Ординатура / Офтальмология / Английские материалы / Small Incision Cataract Surgery (Manual Phaco)_Singh_2002

automated suction during routine vitrectomy) is replaced by a shorter tubing (about 2.5 cm in length) carrying at its one end a tuberculin syringe to enable manual suction. The vitreous cutter is placed in the anterior vitreous and cutting is actuated. Simultaneously, an assistant begins to withdraw the piston on the tuberculin syringe so as to induce suction. The procedure is stopped after 0.2 to 0.3 ml of sample has been obtained. The lost volume may be replaced by injecting saline (if no further vitrectomy has been planned) or by opening the preplaced infusion line when a vitrectomy has already been planned.

Whenever pars plana vitrectomy has been undertaken in a patient with endophthalmitis, the irrigating fluid admixed with the vitreous present in the cassette may also be sent in a sterile manner to the laboratory. The fluid can also be suctioned with aseptic precautions across a 0.45 μ millipore filter and the filter sent. In the laboratory the filter should be divided into fragments each of which is then transferred onto separate culture media. If the culture medium is solid, then care must be used to ensure that the filtered organisms are on the top surface and not trapped between the filter and the agar. An alternative method is to concentrate the vitreous by centrifugation.

The ideal recommended handling of samples obtained has been:

•Placing one drop of fluid on a blood agar plate, streaking carefully with a needle and incubating at 37°C.

•Placing one drop of fluid on chocolate agar plate, streaking carefully and incubating at 37°C in a 4 to 10 per cent CO2 enriched environment.

•Inoculating one drop into Sabourauds medium without any inhibitors and maintaining this at room temperature.

•Inoculate one drop onto thioglycolate broth and mix with the deeper, thicker portion of the broth with sterile cotton tipped applicator. This media is incubated at 37°C and it supports growth of anaerobes and microaerophilic organisms such as P. acnes.

•Place one drop on each of two clean glass slides for Gram and Giemsa stain. The smear made should neither be too thin nor too thick. The former tends to disperse the microbes and cells making microscopic study difficult and the latter takes up heavy staining making it impossible to detect organisms. Ideally the drop should be gently spread on a scrupulously clean glass slide using a clean bacteriological loop. The smear is then allowed to air dry. Heat fixation is to be avoided and absolute methanol used instead for

fixation. Centrifugal cytology has been reported as being superior but is not always necessary to undertake.

•Additional fluid available may be inoculated into brain-heart infusion or cooked meat broth.

A summary about the laboratory confirmation of diagnosis in endophthalmitis including the criteria for ‘laboratory confirmed growth’ is given in Appendix 1.

Since most cases of postoperative endophthal mitis occur in the early days following surgery, one has to pay adequate attention to the integrity of the wound before undertaking procedures like aqueous or vitreous aspiration and intravitreal injection. Most cases need a facial block and either topical anaesthesia or retrobulbar anaesthesia. General anaesthesia is recommended for children, un-cooperative patients and those with profuse congestion of the orbital tissue.

Treatment of Postoperative Bacterial Endophthalmitis

Having made a diagnosis of endophthalmitis, the patient is told about the diagnosis and the therapeutic interventions that may be necessary. He is also informed about the guarded visual prognosis and consent is obtained. Endophthalmitis can be managed as a daycare, outpatient emergency provided patient understands the need for frequent evaluation. Out-station patients and oneeyed patients may need to be admitted.

The three most important determinants in the outcome following endophthalmitis are:

•Time duration between presentation of symptoms, diagnosis of endophthalmitis and the initiation of appropriate treatment. In our country, late presentation of patients to their specialists is a frequent occurrence particularly in rural circumstances and this decreases any chances of visual recovery. In experimental animals it has been shown that intravitreal antibiotics are not able to salvage an eye when given 24 hours or later after introducing an innoculum of an infective organism into the vitreous cavity. It has also been reported that when intravitreal antibiotics are injected more than 48 hours after the infection is established the efficacy of the drug in eradicating the infection is poor.

•Virulence and load of the causative organism. Higher the organismal load entering the eye and greater the virulence of the organism, greater is the risk of developing endophthalmitis. However, the actual number of organisms necessary to incite endophthalmitis in humans is yet to be determined. The most virulent organisms responsible for post-surgical endophthalmitis are the gram-negative bacilli (Pseudomonas,

Klebsiella, Escherichia, Proteus) and gram-positive organisms like Staphylococcus aureus and streptococci.

•Pharmacokinetics and spectrum of activity of the intravitreal drugs: The goal in endophthalmitis management is to rapidly obtain adequate concentrations of the anti-microbial agent and maintain this for a sufficient period of time in the vitreous cavity without causing any toxic effect. This depends on several

factors concerning both the drug (original dose, route of egress from the eye, pH, ionization, molecular size, protein binding) and the ocular tissue (surgical status of the eye: presence or absence of the lens and vitreous, degree of breakdown of the blood retinal barrier).

At presentation, it would be useful to grade the severity of endophthalmitis as severe or not severe. Endophthalmitis is considered to be severe in the presence of the following: vision of inaccurate light projection, afferent pupillary defect, no fundus glow, limbal ring infiltrate (abscess) and cases that have not responded to appropriate intravitreal therapy. Some consider a vision of 20/400 and above as mild cases. However they also caution that vision may not be that useful a parameter to classify the severity of endophthalmitis as some patients may have poorer vision compared to the other milder clinical signs.

The primary objective in endophthalmitis treatment is to rapidly eradicate the colonization of the infecting organism within the vitreous cavity and also to prevent toxin and inflammation mediated damage to vital structures like the optic nerve (nerve fiber layer) and retina. The secondary objectives are to provide symptomatic relief, prevent synechia formation in miosis, remove any opaque membranes in the media (pupillary/vitreous). The ultimate objective is to maximize visual recovery. When the visual potential seems unlikely, one should atleast aim to sustain the structural integrity of the globe so as to prevent cosmetic disfigurement. The follow up of patients with endophthalmitis should be atleast every 12 hours initially and not any longer.

The mainstay of treatment in post-surgical endophthalmitis is administration of broad-spectrum intravitreal antibiotics. Vitrectomy is indicated in a highly selective category of patients and is discussed in detail later on. For a better understanding, the management options in post-surgical endophthalmitis include: Antimicrobial therapy (Intravitreal therapy, topical and systemic therapy); Anti-inflammatory therapy (Intravitreal, topical and systemic corticosteroids and NSAIDs);

Supportive therapy (cycloplegics, anti-glaucoma medication, etc.) and Surgical therapy (Vitrectomy).

ANTIMICROBIAL THERAPY

Earlier on, the most useful route for administering antibiotics to treat intraocular infections used to be very controversial. It is however, now unequivocally established that most antibiotics given systemically do not reach the minimum inhibitory concentrations necessary within the vitreous cavity. This is true despite the presence of a compromised blood ocular barrier in patients with endophthalmitis. Some present day antibiotics (Ciprofloxacin, Sparfloxacillin, and Pefloxacillin) have been shown to achieve significant concentrations in the vitreous cavity following systemic administration. However, the destruction progresses so rapidly in bacterial endophthalmitis that the concentrations may still be inadequate to rapidly curtail further growth of the organisms. The only route that is capable of achieving this objective is the direct administration of antibiotics into the vitreous cavity. Intravitreal route of administration however, has its own limitations and risks.

Intravitreal Antibiotics in Post-surgical Endophthalmitis

It would be most ideal to identify the causative agent, determine its antibiotic sensitivity and then administer specific antibiotics. This is not practical in treating endophthalmitis because the above process takes time and any time delay in these patients worsens the prognosis rapidly. Hence, the most important criterion while deciding on the choice of antibiotic for intravitreal administration is its spectrum of activity against the most common organisms known to produce endophthalmitis and also its known toxicity. The recommendation that one should administer a broad spectrum antibiotic does not mean that obtaining intraocular specimens, laboratory culture and determination of antibiotic sensitivity are no longer necessary. Contrarily, these measures become most important whenever there is a lack of response to the broad-spectrum antibiotic administered empirically at the start of treatment. Today, the preferred intravitreal antimicrobial therapy in postsurgical endophthalmitis is with a combination of two drugs, one having a broad spectrum of activity against gram-positive organisms and the other against gramnegative organisms (preferred combinations are mentioned subsequently). It is important however, that the two drugs to not have any antagonistic tendency. Two drug regimen is the preferred modality of treatment

because there is as yet no single drug that is highly effective against both gram-positive and gram-negative organisms and also has an adequate half life in the vitreous cavity.

Before giving an intravitreal injection any infected sutures or suture abscess should be removed. It is again emphasized that it is necessary to pay adequate attention to the wound integrity and the status of the lens (aphakic/ pseudophakic or phakic). The latter decides the site of pars plana entry and in a aphakic patient with a broken vitreous face, a translimbal route may be adopted. If obvious vitreous herniation or incarceration into the wound is present, then a limited anterior vitrectomy and wound revision may also be planned along with the intravitreal injection. Attention should also be paid to ensure that the intraocular pressure before the injection is not high and there is no pre-existing retinal or choroidal detachment (ultrasonography).

Intravitreal injection should be undertaken with all aseptic precautions by an experienced person or under guidance. The operating room Incharge and sisters should be informed and requested to arrange a trolley with the necessary instruments. It is always useful to have an assistant during the injection. The drugs to be given intravitreally should be prepared afresh by the eye surgeon himself, again with aseptic precautions. This is necessary to ensure proper dosage of the drug. While inadequate concentrations can lead to treatment failure, an excess dose can cause toxic effects on the retina. An informed consent is a must before giving an intravitreal injection.

The choice of anaesthesia should be decided beforehand taking into consideration factors mentioned earlier. In our view a facial block decreases the risk of vitreous upthrust by contraction of the orbicularis oculi during the actual injection of intravitreal drugs and should be used as a routine in all cases with an early postoperative endophthalmitis following conventional cataract surgery. Topical anaesthesia suffices in a large majority of cases with a healthy wound and retrobulbar injection is required less frequently. Peribulbar anaesthesia should be avoided, as it tends to increase orbital volume and pressure over the globe. No digital massage or other forms of mechanical pressure over the globe should be used. If needed, intraocular pressure may be lowered before the injection by giving the patient acetazolamide tablets. This is only rarely necessary as aspiration of intraocular fluids before intravitreal injection for culture and sensitivity serves to decrease the intraocular pressure.

The periocular region is painted with povidone-iodine and the cul-de-sac also washed with a solution of the same. Visualization of the operative site should be adequate. The injection is given transconjunctivally and no peritomy is necessary (if vitreous biopsy is not planned). It is important to choose the quadrant of injection (usually one that increases the ease of injection) and then measure and mark the distance from the limbus (3.0 mm if aphakic, 3.5 mm if pseudophakic and 4.0 mm if phakic) at which the injection is to be given. Aqueous and vitreous samples are then obtained as described earlier. Before actual injection of the drugs, the globe should be stabilized in an atraumatic manner. Use of fixation forceps is fraught with the danger of tearing of the inflammed conjunctiva and haemorrhage in a large number of patients. In cooperative patients, it is often possible to obtain adequate stability of the globe by placing a cotton-tipped applicator in the opposite quadrant. The surgeon then gradually inserts the 26-30 gauge needle on the tuberculin syringe containing the prepared drug, at the previously marked site. The bevel of the needle should be facing upwards towards the surgeon and the direction of penetration should be towards the direction of the anterior or mid-vitreous. The drug should then be injected slowly in a drop by drop manner (achieved by rotating the plunger) and avoiding jet formation. It is prudent to avoid making multiple entries into the eye and so the second injection should be given through the initial needle. The syringe can be carefully replaced with the one containing the second drug by asking the assistant to stabilize the needle by gripping its hub using a forceps. The intraocular pressure is checked at the end of the procedure. Subconjunctival injection of antibiotics is given and the eye patched. There is no need for a prolonged bandaging. Administration of topical drugs may be begun as early as one hour after the procedure. It has been suggested by some that reclining the patient with the head up immediately after the procedure may decrease the risk of the drug settling on the macula and preventing its toxic damage.

In the past, several drugs were used for intravitreal injection and several of these are rarely recommended today. The presently recommended combination therapy of choice in post-surgical bacterial endophthalmitis is:

*This was the preferred antibiotic combination in EVS study.

Preparation of the most frequently and less commonly

used intraocular drugs in the management of postsurgical endophthalmitis is shown in Appendix 2A and Appendix 2B respectively.

Vancomycin is a macrolide antibiotic that is highly effective against most gram-positive organisms including methicillin and cephalosporin resistant strains as well as coagulase negative staphylococci. It has been found to be safe when given even in a dose of 2 mg in 0.1 ml and also has a synergistic effect when used in combination with amikacin. In vitro, vancomycin when combined with ceftazidime in the same syringe is known to produce a precipitate and so they should be injected from separate syringes.

Ceftazidime is a third generation cephalosporin that has been found to have a bactericidal effect against a wide range of gram-negative organisms including pseudomonas. Unlike with aminoglycosides no drug resistance has so far been reported, no retinal toxicity has been found in the recommended intravitreal dose and it has been found more effective in acidic and hypoxic conditions. It has no activity against gram-positive organisms.

Amikacin is the preferred aminoglycoside because it is effective against gram-negative organisms resistant to other aminoglycosides and its retinal toxicity is four times less than that with gentamicin. Since ceftazimide has a similar bactericidal effect and no retinal toxicity exists, many now prefer this drug to amikacin in the first line management of postoperative bacterial endophthalmitis. Gentamicin is only rarely recommended because of the increased likelihood of macular infarction and also because of a high degree of drug resistance.

Quinolones such as ciprofloxacin have also been evaluated as a single drug treatment regimen in postoperative endophthalmitis. However, they suffer from the disadvantage that their half-life in the vitreous cavity is less and so a repeat injection becomes necessary within 12-24 hours in order to obtain a therapeutic response.

Penicillins, erythromycin and even the first and second generation cephalosporins are no longer recommended as the first line drugs in the management of endophthalmitis because of the existence of a significant degree of drug resistance and their limited range of anti-bacterial activity.

Intravitreal treatment carries with it a risk of the following complications: elevated intraocular pressure,

intraocular hemorrhage (including hyphema), drug induced retinal toxicity and retinal detachment. In phakic eyes an added risk is that of cataract due to contact by the needle.

Intravenous Antibiotics in

Post-surgical Bacterial Endophthalmitis

Even though it is possible that the vitreous levels of antimicrobial drugs given systemically may increase in endophthalmitis because of the associated inflammation and resultant breakdown of the blood-retinal barrier, the role of “Intravenous antibiotics alone” in the management of endophthalmitis, is at best, supportive and not primary. This is supported by several observations:

•Animal studies have found poor intraocular penetration of most parenterally administered antibiotics. Exceptions are some newer antibiotics like imipenem, cephazolin, ciprofloxacin and sparfloxacin and pefloxacin. Aminoglycosides in particular have very poor intraocular penetration following parenteral administration.

•73 per cent of 103 patients with endophthalmitis did not regain any useful vision after conventional treatment (systemic antibiotics alone)

•Study by Pavan et al showed good results with the use of only intravitreal drugs without parenteral antibiotics

•Conclusion from EVS study that whether or not systemic antibiotics are used there is no difference in final visual acuity

Added to the above observations, systemic administration of antibiotics also has the disadvantages of cost effectiveness and potential for systemic adverse drug reactions. Another concern with the irrational use of newer antibiotics parenterally is the risk of encouraging drug resistance. Despite these facts however, a large majority of eye specialists involved in treating endophthalmitis still prefer to use parenteral antibiotics in addition to intravitreal injection, as it is possible that it may help in augmenting and sustaining an adequate concentration of the antibiotics in the vitreous cavity for a more prolonged period. The recommended dose of antibiotics for systemic administration in endophthalmitis is given in Appendix 3A.

Topical and Subconjunctival Antibiotics in

Post-surgical Bacterial Endophthalmitis

For topical medication in endophthalmitis, a combination of two drugs is preferred, one having a predominant effect on the gram-positive organisms and the other against

gram-negative organisms. The frequency of administration is every hour (with each drug used alternately) in the initial phases of treatment. This is modified depending on the response to overall measures. In the presence of a corneal ulcer or wound abscess, fortified eyedrops are recommended. The antibiotic drugs used in the EVS study were vancomycin (50 mg/ml) and amikacin (20 mg/ml). The method of preparing fortified eyedrops is given in Appendix 3B.

In a patient complaint to the regimen of topical medication prescribed, subconjunctival antibiotic injection may have only a limited role. Subconjunctival injection is not routinely used by us in managing patients with endophthalmitis. In addition to patient discomfort, tearing of conjunctiva and subconjunctival hemorrhage, there is also a risk of intraocular injection of the drug this procedure. The recommended dose of commonly used antibiotics for subconjunctival injection is shown in Appendix 3C.

ANTI-INFLAMMATORY THERAPY: ROLE OF CORTICOSTEROIDS

One of the major causes of tissue damage in bacterial endophthalmitis is the release of inflammatory mediators in large quantities. This occurs because of invasion by the organisms and also because of several endotoxins released by them. These have an ability to stimulate the complement pathway and also the arachidonic pathway releasing both leukotrienes and prostaglandins. These inflammatory mediators have chemotactic properties and so attract polymorphonuclear leukocytes and macrophages into the vitreous cavity. Proteolytic and collagenolytic enzymes released from leukocytes are also harmful to the highly sensitive intraocular tissues.

Corticosteroids decrease the risk of tissue damage resulting from the above mechanism by inhibiting both the lipo-oxygenase and cyclo-oxygenase pathways of arachidonic acid metabolism. Corticosteroids remain the most potent anti-inflammatory agents known. For being most useful they have to be given early and in adequate doses. They should however not be used whenever a fungal infection is suspected as they enhance fungal growth by decreasing the defense mechanisms within the body.

In post-surgical bacterial endophthalmitis, corticosteroids may be given systemically, injected into the vitreous cavity and also used as topical drops and for subconjunctival injection. The use of systemic corticosteroids is well-accepted but there is a persisting controversy regarding the necessity and role of intravitreal corticosteroids in improving the visual results in endo-

phthalmitis. In addition, it is not known if oral or intravenous administration is as effective as intravitreal injection.

The disadvantages of intravitreal corticosteroids is the possibility that it may reduce the ability of the eye to sterilize the innoculum of microorganisms it has encountered. They have been found to have little effect on the damaging effect of bacterial toxins on the retina. Infact, one experimental study on S. aureus endophthalmitis, has shown increase in inflammation, retinal necrosis and corneal opacification following intravitreal injection of steroids. In contrast to the above findings, there are reports in literature which show clearing of the vitreous in 9 of 20 patients given 1200 μg of corticosteroids intravitreally.

Hence, the use of intravitreal injection of steroids remains a decision left to the discretion of the surgeon.

The recommended dose of corticosteroids in bacterial endophthalmitis for intravitreal, systemic and subconjunctival injection is given in Appendix 4.

SUPPORTIVE THERAPY

Certain medications act as adjuncts in the management of patients with endophthalmitis. These include cycloplegics and drugs to lower any elevation of intraocular pressure that may be seen in some patients. Cycloplegics are an important part of the treatment and as a general guideline, the strongest cycloplegic is to be prescribed in severe cases of endophthalmitis. We use atropine 1% ointment 8 hourly initially and then change to either, homatropine 2% drops 4-8 hourly or to a combination of atropine and prednisolone eyedrops 6 hourly. Apart from enabling control of inflammation and relieving ciliary spasm, cycloplegics prevent synechia formation in miosis and increase the chances of having a dilated pupil. Presence of a dilated pupil not only enables better clinical evaluation but also becomes an asset if a need for performing a vitrectomy arises.

In patients with elevated intraocular pressure, drugs such as oral acetazolamide and timolol may be prescribed. A vitreous tap before giving an intravitreal injection may also help to lower the intraocular pressure at least transiently in some patients.

VITRECTOMY IN POST-SURGICAL ENDOPHTHALMITIS

Vitrectomy is the second line of management approach in endophthalmitis with specific and definite indications. It is technically more demanding, needs experience and has a potential to lead to complications, particularly retinal detachment. Also, it may not always be rewarding

in terms of the functional outcome. Vitrectomy for endophthalmitis may be necessary at two stages, primary, during the acute infection or secondary in the resolved phase for vitreous opacification or membranes. Vitrectomy in the later stage is less demanding and less likely to cause complications in comparison to vitrectomy the primary, acute stage of endophthalmitis.

The timing “When to do vitrectomy” in endophthalmitis is difficult to decide and a controversial issue because choosing a time that is neither too late nor too early is subjective, usually based on the surgeon’s past experiences and also because the benefit versus safety window for this procedure is very narrow.

The advantages of vitrectomy are that it decreases the infectious, toxic and inflammatory load; provides adequate undiluted specimen for culture studies; increases antibiotic concentration within the eye and by removing media opacities enables a more rapid visual recovery. In reality however the functional outcome may be less than that theoretically possible because of a surgical bias in undertaking vitrectomy only in the more severe and advanced cases. As a result of this bias not only does the surgery become more difficult but also the risk of complications like retinal detachment increases and so also the possibility of a relapse.

Peroperative problems that a surgeon faces during vitrectomy for endophthalmitis are a poor visualization due to an edematous cornea or IOL membranes, a ring abscess around the limbus that makes any attempted suturing of an incision here impossible, an absent and often adherent and “Sticky” vitreous without PVD, an inflamed or necrotic retina which is easily liable to develop tears and the risk of operating in the presence of congested choroidal vasculature.

During surgery, a 6 mm infusion cannula should be preferred, the MVR blade used for making the sclerotomies must be sharp and the cutter must not cause a drag on the vitreous fibrils. The three cardinal principles to be remembered while doing vitrectomy for endophthalmitis are: Use maximal cutting rate, minimal suction and do not attempt to induce a PVD if it is not already present. No attempt should be made to go very close to the retina. At the end of vitrectomy, it is generally recommended that intravitreal antibiotic injections be given (1/10th of normal recommended dose).

The indications recommended in literature for undertaking immediate vitrectomy are severe cases, cases in which gram-negative organisms are seen on a smear examination of the vitreous aspirate and in cases showing

no response to medical treatment (intravitreal injection). A severe case has been defined as one in which there is a total absence of red reflex, an inaccurate projection of light, an afferent pupillary defect, a corneal ring infiltrate and a patient worsening 24-48 hours after an intravitreal injection.

The most common approach that mostly decides the time that vitrectomy should be undertaken is a worsening despite a proper intravitreal injection or a lack of response to two repeat intravitreal injections. This is also the approach adopted at our centre.

The only prospective randomized controlled evaluation of intravitreal and vitrectomy procedures for endo– phthalmitis is the endophthalmitis vitrectomy study. Details of this study are discussed in the subsequent section. The goal of pars plana vitrectomy in the EVS was to remove at least 50 per cent of vitreous gel in eyes with no vitreous separation.

The EVS did not answer the question of when to undertake vitrectomy, however, it clearly answered the question of when to undertake additional surgery be it a repeat intravitreal or repeat vitrectomy. According to this study a repeat intervention should be undertaken in eyes doing poorly 36-60 hours after the first intervention (vitrectomy or intravitreal). At 3-9 months after the interventions, when they assessed the visual acuity and media clarity in the various groups, they found that:

•If initial vision is hard movements or better there was no difference in the outcome between immediate vitrectomy and intravitreal injection groups.

•However, if the initial vision was only light perception, the final visual acuity and media clarity was substan-

tially better in patients undergoing vitrectomy.

This study indirectly indicates that one very certain indication to immediately undertake pars plana vitrectomy in patients with postoperative (post-cataract surgery) endophthalmitis is when the initial vision is light perception only.

ENDOPHTHALMITIS VITRECTOMY STUDY

Endophthalmitis Vitrectomy Study (EVS) was a multicentric study undertaken in the United States and involving 420 patients who had developed bacterial endophthalmitis within 6 weeks of cataract surgery or secondary IOL implantation. The primary objective of the study was to determine the role of early pars plana vitrectomy in comparison to intravitreal injection alone (TAP) in patients with endophthalmitis and also to identify the role of systemic antibiotic treatment in these cases.

The main objective outcome measures determined were improvement in visual acuity and improvement in media clarity at the end of 3-9 month follow-up.

The other criteria for inclusion in the study was a visual acuity of at least or more than light perception and less than 20/50 and with a relatively clear cornea and anterior chamber. Patients with a history of other intraocular surgeries, presentation after 6 weeks, fungal endophthalmitis, trauma and age below 18 years were not included in the study. Other exclusion criteria included previous intraocular antibiotic administration, other causes of poor vision, presence of retinal or choroidal detachment and drug sensitivity to lactams.

The patients were randomly categorized into four groups: PPV with systemic antibiotics, PPV without systemic antibiotics, TAP with systemic antibiotics and TAP without systemic antibiotics.

An important factor during this study was in the assessment of vision when it was less than finger counting at 1m. As mentioned earlier light projection was tested at a distance of 90 cm with the strongest intensity of light from an indirect ophthalmoscope. Hand movements vision was tested at a distance of 60 cm with a light source from behind the patient. The stimulus of hand movements was presented five times and the response was recorded as positive if four of these were correctly identified. Media clarity was graded as indicated earlier.

Patients were only admitted for management when it was felt that topical medication may not be administered regularly, in presence of transportation difficulties and whenever pars plana surgery was contemplated.

The most common symptoms at presentation was blurred vision (94%) followed by pain (74%). The mean interval between surgery and onset of symptoms was 4 days and visit to a vitreous surgeon, 6 days. Treatment was initiated in all patients within 6 hours of presentation and after having obtained diagnostic samples from the eyelid, anterior chamber (0.1 ml using 25-27 gauge needle) and vitreous (0.2 ml by aspiration or biopsy). These specimens were cultured on chocolate agar (37°C in CO2), fresh enriched thioglycolate (37°C) and fresh Sabourauds dextrose agar (25°C). Laboratory confirmed growth was defined as mentioned earlier. and using this definition, no growth was obtained in 18 per cent cases, equivocal growth was seen in 13 per cent and positive growth in 69 per cent. The breakup of those with a positive culture was: gram-positive, coagulase negative species in 47 per cent, other gram positive organisms in 16 per cent, gram-negative in four per cent and more than one species in three per cent cases. When culture

growth was co-related with the final visual acuity it was found that in patients with a gram-positive, coagulase negative organismal growth, 62 per cent achieved more than 20/50 vision while only 55 per cent did so in patients with no growth on culture.

The intravitreal drugs used in this study were vancomycin (1000 μg in 0.1 ml) and amikacin (400 μg in 0.1 ml). No intravitreal corticosteroids were given. For systemic (parenteral) administration the chosen drugs were ceftazidime (2 gm every 8 hourly) and amikacin (7.5 mg/kg initially and then 6 mg/kg every 12 hours). In those allergic to lactams, ciprofloxacin 750 mg orally twice daily was used as an alternative. Systemic medication was given for a period of 5-10 days and left to the treating physicians discretion.

For subconjunctival injection the drugs chosen were vancomycin (25 mg/0.5 ml), ceftazidime (100 mg/0.5 ml) and dexamethasone 6 mg/ 0.25 ml. Topical antibiotic medications included vancomycin (50 mg/ml) and amikacin (20 mg/ml) given every 4 hourly routinely or alternately every 1 hourly if wound leak or infection was present.

The surgical interventions performed were vitreous tap and intravitreal injection or vitreous biopsy and intravitreal injection or pars plana vitrectomy and intravitreal injection. The goal of vitrectomy in eyes with no obvious vitreous separation was to remove atleast 50 per cent of the vitreous gel.

Additional surgery (re-vitrectomy, re-vitreous tap or vitrectomy) was undertaken in eyes doing poorly 36-60 hours after the first intervention. Signs of worsening were an absent red reflex or increasing opacification, a 1 mm increase in the height of hypopyon, development of a corneal ring infiltrate and worsening pain.

The major conclusions of the EVS study were:

•If initial vision is hand motions or better then there is no difference in outcome between immediate vitrectomy or intravitreal antibiotics

•If initial vision is only light perception then final visual acuity and media clarity are substantially better in patients undergoing vitrectomy and intravitreal injection as compared to intravitreal injection alone

•Whether or not systemic antibiotics are used there is no difference in final visual acuity

In subsequent reports from the EVS group other important observations were made. These include:

•The vitreous is a richer source of laboratory confirmed growth

•Gram stain should not determine the choice of antibiotic drugs

•Vitrectomy with culture of vitrectomy cassette fluid did not produce significantly more positive cultures than vitreous tap or biopsy material and so the procedure (i.e. vitrectomy) should not be performed solely to improve the microbiological yield.

•Secondary or anterior chamber IOL implantation was associated with a possible shift in the spectrum of organisms isolated towards gram-positive organisms other than coagulase negative micrococci.

•Vancomycin was active against all gram-positive isolates tested; amikacin and ceftazidime showed equivalent activity against gram-negative isolates

•A positive Gram stain or infection with species other than gram-positive, coagulase negative micrococci is significantly associated with poorer visual outcomes.

•Although visual prognosis is strongly associated with the type of infecting organism and Gram stain positivity, presenting visual acuity remains more powerful than microbiologic factors in predicting visual outcome and favorable response to vitrectomy.

POST-SURGICAL FUNGAL ENDOPHTHALMITIS Clinical Features

Patients may present either with a blurring or decrease in vision associated with some degree of redness or with complaints of floaters. Most cases of fungal endophthalmitis following intraocular surgery usually appear several weeks to months later. The characteristic feature is a relative lack of symptoms compared to the signs on examination of the eye. The presentation is that of chronic endophthalmitis with indolent inflammation. The anterior chamber may show a fixed hypopyon and a fibrinous mesh-like exudation. There is variable degree of corneal oedema, the intraocular pressure is frequently elevated and the vitreous may show snow balls and fluffy opacities. Infrequently the fungal colonies may be misdiagnosed as retained lens matter and prescribed steroids. Fundus glow is variable but vision may be less than expected. Clinically it may be sometimes difficult to distinguish fungal endophthalmitis from P. acnes endophthalmitis.

Confirmation of Diagnosis

Material for culture in fungal endophthalmitis is as for post-surgical bacterial endophthalmitis. However it may be more difficult to aspirate as the colonies are usually tenacious. The routinely used culture media for fungal growth is Sabourauds dextrose agar. Although colonies may begin to appear within a few days, it should be

observed for atleast 2 weeks before reporting it as negative. If Sabourauds media is unavailable or the specimen is inadequate, the laboratory may be told to hold the blood agar plate for 3 weeks and observe for fungi.

MANAGEMENT OF POSTOPERATIVE FUNGAL ENDOPHTHALMITIS

The objectives of treatment in fungal endophthalmitis remains the same as that in bacterial endophthalmitis. However, the results are not gratifying because of several factors like delayed diagnosis, lack of non-toxic fungicidal drugs and the inadequacy of intravitreal injection alone in the treatment. The only agreement is that steroids in form are absolutely contraindicated in fungal endophthalmitis.

Systemic Antifungal Therapy

Unlike metastatic endophthalmitis caused by fungi, postoperative fungal endophthalmitis poses a peculiar problem in deciding the route of drug administration. In the former, there is an associated fungaemia and usually extraocular sites of fungal colonization, hence, systemic antifugal therapy is easily justified. This is not so in the case of postoperative fungal endophthalmitis wherein the fungal colonization is limited to the intraocular cavities. Since most anti-fungal drugs have the potential to cause significant adverse effects, the role of systemic therapy in such a situation may seem questionable. Moreover, not all of these drugs have a good enough penetration into the intraocular cavities. The problem has been solved to a certain degree by the discovery of certain anti-fungal drugs causing fewer systemic adverse effects and also having a better intraocular penetration.

Exact management guidelines for the management of exogenous fungal endophthalmitis are not available in literature. The relative paucity of controlled studies on the treatment approach in this form of endophthalmitis is possibly related to the infrequent occurrence of fungal endophthalmitis in comparison to bacterial endophthalmitis after intraocular surgery.

There are three major groups of anti-fungal drugs available in the market and it is important for us to know their limitations, advantages and potential for toxicity. These three groups are: Polyenes (Amphotericin B), Azoles (Ketoconazole, Miconazole, Fluconazole, Itraconazole) and Fluocytosine.

Amphotericin-B Parenteral amphotericin-B has been considered the treatment of choice in intraocular fungal

infections. However, this form of treatment has several disadvantages such as poor intraocular penetration and the potential for both systemic adverse effects and retinal toxicity.

Amphotericin-B may be fungistatic or fungicidal depending on the concentration of the drug within the tissues and the susceptibility of the fungi. It destroys fungi by causing changes in the permeability of their cell membranes. It is recommended that before starting patients on systemic amphotericin-B, a test dose be given. This is usually with 1mg of the drug in 20 ml of 5 per cent dextrose (not normal saline) given intravenously over half an hour. Therapy should not be undertaken if the patient develops serious side effects like hypotension or cardiac arrhythmias with the test dose. In the absence of any adverse effects, treatment can be started one hour later with a dose of 0.7 mg/kg of amphotericin in 500 ml of 5 per cent dextrose by slow intravenous infusion over 2-6 hours. Although more rapid administration (within 1-2 hours) have been advocated, it may carry a greater risk. Previously the subsequent approach was to increase the dose gradually (5-10 mg) each day. However may be better to try and achieve the maximum dose of 0.7- 1.0 mg/kg/day as early as possible. No treatment response is usually seen in the first week. Subsequently, the inflammation begins to gradually subside. Treatment should be continued until the inflammation and any chorioretinal lesions present, show complete resolution.

Toxic adverse effects known to occur with intravenous administration of amphotericin-B are anaphylaxis, convulsions, phlebitis, chills, fever, headache, anaemia and thrombocytopenia. Hypotension and cardiac arrhythmias are other serious side effects. The most common and significant adverse effect is nephrotoxicity. This is reported to occur in 80 per cent of patients and so constant monitoring of renal parameters is a must during treatment with parenteral amphotericin-B treatment.

Azole derivatives These newer group of drugs have the advantages of adequate systemic absorption following oral administration, better penetration into the intraocular cavities (fluconazole>ketoconazole>itraconazole) and lesser risk of serious adverse effects. To be effective however, they have to be given early. The only drug shown to be effective in reducing fungal counts when given after 7 days of fungal inoculation in experimental animals has been ketoconazole.

The disadvantage of azole derivatives are that they are not as effective as amphotericin B, resistance to the drug may develop rapidly and they have a significant antagonistic effect when combined with amphotericin-B.

Moreover, they only have a fungistatic effect. In comparative studies with amphotericin-B and fluoconazole in experimental animals it was seen that the initial response (for the first 17 days) to treatment was identical in both the groups. From the 21st day onwards however, the fluconazole group began to again worsen probably because of the development of drug resistance. As combined therapy with amphotericin-B and azole derivatives has been shown toincreasetheriskofdevelopingresistancetoamphotericin B, the use of this form of combination therapy is not recommended.

The usual dose of the usually preferred azole derivatives in the treatment of intraocular fungal infections is: Ketoconazole (400 mg/day in a single or two divided doses) and Fluconazole (200 mg/day in a single or two divided doses).

If azole derivatives are chosen as the first line of treatment in the management of fungal endophthalmitis, it would be probably prudent to not persist with the treatment if no response is observed within the first 7 to 10 days. This is also probably necessary if the condition begins to worsen after an initial response as seen in the experimental study mentioned earlier. Under both these circumstances one should change to treatment with amphotericin B despite its known adverse effects. Constant interaction with an internist to monitor for toxic effects and modify dosage of the drug accordingly however, becomes very essential.

An important part of the management is to remember that corticosteroids by any route are absolutely contraindicated in the management of fungal endophthalmitis.

Flucytosine Treatment with drug alone in the management of fungal infections is not recommended because of the rapid development of drug resistance. However, it may be used in combination treatment with amphotericin B when the intraocular inflammation is severe and resistant to initial treatment. Fluocytosine is given orally in a dose of 50-100 mg/kg/day in four divided doses. This drug can cause hematologic, renal and hepatic toxicity.

Dose of various anti-fungal agents for systemic administration is shown in Appendix 5A.

INTRAVITREAL ANTIFUNGAL THERAPY

Unlike in postoperative bacterial endophthalmitis where the mainstay of treatment is intravitreal administration, in fungal endophthalmitis, this mode of drug administration acts only as an adjunct to systemic medication. This is because, fungi unlike bacteria multiply less rapidly and correspondingly treatment of fungal endophthalmitis requires a prolonged duration (in weeks and not days)

of adequate concentrations of the anti-fungal agent within the vitreous cavity. This objective cannot be achieved with intravitreal administration. A probably more efficacious way of achieving this objective in future could be the development of sustained release intraocular devices similar to those presently in use to treat patients with cytomegalovirus retinitis (e.g. Ganciclovir intraocular device—GIOD).

The method of injection and the precautions while administering an intravitreal injection is the same as those described under postoperative bacterial endophthalmitis. However, the follow-up guidelines (e.g. when to repeat the injection/how long to wait before vitrectomy, etc.) following this injection are not clearly defined. Steroids are absolutely contraindicated.

Preparation of Amphotericin-B for intravitreal injection is given in Appendix 5B.

Vitrectomy in Fungal Endophthalmitis

The only differences with regard to vitrectomy for bacterial and fungal endophthalmitis are that in the latter it is generally recommended that it be performed early and an antifungal injection (usually Amphotericin B) given into the vitreous at the end of the surgery.

At our center, in cases of fungal endophthalmitis, we prefer to initially give a trial of azole derivatives (Fluconazole/Ketoconazole) for 10-14 days. If this fails, we subject them to vitrectomy and continue with the above drugs. If still the exudation reappears an intravitreal injection of Amphotericin-B 5 μg is given. In general the prognosis has not been very satisfactory in patients with postoperative fungal endophthalmitis.

PROPIONIBACTERIUM ACNES ENDOPHTHALMITIS

In patients with chronic postsurgical endophthalmitis,

Propionibacterium acnes has been reported to be the most common isolate. Propionibacterium acnes is an anaerobic, gram-positive bacillus that is normally present in the conjunctival sac as a commensal. Though considered to have no pathogenic potential a few decades ago, this organism is now known to a frequent cause of chronic bacterial endophthalmitis. A few cases of endophthalmitis by Propionibacterium granulosum have also been reported.

Propionibacterium acnes has an ability to stimulate the immune system but being resistant to killing by the polymorphs and monocytes, it remains intracellularly after phagocytosis by macrophages. The organism probably acts as an adjuvant in stimulating an immune

response against soft lens matter remaining after cataract surgery.

Propionibacterium acnes endophthalmitis presents as a low grade, smoldering type of intraocular inflammation following cataract surgery. Because of this pattern of inflammation, it tends to be missed in its early stages. The inflammation may show an initial response to steroids but subsequently begins to recur. Endophthalmitis caused by P. acnes has certain classical clinical features such as the presence of whitish plaques in relation to the capsular bag, history of unexpected inflammation after YAG capsulotomy and the frequent tendency to relapse following initial response. The whitish plaques are composed of colonies of P. acnes. Laboratory confirmation of the diagnosis is also difficult and often delayed because the organism begins to show up on anaerobic culture media only after about 2 weeks. The whitish plaques are formed by both sequestered colonies of the organism as well as inflammatory cells.

Acute cases of endophthalmitis caused by P. acnes infection have also been reported. These cases are few and they differ from the chronic form in showing a gratifying response to treatment. Recurrences are not known to occur following acute infection with P. acnes infection. This is probably because the organisms have not become sequestered within the capsular bag.

For laboratory confirmation of P. acnes infection, the organism should be cultured on anaerobic media and observed for growth for atleast 14 days.

Management of endophthalmitis caused by P. acnes poses a peculiar challenge. Although the virulence of the organism is low and it does not usually produce a severe inflammatory response, it is difficult to eradicate because it can remain sequestered within the capsular bag. It would probably not be possible for the antibiotics to reach these spaces and remain there at the needed concentrations for a sufficient duration of time. Controversies exist in the management approach in this form of endophthalmitis due to clinical variables such as the severity of infection.

Treatment approaches that have been indicated in literature are intravitreal antibiotics alone in mild cases, to vitrectomy, total capsulectomy and IOL explantation along with intraocular and systemic antibiotics in very severe cases. Intravitreal injection for P. acne endophthalmitis has to be given into the capsular bag to be of some value. The objective of total capsulectomy and IOL explantation is to eradicate colonies of P. acnes sequestered within the confines of the capsular bag. Surgical measure like partial capsulectomy and retaining of the IOL may provide only transient respite. The antibiotic of