Deferring planned neck dissection following chemoradiation for stage IV head and neck cancer: the utility of PET-CT.

OBJECTIVE

To determine whether combined positron emission tomography and computed tomography (PET-CT) may be of value in deferring planned neck dissections for patients with advanced head and neck squamous cell carcinoma (HNSCC).

STUDY DESIGN

Observational study of patients with de novo cervical > or =N2 regional spread of HNSCC in a tertiary care academic medical center.

METHODS

Forty-three patients were identified who underwent post-treatment PET-CT within 6 months of completing neoadjuvant chemotherapy combined with radiation therapy (CRT). The PET-CT was "positive" if the radiologist recommended tissue sampling or resection of cervical lymph nodes, or if there was progressive neck disease in the setting of distant metastatic disease. Patients who had positive PET-CT underwent confirmatory biopsy given clinical suspicion for regional cervical metastasis without distant disease. Patients with negative PET-CT were followed clinically and radiographically for a minimum of 5 months (median 18.1 months) after CRT.

RESULTS

Ten (22%) of the 43 post-treatment PET-CT studies were positive. Seven of the 10 PET-CT scans (70% of positives) were true-positive given histologically-confirmed residual viable tumor or progressive disease including disease in the neck. The 3 remaining studies (30% of positives) were false-positive PET-CT results, given resolution of fluorodeoxyglucose (FDG) avidity on subsequent imaging or tissue sampling demonstrating absence of viable tumor cells. Of the 33 patients with negative PET-CTs in the neck, 1 patient had absence of FDG-avidity in the setting of malignant disease in the neck (3% false negatives); otherwise, patients with an initially negative PET-CT scan had no recurrences during the study (97% true negatives). This corresponds to a sensitivity of 87.5% (7/8), a specificity of 91% (32/35), a positive predictive value of 70% (7/10), a negative predictive value of 97% (32/33), and accuracy of 91% (39/43) for PET-CT scans in the detection of cervical metastatic disease after CRT. Overall, 37 (86%) of 43 patients were spared neck dissection using this technology without evidence of recurrent disease in the neck at extended follow-up.

CONCLUSIONS

Our results suggest that planned neck dissection after CRT for HNSCC may be deferred in favor of serial PET-CT imaging, and that sampling of areas of suspicious FDG-avid uptake can be rationally considered prior to therapeutic neck dissection. These data also suggest that negative PET-CT scans are highly reliable for the absence of residual cervical nodal disease.