Unfavorable fractures in bilateral sagittal split osteotomy: a systematic review of classification and management strategies

INTRODUCTION

Bilateral sagittal split osteotomy (BSSO) is a commonly performed technique in orthognathic surgery due to its broad contact area, which enhances post-surgical stability in both mandibular advancement and retraction procedures. Since its development by Trauner and Obwegeser [1], various surgeons have modified BSSO to reduce the occurrence of complications while refining the overall surgical approach [2-5]. Even with the advancements and widespread use, this procedure requires meticulous surgical skills, and potential complications can arise [6-8]. The most common complications include accidental fractures of the distal or proximal segments as well as damages to the inferior alveolar nerve, leading to sensory disturbances in the lower lip, post-surgical infections, bleeding, inadequate fixation, the need for osteosynthesis material removal, and the risk of relapse [8-12].
A bad split, an unfavorable and unexpected split pattern, is a potential complication during BSSO. A 20-year study by Falter et al. [12] showed that bad splits were observed in 14 out of 2,005 BSSO sites. Similarly, a retrospective study by Jiang et al. [13] indicated bad splits in 40 cases (4.149%) out of 484 individuals. Also, a systematic review was conducted by Steenen et al. [14], analyzing 33 studies published between 1971 and 2015, and reported 458 instances of bad splits occurring within 19,527 sagittal ramus osteotomies performed in 10,271 patients. The overall prevalence of bad splits among sagittal splits was estimated at 2.3%, with individual studies reporting rates ranging from 0.21% to 22.72% per patient, and from 0.2% to 14.6% per BSSO [21, 22].
Bad splits may occur as a result of a structurally thin mandibular ramus, an elevated mandibular lingula, the presence of third molars, incorrect angulation of the osteotome, or even the surgeon’s lack of experience or insufficient focus during the procedure [4, 11, 15-17]. Some researchers indicated that the primary contributing factor to bad splits might be the surgical technique. Moreover, the selection of instruments during the procedure and differences in tools used, such as a chisel versus a spreader, or fine chisels versus heavier ones as well as the number and order of strikes, are considered significant variables [17]. An incomplete transection of the mandible’s inferior border may also be an influencing factor [18], but the reported results remain inconsistent. Precious et al.’s study [15] demonstrated that the existence of third molars (3M) contributes to a higher likelihood of bad splits, whereas Camargo et al. [19] concluded that 3M do not elevate this risk, even among older individuals. Kriwalsky et al. [20] showed that elderly patients have a greater predisposition to bad splits, but Chrcanovic and Freire-Maia [21, 23] found that adolescents face a higher risk of bad splits compared with adults. Therefore, further studies are necessary to confirm the impact of these risk factors, as the existing literature lacks a clear agreement regarding specific combination of factors, which increase an individual’s susceptibility to a bad split [22]. Proper and precise management is crucial, although rare complications associated with these surgeries do occur and must not be neglected [4, 24-26].
Unfavorable fractures have been classified into various categories by different authors [8, 14, 34-36, 49, 50]. However, only some of these articles provided detailed strategies for managing the fractures and addressed the associated challenges. While these classifications offer a framework for understanding different types of fractures, the practical guidance on handling them effectively is often either very limited or entirely absent in several studies. The presented approaches highlighted both the theoretical and clinical aspects, ensuring that practitioners are equipped with comprehensive knowledge to tackle the complexities of unfavorable fractures. An undesired fracture can lead to post-operative instability or impaired function of the mandible, which may contribute to the development of temporomandibular joint (TMJ) disorders [31]. In addition, there is a risk of infection, bone necrosis, and prolonged healing time [21, 31].

OBJECTIVES

The main aim of this systematic review was to categorize unfavorable fractures occurring during BSSO procedures, and evaluate the efficacy and constraints of classification systems. Additionally, the current review aimed to identify and understand the challenges encountered in repairing unfavorable fractures, emphasizing the role of anatomical variations and surgical techniques. Furthermore, the study evaluated various management strategies employed for handling bad splits, focusing particularly on the effectiveness of different stabilization techniques, such as the use of osteosynthesis plates and screws. Another key objective was to analyze the incidence rates and patterns of unfavorable fractures by providing a statistical overview of bad split occurrences from various studies, and to identify common fracture patterns and their clinical implications. In addition, this review aimed to recommend practical treatment guidelines based on the evidence, propose strategies for managing unfavorable fractures, and suggest improvements for pre-operative planning and stabilization methods to reduce the risk of complications. Finally, the study identified deficiencies in the existing studies, and proposed directions for additional investigation, underscoring the necessity of continued research to enhance management strategies and optimize surgical outcomes.

MATERIAL AND METHODS

This systematic review was carried out as per the guidelines of PRISMA statement [27].

Search strategy

A detailed search strategy (Table 1) was applied in databases, including PubMed, Scopus, and Cochrane Library on July 3, 2024. Boolean operators (AND, OR, NOT) were used to combine search terms appropriately. The sole limitation was that the articles considered were published within the last 20 years.

Eligibility criteria

Case reports as well as retrospective and prospective studies on unintended splits during BSSO procedures from the past twenty years were considered, irrespective of their inclusion criteria of control groups.

Study selection

After initially evaluating articles’ eligibility by reviewing titles and abstracts in a standardized way, relevant studies were obtained, their full-texts read to confirm eligibility, screened for duplicates, and assessed for relevance.

Data extraction

Data gathered involved authors, publication date, patient number, demographic and medical details, technique applied for BSSO, types and classifications of unanticipated fractures, treatment methods, and final results.

RESULTS

Eleven articles were reviewed for this investigation following PRISMA guidelines. The PRISMA flow diagram for the study is illustrated Figure 1. A total of 233 records were initially identified, with additional 2 articles selected through manual search. After duplication removal and screening based on relevance, 58 full-text articles were assessed, resulting in 11 studies included for the final review. These comprised 4 retrospective cohort studies [4, 33-35], 4 retrospective chart reviews [8, 12, 28, 32], 1 prospective cohort study [29], and 2 case reports [30, 31]. The results of data extraction are detailed Table 2.

Summary of key findings

Among the first eight studies, Acebal-Bianco et al. [28] documented the lowest occurrence of bad splits, with a rate of 1.0% per patient and 0.5% per split side, utilizing the Dal Pont-Hunsuck-Simpson-Epker modifi­cation method. In contrast, the highest bad split incidence was reported by Borstlap et al. [29], with an incidence rate of 9% per patient and 4.5% per split side, using the Obwegeser and Dal Pont modification with or without the Hunsuck modification technique. Across a total group size of 3,988 patients, there were 85 occurrences of bad splits. The majority of these unfortunate fractures occurred in the proximal segment, followed by the distal segment, the coronoid process, and the condylar process (Table 2).
Data on proximal segments involvement were mainly reported by Acebal-Bianco et al. [28], Teltzrow et al. [8], Falter et al. [12], Lloyd et al. [31], and Mensink et al. [32], while distal segment fractures were mostly documented by Mehra et al. [4] and Borstlap et al. [29]. Unfavorable fractures of coronoid process were documented twice, in the studies by Acebal-Bianco et al. [28] and Teltzrow et al. [8]. Similarly, bad splits of condylar process were noted twice, by Mensink et al. [32] and Teltzrow et al. [8]. Lee et al. [33] and Jiang et al. [34] concentrated solely on assessing buccal segment fractures in the proximal region. In the research by Jiang et al. [34], an innovative classification system was introduced for adverse fractures affecting the proximal segment. In the latest article by Chandegra et al. [35], there were 22 bad splits among 311 individuals, who underwent BSSO procedures. All unintended fractures were assessed and classified using a newly developed classification system, which provides a better understanding of the management process for unanticipated splits.
In the management of bad splits’ complications during BSSOs, all authors consistently employed osteosynthesis plates [4, 8, 12, 28-35]. The results of managing these adverse effects were generally positive; how­ever, some complications were also reported [28, 32-34]. Acebal-Bianco et al. [28] used osteosynthesis plates and bicortical screws with releasing of maxillomandibular fixation (MMF) at the end of the surgery, resulting in infections and complications, such as hematoma, TMJ’s tenderness, displacement of the disc, lower lip’s sensory impairment, and irregular contour of the mandibular angle.
Mehra et al. [4] used 2 mm bicortical screws, additional screws, with a 1.1 mm bone plate stabilized using mono-cortical screws, and reported no post-operative complications related to infection. Borstlap et al. [29] managed complications with tight elastics, additional plates, and a splint, leading to positive patients’ outcomes without neurosensory and skeletal issues. Teltzrow et al. [8] employed osteosynthesis plates with MMF, and observed stable bone healing. Kincaid et al. [30] combined osteosynthesis plates, positional screws, mono-cortical and bicortical screws, 4-week rigid MMF, and heavy elastics, resulting in stable occlusion without TMJ symptoms. Falter et al. [12] reported good functional occlusion six months post-operatively using osteosynthesis plate. A case report study by Lloyd et al. [31] demonstrated 2 cases. In the first patient, a plate for osteosynthesis and MMF was used. In the second case, similar procedures as in the first patient were performed, but MMF was released at the end of surgery. The post-operative outcomes were positive, with proper fracture healing, restored function, and no facial asymmetry, leading to good cosmetic results.
Mensink et al. [32] employed osteosynthesis plates and MMF, and observed successful recovery despite some neurosensory disturbances in two patients. Lee et al. [33] used mini-plates, mono-cortical screws, and additional bicortical screws, reporting delayed bone healing. The presence of a bad split caused changes in patients’ cephalometries. Jiang et al. [34] used MMF rigid fixation with osteosynthesis plates and bicortical screws, demonstrating satisfactory outcomes, although improper alignment of the condylar process was observed in patients, who experienced fractures corresponding to a specifically defined type of fracture as per novel classification system for proximal segments.
Chandegra et al. [35] documented the application of mini-plates, mono-cortical, and bicortical screws as well as potential implementation of vertical sub-sigmoid osteotomy (VSSO) in cases with type 2 fractures, according to the novel classification system of bad splits, along with tight elastics and the final wafer. Post-operative patient outcomes were not evaluated.

DISCUSSION

In BSSO, adverse fractures pose serious complications, potentially leading to failed surgical outcomes. These fractures vary greatly due to specific factors, such as individual anatomical differences in the mandibular ramus, existence of third molars, surgeon’s level of experience, bone density, age, and sex [22]. Despite the improvement in surgeons’ skills and knowledge, along with advancements in specialized instruments, managing these unanticipated fractures remains a difficult challenge for professionals.
The current systematic review focused on presenting the classification and management of unfavorable fractures, known as bad splits, occurring during BSSO. Conventionally, bad splits have been categorized according to the anatomical site of the fracture, encompassing fractures of the proximal segment, distal segment, condylar process, and coronoid process [14]. This classification system has been used by various authors, including Acebal-Bianco et al. [28], Mehra et al. [4], Borstlap et al. [29], Teltzrow et al. [8], Falter et al. [12], Lloyd et al. [31], Mensink et al. [32], and Lee et al. [33].
Recent studies by Jiang et al. [34] and Chandegra et al. [35] have introduced new classification systems for bad splits. Jiang et al. [34] developed a new system specifically for proximal segment fractures, dividing them into three types based on the nature of fracture and the corresponding treatment approach. The illustration shown in Figure 2 was created based on a figure from the article.
Type 1 fractures, referred to as transverse fractures of the mandibular ramus, occur when the ramus is separated into two complete sections along a horizontal plane. These fractures are particularly difficult to manage, and often necessitate halting BSSO surgery. The treatment involves applying MMF while exercising extreme caution to minimize fracture movement, thereby reducing the risk of additional neurovascular injury or delayed bone healing. In some cases, the osteotomies may need to be performed again after a recovery period to ensure proper healing and alignment. When an excessively deep horizontal osteotomy on the medial ramus results in the condyle and coronoid process is partly involved in the same proximal segment, maintaining precise control of the condyle’s position becomes more difficult [34, 36]. The segments of the ramus can be positioned and secured through either an external or internal technique. Nonetheless, the limited surgical space inside the mouth prevents the completion of rigid internal fixation, and patients might reject the external scarring.
Jiang et al. [34] outlined the potential strategies for managing this type of fractures with the associated challenges. One approach is to initially perform a BSSO procedure on the opposite side, and maintain MMF in pre-determined occlusion for a period of 4 to 6 weeks in cases requiring mandibular repositioning. Another strategy involves utilizing trans-oral plating in combination with a trans-buccal trocar technique, to facilitate accurate alignment and promote proper fracture healing. Alternatively, the procedure may be discontinued, followed by the application of MMF. It is essential to limit movement during both the surgical procedure and fixation phase, to avoid additional damage. An extended healing period of 4 to 6 weeks is often required to ensure sufficient bone recovery. After this time, the osteotomies can be performed again. Stabilization of the maxilla and mandible in the intended occlusion can be achieved with a low possibility of condylar misalignment in cases of mandibular repositioning. However, if mandibular advancement is necessary, ensuring accurate condylar positioning within the proximal segment may prove difficult, potentially leading to condylar displacement and significant occlusal discrepancies, which might require further surgical intervention.
Similarly, Van Sickels and Salman [36] in “Complications in Orthognathic Surgery” from “Peterson’s Principles of Oral and Maxillofacial Surgery, Vol. 1” (4^th ed.) indicated that when the proximal and distal segments do not make contact after adjusting the occlusion in cases of significant mandibular advancements, the large proximal fragment should be secured using rigid fixation, preferably in its original position. If fixation must be performed externally on a separate surface, it would require temporary disarticulation. The fragment should then be reattached to the remaining proximal segment (Figure 3), potentially utilizing a percutaneous trans-buccal technique for screw insertion or specialized right-angled instruments. At this stage, handling the proximal segment as a single unit becomes more feasible, allowing it to remain stabilized in its designated position relative to the distal segment, using multiple plates or bicortical screws. Proper condylar positioning can be ensured by exerting controlled posterior, superior, and vertical pressure on the proximal segment. Additionally, a proximal-distal clamp (e.g., Jeter-Van Sickels) may be applied before inserting bicortical screws or alternatively, the coronoid process can be clamped to secure the proximal fragment before screw placement.
Type 2 fractures are unintentional fractures involving detached bone fragments, where an extra fragment becomes dislodged throughout division of the proximal and distal segments.
Depending on the location and size of the fractured bone pieces, the complexity of handling detached bone segments differs. The separation of the proximal and distal sections of the mandible should be fully completed before proceeding. When a buccal fragment becomes dislodged, it is usually due to insufficient or incomplete osteotomy at the inferior margin of the lateral vertical cut [18, 36]. To properly finalize the split, a deep groove must be formed at the inferior border, linking it to the initial osteotomy as it progresses into the external oblique ridge [36]. By carefully adjusting and manipulating the segments or if necessary, performing additional bone sectioning, the intended division can be achieved. Securing the distal segment with intermaxillary fixation (IMF) facilitates the stabilization of the detached buccal plate fragment using screws and plates, as described by Van Sickels and Salman [36] (Figure 4). To prevent sequestration, small bone fragments that lost their periosteal attachment, should be removed [34, 37]. Larger fractures with an intact periosteum must be immediately stabilized and reduced using plate osteosynthesis [18]. Moreover, the author of the classification [34] emphasized that fractures in the distal portion can be managed by realigning the proximal and distal sections, and subsequently securing the detached distal fragment to the mandible in its adjusted position. Small bone fragments should be meticulously extracted and eliminated to minimize the risk of infection. In uncommon cases, the detached fragment may become lodged beneath the masseter muscle or near the posterior margin of the mandible, making it difficult for the surgeon to access. In such situations, leaving the fragment undisturbed without additional intervention is a viable option. Ensuring stable fixation of the proximal and distal portions in all cases is crucial to prevent potential complications.
Type 3 are classified as unexpected fractures’ configurations, which do not involve detached bone fragments; adverse fractures occur when the proximal and distal segments can be effectively separated, but the fracture pattern deviates from the Hunsuck standard.
Attention should be given to thorough inspection and fastidious care. Type 3 adverse fractures commonly develop at the posterior region of the mandibular angle or along the lower border of the mandible [38]. These fractures can result from an unsplit mandibular angle due to improper chisel direction, or an unsplit inferior border from inadequate vertical osteotomy depth. Although fracture patterns may vary, the approach for the management of type 3 fractures remains consistent. A meticulous examination is essential during assessment, and ensuring a clear distinction between the proximal and distal sections, with bite alignment and condyle positioning verification, is crucial. Even if the fracture configuration is sub-optimal, the stabilization process can still be performed successfully [39, 40], and if these conditions are met, the surgery can proceed effectively. In cases where a large buccal plate fracture is present, treatment choices are more restricted due to reduced surgical visibility. Extensive extra-oral incisions in the facial or neck region for direct access to the fractured segments, can increase the risk of facial nerve damage and potential scarring if immediate repair is attempted [30]. To mitigate these complications, an endoscopic approach can be utilized for improved visualization of the lateral ramus and condylar segments, minimizing the disadvantages associated with extra-oral open reduction [41-44].
Chandegra et al. [35] introduced a new classification system for bad splits, and described specific techniques for managing such fractures. Unfavorable cracks were categorized into three distinct types, excluding splits of the coronoid processes, as they do not affect BSSO handling [35]. The image (Figure 5) was created based on a figure from the article.
Type 1 splits includes fractures, in which the mandibular condyle is either fully fractured or remain connected to the proximal segment of the split that does not bear teeth. In this type of fracture, the osteotomy can still proceed, allowing for mandibular advancement or repositioning, with the final wafer applied. According to the article, the authors managed every type 1 fracture by fully performing the BSSO procedure, separating the proximal and distal segments, and handling it as a standard sagittal split osteotomy using the Obwegeser technique and Dal-Pont and Hunsuck modifications. They suggested managing the condylar fracture following fracture management principles. Type 1 splits may include partial fragmentation of the buccal cortex, which can be stabilized using osteosynthesis plates and screws. The initial step involves completing the sagittal split and then placing the final wafer to achieve the intended occlusion. Osteosynthesis plates and screws are utilized to secure the fractured buccal segments as well as the proximal and distal fragments. For an isolated fractured condyle, initial conservative management is recommended, followed by post-operative evaluation to determine if an open reduction and internal fixation or a closed approach is necessary.
Van Sickels and Salman [36] suggested that for an isolated condylar fracture, where the condylar fragment is detached from the proximal segment, rigid fixation is necessary to form a single proximal segment (Figure 6). After successfully securing the proximal segment, both the proximal and distal sections can be fixed following the standard stabilization protocol used in a typical BSSO procedure.
Steenen et al. [14] indicated that aligning bone fragments and using semi-rigid plating is the most effective method for treating this type of fracture. Such complex procedure often demands expertise in open reduction and internal fixation for condylar fractures as well as transcutaneous access. In some cases, it might be best to discontinue the procedure and attempt it again after bone consolidation.
Proper condylar positioning is challenging, especially when managing a small condylar stump. Trans-buccal percutaneous incisions can facilitate the placement of a plate on the condylar fragment, allowing for controlled adjustment of the condyle during drilling though the distal segment and application of screws for stabilization. Alternatively, an endoscope-assisted approach can be applied to improve visibility and ensure secure fixation within this confined surgical area [41-44].
Type 2 splits are fractures, when the mandibular condyle remains connected to the distal (tooth-bearing) segment, typically along the lower border of the mandible, accompanied by a fracture of the buccal cortical plate.
The clinician must attempt to separate the remaining portion of the mandible. If this is not possible, transitioning to a vertical sub-sigmoid split osteotomy (VSSO) becomes necessary, as a BSSO cannot be successfully performed when the mandibular condyle remains attached to the tooth-bearing segment. A VSSO serves to detach the condyle from the tooth-bearing segment. The fractured buccal bone is subsequently positioned between the separated condyle and the distal tooth-bearing segment, and secured in place using a bicortical screw that anchors it to the lingual bone. This approach facilitates proper bone-to-bone contact, promoting optimal healing. To minimize the risk of TMJ complications, ensuring accurate alignment of the mandibular condyles within the glenoid fossa during fixation is essential. Furthermore, in some cases, the coronoid process may need to be excised to provide better access to VSSO achievement.
A well-documented drawback of VSSO is its applicability to specific cases only. It is primarily effective for slight adjustments in mandibular positioning, typically within a 1-2 mm range, or for mandibular retraction [45, 46]. Another constraint associated with VSSO is its inability to accommodate significant mandibular advancements.
For each type 2 split salvage procedure, the authors of the classification utilized a final wafer to ensure precise alignment of the maxilla and mandible, according to the intended post-operative occlusion [47]. Placing the fractured buccal segment between the proximal and distal osteotomy sections contributes to segment stabilization, and enables healing by optimizing direct bone contact.
Type 3 splits emerge when the lingual plate is fractured.
Over 11 years of research on bad splits, Chandegra et al. [35] did not observe fractures of the lingual plate. They propose aligning the proximal and distal segments of BSSOs using lingual osteotomies to reduce the risk of a relapse and prevent condylar torque, thereby decreasing the likelihood of TMJ indications. This approach avoids using fixation, and employs osteosynthesis plates and mono-cortical screws to stabilize the buccal segments, referring to a research by Ellis, 3^rd ed. [48].
According to Van Sickels and Salman [36], lingual fractures often occur due to the presence of an impacted third molar, which results in a thin and structurally weak cortical plate in the region surrounding the third molars [14]. Additionally, an insufficiently completed osteotomy of the medial horizontal ramus or the sagittal section of BSSO can contribute to fracture formation, leading to a wedging effect on the medial aspect of the mandible. To minimize the risk of such fractures, the authors advise removal of third molar at least nine months before performing BSSO. If an unexpected fracture occurs, the osteotomy should still be completed as originally planned. The free lingual segment, which generally maintains its muscular and vascular attachment, is not a significant obstacle. The distal segment should be stabilized with IMF, while the free lingual plate fragment should be repositioned anteriorly to establish contact with the distal segment and affixed to the proximal segment using bicortical screws. In order to secure both the proximal and distal segments, one or more plates may be used to the buccal part of the osteo­tomy site.
Steenen et al. [14] suggested that in cases with vertical fractures of the lingual plate (Figure 5), the osteo­tomy can be finalized while leaving the lingual plate detached. Stabilization can be achieved using buccal plates and mono-cortical screws, with additional option of securing the lingual fragment with one or two bicortical screws. Once the proper alignment is established, fixation may not always be necessary. For horizontal fractures (Figure 5), the surgical process remains unaffected, and stabilization can be carried out during the same procedure through plate osteosynthesis, or by securing the upper border with bicortical screws. If required, bicortical screws can also be placed at the lower border to anchor the lingual fragment.
Teltzrow et al. [8] categorized problematic fracture patterns into three types, i.e., the buccal cortex, the coronoid process, and the condylar process. Later, Plooij et al. [49] developed the lingual split scale using 3D-CT after BSSO, introducing four categories based on the fracture line’s path along the lingual side of the ramus. Subsequently, Muto et al. [50] further classified fractures occurring in the lingual surface, posterior border, and buccal surface of the mandibular ramus into five distinct split patterns. Also, Steenen et al. [14] created a classification system after reviewing 33 studies, categorizing unexpected splits into four types: type 1 – proximal segment fractures, type 2 – distal segment fractures, type 3 – coronoid process fractures, and type 4 – condylar process fractures. These fracture patterns were further divided into sub-categories, i.e., 1a-f, 2a-b, 3, and 4.
Nevertheless, the presented classifications primarily focus on the anatomical location of unfavorable fractures, and do not provide comprehensive information regarding clinical management. Additionally, they offer only limited practical treatment guidelines within own frameworks.

CONCLUSIONS

This review emphasizes the classification and mana­gement of unfavorable fractures, known as bad splits, occurring in BSSO procedures. Despite advancements in surgical techniques and tools, the management of such fractures remains challenging due to variability in individual anatomical structures and other factors.
Numerous authors have developed classification systems to better understand and manage these fractures. While the majority of these systems primarily focus on anatomical locations, they often lack comprehensive clinical management guidelines. Effective management strategies have been proposed, including the use of osteo­synthesis plates, mono-cortical and bicortical screws as well as specific techniques, to align the segments and ensure proper healing.
For better clinical outcomes, integrating practical treatment guidelines into these classification systems is essential. This approach will provide practitioners with the necessary tools to address the complexities associated with unfavorable fractures effectively. In conclusion, while significant progress has been made in the classification and treatment of bad splits, further studies and refinement of strategies’ management are necessary to improve patient outcomes.

Disclosures

1. Institutional review board statement: Not applicable.
2. Assistance with the article: None.
3. Financial support and sponsorship: None.
4. Conflicts of interest: The authors declare no potential conflicts of interest concerning the research, authorship, and/or publication of this article.

References