Infected Rathke cleft cysts: distinguishing factors and factors predicting recurrence

Abstract

Background: Rathke cleft cysts (RCCs) are benign sellar lesions that are generally asymptomatic but sometimes warrant transsphenoidal drainage. Small case reports have described infected RCCs, but this phenomenon remains uncharacterized.

Objective: We reviewed RCCs over 23 years at our institution to determine factors predicting infection and recurrence.

Methods: We retrospectively reviewed the magnetic resonance images, laboratory results, microbiology, and pathology of 176 RCC patients (1985-2008) who underwent initial operation at our institution (n = 170) or at another institution followed by recurrence managed at our institution (n = 6).

Results: There were 3 RCC categories: cysts cultured intraoperatively during initial surgery (n = 21), cysts not cultured during initial surgery but cultured during subsequent surgery (n = 9), and cysts that were never cultured (n = 146). Cultured cysts were larger (1.6 vs 1.2 cm; P = .002) and had more frequent pituitary dysfunction (76% vs 30%; P < .001) than noncultured cysts. Restricted diffusion was also more common in cultured cysts (50% vs 0%; P = .02). Of cysts cultured at initial or subsequent surgery, 48% and 44%, respectively, had positive cultures (n = 14) and were treated with antibiotics. The most common organisms were Staphylococcus epidermidis (64%) and Propionibacterium acnes (57%). Kaplan-Meier recurrence rates were 13% (culture positive/antibiotic treated), 31% (culture negative/not antibiotic treated), and 9% (noncultured) (P = .002, cultured vs noncultured; P = .002, culture negative/not antibiotic treated vs non-cultured; P = .5 culture positive/antibiotic treated vs noncultured).

Conclusion: Suspected RCC infection, regardless of culture results, is a strong predictor of recurrence and may warrant antibiotic treatment. With antibiotic treatment, the recurrence rate of infected RCC approaches that of noninfected cysts. The higher recurrence rates reported in other series may reflect underrecognition of occult infection.