A waiver of informed consent was granted by our institution's committee on human research to conduct this retrospective review. We identified patients with Stage III-IV oropharyngeal carcinoma who underwent chemoradiation therapy between April 2000 and September 2004 and underwent longitudinal follow-up care at the University of California, San Francisco (UCSF) Comprehensive Cancer Center. Patients who received clinical and radiographic follow-up at outside institutions were excluded. Prior to treatment, all patients underwent complete history and physical examination, panendoscopy and biopsy, CT or MRI of the head and neck region, laboratory studies and chest radiography, dental evaluation, and nutritional, speech and swallowing evaluation.
Recommended post treatment surveillance included clinical history and physical examination at a frequency of every 1-2 months for the first year, 2-3 months for the second year, 3-4 months for the third and fourth years, and 6 months for the fifth year. Chest radiograph and thyroid function studies were obtained yearly. All patients underwent a post treatment baseline MRI approximately two months after completion of chemoradiation therapy. Surveillance MRIs were performed approximately every 3-4 months in the first year post treatment and approximately every 6 months in subsequent years. All patients included in this study adhered closely to this protocol. MRI sequences included axial, coronal, and sagittal pregadolinium T1; fat-saturated axial fast spin echo T2; and fat-saturated axial and coronal postgadolinium T1. Adjunctive studies, such as PET/CT scans, were performed as clinical evaluation dictated, but were not specifically analyzed in this study. References to radiographic investigations imply MRI scans for the purposes of this study.
Patient charts were retrospectively reviewed for disease staging, treatment and surveillance information, and recurrence. For patients who experienced locoregional failure, charts were reviewed for description of recurrence, imaging and physical exam findings, and patient symptomatology at time of diagnosis. Time until treatment failure was defined as the difference from chemoradiation completion until radiographic or definitive clinical evidence of recurrence, whichever came first. Radiographic failure was defined by language within the finalized MRI report specifically implicating the presence of tumor or suggesting tumor presence by interval growth or concerning radiographic features such as signal characteristics. Additionally, for patients with clinical symptoms or questionable findings on MRI scans, review at tumor board by the multidisciplinary head and neck team routinely occurred, including the presence of an experienced head and neck radiologist. These definitions were then used to calculate MRI diagnostic accuracy measurements, including sensitivity, specificity, positive predictive value, and negative predictive value. Disease persistence versus recurrence was defined as disease identified within or after a 3-month-time frame post treatment, respectively. Invasive procedures performed within 4 months of chemoradiation completion, such as neck dissection, were considered to be an extension of primary treatment rather than investigations into disease recurrence.
Expense estimates were obtained from our institution's Department of Radiology. We based our estimates from current charge data for an MRI neck with gadolinium, which was $4,942 as of April, 2010.