Clinical management of an outbreak of nutritionally variant streptococcus endophthalmitis following intravitreal bevacizumab injection

BACKGROUND: The management of an outbreak of endophthalmitis associated with intravitreal bevacizumab represents a challenging real-time process involving identification of cases, treatment and mitigation measures during the outbreak. We summarize the clinical presentation and management of a cluster of endophthalmitis cases from contaminated bevacizumab, in addition to mathematical probabilistic assessment of the number of cases that define an outbreak. METHODS: A retrospective study was conducted to assess the management of an endophthalmitis outbreak after intravitreal bevacizumab (IVB) administration. Demographic data, clinical information, individual patient management and public health reporting measures were reviewed. Outcomes of patients who received prophylactic antibiotics for endophthalmitis prevention were also reviewed. Binomial tail probability calculations were performed to determine the likelihood of clusters of endophthalmitis that could inform when an outbreak was evolving that would warrant more public health notification measures and communication. RESULTS: Forty-five eyes of 42 patients who received IVB from a single batch were reviewed. Four cases of endophthalmitis from Granulicatella adiacens, a nutritionally-variant Streptococcus species, were treated successfully with intravitreal antibiotics ± vitrectomy. Thirty-four of the remaining 41 eyes were treated with prophylactic intravitreal vancomycin with no additional cases of endophthalmitis. Outbreak management also included CDC, ASRS and public health authority notification. Binominal tail probabilities demonstrated the rarity of clusters from a single batch (i.e. ~ 1/10,000 for 2 cases; 1/2 million for 3 cases). However, given the U.S. scale of IVB administration, there is an 87% chance of a cluster ≧ 2 and a 1% chance of a cluster ≧ 3 cases annually, which may guide outbreak management. A process diagram was developed to incorporate patient management and public health measures when an outbreak is suspected. CONCLUSION: Intravitreal antibiotics and vitrectomy were effective in the individual management of cases of endophthalmitis, and no serious adverse events occurred with prophylactic intravitreal vancomycin for at-risk eyes. Best practices for outbreaks should be evaluated, given their likelihood within the U.S. and the sight-threatening consequences of endophthalmitis.