Introduction: Tonsillectomy is one of the commonest operations performed in the United Kingdom, the majority in children for either recurrent tonsillitis or obstructive sleep apnoea. In adults, the indications are similar but the proportion of patients undergoing surgery for the suspicion of malignancy is higher.

A number of studies have examined the malignancy rate within tonsil specimens, and demonstrated that the number of cancers picked up across adults and children in tonsils removed for recurrent tonsillitis is very low. The majority of tumours are found in tonsils where there was a high index of suspicion (rapidly enlarging/history of smoking/presence of neck metastases). In our department, tonsils are only sent for histological examination if there is anything to suggest a non-benign diagnosis, including asymmetry. To date, there have been no studies examining the rate of malignancy in tonsils removed purely for the reason of unequal size, and the proportion of patients undergoing surgery for this indication is unknown.

Objectives: To identify the proportion of patients undergoing tonsillectomy for asymmetrically sized tonsils, and to correlate the malignancy rate to other symptoms.

Methods: All patients undergoing tonsillectomy for any indication at University Hospital Birmingham between July 2008 and July 2015 were included. Case notes were examined and correlated with histological diagnoses.

Results: 1263 patients underwent tonsillectomy in the study period, of which 139 (11%) had specimens sent for histology: this subgroup was further analysed. The number of malignancies was 23 (16.5%), including squamous cell carcinoma (n=16, 11.5%) and lymphoma (n=7, 5%). Primary indications for surgery were recurrent quinsy/tonsillitis (38.8%), suspicion of malignancy based on appearance (24.6%), unknown primary (11.5%), asymmetry alone (11.5%), unilateral throat pain (8.6%), OSA (3.6%), known malignancy (0.7%) and abnormal imaging (0.7%). In 6 patients a decision was made to send the tonsils for analysis based on abnormal findings intra-operatively. Asymmetry was additionally a part indication for 49 patients (39.8%).

In the asymmetry-alone group, the average age was 39.4 years (range 18-61) and mean time from clinic visit to operation 38 days (range 8-168). 9 patients underwent cross-sectional imaging prior to surgery. No patients were found to have malignancy in this group (specificity=0%). In addition, no malignancies were detected in patients with a history of recurrent tonsillitis/quinsy and asymmetry (n=22).

In the group with asymmetrical tonsils and another worrying feature, there were 5 malignancies (2 with neck lumps and abnormal looking tonsil, 1 unknown primary, 1 clinically malignant tonsil and one associated with worsening unilateral throat pain), a specificity of 10.2%. All had received cross-sectional imaging prior to surgery.

Conclusions: No malignancies were found in patients presenting with either purely asymmetrical tonsils, or with recurrent tonsillitis/quinsy and asymmetry subsequently noted by the clinician. It would be reasonable, therefore, to limit those undergoing tonsillectomy for asymmetry to those with an additional worrying feature (abnormal surface, unilateral throat pain, rate of enlargement, B symptoms, bleeding), or for those with suspicious imaging. Further studies will continue to assess the role of observation versus surgery in this patient group.