Predicting Outcomes Using Pre- and Post-treatment PET/CT in Locoregionally Advanced Non-oropharyngeal Head and Neck Cancer

Abstract

To determine if pre- and post-treatment positron emission tomography and computed tomography (PET/CT) is predictive of outcomes including overall survival (OS) and development of metastasis in non-oropharyngeal head and neck cancers. We gathered pre- and post-treatment PET/CT data on 47 patients with non-oropharyngeal head and neck cancer from 2005-2016 who were treated with definitive chemoradiation. Post-treatment PET/CT was obtained 3 months after the completion of treatment. We recorded the primary tumor size, dominant lymph node size, max SUV of both the primary tumor and the dominant node. Analysis was performed to determine the predictive utility of each parameter for OS and the development of metastasis. Forty-seven patients with non-oropharyngeal head and neck cancer underwent pre- and post-treatment PET/CT. The primary tumor sites were hypopharynx (9), larynx (24), nasopharyngeal (9), oral cavity (3), and paranasal sinus (2). Four patients were stage II, 17 were stage III, 18 were stage IVA, and 7 were stage IVB. For the primary tumor, the mean size was 3.1 cm (range, 1.0 – 7.5 cm) and mean max SUV was 21.85 (range, 5.9 – 48.1). The dominant node was used representing all nodal disease. For the dominant node, the mean size was 1.94 cm (0.7 cm – 5.8 cm) and mean max SUV was 12.6 (2.3 – 28.5). Patients were most commonly treated with simultaneous in-field boost SIB-IMRT technique with median dose 6750 cGy with concurrent weekly cisplatin. Mean follow-up time was 43.6 months. Of the 12 patients who experienced a treatment failure, 3 failed locally only, 4 failed distantly only, 1 failed locally and distantly, 1 failed distantly and had nodal failure, 2 had isolated nodal failure, and 1 failed synchronously at all three sites. Eighteen patients died during the follow-up period. On the pre-treatment PET/CT, only the max SUV of the dominant node (HR 1.06, p=0.02) was predictive of OS. The decrease from pre- to post-treatment of the nodal SUV (HR 1.06, p=0.03) was also predictive of OS. On post-treatment PET/CT, the SUV of the primary tumor (HR 1.13, p=0.03) and the size of the primary tumor (HR 3.05, p=0.02) were predictive for OS. No pre-treatment PET/CT data was predictive for the development of metastasis. Only the post-treatment size of the primary tumor (HR 18.3, p<0.01) was predictive for the development of metastasis. In non-oropharyngeal head and neck cancer, both pre-treatment and post-treatment PET/CT are valuable for predicting OS. Only post-treatment PET/CT data is useful for predicting the development of metastasis.