The causes of postoperative inflammation need to be divided into those occurring acutely and those occurring after the passage of time.
Fungal endophthalmitis may occur sometime after surgery.
Lens-induced uveitis may occur spontaneously in an eye with a hypermature lens or may follow ocular trauma or cataract surgery.
Intraocular inflammation can occur after any ocular surgical procedure. It is especially important in the postoperative patient to differentiate infectious causes of uveitis from other causes of intraocular inflammation because bacterial and fungal endophthalmitis require prompt treatment with specific antimicrobial therapy. The major causes of postsurgical uveitis are listed in Box 18-1 .
DAY 1 to DAY 30
Acute aerobic bacterial endophthalmitis
Increased activity of previous uveitis
Phacogenic (lens-related) uveitis
Toxic reaction to intraocular lens
DAY 15 to 2 YEARS
Propionibacterium acnes or other anaerobic endophthalmitis
Low virulence aerobic bacterial endophthalmitis
Phacogenic (lens-related) uveitis
Toxic reaction to intraocular lens
Iris–ciliary body irritation related to physical contact with intraocular lens
Glaucoma drainage device
New onset of idiopathic uveitis
Patients with a preexisting uveitis often have an exacerbation of intraocular inflammation after surgery. The flare-up usually occurs 3 days to a week postoperatively in patients who receive subconjunctival corticosteroids at the end of the surgical procedure. In patients who do not receive steroids the inflammation may occur earlier. Occasionally, the uveitis can be severe enough to be confused with an infectious endophthalmitis, although hypopyon and severe pain are rare in patients with noninfectious postsurgical uveitis. In contrast, low-grade intraocular inflammation is common after many ocular procedures, especially after cataract extraction. This inflammation appears to be mediated by prostaglandins and can be inhibited by prostaglandin inhibitors. In contrast, a rat model of bacterial endophthalmitis showed that a wide range of proinflammatory cytokines are released, including tumor necrosis factor (TNF)-α, interleukin-1β, and the rat equivalent of interleukin-8. During these severe infections, a better idea is emerging of the host/pathogen interactions; the eye upregulates αB crystallin in an attempt to prevent apoptosis of the retinal cells. However, some infectious agents, such as Propionibacterium acnes , may produce a low-grade postsurgical endophthalmitis that may take months to develop.
Acute bacterial endophthalmitis
Most cases of bacterial endophthalmitis occur after intraocular surgery and are a surgical emergency. Fortunately, postoperative endophthalmitis is a rare condition. Allen and Mangiaracine reported 22 occurrences of endophthalmitis in a series of 20 000 ocular operations (0.11%), and nine occurrences in a second series of 16 000 operations (0.056%). The overall prevalence has been estimated to be 0.07%. Post-keratoplasty endophthalmitis is associated with grafts from those who die in hospital from cancer. The rate for endophthalmitis after pars plana vitrectomy is low, reported at 0.039%, with a recent publication reporting the incidence of infection after 20 and 25 gauge vitrectomies to be the same (0.03%). However, the visual outcome of these cases is usually not good. For cataract extraction older studies suggested a rate between 0.07% and 0.12%5; , 75% of patients present within 11 days of cataract surgery or a secondary implant. West et al. suggested that the rate has been increasing over the years, from 0.1% in the 1990s to 0.2% for 2000–2003. Ng and co-workers reported that in their series of endophthalmitis secondary to cataract surgery 48% occurred after phacoemulsification and 39% after planned extracapsular extraction. In Sweden, the rate was calculated to be 0.048%, with a higher rate seen with a clear cornea approach (0.053%) versus a sclerocorneal incision (0.036%); the difference was statistically significantly different. Thoms and coworkers reviewed 815 clear cornea cataract extraction cases, almost equally divided between sutured and unsutured, and found that there were five cases of culture-positive endophthalmitis in the unsutured group and none in the sutured group (p = 0.022). In addition, those who did not receive topical antibiotics until the day after surgery and those who did not receive 5% povidone-iodine drops immediately after wound closure were at greater risk. In a survey sent to 4000 American Society of Cataract and Refractive Surgery (ASCRS) members, of the 1312 who responded, 30% said that they used intracameral antibiotics either via injection or in the irrigating solution. Wound problems appear to be an important contributing factor in about 25% of occurrences of endophthalmitis, and infectious endophthalmitis may be a more common problem after glaucoma surgery, with late-onset endophthalmitis after trabeculectomy and antimetabolites seen in 2.7–8% of patients. Others have reported endophthalmitis associated with glaucoma drainage implants, usually due to exposure of the drainage tube. Kuang et al. evaluated 988 trabeculectomy procedures performed over a 5-year period. In their experience there was one case of early endophthalmitis, but six cases (0.6%) that had a late onset. Mitomycin use increased the risk of late onset infection. It has been reported after Ahmad valve placement. Endophthalmitis was noted in one study in nine of 542 eyes (1.7%), and the rate was five times higher in children.
The interval between surgery and the onset of inflammation is sometimes helpful in distinguishing bacterial endophthalmitis from other causes of postoperative inflammation. Most aerobic bacterial infections occur within 11 days after surgery. Pain and diminished vision are accompanied by conjunctival hyperemia, chemosis, and lid edema. Anterior uveitis with fibrin, hypopyon, and vitritis is also characteristic of a bacterial endophthalmitis. These symptoms and signs should prompt a diagnostic evaluation that includes culture of intraocular fluids, usually obtained during vitrectomy surgery. Even in patients with endophthalmitis after anterior segment surgery, cultures of the vitreous yield positive cultures more frequently than those of the aqueous. The most common organisms are the Gram-positives, particularly coagulase-negative staphylococcus. Methicillin-resistant Staphylococcus aureus (MRSA) needs to be considered in the diagnosis. Deramo and colleagues reported that in 64 cases of acute endophthalmitis secondary to cataract extraction, 33/164 were culture positive, and 6/33 (18.2%) were MRSA. All were sensitive to gentamicin and vancomycin.
The management of bacterial endophthalmitis in the past included vitrectomy, intravitreal, systemic, subconjunctival and topical antibiotics, and corticosteroids. , The approach to postoperative endophthalmitis has been helped by the results of the Endophthalitis Vitrectomy Study, in which all patients received intravitreal amikacin and vancomycin. One group was randomized to pars plana vitrectomy and the other group received a vitreous tap. Each of these larger groups was further subdivided into those receiving systemic antibiotics and those who did not. In general, patients did reasonably well in terms of visual results. It was noted that vitrectomy was of help in patients presenting with poor vision. Systemic antibiotics did not seem to help the outcome. Therefore, the approach today is usually a vitrectomy and the use of broad-spectrum antibiotics given intravitreally, such as vancomycin and ceftazidime or vancomycin and amikacin. However, only 75% of eyes with a clinical picture consistent with bacterial endophthalmitis will have positive bacterial culture results. The polymerase chain reaction is being used more and more in these situations, often with panbacterial 16S rDNA gene primers or cytochrome P-450L primers for fungus. PCR techniques have developed rapidly and have real clinical importance. Chiquet and coworkers used eubacterial PCR primers for bacterial detection in 100 acute post-cataract endophthalmitis cases. When they evaluated the detection rate of culture versus PCR when using aqueous there was little difference (38.2% vs 34.6%) from when a vitreous tap was used. However, with vitrectomy the PCR positivity rate was 70% for PCR versus 9% for culture. When PCR and culture were combined, the detection rate was 68% from a vitreous tap, 72% from a vitrectomy specimen, and 47% from an aqueous sample.
These patients should, however, initially be treated as though they have bacterial endophthalmitis until their condition has improved or until other clinical information suggests an alternative diagnosis. Although sterile endophthalmitis can occur in the early postoperative period and may mimic a bacterial infection, the consequences of untreated bacterial endophthalmitis are so severe that culturing and antibiotic therapy should not be delayed.
What about patients with intraocular foreign bodies? Clearly this subject is too complex to discuss fully here. However, in a retrospective study of 589 eyes with an intraocular foreign body seen over a 12-year period at the King Khaled Eye Hospital, 44 (7.5%) developed clinical evidence of endophthalmitis. A positive culture, however, was obtained on 17 eyes. The issue of whether prophylactic use of antibiotics in cases of intraocular foreign bodies will diminish the risk of infection was addressed by Soheilian and coworkers, who performed a randomized study on 346 eyes with intraocular foreign bodies. In addition to the surgery (repair of the globe and removal of the object) the eyes were randomized to receiving either 40 µg of gentamicin sulfate and 45 µg of clindamycin intravitreally, or BSS injected intravitreally; 8/167 of those receiving BSS developed endophthalmitis, compared to 1/179 in the antibiotic group. Some have questioned the specifics of the study, but the larger issue regarding the value of antibiotics in these circumstances is strengthened.
It is useful to add a comment about intravitreal injections, which are being used more and more frequently. Contamination from the ocular surface is a real possibility. De Caro et al. cultured sites and needles from patients undergoing intravitreal injections and found that 2% (2/114) of needles were contaminated with bacteria. Injection site prophylaxis does make a big difference, with positive cultures found in 43% of the sites tested without prophylaxis and 13% positive after prophylaxis. This is still a relatively rare occurrence clinically. Although endophthalmitis hase been reported after bevacizumab intravitreal injections, in one study of 5233 consecutive injections only one case of acute endophthalmitis was found.
Chronic bacterial endophthalmitis
Although most patients with postoperative bacterial endophthalmitis present with the acute onset of severe inflammation, loss of vision and pain, some bacteria can cause a chronic indolent intraocular infection. These organisms are more difficult to culture and grow more slowly.
Propionibacterium acnes has been recognized as an organism that can cause a chronic postoperative uveitis, especially after cataract extraction with intraocular lens placement. It has also been reported after a cataract extraction and Molteno placement in a 7-year-old with congenital glaucoma. The typical patient with P. acnes endophthalmitis presents with chronic low-grade anterior segment inflammation that starts from 2 months to 2 years after cataract surgery. Topical corticosteroids often suppress the inflammation early in the course of the disease, but do not prevent the protracted low-grade inflammatory response. White plaques composed of bacteria are often seen between the intraocular lens and the posterior capsule or on the intraocular lens haptics ( Fig. 18-1 ). The clinical course of P. acnes endophthalmitis may resemble that of pseudophakic phacoanaphylaxis, , and it is not clear whether some of these patients actually had undiagnosed P. acnes infections.
In addition to the predominance of polymorphonuclear cells, the inflammatory infiltrate in eyes with P. acnes endophthalmitis includes a large number of macrophages (10–15%). The presence of residual lens material in some patients suggests that lens-induced uveitis may be contributing to the inflammation. , Others have suggested that P. acnes immunopotentiation may be the result of inhibition of suppressor T cells caused by stimulated macrophages.
Diagnosis of P. acnes endophthalmitis is based on culture data. It is important to instruct the microbiology laboratory not to discard the cultures after the normal 3–4 days, because the organism is slow growing and cultures should be maintained by the laboratory for 14 days before they are read as negative. Therapy for P. acnes endophthalmitis usually involves a combination of vitrectomy with posterior capsulectomy, intravitreal vancomycin, and sometimes systemic cephalosporins. Although the intraocular lens does not need to be removed at the time of initial therapy, recurrent disease is managed with repeat courses of intravitreal antibiotics and removal of the residual posterior capsule and the intraocular lens. Clark and coworkers reported the following results in their study of eyes with P. acnes : 100% of eyes treated with intraocular antibiotics only had persistent disease; 50% of eyes treated with pars plana vitrectomy and antibiotics had persistent disease; and 14% of eyes treated with pars plana vitrectomy, antibiotics, and subtotal capsulotomy had persistent disease. None of the eyes receiving a total capsulotomy and removal of the intraocular lens had recurrent disease.
Despite the delay in diagnosis and therapy of this chronic endophthalmitis, the visual prognosis is usually good. P. acnes infection has been reported after Ahmed valve placement. Other organisms, including Propionibacterium granulosum , Achromobacter , Pseudomonas oryzihabitans , Mycobacterium , and Staphylococcus epidermidis may also cause a chronic bacterial endophthalmitis. However, in addition to these bacteria, fungal endophthalmitis may also cause chronic smoldering postoperative intraocular inflammation.
Fungal endophthalmitis is a relatively rare condition; however, as our diagnostic abilities improve, the incidence of this condition appears to be increasing. Recent studies show that fungus accounts for 8–13% of culture-positive isolates; in some reports the Candida species predominate, , in others Aspergillus . Fungal infections may not become symptomatic until months after surgery. They often start with low-grade iridocyclitis or vitritis, and some may have corneal involvement. Symptoms of blurred vision, redness, and pain gradually increase. Corticosteroid therapy can temporarily ameliorate the symptoms but should be avoided. With time the infection progresses, and ‘fluff balls’ may form in the vitreous in addition to anterior uveitis and vitritis. A discrete fungal mass may be present in the vitreous or anterior chamber in some eyes. There may be a ‘string of pearls’ in the posterior segment, a mass in the iris, a corneal infiltrate, or a necrotizing scleritis. The risk factors for and ocular characteristics of postoperative and endogenous fungal infections are listed in Table 18-1 .
|Organism||Risk Factors||Ocular Characteristics|
|Candida sp.||IV drug use, chronic IV therapy, surgery||Yellow-white chorioretinal lesions, vitreous fluff balls|
|Aspergillus sp.||IV drug use, immunosuppression, surgery||Yellow-white chorioretinal lesions, vitreous fluff balls|
|Coccidioides immitis||Southwestern United States||Multiple yellow-white chorioretinal lesions, punched-out choroidal lesions; often associated with pulmonary C. immitis infection|
|Cryptococcus neoformans||Immunosuppression, lymphoma||Small or large yellow-white chorioretinal lesions; often associated with meningitis|
|Histoplasma capsulatum||Immunosuppression||Panuveitis with iris involvement|
|Blastomyces dermatitidis||Systemic North American blastomycosis||Panuveitis or multiple small chorioretinal lesions|