Mi■iihiiiiiii41@4
`Results of the Endophthalmitis Vitrectomy Study
`
`A Randomized Trial of Immediate Vitrectomy and of Intravenous Antibiotics
`for the Treatment of Postoperative Bacterial Endophthalmitis
`
`Endophthalmitis Vitrectomy Study Group
`
`Obiective: To determine the roles of immediate pars plana
`vitrectomy (VIT) and systemic antibiotic treatment in the
`management of postoperative endophthalmitis.
`
`Design: Investigator-initiated, multicenter, random(cid:173)
`ized clinical trial.
`
`Setting: Private and university-based retina-vitreous
`practices.
`
`Patients: A total of 420 patients who had clinical evi(cid:173)
`dence of endophthalmitis within 6 w~eks after cataract
`surgery or secondary intraocular lens implantation.
`
`Interventions: Random assignment according to a 2 X 2
`factorial design to treatment with VIT or vitreous tap or
`biopsy (TAP) and to treatment with or without sys(cid:173)
`temic antibiotics (ceftazidime and amikacin).
`
`Main Outcome Measures: A 9-month evaluation of
`visual acuity assessed by an Early Treatment Diabetic Reti(cid:173)
`nopathy Study acuity chart and media clarity assessed both
`clinically and photographically.
`
`Results: There was no difference in final visual acuity or
`media clarity with or without the use of systemic antibiot(cid:173)
`ics. In patients whose initial visual acuity was hand motions
`or better, there was no difference in visual outcome whether
`or not an immediate VII was performed. However, in the
`subgroup of patients with initial light perception-only vi(cid:173)
`sion, VIT produced a threefold increase in the frequency of
`achieving 20/40 or better acuity (33% vs 11 % ) , approximately
`a twofold chance of achieving 20/100 or better acuity (56%
`vs 30%), and a 50% decrease in the frequency of severe vi(cid:173)
`sual loss (20% vs47%) over TAP. ln this group of patients,
`the difference between VIT and TAP was statistically signifi(cid:173)
`cant (P<.001, log rank test for cumulative visual acuity scores)
`over the entire range of vision.
`
`Conclusions: Omission of systemic antibiotic treat(cid:173)
`ment can reduce toxic effects, costs, and length of hos(cid:173)
`pital stay. Routine immediate VIT is not necessary in pa(cid:173)
`tients with better than light perception vision at
`presentation but is of substantial benefit for those who
`have light perception-only vision.
`
`(Arch Ophthalmol. 1995; 113: 14 79-1496)
`
`C ERT AIN ASPECTS of the man(cid:173)
`
`agement of bacterial en(cid:173)
`dophthalmitis after cataract
`extraction, such as injection
`of intravitreal antibiotics,
`have been widely accepted. Other aspects of
`management are controversial. There have
`been no clear data as to whether pars plana
`vitrectomy (VIT) should be used in the ini(cid:173)
`tial management of endophthalmitis. Imme(cid:173)
`diate VIT for endophthalmitis offers several
`theoretical advantages, including removal of
`the infecting organisms and the toxins they
`produce, removal of vitreous membranes that
`could lead to subsequent detachment of the
`retina, clearing of vitreous opacities, collec(cid:173)
`tion of abundant material for culture, and pos(cid:173)
`sibly better distribution of intravitreal anti(cid:173)
`biotics. In some experimental animal stud(cid:173)
`ies, VIT offered advantages over the use of
`intraocularantibiotics alone.1·3 However, past
`
`ARCH OPHTHALMOLNOL 113. DEC 1995
`1479
`
`data from human studies have not shown VlT
`with intravitreal antibiotics to be superior to
`treatment with intravitreal antibiotics alone.
`In these studies, eyes that underwent VIT
`were not randomly selected and were those
`with the worst clinical presentation. 4-8Because
`of this selection bias, the place of pars plana
`VIT in the initial treatment of patients with
`endophthalmitis after cataract surgery re(cid:173)
`mained in doubt.
`Although the intraocular penetration
`of most antibiotics is poor after systemic ad(cid:173)
`ministration, drugs given by this route have
`remained part of the routine management
`of bacterial endophthalmitis.9 Although
`
`See Methods on next page
`
`Participants and clinical
`and support centers of the
`Endopltthalmitis Vitrectomy
`Study trials are listed on pages
`1493 to 1495.
`
`Downloaded From: http://archopbt.jamanetwork.com/ by Andrew Calman on 11/19/2015
`
`Novartis Exhibit 2310.001
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`METHODS
`
`PROTOCOL AND PROCEDURES
`
`Study Organization
`
`Twenty-four centers across the United States participated
`in this clinical trial. Patients were enrolled between Feb(cid:173)
`ruary 1990 and January 1994, and follow-up was com(cid:173)
`pleted in December 1994. Statistical design, data manage(cid:173)
`ment, study communications, and data analysis were carried
`out by the Coordinating Center (University of Pittsburgh,
`Pa). A Photographic Reading Center (University of Wis(cid:173)
`consin, Madison) was responsible for evaluating fund us pho(cid:173)
`tographs. Scientific direction for the study was the respon(cid:173)
`sibility of the Study Chair in collaboration with the Executive
`Committee. An independent Data and Safety Monitoring
`Committee met annually and reported to the National Eye
`Institute. The research protocol was approved by institu(cid:173)
`tional review boards at each participating clinical center.
`
`Patient Selection
`
`Patients were eligible for study entry if they had clinical signs
`and symptoms of bacterial endophthalmitis within 6 weeks
`after cataract surgery or secondary intraocular lens implan(cid:173)
`tation. Eligibility required the following: visual acuity of light
`perception (LP) or better and worse than 36 letters at 4 m
`(equivalent to approximately 20/50 or worse) on an Early
`Treatment Diabetic Retinopathy Study (ETD RS) acuity chart1\
`cornea and anterior chamber of the involved eye clear enough
`to allow visualization of at least some part of the iris; the cor(cid:173)
`nea clear enough to perform pars plana VIT; and a hypo(cid:173)
`pyon or sufficient clouding of the anterior chamber or vitre(cid:173)
`ous to obscure a view of second-order retinal arterioles.
`Patients were excluded for any of the following reasons: known
`eye disease limiting visual acuity to 20/100 or worse before
`the development of cataract, prior intraocular surgery other
`than cataract or intraocular lens surgery, prior penetrating
`ocular trauma, previous injection of intravitreal antibiotics,
`prior pars plana VIT, retinal detachment or choroidal de(cid:173)
`tachment that was moderately high as judged by indirect oph(cid:173)
`thalmoscopy or ultrasound, probable intolerance to any study
`drugs (with the exception of penicillin allergy, in which case
`alternatives to [3-lactam drugs were used), strong suspicion
`of fungal endophthalmitis, age younger than 18 years, un(cid:173)
`suitability for surgery, or likelihood that the patient would
`not return for follow-up visits. A total of 1283 patients with
`endophthalmitis were screened, 855 of whom had endoph(cid:173)
`thalmitis within 6 weeks after cataract extraction or second(cid:173)
`ary lens implantation. Of these, 510 met eligibility criteria,
`and 420 agreed to participate and were enrolled. Written in(cid:173)
`formed consent was obtained from each patient.
`
`Examination
`
`The initial examination was performed before randomiza(cid:173)
`tion. Best refracted vision was determined using an ETDRS
`acuity chart. lf no letters could be read on the chart at 4 m ,
`then at 1 m, vision was tested for the ability to count fingers.
`If the patient was unable to count fingers, vision was tested
`for the ability to recognize hand motions. For this, the pa(cid:173)
`tient's opposite eye was occluded, and a light source, such
`
`as a lamp used for near vision, was directed from behind the
`patient to the examiner's hand that either was stationary or
`was moved at one motion per second in a horizontal or ver(cid:173)
`tical direction at a distance of 60 cm from the eye. The pa(cid:173)
`tient was asked to identify whether the examiner's hand was
`still, moving sideways, or moving up and down. The presen(cid:173)
`tation was repeated five times , and hand-motion visual acu(cid:173)
`ity was considered present if the patient was able to identify
`the examiner's action on at least four of the presentations. lf
`the examiner was not convinced that hand motions could be
`detected, LP was tested at 0.9 m with an indirect ophthal(cid:173)
`moscope set at maximum intensity.
`
`Treatment Assignment
`
`Eligible patients who provided consent were immediately
`randomly assigned according to a 2 X2 factorial design to
`one of four treatment groups: initial VIT with IV antibiot(cid:173)
`ics, initial VIT without IV antibiotics , initial TAP with IV
`antibiotics, or initial TAP without IV antibiotics.
`
`Initial Procedure
`
`Treatment was begun within 6 hours of the initial examina(cid:173)
`tion. Eyelid cultures were obtained from the affected eye. All
`the patients had a 0.1-mL anterior chamber sample ob(cid:173)
`tained with a 25- to 27-gauge needle and syringe. Patients
`assigned to the VIT groups underwent a three-port pars plana
`VIT. An initial undiluted vitreous specimen was obtained af(cid:173)
`ter placing all sclerotomies, but before turning on the infu(cid:173)
`sion fluid. The VIT cutter was introduced into the midvitre(cid:173)
`ous, and O .2 to O .5 ml of vitreous gel was excised and aspirated
`into a syringe, using manual suction with a high cutting rate.
`Once this sample was obtained, the infusion was turned on
`and the VIT procedure was continued with automated suc(cid:173)
`tion and collection into a VIT cassette. When necessary for
`visualization, the anterior chamber was cleared using any one
`of a variety of techniques. 15 If there was no posterior vitre(cid:173)
`ous separation, no attempt was made to induce a vitreous de(cid:173)
`tachment, and 1he posterior cortical vitreous was not aggres(cid:173)
`sively removed. It was a goal of surgery to remove at least
`50% of the vitreous gel in eyes with no vitreous separation.
`Patients assigned to the TAP groups had, at the discre(cid:173)
`tion of the operating surgeon , a vitreous specimen collected
`either by trans-pars plana vitreous needle aspiration or by
`vitreous biopsy through a single sclerotomy using a VIT in(cid:173)
`strument.16 A vitreous sample of 0.1 to 0.3 ml was col(cid:173)
`lected. If the surgeon chose needle aspiration and an ad(cid:173)
`equate sample could not be safely obtained with that technique,
`a vitreous biopsy using a VIT instrument was performed.
`
`Study Medications
`
`All the study patients received a standard antibiotic regimen
`that was chosen by agreement among the investigators and in(cid:173)
`fectious disease consultants before recruitment began. The EVS
`was not designed to test the efficacy of specific antibiotics or
`other drugs. The goal was to use what were judged to be the
`best available drugs. The antibiotic choices were reviewed an(cid:173)
`nually, and on each occasion the investigators chose to con(cid:173)
`tinue using the same drugs throughout the course of the study.
`After the initial VIT or TAP procedure , all the
`patiems received intravitreal inj ection of amikacin
`(0 .4 mg in 0.1 ml [volume of normal saline solution ])
`
`Downloaded From: http://archopbt.jamanetwork.com/ by Andrew Calman on 11/19/2015
`
`ARCH OPHTHALMOL/VOL I 13, DEC 1995
`1480
`
`Novartis Exhibit 2310.002
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`
`
`and vancomycin hydrochloride (1.0 mg in 0.1 mL). Vanco-
`mycin hydrochloride (25 mg in 0.5 mL), ceftazidime
`(100 mg in 0.5 mL), and dexamethasone sodium phos-
`phate (6 mg in 0.25 mL) were administered by subconjunc-
`tival injection. If the patient was allergic to penicillin, sub-
`conjunctival amikacin (25 mg in 0.5 mL) was substituted
`for ceftazidime. Topical antibiotics (vancomycin hydrochlo-
`ride, 50 mg/mL,alternating with amikacin, 20 mg/mL) were
`administered as frequently as one dropper hourif there was
`evidence of woundinfection or leak, and every 4 hours oth-
`erwise. Topical cycloplegics (1% atropinesulfate or 1/4% sco-
`polamine hydrobromide) and topical corticosteroids (1%
`prednisoloneacetate) were also administered after surgery.
`Systemic corticosteroids (prednisone, 30 mg twice a day for
`5 to 10 days) were administeredorally.
`Patients assigned to the lV antibiotic groups received
`two drugs. The first was ceftazidime, 2 g every 8 hours, in
`mostpatients (1.5 g for patients weighingless than 50 kg).
`Patients who wereallergic to penicillin were given cipro-
`floxacininstead, 750 mgorally twice a day. The second sys-
`temic drug was amikacingiven in a 7.5-mg/kginitial IV dose,
`followed by 6 mg/kg every 12 hours. If the patient’s serum
`
`creatinine concentration exceeded 177 pmol/L (2 mg/dL),
`subsequentdoses of amikacin were based on serum con-
`centrations of the antibiotic. In all patients, serum concen-
`trations were obtained and doses were adjusted to main-
`tain peak amikacin concentrations of 25 g/mL and trough
`concentrations of less than 5 jrg/dL. Therationale for these
`drug choices has been previously reported.” Patients were
`maintained on treatment with the systemic antibiotics for
`5 to 10 days at the physician’s discretion.
`
`Cultures and Stains
`
`Cultures of the anterior chamberfluid and undiluted vitre-
`ous were plated on chocolate agar (37°C in carbon dioxide)
`in freshly reduced, enriched thioglycolate liquid (aerobic at
`37°C) (Baltimore Biological Laboratories, Cockeysville, Md),
`designated BBL 1135, and fresh Sabouraud dextrose agar(in-
`cubated at 25°C). Gram stains were preparedfrom theante-
`rior chamber and undiluted vitreous specimens. The VITef-
`fluent (collected in the VIT cassette) wasfiltered through a
`
`sterile 0.45-m membrane filter. The filter was subse-
`quently divided understerile conditionsinto three pieces. One
`piece was placed on chocolateagar for culture at 37°C in 5%
`to 10% carbon dioxide and one was placed on fresh Sab-
`ouraud dextrose agar for culture at 25°C. Anaerobic culture
`ofthe filtered material was performedin either enrichedthio-
`glycolate broth or anaerobic blood agar enriched with he-
`min and vitamin K at 37°C.
`
`
`Additional Procedures During Initial Hospitalization
`
`The protocolallowedpatients in the TAP groups to have VIT
`and reinjection of intravitreal antibiotics if the eye was do-
`ing poorly 36 to 60 hoursafter theinitial surgery. For such
`additional surgery to be recommended,aneye had to meet
`all the following criteria: (1) visual acuity of less than 5/200
`but LP or better; (2) an absent red reflex or an increase in
`media opacification compared withinitial presentation; (3)
`at least an equivocal growth from theinitial culture; and (4)
`one or moreofthe following: (a) a 1-mm increasein the height
`of the hypopyon, (b) a corneal ringinfiltrate, or (c) worsen-
`ing pain. Similarly, patients assigned to the VIT group who
`
`met the samecriteria 36 to 60 hoursafter the initial proce-
`dure could have repeated VIT (or vitreous aspiration) and
`reinjection ofintravitreal antibiotics. Patients whose eyes did
`not meetthe criteria for reoperation couldstill undergo ad-
`ditional surgery if their physician thoughtit to be in the pa-
`tient’s best interest. Conversely, patients who metcriteria for
`additional surgery were not required to have such surgery if
`it was thoughtnotto bein their best interest.
`
`Late Additional Surgery
`
`At the 3- and 9-month follow-up examinations, patients
`were assessed for remediable factors thatlimited visual acu-
`ity, such as vitreous opacities, macular pucker, or opaci-
`fied posterior capsule. If clinically appropriate, additional
`surgery was encouraged to improve these conditions.
`
`Outcome Evaluation
`
`Primary study end points werevisual acuity and clarity of the
`ocular media. All patients had end-point assessmentat 3- and
`9-monthfollow-upvisits. An additional assessment was made
`at a 12-monthvisit for those patients who hadadditional pro-
`cedures, based on theresults of the 9-monthvisit. Best cor-
`rected visuat acuity was measured after manifest refraction
`using the ETDRS visual acuity charts. Measurement was ob-
`tained by a certified technician maskedto treatmentassign-
`ment. Before data analysis, three thresholds ofvisual out-
`come were chosentoreflect differentlevels of functionalvision:
`20/40 or better, 20/100 or better, and 5/200 orbetter.
`Media clarity was assessed both clinically and photo-
`graphically. Clinical assessment of media clarity was per-
`formed with indirect ophthalmoscopy to classify the me-
`dia as one of the following: (1) better or equal to a 20/40
`view to the retina; (2) clarity worse than a 20/40 view, with
`a second-orderretinal vessel visible; (3) inability to see a
`second-orderretinal vessel, but with someretinalvesselvis-
`ible; (4) inability to see a retinal vessel, but with a red re-
`flex; and (5) no red reflex visible.
`Photographic grading of media clarity was based on
`the 3-monthandfinal (9- or 12-month) follow-up evalu-
`ations. The photographsconsisted of (1) a stereo pair fo-
`cused on the retina and centered halfway between thedisc
`and macula; (2) a single clearest possible photographof the
`retina centered in the same location; and (3) a stereo-
`scopic anterior segment photographic pair to documentthe
`status of the cornea, anterior chamber,intraocularlens (if
`present) as well as to showthe appearanceof the fundus
`red reflex. A masked observer at the EVS Photographic Read-
`ing Center graded the photographs by comparison with two
`preselected standard photographs.
`
`STATISTICAL METHODS
`
`Sample Size Considerations
`
`To determine sample size, the primary end point used to
`define success was a visual acuity of 20/400 orbetteratfi-
`nal follow-up. A success rate of 60% with TAP wasas-
`sumed. A one-tailed test was used because a physician would
`wart to recommend VIT onlyif it were better than TAP
`alone. Given a sample size of 420 patients, if there were
`
`Continued on next page
`
`DownloadedFrom:http://archopht.jamanetwork.com/ by Andrew Calmanon 11/19/2015
`
`
`ARCH OPHTHALMOI/VOQ]1. 113, DEC 1995
`1481
`
`Novartis Exhibit 23 10.003
`Regeneron v. Novartis, IPR202 1-008 16
`
`Novartis Exhibit 2310.003
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`60% success in the TAP group, the rate in the ViT group
`that could be detected with 80% power would be 72%, and
`that with 90% power, 74%.'*
`
`Treatment Group Comparisons
`
`The distribution of baseline characteristics and follow-up
`events were compared amongthe fourtreatment groups us-
`
`ing x’ tests or Brown-Moodmediantests as appropriate.” Be-
`causethetrial used a 2 2 factorial design, logistic regression
`models for each visual acuity threshold were fitted with each
`of the two experimental factors and their interaction as ex-
`planatory variables. Since there was no evidence of an inter-
`action between surgical treatment and systemic antibiotictreat-
`ment, these analyses are not reported. For each threshold of
`visual acuity, dichotomous outcomedifferences among the
`
`four treatment groups weretested with a x’ statistic. In ad-
`dition, two-waytests were performed to compare the VIT and
`
`TAP groups and the IV and NOITVgroups. TheP valuesre-
`ported werenot adjusted for multiple comparisons; therefore,
`
`the nominal P values must beinterpreted with this in mind.
`
`Outcome Evaluation
`
`To examinethefull rangeofvisual acuity outcomes, we con-
`sidered the visual acuity score based on the ETDRSacuity
`chart. Thevisual acuity scores among EVSpatients were not
`normally distributed, so linear models that require an as-
`sumption of normality were not appropriate. A Mantel-
`Haenszel log rank analysis was used with each visual acuity
`score as a stratum. This allowed outcome comparisonsofthe
`proportion of patients with visual acuity scores of more than
`oneletter, more than twoletters, more than threeletters, and
`so on. This analysis of outcomeis parallel to the usuallife-
`table analysis, in which one compares treatment according
`to the proportionofpatients alive at more than 1 year, more
`than 2 years, more than 3 years, and so on. Figures were con-
`structed to present the cumulative proportionofpatients ac-
`cordingto thefinal visual acuity score achieved. The figures
`are presented parallel to usual “survival curves.”After veri-
`fying the assumptionof proportionalhazards, a Cox regres-
`sion model was used to extend analysis of visual acuity out-
`cometo take into account baseline characteristics.
`
`Safety Monitoring
`
`Forissues of safety, the visual acuity score at the 3-month
`follow-up visit was used as the end pointfor the interim moni-
`toring. A threshold of5/200 visual acuity was used to com-
`pare the VIT and TAPtreatment groups. For interim moni-
`toring,the statistic described by O’Brien and Fleming” was
`calculated after 140, 280, and 420 patients had entered the
`trial, and these were reported to the Data and Safety Moni-
`toring Committee. Formalinterim statistical testing was not
`performed for patients treated with systemic antibiotics vs
`those who were not; however, tabulations comparing these
`patients were part of the presentationat the regular Data and
`Safety Monitoring Committee meetings.
`
`
`
`
`
`Subset Analysis
`
`To determine whether one surgical treatment was supe-
`rior to the other for any subset of patients, outcome was
`examined by surgical treatment for each subgroup de-
`fined by clinical presentation. This was carried out for each
`of the four definitions of successful outcomebased on the
`three visual acuity thresholds and mediaclarity level of 20/40
`view orbetter. A logistical model wasfitted with three ex-
`planatory variables. These three explanatory variables were
`VIT treatment, an individualrisk factor defining the sub-
`group, and an interaction term of VIT with the risk factor.
`An interaction P value was calculated. A statistically sig-
`nificant coefficient for the interaction term was inter-
`preted to mean that the association of VIT with outcome
`differed in the subgroup defined bythe risk factor. To ex-
`amine further whether VIT treatment was moreeffective
`than TAPfor a particular subgroup, the Cox model was used
`with the same variables as were used in the logistic model
`described in the previous paragraph. To examinethe con-
`sistency of the subgroup findings, a model wasfitted to ad-
`just for additional factors, once appropriate interaction terms
`were determined.
`
`Risk Factors for Visual Acuity Outcome
`
`To examinethe relation of baseline characteristics to out-
`come, we performed tabulations for each visual outcome by
`each baseline factor. To determine which of these variables
`were independentrisk factors for poor visual outcome,lo-
`gistic regression models werefitted using a backward step-
`ping procedure. Four separate models werefitted for the three
`threshold definitions of visual acuity and for mediaclarity
`outcome. A fifth modelusing the entire range ofvision as an
`outcome wasfitted using Cox regression analysis.
`
`
`
`Patients Analyzed
`
`Baseline characteristics are reported for the 420 EVS pa-
`tients enrolled. Outcomeis reported for the 396 patients
`who completed a final follow-up visit. Twenty-four pa-
`tients did not have final follow-up data: 12 died, five with-
`drew consentto be followed up, and seven werenot will-
`ing to return forthe visit. These patients had been assigned
`in nearly equal numberstoall treatment groups. Among
`the 396 with final visit data, two were missing visual acu-
`ity data and four were missing a clinical assessment of me-
`dia clarity. Thus, final visual acuityis reported in 394 pa-
`tients, and media clarity in 392. Included in the reports was
`information on patients with enucleated eyes whose vi-
`sion wasclassified as no LP. Also includedis one patient
`who died before a scheduled 12-month visit; therefore,
`9-monthdata were considered asfinal. Three patients who
`were entered into thetrial were subsequently noted to have
`had exclusionary criteria. Based onthe principle of “inten-
`tion to treat,” these patients were consideredin the analy-
`sis, although one was amongthe patients whodid not have
`a final follow-up visit.
`
`some newerdrugs,eg, the fluoroquinolones, given intra-
`commonbacteria that are responsible for postoperative en-
`dophthalmitis.’° The 8-lactam drugs and vancomycin hy-
`venously and evenorally have greater penetration into the
`drochloride, whichare agents of choice for infections caused
`humanvitreous,theystill do not reach sufficient intraocu-
`by gram-positive cocci, penetrate relatively poorly, and it
`lar concentrationsto be consideredefficacious against many
`—___eee$$
`ARCH OPHTHALMOL/VOL113, DEC 1995
`1482
`
`Downloaded From:http://archopht.jamanetwork.com/ by Andrew Calman on 11/19/2015
`
`Novartis Exhibit 2310.004
`Regeneron v. Novartis, IPR202 1-008 16
`
`Novartis Exhibit 2310.004
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`To analyze microbiology results, “laboratory con-
`firmed growth”wasdefined asat least semiconfluent growth
`on a solid medium, any growth on two or more media, or
`growth on one medium supported by a positive Gramstain.
`Results showed no growth in 18% of patients, “equivocal
`growth” (defined as growth less than laboratory-
`confirmed growth) in 13%, and laboratory-confirmed
`growth in the remainder. Laboratory-confirmed organ-
`isms were grouped as gram-positive coagulase-negative
`(47% of patients), other gram-positive (16% of patients),
`and gram-negative (4% of patients). More than one spe-
`cies grew in 3%ofpatients, either gram-positive coagulase-
`negative plus other gram-positive or gram-positive coagu-
`lase-negative plus gram-negative. The type of organism was
`evenly distributed across treatment groups (Table 1).
`
`ADVERSE EVENTS AND
`ADDITIONAL PROCEDURES
`
`is not clear that they offer additional benefit over intravit-
`real injections. Disadvantages of systemic antibiotic treat-
`mentinclude adverse effects that may be severe,''? the cost
`of antibiotics, and the hospitalization required for their ad-
`ministration. In a nonrandomized study, Pavan and Brin-
`ser? successfully treated several patients with endophthal-
`mitis without using systemic antibiotics. Consideringtheir
`uncertain efficacy, possible toxic effects, and high cost, the
`role of systemic antibiotics in postoperative endophthal-
`mitis was also examined.
`The Endophthalmitis Vitrectomy Study (EVS), a ran-
`domized, multicenter, clinical trial supported by the
`National Eye Institute of the National Institutes of Health,
`Bethesda, Md, was designed to determine therole of
`immediate pars plana VIT and, separately,the role of sys-
`temic antibiotics in the managementof endophthalmi-
`tis after cataract extraction or secondary intraocular lens
`insertion. The study subjects were 420 patients in whom
`clinical signs of endophthalmitis developed within 6 weeks
`alter cataract surgery or secondary lens implantation. They
`To monitorthesafety of treatments used in the EVS,events
`were randomly assigned to treatment with either imme-
`during follow-up were compared by treatment(Table 2).
`diate pars plana VIT orvitreoustap or biopsy (TAP). They
`At the 36- to 60-hour examination, five eyes had no LP,
`also were randomly assignedto either intravenous (TV)
`four from the TAP group and one from the VIT group.
`antibiotic treatment or no intravenous (NOIV) antibi-
`Immediate complications associated with theinitial EVS
`otic treatment. Outcome was evaluated by visual acuity
`procedures were few and did not vary substantially by
`and clarity of ocular media at 3 months and at 9 to 12
`treatment. Two patients suffered from a dislocated in-
`months, This article reports the main results of the EVS.
`traocular lens and one patient experienced an expulsive
`hemorrhage. Macular infarction was observed in one pa-
`TEEd
`tient who had undergone VIT with IV antibiotics. Renal
`complications were assessed by a change in serum cre-
`BASELINE CHARACTERISTICS
`atinine levels, although these data were missing in a sub-
`stantial numberof patients assigned to the NOIV group.
`Five percentof patients showedan increase in serumcre-
`
`atinine level of 26 pmol/L. or greater (20.3 mg/dL), and
`
`less than 1% showed an increase of 53 pmol/L or greater
`(20.6 mg/dL). There was nostatistical difference in cre-
`atininerise in patients in the IV group vs the NOIV group.
`
`Table 1 presents baseline characteristics for the fourtreat-
`ment groups. Statistical testing indicated that thetreat-
`ment groups were balanced.Statistics forall patients com-
`bined (last column) describe the study patient profile. The
`median age was 75 years, and less than half the patients
`(43%) were men. There wasa history ofdiabetes mellitus
`in 14% and systemic hypertension in 40%. In this popu-
`lation, cataract surgery (with lens implantation in all but
`two patients) preceded theclinical diagnosis of endoph-
`thalmitis in 95% of cases, and secondary lens implanta-
`tion preceded in the remaining 5%. The median time from
`the cataract extraction or secondary lens implantation un-
`til presentation to a study center was 6 days. Presentation
`within 3 daysofthe initiating procedure occurred in 24%,
`within 4 10 7 days in 37%, within 8 to 13 days in 17%, and
`within 2 to 6 weeks in the remaining 22%. Almostall the
`patients had symptoms,with blurred vision being the most
`common. Pain was reported by 74% of patients.
`Study patients hadpoorinitial vision, with 86% hav-
`ing acuity of less than 5/200. Initial visual acuity was LP
`only in 26% of patients. An afferent pupillary defect was
`present in 12%, corneal ring ulceror infiltrate in 5%, and
`hypopyonin 86%. For patients with a hypopyon, the me-
`dian height was 1 mm, with 30% being higher than 1.5
`mm, Mediaclarityat the initial visit was poor. A second-
`orderretinal vessel could be seen by indirect ophthal-
`moscopyin only 10% ofpatients, and in almost 80% of
`patients, no retinal vessel of any type could be seen with
`indirect ophthalmoscopy.A red reflex was absent in 67%
`ofpatients.
`
`For editorial comment,
`see page 1555
`
`As noted above, the protocol allowed for addi-
`tional surgery in the immediate postoperative period if
`the involved eye was doing poorly. At the 36- to 60-
`hour examination, 29 (7%) of the 420 patients met study
`guidelines to be considered for an additional procedure
`(Table 2). These included 6% (14/218) of eyes in the VIT
`group and 7% (15/202) of eyes in the TAP group. Of eyes
`that metcriteria for additional surgery, 86% (25/29) had
`such a procedure, with no statistical difference between
`the VIT and TAP groups. The clinician had the option
`of performing surgeryoutside the guidelinesif in his or
`her judgmentit was in the best interest of the patient.
`Additional procedures were performed in 4% (14/390)
`of patients who did not meet the guidelines, consisting
`of 2% (3/203) of eyes in the VIT group and 5% (9/187)
`of eyes in the TAP group (nonsignificant difference). Thus,
`in total, based on the 36- to 60-hour assessment, an ad-
`ditional procedure was actually performed in 9% of pa-
`tients, representing 7% (16/218)of eyes in the VIT group
`and 11% (23/202) of eyes in the TAP group. The above
`
`ARCH OPHTHALMOL/VOL113, DEC 1995
`1483
`
`Downloaded From:http://archopht.jamanetwork.com/ by Andrew Calmanon 11/19/2015
`
`Novartis Exhibit 23 10.005
`Regeneron v. Novartis, IPR202 1-008 16
`
`Novartis Exhibit 2310.005
`Regeneron v. Novartis, IPR2021-00816
`
`

`

`- if6(36-91) -
`---, 50.9
`- 3
`()
`}'14.3
`35.7
`_ 6.3
`..
`. l
`;8
`
`=~i [I;;~'.;(;
`
`-
`
`69.0
`94.0
`35. -,-
`
`-
`
`-
`
`9
`: :: - 8
`
`9.
`
`:~;i
`
`History of hype/tension
`~ist~ry of g~(!C9©a
`
`:l'tl~f !f ::,;\: ---·
`
`v~sio~ _•_-
`
`s __ ------•-----
`n rari )
`
`_olfi
`
`<<·
`
`r
`0
`
`urgi~twound abnorriialM
`ak •· (!',t pre,~ - · t
`>
`, yonpt: - /_
`-< i<Median (rarige),)nm
`lntraocular presi;ure, mm Hg
`• Median (rang~)<
`
`11~:~tl~~'
`
`Media clarity (indirect ophthalmoscopy)
`
`(
`'--:·-• ,- ·-- -- (
`(.1-9) -
`
`m<1-5o)
`
`19.8
`
`--. 15.5 (0-50) {. -
`
`c· . : -_ ·: i::.tit:r:
`
`66 (16.2)
`
`Downloaded From: http://archopbt.jamanetwork.com/ by Andrew Calman on 11/19/2015
`
`ARCH OPHTHALMOL/VOL 113, DEC 1995
`1484
`
`Novartis Exhibit 2310.006
`Regeneron v. Novartis, IPR2021-00816
`
`

`

` Table. 1, Baseline Characteristics by Treatment Group (cont)
`
`Characteristic
`White blood.cell count, x10%L
`Median (range)
`> 10.0% 107L
`>14.0*104L
`Creatinine level, mol/L [mg/dL]
`
`Median (range}
`
`% of Patients*
`
`
`
`Vitrectomy With
`{V. Antibiotics
`(n=106)
`
`9.4:(3.4-72.0)
`32.4
`9.4
`
`Vitrectomy With
`No IV Antibiotics
`(n=112)
`
`Tap/Biopsy With
`WV Antibiotics
`{n=100)
`
`Tap/Biopsy With
`No IV Antibiotics
`(n=102)
`
`Total,
`No. (%)
`(N=426)
`
`9:3:(3.5-76.0}
`33.0
`10.7
`
`9:2.(1.3-84.0)
`30.0
`9.0
`
`9.7: (3:3-87.0}
`39.2
`10.8
`
`$.3 (4.3-87.0)
`144 (33.6)
`42 (10.0)
`
`88 (35-168)
`[1.0.(0.4-1,9)}
`40:7
`
`88 (44-362)
`{1.0(0.5-4.1)]
`17.0
`
`88 (44-628)
`{1.0 (0.5-7.1)]
`10.8
`
`88 (35-628)
`{1.0:(0:4-7.1}}
`52.(12.4)
`
`88 (35-354)
`{4.0 (0.4-4.0)]
`1.3
`
`10.4
`20.8
`
`