The 14 eligible studies included a total of 1780 patients, of which 294 underwent ND as their single primary treatment for HNSCCUP. There were on average 17.5 (5-63) (median(range)) cases per study. Although all studies met the initial eligibility criteria and presented outcome data relevant to the primary and / or secondary aims, there was significant heterogeneity in reporting parameters, precluding any meta-analysis. There was a male preponderance, ranging from 76-88% across the studies, with an average age of 55-65 years. Four (28.6%) of the studies reported on p16 status, with one (7.1%) also reporting on high-risk HPV status by in situ hybridisation. In these studies, data were incomplete, with anywhere between 19.1% and 86.7% of included patients having status recorded.

The studies were evaluated for quality of diagnostic investigations compared to perceived ‘gold standards’ (FDG PET-CT, MRI, panendoscopy, tonsillectomy, tongue base mucosectomy). No single study included all investigations prior to treatment. PET-CT was performed on all patients in two studies (53,58) and on a proportion (20.0-58.6%) of patients in a further eight studies (50,54,56,57,59–62). Three studies (51,55,63) made no mention of cross-sectional imaging being performed, with the remainder confirming all patients underwent either CT or MRI scan prior to treatment. One study (51) didn’t state if panendoscopy was performed and in two further studies only a proportion (45.1-64.0%) underwent panendoscopy (56,63). Tonsillectomy was undertaken for all patients in nine of the studies; a proportion of patients (5.8-41.7%) in three of the studies (50,56,63), and was not stated in two studies (51,55). Only one study reported on tongue base mucosectomy, stating that all patients had this procedure (52).

All the studies reported data on the nodal status (AJCC/UICC TNM7) of patients treated; however only half provided discernible data on nodal status for patients undergoing ND alone, totalling 132 of the 294 patients (44.9%). Of these, the breakdown of nodal status was: N1 (n= 35,26.5%), N2a (n= 28, 21.2%), N2b (n =58, 43.9%), N2c (n=4, 3.0%), and N3 (n=7, 5.3%).

Follow up duration was reported in 10 studies (n=800), but none specified follow up timeframes for ND only patients, median follow up duration ranged from 31.1 to 83.5 months, 6 studies reported a lower end range of follow up <12 months post treatment, which may impact interpretation of this data. Data for survival, primary emergence, recurrence, or metastasis was not stratified by p16 status in any of the studies.