weeks) peripheral sympathectomy. Baseline data were compared with those of controis (n = 7 extremities). Following surgery, all seven hands had diminished pain; six had ulcer healing and one had ulcer reduction. Isolated cold stress testing demonstrated abnormalities in temperature and laser Doppler fluxmetry response between patients and controls. Although total flow (reflected by temperature) was not significantly increased after surgery, peripheral sympathectomy increased nutritional flow in these patients with combined vasospastic vessels and occlusive injury. The clinical changes observed following peripheral sympathectomy appear to be related to postsurgical correction of abnormal arteriovenous shunting and to improved nutritional blood flow to ischemic areas. This accounts for the resultant diminution of pain and healing of ulcers observed in these patients after surgery. (J Hand Surg 1995; 20A:709-717.)
T h e t r e a t m e n t o f chronic v a s o s p a s t i c digital s y m p t o m s o f cold intolerance and pain associated with ulceration(s) and/or infection(s) that are refractory to oral p h a r m a c o l o g i c interventions is an extremely difficult clinical problem. Historically, c e r v i c o t h o r a c i c s y m p a t h e c t o m y was a c o m m o n treatment if nonoperative m a n a g e m e n t of s y m p t o m s failed.~ 3 However, this treatment provided only
From the Department of Orthopaedic Surgery, Division of Surgical Sciences, and the Departmentof Public Health Sciences, BowmanGray School of Medicine of WakeForest University,Winston-Salem,NC. This work was aided by a grant from the Orthopaedic Research and Education Foundation, Chicago, IL, and the American Society for Surgery of the Hand, Aurora, CO. Receivedfor publicationJuly 29, 1993; acceptedin revisedformFeb. 27, 1995. No benefitsin any formhave been receivedor will be receivedfrom a commercialpartyrelateddirectlyor indirectlyto the subjectof this article. Reprint requests: L. Andrew Koman, MD, Department of Orthopaedic Surgery, Bowman Gray School of Medicine, Medical Center Boulevard,Winston-Salem,NC 27157-1070.
transient i m p r o v e m e n t , and in 1980~ Flatt ~ suKgested the use o f a distal interrupti'ort o f the periarterial sympathetic fibers and digital nerve branches to the c o m m o n and proper digital arteries (peripheral sympathectomy). This technique has orovided significant palliation in selected paflents. 4~ This study evaluates the iatermediate (6 month) effects of peripheral s y m p a t h e c t o m y on imcLoclrculatory physiology in a well-defined subset of patients with combined vasospastic disease ( R a y n a u d ' s phen o m e n o n ) secondary to collagen vascular disease (eg, scleroderma) and 2-level unreconstructable occlusive disease including proximal thrombosis of t i ~ ulnar artery and/or superficial arch and distal thrombosis of a proper or c o m m o n digital artery. This subset of symptomatic patients with vasoocclusive disease refractory to nonoperative interventions provides a unique biologic model for evaluating the m i c r o c i r c u l a t o r y events that follow peripheral sympathectomy.
The Journal of Hand Surgery
709
710 Komanet al. / Microcirculatory Effects of Peripheral Sympathectomy
Materials and Methods Six patients (seven hands) were included in this study. Criteria for inclusion were painful ulcerated digits unresponsive to pharmacologic or other nonoperative management, Raynaud's phenomenon secondary to a systemic process, and arteriographic documentation of unreconstructible arterial lesions. All patients were women; their average age was 40 years (range, 32-45). Diagnoses included scleroderma (n = 5) and discoid lupus (n = 1). In none of the six had pharmacologic interventions (including alpha-blocking agents, beta-blocking agents, and calcium channel-blocking agents) been successful in alleviating symptoms.
Arteriographic Findings All patients had bilateral upper extremity arteriograms performed by the Department of Radiology, Bowman Gray School of Medicine, Wake Forest University. Arteriography was used to evaluate the aortic arch, the proximal vasculature, and the wrist, hand, and digits (including the common and proper digital vessels). The femoral access route was used and vasospasm was controlled with intra-arterial priscoline and/or nitroglycerin. The entire vascular tree from the aorta to the finger tip was evaluated. Selective injections of distal vessels were enhanced by the use of isomolar contrast materials, anfispasm agents (priscoline and nitroglycerin), and subtraction techniques. Areas of occlusion were documented and classified by region (forearm, wrist, hand, and digit), and quality of distal perfusion was graded (Fig. 1, Table 1). In each hand, the digit selected for study was the digit with a nonhealing ulcer or the digits with the most symptoms. Patients were excluded from the study if any occlusive or embolic events were identified above the proximal forearm. Operative Technique A modification of the peripheral sympathectomy technique of Flatt,' Wilgis, 6 and Miller and Morgan 7 was used in all seven hands. During this procedure, the radial artery and ulnar artery at the wrist were exposed under magnification (3.5x) through two parallel longitudinal incisions; the vessels were identified over a length of 2.5 cm (Fig. 2). One to 2 cm of adventitia was dissected from the vessels, and peripheral nerve connections were transected under magnification (6x to 30x). The superficial palmar arch and the three anterior common digital arteries were identified, mobilized, and dissected through an oblique incision in the palm; branches from adjacent
Figure 1. Localization of arterial occlusion was identified in each study patient by division of the hand into seven zones (A through G; proximal to distal).
digital nerves were transected. One to 1.5 cm of adventitia was then dissected under magnification (6.25x to 30x) (Fig. 2). The proper digital arteries were not identified during surgery.
Clinical Evaluation Preoperative signs and symptoms were compared with postoperative signs and symptoms during the study period. Pain and cold intolerance were rated as mild, moderate, or severe; finger tip ulcerations were rated as unchanged, improved, or healed.
Physiologic Evaluation Isolated cold stress testing (ICST) was performed at four time points: preoperatively, and postoperatively at 2 to 8 weeks, at 3 months (___2weeks), and at 6 months (_+2 weeks). All tests were performed with the patient in a thermoneutral environment (22~ to 24~ Data were accumulated during three phases of the ICST: during baseline, during a 20-minute cooling phase, and during 20 minutes of recovery from the cold stress (rewanning). ~
The Journal of Hand Surgery / Vol. 20A No. 5 September 1995 711
Table 1. Individual Patient Data
Patient/ Age/Sex
Location of Ulcers
Location of Probe
1/45/F
Right index
Index
2/33/F
Left index
Middle
3/32/F
Left index
Index
4/44/F
Necrotic left middle
Middle
5/45/F
Left index Left middle
Middle
6/40/F
Left index
Index
Type of Surgery
Level of Occlusion (see Fig. 1) Forearm-A
Right peripheral sympathectomy 12/30/90 ~ - - - ~ Left peripherfl/ "~.//1) sympathe~tomy 3 / 3 / 9 ~ ~ Left periphe?al-------J/j" sympathectomy 2/18/90 Left peripheral sympathectomy; amputation of left middle fingertip through the DIP joint 1/10/91 Left peripheral Ulnar-6 cm sympathectomy from origin 12/13/90 Left peripheral sympathectomy 3/3/92
Hand-C
Digits-DEFG IRE, IUE, MRE, RRE, LUF IRE, IUE, MRE, MUF, LUF TE, IRE, IUE, MRE, RRE, RUF, LUF IRE, MRE, MUE, RRF, RUE, LRE, LUE
No superficial arch (deep palmer arch intact)
I (R&U), D M (R&U), D R (R&U), D L (R&U), D IRF, IUE, MRF, LRF, LUE
IRE, index, radial, zone E; IUE, index, ulnar, zone E; MRE, middle, radial, zone E; RRE, ring, radial, zone E; LUF, little, ulnar, zone F; TE, thumb, zone E; RUF, ring, ulnar, zone F; LRE, little, radial, zone E.
Digital Blood Flow. The microcirculation (total blood flow and thermoregulatory blood flow) was assessed by monitoring digital pulp temperatures and digital pulp cutaneous perfusion, as evaluated by laser Doppler fluxmetry (LDF) before, during, and after ICST9-12; both measurements were recorded simultaneously. The selected digit was symptomatic and had documented distal occlusive disease. Vasomotion was monitored by LDF, using a Periflux Model PF3 perfu-
sion monitor (Perimed, Stockholm, Sweden) and a conventional flow probe attached to one digit of each hand. Laser Doppler fluxmetry evaluates microvascular blood flux, that is, cutaneous perfusion, beneath the skin at a 1.0-mm depth, ~3and also provides a measure of vasomotion in the nutritional capillary loops and most of the superficial arteriovenous anastomoses. Output from the LDF was collected using a custom software package (CIM Solutions, Advance, NC) with
Figure 2. (Top) During modified peripheral sympathectomy technique, the radial and ulnar arteries at the wrist and the superficial palmar arch and three anterior common digital arteries are exposed. The adventitia and direct peripheral nerve connection are then dissected from the vessels under from 6.25x to 30.0x magnification. (Bottom) Close-up view of an artery and associated nerve demonstrating peripheral nerve connections to distal arteries, which are disrupted by the surgical technique.
712
Koman et al. / Microcirculatory Effects of Peripheral Sympathectomy
readings taken at 0.1-second sampling intervals. LDF curves were generated to characterize microvascular flux during the three phases of the test. Digital cooling was accomplished by having the patients insert their hands into a modified refrigeration unit (Sears Roebuck 3.7 cu. ft. refrigerator) equipped with special portholes (Fig. 3). The temperature within the refrigeration unit was maintained at between 5~ and 8~ Digital pulp temperatures of all fingers were recorded simultaneously on-line via small surface temperature probes (#427 thermistors, Yellow Springs Instrument Company, Yellow Springs, OH). Digital temperature readings were monitored and recorded with custom computer software; measurements were taken at 10-second sampling intervals. Temperature curves were generated to characterize digital blood flow during the test. Normal Control Subjects. Digital temperature and LDF measurements during an ICST evaluation were collected from seven female control subjects (seven hands) following the previously described ICST prot o c o 1 . 9-~2 The control subjects were nonsmokers who had no prior history of trauma or vascular disorder in
their hands or arms; their average age was 29 years (range, 18--43). Data Analysis. Data analysis was performed using a univariate repeated-measures analysis of variance for the between-group comparisons over time. Comparisons were made between treatments (ie, before and after peripheral sympathectomy) and between phases of the isolated cold stress test (ie, baseline vs cooling vs rewarming). Temperature and laser Doppler measurements were averaged over five time periods in the following manner. Baseline measurements were avera g e d over the 5-minute baseline period of the test. During the cooling phase of the test, measurements were collected during minutes 6 to 10 (first 5 minutes of cooling) and minutes 20 to 25 (last 5 minutes of cooling). Measurements for each of the two time periods were averaged to give an average for the first and last 5 minutes of cooling. Similarly, during the rewarming phase of the test, measurements from minutes 26 to 31 (first 5 minutes of rewarming) and minutes 40 to 45 (last 5 minutes of rewarming) were collected and averaged. Digital temperatures and LDF measurements were compared between control sub-
Lamer probestors thermi &
Thermistor~ Figure 3. Schematic representation of the equipment necessary to perform the combined isolated cold stress test (ICST) and laser Doppler fluxmetry (LDF) evaluation. Thermistors are attached to all 10 digits and the laser Doppler probe is attached to one digit of each hand. Measurements of digital temperature and LDF are recorded with custom computer software. (From Pollock et al.8 with permission.)
The Journal of Hand Surgery / Vol. 20A No. 5 September 1995
jects and patients before and after peripheral sympathectomy. Microvascular perfusion was correlated with symptoms and ulcer healing.
Results Clinical Results
Following peripheral sympathectomy, all patients reported decreased pain and improved cold tolerance; finger tip ulcerations healed (n = 6) or improved (n = 1). M i c r o v a s c u l a r Results
The normal microvascular response to the cold stress test was characterized by a decrease in digital tempera-
713
ture during cooling, followed by a gradual but incomplete increase in digital temperature during rewarming (Fig. 4A). There was no significant difference in digital temperature between the patients and controls before surgery (p = .0687) or at 6 months after surgery (p = .2133). Laser Doppler fluxmetry responses during stress documented significant differences between controis and patients before surgery (decreased LDF) (Fig. 4B), which no longer had statistical significance (p = .0955) at 6 months (Fig. 4B) These data suggest that peripheral sympathectomy normalizes the digital microcirculatory response to stress in digits compromised by vaso-occlusive involvement, as reflected by LDE Preoperative versus Postoperative. Analysis of digital ICST before and after peripheral sympathec-
30. 9
NOm(~,L CONTIt01~q N - ? DIGITS
25.
9
PATII~: PBB-SUBGI~.'Y
20,
9 P&TIZNTS: 6 MONTHS p O S T - O P N - 6 DIGIT5
N = 7 DIGIT5
15-
10,
5
I
I
' cootma
'
' PA'W~U~MmG
l
A
TIME
350 9 NOla~LI~. CONTI~OI~ N - ? DIGrm 9 PAT/]roTS: Pla-SUltC~3~ N - ? DIG~TS
250.
9 PATI]~TS: 6 MONTHS P 0 6 T - O p N - 6 DIGITS
.~ 200,
150.
100.
50.
0
B
TIME
Figure 4. (A) Graphs of digital temperature response during ICST of normal, control hands (n = 7) and hands of patients with vaso-occlusive disease (n = 7). (B) Graphic representation of the LDF data during ICST of hands in A. Values represent the average value at each point, + SEM.
714
Koman et al. / Microcirculatory Effects of Peripheral Sympathectomy
tomy confirms improvement in microcirculatory response (Fig. 5). There was a significant difference (p = .0188) between the LDF measurements before and after surgery; however, there was no significant difference between preoperative and postoperative digital temperatures. The microcirculatory response to peripheral sympathectomy in one patient is documented in Figure 6, in which the boxed line represents the environmental temperature to which hands were exposed during the three phases of the test. During the preoperative evaluation, baseline (-5 to 0 min) temperatures were below 26~ in all digits on the right hand, with the index finger exhibiting the
lowest temperatures (between 18~ and 20~ (Fig. 6A). During the cooling phase (0 to 20 min), digital temperatures fell rapidly, with the temperature of the index finger again being lowest. During rewarming (20 to 40 min), there was a slow but complete return to baseline temperatures in all but the index digit. LDF was abnormally low throughout the test, reflecting minimal cutaneous perfusion at baseline; LDF measurements of perfusion were not detectable during cooling and were very low during rewarming. Perfusion began to rise above 0 at 35 minutes, Postoperatively, at both 3 and 6 months, all digital temperatures increased throughout all phases of the
?
v
~nna
9
r~rarm I
4, r ~ n
I
2
10
5
PRE-OP N = 7 DIGITS
A
2-8 ~ POST-OP N = 7 DIGITS
3 MONTHS POST-OP N = 5 DIGITS
8 MONTHS POST-OP N6 DIGITS
140 9
bw~hw
120
g
m ~ a
0
rmr~m
2
A
20
B
PRE-OP N ~ 7 DIGITS
2-8 ~ POST-OP N = 7 DIGITS
3 MONTHS POST OP N = 5 DIGITS
8 MONTHS POST-OP N = 5 DIGITS
Figure 5. (A) The average response of digital temperature during the three phases (baseline, cooling, and rewarming) of the ICST. Each of the four graphs represents the average temperature response of the most severely involved digits during ICST. Results from tests performed at four different time intervals are plotted: (1) preoperative, (2) postoperative (2-8 weeks), (3) postoperative (3 months), and (4) postoperative (6 months). (B) The average LDF values of the most severely affected digit during the three phases of ICST are plotted. Preoperative values (baseline) are compared with tests performed postoperatively at (1) 2 to 8 weeks, (2) 3 months, and (3) 6 months. Values represent the average value at each point, _+SEM.
~ ~.g~.~. ~ ~ . ~ ='~
~
"~
el)
~" , ,
9-=
~
=
"
~
~
o ~
E~ ~ ' ~
~- "~
~
9
8- ~ ~:
o
. "~
~E
,,,,~,~-,,,,,,,,,
,,,,
,,,,
,,,,,,,
~
~
~
~
~.~ ~ ~ ~-,~~ ~ ~ ~.~'~
-
~
-~
-
~
,~
~ ~
B"
~ ~ 0
~
~:~
cq
.~ .~
~
~.~
~
9, " ~
~
~
,-
~
~
~
o
~
~
.~
~ .~
~=
~
~
~
9
9
~
Ii
.,i ! ~.~ ~ *'~
0
r
e ~ [ .~~~
.= g ~
(0o) e~nl~uedkuej.
(^UJ) xnl..I jelddO[] .le~e]
~.~~.~
~ '~
716
Koman et al. / Microcirculatory Effects of Peripheral Sympathectomy
test, and the previously observed rapid decline in temperature during cooling was absent (Fig. 6B, C). At 6 months, the index finger temperature remained lower than the other digits but, even during cooling, was still 8~ warmer than it had been preoperatively. Postoperative LDF values at 3 and 6 months were increased over baseline values, indicating improved cutaneous perfusion and vasomotion throughout the entire test (Fig. 6B, C).
Discussion The human digit serves as a thermoregulatory end organ. In this capacity, its blood flow exhibits cyclical increases and decreases, the magnitude and frequency of which' are determined by environmental temperature, t4 Microvascuiar blood flow in the digits is characterized by two components: (1) thermoregulatory flow, which is modulated by arteriovenous anastomoses and contributes to body temperature control; and (2) nutritional blood flow, which is provided by papillary capillaries and maintains tissue viabilityY 6 Chronic arterial insufficiency of the upper extremity often results in digital pain, cold intolerance, and digital tip ulceration, which are refractory to nonoperative management. Peripheral sympathectomy4'6'~2'~7 is performed by isolating the terminal branches of the sympathetic nerves, dividing these terminal branches, and stripping the adventitia from the appropriate arteries. The theoretic basis for employing this surgical procedure is that it produces a more distal interruption of sympathetic innervation to the blood vessels to decrease norepinephrine release at the myoneural junction in the vessel wall, to eliminate vasospasm, and to dilate arterial smooth muscle. 7,9Peripheral sympathectomy has several theoretic advantages over more proximal cervicothoracic sympathectomy: it eliminates sympathetic fibers that bypass the sympathetic trunk~8; its effect lasts longer'9; and it often is effective when central or cervicothoracic sympathectomy has failed. 6,17,18 Peripheral sympathectomy of the digital artery has been advocated when nonoperative treatment has failed and disabling symptoms and/or finger tip ulceration persist? '6 Clinical relief is believed to result from diminished vasospasm, with resultant dilatation of arterial smooth muscle. ~9Although sympathetic nerve interruption decreases the release of norepinephrine at the myoneural junction, 19 that mechanism fails t o explain why intermediate-term palliation is better with distal interruption than with
more proximal sympathectomy, and it does not address the microcirculatory changes that accompany the procedure. Furthermore, published clinical studies have failed to differentiate patients with vasospastic disease alone from those with combined vasospastic and occlusive involvement. The effects of peripheral sympathectomy on microvascular blood flow in vaso-occlusive digits have not, to our knowledge, been described previously. Previous investigators have demonstrated clinical improvement following peripheral sympathectomy, documenting increased digital t e m p e r a t u r e 3,4,6,~7 and improved pulse volume recordings 7 before and after cold stress. Our data suggest that distal peripheral sympathectomy maximizes cutaneous perfusion in the presence of arterial occlusion: temperature, which provides an index of total pulp flow, did not increase significantly over the short term or over a 6month postoperative follow-up period; and LDE which provides an assessment of cutaneous microvascular perfusion and vasomotion, increased significantly. These observations support a possible preferential increase in flow into nutritional rather than thermoregulatory vessels following peripheral sympathectomy. If that is the case, then nutritional flow at sufficient volume to maintain cellular integrity would explain a decrease in symptoms and would allow ulcerations to heal. Conversely, increased total flow with inappropriate arteriovenous shunting proximal to nutritional papillary capillaries would not improve symptoms. However, caution must be used in applying these observations to digits without occlusive disease, since both total flow and nutritional flow may be increased by the partial interruption of sympathetic tone. This finding m a y explain the significant temperature changes reported by others after peripheral sympathectomy. 3,4,14,~7Until some of these points are settled, interpretation of nonstressed temperature data requires caution. 12 Data from this preliminary study illustrate the manner in which microvascular perfusion improves following peripheral sympathectomy. The use of nonimmersion thermal stress is important since nonstressed data are difficult to interpret? 2 ICST, using both temperature and LDF, allows documentation of physiologic effects of peripheral sympathectomy during a prospective follow-up period. This study demonstrates that peripheral sympathectomy has a direct effect on peripheral microcirculation in digits with vaso-occlusive disease. The microcirculatory changes observed following periph-
The Journal of Hand Surgery / Vol. 20A No. 5 September 1995 717 eral s y m p a t h e c t o m y c o i n c i d e d with clinical i m p r o v e m e n t in pain and cold intolerance, and with healing of previously recalcitrant digital ulcers. The data suggest that these clinical findings are secondary to i m p r o v e m e n t in nutritional microcirculatory flow as reflected in increased L D F measurem e n t s . C o m p a r i s o n s between the responses of the vaso-occlusive digits and those of the normal control digits support the observation that the decline o f cutaneous perfusion b e l o w a certain critical level results in ulceration, pain, and cold intolerance. However, if nutritional cutaneous perfusion can be increased a b o v e this critical level, digital circulation is then adequate to heal the ulcers and to reduce pain and cold intolerance. In this study, such i m p r o v e m e n t s were o b s e r v e d even though cutaneous perfusion in these patients was not as high as that in the control subjects. Peripheral sympathectomy performed by adventitial stripping of sympathetic fibers and peripheral neural connections to the radial and ulnar artery at the wrist, the superficial palmar arch, and the three c o m m o n palmar distal arteries results in (1) clinical i m p r o v e m e n t in patients with R a y n a u d ' s phenomenon secondary to systemic disease and acquired and unreconstructible arterial occlusion; (2) i m p r o v e m e n t in cutaneous perfusion, as reflected b y L D F for up to 6 months after surgery; and (3) increases in total flow, as reflected by increased digital temperature during ICST.
References 1. Barcroft H, Walker AJ. Return of tone in blood vessels of the upper limb after sympathectomy. Lancet 1949; 1:1035-9. 2. Koman LA, Urbaniak JR. Ulnar artery insufficiency: a guide to treatment. J Hand Surg 1981;6:16--24. 3. E1-Gammal TA, Blair WF. Digital periarterial sympathectomy for ischemic digital pain and ulcers. J Hand Surg 1991;16B:382-5.
4. Hatt AE. Digital artery sympathectomy. J Hand Surg 1980; 5:550-6. 5. Jones NF. Acute and chronic ischemia of the hand: pathophysiology, treatment and diagnosis. J Hand Surg 1991; 16: 1074-83. 6. Wilgis EFS. The evaluation and treatment of chronic digital ischemia. Ann Surg 1981;193:693-8. 7. Miller LM, Morgan RF. Vasospastic disorders: etiology, reconstruction, and treatment. Hand Clin 1993;9:171-87. 8. Pollock FE Jr, Koman LA, Smith BP, Holden M, Russell GB, Poehling GG. Measurement of hand microvascular blood flow with isolated cold stress testing and laser Doppler fluxmetry. J Hand Surg 1993;18A: 143-50. 9. Koman LA, Nunley JA, Goldman JL, Seaber AV, Urbaniak JR. Isolated cold stress testing in the assessment of symptoms in the upper extremity: preliminary communications. J Hand Surg 1984;9A:305-13. 10. Koman LA, Nunley JA. Thermoregulatory control after upper extremity replantation. J Hand Surg 1986;11A: 548-52. 11. Nunley JA, Penny WH III, Woodbury MA, Koman LA. Quantitative analysis of cold stress performance after digital replantation. J Orthop Res 1990;8:94-100. 12. Koman LA, Smith BP, Smith TL. Stress testing in the evaluation of upper extremity perfusion. Hand Clin 1993;9: 59-83. 13. Tenland T. On laser Doppler flowmetry. Methods and microvascular applications. Link6ping Studies in Science and Technology Dissertations, No. 83. Link6ping, Sweden: Link6ping University, 1982:7-51. 14. BurtOn AC. The range and variability of the blood flow in human fingers and the vasomotor regulation of body temperature. Am J Physiol 1939;127:437-53. 15. Conrad MC, Functional anatomy of the circulation to the lower extremities. Chicago: Year Book Medical Publishers, 1971:190. 16. Fagrell B. Dynamics of skin microcirculation in humans. J Cardiovasc Pharmacol 1985;7(suppl. 3):$53-8. 17. Egloff DV, Mifsud RP, Verdan C. Superselective digital sympathectomy in Raynaud's phenomenon. Hand 1982; 15:110--4. 18. Pick J. The autonomic nervous system: morphological, comparative, clinical and surgical aspects. Philadelphia: JB Lippincott, 1970. 19. Wilgis EFS. Digital sympathectomy for vascular insufficiency. Hand Clin 1985;1:361-7.
T h e t r e a t m e n t o f chronic v a s o s p a s t i c digital s y m p t o m s o f cold intolerance and pain associated with ulceration(s) and/or infection(s) that are refractory to oral p h a r m a c o l o g i c interventions is an extremely difficult clinical problem. Historically, c e r v i c o t h o r a c i c s y m p a t h e c t o m y was a c o m m o n treatment if nonoperative m a n a g e m e n t of s y m p t o m s failed.~ 3 However, this treatment provided only
From the Department of Orthopaedic Surgery, Division of Surgical Sciences, and the Departmentof Public Health Sciences, BowmanGray School of Medicine of WakeForest University,Winston-Salem,NC. This work was aided by a grant from the Orthopaedic Research and Education Foundation, Chicago, IL, and the American Society for Surgery of the Hand, Aurora, CO. Receivedfor publicationJuly 29, 1993; acceptedin revisedformFeb. 27, 1995. No benefitsin any formhave been receivedor will be receivedfrom a commercialpartyrelateddirectlyor indirectlyto the subjectof this article. Reprint requests: L. Andrew Koman, MD, Department of Orthopaedic Surgery, Bowman Gray School of Medicine, Medical Center Boulevard,Winston-Salem,NC 27157-1070.
transient i m p r o v e m e n t , and in 1980~ Flatt ~ suKgested the use o f a distal interrupti'ort o f the periarterial sympathetic fibers and digital nerve branches to the c o m m o n and proper digital arteries (peripheral sympathectomy). This technique has orovided significant palliation in selected paflents. 4~ This study evaluates the iatermediate (6 month) effects of peripheral s y m p a t h e c t o m y on imcLoclrculatory physiology in a well-defined subset of patients with combined vasospastic disease ( R a y n a u d ' s phen o m e n o n ) secondary to collagen vascular disease (eg, scleroderma) and 2-level unreconstructable occlusive disease including proximal thrombosis of t i ~ ulnar artery and/or superficial arch and distal thrombosis of a proper or c o m m o n digital artery. This subset of symptomatic patients with vasoocclusive disease refractory to nonoperative interventions provides a unique biologic model for evaluating the m i c r o c i r c u l a t o r y events that follow peripheral sympathectomy.
The Journal of Hand Surgery
709
710 Komanet al. / Microcirculatory Effects of Peripheral Sympathectomy
Materials and Methods Six patients (seven hands) were included in this study. Criteria for inclusion were painful ulcerated digits unresponsive to pharmacologic or other nonoperative management, Raynaud's phenomenon secondary to a systemic process, and arteriographic documentation of unreconstructible arterial lesions. All patients were women; their average age was 40 years (range, 32-45). Diagnoses included scleroderma (n = 5) and discoid lupus (n = 1). In none of the six had pharmacologic interventions (including alpha-blocking agents, beta-blocking agents, and calcium channel-blocking agents) been successful in alleviating symptoms.
Arteriographic Findings All patients had bilateral upper extremity arteriograms performed by the Department of Radiology, Bowman Gray School of Medicine, Wake Forest University. Arteriography was used to evaluate the aortic arch, the proximal vasculature, and the wrist, hand, and digits (including the common and proper digital vessels). The femoral access route was used and vasospasm was controlled with intra-arterial priscoline and/or nitroglycerin. The entire vascular tree from the aorta to the finger tip was evaluated. Selective injections of distal vessels were enhanced by the use of isomolar contrast materials, anfispasm agents (priscoline and nitroglycerin), and subtraction techniques. Areas of occlusion were documented and classified by region (forearm, wrist, hand, and digit), and quality of distal perfusion was graded (Fig. 1, Table 1). In each hand, the digit selected for study was the digit with a nonhealing ulcer or the digits with the most symptoms. Patients were excluded from the study if any occlusive or embolic events were identified above the proximal forearm. Operative Technique A modification of the peripheral sympathectomy technique of Flatt,' Wilgis, 6 and Miller and Morgan 7 was used in all seven hands. During this procedure, the radial artery and ulnar artery at the wrist were exposed under magnification (3.5x) through two parallel longitudinal incisions; the vessels were identified over a length of 2.5 cm (Fig. 2). One to 2 cm of adventitia was dissected from the vessels, and peripheral nerve connections were transected under magnification (6x to 30x). The superficial palmar arch and the three anterior common digital arteries were identified, mobilized, and dissected through an oblique incision in the palm; branches from adjacent
Figure 1. Localization of arterial occlusion was identified in each study patient by division of the hand into seven zones (A through G; proximal to distal).
digital nerves were transected. One to 1.5 cm of adventitia was then dissected under magnification (6.25x to 30x) (Fig. 2). The proper digital arteries were not identified during surgery.
Clinical Evaluation Preoperative signs and symptoms were compared with postoperative signs and symptoms during the study period. Pain and cold intolerance were rated as mild, moderate, or severe; finger tip ulcerations were rated as unchanged, improved, or healed.
Physiologic Evaluation Isolated cold stress testing (ICST) was performed at four time points: preoperatively, and postoperatively at 2 to 8 weeks, at 3 months (___2weeks), and at 6 months (_+2 weeks). All tests were performed with the patient in a thermoneutral environment (22~ to 24~ Data were accumulated during three phases of the ICST: during baseline, during a 20-minute cooling phase, and during 20 minutes of recovery from the cold stress (rewanning). ~
The Journal of Hand Surgery / Vol. 20A No. 5 September 1995 711
Table 1. Individual Patient Data
Patient/ Age/Sex
Location of Ulcers
Location of Probe
1/45/F
Right index
Index
2/33/F
Left index
Middle
3/32/F
Left index
Index
4/44/F
Necrotic left middle
Middle
5/45/F
Left index Left middle
Middle
6/40/F
Left index
Index
Type of Surgery
Level of Occlusion (see Fig. 1) Forearm-A
Right peripheral sympathectomy 12/30/90 ~ - - - ~ Left peripherfl/ "~.//1) sympathe~tomy 3 / 3 / 9 ~ ~ Left periphe?al-------J/j" sympathectomy 2/18/90 Left peripheral sympathectomy; amputation of left middle fingertip through the DIP joint 1/10/91 Left peripheral Ulnar-6 cm sympathectomy from origin 12/13/90 Left peripheral sympathectomy 3/3/92
Hand-C
Digits-DEFG IRE, IUE, MRE, RRE, LUF IRE, IUE, MRE, MUF, LUF TE, IRE, IUE, MRE, RRE, RUF, LUF IRE, MRE, MUE, RRF, RUE, LRE, LUE
No superficial arch (deep palmer arch intact)
I (R&U), D M (R&U), D R (R&U), D L (R&U), D IRF, IUE, MRF, LRF, LUE
IRE, index, radial, zone E; IUE, index, ulnar, zone E; MRE, middle, radial, zone E; RRE, ring, radial, zone E; LUF, little, ulnar, zone F; TE, thumb, zone E; RUF, ring, ulnar, zone F; LRE, little, radial, zone E.
Digital Blood Flow. The microcirculation (total blood flow and thermoregulatory blood flow) was assessed by monitoring digital pulp temperatures and digital pulp cutaneous perfusion, as evaluated by laser Doppler fluxmetry (LDF) before, during, and after ICST9-12; both measurements were recorded simultaneously. The selected digit was symptomatic and had documented distal occlusive disease. Vasomotion was monitored by LDF, using a Periflux Model PF3 perfu-
sion monitor (Perimed, Stockholm, Sweden) and a conventional flow probe attached to one digit of each hand. Laser Doppler fluxmetry evaluates microvascular blood flux, that is, cutaneous perfusion, beneath the skin at a 1.0-mm depth, ~3and also provides a measure of vasomotion in the nutritional capillary loops and most of the superficial arteriovenous anastomoses. Output from the LDF was collected using a custom software package (CIM Solutions, Advance, NC) with
Figure 2. (Top) During modified peripheral sympathectomy technique, the radial and ulnar arteries at the wrist and the superficial palmar arch and three anterior common digital arteries are exposed. The adventitia and direct peripheral nerve connection are then dissected from the vessels under from 6.25x to 30.0x magnification. (Bottom) Close-up view of an artery and associated nerve demonstrating peripheral nerve connections to distal arteries, which are disrupted by the surgical technique.
712
Koman et al. / Microcirculatory Effects of Peripheral Sympathectomy
readings taken at 0.1-second sampling intervals. LDF curves were generated to characterize microvascular flux during the three phases of the test. Digital cooling was accomplished by having the patients insert their hands into a modified refrigeration unit (Sears Roebuck 3.7 cu. ft. refrigerator) equipped with special portholes (Fig. 3). The temperature within the refrigeration unit was maintained at between 5~ and 8~ Digital pulp temperatures of all fingers were recorded simultaneously on-line via small surface temperature probes (#427 thermistors, Yellow Springs Instrument Company, Yellow Springs, OH). Digital temperature readings were monitored and recorded with custom computer software; measurements were taken at 10-second sampling intervals. Temperature curves were generated to characterize digital blood flow during the test. Normal Control Subjects. Digital temperature and LDF measurements during an ICST evaluation were collected from seven female control subjects (seven hands) following the previously described ICST prot o c o 1 . 9-~2 The control subjects were nonsmokers who had no prior history of trauma or vascular disorder in
their hands or arms; their average age was 29 years (range, 18--43). Data Analysis. Data analysis was performed using a univariate repeated-measures analysis of variance for the between-group comparisons over time. Comparisons were made between treatments (ie, before and after peripheral sympathectomy) and between phases of the isolated cold stress test (ie, baseline vs cooling vs rewarming). Temperature and laser Doppler measurements were averaged over five time periods in the following manner. Baseline measurements were avera g e d over the 5-minute baseline period of the test. During the cooling phase of the test, measurements were collected during minutes 6 to 10 (first 5 minutes of cooling) and minutes 20 to 25 (last 5 minutes of cooling). Measurements for each of the two time periods were averaged to give an average for the first and last 5 minutes of cooling. Similarly, during the rewarming phase of the test, measurements from minutes 26 to 31 (first 5 minutes of rewarming) and minutes 40 to 45 (last 5 minutes of rewarming) were collected and averaged. Digital temperatures and LDF measurements were compared between control sub-
Lamer probestors thermi &
Thermistor~ Figure 3. Schematic representation of the equipment necessary to perform the combined isolated cold stress test (ICST) and laser Doppler fluxmetry (LDF) evaluation. Thermistors are attached to all 10 digits and the laser Doppler probe is attached to one digit of each hand. Measurements of digital temperature and LDF are recorded with custom computer software. (From Pollock et al.8 with permission.)
The Journal of Hand Surgery / Vol. 20A No. 5 September 1995
jects and patients before and after peripheral sympathectomy. Microvascular perfusion was correlated with symptoms and ulcer healing.
Results Clinical Results
Following peripheral sympathectomy, all patients reported decreased pain and improved cold tolerance; finger tip ulcerations healed (n = 6) or improved (n = 1). M i c r o v a s c u l a r Results
The normal microvascular response to the cold stress test was characterized by a decrease in digital tempera-
713
ture during cooling, followed by a gradual but incomplete increase in digital temperature during rewarming (Fig. 4A). There was no significant difference in digital temperature between the patients and controls before surgery (p = .0687) or at 6 months after surgery (p = .2133). Laser Doppler fluxmetry responses during stress documented significant differences between controis and patients before surgery (decreased LDF) (Fig. 4B), which no longer had statistical significance (p = .0955) at 6 months (Fig. 4B) These data suggest that peripheral sympathectomy normalizes the digital microcirculatory response to stress in digits compromised by vaso-occlusive involvement, as reflected by LDE Preoperative versus Postoperative. Analysis of digital ICST before and after peripheral sympathec-
30. 9
NOm(~,L CONTIt01~q N - ? DIGITS
25.
9
PATII~: PBB-SUBGI~.'Y
20,
9 P&TIZNTS: 6 MONTHS p O S T - O P N - 6 DIGIT5
N = 7 DIGIT5
15-
10,
5
I
I
' cootma
'
' PA'W~U~MmG
l
A
TIME
350 9 NOla~LI~. CONTI~OI~ N - ? DIGrm 9 PAT/]roTS: Pla-SUltC~3~ N - ? DIG~TS
250.
9 PATI]~TS: 6 MONTHS P 0 6 T - O p N - 6 DIGITS
.~ 200,
150.
100.
50.
0
B
TIME
Figure 4. (A) Graphs of digital temperature response during ICST of normal, control hands (n = 7) and hands of patients with vaso-occlusive disease (n = 7). (B) Graphic representation of the LDF data during ICST of hands in A. Values represent the average value at each point, + SEM.
714
Koman et al. / Microcirculatory Effects of Peripheral Sympathectomy
tomy confirms improvement in microcirculatory response (Fig. 5). There was a significant difference (p = .0188) between the LDF measurements before and after surgery; however, there was no significant difference between preoperative and postoperative digital temperatures. The microcirculatory response to peripheral sympathectomy in one patient is documented in Figure 6, in which the boxed line represents the environmental temperature to which hands were exposed during the three phases of the test. During the preoperative evaluation, baseline (-5 to 0 min) temperatures were below 26~ in all digits on the right hand, with the index finger exhibiting the
lowest temperatures (between 18~ and 20~ (Fig. 6A). During the cooling phase (0 to 20 min), digital temperatures fell rapidly, with the temperature of the index finger again being lowest. During rewarming (20 to 40 min), there was a slow but complete return to baseline temperatures in all but the index digit. LDF was abnormally low throughout the test, reflecting minimal cutaneous perfusion at baseline; LDF measurements of perfusion were not detectable during cooling and were very low during rewarming. Perfusion began to rise above 0 at 35 minutes, Postoperatively, at both 3 and 6 months, all digital temperatures increased throughout all phases of the
?
v
~nna
9
r~rarm I
4, r ~ n
I
2
10
5
PRE-OP N = 7 DIGITS
A
2-8 ~ POST-OP N = 7 DIGITS
3 MONTHS POST-OP N = 5 DIGITS
8 MONTHS POST-OP N6 DIGITS
140 9
bw~hw
120
g
m ~ a
0
rmr~m
2
A
20
B
PRE-OP N ~ 7 DIGITS
2-8 ~ POST-OP N = 7 DIGITS
3 MONTHS POST OP N = 5 DIGITS
8 MONTHS POST-OP N = 5 DIGITS
Figure 5. (A) The average response of digital temperature during the three phases (baseline, cooling, and rewarming) of the ICST. Each of the four graphs represents the average temperature response of the most severely involved digits during ICST. Results from tests performed at four different time intervals are plotted: (1) preoperative, (2) postoperative (2-8 weeks), (3) postoperative (3 months), and (4) postoperative (6 months). (B) The average LDF values of the most severely affected digit during the three phases of ICST are plotted. Preoperative values (baseline) are compared with tests performed postoperatively at (1) 2 to 8 weeks, (2) 3 months, and (3) 6 months. Values represent the average value at each point, _+SEM.
~ ~.g~.~. ~ ~ . ~ ='~
~
"~
el)
~" , ,
9-=
~
=
"
~
~
o ~
E~ ~ ' ~
~- "~
~
9
8- ~ ~:
o
. "~
~E
,,,,~,~-,,,,,,,,,
,,,,
,,,,
,,,,,,,
~
~
~
~
~.~ ~ ~ ~-,~~ ~ ~ ~.~'~
-
~
-~
-
~
,~
~ ~
B"
~ ~ 0
~
~:~
cq
.~ .~
~
~.~
~
9, " ~
~
~
,-
~
~
~
o
~
~
.~
~ .~
~=
~
~
~
9
9
~
Ii
.,i ! ~.~ ~ *'~
0
r
e ~ [ .~~~
.= g ~
(0o) e~nl~uedkuej.
(^UJ) xnl..I jelddO[] .le~e]
~.~~.~
~ '~
716
Koman et al. / Microcirculatory Effects of Peripheral Sympathectomy
test, and the previously observed rapid decline in temperature during cooling was absent (Fig. 6B, C). At 6 months, the index finger temperature remained lower than the other digits but, even during cooling, was still 8~ warmer than it had been preoperatively. Postoperative LDF values at 3 and 6 months were increased over baseline values, indicating improved cutaneous perfusion and vasomotion throughout the entire test (Fig. 6B, C).
Discussion The human digit serves as a thermoregulatory end organ. In this capacity, its blood flow exhibits cyclical increases and decreases, the magnitude and frequency of which' are determined by environmental temperature, t4 Microvascuiar blood flow in the digits is characterized by two components: (1) thermoregulatory flow, which is modulated by arteriovenous anastomoses and contributes to body temperature control; and (2) nutritional blood flow, which is provided by papillary capillaries and maintains tissue viabilityY 6 Chronic arterial insufficiency of the upper extremity often results in digital pain, cold intolerance, and digital tip ulceration, which are refractory to nonoperative management. Peripheral sympathectomy4'6'~2'~7 is performed by isolating the terminal branches of the sympathetic nerves, dividing these terminal branches, and stripping the adventitia from the appropriate arteries. The theoretic basis for employing this surgical procedure is that it produces a more distal interruption of sympathetic innervation to the blood vessels to decrease norepinephrine release at the myoneural junction in the vessel wall, to eliminate vasospasm, and to dilate arterial smooth muscle. 7,9Peripheral sympathectomy has several theoretic advantages over more proximal cervicothoracic sympathectomy: it eliminates sympathetic fibers that bypass the sympathetic trunk~8; its effect lasts longer'9; and it often is effective when central or cervicothoracic sympathectomy has failed. 6,17,18 Peripheral sympathectomy of the digital artery has been advocated when nonoperative treatment has failed and disabling symptoms and/or finger tip ulceration persist? '6 Clinical relief is believed to result from diminished vasospasm, with resultant dilatation of arterial smooth muscle. ~9Although sympathetic nerve interruption decreases the release of norepinephrine at the myoneural junction, 19 that mechanism fails t o explain why intermediate-term palliation is better with distal interruption than with
more proximal sympathectomy, and it does not address the microcirculatory changes that accompany the procedure. Furthermore, published clinical studies have failed to differentiate patients with vasospastic disease alone from those with combined vasospastic and occlusive involvement. The effects of peripheral sympathectomy on microvascular blood flow in vaso-occlusive digits have not, to our knowledge, been described previously. Previous investigators have demonstrated clinical improvement following peripheral sympathectomy, documenting increased digital t e m p e r a t u r e 3,4,6,~7 and improved pulse volume recordings 7 before and after cold stress. Our data suggest that distal peripheral sympathectomy maximizes cutaneous perfusion in the presence of arterial occlusion: temperature, which provides an index of total pulp flow, did not increase significantly over the short term or over a 6month postoperative follow-up period; and LDE which provides an assessment of cutaneous microvascular perfusion and vasomotion, increased significantly. These observations support a possible preferential increase in flow into nutritional rather than thermoregulatory vessels following peripheral sympathectomy. If that is the case, then nutritional flow at sufficient volume to maintain cellular integrity would explain a decrease in symptoms and would allow ulcerations to heal. Conversely, increased total flow with inappropriate arteriovenous shunting proximal to nutritional papillary capillaries would not improve symptoms. However, caution must be used in applying these observations to digits without occlusive disease, since both total flow and nutritional flow may be increased by the partial interruption of sympathetic tone. This finding m a y explain the significant temperature changes reported by others after peripheral sympathectomy. 3,4,14,~7Until some of these points are settled, interpretation of nonstressed temperature data requires caution. 12 Data from this preliminary study illustrate the manner in which microvascular perfusion improves following peripheral sympathectomy. The use of nonimmersion thermal stress is important since nonstressed data are difficult to interpret? 2 ICST, using both temperature and LDF, allows documentation of physiologic effects of peripheral sympathectomy during a prospective follow-up period. This study demonstrates that peripheral sympathectomy has a direct effect on peripheral microcirculation in digits with vaso-occlusive disease. The microcirculatory changes observed following periph-
The Journal of Hand Surgery / Vol. 20A No. 5 September 1995 717 eral s y m p a t h e c t o m y c o i n c i d e d with clinical i m p r o v e m e n t in pain and cold intolerance, and with healing of previously recalcitrant digital ulcers. The data suggest that these clinical findings are secondary to i m p r o v e m e n t in nutritional microcirculatory flow as reflected in increased L D F measurem e n t s . C o m p a r i s o n s between the responses of the vaso-occlusive digits and those of the normal control digits support the observation that the decline o f cutaneous perfusion b e l o w a certain critical level results in ulceration, pain, and cold intolerance. However, if nutritional cutaneous perfusion can be increased a b o v e this critical level, digital circulation is then adequate to heal the ulcers and to reduce pain and cold intolerance. In this study, such i m p r o v e m e n t s were o b s e r v e d even though cutaneous perfusion in these patients was not as high as that in the control subjects. Peripheral sympathectomy performed by adventitial stripping of sympathetic fibers and peripheral neural connections to the radial and ulnar artery at the wrist, the superficial palmar arch, and the three c o m m o n palmar distal arteries results in (1) clinical i m p r o v e m e n t in patients with R a y n a u d ' s phenomenon secondary to systemic disease and acquired and unreconstructible arterial occlusion; (2) i m p r o v e m e n t in cutaneous perfusion, as reflected b y L D F for up to 6 months after surgery; and (3) increases in total flow, as reflected by increased digital temperature during ICST.
References 1. Barcroft H, Walker AJ. Return of tone in blood vessels of the upper limb after sympathectomy. Lancet 1949; 1:1035-9. 2. Koman LA, Urbaniak JR. Ulnar artery insufficiency: a guide to treatment. J Hand Surg 1981;6:16--24. 3. E1-Gammal TA, Blair WF. Digital periarterial sympathectomy for ischemic digital pain and ulcers. J Hand Surg 1991;16B:382-5.
4. Hatt AE. Digital artery sympathectomy. J Hand Surg 1980; 5:550-6. 5. Jones NF. Acute and chronic ischemia of the hand: pathophysiology, treatment and diagnosis. J Hand Surg 1991; 16: 1074-83. 6. Wilgis EFS. The evaluation and treatment of chronic digital ischemia. Ann Surg 1981;193:693-8. 7. Miller LM, Morgan RF. Vasospastic disorders: etiology, reconstruction, and treatment. Hand Clin 1993;9:171-87. 8. Pollock FE Jr, Koman LA, Smith BP, Holden M, Russell GB, Poehling GG. Measurement of hand microvascular blood flow with isolated cold stress testing and laser Doppler fluxmetry. J Hand Surg 1993;18A: 143-50. 9. Koman LA, Nunley JA, Goldman JL, Seaber AV, Urbaniak JR. Isolated cold stress testing in the assessment of symptoms in the upper extremity: preliminary communications. J Hand Surg 1984;9A:305-13. 10. Koman LA, Nunley JA. Thermoregulatory control after upper extremity replantation. J Hand Surg 1986;11A: 548-52. 11. Nunley JA, Penny WH III, Woodbury MA, Koman LA. Quantitative analysis of cold stress performance after digital replantation. J Orthop Res 1990;8:94-100. 12. Koman LA, Smith BP, Smith TL. Stress testing in the evaluation of upper extremity perfusion. Hand Clin 1993;9: 59-83. 13. Tenland T. On laser Doppler flowmetry. Methods and microvascular applications. Link6ping Studies in Science and Technology Dissertations, No. 83. Link6ping, Sweden: Link6ping University, 1982:7-51. 14. BurtOn AC. The range and variability of the blood flow in human fingers and the vasomotor regulation of body temperature. Am J Physiol 1939;127:437-53. 15. Conrad MC, Functional anatomy of the circulation to the lower extremities. Chicago: Year Book Medical Publishers, 1971:190. 16. Fagrell B. Dynamics of skin microcirculation in humans. J Cardiovasc Pharmacol 1985;7(suppl. 3):$53-8. 17. Egloff DV, Mifsud RP, Verdan C. Superselective digital sympathectomy in Raynaud's phenomenon. Hand 1982; 15:110--4. 18. Pick J. The autonomic nervous system: morphological, comparative, clinical and surgical aspects. Philadelphia: JB Lippincott, 1970. 19. Wilgis EFS. Digital sympathectomy for vascular insufficiency. Hand Clin 1985;1:361-7.