Journal of Orthopaedic Research 911-19 Raven Press, Ltd., New York 0 1991 Orthopaedic Research Society
Demineralized Allogeneic Bone Matrix for Cartilage Repair Leif Dahlberg and "Andris Kreicbergs Department of Orthopedics, University Hospital, Lund, Sweden; and *Karolinska Hospital, Stockholm, Sweden
Summary: We tested the chondrogenic potential of demineralized allogeneic bone matrix (DABM) in the repair of osteochondral defects. In 42 adult rabbits, a 5-mm2 or 15-mm2defect was created bilaterally in the intercondylar groove of distal femur. DABM was inserted directly in 37 defects, whereas in 35 it was inserted after previous placement in muscle for 4, 16, or 19 days. Another 12 defects were left empty, serving as controls. Subgroups of animals were killed at 6, 12, 18, and 26 weeks. The distal femora were excised and prepared for histologic evaluation in hematoxylin-eosin and toluidine blue. Cartilage-like repair tissue was observed in the majority of defects. However, there was a great variability in the experimental groups without any clear relationship to type of DABM implant, defect size, or postoperative time. Even individual knees exhibited varying stages of cartilage differentiation. Overall, DABM placed in muscle for 19 days appeared to yield the best repair of the defects. The most consistent findings of the present study were bone formation in the marrow of distal femur and, notably, the absence of bone differentiation toward the joint surface. Hence, it seems that the synovial environment prevents bone formation otherwise induced by DABM in vascular tissue. Although tissue formed in articular defects supplemented with DABM is of cartilaginous differentiation, which is retained over time, it is of highly variable quality. Hence, the described approach has to be optimized before it can be applied for the purpose suggested. Key Words: Joint cartilage-Regeneration-Bone induction-Bone matrix.
The low regenerative capability of articular cartilage has generated extensive research on various procedures for either enhancing cartilage formation in situ or replacing pathologic tissue by autogeneic transplantation (1,3-12). The use of chondrogenic tissue, such as perichondrium or periosteum, seems to be the most promising method (7,15,17,18,20). Principally? these procedures entail a combination of transplantation and stimulation of chondrogene-
sis. Postoperative continuous passive motion has been claimed to improve the results (14?18,19). So far, however, no method of cartilage repair has gained wide acceptance in clinical practice. This is probably because the experimental and clinical results are highly variable in spite of strictly standardized procedures (2,5,7,9,11,12). The reason for the variable biologic response remains obscure. It is well established that the bone induction principle introduced by Urist and further developed by Reddi can be used to mimic normal enchondral ossification (16,22). The phenomenon can be elicited by placing demineralized allogeneic bone matrix (DABM) both orthotopically and heterotopically . The formation of an ossicle with bone marrow tis-
Received December 19, 1989; accepted July 2, 1990. Address correspondence and reprint requests to Leif Dahlberg at Department of Orthopedics, University Hospital, S-221 85 Lund. Sweden.
11
L . DAHLBERG AND A . KREICBERGS
12
sue after approximately 3 weeks is preceded by a chondral phase (16,22,23,24,25). Few attempts have been made to use the chondrogenic potential of implanted bone matrix. So far it remains unknown whether osteogenesis can be prevented during the preceding chondral phase. If so, it has not been established whether the chondral tissue can develop to articular cartilage and remain in a synovial environment. We created deep defects in the articular cartilage of rabbit knees and transplanted DABM to fill the defects from underlying cancellous bone to the joint surface. Two fundamental issues were addressed: (a) Would the avascular synovial environment prevent bone formation? (b) Would the resultant tissue exhibit a hyaline cartilage differentiation? MATERIAL AND METHODS
The study included 42 adult white rabbits (3.1-4.3 kg). In each animal an articular defect was made in both knees. Surgery was performed under anesthesia with an i.m. injection containing a mixture of fluanizonium (10 mg/ml) and fentanyl citrate (0.315 mg/ml) (Hypnorm, Leo, Helsingborg, Sweden) 0.3 mg/kg combined with 1.5 mg/kg diazepam (5 mglml) (Valium, Roche, Basel, Switzerland). Arthrotomy was done by a medial parapatellar incision, and the patella was dislocated laterally. A defect of the articular cartilage (5 mm2 or 15 mm2) was created in the intercondylar groove of distal femur deep (4 mm) into bleeding cancellous bone. Large defects were made by drilling two adjacent holes (slightly overlapping) with a hand drill (diameter 3.2 mm). A chisel was used to smooth the edges and to obtain a rectangular form of the defect (Fig. 1). The small defects were made circular with a hand drill (diameter 2.5 mm). In 72 knees the defect was filled with DABM, whereas in 12, serving as controls, the defect was left empty. DABM was prepared from rabbit long bones according to the method of Urist. After removal of the epiphyses and marrow tissue, the specimens were cut into pieces, approximately 1 x 10 mm, demineralized in 0.6 M HC1 for 24 h at 4"C, rinsed three times in water, defatted in ch1oroform:methanol (1: 1) for 2 h and rinsed three more times in water. Finally, the preparations were lyophilized to constant weight, apart from a few specimens used as fresh transplants, The experimental series was divided into one group of 37 articular defects treated by direct
J Orthop Res, Vol. 9, No. 1 , 1991
FIG.1. Representative photograph showing size and form of large (15 mm') defects. Left defect filled with DABM.
DABM transplantation and another 35 treated with DABM transplants previously placed in muscle pouches of the abdominal wall. DABM transplanted directly was fresh, hydrated in water for 2 h, or lyophilized. DABM placed in muscle pouches was transplanted to the defects after 4, 16, or 19 days. The two main groups were further divided according to defect size and time until killing. Tables 1, 2, and 3 summarize the experimental groups. To assess the fate of DABM implants in muscle pouches after various times, control specimens were harvested after 4, 7, 13, 16, and 19 days of implantation for histologic analysis. To check the bone inductive capability of DABM, pieces of DABM were placed in the abdominal muscle wall for varying times and then analyzed macroscopically and histologically. Postoperatively, the animals were housed in cages and allowed to move freely. At 6, 12, 18, and
13
JOINT CARTILAGE REPAIR TABLE 3. Controls
TABLE 1. DABM transplanted directly
Defect size
Defect size 15 mm2
5 mm2
Type of DABM
n
weeks
n
Lyophilized
4
6
FresWhydrated
6
6
10" 2 2 8' 5
~
~
_
_
_
5 mm2
weeks
weeks
n
weeks
6 12 18 6 26
5
6
4 3"
6 26
'
26 weeks subgroups of the animals were killed with an overdose of i.v. pentobarbital sodium (60 mg/ml, ACO, Stockholm, Sweden) (Tables 1,2,3). In each animal the distal femur was excised bilaterally, and the whole defect site with surrounding normal tissue was chiseled en bloc from the intercondylar groove. The specimens were fixed in 10% buffered neutral formalin and subsequently demineralized in 44% formic acid and 20% sodium citrate 1:l for an average of 3 weeks. After paraffin embedding, 8-Fm sections were cut and stained in hematoxylin-eosin and toluidine blue, the latter stain serving as a marker of cartilaginous matrix. RESULTS
Eight of 84 knees had to be excluded because of infection in one case, patellar dislocation in three cases, and failure at specimen preparation in four cases (Tables 1,2,3). This left, for final evaluation, 33 knees with DABM directly transplanted and 32 knees with DABM placed in muscle before transplantation, apart from 11 sham-operated knees. In control specimens of DABM left in muscle pouches of the abdominal wall for 6 weeks, ossicles with red bone marrow consistently developed. In control specimens placed in muscle pouches for 4 7 , 13, 16, TABLE 2. DABM previously placed in muscle Defect size 5 mm2
15 mm2
Days in muscle
n
weeks
n
weeks
4
2
6
16 19
3 2
6 6
10" 2 2 4" 8" 2
6 12 18 6 6 26
a
" One rabbit was excluded.
_ _____ _
Three rabbits were excluded. One rabbit was excluded.
a
15 mm2
n
One rabbit was excluded.
and 19 days toluidine blue-positive cells and matrix emerged on day 13. This mesenchymal cartilage tissue became more differentiated and abundant with time and remained at least until day 19. DABM Transplanted Directly
Macroscopically all defects appeared well healed. They exhibited a fairly even surface, although it had a mosaic-like pattern. At palpation the repair tissue seemed softer than the surrounding articular cartilage. Microscopically, the repair tissue was highly variable with respect to differentiation and distribution. The variability was not related to type of DABM, defect size, or time after surgery. In some knees there was a clear cartilage differentiation at the defect site, whereas in others the tissue was more of the fibrocartilage type (Fig. 2). According to toluidine blue, the tissue close to the DABM pieces was more cartilage-like than the regenerative tissue at some distance from the DABM transplants (Fig. 3). In four of the knees equally distributed in the subgroups, even in those exhibiting cartilage repair tissue of seemingly high differentiation, there was, nevertheless, a central defect of varying size. Some also had narrow crevices at the junction between new and original articular cartilage. Surprisingly, no clear difference could be detected between knees with 5-mm2 and those with 15-mm2 defects. On the whole, there was a considerable variability of the repair tissue not only within each of the experimental subgroups, but sometimes also within the same knee, according to analysis of serial sections. A predominant feature in this experimental group was the occurrence of DABM remnants in the defects. In 14 of 15 knees repaired with freeze-dried DABM, there were still visible remnants of DABM, even as long as 12 and 18 weeks after transplantation. Notably, only six of 13 knees repaired with fresh or hydrated DABM had such remnants after
J Orthop Res, Vol. 9, No. 1 , 1991
14
L . DAHLBERG AND A . KREICBERGS
FIG. 2. Mixture of cartilage and fibrocartilage in a 15-mm2 defect, 6 weeks after implantation, supplemented with fresh DABM. No remnants of DABM. Partial cluster formation. Completely healed junctions. (H&E, x4)
6 weeks; after 26 weeks one of 5 had remnants (Fig. 4). The only consistent feature in all subgroups was the complete absence of osteogenic differentiation in the defects toward the joint surface. In contrast, deeper in the subchondral bone the DABM pieces were consistently found to have induced new bone formation (Fig. 3). Transplantation of DABM Previously Placed in Muscle Macroscopically, all 32 defects appeared healed with an even, rather firm surface. The mosaic pat-
FIG. 3. Large remnant of lyophilized DABM 6 weeks after implantation in a 15-mm2 defect. Heavily toluidine blue-stained cells adjacent to implant. (Toluidine blue, x4)
J Orthop Res, Vol. 9, No. 1 , 1991
tern seemed less pronounced the longer the DABM transplant had been in the muscle pouch. Microscopically, the results of cartilage repair with DABM transplanted after placement in muscle appeared better than with DABM transplanted directly. Nevertheless, the repair tissue both within individual subgroups and within individual knees was as heterogeneous as that observed in knees with DABM transplanted directly. Thus, the regenerative tissue in the articular defects exhibited a differentiation ranging from fibrous to cartilaginous regardless of defect size or post-transplantation time (Figs. 5,6,7). A central defect occurred in the repair tissue of four 15-mm2defects (Fig. 7). There
JOINT CARTILAGE REPAIR
15
FIG. 4. A 15-mm2defect 26 weeks after implantation of hydrated DABM. Partly healed tidemark. Cluster formation in the middle. No fibrillations or crevices. (H&E, x4)
was, however, a difference between the different subgroups according to time in muscle before transplantation. Thus, the most homogeneous tissue with respect to cartilage differentiation was noted in the group of knees receiving DABM after 19 days in muscle. Even a tidemark had formed in one of these knees (Fig. 8). The lowest differentiation, mostly of the fibrous type, was observed in the 16-day group but surprisingly not in the 4-day group. Compared with the previous experimental group, the limited occurrence of DABM remnants was the most striking difference. This was observed in 10 of 15 knees in the 4-day group, three of 6 in the 16-day group, and only two of 9 in the 19-day group. After 12, 18, and 26 weeks, the DABM pieces, commonly
still present after 6 weeks, had almost disappeared (Fig. 9). In the experimental group as a whole the only consistent findings were the absence of bone differentiation toward the joint surface and new bone formation in the deeper part of the defects, i.e., in the marrow, as noted in the previous experimental group. CONTROLS Surprisingly, the 5-mm2 and 15-mm2 defects left empty for 6 weeks occasionally exhibited cartilagelike tissue similar to that seen in the two main experimental groups (Fig. 10). However, the repair
FIG. 5. Large piece of DABM still present 6 weeks after implantation in a 15 mm defect. DABM previously placed in muscle pouch for 4 days. Abundant fibrous tissue and poor junctional healing. (H&E, x4)
J Orthop Res, Vol. 9, No. I , 1991
L. DAHLBERG AND A . KREICBERGS
16
FIG. 6. Same knee as Fig. 5. Fewer DABM rem-
nants and more cartilage-like tissue. Good junctional healing. (Toluidine blue, x4)
tissue was mostly of inferior quality, i.e., fibrous (Fig. 11). A central defect was observed in three of the 15-mm2and two of the 5-mm2defects (Fig. 9). DISCUSSION The present study shows that allogeneic bone matrix has a chondrogenic capability, which is enhanced in the synovial environment. Although the yield in joint defects is cartilage-like, it is mostly of insufficient amount and differentiation to restore the articular surface. To our knowledge this is the first study exploring the potential of allogeneic bone matrix for joint cartilage repair (21). The results show that the bone
FIG. 7. Variable repair tissue in a 15-mm' defect after 6 weeks. DABM previously placed in muscle pouch for 4 days. Cartilage-liketissue to the left, no crevices at the junctions. (H&E, x4)
J Orthop Res, Vol. 9, No. I, 1991
induction principle as described by Urist may be used in an avascular environment to produce cartilaginous tissue and also to retain this differentiation. However, the utility of the method appears to be disappointing. No matter which modifications were tried in optimizing the cartilage yield, the resultant tissue was highly variable and unpredictable. More important, no approach was found to give fully differentiated joint cartilage, as assessed histologically. Apart from deficiencies in maturity, the regenerative tissue was often insufficient quantitatively, as reflected by a remaining central defect. It is surprising that no clear difference could be detected between large and small defects. Moreover, crevices at the junction between new and old carti-
JOINT CARTILAGE REPAIR
17
FIG. 8. Cartilage like repair tissue in a 15-mm' defect after 6 weeks. DABM previously placed in muscle pouch for 19 days. Junctions healed. Tidemark is restored. (H&E, x4)
lage occurred without any clear relationship to defect size, type of DABM, or time after transplantation. The lack of objective criteria and the heterogeneity of the repair tissue made it very difficult to establish clear-cut differences between the various subgroups. The problem of reaching conclusions about the quality of the regenerative tissue also applied to individual knees. Unexpectedly, it even proved difficult to point to clear differences between some DABM- and sham-operated knees, since the latter occasionally exhibited a substantial cartilaginous regeneration (Fig. 10). Most investigators of cartilage repair seem to consider that defects larger than 6 mm2 rarely heal spontaneously and
therefore are appropriate as controls (2,12). The considerable variability of the resultant tissue within experimental groups treated identically, and also within individual knees, illustrate the extreme difficulties in evaluating experiments on cartilage regeneration. In our opinion, studies on joint cartilage repair should be questioned, unless the results are presented in full detail. Overall it appeared that DABM placed in muscle before transplantation generated more cartilaginous tissue in the defects than DABM transplanted directly. One major reason for this could be that remnants of the latter preparation at 6 weeks still occupied part of the defects. To some extent this might have prevented repair tissue from developing in the
FIG. 9. Healed 15-mm2 defect after 26 weeks. DABM previously placed in muscle pouch for 19 days. Note cluster formation and absence of fibrillation. (H&E, x4)
J Orthop Res, Vol. 9, No. 1, 1991
18
L . DAHLBERG AND A . KREICBERGS
FIG. 10. Mostly cartilage-like tissue in a control 15-mm2defect after 6 weeks, with healed junctions, although surface is slightly depressed. (x4)
defects. It seems less likely that there exists a difference in the chondrogenic capability between the two DABM preparations. Among the subgroups of knees receiving DABM preparations previously placed in muscle for various times, we have the impression that those placed in muscle for 19 days gave the best results. However, in the absence of quantitative data this cannot be substantiated. Although histologic analysis probably is adequate in assessing gross differences in quality of repair tissue, it cannot be used to detect discrete differences. Moreover, there is today no histology staining method that includes a marker of fully differentiated joint cartilage. The reason that DABM left 19 days in muscle probably is better
FIG. 11. Control sham-operated knee, where a 5-mm2 defect was created. Abundant fibrous tissue with a central defect 6 weeks postoperative. (X4)
J Orthop Res, Vol. 9, No. I, 1991
than other preparations remains obscure. It might be explained by optimum recruitment of chondrogenic cells (number and differentiation) before transplantation. Analysis of control specimens placed in muscle for different times seemed to support this assumption. From experiments on heterotopic bone formation after transplantation of DABM from one tissue to another, Urist reported that optimum induction occurred if the transplantation was done early or late; DABM transplantation days 6-12 almost failed to yield an ossicle (23). The results of the present study seem to agree with the observations made by Urist. Thus, the 4-day and particularly the 19-day transplants produced better cartilage than the 16-
JOINT CARTILAGE REPAIR
day transplants. Presumably, 4 days in muscle was too short a time to recruit any major number of mesenchymal stem cells for transplantation to joint defects; the cells recruited for the observed cartilage regeneration probably originated from the surrounding subchondral bone or marrow. Transplants placed in muscle for 16 days may be assumed to have recruited chondro- and osteocompetent cells, but hypothetically the cells did not survive the transplantation, i.e., excision from muscle and subsequent insertion into joint cartilage defects. In contrast, the 19-day transplants appeared to contain chondrogenic cells capable of surviving transplantation. Interestingly, it proved impossible to dissect DABM bluntly from muscle after 19 days. Most surprising, there were no signs of newly formed bone adjacent to the 19-day transplants prior to insertion in the defects, although heterotopic ossicle formation is reported to occur at approximately 3 weeks (24). Thus, in our experimental model, DABM left 19 days in muscle seemed to represent the optimum for cartilaginous yield. The most important aspect of the present study is not the repair procedure per se, but rather the biologic response elicited. Thus, from our experiment, it is obvious that DABM placed in an avascular synovial environment does not induce bone formation, as observed in marrow tissue and subchondral bone. Instead, the tissue formed in articular defects supplemented with DABM is of cartilaginous differentiation, which, moreover, is retained over time. Unfortunately, the quality of the repair tissue is highly variable and, in general, too low to restore joint surface integrity. Although it seems that DABM has a chondrogenic potential, which may be used for cartilage repair, the approach has to be optimized by some modification before it can be used for the purpose suggested. Acknowledgment: This study was supported by research funds from the Karolinska Institute. The authors thank Dick HeinegHd and Ove Enquist for advice, and Folke Carell for technical assistance.
REFERENCES 1. Ashton JA, Bentley G: Repair of articular surfaces by allografts of articular and growth-plate cartilage. J Bone Joint Surg 68B:29-35, 1986 2. Buckwalter J, Rosenberg L, Coutts R, Hunziker E, Reddi AH, Mow V: Articular cartilage: injury and Repair. In: Injury and Repair of the Musculoskeletal Soft Tissue, ed by
19
SL-Y Woo, J Buckwalter, Park Ridge, IL, American Academy of Orthopaedic Surgeons, 1987, pp 465482 3. Calandruccio RA, Gilmer WS Jr: Proliferation, regeneration and repair of articular cartilage of immature animals. J Bone Joint Surg 44A:431455, 1962 4. Campell CJ: The healing of cartilage defects. Clin Orthop 64:4543, 1969 5. Chesterman PJ, Smith U: Homotransplantation of articular cartilage and isolated chondrocytes. J Bone Joint Surg 50B: 184-197, 1968 6. Convery FR, Akeson WH, Keown GH: The repair of large osteochondral defects. Clin Orthop 82:253-262, 1972 7. Enquist 0, Johansson SH: Perichondrial arthroplasty: a clinical study in twenty-six patients. Scand J Plast Reconstr Surg 14:71-87, 1980 8. DePalma AF, McKeever CD, Subin DK: Process of repair of articular cartilage demonstrated by histoloy and autoradiography with triated thymidine. Clin Orthop 48:229-242, 1966 9. Grande DA, Pitman MI, Peterson L, Menche D, Klein M: The repair of experimentally produced defects in rabbit articular cartilage by autologous chondrocyte transplantation. J Orthop Res 7:208-218, 1989 10. Itay S, Abramovic A, Nevo Z: Use of cultured embryonal chick epiphyseal chondrocytes as grafts for defects in chicken articular cartilage. Clin Orthop 220286303, 1987 11. Key JA: Experimental arthritis: the changes in joints produced by creating defects in the articular cartilage. J Bone Joint Surg 13:725-739, 1931. 12. Mankin HJ: The response of articular cartilage to mechanical injury. J Bone Joint Surg 64A:460466, 1986 13. Mitchell N, Shepard N: The resurfacing of rabbit articular cartilage by multiple perforations through the subchondral bone. J Bone Joint Surg 58A:23&233, 1976 14. O’Driscoll B, Salter RB: The induction of neochondrogenesis in free intra-articular periosteal autografts under the influence of continuous passive motion. J Bone Joint Surg 66A:1248-1257, 1984 15. Ohlsen L, Widenfalk B: The early development of articular cartilage after perichondrial grafting. Scand J Plnst Reconstr Surg 17:163-177, 1983 16. Reddi AH: Matrix induced enchondral bone differentiation. In:Mechanics of Growth Control, ed by RO Becker, Illinois, Thomas Springfield, 1981, pp 4 3 5 4 6 17. Rubak JM: Reconstruction of articular cartilage defects with free periosteal grafts. Acta Orthop Scnnd 53:175-180, 1982 18. Rubak JM, Poussa M, Ritsila V: Effects of joint motion on the repair of articular cartilage with free periosteal grafts. Acta Orthop Scand 53:187-191, 1982 19. Salter RB, Simmonds DF, Malcolm BW, Rumble EJ, MacMichael D, Clements ND: The biological effect of continuous passive motion on the healing of full-thickness defects in articular cartilage. J Bone Joint Surg 62A:1232-1251, 1980 20. Skoog T, Ohlsen L, Sohn SA: Perichondrial potential for cartilagenous regeneration. Scand J Plast Reconstr Surg 6~123-125,1972 21. Syftestad G, Caplan A: A 31000 dalton bone matrix protein stimulates chondrogenesis in chick limb bud cell cultures. Trans Orthop Res SOC 11:278, 1986 22. Urist MR: Bone: formation by autoinduction. Science 150: 893-899, 1965 23. Urist MR, Hay H, Dubac F, Buring K: Osteogenetic competence. Clin Orthop 64:19&200, 1969 24. Urist MR, Silverman BF, Buring K, Dubac F, Rosenberg JM: The bone induction principle. Clin Orthop 53:243-283, 1967 25. Urist MR: The origin of articular cartilage: investigation in quest of chondrogenic DNA. In: Cartilage, vol. 2, ed by K Hall, New York, Academic Press, 1983, pp 1-85
J Orthop Res, Vol. 9, No. 1 , 1991
Demineralized Allogeneic Bone Matrix for Cartilage Repair Leif Dahlberg and "Andris Kreicbergs Department of Orthopedics, University Hospital, Lund, Sweden; and *Karolinska Hospital, Stockholm, Sweden
Summary: We tested the chondrogenic potential of demineralized allogeneic bone matrix (DABM) in the repair of osteochondral defects. In 42 adult rabbits, a 5-mm2 or 15-mm2defect was created bilaterally in the intercondylar groove of distal femur. DABM was inserted directly in 37 defects, whereas in 35 it was inserted after previous placement in muscle for 4, 16, or 19 days. Another 12 defects were left empty, serving as controls. Subgroups of animals were killed at 6, 12, 18, and 26 weeks. The distal femora were excised and prepared for histologic evaluation in hematoxylin-eosin and toluidine blue. Cartilage-like repair tissue was observed in the majority of defects. However, there was a great variability in the experimental groups without any clear relationship to type of DABM implant, defect size, or postoperative time. Even individual knees exhibited varying stages of cartilage differentiation. Overall, DABM placed in muscle for 19 days appeared to yield the best repair of the defects. The most consistent findings of the present study were bone formation in the marrow of distal femur and, notably, the absence of bone differentiation toward the joint surface. Hence, it seems that the synovial environment prevents bone formation otherwise induced by DABM in vascular tissue. Although tissue formed in articular defects supplemented with DABM is of cartilaginous differentiation, which is retained over time, it is of highly variable quality. Hence, the described approach has to be optimized before it can be applied for the purpose suggested. Key Words: Joint cartilage-Regeneration-Bone induction-Bone matrix.
The low regenerative capability of articular cartilage has generated extensive research on various procedures for either enhancing cartilage formation in situ or replacing pathologic tissue by autogeneic transplantation (1,3-12). The use of chondrogenic tissue, such as perichondrium or periosteum, seems to be the most promising method (7,15,17,18,20). Principally? these procedures entail a combination of transplantation and stimulation of chondrogene-
sis. Postoperative continuous passive motion has been claimed to improve the results (14?18,19). So far, however, no method of cartilage repair has gained wide acceptance in clinical practice. This is probably because the experimental and clinical results are highly variable in spite of strictly standardized procedures (2,5,7,9,11,12). The reason for the variable biologic response remains obscure. It is well established that the bone induction principle introduced by Urist and further developed by Reddi can be used to mimic normal enchondral ossification (16,22). The phenomenon can be elicited by placing demineralized allogeneic bone matrix (DABM) both orthotopically and heterotopically . The formation of an ossicle with bone marrow tis-
Received December 19, 1989; accepted July 2, 1990. Address correspondence and reprint requests to Leif Dahlberg at Department of Orthopedics, University Hospital, S-221 85 Lund. Sweden.
11
L . DAHLBERG AND A . KREICBERGS
12
sue after approximately 3 weeks is preceded by a chondral phase (16,22,23,24,25). Few attempts have been made to use the chondrogenic potential of implanted bone matrix. So far it remains unknown whether osteogenesis can be prevented during the preceding chondral phase. If so, it has not been established whether the chondral tissue can develop to articular cartilage and remain in a synovial environment. We created deep defects in the articular cartilage of rabbit knees and transplanted DABM to fill the defects from underlying cancellous bone to the joint surface. Two fundamental issues were addressed: (a) Would the avascular synovial environment prevent bone formation? (b) Would the resultant tissue exhibit a hyaline cartilage differentiation? MATERIAL AND METHODS
The study included 42 adult white rabbits (3.1-4.3 kg). In each animal an articular defect was made in both knees. Surgery was performed under anesthesia with an i.m. injection containing a mixture of fluanizonium (10 mg/ml) and fentanyl citrate (0.315 mg/ml) (Hypnorm, Leo, Helsingborg, Sweden) 0.3 mg/kg combined with 1.5 mg/kg diazepam (5 mglml) (Valium, Roche, Basel, Switzerland). Arthrotomy was done by a medial parapatellar incision, and the patella was dislocated laterally. A defect of the articular cartilage (5 mm2 or 15 mm2) was created in the intercondylar groove of distal femur deep (4 mm) into bleeding cancellous bone. Large defects were made by drilling two adjacent holes (slightly overlapping) with a hand drill (diameter 3.2 mm). A chisel was used to smooth the edges and to obtain a rectangular form of the defect (Fig. 1). The small defects were made circular with a hand drill (diameter 2.5 mm). In 72 knees the defect was filled with DABM, whereas in 12, serving as controls, the defect was left empty. DABM was prepared from rabbit long bones according to the method of Urist. After removal of the epiphyses and marrow tissue, the specimens were cut into pieces, approximately 1 x 10 mm, demineralized in 0.6 M HC1 for 24 h at 4"C, rinsed three times in water, defatted in ch1oroform:methanol (1: 1) for 2 h and rinsed three more times in water. Finally, the preparations were lyophilized to constant weight, apart from a few specimens used as fresh transplants, The experimental series was divided into one group of 37 articular defects treated by direct
J Orthop Res, Vol. 9, No. 1 , 1991
FIG.1. Representative photograph showing size and form of large (15 mm') defects. Left defect filled with DABM.
DABM transplantation and another 35 treated with DABM transplants previously placed in muscle pouches of the abdominal wall. DABM transplanted directly was fresh, hydrated in water for 2 h, or lyophilized. DABM placed in muscle pouches was transplanted to the defects after 4, 16, or 19 days. The two main groups were further divided according to defect size and time until killing. Tables 1, 2, and 3 summarize the experimental groups. To assess the fate of DABM implants in muscle pouches after various times, control specimens were harvested after 4, 7, 13, 16, and 19 days of implantation for histologic analysis. To check the bone inductive capability of DABM, pieces of DABM were placed in the abdominal muscle wall for varying times and then analyzed macroscopically and histologically. Postoperatively, the animals were housed in cages and allowed to move freely. At 6, 12, 18, and
13
JOINT CARTILAGE REPAIR TABLE 3. Controls
TABLE 1. DABM transplanted directly
Defect size
Defect size 15 mm2
5 mm2
Type of DABM
n
weeks
n
Lyophilized
4
6
FresWhydrated
6
6
10" 2 2 8' 5
~
~
_
_
_
5 mm2
weeks
weeks
n
weeks
6 12 18 6 26
5
6
4 3"
6 26
'
26 weeks subgroups of the animals were killed with an overdose of i.v. pentobarbital sodium (60 mg/ml, ACO, Stockholm, Sweden) (Tables 1,2,3). In each animal the distal femur was excised bilaterally, and the whole defect site with surrounding normal tissue was chiseled en bloc from the intercondylar groove. The specimens were fixed in 10% buffered neutral formalin and subsequently demineralized in 44% formic acid and 20% sodium citrate 1:l for an average of 3 weeks. After paraffin embedding, 8-Fm sections were cut and stained in hematoxylin-eosin and toluidine blue, the latter stain serving as a marker of cartilaginous matrix. RESULTS
Eight of 84 knees had to be excluded because of infection in one case, patellar dislocation in three cases, and failure at specimen preparation in four cases (Tables 1,2,3). This left, for final evaluation, 33 knees with DABM directly transplanted and 32 knees with DABM placed in muscle before transplantation, apart from 11 sham-operated knees. In control specimens of DABM left in muscle pouches of the abdominal wall for 6 weeks, ossicles with red bone marrow consistently developed. In control specimens placed in muscle pouches for 4 7 , 13, 16, TABLE 2. DABM previously placed in muscle Defect size 5 mm2
15 mm2
Days in muscle
n
weeks
n
weeks
4
2
6
16 19
3 2
6 6
10" 2 2 4" 8" 2
6 12 18 6 6 26
a
" One rabbit was excluded.
_ _____ _
Three rabbits were excluded. One rabbit was excluded.
a
15 mm2
n
One rabbit was excluded.
and 19 days toluidine blue-positive cells and matrix emerged on day 13. This mesenchymal cartilage tissue became more differentiated and abundant with time and remained at least until day 19. DABM Transplanted Directly
Macroscopically all defects appeared well healed. They exhibited a fairly even surface, although it had a mosaic-like pattern. At palpation the repair tissue seemed softer than the surrounding articular cartilage. Microscopically, the repair tissue was highly variable with respect to differentiation and distribution. The variability was not related to type of DABM, defect size, or time after surgery. In some knees there was a clear cartilage differentiation at the defect site, whereas in others the tissue was more of the fibrocartilage type (Fig. 2). According to toluidine blue, the tissue close to the DABM pieces was more cartilage-like than the regenerative tissue at some distance from the DABM transplants (Fig. 3). In four of the knees equally distributed in the subgroups, even in those exhibiting cartilage repair tissue of seemingly high differentiation, there was, nevertheless, a central defect of varying size. Some also had narrow crevices at the junction between new and original articular cartilage. Surprisingly, no clear difference could be detected between knees with 5-mm2 and those with 15-mm2 defects. On the whole, there was a considerable variability of the repair tissue not only within each of the experimental subgroups, but sometimes also within the same knee, according to analysis of serial sections. A predominant feature in this experimental group was the occurrence of DABM remnants in the defects. In 14 of 15 knees repaired with freeze-dried DABM, there were still visible remnants of DABM, even as long as 12 and 18 weeks after transplantation. Notably, only six of 13 knees repaired with fresh or hydrated DABM had such remnants after
J Orthop Res, Vol. 9, No. 1 , 1991
14
L . DAHLBERG AND A . KREICBERGS
FIG. 2. Mixture of cartilage and fibrocartilage in a 15-mm2 defect, 6 weeks after implantation, supplemented with fresh DABM. No remnants of DABM. Partial cluster formation. Completely healed junctions. (H&E, x4)
6 weeks; after 26 weeks one of 5 had remnants (Fig. 4). The only consistent feature in all subgroups was the complete absence of osteogenic differentiation in the defects toward the joint surface. In contrast, deeper in the subchondral bone the DABM pieces were consistently found to have induced new bone formation (Fig. 3). Transplantation of DABM Previously Placed in Muscle Macroscopically, all 32 defects appeared healed with an even, rather firm surface. The mosaic pat-
FIG. 3. Large remnant of lyophilized DABM 6 weeks after implantation in a 15-mm2 defect. Heavily toluidine blue-stained cells adjacent to implant. (Toluidine blue, x4)
J Orthop Res, Vol. 9, No. 1 , 1991
tern seemed less pronounced the longer the DABM transplant had been in the muscle pouch. Microscopically, the results of cartilage repair with DABM transplanted after placement in muscle appeared better than with DABM transplanted directly. Nevertheless, the repair tissue both within individual subgroups and within individual knees was as heterogeneous as that observed in knees with DABM transplanted directly. Thus, the regenerative tissue in the articular defects exhibited a differentiation ranging from fibrous to cartilaginous regardless of defect size or post-transplantation time (Figs. 5,6,7). A central defect occurred in the repair tissue of four 15-mm2defects (Fig. 7). There
JOINT CARTILAGE REPAIR
15
FIG. 4. A 15-mm2defect 26 weeks after implantation of hydrated DABM. Partly healed tidemark. Cluster formation in the middle. No fibrillations or crevices. (H&E, x4)
was, however, a difference between the different subgroups according to time in muscle before transplantation. Thus, the most homogeneous tissue with respect to cartilage differentiation was noted in the group of knees receiving DABM after 19 days in muscle. Even a tidemark had formed in one of these knees (Fig. 8). The lowest differentiation, mostly of the fibrous type, was observed in the 16-day group but surprisingly not in the 4-day group. Compared with the previous experimental group, the limited occurrence of DABM remnants was the most striking difference. This was observed in 10 of 15 knees in the 4-day group, three of 6 in the 16-day group, and only two of 9 in the 19-day group. After 12, 18, and 26 weeks, the DABM pieces, commonly
still present after 6 weeks, had almost disappeared (Fig. 9). In the experimental group as a whole the only consistent findings were the absence of bone differentiation toward the joint surface and new bone formation in the deeper part of the defects, i.e., in the marrow, as noted in the previous experimental group. CONTROLS Surprisingly, the 5-mm2 and 15-mm2 defects left empty for 6 weeks occasionally exhibited cartilagelike tissue similar to that seen in the two main experimental groups (Fig. 10). However, the repair
FIG. 5. Large piece of DABM still present 6 weeks after implantation in a 15 mm defect. DABM previously placed in muscle pouch for 4 days. Abundant fibrous tissue and poor junctional healing. (H&E, x4)
J Orthop Res, Vol. 9, No. I , 1991
L. DAHLBERG AND A . KREICBERGS
16
FIG. 6. Same knee as Fig. 5. Fewer DABM rem-
nants and more cartilage-like tissue. Good junctional healing. (Toluidine blue, x4)
tissue was mostly of inferior quality, i.e., fibrous (Fig. 11). A central defect was observed in three of the 15-mm2and two of the 5-mm2defects (Fig. 9). DISCUSSION The present study shows that allogeneic bone matrix has a chondrogenic capability, which is enhanced in the synovial environment. Although the yield in joint defects is cartilage-like, it is mostly of insufficient amount and differentiation to restore the articular surface. To our knowledge this is the first study exploring the potential of allogeneic bone matrix for joint cartilage repair (21). The results show that the bone
FIG. 7. Variable repair tissue in a 15-mm' defect after 6 weeks. DABM previously placed in muscle pouch for 4 days. Cartilage-liketissue to the left, no crevices at the junctions. (H&E, x4)
J Orthop Res, Vol. 9, No. I, 1991
induction principle as described by Urist may be used in an avascular environment to produce cartilaginous tissue and also to retain this differentiation. However, the utility of the method appears to be disappointing. No matter which modifications were tried in optimizing the cartilage yield, the resultant tissue was highly variable and unpredictable. More important, no approach was found to give fully differentiated joint cartilage, as assessed histologically. Apart from deficiencies in maturity, the regenerative tissue was often insufficient quantitatively, as reflected by a remaining central defect. It is surprising that no clear difference could be detected between large and small defects. Moreover, crevices at the junction between new and old carti-
JOINT CARTILAGE REPAIR
17
FIG. 8. Cartilage like repair tissue in a 15-mm' defect after 6 weeks. DABM previously placed in muscle pouch for 19 days. Junctions healed. Tidemark is restored. (H&E, x4)
lage occurred without any clear relationship to defect size, type of DABM, or time after transplantation. The lack of objective criteria and the heterogeneity of the repair tissue made it very difficult to establish clear-cut differences between the various subgroups. The problem of reaching conclusions about the quality of the regenerative tissue also applied to individual knees. Unexpectedly, it even proved difficult to point to clear differences between some DABM- and sham-operated knees, since the latter occasionally exhibited a substantial cartilaginous regeneration (Fig. 10). Most investigators of cartilage repair seem to consider that defects larger than 6 mm2 rarely heal spontaneously and
therefore are appropriate as controls (2,12). The considerable variability of the resultant tissue within experimental groups treated identically, and also within individual knees, illustrate the extreme difficulties in evaluating experiments on cartilage regeneration. In our opinion, studies on joint cartilage repair should be questioned, unless the results are presented in full detail. Overall it appeared that DABM placed in muscle before transplantation generated more cartilaginous tissue in the defects than DABM transplanted directly. One major reason for this could be that remnants of the latter preparation at 6 weeks still occupied part of the defects. To some extent this might have prevented repair tissue from developing in the
FIG. 9. Healed 15-mm2 defect after 26 weeks. DABM previously placed in muscle pouch for 19 days. Note cluster formation and absence of fibrillation. (H&E, x4)
J Orthop Res, Vol. 9, No. 1, 1991
18
L . DAHLBERG AND A . KREICBERGS
FIG. 10. Mostly cartilage-like tissue in a control 15-mm2defect after 6 weeks, with healed junctions, although surface is slightly depressed. (x4)
defects. It seems less likely that there exists a difference in the chondrogenic capability between the two DABM preparations. Among the subgroups of knees receiving DABM preparations previously placed in muscle for various times, we have the impression that those placed in muscle for 19 days gave the best results. However, in the absence of quantitative data this cannot be substantiated. Although histologic analysis probably is adequate in assessing gross differences in quality of repair tissue, it cannot be used to detect discrete differences. Moreover, there is today no histology staining method that includes a marker of fully differentiated joint cartilage. The reason that DABM left 19 days in muscle probably is better
FIG. 11. Control sham-operated knee, where a 5-mm2 defect was created. Abundant fibrous tissue with a central defect 6 weeks postoperative. (X4)
J Orthop Res, Vol. 9, No. I, 1991
than other preparations remains obscure. It might be explained by optimum recruitment of chondrogenic cells (number and differentiation) before transplantation. Analysis of control specimens placed in muscle for different times seemed to support this assumption. From experiments on heterotopic bone formation after transplantation of DABM from one tissue to another, Urist reported that optimum induction occurred if the transplantation was done early or late; DABM transplantation days 6-12 almost failed to yield an ossicle (23). The results of the present study seem to agree with the observations made by Urist. Thus, the 4-day and particularly the 19-day transplants produced better cartilage than the 16-
JOINT CARTILAGE REPAIR
day transplants. Presumably, 4 days in muscle was too short a time to recruit any major number of mesenchymal stem cells for transplantation to joint defects; the cells recruited for the observed cartilage regeneration probably originated from the surrounding subchondral bone or marrow. Transplants placed in muscle for 16 days may be assumed to have recruited chondro- and osteocompetent cells, but hypothetically the cells did not survive the transplantation, i.e., excision from muscle and subsequent insertion into joint cartilage defects. In contrast, the 19-day transplants appeared to contain chondrogenic cells capable of surviving transplantation. Interestingly, it proved impossible to dissect DABM bluntly from muscle after 19 days. Most surprising, there were no signs of newly formed bone adjacent to the 19-day transplants prior to insertion in the defects, although heterotopic ossicle formation is reported to occur at approximately 3 weeks (24). Thus, in our experimental model, DABM left 19 days in muscle seemed to represent the optimum for cartilaginous yield. The most important aspect of the present study is not the repair procedure per se, but rather the biologic response elicited. Thus, from our experiment, it is obvious that DABM placed in an avascular synovial environment does not induce bone formation, as observed in marrow tissue and subchondral bone. Instead, the tissue formed in articular defects supplemented with DABM is of cartilaginous differentiation, which, moreover, is retained over time. Unfortunately, the quality of the repair tissue is highly variable and, in general, too low to restore joint surface integrity. Although it seems that DABM has a chondrogenic potential, which may be used for cartilage repair, the approach has to be optimized by some modification before it can be used for the purpose suggested. Acknowledgment: This study was supported by research funds from the Karolinska Institute. The authors thank Dick HeinegHd and Ove Enquist for advice, and Folke Carell for technical assistance.
REFERENCES 1. Ashton JA, Bentley G: Repair of articular surfaces by allografts of articular and growth-plate cartilage. J Bone Joint Surg 68B:29-35, 1986 2. Buckwalter J, Rosenberg L, Coutts R, Hunziker E, Reddi AH, Mow V: Articular cartilage: injury and Repair. In: Injury and Repair of the Musculoskeletal Soft Tissue, ed by
19
SL-Y Woo, J Buckwalter, Park Ridge, IL, American Academy of Orthopaedic Surgeons, 1987, pp 465482 3. Calandruccio RA, Gilmer WS Jr: Proliferation, regeneration and repair of articular cartilage of immature animals. J Bone Joint Surg 44A:431455, 1962 4. Campell CJ: The healing of cartilage defects. Clin Orthop 64:4543, 1969 5. Chesterman PJ, Smith U: Homotransplantation of articular cartilage and isolated chondrocytes. J Bone Joint Surg 50B: 184-197, 1968 6. Convery FR, Akeson WH, Keown GH: The repair of large osteochondral defects. Clin Orthop 82:253-262, 1972 7. Enquist 0, Johansson SH: Perichondrial arthroplasty: a clinical study in twenty-six patients. Scand J Plast Reconstr Surg 14:71-87, 1980 8. DePalma AF, McKeever CD, Subin DK: Process of repair of articular cartilage demonstrated by histoloy and autoradiography with triated thymidine. Clin Orthop 48:229-242, 1966 9. Grande DA, Pitman MI, Peterson L, Menche D, Klein M: The repair of experimentally produced defects in rabbit articular cartilage by autologous chondrocyte transplantation. J Orthop Res 7:208-218, 1989 10. Itay S, Abramovic A, Nevo Z: Use of cultured embryonal chick epiphyseal chondrocytes as grafts for defects in chicken articular cartilage. Clin Orthop 220286303, 1987 11. Key JA: Experimental arthritis: the changes in joints produced by creating defects in the articular cartilage. J Bone Joint Surg 13:725-739, 1931. 12. Mankin HJ: The response of articular cartilage to mechanical injury. J Bone Joint Surg 64A:460466, 1986 13. Mitchell N, Shepard N: The resurfacing of rabbit articular cartilage by multiple perforations through the subchondral bone. J Bone Joint Surg 58A:23&233, 1976 14. O’Driscoll B, Salter RB: The induction of neochondrogenesis in free intra-articular periosteal autografts under the influence of continuous passive motion. J Bone Joint Surg 66A:1248-1257, 1984 15. Ohlsen L, Widenfalk B: The early development of articular cartilage after perichondrial grafting. Scand J Plnst Reconstr Surg 17:163-177, 1983 16. Reddi AH: Matrix induced enchondral bone differentiation. In:Mechanics of Growth Control, ed by RO Becker, Illinois, Thomas Springfield, 1981, pp 4 3 5 4 6 17. Rubak JM: Reconstruction of articular cartilage defects with free periosteal grafts. Acta Orthop Scnnd 53:175-180, 1982 18. Rubak JM, Poussa M, Ritsila V: Effects of joint motion on the repair of articular cartilage with free periosteal grafts. Acta Orthop Scand 53:187-191, 1982 19. Salter RB, Simmonds DF, Malcolm BW, Rumble EJ, MacMichael D, Clements ND: The biological effect of continuous passive motion on the healing of full-thickness defects in articular cartilage. J Bone Joint Surg 62A:1232-1251, 1980 20. Skoog T, Ohlsen L, Sohn SA: Perichondrial potential for cartilagenous regeneration. Scand J Plast Reconstr Surg 6~123-125,1972 21. Syftestad G, Caplan A: A 31000 dalton bone matrix protein stimulates chondrogenesis in chick limb bud cell cultures. Trans Orthop Res SOC 11:278, 1986 22. Urist MR: Bone: formation by autoinduction. Science 150: 893-899, 1965 23. Urist MR, Hay H, Dubac F, Buring K: Osteogenetic competence. Clin Orthop 64:19&200, 1969 24. Urist MR, Silverman BF, Buring K, Dubac F, Rosenberg JM: The bone induction principle. Clin Orthop 53:243-283, 1967 25. Urist MR: The origin of articular cartilage: investigation in quest of chondrogenic DNA. In: Cartilage, vol. 2, ed by K Hall, New York, Academic Press, 1983, pp 1-85
J Orthop Res, Vol. 9, No. 1 , 1991
