The Clinical Analysis of Endogenous Endophthalmitis

144The Clinical Analysis of EndogenousEndophthalmitisLingyi Liang, Xiaofeng Lin, Ayong Yu, Aihua Lin* , Zhaohui YuanZhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou 510060, China*Department of Statistics , Sun Yal-sen University, Guangzhou 510089, ChinaPurpose : To study the clinical characteristics, therapeutic eficacy and investigate pro-gnostic factors of endogenous endophthalmitis.Methods: Twenty-eight cases (28 eyes) of endogenous endophthalmitis were surveyedretrospectively. The clinical characteristics, primary infection foci, predisposing systemicdisease, complications ，pathogens examination, therapeutic options and efficacy wereanalysed.Results: The endogenous endophthalmitis occurred more frequently in the right eye thanin the left one. The respiratory tract was the most common primary foci. The positive rate ofpathogens culture was higher in vitreous sample than that in other tissues. Cataract andretinal detachment were the common complications. The visual improvement and infectioncontrol were achieved in 13 eyes (46.43% ). These 13 patients received treatment (3.77土2.49)days after onset of endophthalmitis being much earlier than that of others [(10.13土4.98)days, P=0.002]. The prognosis was relevant to the type of the disease. The anteriorsegment inflammation type (anterior type) had better prognosis than posterior segmentinfammation type(posterior type) and that of inflammation in both parts(mix type) (P<0.05). There were no significant relation between the prognosis and the age, predisposingsystemic disease, vitreous antibiotic injection and vitrectomy (P> 0.05).Conclusions : Endogenous endophthalmitis is a vital ocular emergency. Early diagnosisand effective treatment combination with systemic and local antibiotics are of significantvalue. The anterior type is prone to have better outcome than the others. Eye Science2004;20:144-148.Key words: endogenous endophthalmitis; complication; treatment; prognosisE ndogenous endophthalmitis is a vital ocularThere were totally twenty-eight endogenousemergency being subjected to misdiagnosis endophthalmitis cases (28 eyes) in our hospitaldue to its low incidence. Therefore, the endoge- from 1998 to March 2003，which accounted fornous endophthalmitis has poor therapeutic effic- 17.03% among the endophthalmitis in patientsacy and is prone to blindness, even other severe during the study period, 17 males and 11 females.complications such as orbital cellulitis，eyeball The patients' age ranged from 2 monthes to 59perforation，sepsis and so on. There have been (26.33+2.71 )years. Thirteen patients (46. 41%)only case reports of this entity in China. To were younger than 15 years. All cases were mon-understand this disease better, twenty-eight con- ocular involved, 16 in the right eye and 12 insecutive cases (28 eyes) in our hospital from the left one (right:left = 1.3:1 ). The visual acuities1998to March 2003. were surveyedretrospectively. were no light perception (NLP) to 0.3 beforetreatment and NLP to 0.6 after treatment.MethodsThe patients received effective treatment 2 ~20General materials(7.36中国煤化工of endogenousTYHCNMHGLiangLY, etalEndogenous Endophthalmitisendophthalmitis.(combined with antifungal for suspected fungalEndogenous endophthalmitis is defined asendophthalmitis), steroid and madriatics werethe infectious purulent uveitis and retinitisfrequently used by all patients. 16 patients recei-excluding ocular trauma and surgery'l. Accordingved vitreous injection of 0.1 ml solution of ceph-to the location of infection the endophthalmitisazolin (2.25 mg) or tobramycin (0.4 mg) or mico-can be divided into three types: anterior segmentnazole (20 μg). Vitrectomy was performed in 15inflammation type( anterior type), posterior seg-patients. Eventually, visual improvement as wellment inflammation type( posterior type) and thatas infection control were obtained in 13 patients,of inflammation in both parts( mix type). Amongand infection control without visual improvementthe 28 eyes there were 7 with anterior type (25% )in 12 patients. Visual loss in the rest 3 patientswhich had characteristics of corneal keratic prec-was due to the complicating orbital cellulitis,ipitates, hypopion with or without mild vitreousperforation of cormeal ulcer and ocular atrophy.opacity, 9 with posterior type (32.14%) whichhad characteristics of vitreous pus, retinal puru]-Observation itemsent effusion and mild inflammatory response inThe distribution of sex, age and eye, theanterior chamber, 12 with mix type (42.56% )clinical characteristics and the therapeutic effc-which had characteristics of hypopion and vitre-acy were observed. The positive rate of pathogensous pus. The criteria of infection control are:1) culture of varied samples was compared. Thethe alleviation or elimination of ocular symp-prognostic factors via t test, separated chi-squaretoms; 2) gradul reduction of intraocular empy- test and Fisher's exact test were analyzed. P<0.05ema, and 3) iris or vitreous neovascularization.is the point to judge significance.The patients were examed with slit lamp,ophthalmoscope，tonometer and ocular B-scan.Resultsfundus fluorescein angiography and peripheralserum immunology (RF, HLA-B27, HLA-DR,Clinical featuresHLA-DW, etc.) were perfomed to exclude non-The patients had cute blurred vision with redinfectious uveitis. Other examinations such as eye and pain, without history of ocular trauma orchest X -ray, blood biochemistry test, blood rou-surgery. Hypopion and/or vitreous abscess, ortine test, routine urinalysis, faces routine testpurulent exudation in retina was noted in allwere taken to figure out primary infection focipatients. Systemic disease or depressed immuneand predisposing diseases. Seven blood samplesstate before or on the onset of endogenous endop-of fever patients and 17 intraocular liquidhthalmitis were found in 16 patients (57.16%)，(aqueous humor, vitreous) samples were coll-among which there were 8 hepatitis, 4 diabetes ,ected for bacteria and fungi culture.4 on menses, 1 (child) measles and 1 hyperte-nsion. Primary infection foci were found in 19Therapeutic options and efficacypatients (67.86% ) , among which 8 patientsAll patients received antibacterial via intr-(28.57% ) had respiration infection (bronchitis ，avein infection once the endogenous endo-pneumonia), 4 ( 14.29% )oral infection ( perico-phthalmitis was identified. (Cephazolin 4~5g/d ronitis, decayed tooth), 4 (14.29% ) urinogenitalfor adult and 50 mg* kg-.d-) for child, tobram-infection (urethtitis, renal abscess), 1(3.57%)ycin 240 mg/d for adult and 4 mg*kg-+.d+) for dermatitis, 1(3.57% )ymphadenitis, 1(3.57%)child，, steroid ( dexamethasone) 10 mg/d for ad-tympanitis.Fever before or on the onset ofults). Two patients were suspected fungal endop-endogenous endophthalmitis was found in 19hthalmitis and treated with antibacterial andpatients (67.86% )，leukocytosis in 20 patientsantifungal without steroid. Eyedrops of antibiotics (7中国煤化工MHCNMHG146EYE SCIENCEVol.20 No.3 2004Positive rate of pathogens culturebrachiocephalic artery and close to heart mayAll of the blood cultrues and aqueous hum-contribute to this difference. In our study theor cultures were negative. Five of 17 vitreous cul-incidence rate in the right eye to that in the leftture were positive(29.41%).was 1.3:1.ComplicationsDifferential diagnosisComplicated cataract was found in 20 eyesEndogenous endophthalmitis should be diff-(7 1.43%), retinal detachment in 10eyes (35.71%),erentiated from non-infectious uveitis. The formersecondly glaucomain 2eyes (7.14%), keratitis inusually has much more severe eye irritation and2 eyes (7.14%), choroid detachment in 1 eyeintraocular purulent effusion，and is often acco-(3.56%), orbital cellulitis in 1 eye (3.56%) andmpanied with fever, leukocytosis and primaryocular atrophy in 1 eye (3.56%).infectious foci，but no specific finding by FFAand peripheral serum immunology. Thus， non-Prognostic factors (see chart 1)infectious uveitis was ruled out in our 28Infection control with visual improvementpatients.was considered as good prognosis. Thirteen pati-ents with good prognosis went to a doctor (3.77土Predisposing diseases and primary infections2.49)days after the onset，the rest 15 (10.13土The patients were usually in poor health. In4.98)days after the onset. The difference wasour study systemic diseases were found in 16significant (P=0.002).The anterior type patients patients (57.16% ), among which hepatitis washad good therapeutic eficacy, and their progno- most common (8 cases, 28.57%) and4of 10sis was significantly different from the other twofemale adults were on menses. The most commontypes (P=0.029). No significant difference ofpredisposing diseases with endogenous endopht-prognosis was found between posterior type andhalmitis in developed country were diabetes andmix type (P=0397).Prognosis was not correlatedchronic renal dysfunction B. The above differenceto age, the existence of systemic disease, whether is concomitant with the different disease proport-vitreous antibiotics injection or vitrectomy waion in different countries.performned or not(P > 0.05).The pathogen of endogenous endophthalm-itis comes from the other infectious foci of bodyDiscussiondestroying the blood-ocular barrier via bloodcirculation and cause severe purulent uveitis.IncidencePrimary infection foci were found in 19 patientsThe incidence of endogenous endophtha-(67.86% ) , among which respiration tract was thelmitis is rare but increasing recently(2. Endoge-most common one, followed by oral and urinoge-nous endophthalmitis has not been reportednital tract. So it is important to perform systemicsystemically in China. Twenty-one of 28 patientsexamination for endogenous endophthalmitisin our research arised from 2000, which may be patients to figure out the coexisting systemicrelated to the increasing drug abuse and the usediseases and primary infectious foci.of invadable diagnostic and therapeutic methods.Some of the pediatric patients had used retainedIdeological examinationsvenous needles in other hospitals.Pathogens exam and drug sensitivity tesIt was reported that the incidence of endog-should be performed before the use of antibioticenous endophthalmitis in the right eye was as twoin suspected cases. However, the positive rate oftimes as that in the left eye [D]. The right common cultures was low which may be related to thecarotid artery being straight away from rightimpro中国煤化工he samples orYHCNM HGLiangLY, el al _Endogenous Endophthalmitis147because the antibiotic had been used before. Theside effects of steroid had to be monitoredocular damage caused by anaerobes or toxemiacarefully because of the depressed immunity ofalso attributed to this phenomenon[4]. Among thethe patients. Steroid was contraindicated in sus-different tissues we took，the vitreous samplespected fungal infection. The combination of cep-had the highest positive rate because of thehazolin and tobramycin was advocated before the“culture medium”role of vitreous. Fungus is an culture result was reported because of their broadimportant pathogen of endogenous endophtha-antibacterial spectrum, relatively high permeab-lmitis accounting for more than 50% of the casesility to the blood-ocular barrier and strong effect.and usually has poor prognosislil. But only 1Many doctors considered the ideal intrao-patient in our study was confirmed with fungalcular drug concentration might be achieved viainfection. W e speculate that fungal endophthalm-systemic route because of the damaged blood-itis should have been misdiagnosed in someocular barrier in endogenous endophthalmitis (26.].patients with poor prognosis .The misdiagnosisWe found the prognosis was not signifcantlymay due to the atypical signs coexisting withcorrelated to vitreous injection or vitrectomy. Thisbacterial infection and the low positive rate ofis consistency with what W ong|2 reported. The rolepathogens culture.and the indication of vitreous injection andvitrectomy in the treatment of endogenousComplicationsendophthalmitis remained an open question2,51 toCataract and retinal detachment werebe answered by further research in large samples.common complications in our study. Purulent andbacterial emboli，which cause retinal ischemia,Prognosiswere found in the retinal blood vessels throughThe results are not so saisfactory(.21. For onepathological examination of the enucleated eye-hand, many patients were in depressed immunity.ball. Retinal tear and vitreoretinal traction andFor the other,，the low positive rate of pathogensresult in retinal detachment.culture to some extend results in the confusion ofantibiotics choose. The visual improvement andTreatmentsinfection control were achieved in only 13 eyesEffective antibacterial was used systemic-(46.43%). The period from the onset of diseaseally and locally as soon as the samples wereto receiving effective treatment was (3.77+2.49)collected. Combination with the steroid to allevi- dyas in these 13 patients while more than 10ate the intraocular damage and proliferationdays in 3 patients who failed to infection controlresulted from inflammation is advocated. Theand unfortunately lost vision. The prognosis wasChart 1. analysis of prognostic factors(Fisher's exact test)factorsgood prognosis(eyes) bad prognosis(eyes)P valueage<15y9>15y60.943systemicwithdiseasewithout0.276Vitrectomyyesno0.988vitreousinjection1o0.0943Typeanteriorposterior50.029mix中国煤化工_YHCNMHG148EYE SCIENCEVol.20 No.3 2004relevant to the type of the disease. The anterior2. Wong JS, Chan TK, Lee HM, et al. Endogenoustype had better prognosis than posterior type andbacterial endophthalmitis an East Asian experiencemix type for the higher permeabilty of the blood-and a reappraisal of a severe ocular affliction.Ophthalmology, 200; 107(8): 1483-1491.ocular barrier in the anterior segment than that ofblood-retina barrierl4. Accordingly the concentra-3. Annabelle A, Donald J. Endogenous endophthalmi-tis. Foundations of clinic ophthalmology. USA,tion of medicine in anterior segment is higherphiladelphia: Lippincott Williams&Wilkins. 1999: .than that of the posterior segment. The pathogens3120-3125.in the anterior segment are more likely to be4. Greenwald MJ, Wohl LG ,Sell CH. Metastatic bact-cleared away because of the rapid circulationerial endophthalmitis: a contemporary reappraisal.and refresh of aqueous humor. In conclusion ,Suro Ophthalmol, 1986; 31(2): 81-101.early diagnosis and the effective combination of5. Okada AA，Johnson PJ, Liles WC, et al. Endog-systemic and local therapy when the pathologicalenous bacterial endophthalmitis report of a ten-year retrospective study. Ophthalmology， 1994;changes are limited in anterior segment play an101(5): 832-838.important role in infection control, complication5. Shrader SK, Band JD, Lauter CB, et al. The clinicalprevention and visual improvement.spectrum of endophthalmitis: incidence , predispos-ing factors，and features influencing outcome. JReferencesInfect Dis, 1990; 162(1): 115-120.7. Zhang SC, Gao RL, HuJ, et al. Endophthalmitis1. Gu XQ. Infectious Uveitis. In: Li FM eds. System ofAfter Pas Plana Vitrectomy. Acad JSums, 1998; 19Ophthalmology .Beijing: People's Hygiene Publish(4):292-294.House, 1996: 2135-2147.(Editor:Qifan Wang)中国煤化工MYHCNMHG