Introduction: Accurate definition of the presence and extent of large nerve perineural spread (PNS) is a vital component in planning appropriate surgery and radiotherapy for head and neck cancers. Our research aimed to define the sensitivity and specificity of 3T MRI in detecting the presence and extent of large nerve PNS, compared with histologic evaluation.
Methods: Retrospective review of surgically proven cases of large nerve PNS in patients with preoperative 3T MRI performed as high resolution neurogram.
Results: 3T MRI had a sensitivity of 95% and a specificity of 84%, detecting PNS in 36 of 38 nerves and correctly identifying uninvolved nerves in 16 of 19 cases. It correctly identified the zonal extent of spread in 32 of 36 cases (89%), underestimating the extent in three cases and overestimating the extent in one case.
Conclusion: Targeted 3T MRI is highly accurate in defining the presence and extent of large nerve PNS in head and neck cancers. However, there is still a tendency to undercall the zonal extent due to microscopic, radiologically occult involvement. Superficial large nerve involvement also remains a difficult area of detection for radiologists and should be included as a 'check area' for review. Further research is required to define the role radiation-induced neuritis plays in the presence of false-positive PNS on MRI.
Keywords: 3T MRI; diagnosis; head and neck cancer; perineural spread.
© 2015 The Royal Australian and New Zealand College of Radiologists.