`Retinal Toxicity, Clearance, and Interaction With Gentamicin
`Stephen C. Pflugfelder, MD; Eleut Hern\l=a'\ndez;Steven J. Fliesler, PhD; Juana Alvarez;
`Maureen E. Pflugfelder, MT (ASCP); Richard K. Forster, MD
`
`\s=b\Some of the gram-positive isolates
`from exogenous bacterial endophthalmi-
`tis cases treated at our institution have
`been found to be resistant to either cefa-
`zolin sodium, gentamicin sulfate, or both.
`However, all of these isolates have been
`sensitive to vancomycin. These findings
`prompted us to reevaluate the retinal tox-
`icity and clearance of intravitreal vanco-
`mycin in pigmented rabbits. Doses up to 2
`mg were found to be nontoxic in phakic
`and aphakic-vitrectomized eyes. Clear-
`ance was determined in phakic and
`aphakic-vitrectomized rabbit eyes with or
`without intact lens capsules. The antibiot-
`ic was cleared most slowly in phakic eyes.
`Aphakic-vitrectomized eyes without an
`lens capsule cleared antibiotic
`intact
`most rapidly, while aphakic-vitrectomized
`eyes with intact capsules exhibited an
`In addition,
`intermediate clearance rate.
`the interaction between vancomycin and
`gentamicin on gram-positive endophthal-
`mitis isolates was found to be additive or
`synergistic depending on the bacterial
`species. Based on these data, we recom-
`mend the combination of vancomycin and
`an aminoglycoside as the initial antibiotic
`therapy for exogenous bacterial endoph-
`thalmitis.
`(Arch Ophthalmol 1987;105:831-837)
`
`Oixty-six percent of
`the exogenous
`^ endophthalmitis cases from our
`institution have been caused by gram-
`positive bacteria, and 25% have been
`by
`caused
`streptococcal
`species.13
`Antibiotic sensitivities for these iso¬
`lates indicate that some of the orga-
`
`Accepted for publication Feb 26, 1987.
`From the Department of Ophthalmology, Uni-
`versity of Miami School of Medicine.
`Reprint requests to Bascom Palmer Eye Insti-
`tute, PO Box 016880, Miami, FL 33101 (Dr Pflug-
`felder).
`
`nisms, particularly streptococci, are
`resistant to cefazolin sodium and geñ¬
`tamicin sulfate,
`the two antibiotics
`most often recommended for initial
`therapy.45 Yet, 100% of
`intravitreal
`the gram-positive endophthalmitis
`isolates
`(including
`Streptococcus,
`Staphylococcus, and Bacillus species)
`were sensitive to vancomycin.
`Vancomycin has previously been
`evaluated for endophthalmitis thera¬
`In 1967, Pryor and colleagues6
`py.
`found that therapeutic aqueous levels
`could be achieved after topical or sub¬
`conjunctival administration of vanco¬
`mycin; however, vitreous levels were
`negligible. Homer and associates7
`examined the retinal
`toxicity and
`clearance of intravitreal vancomycin
`in rabbits and found that an intravi¬
`treal dose of 1 mg was nontoxic to the
`sensory retina. Prolonged therapeutic
`levels were achieved after a single
`injection of 1 mg, and
`intravitreal
`vancomycin was found to be effective
`in treating experimental methicillin-
`resistant Staphylococcus aureus endo¬
`phthalmitis. Recently, Smith et al8
`also reported success
`in treating
`methicillin-resistant
`experimental
`Staphylococcus epidermidis endoph¬
`thalmitis with vancomycin.
`In that
`study, a 5-mg intravitreal dose of van¬
`comycin hydrochloride was found to
`be nontoxic in phakic albino rabbits.
`To recommend an intravitreal dose
`for the treatment of exogenous bacte¬
`rial endophthalmitis, we reevaluated
`the retinal
`toxicity of
`intravitreal
`vancomycin in phakic and aphakic-
`vitrectomized pigmented rabbit eyes.
`We also examined the clearance of
`intravitreal vancomycin in phakic and
`aphakic-vitrectomized eyes both with
`and without
`intact
`lens capsules.
`Finally, because of the potential ther¬
`apeutic efficacy of vancomycin in con-
`
`junction with an aminoglycoside anti¬
`biotic, we evaluated the interaction of
`this combination on common endoph¬
`thalmitis isolates by a microdilution
`checkerboard assay.
`MATERIALS AND METHODS
`Toxicity
`All studies were performed using Dutch
`Belted pigmented rabbits (weight range,
`2.5 to 3.5 kg) with an average vitreous
`volume of 1.4 mL. The toxicity of vancomy¬
`cin was evaluated in both phakic and
`aphakic-vitrectomized eyes. The following
`doses were evaluated: in phakic eyes, 0.5,1,
`in aphakic-vitrectomized
`2, 3, and 5 mg;
`eyes, 1, 2, and 5 mg. Three or more eyes
`were examined for each antibiotic dose. All
`doses of antibiotics were prepared by a
`pharmacist in 0.1 mL of preservative-free
`sterile water. Antibiotics were adminis¬
`tered into the vitreous cavity via the pars
`plana, 2 to 3 mm posterior to the limbus,
`using a 1-mL syringe with a 27-gauge
`needle. The injection was performed slowly
`with the bevel of the needle facing anteri¬
`orly. Antibiotic injection was performed
`only in the left eye of each animal;
`the
`right eye served as a control. Rabbits were
`anesthetized with a mixture of ketamine
`hydrochloride (14 mg/kg) and xylazine (7
`mg/kg) prior to the injection. The rabbits
`were killed with a lethal intravenous injec¬
`tion of pentobarbital sodium.
`Lensectomy was performed by a pars
`plana approach using an ultrasonic lens
`fragmenter.9 The anterior lens capsule was
`retained in some eyes for the toxicity and
`clearance portions of the study. Pars plana
`vitrectomy was performed using an auto¬
`mated vitreous suction-cutter.9 Eyes were
`treated with 1% atropine ointment once a
`day for two to four weeks after surgery
`before they were used to evaluate toxicity
`and clearance. Sensory retinal toxicity was
`evaluated by direct observation, electroret-
`inography, and light and electron micros¬
`copy. Indirect ophthalmoscopy of the rab¬
`bit fundus after dilation was performed at
`the following times after intravitreal van¬
`comycin injection: five minutes, 24 hours,
`
`Downloaded From: https://jamanetwork.com/ by Judith Little on 01/07/2022
`
`Exhibit 2066
`Page 01 of 07
`
`
`
`Color Fig 1.—Normal-appearing fundus three days after injection of 2
`mg of vancomycin into vitreous cavity of phakic pigmented rabbit
`eye.
`
`Color Fig 2.—Retinal opacification three days after injection of 5 mg of
`vancomycin into vitreous cavity of phakic pigmented rabbit eye.
`
`Color Fig 4.—Retina
`and pigmented epi¬
`thelium after 5-mg
`intravitreal vancomy¬
`cin dose in aphakic-
`pig¬
`vitrectomized
`mented rabbit eye.
`Note disorganization
`of photoreceptor
`in¬
`and
`segments
`ner
`loss of photoreceptor
`segments.
`outer
`Bar
`represents
`20
`(toluidine blue,
`µ \
`X1000).
`
`Color Fig 3.—Normal-appearing retina and pigment epithelium of
`phakic pigmented rabbit eye after 2-mg intravitreal dose of vancomycin.
`Bar represents 20 Mm (toluidine blue, X1000).
`
`Color Fig 5.—Aqueous vancomycin concen¬
`tration after injection of 2 mg into anterior
`vitreous cavity of rabbit eye. Each point repre¬
`sents average value from two eyes; bars
`indicate lower confidence limit that could be
`indicates that one aphakic-
`calculated.
`1
`lens capsule was
`vitrectomized eye without
`evaluated at 12 hours; 2, that one of two eyes
`had aqueous (or vitreous) vancomycin con¬
`centration greater than 100 mg/L, with point
`representing mean of 100 and lower mea¬
`sured value; and 3, that lower confidence limit
`includes zero. Triangles indicate aphakic-
`lens capsule;
`vitrectomized eyes without
`ovals, aphakic-vitrectomized eyes with intact
`capsule; and x's, phakic eyes.
`
`Color Fig 6.—Vitreous vancomycin concen¬
`tration after injection of 2 mg into anterior
`vitreous cavity of rabbit eye. Each point repre¬
`sents average value from two eyes; bars
`indicate lower confidence limit that could be
`calculated. Triangles indicate aphakic-vitrec¬
`lens capsule; ovals,
`tomized eyes without
`aphakic-vitrectomized eyes with intact cap¬
`sule; and x's, phakic eyes.
`
`Downloaded From: https://jamanetwork.com/ by Judith Little on 01/07/2022
`
`Exhibit 2066
`Page 02 of 07
`
`
`
`48 hours, 72 hours, one week, and two
`weeks. Bright-flash electroretinography
`was performed binocularly on rabbits that
`had been dark-adapted for 30 minutes
`after pharmacologie mydriasis at 72 hours,
`one week, and two weeks following intra¬
`vitreal vancomycin injection. Preinjection
`and postinjection beta-wave amplitudes
`were measured in microvolts. Each eye
`served as its own control. Additionally, the
`noninjected fellow eye and eyes injected
`intravitreally with a sterile balanced salt
`solution also served as controls.
`Eyes were fixed for microscopic analysis
`using a mixture of 2% glutaraldehyde and
`2% paraformaldehyde in 0.1 mol/L of sodi¬
`um phosphate buffer (pH 7.4). This fixa¬
`tive was injected into the rabbit vitreous
`cavity immediately after death. The eyes
`were enucleated,
`the anterior segments
`were removed, and the posterior eyecups
`were immersed in fixative for 24 hours at
`4°C. Specimens were postfixed in 1% aque¬
`then dehydrated
`ous osmium tetroxide,
`through a graded ethanol-water series, and
`finally embedded in epoxy resin. Thick
`sections (1 Mm) were stained with toluidine
`blue for light microscopy;
`thin sections
`(silver-gray) were stained with uranyl ace¬
`tate-lead citrate for electron microscopy.
`The photomicrographs shown herein are
`representative of a given antibiotic dosage
`and were chosen because they exhibited
`good photoreceptor alignment and illus¬
`trated several ultrastructural features in
`one field.
`
`Intraocular Clearance
`The clearance studies were performed on
`three groups of pigmented rabbit eyes:
`phakic, aphakic-vitrectomized with an
`intact lens capsule, and aphakic-vitrectom¬
`ized without a lens capsule. Vancomycin
`hydrochloride (2 mg in 0.1 mL of preserva¬
`tive-free sterile water) was injected into
`the vitreous cavity, and the eyes were
`enucleated immediately following death at
`12, 24, 48, and 72 hours postinjection. Two
`eyes were evaluated at each time point in
`each of the three groups. Aqueous speci¬
`mens were collected from phakic and
`aphakic-vitrectomized eyes with an intact
`lens capsule by removing the fluid with a
`25-gauge needle. Those eyes were then
`frozen in liquid nitrogen, and the vitreous
`cavity fluid was dissected en bloc by the
`technique described by Abel and Boyle.10 In
`aphakic eyes, the fluid in the aqueous and
`vitreous cavities was removed with a 25-
`gauge needle. All specimens were frozen in
`evaporation-proof containers.
`The vancomycin concentration of
`the
`ocular fluid was determined by a fluores¬
`cent polarization immunoassay." Controls
`consisted of specimens supplied by the
`manufacturer and undiluted rabbit vitre¬
`ous mixed with a known concentration of
`vancomycin.
`Combination Antibiotic Testing
`Antibiotic sensitivity tests to vancomy¬
`cin and geñtamicin were performed on 14
`endophthalmitis isolates from our ocular
`microbiology laboratory using a microdilu-
`tion assay.12 Some of
`these isolates had
`
`Fig 1.—Mean electroretinogram beta-wave amplitude (in microvolts) of six groups of pigmented
`rabbit eyes expressed as percentage of beta-wave amplitudes of the normal fellow eyes. Three
`eyes were evaluated in each group. Standard deviation is indicated by bracket. In phakic eyes,
`group A received intravitreal injection of saline (control); group B, 2 mg of vancomycin; and group
`C, 5 mg of vancomycin. In aphakic vitrectomized eyes, group D received intravitreal
`injection of
`saline one week postoperatively (control); group E, 2 mg of vancomycin; and group F, 5 mg of
`vancomycin.
`
`previously been identified as gentamicin-
`resistant. The interaction of these isolates
`to a combination of vancomycin and geñta¬
`micin was determined by a microdilution
`checkerboard assay.13 The combined effect
`of the two antibiotics acting together was
`determined to be either synergistic, indif¬
`ferent, or antagonistic based on the frac¬
`inhibitory
`tional
`(FIC)
`concentration
`index or the appearance of an arithmetic
`inhibitory concentra¬
`plot of the minimal
`tions of the two antibiotics in each well of
`the microdilution checkerboard plate.13
`The combination was synergistic if the FIC
`index was less than 0.5; an FIC greater
`than 2.0 indicated an antagonistic combi¬
`nation. The FIC index was 1.0 if the effect
`of two antibiotics was the same as the sum
`of their individual effects (indifferent or
`additive).
`
`RESULTS
`Toxicity
`The vitreous and retina appeared
`normal by indirect ophthalmoscopy in
`both phakic and aphakic-vitrecto¬
`mized eyes after injection of intravit¬
`real vancomycin doses of up to 2 mg
`(Color Fig 1). In doses greater than 2
`mg, there was immediate clouding of
`the vitreous after the injection, and
`within 24 hours, opacification of the
`retina was noted (Color Fig 2). By two
`weeks the retina cleared; however, the
`retinal pigment epithelium showed
`
`Fig 2.—Preinjection and postinjection electro-
`retinograms of phakic and aphakic-vitrecto¬
`mized pigmented rabbit eyes with intact cap¬
`sule after 2- or 5-mg intravitreal
`injection of
`vancomycin (in 0.1 mL of preservative-free
`sterile water).
`
`areas of pigment clumping and atro¬
`phy·
`There was no electroretinographic
`evidence of toxicity between control
`eyes and eyes (phakic or aphakic-
`vitrectomized) receiving 2 mg of van¬
`comycin intravitreally (Figs 1 and 2).
`There was marked reduction in the
`postinjection alpha-wave and beta-
`
`Downloaded From: https://jamanetwork.com/ by Judith Little on 01/07/2022
`
`Exhibit 2066
`Page 03 of 07
`
`
`
`wave amplitudes in eyes receiving a
`5-mg intravitreal
`dose. Compared
`with control eyes, an intravitreal dose
`of 5 mg resulted in a 45% reduction in
`beta-wave amplitude in phakic eyes
`and an 80% reduction in beta-wave
`amplitude in aphakic-vitrectomized
`eyes (Fig 1).
`There were no gross differences at
`the light or electron microscopic level
`in the appearance of the retina and
`retinal pigment epithelium between
`the control rabbit eyes and either the
`phakic or aphakic-vitrectomized eyes
`that had received 0.5 to 2 mg of vanco¬
`mycin hydrochloride intravitreally
`(Color Fig 3; Figs 3 and 4). Pathologic
`alteration of these tissues, both at the
`light and electron microscopic levels,
`was noted when intravitreal vanco¬
`mycin doses greater than 2 mg were
`administered (Color Fig 4; Fig 5).
`toxicity
`features of
`Ultrastructural
`included the following:
`(1) hypertro¬
`phy of the retinal pigment epithelium
`with abnormal clustering of pigment
`granules in the apical cytoplasm; (2)
`loss of photoreceptor outer segment-
`retinal pigment epithelium interdigi-
`tation due to retraction of apical
`microvilli of retinal pigment epitheli¬
`um; (3) appearance of lucent vacuoles
`(100 to 200 nm in diameter) in the
`pigment epithelium basal
`retinal
`cytoplasm beneath the plasmalemma
`(4) gross disorganization
`infoldings;
`of the photoreceptor outer segments
`with distension and displacement of
`the inner segments past the external
`limiting membrane; (5) pyknosis (es¬
`pecially in the outer retinal
`layers);
`(6) accumulation of cellular
`and
`debris in the subretinal space.
`Intraocular Drug Clearance
`Rabbit aqueous and vitreous speci¬
`mens (seven of each) with vancomycin
`concentrations ranging from 0 to 100
`mg/L served as controls for the fluo¬
`rescent polarization immunoassay.
`The vancomycin concentrations mea¬
`sured in the control specimens by this
`technique were found to differ by 0%
`to 13% (mean, 6%) from the expected
`vancomycin concentrations.
`The ocular clearance of vancomycin
`after an intravitreal injection of 2 mg
`in the three groups of eyes (ie, phakic,
`and aphakic-vitrectomized with intact
`lens capsules, and aphakic-vitrecto¬
`lens capsules) was
`mized without
`examined (Color Figs 5 and 6). Vanco¬
`mycin was cleared most slowly in
`phakic eyes. At 72 hours postinjection,
`the average aqueous and vitreous con¬
`centrations exceeded 70 and 100 mg/L,
`respectively. Antibiotic clearance in
`aphakic-vitrectomized eyes with in-
`
`Fig 3.—Transmission electron micrograph of retinal pigment epithelium and photoreceptor cells
`injection of 2 mg of vancomycin.
`from phakic pigmented rabbit eye three days after intravitreal
`Top, Retinal pigment epithelial cell exhibits normal-appearing nucleus (n), endoplasmic reticulum
`(er), mitochondria (m), and pigment granules (pg). There is close apposition of apical microvilli
`(arrows) with distal rod outer segments (ros). Occasional residual bodies (rb) are observed in
`Intact junctional complex (arrowhead) is shown on
`basal retinal pigment epithelial cytoplasm.
`basolateral borders of two adjacent retinal pigment epithelial cells. Bruch's membrane (bm)
`appears normal. Bottom, Longitudinal view of apical rod inner segment (ris) and basal rod outer
`segment (ros) regions. Disc membranes, including basal "open" discs (arrows), are well aligned
`and nonvesiculated. Rod cell-connecting cilium (c), basal body (bb), ciliary rootlet (cr), and
`mitochondria (m) are intact and appear normal (original magnification, X13 500).
`
`lens capsules was more rapid
`tact
`than in the phakic eyes. The average
`aqueous and vitreous vancomycin con¬
`centrations in these eyes exceeded 25
`and 85 mg/L, respectively, at 48 hours.
`Clearance was most rapid in aphakic-
`vitrectomized eyes without lens cap¬
`sules.
`
`Combination Antibiotic Testing
`
`The results of individual and combi¬
`nation antibiotic sensitivity testing on
`14 representative gram-positive en¬
`dophthalmitis isolates from our insti¬
`tution were analyzed (Table). All
`organisms were found to be sensitive
`
`Downloaded From: https://jamanetwork.com/ by Judith Little on 01/07/2022
`
`Exhibit 2066
`Page 04 of 07
`
`
`
`reported that intravitreal vancomycin
`hydrochloride doses of 5 mg were non¬
`toxic in phakic albino rabbit eyes. In
`that study, posterior segment toxicity
`was evaluated by ophthalmoscopy and
`light microscopy; ultrastructure was
`examined in only one eye. In previous
`studies performed at our institution
`evaluating intraocular antibiotic tox¬
`icity in a rabbit model,1415 visible and
`histologie toxicity noted in pigmented
`rabbit eyes was often not detected in
`albino rabbit eyes at the same intra¬
`vitreal dosage. We felt it was neces¬
`sary to reevaluate the sensory retinal
`toxicity of intravitreal vancomycin in
`both phakic and aphakic-vitrectom¬
`ized pigmented rabbit eyes. We found
`intravitreal administration of
`that
`vancomycin in doses up to 2 mg to be
`nontoxic by all diagnostic criteria,
`including electron microscopy, in both
`of these groups. In pigmented rabbit
`eyes, a 5-mg intravitreal vancomycin
`dose, which was found to be nontoxic
`in albino rabbit eyes by Smith et al,8
`produced irreversible damage to the
`retina and retinal pigment epitheli¬
`um.
`lensectomy and
`We found that
`vitrectomy increased the intraocular
`clearance of vancomycin, but they did
`retinal
`not alter the threshold for
`toxicity produced by the antibiotic. A
`similar phenomenon was found with
`intravitreal aminoglycoside adminis¬
`tration by Talamo et al.16 They pro¬
`posed that toxicity is related to the
`peak concentration of the antibiotic,
`and not the duration of the contact, to
`which the retina is exposed shortly
`injection. Both
`intravitreal
`after
`the retina is
`studies indicate that
`exposed to the same peak antibiotic
`concentration after an injection into
`the vitreous cavity in both vitrectom-
`ized and nonvitrectomized eyes. Since
`the volume of the average human vit¬
`reous cavity is at least twice as great
`as that of the 3-kg pigmented rabbits
`used in this study,17 the peak vitreous
`vancomycin concentration after an
`intravitreal injection in a rabbit eye
`would be greater than in an adult
`human eye after injection of the same
`dose. Therefore, a 2-mg intravitreal
`dose in humans should be well below
`the toxic threshold if the toxic thresh¬
`old of the sensory retina is the same in
`pigmented rabbits and humans.
`All of our gram-positive endoph¬
`thalmitis isolates had a mean inhibi¬
`tory concentration (MIC)
`to vanco¬
`mycin of 4 mg/L or
`less, with the
`majority having an MIC of 0.5 to 1.0
`mg/L. Aqueous and vitreous vanco¬
`mycin levels greatly exceeding the
`MIC of gram-positive organisms were
`
`Fig 4.—Transmission electron micrograph of retinal pigment epithelium (rpe) and photoreceptor
`cells from aphakic-vitrectomized pigmented rabbit eye three days after intravitreal
`injection of 2
`mg of vancomycin. Ultrastructure and distribution of rpe organelles appear normal, m indicates
`mitochondria; bm, Bruch's membrane; rb, residual bodies; pg, pigment granules; ris; basal rod
`inner segment; and c, rod cell-connecting cilium. Additional features include well-organized Golgi
`complex (g) and phagosome (ph). Alignment and ultrastructural
`features of rod cells appear
`normal. Note normal nonvesiculated "open" discs (open arrows) at base of rod outer segments
`(ros). Arrow indicates basal open disc; arrowhead, intact junctional complex (original magnifica¬
`tion, X13 500).
`
`to vancomycin. Staphylococcus epider¬
`midis and Streptococcus species were
`resistant to geñtamicin. These orga¬
`nisms had an additive interaction to
`the combination of vancomycin and
`geñtamicin. Staphylococcus aureus
`and Bacillus cereus were sensitive to
`geñtamicin, and they had a synergis¬
`tic response to the combination of
`vancomycin and geñtamicin.
`COMMENT
`Forster et al4-5 have recommended
`the combination of cefazolin and geñ¬
`intravitreal
`the initial
`tamicin as
`antibiotic therapy for bacterial en-
`
`dophthalmitis. However, some of the
`endophthalmitis isolates at our insti¬
`tution, including an increasing num¬
`ber of coagulase-negative staphylo-
`cocci and some streptococcal species,
`have been resistant to both of these
`antibiotics. All of our gram-positive
`endophthalmitis isolates have been
`sensitive to vancomycin.
`Homer and associates1 evaluated
`the retinal toxicity of vancomycin in
`phakic albino rabbits by ophthalmos¬
`copy, electroretinography, and light
`microscopy in 1975. They found intra¬
`vitreal doses of 1 mg or less to be
`nontoxic. More recently, Smith et al8
`
`Downloaded From: https://jamanetwork.com/ by Judith Little on 01/07/2022
`
`Exhibit 2066
`Page 05 of 07
`
`
`
`tions were present beyond 48 hours in
`aphakic-vitrectomized eyes with in¬
`tact lens capsules; however, vitreous
`concentrations fell to subtherapeutic
`levels by 48 hours in aphakic-vitrec¬
`tomized eyes without an intact lens
`the timing of
`capsule. Therefore,
`repeated intravitreal vancomycin in¬
`jection,
`if it is clinically indicated,
`should depend on the surgical state of
`the eye. Intraocular vancomycin con¬
`centrations
`have previously
`been
`determined by a bioassay.7-8 In this
`study, we used a fluorescent polariza¬
`tion immunoassay to measure aque¬
`ous and vitreous antibiotic levels.11
`This technique has been found to be
`more accurate and reproducible for
`measuring serum antibiotic levels
`than a biological assay,18 and we found
`it to be acceptable for measuring the
`vancomycin concentration in intraoc¬
`ular fluid.
`We routinely treat endophthalmitis
`intravitreal
`with a combination of
`antibiotics to provide coverage for
`both gram-positive and gram-nega¬
`tive causative bacteria. For this rea¬
`son, we evaluated the interaction of
`vancomycin and an aminoglycoside
`in the laboratory to
`(geñtamicin)
`ensure that this combination of anti¬
`biotics was not antagonistic. A micro-
`dilution checkerboard assay using
`common gram-positive endophthal¬
`mitis isolates, including Staphylococ¬
`cus, Streptococcus, and Bacillus spe¬
`cies, was used, and an additive or
`synergistic interaction was found in
`all cases.
`Until recently, we reserved intravit¬
`real vancomycin administration ei¬
`ther for the treatment of endophthal¬
`mitis due to cefazolin- and gentami-
`cin-resistant gram-positive
`organ¬
`isms or for the treatment of late-onset
`endophthalmitis associated with fil¬
`these
`tering blebs because 57% of
`cases were found to have been caused
`by a streptococcal species in a series
`reported by Mandelbaum and col¬
`leagues.2 We are now treating all
`cases of bacterial endophthalmitis at
`our institution with the combination
`of vancomycin and an aminoglycoside.
`We found the intravitreal vancomycin
`dose of 1 mg recommended by Homer
`and associates7 to be clinically effec¬
`tive in the limited number of cases
`treated with vancomycin prior to this
`study, and this is the dosage currently
`used for the treatment of bacterial
`endophthalmitis in phakic and apha¬
`kic (pseudophakic) eyes with intact
`lens capsules at our institution. We
`recommend that a 2-mg intravitreal
`vancomycin dose be used for treat¬
`ment of endophthalmitis in aphakic-
`
`Fig 5.—Transmission electron micrograph of retinal pigment epithelium (rpe) and photoreceptor
`cells from aphakic-vitrectomized pigmented rabbit eye three days after intravitreal
`injection of 5
`mg of vancomycin. The rpe appears hypertrophied and exhibits abnormal clustering of pigment
`granules in apical cytoplasm. Apical rpe microvilli are retracted, resulting in loss of photoreceptor
`interdigitation. The rpe nucleus (n) and Bruch's membrane (bm) appear normal. Although rpe
`their distribution within cytoplasm is abnormal. Row of
`mitochondria appear normal,
`lucent
`vacuoles (arrowheads), each 100 to 200 nm in diameter, is seen in basal cytoplasm just beneath
`rpe plasmalemma infoldings. Photoreceptor cell
`is grossly disorganized. Rod outer
`layer
`segments (ros) are poorly aligned, although discs appear normal. Rod inner segments (ris) are
`distended and displaced. The ris mitochondria are mostly normal
`in ultrastructural detail.
`Remnant of dead cell (open arrow) is seen in subretinal space. Pigment granules (pg) are also
`noticeable (original magnification, X13 500).
`
`Antibiotic Sensitivity Testing
`
`Mean Inhibitory Concentration, mg/L
`
`Geñtamicin
`
`Vancomycin
`
`Interaction!
`
`Organism*
`Streptococcus
`S faecalis (4)
`S viridans ( 1 )
`Group D (1)
`Staphylococcus
`S epidermidis (4)
`S aureus (2)
`Bacillus
` cereus (2)
`* Numbers in parentheses indicate the number of isolates tested,
`tlnteraction of the combination of vancomycin and geñtamicin was determined by a microdilution checker¬
`board assay.
`
`16
`
`16
`
`0.5
`
`0.5
`
`Indifferent
`
`Indifferent
`
`Indifferent
`Synergistic
`
`Synergistic
`
`found 72 hours after
`intravitreal
`injection in phakic rabbit eyes. This
`finding supports those of previous
`
`intraocular vancomycin
`reports of
`clearance in phakic eyes.7·8 Therapeu¬
`tic intraocular vancomycin concentra-
`
`Downloaded From: https://jamanetwork.com/ by Judith Little on 01/07/2022
`
`Exhibit 2066
`Page 06 of 07
`
`
`
`vitrectomized eyes without a lens cap¬
`sule because of the more rapid antibi¬
`otic clearance.
`The combination of vancomycin and
`an aminoglycoside is adequate to treat
`all gram-positive exogenous endoph¬
`thalmitis isolates. This combination
`of antibiotics eliminates the need to
`select a different combination of anti¬
`biotics for each cause of exogenous
`endophthalmitis (postsurgical, filter¬
`traumatic). Ba-
`ing-bleb-associated,
`cillis cereus is one of the most com¬
`mon isolates in traumatic endophthal¬
`mitis.3 " Peyman et al20 recommended
`that clindamycin and geñtamicin be
`used as the initial intravitreal antibi¬
`otics in cases of traumatic endoph¬
`thalmitis to treat infections caused by
` cereus and other common bacteria.
`Our sensitivity tests show that cere¬
`us is equally sensitive to vancomycin
`(MIC, < 0.5 mg/L) and clindamycin
`(MIC, < 0.5 mg/L). Similar sensitivi¬
`ties have also been reported by O'Day
`and colleagues.21 Based on our limited
`clinical experience and the present
`laboratory data, we feel that vaneo-
`
`1. Driebe WT, Mandelbaum S, Forster RK, et
`al: Pseudophakic endophthalmitis. Ophthalmolo-
`gy 1986;93:442-448.
`2. Mandelbaum S, Forster RK, Gelender H, et
`al: Late onset endophthalmitis associated with
`filtering blebs. Ophthalmology 1985;92:964-972.
`3. Affeldt JC, Forster RK, Mandelbaum S, et
`al: Traumatic endophthalmitis. Ophthalmology
`1985;92(suppl):81.
`4. Forster RK, Abbott RL, Gelender H: Man-
`agement of endophthalmitis. Ophthalmology
`1980;87:313-319.
`5. Forster RK: Endophthalmitis, in Duane TM
`(ed): Clinical Ophthalmology. New York, Harper
`& Row Publishers Inc, 1987, vol 4, chap 24.
`6. Pryor JG, Apt L, Leopold IH: Intraocular
`penetration of vancomycin. Arch Ophthalmol
`1962;67:608-611.
`7. Homer P, Peyman GA, Koziol J, et al:
`injection of vancomycin in experi-
`Intravitreal
`staphylococcal endophthalmitis. Acta
`mental
`Ophthalmol 1975;53:311-320.
`8. Smith MA, Sorenson JA, Lowy FD, et al:
`Treatment of experimental methicillin-resistant
`endophthalmitis
`epidermidis
`Staphylococcus
`with intravitreal vancomycin. Ophthalmology
`1986;93:1328-1335.
`'Le System': A unified microsur-
`9. Parel JM:
`in GW Blankenship, M Stirpe, M
`gical center,
`Gonvers, et al (eds): Basic and Advanced Vitre-
`ous Surgery. Padova, Italy, Liviana Press, 1986,
`pp 21-27.
`10. Abel R, Boyle GL: Dissecting intraocular
`tissue for intraocular drug studies. Invest Oph-
`thal Vis Sci 1976;15:216-217.
`11. Principles of automated methods for diag-
`nostic microbiology, in Fingold S, Baron EJ (eds):
`Bailey and Scott's Diagnostic Microbiology. St
`Louis, CV Mosby Co, 1986, pp 141-155.
`12. Jones RN, Barry RI, Gavan TL, et al:
`
`mycin may be as appropriate as clin¬
`damycin for the treatment of
`trau¬
`matic endophthalmitis.
`Vancomycin22 and the aminoglyco¬
`side23 class of antibiotics have differ¬
`ent modes of action on bacterial
`metabolism. However, both vancomy¬
`cin2426 and the aminoglycosides24"·27
`have been reported to cause nephro-
`toxicity and ototoxicity in the host.
`The exact biochemical mechanism of
`toxicity in host tissues is unknown for
`these classes of antibiotics,
`each of
`the evidence suggests that it is
`but
`different. Characteristic lamellar ly-
`inclusions have been noted
`sosomal
`ultrastructurally in tissues exposed to
`toxic concentrations of aminoglyco-
`sides.27 These are felt to be accumula¬
`tions of complex lipids within lyso-
`somes due to a toxic effect on lysoso-
`mal enzymes involved in lipid metabo¬
`lism.27 To our knowledge, this type of
`inclusion has not been detected in any
`tissues exposed to toxic concentra¬
`tions of vancomycin in this or any
`other study.26·28
`Nephrotoxicity and ototoxicity have
`
`References
`
`Susceptibility tests: Microdilution and macrodi-
`lution broth procedures, in Lennette EH, Balows
`A, Hauster WJ, et al (eds): Manual of Clinical
`Microbiology. Washington, DC, American Soci-
`ety for Microbiology, 1985, pp 972-977.
`13. Krogstad DJ, Moellering RC: Combina-
`tions of antibiotics, mechanisms of interaction
`against bacteria, in Lorian V (ed): Antibiotics in
`Laboratory Medicine. Baltimore, Williams &
`Wilkins, 1980, pp 298-341.
`14. Zackery IG, Forster RK: Experimental
`intravitreal gentamicin. Am J Ophthalmol 1976;
`82:604-611.
`15. Fischer JP, Civiletto SE, Forster RK: Tox-
`intravitreally
`icity, efficacy, and clearance of
`injected cephazolin. Arch Ophthalmol
`1982;
`100:650-652.
`16. Talamo JH, D'Amico DJ, Hamminen LA,
`et al: The influence of aphakia and vitrectomy on
`experimental retinal toxicity of aminoglycoside
`antibiotics. Am J Ophthalmol 1985;100:840-847.
`17. Gloor BP: The vitreous, in Moses RA (ed):
`Adler's Physiology of the Eye: Clinical Applica-
`tion. St Louis, CV Mosby Co, 1981, pp 255-276.
`18. Buchanan AG, Witwicki E, Albritton WL:
`Serum aminoglycoside monitoring by enzyme
`immunoassay, biological, and fluorescent immu-
`noassay procedures. Am J Med Technol 1983;
`49:437-441.
`19. Brinton GS, Topping TN, Hyndiuk RA, et
`al: Posttraumatic endophthalmitis. Arch Oph-
`thalmol 1984;104:547-550.
`20. Peyman GA, Carroll CP, Raichand M: Pre-
`vention and management of traumatic endoph-
`thalmitis. Ophthalmology 1980;87:320-324.
`21. O'Day DM, Smith RS, Gregg CR, et al: The
`problem of Bacillus species infection with special
`emphasis on the virulence of Bacillus cereus.
`Ophthalmology 1981;88:833-838.
`22. Barna JCJ, Williams DH: The structure
`
`been found to be potentiated by the
`concurrent use of vancomycin and
`aminoglycosides29'33 in both clinical
`studies of humans and experiments
`with rats.28 It is possible that retinal
`and pigment epithelial toxicity could
`also be potentiated by this combina¬
`tion of antibiotics. Thus far, we have
`not noted any clinically evident poste¬
`rior segment toxicity when these two
`antibiotics are injected together in
`individually nontoxic doses. Neverthe¬
`less, we feel that the toxicity of this
`combination of antibiotics should be
`evaluated in an animal model.
`
`This investigation was supported in part by US
`Service
`Health
`research
`Public
`grant
`2RO1EY06045 and core grant 5P30EY02180 (Na¬
`tional Institutes of Health, National Eye Insti¬
`tute, Bethesda, Md); a center grant
`from the
`National Retinitis Pigmentosa Foundation Ine,
`Baltimore; and an unrestricted departmental
`grant from Research to Prevent Blindness Ine,
`New York.
`We are appreciative of the assistance provided
`to us by William Feuer, MS, Department of
`Ophthalmology, University of Miami School of
`Medicine.
`
`and mode of action of glycopeptide antibiotics of
`the vancomycin group. Annu Rev Microbiol
`1984;38:339-357.
`23. Weisblum B, Davis J: Antibiotic inhibitors
`of the bacterial ribosome. Bacteriol Rev 1968;
`32:493-528.
`24. Appel GP, Neu HC: The nephrotoxicity of
`antimicrobial agents. NEngl J Med 1977;296:722\x=req-\
`728.
`25. Brummett RE: Drug-induced ototoxicity.
`Drugs 1980;19:412.
`26. Fekety R: Vancomycin. Med Clin North