Shoot first, ask questions later
Earlier injection recommended in suspected bacterial endophthalmitis
Prompt intervention using intravitreal antibiotics is essential in cases of suspected bacterial endophthalmitis after cataract surgery, emphasised Prof Jan van Meurs at the 19th EURETINA Congress in Paris.
“The evidence suggests that we really need to act quickly as soon as the patient presents with suspected endophthalmitis. It is a case of ‘shoot first and ask questions later’, as the data shows there is a good outcome for most patients even with antibiotic injections only. For those patients who did badly with little or no light perception or evisceration, the key factor appears to be the type of bacteria involved and the fact that they probably received antibiotic treatment too late for it to be effective,” he said.
Prof van Meurs, Rotterdam Eye Hospital, the Netherlands, added that intravitreal dexamethasone, either with or without preservatives, should no longer be used.
“A study carried out by our research team recently showed that dexamethasone as an adjuvant to intravitreal antibiotics does not improve visual acuity in patients treated for suspected bacterial endophthalmitis after cataract surgery,” he said.
Dr van Meurs’ double-blind, prospective, multi-centre, randomised, placebo-controlled trial included 167 patients over a 10-year period, 81 of whom were treated with dexamethasone and 86 with placebo.
The final best-corrected visual acuity (BCVA) did not differ between the dexamethasone and the placebo group, nor did the number of patients with final vision of no light perception.
“The good news was that after one year 70% of patients could see better than 20/40 and more than 80% could see better than 20/200, which are very good figures. Interestingly, the major prognosticator for bad outcome was the bacterial culture result. When we look at eyes that were eviscerated or did badly in terms of light perception, most of them were Gram-negatives or the more severe type of Gram-positive bacteria,” he said.
Prof van Meurs said that a recent study with a similar design by Nijmegen et al of 126 patients also confirmed the efficacy of antibiotic injections only and the fact that the virulence of the bacteria was the key prognostic factor for a poor outcome.
For the more severe endophthalmitis cases, the timing of the injections may be the key factor, he said.
“This is borne out by a study by Michelle Cellegan et al in 2011 on the efficacy of vitrectomy in improving the outcome of severe endophthalmitis. The major finding was that vitrectomy and antibiotics were useful but that the intervention had to bew quick – any later than four hours after infection rendered both of the treatments ineffective. From a purely practical viewpoint it is much easier to be quick with antibiotics than with a vitrectomy,” he said.
More detailed study of the eyes that performed poorly in Prof van Meurs’ own study also showed that all of the bacteria identified were sensitive to antibiotic treatment but it had simply been administered too late.
“The key point is that the eyes that were eviscerated or had little light perception were already lost on presentation in our current way of dealing with them. The choice of antibiotics is not the issue as even these bacteria were sensitive to these drugs, so the only answer is to move the treatment forward,” he said.
Moving to earlier intervention has yielded dividends in the field of neurology, so perhaps ophthalmology could follow its example, suggested Prof van Meurs.
“It was shown in a large series of patients with meningitis in Sweden that those who received faster intervention died less frequently and had less morbidity. So, we would propose a strategy to give injections earlier than we currently do, setting up a system with direct intravitreal antibiotics either at the A&E at the hospital, or more optimally by the referring ophthalmologist the moment that endophthalmitis is suspected. They could then send the patient in for biopsy and with improved PCR techniques we can confirm the diagnosis without relying on live organisms,” he concluded.
Jan van Meurs: J.vanMeurs@oogziekenhuis.nl