Reducing the unknowns: A systematic review & meta-analysis of the effectiveness of trans-oral surgical techniques in identifying head and neck primary cancer in carcinoma unknown primary

The use of transoral robotic surgery (TORS), transoral laser microsurgery (TLM) and more recently reported transoral endoscopic electrocautery (TOEC) in identifying the primary cancer in head and neck Carcinoma Unknown Primary (CUP) patients have gained popularity. This review aims to assess the effectiveness of TORS, TLM and TOEC. A systematic review and meta-analysis was carried out. EMBASE, MEDLINE and CINAHL databases were searched from inception to September 2020. All primary studies were considered for inclusion. Primary outcome measure was detection rates of primary cancer of the different techniques. Secondary outcome measures were complications and length of hospital stay. 289 studies were identified of which 30 met the inclusion criteria. The primary cancer was identified in 567 /777 patients (pooled results was 64% (95% CI 54-73). The primary identification rates were 45% and 32% in lingual (n = 273) and palatine tonsillectomy (n = 118) respectively. The primary cancer identification rates by surgical techniques are: TORS was 60% (95% CI 49-70), TLM was 80% (95% CI 0.58, 1.01), TOEC was 41% (95% CI 0.05, 0.76). 529/777 (68%) tumours were Human Papilloma Virus (HPV) related. The pooled data of studies that reported on detection rates relating to HPV status were 178/216 (82%) for HPV +ve and 7/59 (12%) for HPV -ve tumours. Coefficient of variation results suggest heterogenous data for TORS and TLM. The commonest complication was haemorrhage (5.3%). The length of reported hospital stay ranged from 1.4 to 7 days. This is the largest systematic review in the subject. The quality of studies and heterogeneity of data limit conclusive findings. Lingual tonsillectomy is an effective procedure in CUP work up. Further larger, multicentre, prospective studies of PET CT negative CUP patients is needed to draw conclusive results.