Fracture line sequestration: An important complication of open mandibular fractures in three horses

The case series by Aßmann et al. (2025) describes three cases of mandibular fractures that developed sequestration of the fracture margin as a complication of the healing process. Two cases presented chronically after attempted conservative management; the third presented acutely. Despite robust stabilisation and debridement, two of the three cases continued to exhibit associated symptoms and were euthanised. The second case, which was presented and stabilised acutely, was in this group. For the first case, a type 1 external fixator was used with Schanz screws, and for the second case, a 4.5/5.0 narrow LCP plate was placed as a type 1 external fixator. One of the chronically presented cases underwent internal stabilisation with a 4.5/5.0 mm narrow LCP plate, continued treatment and eventually resolved after 25 months of treatment with debridement and flushing for over a year after the implants were removed.

Unilateral mandibular fractures tend to have high-success rates within the literature, with most reports indicating 90% or more success with minor complications (Belsito & Fischer, 2001; Henninger et al., 1999; Jansson, 2016). Conservative management of unilateral fractures has been shown to be highly effective; however, when complications do occur, they can be significant. One of 24 horses was euthanised due to persistent difficulty with mastication following conservative management of a mandibular fracture according to the retrospective study by Jansson, (2016). The abundant blood supply and support of the opposite hemimandible are thought to aid in the high level of success with resolution of these fractures. Interestingly, in the report by Aßmann et al. (2025), two of three cases presenting with complications were subsequently euthanised, albeit a function of the owner's decision-making.

The terms delayed union and nonunion are typically used to describe fractures that do not show normal signs of healing at 3 months (Bowers & Anderson, 2024). While none of these fractures were that old, an inappropriate healing response was observed which can be further described using characteristics of these terms. Radiographically, there is a recognised abnormality in bone response categorised as the absence of callus, mild callus and excessive callus, respectively, termed atrophic, oligotrophic or hypertrophic (Bowers & Anderson, 2024). These are thought to be associated with insufficient blood supply to the fracture site (atrophic), impaired osteogenesis (oligotrophic) and excessive motion at the fracture site (hypertrophic). Taking a step further back, there are four main recognised categories of factors that can contribute to delayed healing or the lack thereof—patient, mechanical, surgical and injury-related factors (Bowers & Anderson, 2024). In the presented cases, minimal callus formation was present which would suggest a vascular cause, but the presence of overwhelming infection and contamination are also critical. These factors also contribute to sequestrum formation and the mechanical presence of the sequestra may have contributed to a lack of meaningful bone response.

Blood supply is an important factor in the contribution to sequestrum formation. In horses, the most common area of sequestrum occurrence is the distal limb, where there is minimal soft tissue coverage and limited blood supply (Clem et al., 1988). However, the mandible still represented 4.4% of the lesions in one report (Clem et al., 1988). The presence of a draining tract is the most common complaint associated with sequestra in general, and manifestation of pain is a frequent symptom, which presents as difficulty with mastication in these cases (Firth, 1987). Most cases are treated with sequestrectomy and surgical debridement. Medical management has been successful but resulted in unsightly scarring (Clem et al., 1988). Interestingly, cattle appear to be significantly more prone to sequestration of mandibular fractures than horses, with 11 of 17 cases involving the cheek teeth region developing a sequestrum in one study (Reif et al., 2000). The region of the mandible where the cheek teeth reside is covered externally by the masseter muscle, so this area would be expected to have fewer sequestra compared to the rostral mandible, where the central mandibular blood supply is thought to be more vulnerable.

As discussed in the case series, the inferior alveolar artery runs through the mandibular canal with the inferior alveolar nerve and vein to supply the mandibular cheek teeth and endosteal blood supply of the mandible. Iacopetti et al. (2015) found that in tracing the inferior alveolar nerve, the mandibular canal courses centrally under the 09-11 cheek teeth, has a variable course ventral to the 08, and then courses medially beneath the 06-07 roots to exit the mental foramen. Despite the high level of detail afforded by CT, the entirety of the mandibular canal is difficult to trace, so imaging alone can be insufficient to establish if the canal is fully disrupted (Rawlinson et al., 2018). This further complicates prognostication and the ability to predict which cases will develop these sequestra because of disruption of the inferior alveolar artery. For the second case reported, fistulation of the mental foramen was observed which may have coincided with trauma and necrosis of the structures within the mandibular canal (Aßmann et al., 2025). Damage to either the mental or mandibular foramen that is observed on CT may be the first step in being able to predict these sequestra as a potential complication in mandibular fracture cases.

One other interesting note is that in humans, spontaneous osteonecrosis and sequestra of the mandible have been reported with bisphosphonate use, despite evidence that it does not affect bone-healing time in other anatomical regions (Gao et al., 2021). While bisphosphonates were not used in these cases, insights from this human phenomenon may apply. The prevailing theory is that the rapid turnover and heightened osteoclastic activity of the mandible contribute to necrosis and sequestration when inhibited or disrupted (Rogers et al., 2000). Necrosis is spontaneous; the blood supply is not disrupted and there is not a clear route of bacterial inoculation, but sequestra have been documented, which would suggest that disruption of osteoclastic activity alone may be sufficient to result in clinically relevant sequestra.

The author has experienced a similar imaging phenomenon to the one described in a case at our hospital following an iatrogenic fracture during 309 tooth repulsion through a lateral alveolar wall approach in a 25-year-old Shire mare. The fracture ran obliquely through the entire cortical portion of the horizontal ramus of the mandible but seemed stable and was left to heal by conservative management. The horse was eating well 3 days after surgery, with phenylbutazone at a tapered dose and was discharged with a 2-week course of sulfamethoxazole/trimethoprim. One month following the initial visit, the horse presented with a draining tract of the ventral mandible and a CT was performed to recheck the site. As can be seen in Figure 1, there were two regions of hyperattenuating tracts following the margins. The region shown is the lateral aspect of the mandible, while the drainage originated from a tract on the ventromedial aspect. Figure 2 shows the expanded view of the mandible along with the 3D reconstruction to better demonstrate the course of the fracture line and the associated hyperattenuating tracts. These were presumed to be osteomyelitis and sequestra of the fracture lines. An approach was made to debride the lateral aspect through the original buccotomy site but exposure of the entire fracture line could not be achieved; the ventromedial mandible was also debrided, and the horse was placed on a two-week course of sulfamethoxazole/trimethoprim and discharged. Culture showed growth of Trueperella pyogenes and Arcanobacterium haemolyticum from a mixed culture. Follow-up was performed by the referring veterinarian with no major issues reported. Seven months after debridement, the owner reported no discharge, the tract had healed, and the horse was eating well, maintaining weight and able to perform as a carriage and Ghodi horse for wedding ceremonies. A biopsy of the area was not performed but, will be considered to further explore this phenomenon in future cases.

In this case, the fracture was unilateral, but instability may have been present given the compromise associated with the approach and the repulsion of the tooth in that region. However, the horse quickly resumed its normal appetite following the procedure. The alveolus was packed with rapid-set vinylpolysiloxane (Splash!®, DenMat Holdings, LLC) impression material around the region of the fracture to reduce contamination of the alveolus and fracture line. Some contamination with saliva was expected. It should also be noted that the tooth was extracted due to an apical infection and draining tract, which would have also contributed to contamination of the fracture line. The blood supply in this region may have been affected, but this seems unlikely given the approach lateral to the tooth roots and the expected inferior alveolar artery position from the aforementioned anatomical study (Iacopetti et al., 2015). The inferior alveolar canal could not be easily identified in this region on the patient's CT. It is important to consider the level of disruption of soft tissues not only in the original series, but also in this case, as the surgical approach disrupted the soft tissues. Curiously, however, the fracture line sequestration presented rostral and ventral to the initial surgical site. Using CT to follow more of these cases is warranted, especially with the growing availability of standing sedated head CT options in horses.

One final highlight of this series is the outcome of the third case, which experienced prolonged drainage and resolved after 25 months of treatment (Aßmann et al., 2025). Many owners are not committed to this duration of treatment but may consider it if given the possibility of the mandible eventually healing. The other two cases were understandably euthanised, given the level of treatment afforded up to that point, but it is worth noting that resolution may have also eventually been achieved in these cases. Having an unknown timeline along with an open-ended estimate is a difficult reality for an owner but may be acceptable in some cases where these complications are observed.

There are no funders to report for this submission.

No conflict of interest has been declared.