Lea Alhilali

Squamous cell carcinoma of the head and neck accounts for over 550,000 cases annually worldwide. Primary risk factors include tobacco use, alcohol consumption, and infection with human papillomavirus (HPV) (for oropharyngeal cancer) and Epstein-Barr virus (for nasopharyngeal cancer). Among the primary treatment modalities for this disease is chemoradiotherapy, either in isolation or following surgical resection. These treatments are not without their own morbidities and complications. Osteoradionecrosis (ORN), resulting from the damaging effects of ionizing radiation, can affect anywhere from 5–20% of patients undergoing radiation therapy. Although ORN can affect any bone, it is most often seen involving the mandible. Prompt diagnosis and treatment of this condition is key to assuring a desirable outcome, minimizing morbidity, and maintaining a high quality of life. Given the relatively younger age group affected by HPV-positive head and neck cancer, the importance of prompt and correct diagnosis of ORN, as well as aggressive treatment, will likely only increase in future years.

Unfortunately, both clinically and on imaging, ORN can often be difficult to differentiate from recurrent disease, resulting in a regrettable delay in diagnosis and treatment. With the increasing utilization of PET/CT, which in many ways has revolutionized oncologic imaging, we sought to determine if PET/CT could aid in the differentiation of ORN from recurrent tumor. Predictably ORN (which can often coexist with osteomyelitis) demonstrated an elevated standard uptake value (SUV) compared with background; unfortunately, recurrent tumor also demonstrated significantly elevated SUV values. Although, in general, recurrent tumor demonstrated both higher SUV_mean and SUV_max values compared with ORN, significant overlap existed between the two, which precluded reliable differentiation on a case-by-case basis.

Fortunately, we found several findings on conventional CT imaging that can aid the interpreting radiologist in confidently differentiating ORN from recurrent tumor. The presence of a definable cystic or solid mass in patients was strongly suggestive of tumor recurrence, while the presence of bony sclerosis occurred only in patients with ORN. Intraosseous gas was seen exclusively in patients with ORN, but it was a relatively rare finding even in this subgroup of patients, rendering it less valuable in daily clinical practice. While ORN demonstrated less aggressive bony destruction than tumor recurrence, similar to PET/CT SUV values, there was significant overlap between the degree of bony loss between ORN and tumor recurrence, making it difficult to confidently use this finding for an individual case.

Given the significant variability in PET/CT protocols across institutions at this time, our findings highlight the importance of performing diagnostic-quality CTs (as opposed to low-dose attenuation correction CT) in conjunction with the PET scan, as the findings on conventional CT are more useful in distinguishing ORN from tumor recurrence than PET findings. Our study also shows the need for neuroradiologist interpretation of at least the CT portion of the study (if not the entire study), and may warn against becoming overly reliant on the FDG uptake on the PET portion of the study alone.

Read this article at AJNR.org …