Optimizing Diagnosis and Management of Nocardia Keratitis, Scleritis, and Endophthalmitis: 11-Year Microbial and Clinical Overview

Optimizing Diagnosis and Management of Nocardia Keratitis, Scleritis, and Endophthalmitis: 11-Year Microbial and Clinical Overview

Optimizing Diagnosis and Management of Nocardia Keratitis, Scleritis, and Endophthalmitis: 11-Year Microbial and Clinical Overview Francis Char DeCroo...

1MB Sizes 0 Downloads 67 Views

Optimizing Diagnosis and Management of Nocardia Keratitis, Scleritis, and Endophthalmitis: 11-Year Microbial and Clinical Overview Francis Char DeCroos, MD,1 Prashant Garg, MD,2 Ashok K. Reddy, PhD,3 Ashish Sharma, MD,4,6 Sannapaneni Krishnaiah, MSc, MPS,5 Meeta Mungale, MD,2,7 Prithvi Mruthyunjaya, MD,1 for the Hyderabad Endophthalmitis Research Group4* Objective: To identify clinical factors and microbiological assays that facilitate a rapid diagnosis of Nocardia keratitis, scleritis, and endophthalmitis, and to determine optimal medical and surgical management strategies. Design: Retrospective, consecutive case series. Participants: A total of 111 cases of keratitis, 11 cases of scleritis, and 16 cases of endophthalmitis, all culture-proven Nocardia infections, were identified between January 1999 and January 2010. Intervention: The keratitis cases underwent intensive medical management, and the scleritis and endophthalmitis cases required concurrent surgical intervention for disease control. Corneal and scleral scrapings, as well as undiluted vitreous sample, were submitted for microbiologic evaluation (direct smear and culture). Main Outcome Measures: Historical points, clinical findings, and microbiologic assays that facilitated a prompt Nocardia diagnosis were identified, and management choices were examined for correlation with final acuity. Results: Ocular exposure to soil or plant matter was a common historical point in cases of Nocardia keratitis (48%) and scleritis (45%), respectively. Nocardia keratitis often (38.7%) presented with “wreath”-shaped anterior stromal infiltrate or infiltrate interspersed with elevated, pinhead-sized, chalky lesions. Most patients with scleritis (63.4%) presented with nodular lesions demonstrating pointed abscesses. Nocardia endophthalmitis typically (75%) presented with endoexudates or nodular exudates surrounding the pupillary border. Gram stain and 1% acid-fast stain enabled prompt diagnosis of Nocardia in 64% and 63% of keratitis cases and 45% and 63% of scleritis cases, respectively. Direct smear was usually not revealing in cases of Nocardia endophthalmitis. Isolates from Nocardia keratitis, scleritis, and endophthalmitis demonstrated 97%, 100%, and 90% susceptibility to amikacin, respectively. Nocardia keratitis resolved with medical therapy alone in 82% of cases. Younger age and better initial acuity correlated with improved final acuity in keratitis cases. Outcomes were poor after Nocardia scleritis and endophthalmitis. Conclusions: Early appropriate treatment often results in visual recovery in eyes with Nocardia keratitis. Despite aggressive and prompt surgical intervention, the prognosis for Nocardia scleritis and endophthalmitis is more guarded. Nocardia isolated from ocular infections demonstrate high levels of susceptibility to amikacin. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2011;118:1193–1200 © 2011 by the American Academy of Ophthalmology.

Organisms in genus Nocardia are gram-positive, weakly acid-fast, filamentous bacteria that are a rare but significant cause of ocular infection. These saprophytic bacteria are ubiquitous and found in water, soil, dust, and decaying vegetation worldwide. Although Nocardia are not commensal organisms, they can be found as a saprophyte on the skin and upper respiratory tract.1 Most known for causing pulmonary disease in immunocompromised patients, these organisms have been increasingly identified as a cause of ocular morbidity in immunocompetent individuals.2–5 © 2011 by the American Academy of Ophthalmology Published by Elsevier Inc.

Nocardia ocular infections, such as keratitis, scleritis, and endophthalmitis, are characterized by a protracted clinical course and poor visual outcomes.3,6,7 This pathogen is frequently resistant to typical first-line ocular antibiotics, such as fluoroquinolones,6 and delay in diagnosis can lead to suboptimal outcomes. This delayed diagnosis may be in part due to a clinician’s lack of familiarity with this uncommon pathogen. In reports from multiple series worldwide analyzing keratitis,8 –12 scleritis,13 and endophthalmitis,14 –17 Nocardia isolates were reported at less than 2% frequency. ISSN 0161-6420/11/$–see front matter doi:10.1016/j.ophtha.2010.10.037

1193

Ophthalmology Volume 118, Number 6, June 2011 Expeditious recognition of the clinical signs of Nocardia infection (keratitis, scleritis, or endophthalmitis) combined with appropriate microbiologic diagnostic testing may help clinicians improve anatomic and functional outcomes. In particular, microbiologic data can facilitate both rapid diagnosis and help tailor therapy later in the clinical course. This study focuses on identifying distinctive and characteristic clinical features and appropriate microbiology tests for expeditious diagnosis of Nocardia infection. To this end, our group analyzed 111 cases of Nocardia keratitis, 11 cases of scleritis, and 16 cases of endophthalmitis. All 138 cases originated from various parts of India, and the majority began in South India.

Materials and Methods A retrospective study of consecutive cases of Nocardia keratitis, scleritis, and endophthalmitis managed in an academic setting from January 1, 1999 to January 1, 2010, was performed. All patients were evaluated at the Kallam Anji Reddy campus of L V Prasad Eye Institute (Hyderabad, India) and abstracted from a computerized microbiology database cataloging all ocular infections. The study was performed with institutional review board approval (L V Prasad Eye Institute: LEC08122, Duke Eye: Pro00018692). Only culture-positive cases of Nocardia were included in this study. Details collected included demographic details, nature of inciting event if any, time course until evaluation, treatment before arrival at our center, medical and ocular history, baseline clinical examination, microbiologic identification, antibiotic susceptibilities, medical and surgical treatment, clinical course, and final outcome. All cases had a detailed history and slit-lamp biomicroscopic examination. We made note of any prior trauma and surgery, including cataract surgery, penetrating keratoplasty (PK), and open globe repair. If the fundus could not be visualized, B-scan ultrasonography was performed using the 10-MHz handheld transducer looking for evidence of vitreoretinal or choroidal pathology, as well as for any intraocular foreign body. Microbiology workup consisted of microscopic examination of smears using various staining techniques and inoculation of specimen on culture media that facilitate growth of bacteria, fungi, and Acanthamoeba. The specimens used for microbiology workup consisted of corneal scrapings, scleral scrapings after conjunctival excision and scleral de-roofing, or undiluted vitreous obtained during vitreous biopsy or vitrectomy. Each sample was smeared on clean presterilized glass slides for microscopic examination using Gram stain, Giemsa stain, and potassium hydroxide (KOH) with calcofluor white preparation. One percent or 20% acid-fast stains (Kinyoun stain, Ziehl-Neelsen stain) were used when there was a strong clinical suspicion of Nocardia or atypical mycobacterial infection. The specimens were also inoculated onto sheep blood agar, chocolate agar, Sabouraud dextrose agar, potato dextrose agar, thioglycollate broth, brain heart infusion broth, and non-nutrient agar with Escherichia coli overlay (for Acanthamoeba). All media except Sabouraud dextrose and potato dextrose agar were incubated at 37°C for a period of 7 days. Sabouraud dextrose and potato dextrose agar were incubated at 27°C for a period of 14 days. A positive smear for Nocardia was defined as gram-positive thin branching filaments on Gram stain (Fig 1A, available at http://aaojournal.org) or thin branching filaments on Giemsa stain, KOH preparation, or acid-fast staining (Fig 1B, available at http://aaojournal.org). Because Nocardia is only weakly pos-

1194

itive on acid-fast staining, positive acid-fast staining was not required of the filaments. The culture was considered significant when there was (1) growth of the same organism on 2 or more media; (2) confluent growth at the site of inoculation on 1 solid media; or (3) growth in 1 medium with consistent direct microscopy findings.18 Speciation was performed by biochemical reactions until 2007 and afterward by 16S rRNA gene sequencing.19,20 Antibiotic susceptibility was performed using the E-test21 (AB Biodisk, Stockholm, Sweden) for 20 isolates, and the rest were tested using the disc diffusion method.22 Susceptibility was recorded for amikacin, cefazolin, ofloxacin, gentamicin, vancomycin, gatifloxacin, moxifloxacin, ciprofloxacin, chloramphenicol, and trimethoprim/sulfamethoxazole when available. Initial treatment was started on the basis of smear results and modified according to clinical response, culture, and antibiotic susceptibility testing. The treatment protocol used in the institute during the study period was as follows:

Nocardia Keratitis All patients in whom smear examination did not reveal bacteria, fungi, or parasite received treatment with topical broad-spectrum antibiotics. This was topical ciprofloxacin 0.3% or ofloxacin 0.3% monotherapy until 2002 and a combination of cefazolin 5% with ciprofloxacin 0.3% after 2002. This change was necessary in light of the decreasing efficacy of second-generation fluoroquinolones against gram-positive organisms.12 If direct smear examination revealed acid-fast organisms or thin beaded branching filaments, the patients were administered topical amikacin 2.5%. Topical antibiotic drops were given hourly on an outpatient basis for less severe cases and on an inpatient basis for severe keratitis cases, including those with complications. The patients with Nocardia keratitis who showed slow response or worsening on topical amikacin were also administered oral trimethoprim/sulfamethoxazole. Resolution of keratitis was defined as lack of both epithelial defect and stromal infiltrate. Healing time was defined as time from initial presentation to complete resolution of keratitis. Surgical treatment was considered for all patients not responding to the medical treatment, presenting with large infiltrate threatening scleral involvement, or developing gross thinning or perforation. The surgical options were application of tissue adhesive and bandage contact lens or PK.

Nocardia Scleritis Initial medical therapy was based on clinical diagnosis and direct smear results. Acid-fast organisms or thin, beading, branching filaments suggestive of Nocardia led to initial treatment with topical amikacin. In addition, patients with Nocardia scleritis received systemic therapy with intravenous amikacin or oral trimethoprim/sulfamethoxazole. Treatment was later modified on the basis of clinical response and antibiotic susceptibility of cultured isolated. Surgical debridement was performed to facilitate antibiotic penetration if necrotic tissue was thought to be impeding healing despite appropriate medical treatment of susceptible organisms. Resolution of scleritis was defined as resolution of scleral ulceration along with lack of redness and subjective report of pain. Healing time was defined as time from initial presentation to complete resolution of scleritis. For this study, cases involving both cornea and sclera were classified as follows: (A) Eyes with large scleral lesion with contiguous corneal infiltrate and positive culture only from scleral scraping were classified as scleritis. (B) Eyes with a large corneal lesion with adjacent scleral lesion and positive culture only from corneal scraping were classified as keratitis. (C) Eyes with non-

DeCroos et al 䡠 Strategies for Early Diagnosis and Treatment of Nocardia adjacent lesions were classified according to site of positive culture (i.e., corneal or scleral lesion), and eyes with positive culture from both corneal and scleral lesions were classified in the scleritis subgroup. (D) Eyes with keratitis that later developed scleritis were classified in the keratitis subgroup and designated secondary scleritis.

Nocardia Endophthalmitis All patients were admitted for 3 to 7 days and treated preoperatively with topical antibiotics, steroids, and cycloplegics. Preliminary surgical management depended on severity of initial presentation and was instituted within 6 hours of initial presentation. In eyes with poor visual potential, defined as presenting vision of no light perception or total retinal detachment, vitreous tap was performed as described previously.23 Immediately afterward, intravitreal antibiotic injection was performed with vancomycin (1 mg/ 0.1 ml) and amikacin (400 mg/0.1 ml) or ceftazidime (2.25 mg/0.1 ml). A high initial suspicion of fungal infection in some cases prompted an intravitreal injection of amphotericin B (5 mg/0.1 ml). Eyes presenting with light perception or better vision received core pars plana vitrectomy (PPV) with or without primary lensectomy. Vitreous biopsy was performed as described previously during this intervention.23 Routine anterior chamber sampling was not performed, except in 1 case. After vitrectomy, intravitreal antibiotics were used as described. After initial surgery, the patients received topical antibiotic therapy (amikacin, ciprofloxacin, or gentamicin), and some patients were administered systemic antibiotics (intravenous amikacin or oral ciprofloxacin) and anti-inflammatory agents at the discretion of the managing clinician. Therapies were adjusted according to clinical response. In some cases, further surgery including repeat intravitreal antibiotic injection or intraocular lens (IOL) explantation was required.

Statistical Analysis We performed paired-sample t tests for data that followed the normal distribution to find the differences between baseline and final logarithm of the minimal angle of resolution (logMAR) best-corrected visual acuity (BCVA). We also performed independent samples t test to find the significant difference of mean final visual acuity between groups. Nonparametric Wilcoxon signedrank tests were performed for the data non-normally distributed for related samples. For related samples, the McNemar test was performed for comparisons and the chi-square test was performed for independent groups to assess the associations between groups. Alpha was set to 0.05 for a 2-tailed test. The Statistical Package for the Social Sciences version 16.0 (SPSS Inc, Chicago, IL) was used for all statistical analysis.

Results Nocardia Keratitis Of the 111 culture-proven cases of Nocardia keratitis identified during the study period, 63 were male and 48 were female. The mean age was 36.9⫾15.1 years. The most common presenting symptoms were ocular pain (102 patients), redness (92 patients), and watering (90 patients). Twenty-seven patients (24.3%) reported an agriculture-related occupation. Twenty-two patients (19.8%) endorsed ocular soil exposure, and 32 patients (28.8%) confirmed ocular plant or animal matter exposure. Mean time to presentation from onset of symptoms was 24.5⫾22.2 days. Fiftyone patients (47.2%) presented within 15 days of onset of initial

symptoms, and 57 patients (52.7%) waited more than 15 days. Time to presentation was not recorded for 3 patients. A small subset of eyes had undergone ocular surgery before the onset of infection: LASIK (4), PK (3), extracapsular cataract extraction (2), and phacoemulsification (1). No patient had any known history of immunosuppressive illness. Before referral to L V Prasad Eye Institute, 97 patients received treatment from referring physicians, but no patient reported sustained improvement. The prior treatment regimens consisted of exclusively antibacterial agents (n ⫽ 38 eyes, [51.4%]), exclusively antifungal agents (n ⫽ 11, [14.9%]), exclusively antiviral agents (n ⫽ 3, [4.1%]), or combination antibacterial and antifungal therapy (n ⫽ 22, [29.7%]). The exact nature of treatment before referral for the remaining 23 patients was not available in the medical records. The right eye was affected in 64 patients (57.6%), and the left eye was affected in 47 patients (42.3%). One patient presented with keratitis in both eyes 5 days after bilateral LASIK surgery. The clinical features of all the patients with keratitis are summarized in Table 1 (available at http://aaojournal.org). The mean initial BCVA was 1.5⫾0.99 logMAR units (Snellen equivalent 20/632), improving to 0.9⫾0.97 logMAR units (20/159) at final follow-up. Distinctive and characteristic anterior segment findings6 were appreciated in 52 patients. These were “wreath”-like circular anterior stromal infiltrate in 33 eyes (29.7%, Fig 2A), “pinhead”-size raised yellow infiltrates in 10 eyes (9.0%, Fig 2B), and endoexudates in 9 eyes (8.1%). Nocardia keratitis also rarely presented with dry cotton wool infiltrates mimicking fungal keratitis (2.7%, Fig 2C). Four cases of secondary sclerokeratitis were also observed. Scleral extension was contiguous to the corneal infiltrate, and organisms were only cultured from corneal scraping. On the basis of clinical examination at the first visit, Nocardia was considered as the etiologic agent in 51 patients (45.9%). Although the infection resolved to medical therapy in the majority (82.0%) of patients, a small subset required surgical intervention. All 4 eyes with secondary sclerokeratitis required debridement or tissue adhesive for impending perforation. Of the remaining 107 eyes, 6 (5.6%) underwent tissue adhesive, 3 (2.8%) underwent debridement, 2 (1.9%) underwent therapeutic PK, 2 (1.9%) underwent corneal biopsy, 2 (1.9%) underwent vitreous biopsy and intravitreal antibiotics (vancomycin [1 mg/0.1 ml] and amikacin [400 mg/0.1 ml]), and 1 (0.9%) underwent interface wash with amikacin post-LASIK. During the course of treatment, 2 more eyes (1.9%) required tissue adhesive for corneal perforation and 2 more eyes (1.9%) required therapeutic PK. Mean follow-up was 172.7⫾53 days, with 73 eyes demonstrating complete resolution and 27 eyes demonstrating significant resolution before loss of follow-up. No recurrence was noted in 4 eyes treated with therapeutic PK. One eye was eviscerated, and another patient was advised evisceration for a blind, painful eye but declined. Insufficient follow-up was available for 5 patients to judge the clinical course. Mean healing time (time to complete resolution of infection) was 38.3⫾16.8 days. Final mean BCVA (0.9⫾0.97 logMAR units [Snellen equivalent 20/158]) was significantly improved from initial acuity (P⬍0.0001). Patients with better initial BCVA acuity had better final BCVA. Patients younger than 40 years of age more often (79.1%) had 20/200 or better vision at final follow-up (43.6%) compared with older patients (P⬍0.0001). Time duration to presentation was not correlated with final BCVA. Early presentation was not correlated to wreath-like stromal infiltrates noted on initial examination. Review of microbiology data showed that Gram stain and 1% acid-fast stain most often (64% and 63%) facilitated early diagnosis of Nocardia. The direct smear results are summarized in Figure 3. Of the 111 Nocardia isolates, 51 were speciated as Nocardia asteroides complex, 7 as Nocardia levis, 7 as Nocardia brasiliensis, 6 as Nocardia farcinica, 4 as Nocardia cyriacigeorgica, 4 as

1195

Ophthalmology Volume 118, Number 6, June 2011

Figure 2. Characteristic clinical findings from Nocardia keratitis and scleritis infections. A, “Wreath”-like ulcerative corneal lesion. B, Corneal infiltrate with multiple “pinhead”-shaped lesions. C, Dry “cotton wool” corneal infiltrate mimicking fungal infection. D, Scleral infection presenting with multiple nodular lesions. E, Ulcerative lesion on sclera. F, Inferotemporal scleral nodule with pointing abscess and adjacent ulcerative lesion.

Nocardia beijingensis, 3 as Nocardia abscessus, 2 as Nocardia amamiensis, 2 as Nocardia thailandica, 1 as Nocardia otitidiscaviarum, 1 as Nocardia pneumoniae, and 1 as Nocardia cornea; the remaining 22 isolates were not speciated further. Eighteen cases were associated with polymicrobial infection. The most common associated organism was Staphylococcus epidermidis (55.6%). The results of in vitro antimicrobial susceptibility are shown in Figure 4. The percentage of in vitro antimicrobial susceptibility of the cultured isolates to amikacin, gentamicin, and gatifloxacin was 97%, 85%, and 75%, respectively.

Nocardia Scleritis We identified 11 culture-proven cases of Nocardia scleritis managed during the study period. All the patients were male. The mean

1196

age was 56.8⫾13.0 years. Common presenting symptoms were ocular pain (11 patients), redness (8 patients), and blurred vision (6 patients). Six patients reported an agriculture-related occupation. One patient endorsed ocular exposure to soil, and 4 patients confirmed plant matter exposure. Four patients (36.4%) presented within 15 days of onset of initial symptoms, and 7 patients (63.7%) waited more than 15 days. Mean time to presentation from onset of symptoms was 30.4⫾16.8 days. No patient had any known history of immunosuppressive illness or autoimmune condition. Nine patients (81.8%) received treatment before arrival to study site, but none reported sustained improvement. Treatment consisted of exclusively antibacterial agents (n ⫽ 6 eyes, [54.5%]), exclusively anti-inflammatory drugs (n ⫽ 2, [18.2%]), or combination antibacterial and antifungal therapy (n ⫽ 1, [9.1%]).

DeCroos et al 䡠 Strategies for Early Diagnosis and Treatment of Nocardia 100%, 82%, and 82% in vitro susceptibility to amikacin, gentamicin, and vancomycin, respectively (Fig 4).

Nocardia Endophthalmitis

Figure 3. Positive detection rate of diagnostic laboratory assays for Nocardia keratitis and scleritis. One percent or 20% acid-fast stains ⫽ Kinyoun stain or Ziehl–Neelsen stain; KOH ⫽ potassium hydroxide with calcofluor white preparation.

Eight right eyes and 3 left eyes were affected, and no patient had bilateral scleritis or any other concurrent infection in the other eye. Patients presented with nodular scleritis (45.5%, Fig 2D), ulcerative scleritis (36.4%, Fig 2E), or both simultaneously (18.1%, Fig 2F). Two eyes had undergone prior ocular surgery less than 45 days before presentation: One patient had scleral buckle for posttraumatic retinal detachment, and the buckle was removed for suspected infection and found to have scleral infection. The second patient had undergone PPV for vitreous incarceration after open globe repair, and infection occurred at the site of scleral tear. The mean initial BCVA was 1.4⫾1.05 (20/502). Four cases of concurrent sclerokeratitis were observed. In all but 1 case, scleral extension was contiguous to the corneal infiltrate. In that single case, the clinical picture comprised a temporal scleral nodule with nasal corneal infiltrate, with both the scleral and corneal scraping yielding Nocardia asteroides complex with the same susceptibilities. All 11 patients required some surgical intervention: Eight eyes underwent scleral debridement, 2 eyes underwent conjunctival excision, and 1 eye underwent tissue adhesive application for impending perforation. Topical treatments consisted of amikacin, ciprofloxacin, or cefazolin, with 3 patients receiving intravenous amikacin. Four patients required 2 or more surgical debridements. Mean follow-up was 59.2⫾24.9 days, with 7 eyes demonstrating complete resolution and 3 eyes demonstrating significant resolution before loss of follow-up. No evisceration or enucleation procedure was required. The clinical course of all Nocardia scleritis cases are summarized in Table 2 (available at http://aaojournal.org). Mean healing time was 59.2⫾24.9 days. Final mean BCVA was 1.3⫾1.07 (20/399), which was not statistically different from initial BCVA. Better initial BCVA, younger age, or presentation within 15 days of symptom onset did not correlate with improved final BCVA. Of the 11 Nocardia isolates, 9 were speciated as Nocardia asteroides complex, 1 as Nocardia brasiliensis, and 1 as Nocardia farcinica. Two cases of polymicrobial infection with Staphylococcus epidermidis were noted. The 1% acid-fast stain most often (63%) facilitated early diagnosis of Nocardia. The direct smear results are summarized in Figure 3. Cultured isolates demonstrated

Of the 16 culture-proven cases of Nocardia endophthalmitis identified in our series, 8 occurred in men and 8 occurred in women. The mean age was 58.8⫾18.2 years. Common presenting symptoms were ocular pain (12 patients), redness (10 patients), and blurred vision (9 patients). Five patients reported an agriculture-related occupation. The endophthalmitis occurred after cataract surgery in 14 patients and after ocular trauma in 2 patients. No patients endorsed ocular soil or plant exposure. No patients had any known history of immunosuppressive illness. Four patients presented within 1 week of onset of initial symptoms, whereas 12 patients waited more than 1 week. Mean time to presentation from symptom onset was 21.5⫾20.6 days. Mean duration from surgery to presentation was 47.5⫾49.4 days. Fourteen patients received medical treatment before arrival, but none reported sustained improvement. Ten right eyes and 6 left eyes were affected, and no patient demonstrated concurrent infection in the other eye. The clinical courses of all Nocardia endophthalmitis cases are summarized in Table 3 (available at http://aaojournal.org). The mean initial BCVA was 2.6⫾0.56 logMAR units (Snellen equivalent 20/7962). Twelve eyes demonstrated yellow white anterior chamber exudates suggestive of Nocardia endophthalmitis. One eye presented with pupillary block. All 16 patients required surgical intervention: Six eyes were subjected to PPV, intraocular antibiotic injection (IOAB), and IOL explantation; 4 eyes underwent PPV, IOAB, and anterior chamber wash; 2 eyes underwent PPV and IOAB; 1 eye underwent vitreous biopsy with IOAB and IOL explantation; 1 eye underwent vitreous biopsy with IOAB; 1 eye underwent PPV/IOAB/foreign body extraction; and 1 eye underwent evisceration. The details of the antibiotics used in these patients are given in Table 3 (available at http://aaojournal.org). The pupillary block was successfully relieved with laser peripheral iridotomy. Three patients required more than 1 surgery for final control of infection. One patient was advised evisceration, although the patient declined.

Figure 4. In vitro antibiotic susceptibility rates of Nocardia isolates from keratitis, scleritis, and endophthalmitis infections.

1197

Ophthalmology Volume 118, Number 6, June 2011 Topical treatments consisted of amikacin, ciprofloxacin, gentamicin, or cefazolin, with 2 patients receiving intravenous amikacin. Mean follow-up was 59.9⫾53.8 days, including 3 eyes with no perception of light. Final mean BCVA was 2.1⫾0.89 logMAR units (Snellen equivalent 20/2517). Of the 16 culture-positive cases, 8 were speciated as Nocardia asteroides complex and 8 were not further speciated. No case of polymicrobial infection was noted. Direct microscopic examination of vitreous biopsy samples was negative in all 16 cases. However, anterior chamber aspirate from case 16 was positive both on smear examination (Gram stain, Giemsa stain, acid-fast 1%, and KOH preparation) and on culture. Ninety percent of cultured isolates demonstrated in vitro susceptibility to amikacin. The remaining antibiotic susceptibility results are summarized in Figure 4. There were no statistically significant correlations between final BCVA and any of the following: initial BCVA, age, presentation within 10 days of symptom onset, intravenous amikacin use, or IOL explantation.

Discussion Nocardia is an organism typically found in the environment and is thought to be more prevalent in the soil of South Asia.24,25 Although this bacterium is ubiquitous in nature, it is a rare cause of ocular morbidity. Of 2511 culture-proven bacterial keratitis cases managed at L V Prasad Eye Institute over a 10-year period, Nocardia comprised only 46 (1.8%) isolates.8 The virulence of Nocardia is likely to be mediated by mechanisms such as cell wall toxins,26 secreted exotoxins,27 and host defenses evasion via free radical scavenging.28 Ocular infections by this organism have been described only in smaller series3,7,29 –32 and case reports,33– 43 and this is the first large series spanning multiple subtypes of ocular Nocardia infections. Our study seeks to better define the utility of characteristic clinical findings and the efficacy of rapid microbiologic assays to facilitate early diagnosis and identify appropriate empiric treatment. Nocardia keratitis typically has an indolent course, although fulminant cases have been observed. The majority (52.8%) of our patients presented more than 2 weeks after the onset of symptoms. Many patients with keratitis (40%) presented with patchy anterior stromal infiltrates arranged in a “wreath” pattern or yellow white, raised, pinhead-sized lesions in the infiltrated area. These findings are almost pathognomic of Nocardia keratitis, and a diagnosis of Nocardia was suspected in 46.8% of eyes after initial clinical examination. Microscopic examination of smears established the etiologic diagnosis in 81.1% of cases. Gram stain and 1% acid-fast were positive in 64% and 63% of samples, respectively. The rate of positive smears (both Gram stain and 1% acid-fast) was lower compared with a report by Lalitha et al29 (84% and 69% positive, respectively) and may be due to the more than 2-fold higher rate of prior antibacterial treatment in our population. Most patients with keratitis required prolonged treatment averaging 38 days. However, appropriate diagnosis and extended treatment generally led to favorable visual outcomes. In our study, 50 of 111 patients (45%) had ⱖ20/50 BCVA and 72 of 111 patients (65%) had ⱖ20/200 BCVA at final follow-up consistent with prior smaller series.29,31

1198

Also, a lower percentage of Nocardia keratitis cases (19/ 111, 17%) required surgical intervention when compared with bacterial keratitis in aggregate (799/1849, 43.2%).8 Poor outcome was in particular associated with older age and scleral extension of keratitis. In at least 1 case, poor compliance with medication resulted in scleral extension of a primary corneal infection. This case emphasizes that prompt initiation of treatment with frequent monitoring of clinical course might avoid visually devastating sequelae, such as secondary scleritis after keratitis. Although infectious scleritis accounts for less than 10% of all cases of scleritis,44 prior work by our group noted Nocardia as the second most common cause of infectious scleritis in Southern India.3 In endemic regions, therefore, it is important to consider Nocardia in the differential diagnosis of infective scleritis cases. However, the diagnosis based on clinical examination is difficult because most patients present with nonspecific clinical signs, such as injection, scleral nodule formation, and ulceration. In our study, Nocardia scleritis was suspected in only 1 patient after initial clinical examination. History may provide an important clue, because most (72.7%) of our patients reported history of ocular trauma, and many (45.5%) of these had ocular plant or soil contact. Direct smear examination was instrumental in early diagnosis, particularly 1% acidfast staining, which was positive in 62.5% of cases. Early diagnosis and appropriate medical therapy were insufficient for infection control. All 11 patients of this series required initial surgery, and eventually an aggregate of 19 surgical interventions were performed. An ophthalmologist managing Nocardia scleritis should be aware that early (and possibly repeated) surgical intervention will be required for managing these cases.45– 47 Although the infection was controlled in all 11 patients, we did not observe significant improvement in BCVA from initial presentation. Nocardia endophthalmitis is a more common cause of ocular morbidity in some parts of the world.2,7,32 Clinical examination and microbiologic analysis are crucial in facilitating an early diagnosis. In our series, 75% of patients presented with distinctive anterior chamber findings, such as the exudates on corneal endothelium or nodule-like deposits at the pupillary border. Haripriya et al7 also reported a similarly high incidence (83%) in another large series of Nocardia endophthalmitis cases. The intense anterior chamber involvement coupled with minimal vitritis and other posterior segment findings in exogenous Nocardia endophthalmitis is in sharp contrast with the typical large choroidal lesions with overlying hemorrhage reported in endogenous Nocardia endophthalmitis.48,49 Microbiologic analysis of vitreous and aqueous samples complements clinical examination in establishing a prompt diagnosis of Nocardia endophthalmitis. However, in this cohort of patients, direct smear examination of vitreous biopsy samples did not have any utility in the early identification of Nocardia. Haripriya et al7 also reported a similar experience. The group reported a 31% positive smear rate from anterior chamber samples.7 In the single patient of this series in whom aqueous humor was subjected to microbiology workup, Nocardia was identified both on smear ex-

DeCroos et al 䡠 Strategies for Early Diagnosis and Treatment of Nocardia amination and culture, suggesting that aqueous tap may be of value in this situation. As observed in the scleritis subgroup, eyes with endophthalmitis more frequently required multiple surgical interventions and often required IOL explantation. Intraocular lens explantation was performed in 9 eyes to achieve control of infection. This subgroup included 2 patients who required IOL explantation as secondary procedure after primary PPV/IOAB. On the basis of this observation, we recommend that vitreoretinal surgeons should consider primary IOL explantation for cases of Nocardia endophthalmitis. Overall, outcomes of Nocardia endophthalmitis were poor in this series. We attribute the suboptimal outcomes to delayed presentation and extensive anterior chamber involvement. A BCVA of ⱖ20/400 was obtained in 5 of 16 eyes in our group. Other investigators also had similar experience with 6 of 37 eyes (described across 1 case series and multiple case reports) obtaining final ⱖ BCVA of 20/400.7,32,33,35,37,39,41 Sophisticated Nocardia differentiation techniques have identified species level differences in virulence and antibiotic susceptibility.1,21 In clinical practice, however, this speciation is often delayed, so there is need for effective empiric antiNocardia therapy. Amikacin is a frequently chosen option for Nocardia infections.1,20,29 Our study supports this selection by showing that ocular Nocardia isolates demonstrated a high level of susceptibility to amikacin (89.7%), followed by moderate sensitivity to gentamicin (79.5%), gatifloxacin (65.2%), and vancomycin (64.3%). In vitro studies have demonstrated Nocardia isolates having a 100% susceptibility to linezolid,21,50 and this oral drug may be a useful supplementary treatment in refractory cases of ocular Nocardiosis. With our study or any other using in vitro susceptibility testing, it is essential to remember that breakpoints of in vitro assays are based on predicted bacterial response to concentrations achievable in serum.51 Topical application, however, may achieve higher concentration in the corneal tissue than can be reached in the serum.52,53 In conclusion, this large series highlights diagnostic microbiologic assays and management strategies for ocular infections caused by Nocardia. Nocardia keratitis often presents with characteristic clinical findings, and early identification and appropriate therapy can help preserve vision in these cases. On the contrary, Nocardia scleritis and endophthalmitis often require aggressive surgical and medical management and carry a guarded visual prognosis. Although careful microscopic examination is useful in etiologic diagnosis of keratitis and scleritis cases, it is often inconclusive in endophthalmitis cases. On the basis of in vitro susceptibilities and aggregated clinical experience from our large series, amikacin remains a robust antibiotic choice for management of ocular Nocardia infections.

References 1. Brown-Elliott BA, Brown JM, Conville PS, Wallace RJ Jr. Clinical and laboratory features of the Nocardia spp. based on current molecular taxonomy. Clin Microbiol Rev 2006; 19:259 – 82.

2. Lalitha P, Rajagopalan J, Prakash K, et al. Postcataract endophthalmitis in South India: incidence and outcome. Ophthalmology 2005;112:1885–90. 3. Jain V, Garg P, Sharma S. Microbial scleritis-experience from a developing country. Eye (Lond) 2009;23:255– 61. 4. Bharathi MJ, Ramakrishnan R, Meenakshi R, et al. Microbial keratitis in South India: influence of risk factors, climate, and geographical variation. Ophthalmic Epidemiol 2007;14:61–9. 5. Ramakrishnan R, Bharathi MJ, Shivkumar C, et al. Microbiological profile of culture-proven cases of exogenous and endogenous endophthalmitis: a 10-year retrospective study. Eye (Lond) 2009;23:945–56. 6. Sridhar MS, Gopinathan U, Garg P, et al. Ocular Nocardia infections with special emphasis on the cornea. Surv Ophthalmol 2001;45:361–78. 7. Haripriya A, Lalitha P, Mathen M, et al. Nocardia endophthalmitis after cataract surgery: clinicomicrobiological study. Am J Ophthalmol 2005;139:837– 46. 8. Gopinathan U, Sharma S, Garg P, Rao GN. Review of epidemiological features, microbiological diagnosis and treatment outcome of microbial keratitis: experience of over a decade. Indian J Ophthalmol 2009;57:273–9. 9. Green M, Apel A, Stapleton F. Risk factors and causative organisms in microbial keratitis. Cornea 2008;27:22–7. 10. Norina TJ, Raihan S, Bakiah S, et al. Microbial keratitis: aetiological diagnosis and clinical features in patients admitted to Hospital Universiti Sains Malaysia. Singapore Med J 2008; 49:67–71. 11. Fong CF, Hu FR, Tseng CH, et al. Antibiotic susceptibility of bacterial isolates from bacterial keratitis cases in a university hospital in Taiwan. Am J Ophthalmol 2007;144:682–9. 12. Alexandrakis G, Alfonso EC, Miller D. Shifting trends in bacterial keratitis in south Florida and emerging resistance to fluoroquinolones. Ophthalmology 2000;107:1497–502. 13. Riono WP, Hidayat AA, Rao NA. Scleritis: a clinicopathologic study of 55 cases. Ophthalmology 1999;106:1328 –33. 14. Al-Mezaine HS, Al-Assiri A, Al-Rajhi AA. Incidence, clinical features, causative organisms, and visual outcomes of delayed-onset pseudophakic endophthalmitis. Eur J Ophthalmol 2009;19:804 –11. 15. Fan JC, Niederer RL, von Lany H, Polkinghorne PJ. Infectious endophthalmitis: clinical features, management and visual outcomes. Clin Experiment Ophthalmol 2008;36:631– 6. 16. Cao X, Liu A, Zhang J, et al. Clinical analysis of endophthalmitis after phacoemulsification. Can J Ophthalmol 2007;42:844 – 8. 17. Gupta A, Gupta V, Gupta A, et al. Spectrum and clinical profile of post cataract surgery endophthalmitis in north India. Indian J Ophthalmol 2003;51:139 – 45. 18. Sharma S, Kunimoto DY, Gopinathan U, et al. Evaluation of corneal scraping smear examination methods in the diagnosis of bacterial and fungal keratitis: a survey of eight years of laboratory experience. Cornea 2002;21:643–7. 19. Laurent FJ, Provost F, Boiron P. Rapid identification of clinically relevant Nocardia species to genus level by 16S rRNA gene PCR. J Clin Microbiol 1999;37:99 –102. 20. Reddy AK, Garg P, Kaur I. Speciation and susceptibility of Nocardia isolated from ocular infections. Clin Microbiol Infect 2010;16:1168 –71. 21. Glupczynski Y, Berhin C, Janssens M, Wauters G. Determination of antimicrobial susceptibility patterns of Nocardia spp. from clinical specimens by Etest. Clin Microbiol Infect 2006;12:905– 12. 22. Wallace RJ Jr, Septimus EJ, Musher DM, Martin RR. Disk diffusion susceptibility testing of Nocardia species. J Infect Dis 1977;135:568 –76.

1199

Ophthalmology Volume 118, Number 6, June 2011 23. Sharma S, Jalali S, Adiraju MV, et al. Sensitivity and predictability of vitreous cytology, biopsy, and membrane filter culture in endophthalmitis. Retina 1996;16:525–9. 24. Goel S, Kanta S. Prevalence of Nocardia species in the soil of Patiala area. Indian J Pathol Microbiol 1993;36:53– 60. 25. Randhawa HS, Khan ZU. Pulmonary and systemic nocardiosis. Indian J Chest Dis Allied Sci 1977;19:32– 44. 26. Silva CL, Tincani I, Brandao Filho SL, Faccioli LH. Mouse cachexia induced by trehalose dimycolate from Nocardia asteroides. J Gen Microbiol 1988;134:1629 –33. 27. Mikami Y, Yu SF, Yazawa K, et al. A toxic substance produced by Nocardia otitidiscaviarum isolated from cutaneous nocardiosis. Mycopathologia 1990;112:113– 8. 28. Beaman BL, Black CM, Doughty F, Beaman L. Role of superoxide dismutase and catalase as determinants of pathogenicity of Nocardia asteroides: importance in resistance to microbicidal activities of human polymorphonuclear neutrophils. Infect Immun 1985;47:135– 41. 29. Lalitha P, Tiwari M, Prajna NV, et al. Nocardia keratitis: species, drug sensitivities, and clinical correlation. Cornea 2007;26:255–9. 30. Bharathi MJ, Ramakrishnan R, Vasu S, et al. Nocardia asteroides keratitis in South India. Indian J Med Microbiol 2003;21:31– 6. 31. Sridhar MS, Sharma S, Garg P, Rao GN. Treatment and outcome of Nocardia keratitis. Cornea 2001;20:458 – 62. 32. Srinivasan M, Sundar K. Nocardial endophthalmitis [letter]. Arch Ophthalmol 1994;112:871–2. 33. Atkinson PL, Jackson H, Philpott-Howard J, et al. Exogenous Nocardia asteroides endophthalmitis following cataract surgery. J Infect 1993;26:305– 8. 34. Cacchillo PF, Ciulla TA, Strayer T, et al. Nocardia endophthalmitis following uncomplicated phacoemulsification and implantation of a posterior chamber intraocular lens. Ophthalmic Surg Lasers 2000;31:64 –5. 35. Chen CJ. Nocardia asteroides endophthalmitis. Ophthalmic Surg 1983;14:502–5. 36. Choudhry S, Rao SK, Biswas J, Madhavan HN. Necrotizing nocardial scleritis with intraocular extension: a case report. Cornea 2000;19:246 – 8. 37. Hudson JD, Danis RP, Chaluvadi U, Allen SD. Posttraumatic exogenous Nocardia endophthalmitis. Am J Ophthalmol 2003; 135:915–7. 38. King LP, Furlong WB, Gilbert WS, Levy C. Nocardia asteroides infection following scleral buckling. Ophthalmic Surg 1991;22:150 –2.

39. Papaventsis D, Siafakas N, Kondyli L, et al. Nocardia puris endophthalmitis [letter]. Indian J Med Microbiol 2009;27:168 – 70. 40. Seth RK, Gaudio PA. Nocardia asteroides necrotizing scleritis associated with subtenon triamcinolone acetonide injection. Ocul Immunol Inflamm 2008;16:139 – 40. 41. Zimmerman PL, Mamalis N, Alder JB, et al. Chronic Nocardia asteroides endophthalmitis after extracapsular cataract extraction. Arch Ophthalmol 1993;111:837– 40. 42. Ramos-Esteban JC, Servat JJ, Silva RS, et al. Necrotizing nocardial scleritis after combined penetrating keratoplasty and phacoemulsification with intraocular lens implantation: a case report and review of the literature. Arq Bras Oftalmol 2007;70: 355–9. 43. Sridhar MS, Cohen EJ, Rapuano CJ, et al. Nocardia asteroides sclerokeratitis in a contact lens wearer. CLAO J 2002;28: 66 – 8. 44. Jabs DA, Mudun A, Dunn JP, Marsh MJ. Episcleritis and scleritis: clinical features and treatment results. Am J Ophthalmol 2000;130:469 –76. 45. Hsiao CH, Chen JJ, Huang SC, et al. Intrascleral dissemination of infectious scleritis following pterygium excision. Br J Ophthalmol 1998;82:29 –34. 46. Huang FC, Huang SP, Tseng SH. Management of infectious scleritis after pterygium excision. Cornea 2000;19:34 –9. 47. Lin CP, Shih MH, Tsai MC. Clinical experiences of infectious scleral ulceration: a complication of pterygium operation. Br J Ophthalmol 1997;81:980 –3. 48. Yap EY, Fam HB, Leong KP, Buettner H. Nocardia choroidal abscess in a patient with systemic lupus erythematosus. Aust N Z J Ophthalmol 1998;26:337– 8. 49. Phillips WB, Shields CL, Shields JA, et al. Nocardia choroidal abscess. Br J Ophthalmol 1992;76:694 – 6. 50. Brown-Elliott BA, Ward SC, Crist CJ, et al. In vitro activities of linezolid against multiple Nocardia species. Antimicrob Agents Chemother 2001;45:1295–7. 51. Lorian V, Burns L. Predictive value of susceptibility tests for the outcome of antibacterial therapy. J Antimicrob Chemother 1990;25:175– 81. 52. Kaye SB, Neal T, Nicholson S, et al. Concentration and bioavailability of ciprofloxacin and teicoplanin in the cornea. Invest Ophthalmol Vis Sci 2009;50:3176 – 84. 53. Healy DP, Holland EJ, Nordlund ML, et al. Concentrations of levofloxacin, ofloxacin, and ciprofloxacin in human corneal stromal tissue and aqueous humor after topical administration. Cornea 2004;23:255– 63.

Footnotes and Financial Disclosures Originally received: June 30, 2010. Final revision: October 21, 2010. Accepted: October 25, 2010. Available online: January 26, 2011. 1

5

International Centre for Advancement of Rural Eye Care, Hyderabad Eye Research Foundation, L V Prasad Eye Institute, Hyderabad, India.

Manuscript no. 2010-902.

Duke University Eye Center; Durham, North Carolina.

2

Cornea and Anterior Segment Service, Hyderabad Eye Research Foundation, L V Prasad Eye Institute, Hyderabad, India.

3

Jhaveri Microbiology Centre, Hyderabad Eye Research Foundation, L V Prasad Eye Institute, Hyderabad, India.

4

Smt Kanuri Santhamma Centre for Vitreo-Retinal Diseases, Hyderabad Eye Research Foundation, L V Prasad Eye Institute, Hyderabad, India.

1200

6

Lotus Eye Care Hospital, Coimbatore, Tamil Nadu, India.

7

Vadodara, Gujarat, India.

*The Hyderabad Endophthalmitis Group: Drs Taraprasad Das, Subhadra Jalali, Savitri Sharma, Ajit Babu, and Avinash Pathengay. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Prashant Garg, MD, Cornea and Anterior Segment Service, L V Prasad Eye Institute - Kallam Anji Reddy Campus, L V Prasad Marg, Banjara Hills, Hyderabad - 500 034, India. E-mail: [email protected].