Bisphosphonate-related osteonecrosis of the jaw (BRONJ) has recently emerged as a serious side effect among patients with prominent destructive bone alterations, including osteonecrosis, treated with bisphosphonates. ^,,,, It has been reported that not only bisphosphonates, but also other bone antiresorptive drugs (for example, denosumab) or antiangiogenic agents (for example, bevacizumab), can cause destructive changes, including osteonecrosis of the jaw. ^,,,,,,,,,,, Together with such drugs, their diseases are referred to as medication-related osteonecrosis of the jaw (MRONJ).

The number of publications on MRONJ, including BRONJ, has increased in recent years, ^{,,,,,,,,,,,,,,,,} as have reports on imaging findings in cases of MRONJ. ^,,,,,,,,,,

Cemento-osseous dysplasia (COD) is a fibro-osseous lesion that represents a reactive process in which normal bone is replaced by a poorly cellularised cementum-like material and cellular fibrous connective tissue. ^,,, Regarding the particular pathological conditions of COD, the jaw is prone to infection in affected patients, and osteomyelitis tends to occur in the mandible and maxilla. ^, We hypothesised that MRONJ in patients with COD would show particular imaging findings.

To our knowledge, no studies have attempted to elucidate the characteristics of imaging findings in patients with MRONJ and COD. In the present study therefore, we retrospectively analysed imaging findings in MRONJ among patients with COD from modalities including multidetector row computed tomography (MDCT), magnetic resonance imaging (MRI), panoramic radiographs, and dental radiographs.

We retrospectively analysed 52 patients with non-medication-related osteomyelitis (NMRO) (8 males, 44 females; mean (SD) age 76.9 (8.2) years, range 56–89 years), 13 with NMRO and COD (1 male, 12 females; mean (SD) age 77.1 (10.9) years, range 51–90 years), 130 with MRONJ (34 males, 96 females; mean (SD) age 78.6 (9.5) years, range 46–96 years), and 29 with MRONJ and COD (7 males, 22 females; mean (SD) age 80.8 (9.5) years, range 60–99 years) who had visited the Division of Oral and Maxillofacial Radiology at Kyushu Dental University Hospital between 2009 and 2021.

The inclusion criteria for patients with MRONJ were those meeting definitions of MRONJ by the Japanese Allied Committee on Osteonecrosis of the Jaw and the American Association of Oral and Maxillofacial Surgeons, ^, and for whom a medical and dental history and CT and panoramic radiographs were available. In some cases, dental radiographs and MRI scans were also acquired. Since this study was retrospective, we added documentation stating that we analysed the existing image data for each case to the greatest extent possible.

The inclusion criteria for patients with COD were the presence of periapical, focal, or florid dysplasia in the jaw based on imaging findings according to Cavalcanti et al. Among 42 patients with COD in the present study, 16 were evaluated microscopically on pathological examination after surgery.

Patients with suspected pathological lesions other than osteomyelitis and COD, those with a history of craniofacial malformation or syndrome, or a history of trauma or surgery, were excluded from the analysis. Informed consent was obtained from all patients for the imaging examinations, but individual consent for the retrospective analysis was waived. The study was approved by the Institutional Review Board of Kyushu Dental University (No. 24-28). The distributions of patients in the present study with NMRO, NMRO and COD, MRONJ, and MRONJ and COD, are shown in Table 1 .

MDCT was performed using an Activion 16 CT scanner (Toshiba Co Ltd), and multiplanar reconstruction images were taken at 3.0 mm from the axial plane with 1 mm-thick contiguous sections at the jaw level. All images were obtained with soft tissue and bone-target windows using standard algorithms. For the CT scans, the following parameters were used: 120 kV, 200 mA, 1.0 second per tube rotation, slice thickness of 0.5 mm, field of view of 180 × 180 mm, helical pitch of 0.688, and 5.5 mm/rotation table speed.

Panoramic radiographs were acquired using a panoramic AUTO-1000 EX system (Asahi Roentgen Ind Co Ltd). All images were taken in the incisive occlusion position with the head immobilised by cushioned supports and the Frankfort horizontal plane parallel to the ground. To visualise the maxilla and/or mandibles, all images were acquired using a 1.5-T full-body MR system (Excelart Vantage powered by Atlas, Toshiba) with a head coil (Atlas Head SPEEDER, Canon). T1-weighted, short-tau inversion recovery images were acquired for each patient.

Baba et al, with modifications, compared the prevalence of six imaging findings, each considered inflammatory and effective for evaluating MRONJ on CT, across groups: ^, (1) presence of sequester ( Fig. 1A Supplementary Data ); (2) presence of periosteal reaction ( Fig. 1B Supplementary Data ); (3) presence of pathological fracture ( Fig. 1C Supplementary Data ); (4) presence of spread into soft tissue ( Fig. 1D Supplementary Data ); (5) presence of cortical perforations ( Fig. 1E Supplementary Data ); and (6) presence of separation of cortical and cancellous bone ( Fig. 1F Supplementary Data ).

The images, including the six types of findings mentioned above, were assessed by a single specialist in oral and maxillofacial surgery, certified by the Japanese Society for Oral and Maxillofacial Surgery (MH), who had 20 years of experience.

All statistical analyses were performed using IBM SPSS Statistics for Windows version 23 (IBM Corp). All values of p < 0.05 were considered statistically significant. Categorical variables were compared using the chi squared test. The significance of differences between continuous independent variables was evaluated using a non-parametric test (Mann–Whitney U test) if the data did not follow a normal distribution. Relations between categorical variables were assessed using Pearson’s correlation coefficient. Correlations from the Pearson analysis were classified into the following five grades: very weak (0.00 to ±0.20); weak (±0.20 to ±0.40); moderate (±0.40 to ±0.70); strong (±0.70 to ±0.90); and very strong (±0.90 to ±1.00).