J. BIOMED. MATER. RES.
VOL. 1 1 , PP. 297-314 (1977)
Effect of Collagen Crosslinking on the Rate of Resorption of Implanted Collagen Tubing in Rabbits M. CHVAPIL, J. A. OWEN, and D. S. CLARK, Division of Surgical Biology, Department of Surgery, Arizona Medical Center, Tucson, Arizona 857.24, and S. KOORAJIAN and A. P. GOODMAN, Research Division of Edwards Laboratories, Santa A n a , California 96705
Summary Collagen tubes were tanned with glutaraldehyde for different periods of time. Some were oxidized with periodate and sterilized with either @ C ‘o (1.5 Mrad) or propylene oxide. The tubes were coated with polyfilamentous polyester fabric, filled with x-ray contrast material, and implanted subcutaneously in rats and rabbits. Rate of resorption was ascertained by x-ray procedure of progressive leakage of contrast material. A close relation between tanning time of the collagen fabric-combined prostheses and rate of their resorption in subcutis was found in both rats and rabbits. I n rabbits, however, the implants were resorbed at a significantly faster rate than in rats. No effect of oxidation of collagen on the resorption was observed. Collagen tubes tanned for shorter time periods and sterilized with propylene oxide were more resistant t o degradation than those sterilized with irradiation. This difference was absent, however, with material tanned for longer times. The mechanism of resorption of implanted collagen tubes was studied by morphological methods. The role of inflammatory cells in resorption is documented. The paper indicates the advantages as well as limits of the x-ray method of studying the resorption rate of biodegradable materials.
One of the important features of using collagen as a biodegradable material for construction of various medical prostheses is its controllable resorption (for review see refs. 1 and 2 ) . So, for instance, a long-term patency of collagen-Dacron vessel prosthesis depends on a programmed resorption of the collagen tubing within a definite 297 @ 1977 by John Wiley & Sons, Inc.
298
CHVAPIL ET AL.
time after implantation. Rapid resorption may be fatal, overly slow degradation of collagen tube will slow the rate of capillary ingrowth, formation of pseudoendothelium layer, and replacement of the prosthesis by tissue components mimicking the structure of normal vessel wall^.^^^ It was shown that the rate of collagen resorption at implant sites depends on the number of cohesive forces existing within a certain collagenous structure. Covalent crosslinks are the most important bonds as they contribute mainly to structural and functional stability of the collagenous structure. Thus, solubility, swelling and resorption by collagenase depends on the degree of crosslinking.' The crosslinking which occurs during biological maturation of collagen in the course of aging can be simulated in vitro by various tanning agents1 I n this study we tested the relation between degree of tanning of collagen and its in vivo resorption rate in two species, i.e., in rabbits and rats. We also studied the effect of two methods of sterilization and other treatment of collagen on the resorption kinetics.
MATERIAL AND METHODS Collagen The sample was obtained from Freudenberg Company (Weinheim, Germany) and consisted of purified collagen isolated from bovine skin by repetitive washings in 0.1 N HCI of cleaned and mechanically disintegrated skin. Amino acid analysis of the residual as determined on a Beckman Amino Acid Autoanalyzer produced the following results, given residues are per 1000 total residues: 2.4 Tyr, 97 Hyp, 312 Gly. Xo CySH was detected. This was an indicatian that the resulting collagen was reasonably pure. The collagen was extruded through a rotating head t o produce collagen tubes, 4 mm in diameter, with a wall thickness of 0.1 or 0.2 mm as indicated in appropriate experiments.
Tanning Procedure The extruded collagen tubes were air-dried, and treated for various time periods with 0.05 or O.lOyo glutaraldehyde, then extensively
RESORPTION RATE OF COLLAGEN TUBING
299
washed with distilled water to remove excess of tanning agent. The collagen tubes were then covered with a multifilamentous polyester fabric.
Oxidation A group of tubes were further processed by oxidizing with 1% periodate in 0.1 M Tris buffer (pH 7.5), a t 4°C for 24 hr in order to remove the carbohydrates present in the collagen.
Radio-opaque Filler Approximately 15 mm long segments of tube were filled with Lipidol (Savage Labs., Texas; poppyseed oil, 40% iodine w/w) and tied on both ends with a nonresorbable suture.
Sterilization The tubes were sterilized either by 6oCoirradiation (1.5 Mrad) or by 40/, propylene oxide solution in distilled water for 48 hr.
Animals (Rabbits) Thirty-six male New Zealand white rabbits (Heidt Rabbitry, Tucson, Arizona), weighing 1.3-1.8 kg were divided into two groups, A and B. I n group A, 16 rabbits were implanted with collagenpolyester tubes having a collagen wall thickncss of 0.1 mm, and tanned in o.0570 glutaraldehyde solution for 4, 8, 12, 16, 20, and 24 min a t room temperature. All samples were sterilized with 7-irradiation a t a dose of 1.5 Mrad. Each rabbit was implanted with tubes of different tanning times, according to a definite protocol, alternating the position of a certain implant in the group of rabbits. I n group B, 18 rabbits were implanted with samples having a waH thickness of 0.2 mm and tanned in 0.20% glutaraldehyde solution for 20, 30, 40, and 55 min a t room temperature. Half of the samples had been sterilized by ‘j0Coirradiation, while the other half were sterilized by propylene oxide. Further subdivision occurred through the oxidation of half of the samples from each of the above groups. Each subgroup originally included 9 samples. However, the number of implants of each subgroup varied due to the visible leakage from the tied ends of the tubes prior to implantation. These tubw were eliminated.
300
CHVAPIL ET AL.
Resorption Variability Determination of intra-individual variability of the rate of resorption by the procedure described below was followed in three rabbits, and implanted with eight identically processed prostheses (0.2 mm thick, 40 min tanning with 0.1% glutaraldehyde a t 20°C). We measured the time when 50% of implanted prostheses showed a leakage, clearly indicated in the x-ray films (Table I).
Rats Eighteen 250 g male Sprague-Dawley rats (Charles River) were implanted with the same type of prosthesis as the rabbits in group A. While 6-8 prostheses were implanted in the back region of rabbits, four implants were inserted into similarly positioned subcutaneous pockets of rats.
Implant Procedure Innovar Vet (Pittman Moore) (0.2 m1/2 kg body wt) was administered i.m. 15-20 min prior to surgery. After closely shaving the backs of the rabbits, the area was divided into six (group A) or eight (group B) zones. A 4 mm wide incision was made through to the fascia and a subcutaneous pocket was formed using a Kelly 5 in. curved hemostat. After inserting the grafts into the pocket, the wound was sutured with 000 proline suture (Ethicon) using a single stitch. Each rabbit receivcd randomly positioned tubes treated as described above.
Rate of Collagen-Fabric Tube Resorption This was studied in rabbits or rats immobilized by Innovar-Vet, twice weekly, using a Dynamax Model C-792A x-ray machine (Machlett Labs) with a Westinghouse Alodcl 3400 control panel. Medical x-ray, 14 x 17 type R film (3-M) was used. Diffuse opacity a t implanted tube site was considered evidence of leakage. Still, leakage was recognized only when it was present in x-ray reading of two consecutive time periods. X-Rays taken immediately after surgery were used as controls for comparison.
Histology Representative samplcs were evaluated by standard histological techniques using hematoxylin-eosin and Masson Trichrome stains.
RESORPTION RATE OF COLLAGEN TUBING
301
RESULTS Variability of the Measurements of the Rate of Collagen Tubes Resorption Although the x-ray method used in this study to observe the in vivo resorption of implantcd collagen tubes was first described by Chvapil and Rrajicek in 1963, the error of the method has never been determined. This was now accomplished by implanting eight identical collagen tubings into each of three rabbits (Table I ) . We found that the variation coefficient of the resorption rate of tubings in each rabbit is close to 10% of the mean. Interindividual variability based on resorption time of 50% of implanted identical tubings into three rabbits was 5.670.
Effect of Tanning A definite relation between tanning time of the collagen tubes and their rate of resorption by the subcutaneous tissue was found in both groups of rabbits (A and B). A typical flow sheet of the results is presented in Table I1 indicating the percent of resorbed prostheses in time as a function of the tanning in group A of rabbits. Similar summations of results in group B rabbits are presented in Table 111, which shows also the effects of other treatments, such as sterilization and oxidation. Plotting the time when the leak occurred in days against the tanning time of the collagen tubing as derived from group B experiments, TABLE I Variability of the 1)etermiriation of the Rate of Resorptioii of Implanted Collagen Tubes in Rabbits
Rabbita
Days of leakage of 50% implanted tubes
1 2 3
37.0 4.9 41.5 & 5 . 0 42.4 f 5 . 1
*
Variation coefficientIi 13.2 12.0 12.1
a Eight identical sponges implanted iri each rabbit; x-rays made in 4 day intervals. (X & SD.) Gives SD in percent of the mean value.
CHVAPIL ET AL.
302
TABLE I1 Resorption Studies Showing Percent of Leakage of Collagen-Dacron Implants in Rabbits (Subcutaneous Tissue)s
Days
Tanning Time
4 8 12 16 20 24
14
21
28
35
42
49
56
66
62 62 50 12 0 12
100 62
100 100 100 62 50 50
100 100 100
100 100 100 87 75 75
100 100 100 100 75 75
100 100 100 100 87 75
100 100 100 100 100 87
75 12
0 12
75 62 75
* Sixteen rabbits (male 1.5-2 kg) were implanted with six sponges each. Leakage time ascertained by x-ray. Data is presented as percent of leaking implants from total number of implants for each tanning group.
a progressively slower resorption (progressive leaks) of more crosslinked collagen tubes is shown in Figure 1. As evident in Tables I1 and I11 and in Figure 1, the variability of the results did not exceed 15y0of the mean value, when 8-10 animals were used to test a single variable. Collagen I sorption
T
20 30 40 55
20 30 4 0 55 Tanning Time (min)
Fig. 1. Effect of tanning on the rate of resorption of collagen-fabric tubings. Composite collagen tube-fabric prostheses were implanted in the subcutis of rabbits. Resorption is reported as days after implantation when the first progressive leak of radio-opaque material from the tube was observed i n the x-ray film. Variability given as mean f SEX.
RESORPTION RATE OF COLLAGEN TUBING
303
TABLE I11 Statistically Averaged Peak Resorption (in Days) Graft #
Group
Mean f S.E.
~
Radiation-Sterilized Nonoxidized 20" 30 40 55
21.8 25.3 29.6 34.5
i 2.33 f 3.77 f 3.95 f 5.57
Oxidized 20 30 40 55
21.9 30.7 37.3 43.6
f 2.72 f 4.90 f 4.39 f 8.11
26.1 27.5 35.5 38.8
f f f f
28.5 33.6 34.6 44.6
f 4.52 f 6.23 f 4.88 f 3.07
Propylene Oxide-Sterilized Nonoxidized 20 30 40 55 Oxidized 20 30 40 55
5 6
7 8
(8)
(8) (8) (8)
4.42 3.05 3.11 5.09
Tanning time (in min). Graft # 16 showed only 75% resorption after 77 days. Therefore, only the resorbed grafts were used in calculations. 0 Number grafts used in calculations. a
b
Effects of Sterilization and Oxidation Although evidence was presented that ' W o irradiation may under certain circumstances result in decreasing the structural stability of collagen21 (higher dose, presence of humidity, production of free radicals), we did not observe any statistically significant difference between sterilization of collagen tubes with 6oCo irradiation (1.5 Mrad) and propylene oxide (Table 111). This conclusion obviously reflects a large variability of the results. Combined data including both oxidized and nonoxidized implants, tanned for varying times
CHVAPIL ET AL.
304
TABLE IV Effect of Two Methods of Sterilization on the Resorption Rate of Collagen-Dacron Composite Tubes ~~
Resorption Rate (in days) Sterilization Method Irradiation (@WO, 1.5 Mrad) Propylene oxide
20 mina
+
21.9 1.8;bp 28.0 f 2.1
30 minc
< 0.05
28.0 f 2 . 3 ; p 36.1 f 3 . 2
< 0.05
* Refers to tanning time in min. b Variability is given by X + SEM.
with glutaraldehyde (Table IV) , showed significantly slower resorption of collagen tubes tanned for 20 or 30 min and sterilized with propylene oxide. I n pooled data for tubes tanned for 40 or 55 min, no effect of sterilization method was noticed.
Effect of Species A striking difference in the rate of resorption of similarly treated collagen fabric composite tubes was found between rats and rabbits, as indicated in Figure 2. Only data on collagen tubes tanned for 4 and 8 min are shown. Collagen tubes tanned for 4 min and implanted into rabbits were rcsorbed at a much faster rate than those in rats. About 50% of the implants were resorbed at 13 days in rabbits, in rats at approximately 21 days. The difference of resorption rates in rats and rabbits was highly significant ( p < 0.001) and held for samples tanned for cithcr timc pcriod.
Morphology of the Dynamics of the Resorption For the sake of simplicity, only histological results from A series of rabbit experiments are presented. Similar conclusions were obtained in B series, however. It should be kept in mind that the implanted material consisted of collagen tubing coated with multifilamentous polyester fabric mesh. It has been well established that various fabric materials induce differing tissue reactivity and cell ingrowth.2 After subcutancous implantation, progressive infiltration of the fabric with inflammatory cells consisting mainly of
RESORPTION RATE OF COLLAGEN TUBING
305
late of Resorption of Collagen in Rabbits and Rats
Days after Implantation
Fig. 2. Species difference in the rate of collagen tube resorption. Identical composite collagen tube-fabric prostheses, tanned for 4 and 8 min, were implanted in the subcutis of rabbits and rats. Rate of resorption is given in percent of progressively leaking implants of a total number of tubes implanted. Variability is presented as mean f SEM.
granulocytes and macrophages was seen in the granuloma tissue pcrmeating the fabric multifilamentous meshwork (Fig. 3). With longer time, the spaces between fibers (f) filled with more voluminous granuloma tissue containing mostly macrophages and several fusiform-shaped cells (Fig. 4). In the next stage, depending on the degree of tanning of the collagen tube, a beginning loosening of the continuity of collagen membrane (c) was seen, mainly that portion in contact with the interfibrillar spaced rich in granuloma tissue with macrophages (Fig. 5 ) . I n some loci (see arrow in Fig. S), cells migrated inside the collagen membrane which showed noncontinuity in structure and characteristic change in collagen stainability by Trichrome stain (Fig. 7) from typical blue to red; this
306
CHVAPIL ET AL.
Fig. 3. Composite collagen tube-fabric implant in subcutis of rabbit, 10 days after implantation. (f : polyester fibers; c: collagen membrane.) Inflammatory cells of the reactive granuloma tissue are accumulating mainly in between fabric fibers; collagen membrane intact. 40X.
may indicate the denaturation of collagen, as proposed by Craik and M ~ N e i l . ~ The corrosion of the collagen tube membrane was definitely not uniform over all the tubing. Ultimately the collagen membrane was completely permeated by cells, mainly macrophages, to such a n extent that Lipidol leaked out of the tube (Fig. 8). I n highly tanned collagen tubings, the described degradation of the collagen wall was negligible for periods of 2 months or even longer. In these situations, there was a continuous fibrous layer of condensed collagen fibers formed around the implanted collagen fabric tubing.
RESORPTION RATE OF COLLAGEN TUBING
Fig. 4.
307
Proliferation of inflammatory cells with beginning destruction of outer layer of collagen membrane, 14 days after implantation. 1OOX.
DISCUSSION The rate a t which collagen products are resorbed in vivo depends on several factors, the degree of tanning and the site of implantation*z6J5being the most important. Although tissue collagenases and possibly also cathepsins (B, D) are responsible for the degradation of collagen implant~,’~~17*18 it is mainly the macrophage which has been implicated in the resorption o f collagen a t the cellular l e ~ e l . ~ There ~ ~ ” is enough evidence that macrophage infiltration occurs around implanted collagen ~ u t u r e s ,sponge^,^.^ ~ and vessel prostheses.’OJl Vascular grafts are gradually invaded by inflammatory cells and fibroblasts and replaced by new collagen and
308
CHVAPIL ET AL.
Fig. 5 . Swollen collagen membrane penetrated by inflammatory cells. The compact structure of collagen membrane is loosened. 100X .
connective tissue structures. It has been our personal experience that the entire graft is gradually replaced, regardless of the degree of t a n n i ~ ~ g .However, ~.~ some authors have found graft collagen preserved and r e c e l l ~ l a r i z e d . ~ ~Inflammatory -~~ cells such as granulocytes or macrophages contain collagenase as well as other potent proteases.lg It was also shown that the rate of collagen degradation by collagenase is inversely related to the degree of crosslinking.20 The results of our study, in agreement with the above findings, show that thc rate of the resorption of the implanted collagen material depends on the degree of tanning and crosslinking of collagen structures. Of interest is the species diffcrcnce in the resorption kinetics,
RESORPTION RATE OF COLLAGEN TUBING
309
Fig. 6. Maximum penetration of inflammatory cells with destruction of collagen membrane occurs between fabric fibers. Although collagen membrane is almost completely invaded by mostly macrophages, this membrane still did not leak a t 27 days after implantation. Arrow points to the site of cellular infiltration of collagen membrane. 160X .
for which we could not find any supporting evidence in the available literature. We believe that the ((species difference” does not reflect the difference in age, as both rats and rabbits could be classified as young adults. I n the wound healing problems, it is well known that some species such as man and horse heal-synthesize as well as degrade collagenmuch faster than others (rat, mouse). Our finding of faster degradation of implanted collagen in rabbits than in rats may just reflect higher turnover of collagen in this species.
310
CHVAPIL ET AL.
Fig. 7. Polyester polyfilamentous fibers are surrounded by inflammatory cells with maximum penetration into the collagen membrane between the fibers. On the outer side of the fabric, a continuous sheet of collagen fiber is formed (28 days after implantation). 30X.
Sterilization of collagen related biomaterial by 6oCoirradiation or by propylene oxide is commonly used. Still, some modifications of collagen biomaterial by either method has been indicated.lJ1 I n the case of 6oCosterilization, we observed faster resorption in two groups of collagen tubes, indicating a slight denaturation of the collagen material during sterilization even a t a relatively low dose of 1.5 Mrad. Treatment of collagen with periodate did not affect the resorption characteristics of implanted tubes. Although we described the x-ray method of following the rate of collagen resorption more than 10 years & ~ O , ~ J O we have been using
RESORPTION RATE OF COLLAGEN TUBING
311
Fig. 8. Complete penetration and destruction of collagen membrane by inflammatory cells which invade the membrane in streaks, leaving isolated islands of collagen structures still intact. Originally continuous and compact membrane is disintegrated and leaks. lOOx.
this technique almost on a routine basis since and accumulated new experience which could be summed up as follows. I n rats, a maximum of 4 tubes and, in rabbits, 6-8 tubes can be implanted and evenly located in the subcutis of the dorsum of the animal (Figs. 9a and 9b). Immediately after implantation of collagen tubes filled with x-ray opaque material, the boundaries of tubes should be sharp. Tubes showing a leak through the suture line should be eliminated from final statistics. We also eliminated those leaks that occurred within the first 10 days after implantation because maximum cumulative resorption occurred at two to three times longer that the usual time
312
CHVAPIL ET AL.
Fig. 9. Typical x-ray taken immediately after subcutaneous implantation of collagen-fabric composite material into rats (top) or rabbits (bottom). Note the sharp delineation of implants.
period. We considered these leaks a reflection of mechanical rupture of the wall of the collagen tube, possibly caused during manipulation of the sample before or during implantation. Only progressive leakage from the same tube in two consecutive periods were considered indicative of resorption (Fig. 10). We consider this method satisfactory when discrimination of sensitivity a t 5-7 days would be
RESORPTION RATE OF COLLAGEN TUBING
313
Fig. 10. Various stages of the resorption of collagen tubes in x-ray examination are shown by pictures A to E . Picture F refers to nonspecific leak of opaque material through suture.
requested. Still, resorption time of the same sample should be tested a t least in 8-10 implantation sites. This research was supported by funds from Edwards Laboratories, Santa Ana, California.
References 1.
-U.Chvapil, R. L. Kronenthal, and W. van Winkle, Jr., International Review of Connectzve Tissue Research 6, Academic Press, New York, 1973, p. 1.
2. M. Chvapil and M. Krajicek, Lagenbecks Arch. Klin. Chir., 318, 80 (1967). 3. 31. Chvapil and M. Krajicek, J . Surg. Res., 11, 358 (1963). 4. M.Krajicek, V. Zastava, and RI. Chvapil, J . Surg. Res., 4, 290 (1964). 5. J. E. Craik arid I. R. R. McNeil, Nature, 209, 931 (1966). 6. H. Ungar, Amer. J. Pathol., 29, 973 (1953). 7. T . N . Salthouse, J. A. Williams, and D. A. Willigan, Surgery, 129, 690 (1969). 8. &I.Chvapil, R. Holusa, K. Kliment, and M. Stoll, J . Bzomed. Muter. Res., 3, 315 (1969).
314
CHVAPIL ET AL.
9. D. E. Cutright, W. R. Posey, S. N. Bhaskar, and W. J. Larson, Oral Surg. Oral Med. Oral Pathol., 35, 144 (1973). 10. J. Hejnal, M. Krajicek, and B. Zastava, Vessel Substitutes, Academia Czechoslovakia, 1967. 11. M. Chvapil and M. Krajicek, Vascular Surgery, Butterworths, London, 1970, p. 120. 12. A. D. Potenza, J. Bone Joint Surg., 44,49 (1962). 13. A. D. Potenza, J. Bone Joint Surg., 46, 1462 (1964). 14. L. Klein, Ph.D. Thesis, Boston University, Boston, Mass., 1958. 15. M. J. White, I. Kohno, A. L. Itubin, and K. H. Stenzel, Biomat. Med. Dev., Art. Org., 1(4), 703 (1973). 16. A. Z. Eisen, E. A. Bauer, and J. J. Jeffrey, J. Invest. Dermatol., 55,359 (1970). 17. G. S. Lazarus and J. F. Goggins, Science, 186, 653 (1974). 18. P. B. Robertson, It. B. Ityel, It. E. Taylor, K. W. Shyr, and H. M. Fullmer, Science, 177, 64 (1972). 19. L. M. Wahl, S. M. Wahl, S. E. Mergenhagen, and G. R. Martin, Science, 187, 261 (1975). 20. M. E. Grant and 1).J. Prockop, New Engl. J. Med., 55, 242 (1972). 21. A. J. Bailey, Znt. Rev. Connect. Tissue Rea., 4, 268 (1968).
Received December 8, 1975 Revised July 15, 1976
VOL. 1 1 , PP. 297-314 (1977)
Effect of Collagen Crosslinking on the Rate of Resorption of Implanted Collagen Tubing in Rabbits M. CHVAPIL, J. A. OWEN, and D. S. CLARK, Division of Surgical Biology, Department of Surgery, Arizona Medical Center, Tucson, Arizona 857.24, and S. KOORAJIAN and A. P. GOODMAN, Research Division of Edwards Laboratories, Santa A n a , California 96705
Summary Collagen tubes were tanned with glutaraldehyde for different periods of time. Some were oxidized with periodate and sterilized with either @ C ‘o (1.5 Mrad) or propylene oxide. The tubes were coated with polyfilamentous polyester fabric, filled with x-ray contrast material, and implanted subcutaneously in rats and rabbits. Rate of resorption was ascertained by x-ray procedure of progressive leakage of contrast material. A close relation between tanning time of the collagen fabric-combined prostheses and rate of their resorption in subcutis was found in both rats and rabbits. I n rabbits, however, the implants were resorbed at a significantly faster rate than in rats. No effect of oxidation of collagen on the resorption was observed. Collagen tubes tanned for shorter time periods and sterilized with propylene oxide were more resistant t o degradation than those sterilized with irradiation. This difference was absent, however, with material tanned for longer times. The mechanism of resorption of implanted collagen tubes was studied by morphological methods. The role of inflammatory cells in resorption is documented. The paper indicates the advantages as well as limits of the x-ray method of studying the resorption rate of biodegradable materials.
One of the important features of using collagen as a biodegradable material for construction of various medical prostheses is its controllable resorption (for review see refs. 1 and 2 ) . So, for instance, a long-term patency of collagen-Dacron vessel prosthesis depends on a programmed resorption of the collagen tubing within a definite 297 @ 1977 by John Wiley & Sons, Inc.
298
CHVAPIL ET AL.
time after implantation. Rapid resorption may be fatal, overly slow degradation of collagen tube will slow the rate of capillary ingrowth, formation of pseudoendothelium layer, and replacement of the prosthesis by tissue components mimicking the structure of normal vessel wall^.^^^ It was shown that the rate of collagen resorption at implant sites depends on the number of cohesive forces existing within a certain collagenous structure. Covalent crosslinks are the most important bonds as they contribute mainly to structural and functional stability of the collagenous structure. Thus, solubility, swelling and resorption by collagenase depends on the degree of crosslinking.' The crosslinking which occurs during biological maturation of collagen in the course of aging can be simulated in vitro by various tanning agents1 I n this study we tested the relation between degree of tanning of collagen and its in vivo resorption rate in two species, i.e., in rabbits and rats. We also studied the effect of two methods of sterilization and other treatment of collagen on the resorption kinetics.
MATERIAL AND METHODS Collagen The sample was obtained from Freudenberg Company (Weinheim, Germany) and consisted of purified collagen isolated from bovine skin by repetitive washings in 0.1 N HCI of cleaned and mechanically disintegrated skin. Amino acid analysis of the residual as determined on a Beckman Amino Acid Autoanalyzer produced the following results, given residues are per 1000 total residues: 2.4 Tyr, 97 Hyp, 312 Gly. Xo CySH was detected. This was an indicatian that the resulting collagen was reasonably pure. The collagen was extruded through a rotating head t o produce collagen tubes, 4 mm in diameter, with a wall thickness of 0.1 or 0.2 mm as indicated in appropriate experiments.
Tanning Procedure The extruded collagen tubes were air-dried, and treated for various time periods with 0.05 or O.lOyo glutaraldehyde, then extensively
RESORPTION RATE OF COLLAGEN TUBING
299
washed with distilled water to remove excess of tanning agent. The collagen tubes were then covered with a multifilamentous polyester fabric.
Oxidation A group of tubes were further processed by oxidizing with 1% periodate in 0.1 M Tris buffer (pH 7.5), a t 4°C for 24 hr in order to remove the carbohydrates present in the collagen.
Radio-opaque Filler Approximately 15 mm long segments of tube were filled with Lipidol (Savage Labs., Texas; poppyseed oil, 40% iodine w/w) and tied on both ends with a nonresorbable suture.
Sterilization The tubes were sterilized either by 6oCoirradiation (1.5 Mrad) or by 40/, propylene oxide solution in distilled water for 48 hr.
Animals (Rabbits) Thirty-six male New Zealand white rabbits (Heidt Rabbitry, Tucson, Arizona), weighing 1.3-1.8 kg were divided into two groups, A and B. I n group A, 16 rabbits were implanted with collagenpolyester tubes having a collagen wall thickncss of 0.1 mm, and tanned in o.0570 glutaraldehyde solution for 4, 8, 12, 16, 20, and 24 min a t room temperature. All samples were sterilized with 7-irradiation a t a dose of 1.5 Mrad. Each rabbit was implanted with tubes of different tanning times, according to a definite protocol, alternating the position of a certain implant in the group of rabbits. I n group B, 18 rabbits were implanted with samples having a waH thickness of 0.2 mm and tanned in 0.20% glutaraldehyde solution for 20, 30, 40, and 55 min a t room temperature. Half of the samples had been sterilized by ‘j0Coirradiation, while the other half were sterilized by propylene oxide. Further subdivision occurred through the oxidation of half of the samples from each of the above groups. Each subgroup originally included 9 samples. However, the number of implants of each subgroup varied due to the visible leakage from the tied ends of the tubes prior to implantation. These tubw were eliminated.
300
CHVAPIL ET AL.
Resorption Variability Determination of intra-individual variability of the rate of resorption by the procedure described below was followed in three rabbits, and implanted with eight identically processed prostheses (0.2 mm thick, 40 min tanning with 0.1% glutaraldehyde a t 20°C). We measured the time when 50% of implanted prostheses showed a leakage, clearly indicated in the x-ray films (Table I).
Rats Eighteen 250 g male Sprague-Dawley rats (Charles River) were implanted with the same type of prosthesis as the rabbits in group A. While 6-8 prostheses were implanted in the back region of rabbits, four implants were inserted into similarly positioned subcutaneous pockets of rats.
Implant Procedure Innovar Vet (Pittman Moore) (0.2 m1/2 kg body wt) was administered i.m. 15-20 min prior to surgery. After closely shaving the backs of the rabbits, the area was divided into six (group A) or eight (group B) zones. A 4 mm wide incision was made through to the fascia and a subcutaneous pocket was formed using a Kelly 5 in. curved hemostat. After inserting the grafts into the pocket, the wound was sutured with 000 proline suture (Ethicon) using a single stitch. Each rabbit receivcd randomly positioned tubes treated as described above.
Rate of Collagen-Fabric Tube Resorption This was studied in rabbits or rats immobilized by Innovar-Vet, twice weekly, using a Dynamax Model C-792A x-ray machine (Machlett Labs) with a Westinghouse Alodcl 3400 control panel. Medical x-ray, 14 x 17 type R film (3-M) was used. Diffuse opacity a t implanted tube site was considered evidence of leakage. Still, leakage was recognized only when it was present in x-ray reading of two consecutive time periods. X-Rays taken immediately after surgery were used as controls for comparison.
Histology Representative samplcs were evaluated by standard histological techniques using hematoxylin-eosin and Masson Trichrome stains.
RESORPTION RATE OF COLLAGEN TUBING
301
RESULTS Variability of the Measurements of the Rate of Collagen Tubes Resorption Although the x-ray method used in this study to observe the in vivo resorption of implantcd collagen tubes was first described by Chvapil and Rrajicek in 1963, the error of the method has never been determined. This was now accomplished by implanting eight identical collagen tubings into each of three rabbits (Table I ) . We found that the variation coefficient of the resorption rate of tubings in each rabbit is close to 10% of the mean. Interindividual variability based on resorption time of 50% of implanted identical tubings into three rabbits was 5.670.
Effect of Tanning A definite relation between tanning time of the collagen tubes and their rate of resorption by the subcutaneous tissue was found in both groups of rabbits (A and B). A typical flow sheet of the results is presented in Table I1 indicating the percent of resorbed prostheses in time as a function of the tanning in group A of rabbits. Similar summations of results in group B rabbits are presented in Table 111, which shows also the effects of other treatments, such as sterilization and oxidation. Plotting the time when the leak occurred in days against the tanning time of the collagen tubing as derived from group B experiments, TABLE I Variability of the 1)etermiriation of the Rate of Resorptioii of Implanted Collagen Tubes in Rabbits
Rabbita
Days of leakage of 50% implanted tubes
1 2 3
37.0 4.9 41.5 & 5 . 0 42.4 f 5 . 1
*
Variation coefficientIi 13.2 12.0 12.1
a Eight identical sponges implanted iri each rabbit; x-rays made in 4 day intervals. (X & SD.) Gives SD in percent of the mean value.
CHVAPIL ET AL.
302
TABLE I1 Resorption Studies Showing Percent of Leakage of Collagen-Dacron Implants in Rabbits (Subcutaneous Tissue)s
Days
Tanning Time
4 8 12 16 20 24
14
21
28
35
42
49
56
66
62 62 50 12 0 12
100 62
100 100 100 62 50 50
100 100 100
100 100 100 87 75 75
100 100 100 100 75 75
100 100 100 100 87 75
100 100 100 100 100 87
75 12
0 12
75 62 75
* Sixteen rabbits (male 1.5-2 kg) were implanted with six sponges each. Leakage time ascertained by x-ray. Data is presented as percent of leaking implants from total number of implants for each tanning group.
a progressively slower resorption (progressive leaks) of more crosslinked collagen tubes is shown in Figure 1. As evident in Tables I1 and I11 and in Figure 1, the variability of the results did not exceed 15y0of the mean value, when 8-10 animals were used to test a single variable. Collagen I sorption
T
20 30 40 55
20 30 4 0 55 Tanning Time (min)
Fig. 1. Effect of tanning on the rate of resorption of collagen-fabric tubings. Composite collagen tube-fabric prostheses were implanted in the subcutis of rabbits. Resorption is reported as days after implantation when the first progressive leak of radio-opaque material from the tube was observed i n the x-ray film. Variability given as mean f SEX.
RESORPTION RATE OF COLLAGEN TUBING
303
TABLE I11 Statistically Averaged Peak Resorption (in Days) Graft #
Group
Mean f S.E.
~
Radiation-Sterilized Nonoxidized 20" 30 40 55
21.8 25.3 29.6 34.5
i 2.33 f 3.77 f 3.95 f 5.57
Oxidized 20 30 40 55
21.9 30.7 37.3 43.6
f 2.72 f 4.90 f 4.39 f 8.11
26.1 27.5 35.5 38.8
f f f f
28.5 33.6 34.6 44.6
f 4.52 f 6.23 f 4.88 f 3.07
Propylene Oxide-Sterilized Nonoxidized 20 30 40 55 Oxidized 20 30 40 55
5 6
7 8
(8)
(8) (8) (8)
4.42 3.05 3.11 5.09
Tanning time (in min). Graft # 16 showed only 75% resorption after 77 days. Therefore, only the resorbed grafts were used in calculations. 0 Number grafts used in calculations. a
b
Effects of Sterilization and Oxidation Although evidence was presented that ' W o irradiation may under certain circumstances result in decreasing the structural stability of collagen21 (higher dose, presence of humidity, production of free radicals), we did not observe any statistically significant difference between sterilization of collagen tubes with 6oCo irradiation (1.5 Mrad) and propylene oxide (Table 111). This conclusion obviously reflects a large variability of the results. Combined data including both oxidized and nonoxidized implants, tanned for varying times
CHVAPIL ET AL.
304
TABLE IV Effect of Two Methods of Sterilization on the Resorption Rate of Collagen-Dacron Composite Tubes ~~
Resorption Rate (in days) Sterilization Method Irradiation (@WO, 1.5 Mrad) Propylene oxide
20 mina
+
21.9 1.8;bp 28.0 f 2.1
30 minc
< 0.05
28.0 f 2 . 3 ; p 36.1 f 3 . 2
< 0.05
* Refers to tanning time in min. b Variability is given by X + SEM.
with glutaraldehyde (Table IV) , showed significantly slower resorption of collagen tubes tanned for 20 or 30 min and sterilized with propylene oxide. I n pooled data for tubes tanned for 40 or 55 min, no effect of sterilization method was noticed.
Effect of Species A striking difference in the rate of resorption of similarly treated collagen fabric composite tubes was found between rats and rabbits, as indicated in Figure 2. Only data on collagen tubes tanned for 4 and 8 min are shown. Collagen tubes tanned for 4 min and implanted into rabbits were rcsorbed at a much faster rate than those in rats. About 50% of the implants were resorbed at 13 days in rabbits, in rats at approximately 21 days. The difference of resorption rates in rats and rabbits was highly significant ( p < 0.001) and held for samples tanned for cithcr timc pcriod.
Morphology of the Dynamics of the Resorption For the sake of simplicity, only histological results from A series of rabbit experiments are presented. Similar conclusions were obtained in B series, however. It should be kept in mind that the implanted material consisted of collagen tubing coated with multifilamentous polyester fabric mesh. It has been well established that various fabric materials induce differing tissue reactivity and cell ingrowth.2 After subcutancous implantation, progressive infiltration of the fabric with inflammatory cells consisting mainly of
RESORPTION RATE OF COLLAGEN TUBING
305
late of Resorption of Collagen in Rabbits and Rats
Days after Implantation
Fig. 2. Species difference in the rate of collagen tube resorption. Identical composite collagen tube-fabric prostheses, tanned for 4 and 8 min, were implanted in the subcutis of rabbits and rats. Rate of resorption is given in percent of progressively leaking implants of a total number of tubes implanted. Variability is presented as mean f SEM.
granulocytes and macrophages was seen in the granuloma tissue pcrmeating the fabric multifilamentous meshwork (Fig. 3). With longer time, the spaces between fibers (f) filled with more voluminous granuloma tissue containing mostly macrophages and several fusiform-shaped cells (Fig. 4). In the next stage, depending on the degree of tanning of the collagen tube, a beginning loosening of the continuity of collagen membrane (c) was seen, mainly that portion in contact with the interfibrillar spaced rich in granuloma tissue with macrophages (Fig. 5 ) . I n some loci (see arrow in Fig. S), cells migrated inside the collagen membrane which showed noncontinuity in structure and characteristic change in collagen stainability by Trichrome stain (Fig. 7) from typical blue to red; this
306
CHVAPIL ET AL.
Fig. 3. Composite collagen tube-fabric implant in subcutis of rabbit, 10 days after implantation. (f : polyester fibers; c: collagen membrane.) Inflammatory cells of the reactive granuloma tissue are accumulating mainly in between fabric fibers; collagen membrane intact. 40X.
may indicate the denaturation of collagen, as proposed by Craik and M ~ N e i l . ~ The corrosion of the collagen tube membrane was definitely not uniform over all the tubing. Ultimately the collagen membrane was completely permeated by cells, mainly macrophages, to such a n extent that Lipidol leaked out of the tube (Fig. 8). I n highly tanned collagen tubings, the described degradation of the collagen wall was negligible for periods of 2 months or even longer. In these situations, there was a continuous fibrous layer of condensed collagen fibers formed around the implanted collagen fabric tubing.
RESORPTION RATE OF COLLAGEN TUBING
Fig. 4.
307
Proliferation of inflammatory cells with beginning destruction of outer layer of collagen membrane, 14 days after implantation. 1OOX.
DISCUSSION The rate a t which collagen products are resorbed in vivo depends on several factors, the degree of tanning and the site of implantation*z6J5being the most important. Although tissue collagenases and possibly also cathepsins (B, D) are responsible for the degradation of collagen implant~,’~~17*18 it is mainly the macrophage which has been implicated in the resorption o f collagen a t the cellular l e ~ e l . ~ There ~ ~ ” is enough evidence that macrophage infiltration occurs around implanted collagen ~ u t u r e s ,sponge^,^.^ ~ and vessel prostheses.’OJl Vascular grafts are gradually invaded by inflammatory cells and fibroblasts and replaced by new collagen and
308
CHVAPIL ET AL.
Fig. 5 . Swollen collagen membrane penetrated by inflammatory cells. The compact structure of collagen membrane is loosened. 100X .
connective tissue structures. It has been our personal experience that the entire graft is gradually replaced, regardless of the degree of t a n n i ~ ~ g .However, ~.~ some authors have found graft collagen preserved and r e c e l l ~ l a r i z e d . ~ ~Inflammatory -~~ cells such as granulocytes or macrophages contain collagenase as well as other potent proteases.lg It was also shown that the rate of collagen degradation by collagenase is inversely related to the degree of crosslinking.20 The results of our study, in agreement with the above findings, show that thc rate of the resorption of the implanted collagen material depends on the degree of tanning and crosslinking of collagen structures. Of interest is the species diffcrcnce in the resorption kinetics,
RESORPTION RATE OF COLLAGEN TUBING
309
Fig. 6. Maximum penetration of inflammatory cells with destruction of collagen membrane occurs between fabric fibers. Although collagen membrane is almost completely invaded by mostly macrophages, this membrane still did not leak a t 27 days after implantation. Arrow points to the site of cellular infiltration of collagen membrane. 160X .
for which we could not find any supporting evidence in the available literature. We believe that the ((species difference” does not reflect the difference in age, as both rats and rabbits could be classified as young adults. I n the wound healing problems, it is well known that some species such as man and horse heal-synthesize as well as degrade collagenmuch faster than others (rat, mouse). Our finding of faster degradation of implanted collagen in rabbits than in rats may just reflect higher turnover of collagen in this species.
310
CHVAPIL ET AL.
Fig. 7. Polyester polyfilamentous fibers are surrounded by inflammatory cells with maximum penetration into the collagen membrane between the fibers. On the outer side of the fabric, a continuous sheet of collagen fiber is formed (28 days after implantation). 30X.
Sterilization of collagen related biomaterial by 6oCoirradiation or by propylene oxide is commonly used. Still, some modifications of collagen biomaterial by either method has been indicated.lJ1 I n the case of 6oCosterilization, we observed faster resorption in two groups of collagen tubes, indicating a slight denaturation of the collagen material during sterilization even a t a relatively low dose of 1.5 Mrad. Treatment of collagen with periodate did not affect the resorption characteristics of implanted tubes. Although we described the x-ray method of following the rate of collagen resorption more than 10 years & ~ O , ~ J O we have been using
RESORPTION RATE OF COLLAGEN TUBING
311
Fig. 8. Complete penetration and destruction of collagen membrane by inflammatory cells which invade the membrane in streaks, leaving isolated islands of collagen structures still intact. Originally continuous and compact membrane is disintegrated and leaks. lOOx.
this technique almost on a routine basis since and accumulated new experience which could be summed up as follows. I n rats, a maximum of 4 tubes and, in rabbits, 6-8 tubes can be implanted and evenly located in the subcutis of the dorsum of the animal (Figs. 9a and 9b). Immediately after implantation of collagen tubes filled with x-ray opaque material, the boundaries of tubes should be sharp. Tubes showing a leak through the suture line should be eliminated from final statistics. We also eliminated those leaks that occurred within the first 10 days after implantation because maximum cumulative resorption occurred at two to three times longer that the usual time
312
CHVAPIL ET AL.
Fig. 9. Typical x-ray taken immediately after subcutaneous implantation of collagen-fabric composite material into rats (top) or rabbits (bottom). Note the sharp delineation of implants.
period. We considered these leaks a reflection of mechanical rupture of the wall of the collagen tube, possibly caused during manipulation of the sample before or during implantation. Only progressive leakage from the same tube in two consecutive periods were considered indicative of resorption (Fig. 10). We consider this method satisfactory when discrimination of sensitivity a t 5-7 days would be
RESORPTION RATE OF COLLAGEN TUBING
313
Fig. 10. Various stages of the resorption of collagen tubes in x-ray examination are shown by pictures A to E . Picture F refers to nonspecific leak of opaque material through suture.
requested. Still, resorption time of the same sample should be tested a t least in 8-10 implantation sites. This research was supported by funds from Edwards Laboratories, Santa Ana, California.
References 1.
-U.Chvapil, R. L. Kronenthal, and W. van Winkle, Jr., International Review of Connectzve Tissue Research 6, Academic Press, New York, 1973, p. 1.
2. M. Chvapil and M. Krajicek, Lagenbecks Arch. Klin. Chir., 318, 80 (1967). 3. 31. Chvapil and M. Krajicek, J . Surg. Res., 11, 358 (1963). 4. M.Krajicek, V. Zastava, and RI. Chvapil, J . Surg. Res., 4, 290 (1964). 5. J. E. Craik arid I. R. R. McNeil, Nature, 209, 931 (1966). 6. H. Ungar, Amer. J. Pathol., 29, 973 (1953). 7. T . N . Salthouse, J. A. Williams, and D. A. Willigan, Surgery, 129, 690 (1969). 8. &I.Chvapil, R. Holusa, K. Kliment, and M. Stoll, J . Bzomed. Muter. Res., 3, 315 (1969).
314
CHVAPIL ET AL.
9. D. E. Cutright, W. R. Posey, S. N. Bhaskar, and W. J. Larson, Oral Surg. Oral Med. Oral Pathol., 35, 144 (1973). 10. J. Hejnal, M. Krajicek, and B. Zastava, Vessel Substitutes, Academia Czechoslovakia, 1967. 11. M. Chvapil and M. Krajicek, Vascular Surgery, Butterworths, London, 1970, p. 120. 12. A. D. Potenza, J. Bone Joint Surg., 44,49 (1962). 13. A. D. Potenza, J. Bone Joint Surg., 46, 1462 (1964). 14. L. Klein, Ph.D. Thesis, Boston University, Boston, Mass., 1958. 15. M. J. White, I. Kohno, A. L. Itubin, and K. H. Stenzel, Biomat. Med. Dev., Art. Org., 1(4), 703 (1973). 16. A. Z. Eisen, E. A. Bauer, and J. J. Jeffrey, J. Invest. Dermatol., 55,359 (1970). 17. G. S. Lazarus and J. F. Goggins, Science, 186, 653 (1974). 18. P. B. Robertson, It. B. Ityel, It. E. Taylor, K. W. Shyr, and H. M. Fullmer, Science, 177, 64 (1972). 19. L. M. Wahl, S. M. Wahl, S. E. Mergenhagen, and G. R. Martin, Science, 187, 261 (1975). 20. M. E. Grant and 1).J. Prockop, New Engl. J. Med., 55, 242 (1972). 21. A. J. Bailey, Znt. Rev. Connect. Tissue Rea., 4, 268 (1968).
Received December 8, 1975 Revised July 15, 1976
![[Corneal collagen crosslinking in keratoconus].](/assets/img/hold.png)
