I’m Ed Smith, a Sacramento Eye Injury Attorney. Despite recent advances in the treatment of endophthalmitis, infection from penetrating eye trauma continues to present a clinical challenge. It remains an important cause of vision failure following open globe injuries and may complicate seemingly benign injuries such as small, self-sealing corneal lacerations without associated intraocular damage. About 62 percent of all cases of endophthalmitis occur after eye surgery, ten percent are associated with planned or inadvertent filtering blebs and the rest are due to metastatic spread from other infected areas.
Although the prognosis has been significantly improved by recent refinements in diagnoses, antibiotic therapy, and vitreous surgery, the overall prognosis of traumatic endophthalmitis remains poor compared to that seen in intraocular surgery. The reasons for the poor prognosis include associated damage to vital eye structures, infection with extremely virulent organisms, and delay in diagnosis and treatment of the condition.
Postoperative endophthalmitis is usually caused by gram positive cocci, such as Staphylococcus epidermidis, accounting for about half of these cases. Most of the gram negative organisms are proteus but pseudomonas infections can happen. Fungi are an uncommon cause of infection, resulting in only 100 proven cases. Other organisms are streptococci and Hemophilus species. There is also a high incidence of Bacillus species, in particular Bacillus cereus. These organisms are found in about 20 percent of cases.
Bacillus cereus is probably the most destructive organism found in the eye. It gives off several enzymes and exotoxins that typically cause infections to advance explosively to an unsalvageable state within 24 hours of diagnoses. Only one reported case of infection with this organism retained useful visual acuity following treatment. There is a strong association between B. cereus infections and retained metallic foreign bodies, especially when associated with farm and soil-related injuries. This organism and clostridia are usually the only endophthalmitis-causing agents that produce constitutional symptoms such as fever, sweat, and leukocytosis.
As in postoperative cases, fungi are a relatively uncommon cause of traumatic endophthalmitis. Only 20 cases of posttraumatic fungal endophthalmitis have been reported. Fungi are implicated in about 8 percent of trauma-related endophthalmitis cases. The worst prognosis is usually found in cases involving Fusarium solanae and similar virulent species.
The diagnosis of traumatic endophthalmitis depends on having a high index of suspicion along with the application of appropriate diagnostic procedures and culture techniques. The recognition of endophthalmitis after trauma may be obscured by the consequences of the trauma itself, such as pain, swelling, loss of vision, redness, and opacification of the media of the eye. These can make the usual signs and symptoms difficult to distinguish from a typical inflammatory response to the injury.
These difficulties in diagnosis often contribute to a delay in treatment, which is an important factor in the poorer prognosis in traumatic endophthalmitis compared to surgical cases. Although the mean interval from the time of injury and onset of clinically detectable endophthalmitis is difficult to determine, studies have shown that the interval for bacterial cases was about 4 days, while the interval for fungal cases was about 57 days. The time interval for bacterial cases reflected a better prognosis in organisms of lower virulence. With such organisms, the endophthalmitis may develop more slowly and may remain treatable for a longer period of time after the injury.
The typical symptoms of bacterial endophthalmitis may include eye pain, headache, photophobia, and purulent discharge from the eye. Clinical signs may include fever, proptosis, eyelid swelling, ecchymosis, increased intraocular pressure and a decrease or absence of the red reflex. There may be corneal edema or corneal ring abscess associated with anterior chamber reaction, including flare, cells and fibrin within the anterior chamber. Contrary to popular belief, air in the anterior chamber is not always associated with Clostridium species and B. cereus has also been reported to have this sign.
Features of a B. cereus infection include a rapid and explosive onset, usually within 24 hours after the injury, associated with bruising, periorbital swelling, extreme proptosis, low grade fever, and leukocytosis. There may be a peripheral corneal ring infiltrate which may progress to an abscess within hours. When the infection has reached this state, the eye has already suffered irretrievable loss of vision.
If there is evidence of a retained foreign body, the appropriate x-rays and ultrasound studies must be obtained. They are especially important in cases where metal has been struck against metal. CT scanning is perhaps the single most sensitive technique for identifying these foreign bodies.
In one study of endophthalmitis, there was an incidence of infection in 10.7 percent of cases with a retained foreign body and a 5.2 percent incidence without a retained foreign body. Ultrasound is important where the foreign body is not radiopaque and may also show up a retinal detachment. The combination of retinal detachment and endophthalmitis carries an extremely poor prognosis.
The diagnosis ultimately rests with the demonstration of infectious organisms within the eye by doing appropriate cultures and stains of the eye. If the examination is suspicious for an infection, vitreous and aqueous samples must be obtained for culture prior to therapy. Although a vitreous culture is a much better culture medium for microorganisms than is aqueous, and the rate of recovery of organisms from the vitreous is much higher, in some instances, the aqueous cultures are positive and the vitreous cultures are negative. This is why both cultures need to be taken.
The culture should be taken in the operating room but can be obtained in the office if the patient is cooperative and the trauma very limited. The choice of anesthesia will vary according to the trauma but a preference for general anesthesia in the management of most cases of eye trauma is preferred.
The aqueous culture can be obtained by a simple anterior chamber tap with a 25-27 gauge needle on a tuberculin syringe, taking out about 0.1 ml o fluid. The vitreous can be obtained by needle aspiration or vitrectomy instrumentation, the latter limited to the operating room setting. The needle aspiration is performed using a 22-gauge needle at the limbus and through the pupil, taking out 0.2-0.3 ml of fluid into a 3 ml syringe.
The samples should be examined by gram staining but because a gram stain is positive in only about 60 percent of culture-positive cases, presumptive therapy for bacterial endophthalmitis for suspected cases should not be withheld solely because of a negative gram stain. Use of fungal stains should also be undertaken to rule out fungal infection of the eye.
Samples should be promptly plated on a fresh petri dish containing blood agar, chocolate agar, and Thioglycolate broth. The dishes should be incubated at 37 degrees Celsius. A separate blood agar plate should be incubated at room temperature for fungi. One plate containing cooked meat medium should be incubated for anaerobic bacteria.
Once the appropriate specimens have been obtained for culture, presumptive antibiotic therapy for endophthalmitis should be offered by intravitreal, topical, systemic, and periocular routes.
Antibiotic Prophylaxis and Therapy
The management of traumatic endophthalmitis includes three different clinical settings: 1) prophylaxis of endophthalmitis, 2) initial management of suspected endophthalmitis and 3) subsequent management of culture positive infections. The specific elements of management include antibiotics, vitrectomy, and steroids. These clinical settings and elements of management are highly interrelated.
The routine use of intravenous, periocular, and topical antibiotics is indicated in virtually all cases of penetrating ocular trauma, encompassing the clinical setting of prophylaxis, presumptive treatment, and ongoing management of traumatic endophthalmitis. The choice of antibiotics depends on the breakdown of the ocular/blood barrier so that the actual intraocular drug levels can be at a therapeutic level. IV antibiotics can be used for 3-5 days to up to 10-14 days.
Antibiotics selected to cover for a broad spectrum of gram positive and gram negative organisms typically include vancomycin or a cephalosporin for gram positive coverage and an aminoglycoside for gram negative coverage. Some organisms have been found to be resistant to cephalosporins but have been found to be sensitive to vancomycin. Gram negative coverage can be undertaken with gentamycin or amikacin.
Antibiotic regimens are altered if fungal organisms are observed in a fresh smear of vitreous or aqueous. Detection of fungal organisms on the smear at the time of specimen collection is advantageous and may permit the use of antifungal medications shortly after the injury. In more typical cases, the initial smear is not definitive, and presumptive antibacterial therapy is offered instead. The therapy for fungi becomes secondary and is started promptly upon the finding of a positive fungal culture.
Corticosteroids have been increasingly used by systemic, subconjunctival, topical and intravitreal routes. Systemic, topical, and subconjunctival steroids have long been found to be effective in reducing the ocular inflammation in endophthalmitis. Prophylaxis with steroids should be avoided as this can mask the signs and symptoms of the infection. In cases of significant anterior chamber inflammation after penetrating trauma, topical steroids may be cautiously used.
The prognosis of traumatic endophthalmitis, although better over the past few decades, is still dismal and remains the most dreaded complication of penetrating trauma. It is considered successful if the patient retains a 20/400 or better visual status. A review of all reported cases since 1948 revealed a 66 percent rate of enucleation of the eye or no light perception but these studies did not differentiate between traumatic cases and other causes of endophthalmitis. Still, traumatic cases have a significantly worse prognosis when compared to cases that occur following surgery.
There are several important prognostic indicators for endophthalmitis in general. Favorable outcomes are associated with negative cultures or cultures with less virulent organisms. Prompt versus delayed therapy seems to make a difference and things like good visual acuity, a positive red reflex, absence of retinal detachment, and absence of nonsurgical trauma also make a difference. Poor prognostic indicators include a delay in diagnosis, a more virulent organism, and the extent of the associated injury.
With Bacillus species, a delay in diagnosis of even a few hours can mean the difference between successful outcome and loss of the eye. Visual results in fungal cases are seemingly superior but the issue is confounded by the extent of injury since most reported fungal cases had trauma limited to a corneal laceration.
Member of Million Dollar Advocates Forum.