Streptococcus pneumoniae endophthalmitis: a study of 36 cases with special reference to antibiotic resistance and treatment options

ABSTRACTPatients (n = 36) diagnosed with pneumococcal endophthalmitis from six Spanish hospitals between 1986 and 2004 were studied retrospectively. The diagnosis was based on clinical findings, ophthalmological examination, and isolation of Streptococcus pneumoniae from vitreous and/or aqueous humours of 19 patients (definite diagnosis), and from other ocular specimens of 17 patients (probable diagnosis). The mean (± SD) age was 69.3 (± 16.5) years (range 1.5–89 years), and 20 (55.5%) patients were male. The origin of endophthalmitis was considered exogenous for 34 (94.5%) patients. The most common predisposing factors were previous ocular surgery (n = 25, 69.4%), ocular trauma (n = 5, 13.9%), and close-to-eye radiotherapy (n = 3, 8.3%). Eleven (30.5%) patients underwent evisceration as the first therapeutic measure (primary evisceration), and evisceration was performed after antibiotic treatment failure (secondary evisceration) for six (16.7%) patients. Primary evisceration was performed more commonly (63.6%) during 1998–2004, while secondary evisceration was only performed during 1986–1997. Eighteen (50%) patients received intra-vitreous antibiotics (mainly vancomycin), and 31 (86.1%) patients were given systemic antibiotic therapy. The most frequent pneumococcal serogroups isolated were 6, 19, 9, 15 and 23. Pulsed-field gel electrophoresis analysis of 23 isolates revealed that four belonged to the international clones Spain23F-1, Spain6B-2, Spain9V-3 and Sweden15A-25. Non-susceptibility rates (i.e., intermediately-resistant and resistant) were: co-trimoxazole, 44.8%; penicillin, 33.3%; tetracycline, 31.0%; erythromycin, 21.9%; chloramphenicol, 17.9%; rifampicin, 7.4%; cefotaxime, 5.9%; and levofloxacin, 0%. Although uncommon, pneumococcal endophthalmitis is a medical emergency because of the often aggressive clinical course, poor visual outcome and need for evisceration in a large proportion of patients.