The bacterial community and local lymphocyte response are markedly different in patients with recurrent tonsillitis...

Obstructive sleep apnea (OSA) is now a more common indication for tonsillectomy than recurrent tonsillitis (RT) [1,2]. Few studies have addressed possible differences in pathogenesis between these two conditions. Children with RT and OSA are often being treated in the community with multiple courses of antibiotics before surgery. Current understanding of the role of bacteria in disorders of the tonsils is mainly based on the culture of tonsil swabs. Swab cultures reflect only a very small fraction of the bacteria present on the mucosal surface and may not represent the bacterial communities within the tonsil crypts [3,4]. This study aimed to evaluate the local lymphocyte response and associations with bacterial community composition using molecular techniques of the tonsils removed from children for RT or OSA.

The palatine tonsils were removed by extracapsular dissection from 24 patients with age range one to ten years, 14 of whom had RT and 10 had OSA. The fixed tonsil tissues were evaluated for bacteria by Gram-staining and presence of connective tissue by safranin staining. B lymphocytes and T lymphocytes were also measured immunohistochemically. Finally, previously published bacterial community data for this cohort were reassessed for associations with RT and OSA, and with the observed lymphocyte patterns.

In tonsils from patients with RT, large micro-colonies of bacteria were observed in the tonsil crypts, and a large number of B and T lymphocytes were noted immediately adjacent to the tonsil crypt itself. In marked contrast, the tonsils from patients with OSA had no bacteria identified, and no significant skewing of lymphocytes based on site (such as follicles or crypts). We observed that the majority of lymphocytes surrounding the bacterial micro-colonies were B lymphocytes with a mean ratio of 109:55 (B lymphocytes: T lymphocytes). Bacterial community diversity was not different between the two cohorts; however, there were significant differences in bacterial community composition. Children with RT had a higher relative abundance of members from the genera Parvimonas, Prevotella, and Treponema. While children with OSA had a higher relative abundance of Haemophilus, and Capnocytophaga.

These results demonstrate significant differences in the local lymphocyte response and bacterial community composition in tonsil tissue between RT and OSA patients. It suggests that the response to antibiotics used in the treatment of these two conditions may be different. Furthermore, the presence of lymphocytes in RT within the tonsil crypt outside the tonsil epithelium is a unique observation of the location of these cells.