Osteoradionecrosis of the jaw is a common radiation-induced complication that may be observed in oral cancer patients. Several classifications and staging systems have been proposed for osteoradionecrosis of the mandible based on clinical symptoms, radiological findings, and/or the response to diverse treatments. However, none has been universally accepted because of their individual deficiencies. The aim of this study was to introduce a new clinical classification that can be applied to the treatment of osteoradionecrosis in an easier and more acceptable way, through a retrospective analysis of patients with osteoradionecrosis of the mandible. A review was conducted of 99 patients diagnosed with osteoradionecrosis of the mandible in the study institution between 2000 and 2013. A novel classification was established on the basis of bone necrosis and soft tissue defects. A new staging system with four different stages (stage 0, stage I, stage II, and stage III) is proposed. We believe that this new classification and staging system is easier and more acceptable for clinical evaluation than previous ones.
Radiotherapy has been an indispensable primary treatment or adjuvant treatment for malignant oral and maxillofacial tumours. Although it plays an ameliorative role in local control of the tumour and in improving the survival rate, late adverse effects such as osteoradionecrosis of the jaw (ORNJ) cannot always be avoided and may have a tremendous psychological and physical impact on the patient.
Regaud first reported ORNJ in 1922, and a great deal of attention has since been paid to this clinical challenge. In 1983, Marx defined ORNJ as “an area greater than 1 cm of exposed bone in a field of irradiation that had failed to show any evidence of healing for at least six months”. Since then, several different theories and definitions relating to ORNJ have been proposed by various researchers.
Despite the lack of a standard and unified definition, ORNJ is characterized by different extents of bone necrosis combined with soft tissue defects. These varying degrees of bone destruction make it essential that a classification and staging system be devised to help provide a strategy for the clinical treatment of osteoradionecrosis of the mandible (ORNM).
In 1983, Marx reported the Wilford Hall hyperbaric oxygen (HBO) osteoradionecrosis protocol, consisting of three stages based on the response to HBO therapy, which was applied due to the aetiological hypothesis of ‘hypoxia, hypocellularity, and hypovascularity’. This classification was validated by its excellent therapeutic outcomes and wide utilization by subsequent investigators. Nevertheless, this basic grading system has its limitations, including: (1) it is not applicable in patients not receiving HBO, (2) it lacks a description of the pathophysiology of ORNM, and (3) the effect of HBO therapy is now in question. Slightly different to Marx, Tsai et al., using a classification of four stages, categorized patients who required HBO therapy as being stage III, while various alternative treatments were needed for the other stages. This classification is also not ideal for clinical evaluation.
There are many controversies regarding the aetiology and pathophysiology of ORNJ. In 2014, Lyons et al. proposed a new classification based on the therapeutic effects of the anti-fibrotic drug pentoxifylline, which was prescribed to patients as a result of the fibro-atrophic theory that was elucidated by Delanian and Lefaix. This theory has a close association with anti-fibrosis and may lose its grading efficiency when the anti-fibrotic factor is removed. Epstein et al. were the only authors to elaborate their classification according to the status of ORNM progression. This was a completely new idea that was quite different from the other theories and provided an instructional way to determine when the management of ORNM should be performed. Clinically, it is difficult to judge how ORNM is progressing; the essential nature of the disease has been ignored. Some investigators established their staging systems based upon the extent of bone destruction described in the vertical direction, such as inferior mandible border involvement, inferior alveolar nerve canal involvement, and local or diffuse involvement. These methods provided quantitative analysis for the necrotic bone, but were unsuitable for that originating from the central bone.
From this review, it can be seen that these proposed classifications were available and deemed suitable for application in the investigators’ own institutions, although their limitations are evident. The aim of this study was to establish a new clinical classification that divides ORNM into different stages of escalating degrees, and that can be applied to the treatment of ORNM in an easier and more acceptable way.
Patients and methods
The files of patients diagnosed with ORNM in the department of oral maxillofacial and head and neck oncology of the study institution during the period 2000–2013 were collected retrospectively. In this study, ORNM was defined as radiological evidence of bone necrosis within the irradiated region of the mandible, as proposed by Store and Boysen. To be included in this classification, patients had to fulfil the following criteria: (1) diagnosed with osteoradionecrosis of the mandible without primary tumour recurrence; (2) received single, prior radiation in the study institution; (3) had no severe systemic comorbidities (e.g., coronary heart disease, mental illness, or cachexia); (4) received subsequent treatments in the study institution (eliminating the interference of other organizations); and (5) had complete treatment and follow-up information.
Patients with different malignant tumours of the head and neck region were treated with radiotherapy in the study institution. The duration of follow-up covered the period from the completion of radiotherapy to the day of analysis, divided into time intervals of 3 months. As patients developed ORNM during the research period, the specific sites affected were recorded. The clinical manifestations were also summarized, including spontaneous pain (mild: intermittent pain that patients could endure; moderate: persistent pain that patients could bear; severe: pain that could not be tolerated), bone exposure, skin fistula, limitation in mouth opening, numbness, and pathological fracture. The maximal diameter of the bone lesion and the status of the pathological fracture were assessed by computed tomography (CT), i.e., the maximal diameter was determined by the maximum length acquired from the CT images. The buccolingual and anteroposterior diameters were measured on the transverse section, and the vertical diameter was measured on the coronal plane. Through sliding the bone window in the CT software, it was possible to determine the maximal length in different directions; the maximal length was then selected.
The treatment records were carefully collected, including information on conservative therapy (e.g., antibiotics, oral hygiene control, and immunity enhancement), sequestrectomy, and segmental mandibulectomy with or without vascularized tissue reconstruction. The treatment outcomes were recorded as ‘resolved’, ‘improved’, ‘stable’, or ‘progressed’. ‘Resolved’ relates to the status in which the patient is asymptomatic without progression and has relatively good function. ‘Improved’ relates to the status in which the patient has relief from their symptoms and where the necrotic lesion is noted on radiography to be decreased. ‘Stable’ signifies that the disease has neither progressed nor improved. ‘Progressed’ indicates deterioration in the condition of the patient.
Based on these clinical features of ORNM and previous classification schemes, a new classification system was developed that classifies all patients into different groups demonstrating varying disease severity. The clinical features, corresponding therapies, and outcomes of patients in the distinct stages are described below.
Demographic and clinical characteristics
A total of 120 patients (maxilla n = 11, mandible n = 109) suffering from ORNJ were registered in the electronic database of the study institution. Ten ORNM patients were excluded as they did not fulfil the study inclusion criteria, hence 99 patients with ORNM (male n = 72, female n = 27) were ultimately analysed in this study. The age at diagnosis of these 99 patients ranged from 26 to 78 years, with a mean and median age of 54.5 and 55 years, respectively. The primary malignant tumours are listed in Table 1 .
|Locations||No. of cases ( N = 99)||Percentage (%)|
|Oral cavity malignant tumours|
|Floor of the mouth||11||11.1|
|Pharynx malignant tumours|
|Salivary gland malignant tumours|
The X-ray doses resulting in ORNM were within the range of 60 to 74 Gy, with a mean and median dose of 65.5 and 68 Gy, respectively. All patients underwent external beam irradiation with Cobalt-60 irradiation (25.8%) or with a linear accelerator (74.2%). Most of the ORNM ( n = 55, 55.6%) had occurred within the first 12 months of the post-irradiation period; 27 patients (27.3%) were diagnosed at 12–60 months post-irradiation and 17 patients (17.2%) at 5 years after receiving the radiation therapy ( Table 2 ). ORNM in the mandibular body without ramus involvement was observed in 55 patients (55.6%), involvement of the ramus alone occurred in seven patients (7.1%), and involvement of the mandibular body combined with the ramus was observed in 37 patients (37.4%).
|Time (months)||No. of cases ( N = 99)||Percentage (%)|
Radiographic examination was an indispensable approach to evaluating the mineralization changes and even the formation of necrotic bone or further fracture of the mandible. No distinct changes were found on radiography in eight patients (8.1%) who complained of spontaneous bony pain (mild n = 5, moderate n = 3). Ninety-one patients (91.9%) were found to have necrotic bone – 14 (14.1%) with a maximal diameter <2 cm and 65 (65.7%) with a maximal diameter >2 cm; pathological fracture was observed in 12 patients (12.1%) ( Table 3 ).
|Variables||No. of cases, n (%)||Stage 0||Stage I||Stage II||Stage III|
|Bone exposure||65 (65.7)||2||8||55||–|
|Pathological fracture||12 (12.2)||0||0||0||12|
|Skin fistula||55 (55.6)||0||2||43||10|
|No pain||22 (22.2)||0||6||16||0|
|Mouth opening limitation|
|No distinct change||8 (8.1)||8||0||0||0|
|Maximal diameter <2 cm||14 (14.1)||0||14||0||0|
|Maximal diameter >2 cm||65 (65.7)||0||0||65||0|
|Pathological fracture||12 (12.1)||0||0||0||12|
Bone exposure and pathological fracture were typical features of ORNM and these symptoms were found in 65 (65.7%) and 12 patients (12.1%), respectively. In this study, bone exposure was not defined as a condition in which fractured bone was exposed to the oral cavity; this condition was instead assigned to the category of pathological fracture. Radiographic changes seemed not to be associated with the status of bone exposure in the oral cavity. Two patients without distinct changes on radiography had exposed bone; the vitality of such bone should be evaluated further by bone dynamic testing. In contrast, 16 patients with evident bone alterations (except for bone fracture) had an intact mucosa covering the underlying necrotic bone. More surprisingly, one patient with a pathological fracture had an intact covering mucosa.
Fifty-five patients (55.6%) developed skin fistulas, which indicated that a connection existed between the oral cavity and the maxillofacial region. This soft tissue disease played an important role in the selection of various reconstructive methods. There was also no correlation between external skin fistulas and bone exposure, as evidenced by six patients with necrotic bone that was not exposed to the oral cavity who were discovered to have skin fistulas. Seventy-seven patients (77.8%) suffered from spontaneous pain (mild n = 26, moderate n = 31, severe n = 20). The majority of the patients ( n = 59, 59.6%) had a limitation of mouth opening, to varying extents ( Table 3 ).
New classification and staging system
A summary of the previous classifications and staging systems for ORNM is shown in Table 4 . The proposed new classification and staging system, based upon the retrospective study of 99 ORNM patients and the review of the previous classification methods, is shown in Tables 5 and 6 ; a specific explanation of each stage is given below.
B0S0: No evident signs or only osteolytic images on radiography; however, the patient suffers from typical ORNM-related symptoms (bone exposure or pain).
B1S0: The maximal diameter of the lesion on radiography is <2 cm and there is no mucosa or skin defect.
B1S1: The maximal diameter of the lesion on radiography is <2 cm and there is an intraoral mucosa defect or external skin fistula alone.
B1S2: The maximal diameter of the lesion on radiography is <2 cm and there is a through-and-through defect ( Fig. 1 ).
B2S0: The maximal diameter of the lesion on radiography is >2 cm and there is no mucosa or skin defect ( Fig. 2 ).
B2S1: The maximal diameter of the lesion on radiography is >2 cm and there is an intraoral mucosa defect or external skin fistula alone.
B2S2: The maximal diameter of the lesion on radiography is >2 cm and there is a through-and-through defect ( Fig. 3 ).
B3S0: A pathological fracture is identified on radiography and there is no mucosa or skin defect ( Fig. 4 ).
B3S1: A pathological fracture is identified on radiography and there is an intraoral mucosa defect or external skin fistula alone.
B3S2: A pathological fracture is identified on radiography and there is a through-and-through defect ( Fig. 5 ).
|Investigators||Location||No. of patients||Grading systems|
|Marx 1983||Mandible||58||Grade I: 30 dives of HBO was used to attain mucosa recovery|
|Grade II: Grade I non-responders who needed a transoral alveolar sequestrectomy|
|Grade III: Grade II non-responders; a bone resection was needed|
|Grade IV: An additional 30 dives of HBO was given to patients who needed a bone graft|
|Coffin 1983||Mandible||22||Minor: Small sequesters that may separate spontaneously over several weeks|
|Major: Bone necrosis extending to the entire thickness of the jaw; pathological fracture sometimes present|
|Morton 1986||Mandible||39||Minor: Bone exposure with ulceration and a history of spontaneously resolving bony spicules|
|Moderate: Small sequester limited in nature and resolving spontaneously|
|Major: Large area of exposed bone and sequester; bone fracture and fistula|
|Epstein et al. 1987||Mandible||26||Stage I: Resolved/healed, with or without pathological fracture|
|Stage II: Chronic/persistent non-aggressive, with or without pathological fracture|
|Stage III: Active/progressive, with or without pathological fracture|
|Glanzmann and Grätz 1995||Mandible||189||Stage 1: Bone exposure without signs of infection and persisting for at least 3 months|
|Stage 2: Bone exposure with infection or sequester and without the signs of stage 3–5|
|Stage 3: Bone necrosis treated with mandibular resection with a satisfactory result|
|Stage 4: Bone necrosis with persisting problems despite mandibular resection|
|Stage 5: Death due to ORNJ|
|Clayman 1997||Mandible||Undefined||Type I: Presenting with bone lysis under intact gingiva or mucosa|
|Type II: Bone exposure with secondary contamination; an aggressive form|
|Store and Boysen 2000||Mandible||73||Stage 0: Mucosa defect only|
|Stage I: Radiological evidence of necrotic bone with intact mucosa|
|Stage II: Positive radiographic findings with denuded bone intraorally|
|Stage III: Exposure of the necrotic bone; skin fistula and infection|
|Schwartz and Kagan 2002||Mandible||Undefined||Stage I: Superficial involvement of the mandible only|
|Stage II: Localized involvement of the mandible, with or without soft tissue necrosis|
|Stage III: Diffuse involvement of the mandible, with or without soft tissue necrosis|
|Notani et al. 2003||Mandible||87||Stage I: ORN confined to alveolar bone|
|Stage II: ORN limited to the alveolar bone and/or above the level of the inferior alveolar canal|
|Stage III: ORN under the lower part of the inferior alveolar canal, with fistula or bone fracture|
|Tsai et al. 2013||Mandible||30||Stage I: Minimal bone exposure with conservative management only|
|Stage II: Minor debridement required|
|Stage III: HBO needed|
|Stage IV: Major surgery needed|
|Karagozoglu et al. 2014||Mandible||31||Stage 0: Bone exposure more than 1 month; no distinct changes on imaging|
|Stage I: Bone exposure with no distinct changes on imaging, with or without symptoms|
|Stage II: Bone exposure with distinct changes on imaging, with no involvement of the lower mandible border|
|Stage III: Necrotic bone involving the lower border of the mandible|
|Lyons et al. 2014||Mandible||85||Stage 1: <2.5 cm length of bone affected; asymptomatic|
|Stage 2: >2.5 cm length of bone affected; involving fracture or the inferior dental nerve|
|Stage 3: >2.5 cm length of bone affected; symptomatic, with no other features|
|Stage 4: >2.5 cm length of bone affected; bone fracture and involving inferior dental nerve or fistula|
|Bone necrosis (B)||B0:||No distinct changes or just osteolytic images on radiography, but patients suffering from typical ORNM-related symptoms (bone exposure or pain)|
|B1:||Maximal diameter of the lesion on radiography <2.0 cm|
|B2:||Maximal diameter of the lesion on radiography ≥2.0 cm|
|Soft tissue defect (S)||S0:||Mucosal and skin integrity|
|S1:||Intraoral mucosal defect or external skin fistula|
|S2:||Intraoral mucosal and external skin defect; through-and-through defect|
|Stage 0||B0S0||Conservative therapy|
|Stage I||B1S0, B1S1, B1S2||Sequestrectomy and primary closure|
|Stage II||B2S0, B2S1, B2S2||Marginal resection without reconstruction, or segmental resection combined with osteocutaneous flap reconstruction; an additional soft tissue flap should be used in patients with an orocutaneous fistula|
|Stage III||B3S0, B3S1, B3S2||Segmental resection combined with osteocutaneous flap reconstruction; an additional soft tissue flap should be used in patients with an orocutaneous fistula|