Malignancies in the posterior third of the oral cavity, base of the tongue, and oropharyngeal primaries pose a major surgical challenge. To facilitate free resection margins and adequate exposure for reconstruction, e.g. by using microvascular transplants, the transoral approach must be rejected, requiring alternative approaches that compromise functional outcome and patient QoL [30]. TORS can be a valid alternative; however, its availability is limited and poor applicability in cases of larger tumors, especially with related trismus, must be considered [29]. This study aims to provide a novel feasible approach that allows adequate access while minimizing patient debilitation.
GGM, the largest and strongest extrinsic tongue muscle, plays essential roles in tongue mobility, thereby allowing complex voluntary movements, such as the oral preparatory phase of swallowing or speaking [10]. The nonvoluntary pharyngeal phase of swallowing is characterized by hyolaryngeal elevation, which is promoted by the mylohyoid and GHM [10]. Because achieving the correct refixation of the mylohyoid is a challenge due to its fanned-out shape, GHM integrity is important for preserving hyolaryngeal elevation [10, 22].
MLR is characterized by the detachment of both GGM and GHM; reapproximation can only be performed through the use of sutures [8, 26]. This adaptation might be inadequate, because patients subconsciously activate the involved musculature immediately after surgery, which explains the clinical and functional limitations that can be observed with traditional MLR [9].
Therefore, keeping the functionality of both GGM and GHM intact is a pivotal step to optimize swallowing and speaking after MLR and involves proper refixation of the GGM and GHM to the mandible.
GGM advancement procedures and their technical variations are used for treating obstructive sleep apnea surgically [7, 23]. The aim is to apply traction to the GGM by advancing a bony mandibular segment, including the genial tubercle, and to fixate it in an anterior position through plating [7]. This is effective for preventing the base of the tongue from collapsing into the pharynx occluding the posterior airway space [23]. The use of virtually-planned cutting guides is recommended for reducing the risk of failure or injury due to the close proximity to adjacent teeth and lack of visibility on the lingual aspect [7]. Moreover, the precise location of the genial tubercle cannot be predicted reproducibly due to anatomical variations [6]. However, despite its varying shape, some form of mental spine can be found in roughly 98% of the cases [25]. In concert with the MLF, which according to Wang et al. (2015) could be identified radiographically in 97% of the cases and the GGM the area of interest could be identified reproducibly [28].
In our study, no injury to relevant structures or tooth roots could be observed. The workflow of virtual design and 3D-printing of cutting guides is increasingly integrated in orthognathic surgery as well as head and neck reconstructive oncology and will likely prevail in the future [1, 6, 21].
Therefore, the clinically established procedure of GGM advancement was modified to fulfill the specific requirements of this PT approach.
To account for the anatomical vectors of force, designing a retentive shape is required (Figs. 3–6). Notably, before the osteotomy, force vectors of GGM and GHM are directed anteriorly and superiorly, which reversely forces the bony segment posteriorly and inferiorly after osteotomy [22].
Consequently, proper virtual planning is pivotal to not only avoid injury to anatomical structures but also assure stable and wiggle-free repositioning of the segment, thereby preventing it from being pulled posteriorly and inferiorly. Hence, we believe that plating will be unnecessary if the bone cuts are performed with a very delicate saw. This leads to the transformation of tensile forces of the muscles to compressive forces to the osteotomy lines, analogous to “natural strain of compression” described by Champy et al., 1978 [4].
Therefore, immediate postoperative muscle activation and logopedic functional exercise should be possible and even advantageous. Still, a specific regimen for postoperative care must be carefully established, not to jeopardize the potential benefit of this technique.
The findings of this study revealed that the segments could be repositioned stably and could not be elevated from their position when the mouth of the anatomical specimen was closed (supplemental video material). This simulates the forces applied when activating GGM and GHM. Avoiding the application of osteosynthesis material is desirable, because it can cause severe complications, as seen in MS [8]. However, mandibular continuity is intact and if stable repositioning cannot be achieved, a delicate miniplate should suffice. A highly atrophic mandible is a potential contraindication. In these cases, mandibular continuity cannot be preserved or the risk of fracture is excessive. The relatively large number of edentulous and highly atrophic mandibles in this cohort may arise from the high average age of almost 90 years.
Another anatomical foundation includes adequate blood supply to the bony segment to avoid sequestration and ensure postoperative consolidation. Studies have reported the presence of a lingual foramen and the corresponding canal in the midline of the symphysis in 97–100% cases [15, 28]. These common canals can be detected through three-dimensional radiographic imaging despite having diameters ranging from 0.25 to 1.60 mm [13, 28]. Interestingly, although the respective foramina were described several times, they were named inconsistently throughout literature [19]. Here, we have used medial lingual foramen (MLF) based on the nomenclature proposed by Tagaya et al., 2009 and Nakajima et al., 2014 [19, 27]. Despite contradicting reports on the presence of nerves and venous vessels in the corresponding medial lingual canals, broad consensus exists that an arterial branch of the sublingual or submental artery passes through the MLF [15, 17, 24].
In this study, we found that MLF was present in all three cases; in two cases, the medial lingual artery (MLA) could be dissected (Fig. 8a–c). In these two cases, the MLA ran protected between the bellies of GGM and GHM. According to the literature, it is supplied by the submental and sublingual artery; however, many anastomoses between the lingual, facial, submental, and sublingual as well as the inferior alveolar arteries have been described [11, 16, 19]. Therefore, MLA is probably fed by the sublingual rete arteriosum to which the facial as well as the lingual artery contribute [19]. Another source of supply for the bony segment include the intact lingual periosteum and muscle attachment, which could compensate for the rare but potential lack of an MLF [5].
Considering the constant existence of MLF, a stable blood supply to the bony segment is likely. Due to its location central to mental spines, the associated vessel inserts are protected between the bellies of GGM and GHM (Figs. 8 and 9). By leaving this lingual muscle cuff as well as the lingual periosteum of the segment intact, we have demonstrated a technique that facilitates the preservation of pedicled blood supply. Hence, it can be considered a pedicled bone flap. Owing to the delicacy of MLA, this approach may prevent shearing of the vessel. Considering that GGM and GHM are separated for genioglossus advancement, it must be assumed that this specific blood supply can be compromised during this procedure [6, 12]. However, reproducible success of the procedure may be explained by random pattern supply through the muscle and periosteum [7, 12]. In contrast to that in orthognathic surgery, optimal blood supply of the segment is even more important because of the frequent necessity of adjuvant radiotherapy in oral and oropharyngeal cancer [18]. By contrast, the bony segment is positioned far from the tumor bed and a minimized local effect can be achieved by the steep dose gradient of modern radiotherapy techniques, which reduce the risk of sequestration of the segment further [18].
Potential downsides of this work derive from its nature as a study on anatomical specimens resulting in limited transferability to live patients.
The adequate anterior refixation of GGM and GHM as major protagonists for tongue and hyoid movement, and hence, swallowing and speaking will likely ameliorate the functional outcome in patients. Improvements of these functions would accelerate decannulation, independent food intake, social interaction and reintegration, and shorten hospitalization; consequently, improving patient QoL. A modification in terms of a unilateral approach leaving the contralateral floor of the mouth attached is also conceivable. We have provided the justification for further evaluation of this modified PT approach in clinical studies.