Alloplastic TJR has been widely used as a major method of TMJ reconstruction since material and design improvements in the 1990s.1 However, infection and HO are the main causes of prosthesis revision or replacement.10 There were few reports on the TMJ HO after alloplastic TJR compared to the one from orthopedics, especially after Wolford11 proposed using abdominal periumbilical fat to fill the dead space around the joint prosthesis in 1997. The incidence of HO has been significantly reduced than before.12 In this study, we used subcutaneous or abdominal periumbilical fat for TJR. CT follow-up showed that HO was happened only in the medial side of the prosthesis.
In 1993, Turlington and Durr proposed TMJ HO grading system8 according to Brooker’s THA HO classification.11 It is as follows: Grade 0: no bone islands visible; Grade 1: Islands of bone visible within soft tissue around joint; Grade 2: Periarticular bone formation; Grade 3: Apparent bony ankylosis. Grades 1, 2, and 3 were further classified as symptomatic (S) and asymptomatic (A). Symptomatic ossification includes severe pain, decreased interincisal opening (15 mm or less), closed locking of the jaw, or decreased lateral or protrusive movement. In this study, we referred the above classification and described the HO according to its location based on coronal CT reconstruction. Our results showed that the incidence of HO after alloplastic TJR was 47.5%. Among them, 56.3% of HO was bone islands in the soft tissue medial to the condylar prosthesis. 33.3% bone spurs were grown from the mandibular ramus. Only 6.3% of the HO came from the medial side of the fossa and 4.2% developed ankylosis.
The incidence of HO among TMJ ankylosis, osteoarthritis and tumors had no significant difference. By measuring postoperative TJR space, we found that type IV HO had significantly reduced space with an average of 9.1mm than the other three types, which is a risk factor for ankylosis. Studies have shown that when the gap between bone stumps was less than 10mm, ankylosis was more likely to occur.13 When the bone defect is larger than critical-sized defect (CSD), osseous connection will not formed. Animal experiments on dogs with similar mandibular size as human showed that CSD is about 15mm.14,15 At present, there is no requirement for the minimum TJR space when implanting the prosthesis. Although the position of condylectomy is suggested to be at the level of sigmoid notch with removal of coronoid process, for patients with short mandibular ramus, sacrificing a certain joint space to provide sufficient bone support for the mandibular prosthesis may increase the risk of ankylosis. Therefore, it is recommended to use a customized prosthesis with mandibular body extension instead of a standard prosthesis which only fixes the mandibular ramus.
We also analyzed the relevance between HO type and clinical signs and symptoms. Except for type IV HO, which was prone to cause ankylosis and requires revision or replacement of the prosthesis, the clinical symptoms (pain and limited mouth opening) of the other 3 types of HO were mild. This was similar to the report after THA, although the incidence of small-volume HO can be up to 50%, only 10–20% of the patients have significant discomfort due to the severely affected joint mobility.16,17 In addition, we found that pain scores were significantly higher in HO patients than non-HO patients both before and after operation. High pain scores may reflect local inflammation around the joint which may affect bone metabolism and lead to the occurrence of HO after surgery.18–20 HO is also a major cause of postoperative pain. So it is important to prevent HO after TJR surgery.
Studies on the etiology of HO have shown that surgical trauma can cause inflammation and activate mesenchymal stem cells in tissues, thus differentiating into osteogenesis.21–23 In addition, the tension of masticatory muscle can also lead to the bone formation.24,25 Tendons and ligaments may ossified,26 and disc ossification was also reported after operation.27In this study, HO occurred on the medial side of the joint, with bone islands possibly derived from the external pterygoid muscle or the articular disc, as well as bone spurs growing from the ramus stumps after condylectomy. The mechanism of HO formation after alloplastic TJR remains to be further investigated.
Except periarticular autogenous fat grafting to prevent HO, postoperative radiotherapy and oral non-steroidal anti-inflammatory drugs (NSAIDs) such as indomethacin, celecoxib and bisphosphonates are also reported effective.17 In 1993, Durr et al.8 found that early postoperative administration of 10 Gy radiation five times a day could prevent 67% of HO after TJR with a history of ankylosis. Jensen et al.28 also demonstrated that postoperative radiotherapy could prevent long term HO reformation in 50% of the TJR patients. NSAIDs is another method to prevent HO and relief pain by inhibiting the synthesis of inflammatory factor Prostaglandin E2 (PGE2).29 Bhatt et al.30 found that indomethacin was effective in the prevention of HO after recurrent ankylosis. Naylor et al.31 found that celecoxib significantly reduced the incidence of HO from 14.3–4.3% after THA. Ouyang et al.32 proved that celecoxib was effective in post-traumatic TMJ HO in animal models. These methods above can be prophylactic used in high-risk patients such as ankylosing spondylitis, hypertrophic osteoarthritis, and recurrent HO.
This study involved patients with more than 1 year follow-up. HO can be shown from CT scan 3 months after operation and matured without change around 6–12 months after operation. In the future, quantitative measurement of HO and long-term follow-up can be taken to observe HO development and the relationship with inflammation. The incidence of HO in the customized TMJ prosthesis with different materials will be studied and compared with the standard TJR.
In conclusion, HO happened in various degrees after alloplastic TJR with fat graft. Most of which have little impact on patients' mouth opening or quality of life. However, type Ⅳ HO is prone to cause ankylosis, which need surgical removal to improve MIO and pain relief. Sufficient TJR space may reduce the risk of ankylosis.