Abstract
Endophthalmitis in newborns and preverbal children constitutes only a small proportion of all endophthalmitis visiting a hospital. While most principles and practice of evaluation and managing endophthalmitis in this vulnerable group are similar to other cases of endophthalmitis described in this book, important considerations are highlighted in this chapter.
Access provided by CONRICYT-eBooks. Download chapter PDF
Similar content being viewed by others
Keywords
- Neonatal
- endophthalmitis
- pediatric
- endogenous
- umblical sepsis
- blood culture
- Systemic sepsis
- Neonatologist
- topical anaesthesia
- intraocular antibiotics
- candida
- Pseudomonas
Endophthalmitis in newborns and preverbal children constitutes only a small proportion of all endophthalmitis visiting a hospital. While most principles and practice of evaluation and managing endophthalmitis in this vulnerable group are similar to other cases of endophthalmitis described in this book, important considerations are highlighted in this chapter.
Neonatal and infantile endophthalmitis is mostly from an endogenous source of infection and is the main focus of this chapter. Rarely it could be following post-accidental and non-accidental trauma, post-intraocular surgery, or microbial keratitis especially after keratomalacia, gonococcal keratoconjunctivitis, or congenitally anesthetic cornea. Trauma remains the most important cause in older children followed by post-intraocular surgical infections that are dealt with elsewhere in this book and are mostly similar to adult cases.
Endogenous endophthalmitis in neonates is a rare clinical situation and constitutes 0.1–4% of all endogenous endophthalmitis cases [1, 2]. In the USA, the incidence of endophthalmitis reported from one neonatal intensive care unit (NICU) was 0.14% of all admitted babies [3]. Endogenous endophthalmitis is more common in East Asia and India with large series reported that included the neonates [2, 4].
Clinical Presentations and Common Organisms
Endophthalmitis in newborns presents as two distinct clinical presentations, and a third rare type mostly depending upon the virulence of the causative organisms and treatment given. A fulminant, acute inflammation (Fig. 12.1), which is mostly, though not always, due to virulent gram-negative organisms, is the commonest. The causative organisms include Pseudomonas, Klebsiella, E. coli, Acinetobacter, etc. [1,2,3,4]. The gram-positive organisms include methicillin-resistant Staphylococcus aureus (MRSA), Bacillus and Streptococcus, and rarely vancomycin-resistant Staphylococcus epidermidis (VRSE) [5]. Rare cases with fulminant picture may be of fungal etiology such as Fusarium (Fig. 12.2) from a ventriculoperitoneal (VP) shunt [2] or virulent Herpes simplex virus (HSV) infections [3].
The second clinical picture is of a very low-grade retinal infiltrate, diagnosed on routine retinal screening. This presents as a single or multiple yellow-white lesions varying from pinhead to few millimeter size, minimally elevated from the retinal surface or as a small subretinal abscess with indistinct borders. The most common causative organism is Candida albicans and Candida tropicalis (Fig. 12.3). Vitritis is minimal or absent to start with but gradually becomes manifested as the disease progresses. This characteristic clinical diagnosis is sufficient to start antifungals as the lesions are very classic and typical. Low-grade viral retinal infections from HSV and cytomegalovirus (CMV) also start as mild to moderate diffuse retinitis without vitritis (Fig. 12.4) and later can progress to severe vitritis and sometimes keratitis [6].
Rarely a third type of clinical picture, partially resolved bacterial endogenous endophthalmitis presents as a less severe infection, after having received parenteral antibiotics during septicemia. These present mostly as a low-grade partly organized subretinal abscess (Fig. 12.5) or vitritis or organized leucocoria (Fig. 12.6) resembling a pseudoglioma due to retinal necrosis and retinal detachment. Another rare clinical presentation is a localized anterior chamber granuloma or a very focal iris abscess (authors’ experience) that can respond to systemic antimicrobials (anterior endophthalmitis).
Risk factors: Prematurity with low birth weight and hospitalization has been associated in 80% babies presenting with neonatal endophthalmitis [1,2,3,4]. Rarely, healthy and normal-weight term babies without any apparent risk factor get such infections [2,3,4]. The common risk factors are listed in Table 12.1. In the USA, over the years, possibly due to implementation of strict asepsis protocols and attention to the risk factors, there has been a declining incidence of neonatal endogenous endophthalmitis; it reduced from 8.7 per 100,000 babies in 1986 to 4.42 per 100,000 babies in 2006 [7]. In a prospective case series of neonatal candidemia with a prevalence of 1.1 per 100 NICU discharges, endophthalmitis was seen in 2 of 20 babies where eye was evaluated [8].
Diagnostic criteria: Cultures may not be always positive. Hence only laboratory-based confirmation is not adequate to identify all infected endophthalmitis cases, especially of endogenous origin [1,2,3,4,5,6,7]. Diagnostic criteria are given in the Table 12.2.
Investigations
-
1.
All cases suspected to have neonatal and infantile endophthalmitis should have a careful ocular ultrasonography (USG-B) scan with corresponding A-Scan. Endophthalmitis eyes in early stages may have normal USG but in most cases will show low-intensity echoes. USG helps to confirm diagnosis and identifies poor prognostic indicators like retinal detachment, choroidal detachment, membranous echoes (Fig. 12.1, 12.6), and dense echoes filling most of the vitreous cavity. In all cases careful USG including immersion scan should also be done at low gain to evaluate all areas for any mass lesions or calcifications that would suggest an underlying retinoblastoma, which can masquerade as orbital cellulitis or endophthalmitis (Fig. 12.7). If possible, this is done preferably under general anesthesia as sometimes in a crying and struggling child one may miss scanning some areas that contain a hidden underlying tumor. Intraocular cysticercosis that can masquerade as endophthalmitis is also very well diagnosed by USG in most cases. USG is much more sensitive and specific than CT scan or MRI in the diagnosis of neonatal and pediatric endophthalmitis.
-
2.
Find source of infection: Detailed antenatal history must include episodes of fever, septicemia, or vaginal discharge or eruptions in the mother, the mode, and setting of delivery including rural or home deliveries and puerperal sepsis in mother, details of postnatal hospitalization, culture from tubes/ cannulas, etc. for any infective focus like septic arthritis, meningitis, liver abscess, pneumonia, blood cultures, or cultures from urine or other infected body fluids. Evaluation of umbilicus may reveal the source (Fig. 12.8). History should also record any septic epidemic in the nursery at the time of stay of the index case. Thorough systemic evaluation is essential. Past culture and neonatal course reports should be reviewed in consultation with the child healthcare givers.
-
3.
Detailed evaluation of fellow eye including dilated indirect ophthalmoscopy fundus examination at each visit, both in the clinic and whenever the opportunity arises, under general anesthesia additionally, is necessary (Fig. 12.5).
-
4.
Screen other twin/triplets nursery babies. Examination of mother including vaginal swabs, when indicated, helps.
-
5.
All babies with candidemia/bacteremia need dilated fundoscopy, especially if having prolonged infection for more than 2 weeks.
-
6.
All premature babies also need regular fundoscopy for associated ROP; this will need comanagement.
-
7.
Baby must be co-evaluated for other systemic problems that can be life threatening such as septicemia, meningitis, endocarditis, peritonitis, septic liver abscess, etc.
Differential diagnosis: Endogenous endophthalmitis can mimic many other ocular conditions in newborns and small babies. Each of these conditions need attention to clinical, laboratory, imaging, and ancillary testing to arrive to a reasonable diagnosis. Table 12.3 provides other differential diagnosis.
Management: In absence of any large randomized trials or large series of cases of this rare condition, the broad principles of endophthalmitis management are the same in neonates and small children as in adults. These include prompt clinical and microbiological diagnosis and initiation of empirical intravitreal antimicrobials followed by specific modification of antimicrobials based on clinical response and laboratory diagnostic reports. There are however numerous challenges specific to neonatal and infantile endophthalmitis cases (Table 12.4).
Modifications are needed from adult protocols to address some of these challenges. For example, babies who are sick and not fit for general anesthesia could undergo aqueous/vitreous tap under topical anesthesia with intravitreal antibiotics (and steroids, if so considered) in cases presenting with fulminant endophthalmitis (Fig. 12.2). This provides adequate identification of infecting agent and appropriate antibiotic selection; it also provides control of infection that helps to avoid evisceration even in eyes presenting as panophthalmitis. Since most cases are nosocomial and multidrug resistant, microbiological work-up becomes essential. In fungal retinal infiltrate, only intravitreal amphotericin B is given (half adult dose), and no sample is taken as the anterior vitreous or aqueous will not show any results in this scenario, and treatment is based on the typical clinical picture. If child is fit for general anesthesia, then depending on clinical severity, lens-sparing vitrectomy or vitrectomy with lensectomy is done with sclerotomy at a distance of 1.0 mm from limbus in neonates and infants due to underdevelopment of pars plana in these eyes. In one of the largest published series on neonatal endophthalmitis, all eyes with Candida infection could be salvaged by intravitreal amphotericin B under topical anesthesia followed promptly by lens-sparing vitrectomy (Fig. 12.3, 12.5) as soon as vitritis developed [2].
Unlike adults, in neonates, vitreous involvement can lead to rapid folding, stretching, and anterior elevation of the retina (Figs. 12.3 and 12.5). Early surgical intervention can help to flatten these folds before they get elevated far enough to touch the lens, which would necessitate lensectomy. The principles are similar to vitrectomy for ROP-related stage 4A detachments, that progress to require lensectomy if not operated urgently. Once lensectomy is needed, the prognosis becomes poor not only due to challenge of managing a unilateral aphakia and amblyopia at this young age but also due to the high risk of secondary glaucoma [4]. Hence early surgery appears more favorable [2].
Results: Most gram-negative fulminant bacterial cases are reported to resolve with phthisis bulbi or need evisceration. Evisceration can be avoided by managing with intraocular antibiotics and steroids in the eyes that present with fulminant infection [2]. Occasional good outcomes are reported in few cases of fulminant infection, especially those who are diagnosed and managed by immediate surgery. This requires vigilance, high degree of suspicion by treating pediatrician, and a whole lot of coordination to get baby rapidly fit and taken up for surgery [5]. We reported a large series of 31 eyes of 26 babies of neonatal endophthalmitis; in this series all the eyes with suspected Candida could be salvaged with good visual outcomes while all the bacterial fulminant eyes became phthisical, but none progressed to evisceration [2]. High mortality from septicemia or meningitis has been reported in some series, and this would depend on the causative organism and clinical situation [8,9,10]. In a study on long-term outcomes of neonatal Candida endophthalmitis treated by systemic therapy alone (intravenous amphotericin B/oral fluconazole), 7 of 11 eyes achieved good outcomes. All three eyes that had poor outcome were due to vitreous traction and macular lesions that did not undergo prompt surgery [11]. In some cases, fungal infections can present as a lens abscess that is possibly a result from hematogenous spread through the persistent tunica vasculosa lentis in premature babies [12]. These eyes have poor outcomes due to inability of systemic drugs to reach the poorly vascularized lens substance as the hyaloid system regresses, leaving a nidus of infection within the substance of the lens [12]. Prompt lensectomy and intravitreal antifungals would be needed in such cases [12].
Few cases of coexisting ROP in the setting of intraocular infection have been reported. There may be more progression of ROP due to inflammatory cytokines and angiogenic factors [12]. Treatment would also be challenging in view of media haze. Successful management of such cases by surgery or laser has been reported.
Guidelines for surveillance: All babies who are in hospital and diagnosed as having candidemia/bacteremia need regular fundoscopy to detect any ocular spread. This could be weekly in case of candidemia [12] and daily in case of gram-negative septicemia. Any inflammation around eyes (lid edema/erythema, conjunctival injection, ocular discharge) requires pupil dilatation and fundoscopy. The widespread practice of using empirical antibiotic eye drops with a provisional suspicion of a “simple mucopurulent conjunctivitis” without a “red glow” could lead to a delayed diagnosis and loss of a salvageable eye.
References
Okada AA, Johnson RP, Liles WC, et al. Endogenous bacterial endophthalmitis. Report of a ten-year retrospective study. Ophthalmology. 1994;101:832–8.
Jalali S, Pehere N, Rani PK, et al. Treatment outcomes and clinico-microbiological characteristics of a protocol-based approach for neonatal endogenous endophthalmitis. Eur J Ophthalmol. 2014;24:424–36.
Aziz HA, Berrocal AM, Sisk RA, et al. Intraocular infections in the neonatal intensive care unit. Clin Ophthalmol. 2012;6:733–7.
Wong JS, Chan TK, Lee HM, Chee SP. Endogenous bacterial endophthalmitis: an East Asian experience and a re-appraisal of a severe ocular affliction. Ophthalmology. 2000;107:1483–91.
Relhan N, Albini T, Pathengay A, Flynn H Jr. Bilateral endogenous endophthalmitis caused by vancomycin resistant Staphylococcus epidermidis in a neonate. J Ophthalmic Inflamm Infect. 2015;5:11. https://doi.org/10.1186/s12348-015-0039.
Gupta A, Rani PK, Bagga B, et al. Bilateral Herpes Simplex-2 acute retinal necrosis with encephalitis in premature twins. JAAPOS. 2010;14:541–3.
Moshfeghi AA, Charalel RA, Hernandez-Boussard T, et al. Declining incidence of neonatal endophthalmitis in the United States. Am J Ophthalmol. 2011;151:59–65.
Rodriguez D, Almirante B, Park BJ, et al. Candidemia in neonatal intensive care unit: Barcelona, Spain. Barcelona Candidemia Project Study Group. Pediatr Infect Dis J. 2006;25:224–9.
Chen JY. Neonatal candidiasis associated meningitis and endophthalmitis. Acta Paediatr Jpn. 1994;36:261–5.
Basu S, Kumar A, Kapoor K, et al. Neonatal endogenous endophthalmitis: report of six cases. Paediatrics. 2013;131 https://doi.org/10.1542/peds.2011-3391.
Annable WL, Kachmer ML, Di Marco M, De Santis D. Long-term follow-up of Candida endophthalmitis in the premature Infant. J Pediatr Ophthalmol Strabismus. 1990;27:103–6.
Baley JE, Ellis FJ. Neonatal candidiasis: ophthalmological infections. Semin Perinatol. 2003;27:401–5.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Frequently Asked Questions
Frequently Asked Questions
-
1.
Should vitrectomy/intravitreal antimicrobials be done early or late in neonatal endophthalmitis when retina is visualized?
A: We believe that under topical anesthesia, neonates can receive intravitreal antimicrobials, just like adults, and need not wait for systemic status to improve for general anesthesia fitness. This allows lens-sparing vitrectomy/intravitreal approach early. Delaying treatment could necessitate lensectomy that could progress to unilateral amblyopia and glaucoma.
-
2.
When, how often and who should do fundoscopy in babies diagnosed to having bacteremia/candidemia in the nursery.
A: All babies with systemic infection should undergo mandatory dilated indirect ophthalmoscopy at least weekly in active phases of infection and promptly in case of any redness or sign of ocular inflammation during follow-up.
-
3.
When, how often and who should do fundoscopy in the nursery in babies who get red eye and /or mucopurulent discharge from the eyes?
A: All such babies need a red glow fundus test by physician after dilating pupils with 1% tropicamide on daily rounds. Any suspicion should prompt urgent evaluation by a competent ophthalmologist using indirect ophthalmoscopy.
Rights and permissions
Copyright information
© 2018 Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Jalali, S. (2018). Endophthalmitis in Newborns and Preverbal Children. In: Das, T. (eds) Endophthalmitis . Springer, Singapore. https://doi.org/10.1007/978-981-10-5260-6_12
Download citation
DOI: https://doi.org/10.1007/978-981-10-5260-6_12
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-10-5259-0
Online ISBN: 978-981-10-5260-6
eBook Packages: MedicineMedicine (R0)