Hallene Maragh, Barbara S. Meyer, David Davenport, Jeffrey D. Gould, and Julia K. Terzis
MORPHOFUNCTIONAL EVALUATION OF FIBRIN GLUE VERSUS MICROSUTURE Downloaded by: National University of Singapore. Copyrighted material.
NERVE REPAIRS ABSTRACT Complications of microsuture repair of peripheral nerves include mechanical trauma, foreign body reaction, impairment of vascularity, and possible obstruction to successful sprouting. In addition, there are occasions when it is virtually impossible to perform a suture repair because of limited exposure and/or very fine structures, such as are encountered in pediatric cases. These situations have continued to provide the impetus for evaluating alternative methods of nerve coaptation. Recently, the use of tissue glue has gained in popularity as a technique for sutureless nerve repairs. We decided to test the efficacy of fibrin glue repair versus microsuture coaptation in the rat sciatic model. The repair sites were assessed for tensile strength, by quantitative morphometry, and by electrophysiologic studies. Tensile strength findings revealed that at two, four, and eight weeks after surgery, there was no significant difference between the two repair techniques, although there was a trend toward a stronger hold in the microsuture repairs. Electrophysiologic recordings revealed that conventional microsuture repairs had significantly faster conduction velocities, larger area under the curve, and higher peak amplitudes. The onset and peak latencies were comparable, revealing that the axonal quality of at least a certain number of axons was similar electrically. Axonal counts both proximal and distal to the repair showed no significant difference, although there was an overall suggestion of superiority in the number of myelinated axons in the suture repair. In the continued effort to evaluate alternative methods to microsuture coaptation, this study was undertaken to test the efficacy of fibrin glue nerve repair of peripheral nerves, without any suture-assisted technique. In turn, these sutureless repairs were to be compared to conventional microsuture coaptation. A previous study carried out in our laboratory, comparing laser repair of the rat sciatic nerve with conventional microsuture, yielded comparable results.1 There is little doubt that physical and biologic factors influence functional recovery of nerve repairs. Fibrin glue might eliminate the trauma of sutures, the long-term foreign body reaction to the material and, if applied evenly at the site of coaptation, might reduce aberrant sprouting from the site of repair. In addition, it might make certain microneural coaptations faster and easier to perform. In Europe, fibrin glue has gained wide acceptance.
Surgeons are using this material clinically, although there is no definite basic science research to substantiate the claimed advantages of this technique.2-8 Few surgeons have been willing to use tissue glue without using an ancillary relief suture, because of fear of dehiscence 9 -' 5 In the rat sciatic model, transection of the nerve results in a gap, and the ends can be easily approximated without tension after some mobilization. Fascicular orientation in the rat model can be easily achieved by an epineurial repair. Placing two or more microsutures may create a good repair, making the addition of tissue glue redundant. Fibrin adhesives have to be compared independently as a nerve-repair technique in a model where minimal tension exists. To decide the efficacy of fibrin glue as an alternative to conventional microsuture repair, and to determine any detrimental effects, were the purposes of our direct, double-blind comparison of the two techniques.
Microsurgical Research Center, Eastern Virginia Medical School, Medical College of Hampton Roads, Norfolk, VA Reprint requests- Dr. Terzis, Microsurgical Research Center, 700 Olney Rd., P.O. Box 1980, Norfolk, VA 23501 Accepted for publication April 17,1990 Copyright © 1990 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
331
MATERIALS AND METHODS The right and left sciatic nerves of 25 male SpragueDawley rats (402 to 543 g) were exposed, sectioned, and then repaired, using either tissue glue coaptation or standard microsuture nerve repair techniques. Each animal served as its own control, with sides for each type of repair randomly assigned. All repairs were carried out by the same surgeon (H.M.). Table 1 presents a summary of the surgical groups and the assessment parameters used to measure regeneration and tensile strength. Preoperatively, each rat was anesthetized intraperitoneally with sodium pentobarbitol. Procedures began with nerve exposure for a length of 1.5 cm distal to the sciatic notch. The inferior gluteal motor nerve was identified by delivering .5 ma to that division, using a direct current stimulator. At about 5 mm distal to the inferior gluteal, the sciatic was transected with serrated scissors. Spontaneous mushrooming of intrafascicular neural contents, which often followed nerve sectioning and complicated the coaptation procedure, was trimmed as needed. The distal nerve stump was mobilized for 1 to 1.5 cm and thus it was possible to manipulate it with ease and coapt it with the proximal stump. The stumps were oriented and held as closely as possible until a natural fibrin clot was formed. Subsequently, repairofthe nerve was completed, using standard microsuture or fibrin glue techniques. The tissue glue used for our study can presently be found in the form of Tisseel®, a fibrinogen-based mixture with a two-component sealant (a product of IMMUNO, Postfach 30, Industriestrasse 72, A-1221 Wien, Europe). The first component consists of highly concentrated fibrinogen, factor XIII, plasma fibronectin, and aprotinin. The second portion contains a mixture of bovine thrombin ("fast" thrombin) and calcium chloride. Before application to the outer surface of adjoined nerves, the two sealant elements were prepared separately.
Table 1.
OCTOBER 1 9 9 0
Use of the adhesive involved placing one drop of both fibrinogen glue segments onto a small spatula and mixing them thoroughly. The resulting compound was immediately transferred, using a microdissector, to the anterior epineurial surface of the coapted ends. One minute after application to the anterior surface, the nerve was gently turned to expose its posterior surface. A similar mixture of the adhesive was used, insuring that no glue entered the appositional surface of the nerve ends. The compound was recreated for each glue repair. The microsuture technique applied in 18 animals required equidistant placement of three 10-Omonofilament nylon sutures (120° apart) to complete the epineurial repairs. Following the completion of one repair, the contralateral nerve was then repaired with the alternative technique. On completion of nerve repairs, wound closure was achieved by approximating the muscles of the region and skin closure in layers. The animals were returned to their cages after recovery from the anesthesia. There was no immobilization of the legs at any time postoperatively. The two types of coaptation were studied by measuring tensile strength of the repair or by assessing nerve regeneration, using electrophysiologic analysis of the conductive properties of the repaired nerve, followed subsequently by quantitative morphometry of the repaired axons. Tensile strengths were assessed at two, four, and eight weeks following surgery. Functional assessment was determined on the second set of animals by means of compound action potential (CAP) recordings, according to the electrophysiologic arrangement depicted in Figure 1, and by quantitative morphometry which yielded the number of axons that successfully crossed the repair site. The high level of the repair unfortunately resulted in edema of the foot and pressure ulcerations on both heels, an occurrence also reported by the Smahel,14 Moy,15 Matras,16 and Ventura17 groups. This excluded the possibility of toe spread measurements. These
Population Breakdown and Distribution
Population: 25 animals (n = 25) Suture: 24 nerves evaluated Tissue Glue: 23 nerves evaluated* Tensile Strength: 30 nerves Suture: 15 Tissue Glue: 15* Compound Action Potentials: 17 nerves Suture: 9 Tissue Glue: 8* Histology: Proximal: 26 Distal: 25 Common Peroneal: 24 332
*The tissue-glue repaired nerves in three animals dehisced and were eliminated.
SCIATIC NOTCH PERONEAL BRANCH
Figure 1. Electrophysiologic arrangement used for CAP recordings (SE: Stimulating Electrode; RE: Recording Electrode).
Downloaded by: National University of Singapore. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 4
FIBRIN GLUE VERSUS MICROSUTURE/MARAGH, MEYER, DAVENPORT, ET A L
QUANTITATIVE MORPHOLOGY.
Following electro-
physiologic evaluations, the nerves of each animal were harvested for quantitative and qualitative assessment of nerve regeneration across the repair. The rats, still under heavy anesthesia, were perfused using Kar-
novsky's Solution (pH 7.4). Subsequently, both sciatic nerves were harvested and processed for morphometry. Figure 2 depicts the biopsy sites diagrammatically. Seven millimeters rostral and caudal to the repair site, 1-mm sections were collected. These segments were placed in a small vial containing the perfusion solution. An additional 1-mm section was taken from the common peroneal branch at the point of trifurcation of the sciatic nerve. The blocks were then postfixed in osmium tetroxide (1.0 percent in 0.1 percent cacodylate buffer, pH 7.4), and embedded in plastic resin, cut at l|x, and stained with methylene blue. Assessment of the number and area of myelinated axons involved three locations: proximal and distal to the repair and from the peroneal branch (see Fig. 2). Every fourth field of each arbitrarily chosen section was studied, as the investigator reviewed the section from left to right and from top to bottom. Using a drawing tube allowed traced myelinated fibers to be counted through a Zeiss microscope at 1000X. Thus, the total cross-sectional area of each sciatic nerve was studied by computer analysis of tracings made at 100x magnification that were then entered into the computer through a digitizing tablet. A ratio comparing fibers counted and the cumulative area from the fields where counts were derived (as measured by a micrometer) with the total area of the section, allowed realization of the total number of nerve fibers for the section. Average fiber areas for each coaptation method were acquired through analysis of the initial tracings.
Downloaded by: National University of Singapore. Copyrighted material.
behavioral observations would have provided a noninvasive means of assessing functional recovery.118 TENSILE STRENGTH RECORDINGS. TO measure the tensile strength of the repaired sciatic nerve in 28 nerves, the Model 7D Polygraph and DC. driver amplitude was used in conjunction with an FT03C model Force-Displacement transducer. The machine was calibrated before the testing of each nerve. The sciatic was removed from above the sciatic notch to an equal point above the trifurcation. A small hemostat was attached to the force transducer and another to a revolving motor rod. The nerve, approximately 2.5 cm in length, was held between each hemostat with no tension applied. After a base-line reading was established on the polygraph, the motor rod was then allowed to revolve, placing tension on the nerve. The polygraph measured the gradual addition of stress until separation of the nerve occurred. The force necessary to cause rupture of the repair site was referred to as the tensile strength. After testing both sciatics, the animal was sacrificed. CAP EVALUATIONS. Seventeen nerve CAP recordings were taken after a two-month interval. Each animal was anesthetized, as described earlier, and the sciatic nerve was exposed from the sciatic notch to beyond the point of trifurcation distally. The nerve was severed proximal to the sciatic notch to eliminate reflex activation via the spinal cord. Proximal branches of the sciatic nerve were transected to prevent muscular contraction. This removed the possibility of interference from this source and allowed for a monophasic CAP waveform. A bipolar stimulating electrode was then positioned under the sciatic nerve proximal to the repair site. The recording bipolar electrode was placed under the common peroneal branch in the popliteal fossa, (see Fig. 1). Electrical stimuli (from a Grass S48 Stimulator) were applied to the nerve in single square pulses of .05 msec duration. The strength of the excitation was gradually increased from threshold voltages until a "maximum" CAP waveform amplitude was achieved. Ten of these saturated CAP responses for each nerve were collected and averaged by a Cromemco System III computer equipped with an analog-todigital converter. The data were stored for subsequent analyses using a waveform editing analysis program. Using this program, onset and peak latencies, waveform areas, and peak amplitudes were determined. The distance between the stimulating and recording electrodes was measured in each case and this information, along with the latency data, were converted to conduction velocities to be used later in statistical evaluation.
RESULTS TENSILE STRENGTH. Data collected from both assessment groups (tensile strength and CAP/Histology evaluation) for all repairs were analyzed through an analysis of variance (ANOVA). The evaluation of tensile strength results concluded that, at two and four weeks, there was a trend toward stronger coaptations achieved by the suture groups, compared to the nonsuture repairs, although these data were not statistically significant. However, at eight weeks, the tensile strength was comparable in both repair groups (Table 2).
REPAIR SITE
t
(CAP.HISTO)
t
(HISTO)
t
(CAP.HISTO)
Figure 2. Biopsy sites (CAP: Compound Action Potential recordings; HISTO: Histologic evaluation).
333
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 4
Table 2. Tensile Strength Averages (± S.D.) for Suture and Tissue-Glue Repairs at Two, Four, and Eight Weeks Two weeks Suture (5 nerves) Glue (4 nerves) Four weeks Suture (5 nerves) Glue (5 nerves) Eight weeks Suture (5 nerves) Glue (4 nerves)
203.40 g ±61.64) 179.57 g ±65.35) 204.52 g ±96.12) 171.88 g ±96.19)
Onset latency (msec) Peak latency (msec) Peak amplitude (mV) Area (mV)(msec) Conduction velocity (Ms) 334
Quantitative Morphometric Means Suture
Proximal Thickly myelinated Thinly myelinated Total Distal Peroneal branch (distal)
5864 2972 8836 6177 2386
(±330) (±184) (±338) (±704) (±367)
Glue 5481 3341 8822 5357 1690
(±275) (±537) (±624) (±642) (±386)
Summary of CAP Waveform Data Suture
Glue
0.8672 (±0.4723)
0.9433 (±0.8743)
1.4027 (±0.7467)
1.4009 (± 1.2369)
0.4543 (±0.3231)*
0.1827 (±0.2939)*
0.0005 (±0.0004)* 24.167 (±13.374)*
0.0001 (±0.0002)* 13.090 (± 11.578)*
'Indicates differences were statistically significant.
neal branch also provided findings (distal to the repair) which were not significant in a direct comparison of the two types of nerve repair. For that branch, the average diameters measured were 0.20 \xm ± 0.02 for microsutured nerves and 0.29 |xm ± 0.06 for the fibrin glued sciatics. DEHISCENCE. We also discovered the failure of the fibrin glue to sustain neural coaptation in three of 23 tissue glue repairs or 13 percent. There was no dehiscence in any of the nerves coapted with microsutures. AUTOTOMY. For nearly all cases of heel edema, there was autotomy directed at the toes of one or both feet. This finding did not support behavioral testing in these animals.
DISCUSSION This study tried to address the following questions: Does the tissue glue adhesive, which may accidentally fall between the nerve stumps, impede axon growth and successful regeneration? Does the adhesive provide comparable or even greater tensile strength than that afforded by standard microsuture? Must the fibrin clot junction be supplemented by additional means? Analysis of the tensile strength data did not show any statistical difference between the two methods. However, at two and four weeks, there was a tendency toward stronger bonding accomplished by the conventional suture. This trend disappeared by eight weeks, when the tensile strengths of both coaptation types were comparable. Fibrin clots apparently can provide adequate strength when allowed to set completely and when tension is maintained at a minimum shortly after repair. Our three cases of dehiscence show that tension immediately after repair must be avoided, if a long-lasting reattachment of nerve ends is desired. Tension reduction seems to be a concern only immediately after repair. Following that period of solidification, the tissue adhesive demonstrated adequate bonding.
Downloaded by: National University of Singapore. Copyrighted material.
258.98 g (±90.05) 251.15 g (±123.87)
COMPOUND ACTION POTENTIALS. For CAP recordings analysis, the use of ANOVA was supplemented by Duncan's statistical test. Our findings can be reviewed in Table 3. They reveal that the onset and peak latencies were not statistically significant. However, the areas under the curve (p < .01), peak amplitudes {p = .05), and conduction velocities at onset {p = .05) were found to favor the suture group significantly (Table 3). HISTOLOGY OF THE REPAIR SITE. Quantitative morphometric means for both the microsuture and tissue glue groups indicated no statistically significant difference either proximal or distal to the repair site (Table 4). In studying the nerve cross sections with the method used previously, two predominant populations of fibers were present in the proximal sections only: normally myelinated fiber groups, and those which were finely myelinated.' Axon fibers found at the distal position consisted of one finely myelinated group (Fig. 4). Tallies for the peroneal branch (distal to the coaptation of the nerve) also showed no statistical difference between the two techniques (see Table 4). The axonal areas proximal to the site of the repair proved not to hold statistical significance (0.74 |xm2 ± 0.04 for the microsutured nerves versus 0.75 |xm2 ± 0.05 for the fibrin coaptations). At 7 mm distal to the repair, areas proved again to be statistically equivalent for the microsuture (0.61 |xm2 ± 0.08) and tissue glue (0.47 (xm2 ± 0.06) procedures. When the findings from the two nerves were compared, again no statistical difference was present. Further, measurement of axonal areas of the pero-
Table 3.
Table 4.
OCTOBER 1990
FIBRIN GLUE VERSUS MICROSUTURE/MARAGH, MEYER, DAVENPORT, ET AL.
cases of tension-free repairs would be redundant. Often, other authors added a relief suture, either temporarily or as a permanent feature. 71415 The research goal of fibrin adhesive experiments is to discover if replacement of suture with a better technique is possible. The electrophysiologic data in this study showed that nerve conductivity was superior for the microsuture repairs (see Fig. 3). When reviewing peak amplitudes, areas, and conduction velocities, these were significantly different in favor of the conventional suture technique. Moy et al.'5 noted the same disparity in CAP amplitudes at 18 weeks. Thus, the CAP data suggest a lower probability of neural regeneration in the tissue glue repairs. Whether this finding of diminished conductivity is due to the effects of the interposition of a foreign protein between the nerve ends remains unknown. Another possibility is that the axonal penetration through tissue glue repairs may be subjected to early tension forces that are not present at the suture repairs. These forces, due to early mobilization of the extremity following repair, may invite scar formation that hinders nerve regeneration. Microsuture repair has the advantage of precisely positioning and holding the two nerve ends. Thus,
4.88 C400) mSEC
C4003
Downloaded by: National University of Singapore. Copyrighted material.
The dehiscence findings of this study correspond in part with previous reports. Cruz et al.,]3 in a similar comparison of tissue glue and microsuture material, reported an 80 percent dehiscence rate for nerves adhered by means of a fibrin clot. The study conducted by Haase et al22 found total adhesive coaptation dehiscence, although no other study reported such an extreme failure rate. 5610 Perhaps immobilization immediately after repair should be recommended as it was by Matras. 111619 Several other reports convey the same warning of unchecked tension leading to dehiscence shortly after glue repair. 78202123 Specifically, the surgical area must remain free of stress forces for at least a two-week period, and even longer when comparable tensile strengths are provided by the two groups. It should be emphasized that, in the present study, the animals were not immobilized after the nerve repair procedures, but were allowed to move freely in their cages. Despite the lack of immobilization, only three nerves dehisced in the present study. Apparently, a generous degree of mobilization of the nerve ends can generate a tension-free environment at the coaptation site. Although an adjuvant is required, use of suture in these
6.10 mSEC
B
4.88
e
C4003
-.400
FILE LI 5
RUN
5
6.10 mSEC
4.88
FILE R15
RUN
6.10 C400) mSEC
5
Figure 3. Examples of actual computer tracings from recordings obtained from two animals. CAP recordings across suture repairs are depicted in A and C while CAPs obtained across the tissue glue repairs are shown in B and D. Note superiority in peak amplitude, area, and conduction velocity of the suture-repaired nerves. (All recordings obtained at same gain (5000).)
335
OCTOBER 1990
r B
Figure 4. Photomicrographs depicting the axonal regeneration in the two sciatic nerves of the same animal using the two repair techniques. A, Cross section of fibrin adhesive repaired nerve 7 mm distal to the repair site. B, Microsuture repair at same site (methylene blue, X800).
336
there is some certainty that the neural coaptation will be of a permanent nature. In contrast, the gluing technique attempts to preserve only approximation of the nerve ends. In either case, data showing conduction velocity differences imply both that apposition by fibrin glue resulted in poorer myelination, and that regrowth across the lesion by proximal axons occurred through a more inadequate coaptation. These findings differ from those reported by Smahel et al.]4 who showed statistically comparable CAP results in the same experimental model. However, in Smahel's study, a relief suture and silicone tubing was placed across the junction site in the glued repairs, while no such adjunctive measures were used in our study. Our collected histologic data seem to encourage the conclusion of CAP findings. Overall, the quantitative morphometry of the two repairs was comparable, although there were discrepancies in the axon counts that favored the suture repair. There were, however, no statistical differences in the morphologic data. These morphologic findings agree with the reports of Smahel' 4 and Matras16 but contradict the early study of Singer21 and the clinical findings of Egloff6 which postulated that axonal growth was stunted with fibrin adhesive use.
CONCLUSIONS A double-blind study comparing sutureless versus microsuture nerve repairs in the rat sciatic nerve yielded the following results: Adequate tensile strength can eventually be accomplished by both types of repair, if a tension-free environment is provided in the early postoperative period. Once the technique of using tissue glue is mastered, it can offer certain advantages in the clinical arena, including the possibility of executing nerve repairs more rapidly than with the microsuture technique. It can facilitate inaccessible nerve repairs, such as within the bony confines of the skull or within difficult to reach bony canals. It can also make nerve approximation possible between extremely diminutive nerves. However, this study has shown that the interposition of tissue glue between the nerve ends results in a somewhat smaller population of comparable diameter regenerating axons reaching the distal nerve branches; more important, the conductive properties of these axons seemed to be compromised in the tissue glue repairs. Sutureless repairs hold great promise in the fu-
Downloaded by: National University of Singapore. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 4
FIBRIN GLUE VERSUS MICROSUTURE/MARAGH, MEYER, DAVENPORT, ET AL. 10. Tarlov IM, Benjamin B: Plasma clot and silk suture of nerves. Surg Gynecol Obstet 76:366, 1943 11. Matras H, Dinges HP, Mamoli B, Lassmann H: Non-sutured nerve transplantation: A report on animal experiments. ) Maxillofac Surg 1:37, 1973 12. Becker CM, Gueuning CO, Graff GL: Sutures or fibrin glue for divided rat nerves: Schwann cell and muscle metabolism. Microsurg 6:1, 1985 13 Cruz NI, Debs N, Fiol RE: Evaluation of fibrin glue in rat sciatic nerve repairs. Plast Reconstr Surg 78:369, 1986 14. Smahel H, Meyer VE, Bachem U: Gluing of peripheral nerves with fibrin: Experimental studies. J Reconstr Microsurg 3: 211, 1987 REFERENCES 15. Moy 01, Peimer CA, Koniuch MP, et al: Fibrin seal adhesive versus nonabsorbable microsuture in peripheral nerve repair. I Hand Surg 13A:273, 1988 Maragh H. Hawn RS, Gould ID, Terzis IK: Is laser nerve repair 16. Matras H, Braun F, Lassmann H, et al.: Plasma clot welding of comparable to microsuture coaptation? I Reconstr Microsurg 4:189, 1988 nerves: Experimental report. I Maxillofac Surg 1:236, 1973 Miehlke A: Surgery of the Facial Nerve, 2nd ed., Philadelphia: W. B.17. Ventura R, Torri G, Campari A, etal: Experimental suture of the Saunders Co., 1973 peripheral nerves with "fibrin glue". Ital I Orthop Traumatol 6:407, 1980 Wigand ME, Thumfart W: Neurosynthesis of the facial nerve: 18. Hasewaga K: A new method of measuring functional recovery Electrical versus clinical results. In: Samii M, lannetta PJ (eds): The Cranial Nerves. Berlin: Springer-Verlag, 1981 in unanesthetized and unrestrained rats. Experientia 34: Kuderna H: The use of fibrin sealing in peripheral nerve repair 272, 1978 In: Skjoldborg H (ed): 1MMUN0 Scientific Workshop 82. 19. Matras H: Fibrin seal: State of the art. I Oral Maxillofac Surg 43: Tisseel®/Tissucol®-Symposiutn: Areas of Application, Problems and 605, 1985 Perspectives in Current Surgery, Scanticon-Aarhus, Denmark,20. Tarlov IM, Denslow C, Swarz S, Pineles D: Plasma clot suture of 1982 nerves Arch Surg 47:44, 1943 Osgaard O: Tisseel® in peripheral nerve and cranial nerve 21. Singer M: The combined use of fibrin film and clot in end-tosurgery In: Skjoldborg H (ed): IMMUNO Scientific Workshop end union of nerves. I Neurosurg 2:102, 1945 '82: Tisseel®/Tissucol®-Symposium: Areas of Application, Problems 22. Haase I, Andreasen H, Falk E, et al.: Use of fibrin seal (tisseel/ and Perspectives in Current Surgery, Scanticon-Aarhus, Dentissucol) in peripheral nerve surgery: An experimental rat mark, 1982 model In: Schlag G, Redl H (eds): Fibrin Sealant in Operative Egloff DV, Narakas A: Nerve anastomoses with human fibrin: Medicine: Ophthalmology-Neurosurgery, Vol. 2, Berlin: SpringerPreliminary clinical report (56 cases) Ann ChirMain 2:101, Verlag, 1986 1983 23 Matras H, Kuderna H: Glueing nerve anastomoses with clotting Egloff, DV, Narakas A, Bonnard C: Results of nerve grafts with substances. In: Marchac D, Hueston |T (eds): Transactions of tissucol (tisseel) anastomosis. In: Schlag G, Red! H (eds): the Sixth International Congress of Plastic and Reconstructive SurFibrin Sealant in Operative Medicine-. Ophthalmology-Neurogery. Paris, August 24-29, 1975. Paris: Masson, 1976 surgery, Vol 2, Berlin: Springer-Verlag, 1986 24. Vinazzer H: Fibrin sealing: Physiologic and biochemical background Fac Plast Surg 2:291, 1985 Palazzi S: The use of fibrin sealant in nerve adhesions (peripheral nerves and plexus brachialis). In: Schlag G, Redl H (eds): Fibrin Sealant in Operative Medicine. OphthalmologyWe extend our gratitude to Mary Jacobs for electroNeurosurgery, Vol 2, Berlin: Springer-Verlag, 1986 physiologic recordings, to Susan Downing and Jeff Dupree for Young ]Z, Medawar PB: Fibrin suture of peripheral nerves: histologic assistance, and to IMMUNO for use of their prodMeasurement of the rate of regeneration. Lancet 239:126, uct and literature assistance. 1940
337
Downloaded by: National University of Singapore. Copyrighted material.
ture of clinical microneurovascular surgery. At present, use of tissue glue seems to be associated with some adverse effects. An intense effort in the future would be welcomed, to try to substitute glues prepared from the patient's own serum for existing fibrin adhesives.
MORPHOFUNCTIONAL EVALUATION OF FIBRIN GLUE VERSUS MICROSUTURE Downloaded by: National University of Singapore. Copyrighted material.
NERVE REPAIRS ABSTRACT Complications of microsuture repair of peripheral nerves include mechanical trauma, foreign body reaction, impairment of vascularity, and possible obstruction to successful sprouting. In addition, there are occasions when it is virtually impossible to perform a suture repair because of limited exposure and/or very fine structures, such as are encountered in pediatric cases. These situations have continued to provide the impetus for evaluating alternative methods of nerve coaptation. Recently, the use of tissue glue has gained in popularity as a technique for sutureless nerve repairs. We decided to test the efficacy of fibrin glue repair versus microsuture coaptation in the rat sciatic model. The repair sites were assessed for tensile strength, by quantitative morphometry, and by electrophysiologic studies. Tensile strength findings revealed that at two, four, and eight weeks after surgery, there was no significant difference between the two repair techniques, although there was a trend toward a stronger hold in the microsuture repairs. Electrophysiologic recordings revealed that conventional microsuture repairs had significantly faster conduction velocities, larger area under the curve, and higher peak amplitudes. The onset and peak latencies were comparable, revealing that the axonal quality of at least a certain number of axons was similar electrically. Axonal counts both proximal and distal to the repair showed no significant difference, although there was an overall suggestion of superiority in the number of myelinated axons in the suture repair. In the continued effort to evaluate alternative methods to microsuture coaptation, this study was undertaken to test the efficacy of fibrin glue nerve repair of peripheral nerves, without any suture-assisted technique. In turn, these sutureless repairs were to be compared to conventional microsuture coaptation. A previous study carried out in our laboratory, comparing laser repair of the rat sciatic nerve with conventional microsuture, yielded comparable results.1 There is little doubt that physical and biologic factors influence functional recovery of nerve repairs. Fibrin glue might eliminate the trauma of sutures, the long-term foreign body reaction to the material and, if applied evenly at the site of coaptation, might reduce aberrant sprouting from the site of repair. In addition, it might make certain microneural coaptations faster and easier to perform. In Europe, fibrin glue has gained wide acceptance.
Surgeons are using this material clinically, although there is no definite basic science research to substantiate the claimed advantages of this technique.2-8 Few surgeons have been willing to use tissue glue without using an ancillary relief suture, because of fear of dehiscence 9 -' 5 In the rat sciatic model, transection of the nerve results in a gap, and the ends can be easily approximated without tension after some mobilization. Fascicular orientation in the rat model can be easily achieved by an epineurial repair. Placing two or more microsutures may create a good repair, making the addition of tissue glue redundant. Fibrin adhesives have to be compared independently as a nerve-repair technique in a model where minimal tension exists. To decide the efficacy of fibrin glue as an alternative to conventional microsuture repair, and to determine any detrimental effects, were the purposes of our direct, double-blind comparison of the two techniques.
Microsurgical Research Center, Eastern Virginia Medical School, Medical College of Hampton Roads, Norfolk, VA Reprint requests- Dr. Terzis, Microsurgical Research Center, 700 Olney Rd., P.O. Box 1980, Norfolk, VA 23501 Accepted for publication April 17,1990 Copyright © 1990 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
331
MATERIALS AND METHODS The right and left sciatic nerves of 25 male SpragueDawley rats (402 to 543 g) were exposed, sectioned, and then repaired, using either tissue glue coaptation or standard microsuture nerve repair techniques. Each animal served as its own control, with sides for each type of repair randomly assigned. All repairs were carried out by the same surgeon (H.M.). Table 1 presents a summary of the surgical groups and the assessment parameters used to measure regeneration and tensile strength. Preoperatively, each rat was anesthetized intraperitoneally with sodium pentobarbitol. Procedures began with nerve exposure for a length of 1.5 cm distal to the sciatic notch. The inferior gluteal motor nerve was identified by delivering .5 ma to that division, using a direct current stimulator. At about 5 mm distal to the inferior gluteal, the sciatic was transected with serrated scissors. Spontaneous mushrooming of intrafascicular neural contents, which often followed nerve sectioning and complicated the coaptation procedure, was trimmed as needed. The distal nerve stump was mobilized for 1 to 1.5 cm and thus it was possible to manipulate it with ease and coapt it with the proximal stump. The stumps were oriented and held as closely as possible until a natural fibrin clot was formed. Subsequently, repairofthe nerve was completed, using standard microsuture or fibrin glue techniques. The tissue glue used for our study can presently be found in the form of Tisseel®, a fibrinogen-based mixture with a two-component sealant (a product of IMMUNO, Postfach 30, Industriestrasse 72, A-1221 Wien, Europe). The first component consists of highly concentrated fibrinogen, factor XIII, plasma fibronectin, and aprotinin. The second portion contains a mixture of bovine thrombin ("fast" thrombin) and calcium chloride. Before application to the outer surface of adjoined nerves, the two sealant elements were prepared separately.
Table 1.
OCTOBER 1 9 9 0
Use of the adhesive involved placing one drop of both fibrinogen glue segments onto a small spatula and mixing them thoroughly. The resulting compound was immediately transferred, using a microdissector, to the anterior epineurial surface of the coapted ends. One minute after application to the anterior surface, the nerve was gently turned to expose its posterior surface. A similar mixture of the adhesive was used, insuring that no glue entered the appositional surface of the nerve ends. The compound was recreated for each glue repair. The microsuture technique applied in 18 animals required equidistant placement of three 10-Omonofilament nylon sutures (120° apart) to complete the epineurial repairs. Following the completion of one repair, the contralateral nerve was then repaired with the alternative technique. On completion of nerve repairs, wound closure was achieved by approximating the muscles of the region and skin closure in layers. The animals were returned to their cages after recovery from the anesthesia. There was no immobilization of the legs at any time postoperatively. The two types of coaptation were studied by measuring tensile strength of the repair or by assessing nerve regeneration, using electrophysiologic analysis of the conductive properties of the repaired nerve, followed subsequently by quantitative morphometry of the repaired axons. Tensile strengths were assessed at two, four, and eight weeks following surgery. Functional assessment was determined on the second set of animals by means of compound action potential (CAP) recordings, according to the electrophysiologic arrangement depicted in Figure 1, and by quantitative morphometry which yielded the number of axons that successfully crossed the repair site. The high level of the repair unfortunately resulted in edema of the foot and pressure ulcerations on both heels, an occurrence also reported by the Smahel,14 Moy,15 Matras,16 and Ventura17 groups. This excluded the possibility of toe spread measurements. These
Population Breakdown and Distribution
Population: 25 animals (n = 25) Suture: 24 nerves evaluated Tissue Glue: 23 nerves evaluated* Tensile Strength: 30 nerves Suture: 15 Tissue Glue: 15* Compound Action Potentials: 17 nerves Suture: 9 Tissue Glue: 8* Histology: Proximal: 26 Distal: 25 Common Peroneal: 24 332
*The tissue-glue repaired nerves in three animals dehisced and were eliminated.
SCIATIC NOTCH PERONEAL BRANCH
Figure 1. Electrophysiologic arrangement used for CAP recordings (SE: Stimulating Electrode; RE: Recording Electrode).
Downloaded by: National University of Singapore. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 4
FIBRIN GLUE VERSUS MICROSUTURE/MARAGH, MEYER, DAVENPORT, ET A L
QUANTITATIVE MORPHOLOGY.
Following electro-
physiologic evaluations, the nerves of each animal were harvested for quantitative and qualitative assessment of nerve regeneration across the repair. The rats, still under heavy anesthesia, were perfused using Kar-
novsky's Solution (pH 7.4). Subsequently, both sciatic nerves were harvested and processed for morphometry. Figure 2 depicts the biopsy sites diagrammatically. Seven millimeters rostral and caudal to the repair site, 1-mm sections were collected. These segments were placed in a small vial containing the perfusion solution. An additional 1-mm section was taken from the common peroneal branch at the point of trifurcation of the sciatic nerve. The blocks were then postfixed in osmium tetroxide (1.0 percent in 0.1 percent cacodylate buffer, pH 7.4), and embedded in plastic resin, cut at l|x, and stained with methylene blue. Assessment of the number and area of myelinated axons involved three locations: proximal and distal to the repair and from the peroneal branch (see Fig. 2). Every fourth field of each arbitrarily chosen section was studied, as the investigator reviewed the section from left to right and from top to bottom. Using a drawing tube allowed traced myelinated fibers to be counted through a Zeiss microscope at 1000X. Thus, the total cross-sectional area of each sciatic nerve was studied by computer analysis of tracings made at 100x magnification that were then entered into the computer through a digitizing tablet. A ratio comparing fibers counted and the cumulative area from the fields where counts were derived (as measured by a micrometer) with the total area of the section, allowed realization of the total number of nerve fibers for the section. Average fiber areas for each coaptation method were acquired through analysis of the initial tracings.
Downloaded by: National University of Singapore. Copyrighted material.
behavioral observations would have provided a noninvasive means of assessing functional recovery.118 TENSILE STRENGTH RECORDINGS. TO measure the tensile strength of the repaired sciatic nerve in 28 nerves, the Model 7D Polygraph and DC. driver amplitude was used in conjunction with an FT03C model Force-Displacement transducer. The machine was calibrated before the testing of each nerve. The sciatic was removed from above the sciatic notch to an equal point above the trifurcation. A small hemostat was attached to the force transducer and another to a revolving motor rod. The nerve, approximately 2.5 cm in length, was held between each hemostat with no tension applied. After a base-line reading was established on the polygraph, the motor rod was then allowed to revolve, placing tension on the nerve. The polygraph measured the gradual addition of stress until separation of the nerve occurred. The force necessary to cause rupture of the repair site was referred to as the tensile strength. After testing both sciatics, the animal was sacrificed. CAP EVALUATIONS. Seventeen nerve CAP recordings were taken after a two-month interval. Each animal was anesthetized, as described earlier, and the sciatic nerve was exposed from the sciatic notch to beyond the point of trifurcation distally. The nerve was severed proximal to the sciatic notch to eliminate reflex activation via the spinal cord. Proximal branches of the sciatic nerve were transected to prevent muscular contraction. This removed the possibility of interference from this source and allowed for a monophasic CAP waveform. A bipolar stimulating electrode was then positioned under the sciatic nerve proximal to the repair site. The recording bipolar electrode was placed under the common peroneal branch in the popliteal fossa, (see Fig. 1). Electrical stimuli (from a Grass S48 Stimulator) were applied to the nerve in single square pulses of .05 msec duration. The strength of the excitation was gradually increased from threshold voltages until a "maximum" CAP waveform amplitude was achieved. Ten of these saturated CAP responses for each nerve were collected and averaged by a Cromemco System III computer equipped with an analog-todigital converter. The data were stored for subsequent analyses using a waveform editing analysis program. Using this program, onset and peak latencies, waveform areas, and peak amplitudes were determined. The distance between the stimulating and recording electrodes was measured in each case and this information, along with the latency data, were converted to conduction velocities to be used later in statistical evaluation.
RESULTS TENSILE STRENGTH. Data collected from both assessment groups (tensile strength and CAP/Histology evaluation) for all repairs were analyzed through an analysis of variance (ANOVA). The evaluation of tensile strength results concluded that, at two and four weeks, there was a trend toward stronger coaptations achieved by the suture groups, compared to the nonsuture repairs, although these data were not statistically significant. However, at eight weeks, the tensile strength was comparable in both repair groups (Table 2).
REPAIR SITE
t
(CAP.HISTO)
t
(HISTO)
t
(CAP.HISTO)
Figure 2. Biopsy sites (CAP: Compound Action Potential recordings; HISTO: Histologic evaluation).
333
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 4
Table 2. Tensile Strength Averages (± S.D.) for Suture and Tissue-Glue Repairs at Two, Four, and Eight Weeks Two weeks Suture (5 nerves) Glue (4 nerves) Four weeks Suture (5 nerves) Glue (5 nerves) Eight weeks Suture (5 nerves) Glue (4 nerves)
203.40 g ±61.64) 179.57 g ±65.35) 204.52 g ±96.12) 171.88 g ±96.19)
Onset latency (msec) Peak latency (msec) Peak amplitude (mV) Area (mV)(msec) Conduction velocity (Ms) 334
Quantitative Morphometric Means Suture
Proximal Thickly myelinated Thinly myelinated Total Distal Peroneal branch (distal)
5864 2972 8836 6177 2386
(±330) (±184) (±338) (±704) (±367)
Glue 5481 3341 8822 5357 1690
(±275) (±537) (±624) (±642) (±386)
Summary of CAP Waveform Data Suture
Glue
0.8672 (±0.4723)
0.9433 (±0.8743)
1.4027 (±0.7467)
1.4009 (± 1.2369)
0.4543 (±0.3231)*
0.1827 (±0.2939)*
0.0005 (±0.0004)* 24.167 (±13.374)*
0.0001 (±0.0002)* 13.090 (± 11.578)*
'Indicates differences were statistically significant.
neal branch also provided findings (distal to the repair) which were not significant in a direct comparison of the two types of nerve repair. For that branch, the average diameters measured were 0.20 \xm ± 0.02 for microsutured nerves and 0.29 |xm ± 0.06 for the fibrin glued sciatics. DEHISCENCE. We also discovered the failure of the fibrin glue to sustain neural coaptation in three of 23 tissue glue repairs or 13 percent. There was no dehiscence in any of the nerves coapted with microsutures. AUTOTOMY. For nearly all cases of heel edema, there was autotomy directed at the toes of one or both feet. This finding did not support behavioral testing in these animals.
DISCUSSION This study tried to address the following questions: Does the tissue glue adhesive, which may accidentally fall between the nerve stumps, impede axon growth and successful regeneration? Does the adhesive provide comparable or even greater tensile strength than that afforded by standard microsuture? Must the fibrin clot junction be supplemented by additional means? Analysis of the tensile strength data did not show any statistical difference between the two methods. However, at two and four weeks, there was a tendency toward stronger bonding accomplished by the conventional suture. This trend disappeared by eight weeks, when the tensile strengths of both coaptation types were comparable. Fibrin clots apparently can provide adequate strength when allowed to set completely and when tension is maintained at a minimum shortly after repair. Our three cases of dehiscence show that tension immediately after repair must be avoided, if a long-lasting reattachment of nerve ends is desired. Tension reduction seems to be a concern only immediately after repair. Following that period of solidification, the tissue adhesive demonstrated adequate bonding.
Downloaded by: National University of Singapore. Copyrighted material.
258.98 g (±90.05) 251.15 g (±123.87)
COMPOUND ACTION POTENTIALS. For CAP recordings analysis, the use of ANOVA was supplemented by Duncan's statistical test. Our findings can be reviewed in Table 3. They reveal that the onset and peak latencies were not statistically significant. However, the areas under the curve (p < .01), peak amplitudes {p = .05), and conduction velocities at onset {p = .05) were found to favor the suture group significantly (Table 3). HISTOLOGY OF THE REPAIR SITE. Quantitative morphometric means for both the microsuture and tissue glue groups indicated no statistically significant difference either proximal or distal to the repair site (Table 4). In studying the nerve cross sections with the method used previously, two predominant populations of fibers were present in the proximal sections only: normally myelinated fiber groups, and those which were finely myelinated.' Axon fibers found at the distal position consisted of one finely myelinated group (Fig. 4). Tallies for the peroneal branch (distal to the coaptation of the nerve) also showed no statistical difference between the two techniques (see Table 4). The axonal areas proximal to the site of the repair proved not to hold statistical significance (0.74 |xm2 ± 0.04 for the microsutured nerves versus 0.75 |xm2 ± 0.05 for the fibrin coaptations). At 7 mm distal to the repair, areas proved again to be statistically equivalent for the microsuture (0.61 |xm2 ± 0.08) and tissue glue (0.47 (xm2 ± 0.06) procedures. When the findings from the two nerves were compared, again no statistical difference was present. Further, measurement of axonal areas of the pero-
Table 3.
Table 4.
OCTOBER 1990
FIBRIN GLUE VERSUS MICROSUTURE/MARAGH, MEYER, DAVENPORT, ET AL.
cases of tension-free repairs would be redundant. Often, other authors added a relief suture, either temporarily or as a permanent feature. 71415 The research goal of fibrin adhesive experiments is to discover if replacement of suture with a better technique is possible. The electrophysiologic data in this study showed that nerve conductivity was superior for the microsuture repairs (see Fig. 3). When reviewing peak amplitudes, areas, and conduction velocities, these were significantly different in favor of the conventional suture technique. Moy et al.'5 noted the same disparity in CAP amplitudes at 18 weeks. Thus, the CAP data suggest a lower probability of neural regeneration in the tissue glue repairs. Whether this finding of diminished conductivity is due to the effects of the interposition of a foreign protein between the nerve ends remains unknown. Another possibility is that the axonal penetration through tissue glue repairs may be subjected to early tension forces that are not present at the suture repairs. These forces, due to early mobilization of the extremity following repair, may invite scar formation that hinders nerve regeneration. Microsuture repair has the advantage of precisely positioning and holding the two nerve ends. Thus,
4.88 C400) mSEC
C4003
Downloaded by: National University of Singapore. Copyrighted material.
The dehiscence findings of this study correspond in part with previous reports. Cruz et al.,]3 in a similar comparison of tissue glue and microsuture material, reported an 80 percent dehiscence rate for nerves adhered by means of a fibrin clot. The study conducted by Haase et al22 found total adhesive coaptation dehiscence, although no other study reported such an extreme failure rate. 5610 Perhaps immobilization immediately after repair should be recommended as it was by Matras. 111619 Several other reports convey the same warning of unchecked tension leading to dehiscence shortly after glue repair. 78202123 Specifically, the surgical area must remain free of stress forces for at least a two-week period, and even longer when comparable tensile strengths are provided by the two groups. It should be emphasized that, in the present study, the animals were not immobilized after the nerve repair procedures, but were allowed to move freely in their cages. Despite the lack of immobilization, only three nerves dehisced in the present study. Apparently, a generous degree of mobilization of the nerve ends can generate a tension-free environment at the coaptation site. Although an adjuvant is required, use of suture in these
6.10 mSEC
B
4.88
e
C4003
-.400
FILE LI 5
RUN
5
6.10 mSEC
4.88
FILE R15
RUN
6.10 C400) mSEC
5
Figure 3. Examples of actual computer tracings from recordings obtained from two animals. CAP recordings across suture repairs are depicted in A and C while CAPs obtained across the tissue glue repairs are shown in B and D. Note superiority in peak amplitude, area, and conduction velocity of the suture-repaired nerves. (All recordings obtained at same gain (5000).)
335
OCTOBER 1990
r B
Figure 4. Photomicrographs depicting the axonal regeneration in the two sciatic nerves of the same animal using the two repair techniques. A, Cross section of fibrin adhesive repaired nerve 7 mm distal to the repair site. B, Microsuture repair at same site (methylene blue, X800).
336
there is some certainty that the neural coaptation will be of a permanent nature. In contrast, the gluing technique attempts to preserve only approximation of the nerve ends. In either case, data showing conduction velocity differences imply both that apposition by fibrin glue resulted in poorer myelination, and that regrowth across the lesion by proximal axons occurred through a more inadequate coaptation. These findings differ from those reported by Smahel et al.]4 who showed statistically comparable CAP results in the same experimental model. However, in Smahel's study, a relief suture and silicone tubing was placed across the junction site in the glued repairs, while no such adjunctive measures were used in our study. Our collected histologic data seem to encourage the conclusion of CAP findings. Overall, the quantitative morphometry of the two repairs was comparable, although there were discrepancies in the axon counts that favored the suture repair. There were, however, no statistical differences in the morphologic data. These morphologic findings agree with the reports of Smahel' 4 and Matras16 but contradict the early study of Singer21 and the clinical findings of Egloff6 which postulated that axonal growth was stunted with fibrin adhesive use.
CONCLUSIONS A double-blind study comparing sutureless versus microsuture nerve repairs in the rat sciatic nerve yielded the following results: Adequate tensile strength can eventually be accomplished by both types of repair, if a tension-free environment is provided in the early postoperative period. Once the technique of using tissue glue is mastered, it can offer certain advantages in the clinical arena, including the possibility of executing nerve repairs more rapidly than with the microsuture technique. It can facilitate inaccessible nerve repairs, such as within the bony confines of the skull or within difficult to reach bony canals. It can also make nerve approximation possible between extremely diminutive nerves. However, this study has shown that the interposition of tissue glue between the nerve ends results in a somewhat smaller population of comparable diameter regenerating axons reaching the distal nerve branches; more important, the conductive properties of these axons seemed to be compromised in the tissue glue repairs. Sutureless repairs hold great promise in the fu-
Downloaded by: National University of Singapore. Copyrighted material.
JOURNAL OF RECONSTRUCTIVE MICROSURGERY/VOLUME 6, NUMBER 4
FIBRIN GLUE VERSUS MICROSUTURE/MARAGH, MEYER, DAVENPORT, ET AL. 10. Tarlov IM, Benjamin B: Plasma clot and silk suture of nerves. Surg Gynecol Obstet 76:366, 1943 11. Matras H, Dinges HP, Mamoli B, Lassmann H: Non-sutured nerve transplantation: A report on animal experiments. ) Maxillofac Surg 1:37, 1973 12. Becker CM, Gueuning CO, Graff GL: Sutures or fibrin glue for divided rat nerves: Schwann cell and muscle metabolism. Microsurg 6:1, 1985 13 Cruz NI, Debs N, Fiol RE: Evaluation of fibrin glue in rat sciatic nerve repairs. Plast Reconstr Surg 78:369, 1986 14. Smahel H, Meyer VE, Bachem U: Gluing of peripheral nerves with fibrin: Experimental studies. J Reconstr Microsurg 3: 211, 1987 REFERENCES 15. Moy 01, Peimer CA, Koniuch MP, et al: Fibrin seal adhesive versus nonabsorbable microsuture in peripheral nerve repair. I Hand Surg 13A:273, 1988 Maragh H. Hawn RS, Gould ID, Terzis IK: Is laser nerve repair 16. Matras H, Braun F, Lassmann H, et al.: Plasma clot welding of comparable to microsuture coaptation? I Reconstr Microsurg 4:189, 1988 nerves: Experimental report. I Maxillofac Surg 1:236, 1973 Miehlke A: Surgery of the Facial Nerve, 2nd ed., Philadelphia: W. B.17. Ventura R, Torri G, Campari A, etal: Experimental suture of the Saunders Co., 1973 peripheral nerves with "fibrin glue". Ital I Orthop Traumatol 6:407, 1980 Wigand ME, Thumfart W: Neurosynthesis of the facial nerve: 18. Hasewaga K: A new method of measuring functional recovery Electrical versus clinical results. In: Samii M, lannetta PJ (eds): The Cranial Nerves. Berlin: Springer-Verlag, 1981 in unanesthetized and unrestrained rats. Experientia 34: Kuderna H: The use of fibrin sealing in peripheral nerve repair 272, 1978 In: Skjoldborg H (ed): 1MMUN0 Scientific Workshop 82. 19. Matras H: Fibrin seal: State of the art. I Oral Maxillofac Surg 43: Tisseel®/Tissucol®-Symposiutn: Areas of Application, Problems and 605, 1985 Perspectives in Current Surgery, Scanticon-Aarhus, Denmark,20. Tarlov IM, Denslow C, Swarz S, Pineles D: Plasma clot suture of 1982 nerves Arch Surg 47:44, 1943 Osgaard O: Tisseel® in peripheral nerve and cranial nerve 21. Singer M: The combined use of fibrin film and clot in end-tosurgery In: Skjoldborg H (ed): IMMUNO Scientific Workshop end union of nerves. I Neurosurg 2:102, 1945 '82: Tisseel®/Tissucol®-Symposium: Areas of Application, Problems 22. Haase I, Andreasen H, Falk E, et al.: Use of fibrin seal (tisseel/ and Perspectives in Current Surgery, Scanticon-Aarhus, Dentissucol) in peripheral nerve surgery: An experimental rat mark, 1982 model In: Schlag G, Redl H (eds): Fibrin Sealant in Operative Egloff DV, Narakas A: Nerve anastomoses with human fibrin: Medicine: Ophthalmology-Neurosurgery, Vol. 2, Berlin: SpringerPreliminary clinical report (56 cases) Ann ChirMain 2:101, Verlag, 1986 1983 23 Matras H, Kuderna H: Glueing nerve anastomoses with clotting Egloff, DV, Narakas A, Bonnard C: Results of nerve grafts with substances. In: Marchac D, Hueston |T (eds): Transactions of tissucol (tisseel) anastomosis. In: Schlag G, Red! H (eds): the Sixth International Congress of Plastic and Reconstructive SurFibrin Sealant in Operative Medicine-. Ophthalmology-Neurogery. Paris, August 24-29, 1975. Paris: Masson, 1976 surgery, Vol 2, Berlin: Springer-Verlag, 1986 24. Vinazzer H: Fibrin sealing: Physiologic and biochemical background Fac Plast Surg 2:291, 1985 Palazzi S: The use of fibrin sealant in nerve adhesions (peripheral nerves and plexus brachialis). In: Schlag G, Redl H (eds): Fibrin Sealant in Operative Medicine. OphthalmologyWe extend our gratitude to Mary Jacobs for electroNeurosurgery, Vol 2, Berlin: Springer-Verlag, 1986 physiologic recordings, to Susan Downing and Jeff Dupree for Young ]Z, Medawar PB: Fibrin suture of peripheral nerves: histologic assistance, and to IMMUNO for use of their prodMeasurement of the rate of regeneration. Lancet 239:126, uct and literature assistance. 1940
337
Downloaded by: National University of Singapore. Copyrighted material.
ture of clinical microneurovascular surgery. At present, use of tissue glue seems to be associated with some adverse effects. An intense effort in the future would be welcomed, to try to substitute glues prepared from the patient's own serum for existing fibrin adhesives.
