This study aimed to assess if the type of bone adjacent to the intersegmental gap affects long-term osseous healing after mandibular reconstruction with patient-specific 3D-printed titanium reconstruction plates. Volumetric assessment of repetitive CBCT scans, as well as µCT and histological analyses, was used to quantitatively and qualitatively evaluate bone healing. To the authors’ knowledge, this is the first study using this combination of assessments to analyze bone healing at the mandible.

The results of this study show no significant differences in volume change between fibula/fibula and fibula/mandible interfaces. Therefore, there seems to be no relevant impact of the bone adjacent to the intersegmental gap in mandibular reconstruction on the rate of osseous union in the long term. This finding is contrasted by that of the study of Yoda et al., who stated that bone adjacent to the gap influences the bone healing process [8]. In that study, CT data of a 66-year-old patient who received mandibular reconstruction with a FFF were analyzed and a finite element (FE) model was created to quantify mechanobiological responses. Hereby, it was shown that cortical bone formed earlier at the osteotomized interface between fibular and fibular bone. However, there was only one patient included in the study and, more importantly, the bone interfaces were at different anatomical sites, resulting in different mechanics. Disregard of anatomical sites and therefore differing mechanical stimuli were also shortcomings of other studies: Swendseid et al. analyzed maxillary and mandibular defects in 104 flap reconstructions and found a significantly higher rate of complete unions between osteotomized free flap segments compared with flap segments and native bone (65% versus 53%) [9]. However, fibula, scapula and radial forearm flaps as well as maxillary and mandibular defects were combined in the analysis. Also, Knitschke et al. combined maxillary and mandibular defects in the analysis of 133 patients who underwent reconstructions using FFFs [10]. Further shortcomings of both studies were the lack of multivariate analyses concerning the difference in osseous union between different bone interfaces, subjective radiological analyses without the volumetric methodology and the combination of different anatomical regions. Furthermore, the initial gap size was not taken into account, although its importance and confounding influence has been previously demonstrated [5, 15].

The present study overcomes these limitations by using the volumetric assessment of intersegmental bone volumes using repetitive CBCT scans, as previously described [5]. This methodology allows an objective evaluation of radiological pseudarthrosis, which is a relevant improvement compared to the subjective, visual determinations by other studies [3, 15,16,17,18]. Furthermore, the present study minimized the effect of mechanical confounders by only including patients who underwent reconstructions using patient-specific 3D-printed titanium reconstruction plates and multi-segment reconstructions with two intersegmental gaps in the anterior region. This allowed analysis of the interfaces between mandible/fibula and fibula/fibula in a similar anatomical region within the same patient. The reduction of mechanical confounders is especially important, because mechanics have been previously described as an important factor in bone healing [19]. In mandibular reconstruction, this has been proven by the relevant influence of different types of osteosynthesis plates [20] and the gap site (anterior versus posterior) [5] on the development of pseudarthrosis. Postoperative non-occlusion as a further mechanical variable did not significantly influence non-union rates in a previous study [3]. In this study, postoperative occlusion was only present in one patient. Therefore, this study cannot analyze the impact of dentition and further studies using adequate bite force measurement devices are needed [21].

Compared to the previously mentioned studies the use of multivariate analysis is a further strength of the current study. Radiotherapy, which had already been identified as a risk factor for pseudarthrosis [22], was confirmed as an independent risk factor for diminished osseous healing. Also, the relevant influence of time on osseous healing was confirmed by the present study and its influence was taken into account using the regression analysis [10]. The insignificant impact of the initial gap size on osseous healing shown by the multivariate analysis presumably resulted from the limited sample size of the present study. The relevant influence of the indication for surgery and wound healing disorders on osseous healing may also be a result of this and need to be further investigated by studies focusing specifically on these factors. Despite the inclusion of all these factors in the regression analysis, the type of bone adjacent to the gap remained an insignificant factor in the development of pseudarthrosis (Fig. 4).

Clinical demonstration of osseous healing after mandibular reconstruction. A 65-year-old female patient underwent plate removal 11 months after having received a two-segment fibula free flap. There was complete osseous union at both anterior intersegmental gaps. Bleeding from the former drill holes after plate removal indicated good vascularization of the flap. Created with Biorender.com

This finding is remarkable considering that differences between the mandible and other bones are clinically apparent. For example, cherubism and antiresorptive agent-related osteonecrosis of the jaw (ARONJ) are both pathognomonic for the jaw bone [23, 24]. In order to reveal the specifics of the mandibular bone compared to other bones, morphogenetical and morphological analyses were the focus of previous studies: The mandible originates from neural crest mesenchyme (NCM), while the fibula bone has a mesodermic origin [25]. Findings of greater osteogenic potential of the mandible compared to other skeletal bones have been previously described and might be related to the difference in morphogenesis [24, 26]. Mandible bone marrow stromal cells formed 70% larger bone nodules with a three-fold more mineralized bone after implantation into nude mice, in comparison to cells derived from the long bone [24]. In another study, orofacial human bone marrow stromal cells proliferated more rapidly in vitro, indicating orofacial marrow stromal cells as a unique cell population [26]. Also, in comparison to the maxilla, the mandible demonstrated a higher bone remodeling dynamic in animal experiments with skeletally mature dogs [27].

Generally, the mandible is assumed to present higher amounts of collagen compared to the long bone, but posttranslational modifications of collagen, such as intermolecular crosslinking and lysine hydroxylation, are less mature in the mandible, which is meant to allow more flexibility and better resistance to constant exercise [23]. Bone mineral density (BMD) as a measure of bone strength is suggested to be higher in the body and symphysis area of the mandible compared to the spine and hip [28].

However, there are also studies showing the opposite: Rothweiler et al. compared the three-dimensional microstructure of alveolar and iliac bone and found a greater distance from mineralized tissue to the closest pore-vessel boundary in alveolar bone, which was associated with a worse regenerative potential of the alveolar bone [29]. Also, the more rapid proliferation of orofacial human bone marrow stromal cells in vitro could not be proven in vivo, where iliac crest cells formed more compacted bone and were more responsive to osteogenic induction [26]. This demonstrates that the specific properties of the mandible are still a matter of discussion. Even though there might be advantages in comparison with other bones, the influence of these distinct properties on long-term osseous healing seems to be neglectable.

In the present study, the preliminary histological observation revealed similar microstructures of the intersegmental bones of both interframes, although the fibula–fibula intersegmental bone seems to maintain the flap’s native morphological characteristics. Considering the macroscopic differences in shape and size between the mandibular and fibular bone, the preliminary morphological differences in osseous union between the intersegmental bones, as observed in the present study, indicate that the morphological differences can be devalued compared to the influence of mechanics or radiotherapy.

Several limitations account for these findings. Pseudarthrosis was only evaluated radiologically and the biomechanics of the intersegmental callus remain unknown. Due to strict inclusion criteria, the number of patients included is relatively small. Following our in-house standard, not all patients received two consecutive CBCTs because most patients suffered from malign tumors and routine tumor staging usually included a CT as staging, which limited the number of patients. Furthermore, there are varying time intervals between CBCT imaging and the retrospective study design. And although the methodology allowed a reduction of the impact of artefacts, interpretation may still be confounded by metal artefacts. These factors might also explain the insignificant differences in osseous union between superior and inferior parts of the gaps, although the inferior parts receive more axial load, which is beneficial for osseous healing [6].