In this study photooxidized, mushroom shaped, osteochondral grafts of equine origin implanted in sheep were superior in cartilage surface integrity, persistence of grafts and reaction of adjacent host tissue at 6 months after implantation, if compared to equally pretreated bovine, ovine and human grafts. The slower resorption and different cartilage and bone density of the equine grafts was held responsible for their improved overall performance. Additional groups of ovine, human and bovine grafts were subjected to a more intensive cleansing procedure that was shown to reduce the durability of the cartilage surface, but also the subchondral bone. However, if the effect of the species was compared to the cleansing procedure, the cartilage and mainly bone density due to species differences seemed to be more important. The equine grafts showed the best results, even though this group was represented only through grafts that had undergone the additional cleansing procedure.
The human transplants could not be prepared exactly as those of the other species. This was due to official regulatory problems, such that catilage grafts could only be obtained from an official bone bank. This may have been the reason, why these transplants were not of the same and even quality as all other grafts. This was especially true for the cartilage surface that looked macroscopically more hampered, but also for the more uneven thickness of the human cartilage. It may well be that the overall results would have been improved for both groups with human grafts if the same harvesting standard could have been maintained also with the human grafts. The equine grafts (EN) showed the most consistent good quality, although only grafts with the more rigorous cleansing procedure were used.
Surface incongruencies between graft and host cartilage could not always be avoided, although its importance was addressed also by other authors[1, 3, 12, 14, 18–22]. Since biological transplants will never be completely symmetrical also in auto- transplantations, a minimal step at the superficial cartilage layer and/or at the calcified cartilage zone and the tidemark will always occur. Although an attempt has to be made to implant the graft flush with the cartilage surface, a difference up to 0.1–0.3 mm has to be acceptable for clinical reality. As long as the grafts are not left proud of the surface, this should not seriously jeopardize the outcome, at least with photooxodized grafts [4, 5].
Based on our earlier experiments the time point of sacrifice was chosen at 6 months. Cyst formation was most pronounced at this time period and was considered instrumental for graft survival [4, 5]. This was confirmed in this study, where cystic lesions were found in about one third of the grafts. The size of the lesions was not measured radiographically or histologically, since both evaluation techniques represent only two-dimensional techniques. The three dimensional aspect could have been included if microcomputer tomography would have been performed [23–26]. However, it is questionable whether it would have influenced the overall results of this study. The authors felt that calculating the percentage of fibrous versus bony tissue by using histomorphometry reflected the situation of cystic lesions adequately enough to draw valid conclusions.
The score system for semi-quantitative assessment of graft performance was the same as used in our previously published studies [4, 5]. Although it addressed similar aspects as classic scoring systems for grading cartilage [27, 28], it had to be adapted to the specific nature of the grafts. This was especially true in view of the photooxidation process that left an intact matrix as for the structure of the collagen macromolecules, but no living chondrocytes. Cluster formation is considered to be a classic sign of cartilage degeneration [18, 29–32] and was handled the same way in our adapted scoring system (score a = degenerative). However, this may not be correct. As in early osteoarthritis it may be an attempt of the tissue to prevent matrix degradation  and thus, in the case of the photooxidized cartilage actually be a positive sign for viability and repopulation with living cells. Furthermore, each score evaluated may not have the same importance. Thus, the total sum of the score may not truly represent the best graft performance in the overall picture. Surface integrity and graft dislocation are certainly among the most important features for graft survival and there, equine grafts (EN) showed the best results. Nevertheless, as in our previous work, the adapted score system served well its purpose to evaluate graft performance.
Compared to earlier results obtained with photooxidized grafts [3–5] viability and repopulation with new cells of the original photooxidized grafts was relatively low. This was dependent on the implantation period of only 6 months duration in this study, whereas in the previous work graft survival was followed for 12 and 18 months. There, viability, repopulation as well as fusion between host and graft were mainly reported for later time points. Nevertheless, ingrowing cells were noticed also already at 6 months in both studies, but mainly restricted in the deep zone of the cartilage close to the calcified cartilage zone and tidemark. The intensive remodeling of the calcified cartilage including the tide mark suggested that these cells originated from the subchondral bone area.
The intensified cleansing procedure applied on the photooxidized grafts was harmful on structural properties of the grafts. This was already visible macroscopically such, that the cartilage surfaces were roughened at the time of implantation. This may be one of the main reasons, why bovine, ovine and human grafts undergoing this process did not perform at the same high level as their counterpart groups subjected to the original photooxidation process in this study and also in the previously published experiments [4, 5, 34]. Nevertheless, equine, photooxidized, mushroom-shaped, osteochondral grafts showed the best overall results at 6 months after implantation despite the fact that only grafts prepared with the more rigorous cleansing process were used. This was attributed to the slower resorption of the bony part of the graft and high persistence of the cartilage matrix. The slower bone resorption may have resulted in better mechanical stability and thus, graft incorporation in the subchondral bone of the host. The integrity of the cartilage matrix may be due to the relative high density and strong mechanical properties of equine cartilage [35–37]. Although these features were not measured in the current study, the denser appearance of the equine bone in histology sections and the fact that harvesting the plugs with the same instrumentation resulted in a much higher resistance to the hollow drill bit support this assumption on an empirical level. It could be speculated that equine grafts prepared with the original photooxidation process would have resulted in even better performance compared to this study. In the current study, the main interest was focused on the effect of the species differences and graft resistence after implantation, where equine grafts performed best although only grafts with the more vigorous cleansing procedure were tested. Future studies should test the old photooxidation process also for the equine grafts as well as immunogenicity problems that may be related to using equine grafts for other species.