Optimal PET/CT anatomic coverage for staging and diagnosis of head and neck cancer

Objectives PET/CT scanning for the initial staging and diagnosis (ISD) of head and neck cancers (HNc) presently involves torso scanning (TS) from the ear to the mid thigh region. However the clinical value of TS is unclear. We re-evaluated scans performed for these indications to determine the additional clinical value of scanning inferior to the arch of the aorta.

Methods From 2004-2007 we identified 627 consecutive scans having an initial TS for HNc. 599 had HNc without a previously diagnosed concurrent malignancy. 28 were excluded for indications other than that of ISD. Scan re-evaluations were limited and extended only to the level of the aortic arch inferiorly. The patients records, biopsy/pathology reports, other imaging and clinical follow up of at least 6 months were used to determine outcome and validate the scan findings.

Results Of the included 599 scans, 66 (11%) had suspicious findings below the arch on the TS necessitating further evaluation and workup. 7 of these were confirmed as false positives by negative biopsy follow up. Excluding cases where M1 disease was already identified on the LS, only 11 patients (1.8%) had M1 disease based on the TS which could have affected staging and that were not seen on the LS.

Conclusions Optimal PET/CT coverage for ISD of HNc needs only to extend inferiorly to the level of the aortic arch as is performed for standard radiological (CT and MRI) scanning. The extended TS scan, provides negligible additional clinical value in our experience (1.8%). The LS reduces scan time, radiation dose and the cost of evaluating false positive findings without negatively impacting patient care.