Indian Journal of Ophthalmology

: 1995  |  Volume : 43  |  Issue : 3  |  Page : 103--116

Postsurgical endophthalmitis: Diagnosis and management

Taraprasad Das1, Mangat R Dogra2, Lingam Gopal3, Subhadra Jalali1, Atul Kumar4, Anupam Malpani5, S Natarajan5, B Rajeev1, Savitri Sharma1,  
1 L.V. Prasad Eye Institute, Hyderabad, India
2 Department of Ophthalmology, Post Graduate Institute of Medical Eduation and Research, Chandigarh, India
3 Sankara Netralaya, Madras, India
4 Dr. Rajendra Prasad Centre for Ophthalmic Sciences, New Delhi, India
5 Aditya Jyot Eye Hospital, Mumbai, India

Correspondence Address:
Taraprasad Das
L.V. Prasad Eye Institute, Road No. 2, Banjara Hills, Hyderabad 500 034


Infectious endophthalmitis following intraocular surgery is a complication that could cause severe visual loss or loss of the eye. The categorisation of the event that led to intraocular infection will help the clinician to predict the infectious agent and begin appropriate therapy. Most of the cases of postsurgical endophthalmitis are seen following cataract surgery. It is important for all ophthalmologists, irrespective of specialisation and areas of interest,to be familiar with the management of endophthalmitis. This review briefly describes the facets of clinical and laboratory diagnosis, pathology, and management. While the different viewpoints in the management of endophthalmitis are mentioned in appropriate places, more attention is paid to present a rational approach to the management of endophthalmitis.

How to cite this article:
Das T, Dogra MR, Gopal L, Jalali S, Kumar A, Malpani A, Natarajan S, Rajeev B, Sharma S. Postsurgical endophthalmitis: Diagnosis and management.Indian J Ophthalmol 1995;43:103-116

How to cite this URL:
Das T, Dogra MR, Gopal L, Jalali S, Kumar A, Malpani A, Natarajan S, Rajeev B, Sharma S. Postsurgical endophthalmitis: Diagnosis and management. Indian J Ophthalmol [serial online] 1995 [cited 2022 Jul 2 ];43:103-116
Available from:

Full Text

Postsurgical endophthalmitis continues to be one of the major devastating complications of intraocular surgery. Entry of an infectious agent into the vitreous cavity is a prerequisite for endophthalmitis to develop, although occasionally severe inflammation in absence of the infecting microorganism could also lead to this problem.

In recent years, the incidence of postsurgical endophthalmitis has decreased considerably due to improved knowledge of the aetiopathogenesis, availability of better diagnostic techniques and wider spectrum of management options. Since cataract surgery is the commonest surgical procedure performed, the majority of the cases of postsurgical endophthalmitis are those following cataract surgery.[1][2][3][4]

Cataract surgery has undergone significant metamorphosis during the last few decades. Intracapsular cataract extraction (ICCE) has been totally replaced by extracapsular cataract extraction (ECCE) and intraocular lens (IOL) implantation is now a standard practice. IOL implantation has created different types of endophthalmitis needing marginal alterations in the management. While the use of antibiotics, particularly intraocular antibiotics, is the mainstay of medical therapy, the wisdom in the use of systemic antibiotics is often questioned. The use of steroids, its appropriate route and timing remain controversial. The role of early vitrectomy in the management of endophthalmitis continues to be a matter of great debate. The ongoing, multicentre prospective study, the Endophthalmitis Vitrectomy Study (EVS)[5] may put to rest some of these controversies.


The majority of the cases of postcataract endophthalmitis occur in the first postoperative week and 88% of the cases develop in the first six weeks of intraocular surgery.[6] Late endophthalmitis following cataract surgery is usually due to fungi or low virulent organisms such as Staphylococcus epidermidis or Propionibacterium acnes.[7]

The baseline evaluation of postsurgical endophthalmitis begins with a high degree of suspicion in the early postsurgical period. The typical clinical characteristics are increasing pain and redness with decreasing visual acuity. Hypopyon is almost always present; lid oedema and raised intraocular pressure are variably present[2],[8][9][10] [Figure:1]A. Occasionally the signs and symptoms may be subtle and misleading in the early stages until the course becomes more fulminant. Hence, intraocular inflammation greater than expected in the postsurgical period should always be viewed with suspicion.

The examination of the suspected patient should include a detailed history, measurement of visual acuity, examination of ocular adnexa, slit-lamp evaluation of the anterior chamber, and examination of the vitreous, either with slit-lamp or indirect ophthalmoscope. Wound leak, necrotic wound margins, suture abscess, vitreous wick, or an inadvertent filtering bleb are often associated with endophthalmitis, and therefore, these should be looked for while evaluating a patient with endophthalmitis. Fundus details may be obscured by clouding of the anterior chamber or due to vitreous opacities [Figure:1]B. The clarity of ocular media could be assessed by indirect ophthalmoscopy as follows:

Grade I Media clarity 6/12 view of the retina.

Grade II Media clarity < 6/12; can visualise second order retinal vessels.

Grade III Can see some retinal vessels.

Grade IV Vessels not seen; red reflex present.

Grade V Red reflex absent.

Visual acuity and clarity of ocular media are the two most important clinical criteria used for the assessment of treatment outcome in the Endophthalmitis Vitrectomy Study.[5] When hazy ocular media prevent optical evaluation of the vitreous and retina, ultrasonography could be of some help. The characteristic ultrasonic features of endophthalmitis are low to medium amplitude vitreous opacities, vitreous membranes, and choroidal thickening. Posterior vitreous detachment and choroidal detachment could variably be present.[11],[12] While ultrasonography alone without clinical correlation is less useful, detection of associated abnormalities such as retinal detachment, dislocated crystalline lens, or retained intraocular foreign body have an important bearing on the visual prognosis and treatment outcome.

Fungal endophthalmitis usually presents between 2 and 4 weeks, or later, following surgery. Persistent iritis might be the only presenting sign.[13] There might be growth of fungus over the iris surface or intraocular lens. Whitish puff balls and vitreous strands are the cardinal signs of fungal infection [Figure:2]A.

Late endophthalmitis due to slow-growing, less virulent bacteria occurs one to 12 months after the intraocular surgery.[7],[14][15][16] Clinically, there is chronic, usually indolent low-grade iridocyclitis. More than half of the patients may have hypopyon and classically a white plaque may be recognised between the posterior capsule and intraocular lens [Figure:3 ]A.

Bacterial endophthalmitis in filtering blebs shows all features of acute infection. On biomicroscopy, the bleb is usually intact at the time of infection, strongly suggesting microbial invasion of the intact conjunctiva over the bleb.[17],[18] Occasionally, a leak is seen at the bleb site which may be the result rather than the cause of infection. The bleb is usually filled with cells and inflammatory debris.

Endophthalmitis following penetrating keratoplasty [Figure:4]A may occur early or late, the incidence of which is reported upto 2%.[19] While in the immediate postsurgical period the infection could be due to contamination of donor material,[20] infectious keratitis or complications of suture removal could cause late onset endophthalmitis.[21]

Endophthalmitis as a complication of vitreoretinal surgery is relatively uncommon. Scleral buckle infection usually appears between the second and seventh day of surgery, and, presents as severe lid oedema with conjunctival chemosis associated with severe ocular pain and headache.[22] Intraocular inflammation develops associated with media haziness, subretinal exudates and localised exudative retinal detachment. In delayed subacute cases fistula and granuloma form, resulting in exposure of the buckle. In addition to other risk factors common to any surgical procedure, diabetes and prolonged duration of surgery have been identified as specific risk factors[23]


Most of the cases of postsurgical endophthalmitis are seen following cataract surgery. Changes in the surgical techniques for cataract removal have altered the pathogenesis of postsurgical endophthalmitis. The initial technique of ICCE resulted in a direct communication between the anterior chamber and vitreous cavity with only an intact anterior hyaloid to prevent intraoperatively introduced microorganism from extending into the vitreous cavity.

Retention of the posterior capsule following ECCE while being associated with a reduced incidence of endophthalmitis has its own associated unique problems.[24] Organisms of low virulence such as Propionibacterium acnes can become sequestered in the retained capsular sac resulting in chronic localised endophthalmitis.[25] Also, retained lens matter could contribute to an increased incidence of phacogenic uveitis, but conversion from ICCE to ECCE has not been associated with an increased incidence of this entity. This could, in part, be explained by the additional need for an infectious agent acting as an adjuvant to sensitize the host to lens protein of self.

The IOL implantation has also increased the potential for postsurgical inflammation due to the IOL acting as an intra-ocular foreign body [Figure:5]. Increased intraocular manipulations, ineffective sterilization techniques and chronic irritation of the iris and ciliary body are some other factors contributing to postsurgical inflammation. In addition, the electrostatic forces on the IOL may attract microorganisms from the periocular tissues during insertion. IOLs, particularly those with polypropylene haptics seem to be associated with an increased risk of endophthalmitis.[26] The IOL haptics may become embedded in the surrounding tissues. Histologically, this erosion appears to be related to the size of the lens and elasticity of the haptics, rather than duration of the implant.[27]

The newer technique of sutureless wound closure leaves a potential route for entry of microorganisms. However, this technique has not proven to be associated with increased risk of postsurgical endophthalmitis.


Infectious agents gain access into the eye during surgery or postoperatively through incision sites. These infectious agents may be cleared from the intraocular cavity or may induce a local inflammatory response culminating in endophthalmitis. The risk of developing endophthalmitis depends on a variety of factors including the size of the inoculum, virulence of the organism, access to the vitreous, and perioperative antibiotic use. The avascular vitreous appears to be a better culture medium than the aqueous. Proliferation of the infectious agents results in an inflammatory response from the surrounding vascular structures and breakdown of the blood-aqueous and blood-vitreous barriers. A well-formed vitreous gel and an intact anterior hyaloid or lens capsule may initially restrict the extension and spread of inflammation and exudates.

In general, bacterial infections tend to cause diffuse liquefactive necrosis of the vitreous and proliferate rapidly. In acute bacterial infections there occurs invasion of polymophonuclear cells that are seen in varying stages of degeneration [Figure:6], a hallmark of suppurative necrosis.

Gram's stain reveals numerous bacteria in the vicinity of the necrotic debris. In the presence of a typical polymorphonuclear infiltration, a negative Gram's stain, however, does not absolutely rule out bacterial infection particularly when the patient is already on intensive antibiotic therapy. In this situation a culture will be more sensitive.

In contrast, mycotic infections of the vitreous progress slowly and remain localised for a longer duration and tend to form multifocal microabscesses. The inflammatory response is granulomatous and is characterized by a predominant mononuclear cell infiltration. A few eosinophils may be present. Vitreous biopsy may show periodic acid Schiff (PAS)-positive yeast forms [Figure:7] or fungal hyphae [Figure:8]. Histologically, the diagnosis is usually established by demonstrating fungal elements using special stains such as Gomori's methenamine silver or periodic acid Schiff.

The inflammatory process tends to get exacerbated once these infiltrates reach vascular structures such as the iris or ciliary body. Although usually granulomatous in nature, both Candida and Aspergillus may sometimes induce suppurative inflammation.

Inflammatory exudates in the vitreous cause structural changes leading to liquefaction, opacification, posterior vitreous detachment, and eventually, abscess formation. Inflammation extends to structures such as the iris, ciliary body and retina, resulting in suppurative necrosis. Retinal perivasculitis with lymphocytic infiltration can occur, but in many instances this vascular sheathing is clinically masked by opacification of the media.

Continued inflammation in the region of the ciliary body and vitreous base causes proliferation of the ciliary epithelium accompanied by capillaries and fibrous tissue. The fibrovascular membrane extends along the scaffold provided by the anterior hyaloid or remnants of the lens-iris diaphragm, eventually forming a cyclitic membrane spanning the posterior chamber. The membrane contracts as it grows, thereby exerting traction on the ciliary body. This results in ciliary and choroidal detachment with subsequent hypotony. Similarly, in the acute phase of inflammation, reactive proliferation of fibrovascular tissue occurs into the vitreous from the adjacent ciliary body and retina. This leads to vitreous organisation, shrinkage and traction on the retina and choroid that result in detachment of these structures. These changes eventually lead to phthisis.

Tissue Collection: Specimens obtained following surgery or biopsy for postsurgical endophthalmitis could include anterior chamber fluid, vitreous cavity contents, iris, IOL, remnants of lens and lens capsule, and at times the eviscerated contents of the globe or an enucleated globe. Care should be taken to use a sterile vitrectomy cassette. Should the processing of these samples be delayed, they should be fixed with an equal volume of 95% ethy lalcohol or left unfixed over night at 4C. A portion of the specimen should also be submitted for microbiologic evaluation. For routine processing, the iris tissue, explanted IOL, eviscerated contents, and enucleated globe are fixed in buffered 10% formalin. Rarely, if immunohistochemistry is indicated, the tissue can either be snap frozen or fixed in appropriate fixatives depending on the antigen to be detected. For electron microscopy studies, immediate fixation in a paraformaldehyde-glutaraldehyde fixative is ideal.

Tissue Processing: Air-dried smears provide a quick technique to detect infectious agents in aqueous or vitreous smears. Thin smears that avoid cell overlap, stained with Gram, Giemsa or PAS stains provide adequate information. The diluted specimens are first concentrated before being processed further.[28] This could be done by simple centrifugation to obtain a paraffin pellet; by using cytospin to obtain a compact smear; or passing the sample through a Millipore filter and staining or processing the entire filter. Each, however, has its own advantages and suitability.


Both gram-positive and gram-negative bacteria and fungi are known to cause infectious endophthalmitis. Gram-positive aerobic bacteria account for 76 to 90% of cases of culture-positive postsurgical endophthalmitis; 7 to 16% of cases are due to gram-negative bacilli and 3 to 8% of cases are of fungal origin.[2],[4],[6] Staphyloccus epidermidis is the causative agent in 20 to 50% of cases of postsurgical endophthalmitis.[6],[8],[29] Other common gram-positive bacteria include Staphylococcus aureus, Streptococcus pneumoniae, Streptococcus viridans, and Streptococcus pyogenes. Certain gram-positive bacilli such as Propionibacterium acnes has been isolated in chronic low-grade endophthalmitis after ECCE with or without IOL.[14] Among the gram-negative organisms associated with postsurgical endophthalmitis, Pseudomonas aeruginosa is the commonest, while others include Proteus, Haemophilus influenzae, Klebsiella pneumoniae, E. coli, and Enterobacter species.[30][31][32]

Fungal endophthalmitis is rare and is usually caused by both saprophytes and opportunistic pathogens such as Candida, Aspergillus and Fusarium.[13] Endophthalmitis following glaucoma filtering surgery is reported to be caused by Streptococcus species (56%) or H. influenzae (20%).[17,18] Buckle infection following retinal reattachment surgery is usually caused by S. epidermidis.[22] Postkeratoplasty infections are usually due to S.epidermidis and P. aeruginosa; fungal infections are not uncommon.[20],[21]


As with any infection, the cornerstone of management of postsurgical endophthalmitis includes identification of the responsible organism and prompt institution of appropriate therapy.

Collection and Transport of Specimens: Both aqueous and vitreous cultures are recommended in endophthalmitis.[33][34] However, the sensitivity of the culture is increased with the vitreous rather than with the aqueous alone.[34] Cultures of external ocular surface are not of value except in the presence of an open wound or a leaking bleb. The advent of therapeutic vitrectomy has provided an alternate modality of obtaining vitreous in endophthalmitis. The culture sensitivity of the vitrectomy cassette fluid, passed through 0.2 μ Millipore filter, is reportedly higher than vitreous biopsy culture obtained by a needle and syringe.[35]

The aqueous and vitreous specimens are usually sent to the laboratory in the same syringes used for collection, with a rubber cork struck onto the needle [Figure 9]. If the laboratory is not located nearby, the smear and culture media need to be inoculated in the operating room itself and transported. The use of gas pak system would be necessary for incubation (before transport) of certain media for isolation of anaerobic bacteria.[36] Transport media of the conventional kind have no place in the transport of aqueous and vitreous fluids.

Microscopic Examination: Smears for microscopic examination are prepared by placing two or three drops of the aqueous or vitreous onto clean glass slides and air drying them. The air-dried smears are fixed in 95% methylalcohol for 5 minutes prior to staining. The recommended stains are Gram's, Giemsa's, Gomori's methenamine silver and Calcofluor white. While Gram's and Giemsa's stains have been found adequate in most occasions, inclusion of Gomori's methenamine and Calcofluor white stains are especially useful in suspected fungal infection [Figure:10]A, [Figure:10]B.

Smears provide a rapid diagnosis though less specific and sensitive than cultures of the aqueous and vitreous.[8],[30],[37] The experience of two of us (SS, TD) has been similar. In an analysis of 47 consecutive cases of endophthalmitis the sensitivity and specificity of Gram's stain was 66% and 84%, respectively. The sensitivity and specificity of Giemsa's stain (Diff Quick, Bacto lab, Australia) was lower at 40.7% and 80.9%, respectively, than Gram's stain.[38]

Cytopathology[28] in infectious endophthalmitis, in our opinion, is of limited value. In a prospective study by some of us (SS, TD, SJ), polymorphs were seen in both culture-positive and culture-negative cases (96% and 80% respectively), and in both bacterial and fungal infections with no specific preponderance (unpublished data).

Culture Methods and Results: The aqueous and vitreous being normally sterile fluids, do not require selective media for culture. For culture, a host of media must be included [Table:1] for favourable growth of aerobic and anaerobic bacteria and fungi. However, the number of media to be included depends on the volume of sample material available. The culture media must be examined daily for 2 weeks for any microbial growth.

The vitreous biopsy collected in a syringe is inoculated onto various media per se. The vitrectomy specimen collected in a sterile cassette is processed by centrifugation or by passing through 0.2 μ. membrane filter. The filter is removed aseptically and cut into segments for direct inoculation onto the media. While Donahue et al[35] have reported significant sensitivity of culturing cassette fluid alone, we found that processing of both samples (biopsy and cassette fluid) provided greater sensitivity.[38]

The culture is defined as positive when the same organism grows in more than one medium or there is confluent growth on one or more solid media at the inoculation site. Growth in only one liquid medium or scanty growth on one solid medium is generally considered as "uncertain significance." Review of major reports in the literature, however, shows that only 64% of vitreous specimens obtained from eyes with clinical diagnosis of endophthalmitis are culture positive.

Antimicrobial Susceptibility Testing: The standard procedure of Kirby-Bauer disc diffusion technique[39]may be utilized for the antibiotic susceptibility testing of bacteria. The results of this test relate to the level of the drug achievable in the serum, and not to the concentration of the drug in the intraocular tissues.[40] Knowledge of both minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) of antibiotics are useful.[41] A MBC lower than the recommended doses of intraocular antibiotics would indicate effectiveness of the drug against the organism. In our study of 22 vitreous isolates and from antibiotics (gentamicin, vancomycin, cefazolin and ciprofloxacin), MBC was found higher than MIC in all cases though MBC still remained lower than the recommended doses of intravitreal antibiotics in 77% of instances. A good correlation existed between the disc diffusion technique and MIC.[42] Estimation of MBC of drugs against vitreous isolates may be performed routinely and a combined method of testing MIC and MBC has also been described.[43]Antifungal susceptibility tests are not generally performed. Broth dilution and agar dilution techniques and modification of standard tube dilution method have been described.[40]


Effective treatment in the closed space of the vitreous cavity is difficult. Bacterial toxins released inside the vitreous cavity are not washed away by the vascular route, but remain to cause continued damage to the retina and surrounding tissues. For years, the mainstay of treatment of infectious endophthalmitis has been the use of topical and systemic antibiotics. The current approach, however, mainly relies on the use of intravitreal antibiotics and vitrectomy.

 Medical Treatment

Successful medical management of infectious endophthalmitis depends on prompt diagnosis and institution of appropriate antibiotic therapy. The choice and route of antibiotic therapy rests largely on its antibacterial spectrum, pharmacokinetic behaviour of the drug and the ocular tolerance.[44]

Pharmacokinetics: Unlike intravitreal drugs, the ocular penetration by systemic drugs is significantly blocked by the blood-retinal barrier.[45] Drugs are eliminated from the vitreous cavity by the posterior and anterior routes.[46],[47] The beta-lactum antibiotics and steroids are eliminated from the vitreous cavity by the posterior route which involves the retinal pump. The aminoglycosides are eliminated by the anterior route by simple diffusion into the anterior chamber and aqueous outflow channels.

In endophthalmitis, elimination of drugs through the posterior route is decreased due to damage to the retinal pump and the elimination through the anterior route is increased. The surgical status of the eye also contributes to the effect of the drug level and its distribution inside the vitreous cavity.[48] Aphakia and vitrectomy seem to increase the space into which the intravitreal drug is dispersed. This results in lower vitreous drug levels in aphakic and vitrectomised eyes.

The critical stage in antibiotic therapy is before the microbiologic evaluation reports are available to the treating surgeon. Though the most common organism causing postsurgical endophthalmitis is coagulase-negative Staphylococcus, broad spectrum antimicrobial therapy is preferred in order to cover the full range of ocular penetration, topical drugs are less likely to reach measurable concentrations in the vitreous. Despite this limitation, topical antibiotics are useful in the treatment of co-existent problems, such as, suture abscess or corneal ulcer. Routine management consists of instillation of fortified antibiotics, cycloplegics and steroids every one or two hours.

Though subconjunctival drugs have a poor intravitreal penetration, they have a better aqueous penetration than topical drugs. Subconjunctival antibiotics are recommended on initial presentation.

Systemic Antibiotics: In general, systemic antibiotics such as beta-lactum antibiotics and aminoglycosides suffer from poor intravitreal penetration even in inflammed eyes. Recent reports in the literature, however, document good penetration of cefazoline, ciprofloxacin and ceftazidime following systemic therapy.[49][50][51] This, however, was possible after repeated therapy in inflammed eyes. Also, in aphakic and/or vitrectomised eyes relatively higher drug levels were achieved.

Despite occasional reports of successful treatment of postsurgical endophthalmitis without systemic antibiotics, the general practice involves treatment with a combination of various systemic antibiotics.[52] In order to examine more closely the role of systemic antibiotics in the treatment of postcataract endophthalmitis, the Endophthalmitis Vitrectomy Study Group[5] is randomising patients to the "antibiotics" and "no antibiotics" groups. The patients in the "antibiotics" group receive a combination of intravenous ceftazidime and amikacin. Systemic antibiotics are not without side effects and toxicity. The aminoglycosides are nephrotoxic and antifungal antibiotics such as ketaconazole are hepatotoxic. Hence, repeated laboratory and clinical monitoring is required to prevent damage to the kidney or liver.

Intravitreal Antibiotics: Intravitreal administration of antibiotics is considered the principal modality of antibiotic therapy in infectious endophthalmitis. In recent years, cefazoline and gentamicin have been the mainstay of initial treatment regimens for bacterial endophthalmitis. Cefazoline, a first generation cephalosporine, is highly active against many of the gram-positive microorganisms and gentamicin, an aminoglycoside, is the drug of choice against many gram-negative microorganisms. The changing trends in bacterial sensitivity patterns and increasing reports of retinal toxicity due to aminoglycosides[53] have prompted changes in the current recommendation of intravitreal antibiotics to the combined therapy with vancomycin (effective against gram-positive organisms) and amikacin (effective against gram-negative organisms).[41]Of late, there is an increasing tendency to replace amikacin with ceftazidime since amikacin is also not without toxicity and ceftazidime has been found to be highly effective against gram-negative organisms without retinal toxicity.[54]

While the drugs of choice for intravitreal injection are changing, the principle of utilising two drugs, one against gram-positive bacteria and the other against gram-negative bacteria, remains valid. In cases of fungal endophthalmitis the choice of intravitreal antibiotic is restricted to amphotericin B.

The details of initial antibiotic therapy as practised by us and the details of preparation of intravitreal drug therapy are outlined in [Table:2] and [Table:3].

 Surgical Treatment

Vitrectomy is the surgical treatment of choice in endophthalmitis which is equivalent to incision and drainage of abscess anywhere else in the body. Vitrectomy removes the sequestered pockets of infection from the eye. In addition, it removes the inflammatory debris and toxins, clears the media and reduces the risk of late complications related to cellular proliferation on the vitreous matrix.

The precise timing for surgical intervention by vitrectomy in the management of endophthalmitis remains controversial. The risks of vitrectomy must be judged against the potential benefits in each patient. In general, vitrectomy is recommended in the following situations: (1) no improvement or actual worsening of the condition clinically after 48 hours of conservative treatment; (2) moderate to advanced stage of infection (absence of fundal glow) on initial presentation; (3) suspected fungal infection; (4) bleb or trauma associated infection; (5) endophthalmitis associated with retained intra-ocular foreign body; and (6) chronic endophthalmitis with remission and exacerbations.

The surgical technique of vitrectomy is similar to the standard three-port vitrectomy with few exceptions. Placement of infusion cannula and its visualisation through a boggy choroid and hazy ocular media is difficult. It is generally advised to use a longer infusion cannula, such as a 6-mm cannula, and also to use the Wilson's cannula in the initial stage of the surgery till the regular cannula is visible through the pupil. The anterior chamber often contains hypopyon and exudative membrane which interferes with adequate visualisation of the vitreous. In aphakic patients the curved Wilson's cannula and the vitrectomy probe could be brought forward into the anterior chamber through the pupil to remove the debris, while in pseudophakic patients a 26-gauge instrument, such as, a hypodermic needle could be introduced through the limbus to clear the hypopyon and any inflammatory membrane without explantation of the IOL. The IOL need not be sacrificed in all the cases of pseudophakic endophthalmitis as has been shown in clinical[4] and experimental studies.[55] However, explantation of the IOL is necessary in events of gross infection (usually associated with gram-negative organism) and in fungal endophthalmitis. Explantation of the IOL is also advocated in P. acnes endophthalmitis, but the current suggestion is to perform a generous capsulotomy, either to retain or exchange the IOL and to inject intraocular antibiotic near the remaining part of the capsular bag.[56]

In the beginning of vitrectomy, vitreous sample is collected by a short tubing attached to the suction port of the vitrectomy probe. The suction is operated manually by a 5-ml syringe attached to the tubing. For better recovery of microorganisms it is useful to obtain an undiluted vitreous sample before starting the infusion fluid. The infusion is begun once the vitreous sample is collected for microbiologic or pathologic studies. The aim of vitrectomy in endophthalmitis is only to perform a "core" vitrectomy [Figure:11]. No attempt is made for "complete" vitrectomy as this often results in tears and detachment of the inflammed and often necrotic retina which is extremely difficult to treat. At the conclusion of the surgery, intravitreal drugs, antibiotics with or without steroids, are given.

Postoperatively, intensive topical therapy (antibiotics, steroids, and cycloplegics) is continued for a few weeks. The vitrectomy and intravitreal drugs may be repeated if adequate response is not noticed clinically or when the infecting organism is found resistant to the antibiotic used earlier.


Cortiosteroids are important adjuvants to antibiotic therapy and vitrectomy in the management of endophthalmitis. The inflammation accompanying the organismal infection is usually severe and predominantly involves infiltration with polymorphs. Bacteria liberate exotoxins and when killed by polymorphs liberate endotoxins. These toxins have a deleterious effect on the sensitive structures of the eye.[57]In addition, many bacterial enzymes such as collagenase, hyaluronidase, deoxyribonuclease, haemolysin, etc., facilitate tissue invasion. While the destruction followed by scarring may not be a serious issue in treatment elsewhere in the body, scarring could lead to permanent ocular damage. Steroids help limit this damage to the minimum by reducing the severity of inflammatory reaction and granulation tissue formation.[58-63]

 Route of Administration

Intravitreal Route: Dexamethasone is the commonest steroid administered intravitreally. In an experimental study, an intravitreal dose of 400 μ. g of dexamethasone was found nontoxic to the retina.[60] Further, a residual concentration of 0.05 μ g/ml was found retained in the vitreous upto 4 days. There is no precise minimum for the therapeutic effect and even the smallest dose has some beneficial role. It is advisable to inject intravitreal steroids separately and not with other antibiotics, particularly vancomycin, since precipitation of the drug is known to occur.

Topical and Subconjunctival Route: Topical steroids have the potential to penetrate an intact cornea and its efficacy is proportional to the frequency of instillation.[64]

Subconjunctival injection of dexamethasone is more effective and is absorbed through the sclera. However, this route of administration is less preferred due to accompanied discomforts by such an injection in an already painful eye. Depot preparations are usually avoided since the facility to withdraw steroid, if so required, is limited.

Systemic Route: Systemic steroids are recommended in infectious endophthalmitis.[65] However, their administration is withheld until the culture reports are available and specific antibiotics are administered for at least 24 hours. The recommended dose is 1.0 to 1.5 mg/ kg/day prednisolone and should be administered in three divided doses to achieve uniform concentration throughout the day unlike in chronic uveitis where once-a-day or an alternate day regimen is preferred.

Steroids are not used in proven cases of fungal endophthalmitis. The antifungal drugs are weak, and hence, suppression of the immunological reaction by steroids could lead to unhindered growth of fungi.

While it is agreed that steroids have a potentially beneficial role in bacterial endophthalmitis, the choice of the route of administration is not uniform with various clinicians. Most controversies perhaps persist as regards to the intravitreal steroid therapy. Though many retrospective studies have documented beneficial effects of intravitreal steroids in both gram-positive and gram-negative organismal infection[8,66,67], the Endophthalmitis Vitrectomy Study[5] has chosen to administer only systemic and topical steroids, and not intravitreal steroid. In a prospective study by two of us (TD, SJ - unpublished data), intravitreal steroid was found useful in reducing the inflammation and early quietening of the eye though this did not have.a bearing on the final visual outcome at three months review following core vitrectomy and intravitreal injection of antibiotics and steroid.


There is general consensus that in recent years the results of postsurgical endophthalmitis have significantly improved. This is due to the improved understanding of the disease process, pathogenicity of the organisms and host defence mechanisms, coupled with availability of better drugs, intravitreal drug therapy and vitrectomy. Successful management of infectious endophthalmitis depends on the triad of vitreous culture, intraocular antibiotic therapy and vitreous surgery [Figure:2]B-[Figure:4]B.

Specific antibiotic therapy in postsurgical endophthalmitis depends on the vitreous culture results. The current recommended specific antibiotics, based on the sensitivity studies, are listed in [Table:4]. This could, however, change depending on the sensitivity pattern. The precise contribution of pars plana vitrectomy as distinct from intraocular antibiotic therapy is impossible to determine. The accumulated data from the literature do not shed much light. Nevertheless, vitrectomy is usually offered to worse cases of infection and the less severe ones are managed by vitreous biopsy and intraocular antibiotics.

Better visual results have been reported in culture-negative cases of postsurgical endophthalmitis.[4],[6],[37],[68]Staphylococcus species have a relatively good prognosis and cases with S. epidermidis endophthalmitis fare better than those with S. aureus endophthalmitis. Endophthalmitis due to Streptococcus, Bacillus and gramnegative bacteria, especially Pseudomonas aeruginosa carry poor visual prognosis.[4,6,67]


The incidence of endophthalmitis following cataract surgery has declined from 1.14% (before 1950) to the present incidence of less than 0.1%[24,69,70] This is due to a variety of measures taken to prevent and treat this dreaded complication.

The most important measure for prevention of exogenous postsurgical endophthalmitis is careful preoperative evaluation and preparation of the patient. Blepharitis, dacryocystitis, conjunctivitis, contralateral ocular prosthesis, prolonged use of topical corticosteroids and chronic urinary tract infection should caution the physician while planning any kind of intraocular surgery.

Since the preoperative conjunctival culture results usually do not correlate well with the postoperative course of events, conjunctival cultures are not recommended. Use of prophylactic topical antibiotics is claimed to have reduced the incidence of endophthalmitis.[71] However, due to poor ocular penetration the value of this mode of therapy is often questioned. All the same, topical antibiotics can control the ocular surface-related infections, if any. The use of intracameral antibiotics through the infusion fluid is suggested.[72][73][74] For prophylaxis, usually half of the maximum non-toxic dosage of the antibiotic is recommended in the infusion bottle[72] [Table:5].

Use of 5% providine-iodine solution for preoperative preparation of the skin around the eye including the lids and conjunctiva is extremely important in obtaining a relatively safe and sterile area for surgery.[75],[76] Despite all precautions at the time of surgery, the entry of at least a few microorganisms into the eye is inevitable.[77],[78]

The relatively low incidence of endophthalmitis seen in actual practice is therefore attributed solely to the efficacy of the patient's own immunological defences.

Hence, it is important not to compromise the immunological defence system in the immediate pre- and postsurgical period by unjudicious, and often, avoidable use of corticosteroids.


In recent years, there have been opportunities to reduce the risk of intraocular infection. This is because our understanding of the causes, prevention, and treatment of endophthalmitis has vastly improved. We now recognise the importance of the resident microbial flora of the eyelids, conjunctiva, and lacrimal system and the fact that, "opportunistic" organisms can cause serious infection if introduced into the eye. We also have a better understanding of the host defence mechanisms including the antimicrobial nature of body fluids, and more specific, humoral and cellular immune process. We now know better how corticosteroids and antibiotics interfere with the immune system of the eye and impair its local defence mechanisms. While on one hand, clinical trials are being done to rationalise the treatment strategy in infectious postsurgical endophthalmitis,[5]efforts are also being made to design rapid diagnostic modalities, such as polymerase chain reaction[79] and fluorescent Gram's stain.[80] Above all, one should be cognizant of the importance of asepsis in the operating room.


1Forster RK. Symposium: Postoperative endophthalmitis. Etiology and diagnosis of bacterial postoperative endophthalmitis. Ophthalmology 85:320-326, 1978.
2Puliafito CA, Backer HS, Haff J, et al. Infectious endophthalmitis: Review of 36 cases. Ophthalmology 89:921-928, 1982.
3Weber DJ, Hoffman KL, Thoft RA, et al. Endophthalmitis following intraocular lens implantation: Report of 30 cases and review of literature. Rev Infect Dis 8:12-20, 1986.
4Driebe W, Mandelbaum S, Forster RK. Pseudophakic endophthalmitis: Diagnosis and management. Ophthalmology 93:442-447, 1986.
5Doft BH. The endophthalmitis vitrectomy study (EVS) (editorial). Arch Ophthalmol 109:487-489, 1991.
6Olson JC, Flynn HW, Forster RK, et al. Results in the treatment of postoperative endophthalmitis. Ophthalmology 90:692-699,1983.
7Mandelbaum S, Meisler DM. Postoperative chronic microbial endophthalmitis. Int Ophthalmol Clin 33:71-79, 1993.
8Rowsey JJ, Newsome DL, Sexton DJ, et al. Endophthalmitis: Current approaches. Ophthalmology 89:1055-1066,1982.
9Flynn HW Jr, Pflugfelder SC, Culbertson WW, et al. Recognition, treatment and prevention of endophthalmitis. Seminars Ophthalmol 4:69-83,1989.
10Rowsey JJ, Jensen H, Sexton DJ. Clinical diagnosis of endophthalmitis. Int Ophthalmol Clin 27:82-88,1987.
11Chan IM, Jalkh AE, Trempe CL, et al. Ultrasonographic findings in endophthalmitis. Ann Ophthalmol 16:778-784, 1985.
12Dacey MP, Valencia M, Lee MB, et al. Echographic findings in infectious endophthalmitis. Arch Ophthalmol 112:1325-1333,1994.
13Theodore FH. Etiology and diagnosis of fungal postoperative endophthalmitis. Trans Am Acad Ophthalmol Otolaryngol 85:327-334, 1978.
14Meisler DM, Mandalbaum S. Propionibacterium associated endophthalmitis after extracapsular cataract extraction: Review of reported cases. Ophthalmology 96:54-61,1989.
15Ormed LD, Paton BG, Haff J, et al. Anaerobic bacterial endophthalmitis. Ophthalmology 94:799-808,1987.
16Fox GM, Joondeph BC, Flynn HW Jr, et al. Delayed onset pseudophakic endophthalmitis. Am J Ophthalmol 111:163-173, 1991.
17Mandelbaum S, Forster RK, Gelender H, et al. Late onset endophthalmitis associated with filtering blebs. Ophthalmology 92:964-972, 1985.
18Mandelbaum S, Forster RK. Endophthalmitis associated with filtering blebs. Int Ophthalmol Clin 27:107-111, 1987.
19Guss RB, Koenig S, De La Pena W, et al. Endophthalmitis after penetrating keratoplasty. Am J Ophthalmol 95:651-658,1983.
20Pardos GJ, Gallagher MA. Microbial contamination of donor eyes. Arch Ophthalmol 100:1611-1613,1984.
21Weiss JL, Melson JD, Lindstrom RL, et al. Bacterial endophthalmitis following penetrating keratoplasty suture removal. Cornea 3:278-280, 1984.
22Mc Meel JW, Naegele DF, Pollalis S, et al. Acute and subacute infections following scleral buckling operations. Ophthalmology 85:341-349, 1978.
23Bacon AS, Davison CR, Patel BC, et al. Infective endophthalmitis following vitreoretinal surgery. Eye 7:529-534,1993.
24Javitt JC, Vitale S, Canner JK, et al. National outcomes of cataract extraction. Endophthalmitis following inpatient surgery. Arch ophthalmol 109:1085-1089, 1991.
25Meisler DM, Palestine AG, Vastine DW, et al. Chronic Propionibacterium endophthalmitis after extracapsular cataract extraction and intraocular lens implantation. Am J Ophthalmol 102:733-739, 1986.
26Menikoff JA, Speaker MG, Marmor M, et al. A case control study of risk factors for postoperative endophthalmitis. Ophthalmology 98:1761-1768,1991.
27Champion R, McDonnel PJ, Green WR. Intraocular lenses. Histologic characteristics of a large series of autopsy eyes. Surv Ophthalmol 30:1-32,1985.
28Green WR. Diagnostic cytopathology of ocular fluid specimens. Ophthalmology 91:726-749, 1984.
29Wilhelmus KR. The pathogenesis of endophthalmitis. Int Ophthalmol Clin 27:74-81,1987.
30Okumato M. Laboratory diagnosis of endophthalmitis. Int ophthalmol Clin 27:89-94, 1987.
31Allen HF. Postoperative endophthalmitis. Introduction, incidence and aetiology. Ophthalmology 85:317-319,1978.
32Allen HF, Mangiaracine AB. Bacterial endophthalmitis after cataract extraction. II. Incidence in 36,000 consecutive operations with special reference to preoperative topical antibiotics. Arch Ophthalmol 91:3-7, 1974.
33Allansmith MR, Skaggs C, Kimura SJ. Anterior chamber paracentesis. Diagnostic value in postoperative endophthalmitis. Arch Ophthalmol 84:745-748, 1970.
34Koul S, Philipson A, Arvidson S. Role of aqueous and vitreous culture in diagnosing infectious endophthalmitis in rabbits. Acta Ophthalmol 68:466-469, 1990.
35Donahue SP, Kowalski Rp, Jewart BH, et al. Vitreous culture in suspected endophthalmitis. Biopsy or vitrectomy? Ophthalmology 100:452-455,1993.
36Baron EJ, Finegold SM. Cultivation and isolation of viable pathogens. In: Bailey & Scotts's Diagnostic Microbiology, 8th ed. Philadelphia, CV Mosby Co., pp. 87-88,1990.
37Forster RK, Abbott RI, Gelender H. Management of infectious endophthalmitis. Ophthalmology 87:313-319, 1980.
38Sharma S, Das T, Adiraju MV, et al. Diagnosis of infectious endophthalmitis: Vitreous biopsy, membrane filter culture or both? Invest Ophthalmol Vis Sci 36:s631, 1995.
39Bauer AW, Kirby WMM, Sherris JC, et al. Antibiotic susceptibility testing by a standardised single disc method. Am J Clin Pathol 45:493-496, 1966.
40Jones DB, Liesegang TJ, Robinson NM. Laboratory diagnosis of ocular infections. Cumitech 13, Am Soc Microbiol. Washington DC. 24,1981.
41Gardner S. Treatment of bacterial endophthalmitis. I. Ocular therapeutics and management. 2:1-16, 1991.
42Gopinathan U, Adiraju MV, Das T, et al. The minimum inhibitory and bactericidal concentrations of antibiotics for vitreous isolates. Invest Ophthalmol Vis Sci 36:sl55, 1995.
43Baron EJ, Fineglol SM. Methods of testing antimicrobial effectiveness. In: Bailey and Scott's Diagnostic Microbiology, 8th ed. Philadelphia, CV Mosby Co., 1990. pp. 171-194.
44Speaker MG, Menikoff JA. Postoperative endophthalmitis: Pathogenesis, prophylaxis and management. Int Ophthalmol Clin 51-70,1992.
45Baum JL, Peyman GA. Antibiotic administration in the treatment of bacterial endophthalmitis. Surv Ophthalmol 21:332-346, 1977.
46Bazra M, Kane A, Baum J. Pharmakinetics of intravitreal carbenicillin, cefazolonine, and gentamicin in rhesus monkeys. Invest Ophthalmol Vis Sci 24:1602-1606,1983.
47Kane A, Bazra M, Baum J. Intravitreal injection of gentamicin in rabbits. Effect of inflammation and pigmentation on half-life and ocular distribution. Invest Ophthalmol Vis Sci 20:593-597, 1981.
48Ficker L, Meredith TA, Gardner S, et al. Cefazoline levels after intravitreal injection:Effects of inflammation and surgery. Invest Ophthalmol Vis Sci 31:502-505, 1990.
49Martin A, Ficker LA, Aguilar HA, et al. Vitreous cefazolin levels after intravenous injections. Effects of inflammation, repeated antibiotics doses, and surgery. Arch Ophthalmol 108:411-414,1990.
50Kern G, Alhalel A, Bartov E, et al. The intravitreal penetration of orally administered ciprofloxacin in humans. Invest Ophthalmol Vis Sci 32:2388-2392, 1991.
51Aguilar HE, Meredith TA, Shaarawy A, et al.Vitreous cavity penetration of ceftazidime after intravenous administration. Retina 15:154-159,1995.
52Pavan PR, Otzia EE, Huges BA, et al. Exogenous endophthalmitis initially treated without systemic antibiotics. Ophthalmology 101:1289-1297,1994.
53Campochiaro PA, Conway BP. Aminoglycoside toxicity:A survey of retina specialists:Implication of ocular use. Arch Ophthalmol 109:946-950,1991.
54Doft BH, Bazara M. Ceftazidime or amikacin:Choice of intravitreal antimicrobials in the treatment of postoperative endophthalmitis. Arch Ophthalmol 112:17-18, 1994.
55Hopen G, Mondino BJ, Kozy D, et al. Intraocular lenses and experimental bacterial endophthalmitis. Am J Ophthalmol 94:402-407,1982.
56Benson WE. Current management of postsurgical endophthalmitis. In:The Year book of Ophthalmology, Laibson PR, ed. Chicago, Year Book Medical Publishers Inc., 1989, pp. 181-184.
57Jacobs DR, Cohen KB. The inflammatory role of endotoxins in rabbit Gram-negative bacterial endophthalmitis. Invest Ophthalmol Vis Sci 25:1074-1079, 1984.
58Baum J. Intravitreal steroids in endophthalmitis (letter). Ophthalmology 99:301,1992.
59Peyman GA, Schulman JA. Endophthalmitis:Role of steroids. In:Intravitreal Surgery. Norwalk, Appleton and Lange, 1994. pp. 894-896.
60Graham RO, Peyman G. Intravitreal injection of dexamethasone:Treatment of experimentally-induced endophthalmitis. Arch Ophthalmol 92:149-154, 1974.
61Meredith TA, Aguilar E, Miler MT, et al. Comparative treatment of experimental Staphyloccal endophthalmitis. Arch Ophthalmol 108:857-860, 1990.
62Maxwell DP, Bren BD, Diamond JB, et al. Effect of intravitreal dexamethasone on ocular histopathology in a rabbit model of endophthalmitis. Ophthalmology 98:1370-1375,1991.
63Saimy N, Park SS, Ruoff KK, et al. Intravitreal steroids in the treatment of Pneumoccal endophthalmitis in the rabbit. Invest Ophthalmol Vis Sci 35:s3031, 1994.
64Abelson MB, Butrus S. In:Principles and Practice of Ophthalmology:Basics Sciences. Albert DA, Jacobiec FA, ed. Philadelphia, WB Saunders Co. 1994. pp. 1016-1022.
65Forster RK. In:Clinical Ophthalmology. Duane TD, ed. Philadelphia, Lippincott Co., 1994. pp. 13.
66Mandelbaum S, Forster RK. Postoperative endophthalmitis. Int Ophthalmol Clin 27:95-106, 1987.
67Irvine WD, Flynn HW Jr, Miller D, et al. Endophthalmitis caused by Gram-negative organism. Arch Ophthalmol 110:1450-1454,1992.
68Bohigian GN, Oik RJ. Factors associated with a poor visual results in endophthalmitis. Am J Ophthalmol 101:332-341, 1986.
69Allan HF, Mangiaracine AB. Bacterial endophthalmitis after cataract extraction. A study of 22 infections in 20,000 operations. Arch Ophthalmol 72:454-462, 1964.
70Kattan HN, Flynn HW Jr, Pflugfelder SC, et al. Nosocomial endophthalmitis survey:Current incidence of infection after intraocular surgery. Ophthalmology 98:227-238, 1991.
71Allan HF, Mangiaracine AB. Bacterial endophthalmitis after cataract extraction. II. Incidence in 36,000 consecutive operations with special reference to preoperative topical antibiotics. Arch Ophthalmol 91:3-7, 1974.
72Paymen GA, Daun M. Prophylaxis of endophthalmitis. Ophthalmic Surg 25:617-674, 1994.
73Gimble HV, Sun R, De Brof BM. Prophylactic intracameral antibiotic during cataract surgery:The incidence of endophthalmitis and corneal endothelial loss. Eur J Implant Ref Surg 6:280-285, 1994.
74Dickey JB, Thompson KD, Jay WM. Intraocular gentamicin sulfate and post-cataract anterior chamber aspirate cultures. J Cat Refract Surg 20:373-377, 1994.
75Speaker MJ, Menikoff JA. Prophylaxis of endophthalmitis with topical providone-iodine. Ophthalmology 98:1769-1775, 1991.
76Boes DA, Lindquist TD, Fritsche TR, et al. Effects of providone-iodine chemical preparation and saline irrigation on the perilimbal flora. Ophthalmology 99:1569-1574, 1992.
77Dickey JB, Thompson KD, Jay WM. Anterior chamber aspirate cultures after uncomplicated cataract surgery. Am J Ophthalmol 112:278-282, 1991.
78Cohen SM, Benner JD, Landers MB III, et al. Intraocular fluid cultures after primary pars plana vitrectomy. Am J Ophthalmol 114:697-699, 1992.
79Jett BD, Shepard B, Gilmore MS. Detection of Staphylococcus epidermidis and Staphylococcus aureus in vitreous by polymerase chain reaction. Invest Ophthalmol Vis Sci 36:s788, 1995.
80Hahn Tw, Osterhout GJ, O'Brien TP, et al. Utility of a rapid fluorescent Gram stain for recognition of live bacteria in ocular infection. Invest Ophthalmol Vis Sci 36:s797, 1995.