Other / Other / MRI

Osteoradionecrosis

Temporal bone osteoradionecrosis (TBORN) is a delayed complication of head-and-neck radiotherapy, typically presenting years or decades after treatment with exposed devitalized bone in the external auditory canal (EAC). Imaging is performed to characterize the extent of disease and exclude recurrent malignancy.
Look For First
  • Focal or circumferential cortical erosion of the EAC, particularly the anterior and inferior walls, as this is the most common osseous finding
  • Disorganized sclerosis mixed with fragmentation and sequestration confined within the radiation field, which favors ORN over recurrence
  • Full-thickness EAC cortical loss extending beyond the tympanic bone with mastoid air-cell destruction/coalescence indicating diffuse disease
Key Image Findings
  • On high-resolution CT with bone algorithm, EAC erosion appears as focal or circumferential cortical irregularity, usually of the anterior and inferior canal walls, and was present in 18 of 20 patients in the classic series.
  • Sclerosis on CT is disorganized and often mixed with areas of lysis and cortical disruption, which is significantly more common in ORN than in recurrent tumor.
  • Fragmentation and sequestration appear as irregular devitalized bone fragments within the EAC or adjacent temporal bone, supporting necrotic bone diagnosis.
  • Mastoid coalescence on CT shows loss of septa and trabeculae, marking diffuse disease and predicting poor response to conservative treatment.
  • Deep soft-tissue gas appears outside the expected aerated temporal-bone spaces or within phlegmon/collection and suggests advanced necrosis, fistula, or superimposed infection.
  • On contrast-enhanced CT, enhancing tissue around necrotic bone represents granulation or phlegmon and is more common in clinically advanced disease, but a discrete mass requires tumor exclusion.
  • MRI shows low or heterogeneous T1 marrow signal, variable T2/STIR hyperintensity, and patchy enhancement in irradiated temporal bone, with necrotic areas demonstrating reduced enhancement and viable margins showing intense enhancement.
  • Fistulae (canal-mastoid, cutaneous, TMJ-EAC, dural, or CSF) and TMJ erosion indicate diffuse disease; TMJ involvement may produce trismus or synovial otorrhea.
Differential Diagnosis
  • Recurrent head-and-neck carcinoma - presents with discrete measurable solid or cystic enhancing mass or nodular progressive soft tissue, whereas ORN characteristically shows sclerosis and fragmentation without a focal mass; biopsy is required when a progressive mass is present.
  • Necrotizing otitis externa/skull-base osteomyelitis - presents in elderly diabetic patients with severe otalgia and anteromedial infratemporal spread, whereas ORN requires prior irradiation and classically shows exposed EAC bone; the diagnoses overlap because infection can complicate ORN.
  • EAC cholesteatoma - shows a soft-tissue/keratin mass with focal inferior erosion and restricted diffusion on DWI, whereas ORN shows broader disorganized sclerosis and necrosis within the radiation field; both may cause sequestra.
  • Radiation-induced malignancy - presents as a new aggressive mass within the radiation field after long latency, whereas uncomplicated ORN usually lacks a dominant expansile mass; histology must be performed.
  • Central/atypical skull-base osteomyelitis - shows central clival involvement and nasopharyngeal inflammation with relatively little EAC disease, whereas ORN centers on the EAC and extends peripherally; search for mucosal breakdown and tumor recurrence.
  • Keratosis obturans - presents as a keratin plug with smooth generalized EAC widening often bilateral, whereas ORN shows focal invasive erosion and sequestration of exposed necrotic bone.
Discussion

TBORN is fundamentally a delayed radiation effect caused by progressive vascular injury, tissue hypoxia, hypocellularity, and fibrosis; the superficial bony EAC is particularly vulnerable because of thin skin coverage and limited blood supply.

The mean latency from radiotherapy to clinical presentation is approximately 10 years, but the broad range means that even remote radiation history remains relevant to diagnosis.

The critical diagnostic challenge is distinguishing TBORN from recurrent or radiation-induced malignancy; FDG-PET uptake overlaps substantially between the two entities and cannot independently resolve the distinction.

Secondary microbial colonization or true osteomyelitis is common and may drive clinical progression; superficial cultures may reflect colonization rather than deep bone infection, so deep cultures should be obtained when feasible.

The Ramsden classification links anatomic extent with expected behavior: localized disease is confined to the EAC/tympanic bone and may respond to conservative treatment, whereas diffuse disease extending beyond the tympanic bone indicates advanced disease with worse prognosis.

High-resolution CT is superior to MRI for depicting EAC erosion, sclerosis, fragmentation, and sequestration, whereas MRI is superior for marrow replacement, cranial-nerve involvement, and intracranial complications; both modalities are often complementary.

Reporting Pearls

Describe the location and extent of EAC erosion (anterior/inferior, partial or full-thickness), whether it extends beyond the tympanic bone (the key distinguishing feature of diffuse disease), and explicitly compare the presence or absence of a discrete measurable mass versus irregular sclerosis and fragmentation—this architectural distinction is more discriminating than FDG uptake for favoring ORN over recurrence.

Pitfalls
  • Attributing all post-treatment temporal-bone abnormality to radiation necrosis without recognizing that persistent granulation tissue or a new discrete mass may represent recurrent carcinoma, which requires biopsy for definitive diagnosis.
  • Relying on FDG-PET SUV values alone to distinguish ORN from recurrence, as individual SUV ranges overlap substantially; PET must be interpreted with CT morphology and clinical course.
  • Misinterpreting simple post-radiation mastoid effusion or Eustachian-tube dysfunction as ORN; smooth fluid without EAC erosion, sclerosis, fragmentation, or exposed bone does not establish TBORN.
  • Overlooking that cholesteatoma can coexist with ORN and that both entities cause sequestra and EAC erosion; DWI restriction favors keratin, but prior radiation field correlation and broader disorganized sclerosis favor TBORN.