This animal study investigated the healing characteristics around teeth and implants after recession coverage using either a superficial or deep connective tissue graft from the palate or a collagen matrix. We applied descriptive histological and histomorphometrical analyses to evaluate whether differences among the groups exist regarding the healing pattern, epithelial keratinization, and dimensions of soft and hard tissues around teeth and implants.
In terms of keratinization, all groups demonstrated the formation of keratinized epithelium around both teeth and implants. In teeth, the 3 experimental groups obtained similar lengths of the keratinized epithelium, albeit significantly shorter compared to the group with the control teeth. The length of the non-keratinized epithelium was similar for the control and experimental groups. These results imply that the difference of the keratinized epithelium between control and experimental teeth might be strongly influenced by the recession defect that was surgically created. The length of the keratinized tissue around implants could not be determined due to the fact that not all implants demonstrated complete transmucosal healing.
Also, in the minipig model, other studies evaluated the amount of keratinized tissue in response to treatment of gingival recession defects. CAF alone yielded about 1 mm greater width of keratinized tissue compared to CAF + CM [15]. However, comparing CAF + CTG with CAF + acellular dermal matrix (ADM), no differences in the amount of keratinized tissue gain were found [14]. The amount of keratinized tissue averaged 2.66 ± 0.42 mm before CAF + CTG and 3.83 ± 0.47 mm 12 weeks afterwards [14]. In our study, the keratinized epithelium at the experimental teeth measured 0.86 ± 0.92 mm (SCTG), 1.13 ± 0.62 mm (DCTG) and 1.44 ± 0.76 mm (CM). This might be partly due to differences in the histometric evaluation and to the fact that no baseline measurements (i.e., before CAF preparation) of the keratinized epithelium were taken, instead, the values after 8 weeks were compared with a control tooth. Furthermore, in the present study, only mandibular teeth and sites for implant installation were used, whereas the other studies used both maxillary and mandibular sites [14, 15].
The observation that CAF + CTG and CAF + CM resulted in an equivalent amount of keratinized tissue gain is in agreement with clinical studies [19, 20] where keratinized tissue gain averaged 1.26 mm for CAF + CTG and 1.34 mm for CAF + CM [12, 21].
The present study has failed to show that superficial and deep connective tissues display different inherent characteristics to induce keratinization at the recipient site as was suggested by Ouhayoun et al. (1988). However, when interpreting the here presented results, it must be kept in mind that the connective tissue grafts were covered with a rather thick layer of flap which might have hindered the direct influence of cells within the grafts onto the epithelium. Indeed, the results of Ouhayoun et al. (1988) showed that deep connective tissue grafts had not the same ability to induce keratinization as connective tissue grafts that were harvested closer to the epithelium [4]. A recent review with meta-analysis corroborated the superior outcome of superficial grafts, reporting a mean recession coverage of 89.3% for deeper connective tissue grafts and 94.0% for de-epithelialized superficial connective tissue grafts (Travelli et al., 2019). In terms of keratinized tissue gain and recession reduction, better results were found in favor of the superficial graft [22].
Whether inflammatory processes may affect tissue keratinization is still a matter of discussion. Chronic or acute inflammation, experimentally induced in animals, was not able to convert tissue keratinization [23, 24]. On the other hand, a decrease in inflammatory indices correlated with a higher keratinized tissue width and length or in other words a reduction of gingival inflammation allowed sulcular keratinization to occur [25]. In the present study, pocket formation with subgingival calculus formation and inflammatory processes were observed at nearly all (experimental) teeth and around the implants receiving a CTG. In contrast, the implants that received a CM showed no pocket formation and healthy peri-implant soft tissue conditions with minimal (physiologically normal) inflammation. Nevertheless, no difference was observed among inflamed and non-inflamed conditions in terms of epithelial keratinization. One possible explanation for the difference in pocket formation between the CM and the CTG groups at the implants is that the rather voluminous, spherical CTGs substantially lifted the bottom of the vestibulum and may have hampered tight sealing between flap and teeth/implants thus favouring plaque-induced inflammation. Conversely, the less voluminous and rather flat CMs did not result in an elevation of the bottom of the vestibulum and around implants allowed for a undisturbed healing. Epithelial inclusions in soft connective tissue were only found in few teeth receiving SCTG and SDTG. At the teeth receiving CM as well as at control teeth no epithelial inclusions were found. This strongly implies that the epithelial inclusions derived from the connective tissue grafting procedure itself. No epithelial inclusions were found at the implant groups.
One interesting finding was that after 8 weeks of healing, both superficial and deep connective tissue grafts hardly showed any signs of degeneration or integration into the surrounding tissues. This observation was made for both teeth and implants. So far, little is known about the temporal sequence of tissue degradation/integration of transplanted connective tissue grafts from the palate. The seminal studies of Karring et al. (1971) in monkeys not only first addressed the question of the specificity of the epithelium but also described healing from a few days up to 12 months [26]. After 3 months of healing, the transposed tissues had partly degenerated [27, 28]. But here it has to be kept in mind, that the surgical techniques and species differed in the latter and the present study.
In the present study, all experimental groups yielded similar results in terms of biologic width. Of note, at control teeth the BW averaged 5.1 mm which is considerably higher than in other species or in humans [29]. In the SCTG and DCTG groups around teeth, the JE measured 2.51 ± 0.72 mm and 2.21 ± 0.81 mm, what is significantly longer than at control teeth. These results strongly suggest that the surgical manipulation of the soft tissue resulted in a repair process with an apical migration of the JE. Nevertheless, these results are comparable with previous findings in dogs [30] and minipigs, where treatment with CAF alone resulted in 2.79 ± 0.77 mm and CAF + CM in 2.26 ± 0.23 mm of JE [15]. At the implants, the JE was even longer. Also at the implants, the distance cB-PIMM averaged 3.17 ± 0.62 mm and 3.17 ± 0.43 mm for SCTG and DCTG, while a bit less for CM. Here, it might be possible that the connective tissue grafts may induce some kind of bone resorption likewise to root resorptions that have rarely been described [31–33].
Much can be discussed about the limitations of this model. The miniature pig model might not be perfectly suitable for this research question considering that it displays a different and for this type of surgical procedure more challenging anatomy of the vestibulum compared to humans. In particular, the anatomical conditions in the posterior mandible may not be optimal for this kind of investigation. Other researchers have performed coronally advanced flap surgeries after connective tissue or biomaterial transplantations in the minipig in both the mandible and maxilla [14, 15] or in a more anterior position [15]. Consequently, the 3 experimental groups resulted in a deep (CM group), very shallow, missing or directly rising vestibulum (CTG groups). The thickness of the transplanted materials together with the anatomy at these sites may account for the differences between CM and the two CTG groups. Furthermore, harvesting superficial and deep connective tissue from the palate is difficult to standardize. Implant placement and positioning in relation to hard and soft tissues had to be adapted to the anatomical situation and do not fully correspond to the situation in humans which might have been one reason for the saucer-shaped defects to occur. Furthermore, some of the implants resulted in a submerged or semi-submerged healing, while few healed fully transmucosally. During healing, adequate measures of plaque control and postoperative care were not feasible in this animal model. Consequently, tissues around all the teeth and most of the implants showed signs of inflammation and calculus formation on teeth and implants. Horizontal measurements along any level for both teeth and implants were not doable for all samples. The control teeth were not planed but then included in order to have a comparison with normal histomorphometric parameters around teeth (i.e., JE, soft connective tissue height, bone level etc.). However, while the control teeth were molars, all experimental teeth were premolars and thus not fully comparable. Finally, a rather small number of teeth and implants were treated by two surgeons. This might have caused some inter-operator variation.
To better understand the characteristics and effects of superficial and deep connective tissue grafts further studies and more suitable models are warranted.