Abstract
Background
We describe 12 cases of chlamydial ophthalmia neonatorum and the current scientific evidence on its prevention and treatment. The data presented were obtained from the “Maduo” study, a prospective observational study of the relationship between curable sexually transmitted infections and adverse neonatal outcomes at four antenatal clinics in Gaborone, Botswana.
Methods
Infants of mothers with perinatal chlamydia infection were evaluated for chlamydial ophthalmia neonatorum based on clinical presentation of conjunctivitis or positive test via GeneXpert CT/NG assay. Data on 29 infants born to mothers with postnatal C. trachomatis infection were analysed.
Results
12 infants were diagnosed with chlamydial ophthalmia neonatorum. Eight of those cases were confirmed with the GeneXpert CT/NG assay while four were identified as probable cases based on clinical history and presentation. Overall, nine infants presented with signs of conjunctivitis, while three who had a positive diagnostic test result had asymptomatic infection. All but one infant had received ocular 1% tetracycline prophylaxis at birth, and four infants had signs suggestive of chlamydial pneumonia at presentation. Two out of five symptomatic cases whose mothers reported completion of their treatment course with erythromycin had lingering symptoms.
Conclusions
Our findings affirm that the current prophylaxis and treatment modalities for chlamydial ophthalmia neonatorum are inadequate. To the extent feasible in low- and middle-income countries, we recommend implementation of routine C. trachomatis screening and treatment in pregnant women.
Keywords
Introduction
Ophthalmia neonatorum is acute conjunctivitis occurring usually within the first month of life. 1 It carries significant risk of corneal ulceration, corneal perforation, and blindness. Ophthalmia neonatorum is an ongoing public health problem in countries or settings where prevention strategies are inadequate.2–6
The etiology of ophthalmia neonatorum is most commonly due to sexually transmitted bacteria such as Neisseria (N.) gonorrhoeae and Chlamydia (C.) trachomatis but could also arise from non-sexually transmitted bacteria or less often from viral and chemical sources. 7
For neonatal conjunctivitis due to sexually transmitted bacteria, the presence of perinatal maternal infection is the predisposing factor. 2 Consequently, as C. trachomatis infection is prevalent globally, 8 chlamydial conjunctivitis in the newborn is not uncommon.9,10 As high as one in two infants born to mothers with untreated C. trachomatis infection may develop chlamydial conjunctivitis.1,11,12
Chlamydial ophthalmia neonatorum may cause visual impairment from conjunctival scarring, vascularization and pseudo membrane formation.9,13,14 Hemorrhagic conjunctivitis has also been reported. 15
The goal of this report is to describe a series of cases of chlamydial neonatal conjunctivitis and review the evidence on its prevention and treatment.
Methods
The cases were obtained from the “Maduo” study, a prospective observational study conducted in four antenatal clinics in Gaborone, Botswana, to characterize the antenatal burden of sexually transmitted infections (STI) and their association with birth outcomes. 16
Briefly, pregnant women (N = 500) who had not been managed for an STI within the preceding 30 days were recruited at their first antenatal visit between February 2021 and April 2022. Women recruited at screening sites were tested for C. trachomatis and N. gonorrhoeae on a self-collected vaginal swab using the GeneXpert® Chlamydia trachomatis/Neisseria gonorrhoeae (CT/NG) assay (Cepheid, Sunnyvale, CA) at their initial visit and repeat screening in their third trimester. Women recruited at non-screening sites were provided routine antenatal standard-of-care which does not include routine STI screening. Participants were followed throughout pregnancy and asked to return for a final postnatal visit, scheduled 6–8 weeks after delivery to coincide with the infant’s first routine immunization visit. However, because women in Botswana often practice “confinement” during the postpartum period, some mother-infant pairs who presented for their postnatal visit after 8 weeks of delivery were still evaluated.
Postnatal visit procedures, evaluation, and follow-up
Of 500 recruited women, relevant birth history was retrieved from the medical records of 448 women with a live born infant who had presented for the postnatal visit by October 2022. Information was also obtained from mothers about the health of their infant since birth, including likely symptoms of perinatal infection or any therapeutic interventions preceding that postnatal visit.
At their postnatal visit, vaginal swabs were obtained from all mothers and tested for N. gonorrhoea and C. trachomatis using the Gene Xpert platform. If a mother tested positive, clinician-collected ocular and nasopharyngeal swabs were subsequently obtained from her infant and tested. Specimens were not obtained from infants whose mothers did not have a positive postnatal test result. In addition, all infants born to a mother with a positive C. trachomatis or N. gonorrhoea test were clinically evaluated at this visit for signs and symptoms of conjunctivitis and pneumonia, such as eye redness, discharge or eyelid swelling, and cough, fever or increased respiratory rate.
Treatment was provided on the same day of results according to the United States Centers for Disease Control and Prevention (CDC) guidelines. 17 Mothers were treated with oral azithromycin 1 G for C. trachomatis, and 500 mg intramuscular ceftriaxone injection for N. gonorrhoeae. Infants were treated with oral erythromycin (liquid formulation) 50 mg/kg/day for 14 days in four divided doses per day for C. trachomatis, and intramuscular ceftriaxone single-dose 25–50 mg/kg for N. gonorrhoeae.
Follow-up phone calls were placed to mothers a month after their infants received treatment to ascertain resolution of their symptoms, treatment side effects, and overall infant well-being.
Case identification of chlamydial ophthalmia neonatorum
We defined an “exposed infant” as an infant born to a mother who tested positive for C. trachomatis at her postnatal visit.
We defined a “confirmed case” of chlamydial ophthalmia neonatorum as a C. trachomatis-exposed infant (1) with signs of conjunctivitis and a chlamydia-positive ocular or nasopharyngeal swab result, (2) without signs of conjunctivitis (asymptomatic) but a positive ocular swab result.
We defined a “probable case” as a C. trachomatis-exposed infant with clinical signs of conjunctivitis but without a positive swab result.
Ethical approval for the study was provided by the Botswana Ministry of Health and Wellness Health Research Development Committee (HRDC number: 00881) and the Institutional Review Board of University of Southern California (IRB number: HS-21-00245).
Results
The 12 cases of chlamydial ophthalmia neonatorum, gaborone Botswana, identified by postnatal visit (October 2022).
aSVD: spontaneous vaginal delivery.
bConfirmed case: GeneXpert® C. trachomatis-positive ocular and/or nasopharyngeal swab in an infant who has eye signs OR a positive ocular swab result in an infant with asymptomatic infection.
cProbable case: High clinical suspicion based on presence of eye signs at clinical evaluation, with or without history of prior management for chlamydia conjunctivitis.
dCase management: Mothers of infants with probable infection were encouraged to seek care at a local facility. Botswana syndromic care guidelines currently recommend erythromycin (50 mg/kg/day for 14 days) for suspected chlamydial conjunctivitis in the newborn.
Perinatal Characteristics for the Infants born to Mothers with C. trachomatis infection by their Postnatal visit, Gaborone Botswana, October 2022.
aSVD: spontaneous vaginal delivery.
Evidence review
History and Etiology of Ophthalmia Neonatorum
In the 1800s, ophthalmia neonatorum referred to conjunctivitis from N. gonorrhoeae and was the leading cause of blindness globally.18,19 The definition subsequently expanded to include ophthalmia from other bacteria, such as C. trachomatis.2,10
Following the discovery by Dr. Credé in 1881, universal ocular prophylaxis reduced the incidence of neonatal gonococcal conjunctivitis from 10% to <1%.2,10,20–22 One challenge, however, with silver nitrate prophylaxis was the frequent occurrence of chemical conjunctivitis.23,24 This toxicity, together with a need for efficacious prophylaxis against chlamydial ophthalmia neonatorum, has expanded prophylaxis to include other antimicrobial agents such as erythromycin, tetracycline, and povidone iodine.23,25
In many high-income countries like the United States, implementation of routine STI screening and treatment during pregnancy has dramatically reduced the incidence of ophthalmia neonatorum.10,26 As a result, many recommendations have been updated from universal ocular prophylaxis to targeted prophylaxis in infants born to mothers with increased risk of infection.7,27–29
Epidemiology
High income countries
Cases of ophthalmia neonatorum in the United States are now uncommon due to routine STI screening. 30 Similar practices have also made STI-associated ophthalmia neonatorum a rare occurrence in Canada and in Western Europe.31,32
Sub-Saharan Africa
In many sub-Saharan African regions, N. gonorrhoeae and C. trachomatis, the major pathogens implicated in ocular morbidity, are highly prevalent in pregnant women with frequencies of either infection as high as 31%.7,9,33 Studies in Cameroon, Gambia, Gabon, and Kenya suggest that C. trachomatis alone may account for up to a third of cases of ophthalmia neonatorum in sub-Saharan Africa. 9
In Botswana, there are approximately 89,000 pregnant women annually. 34 The few published antenatal STI prevalence studies suggest a baseline antenatal C. trachomatis prevalence ranging from 8% 16,35 to 23%.17,36 There are no published epidemiological estimates for ophthalmia neonatorum.
Diagnosis and evaluation
The confirmation of a suspected case of chlamydial ophthalmia neonatorum can be performed using tissue culture methods or non-culture methods, but culture is challenging due to C. trachomatis being an obligate intracellular pathogen. Direct fluorescent antibody staining is the only non-culture method currently approved by the Food and Drug Administration (FDA) in the United States. However, nucleic acid amplification tests (NAATs) which are less labor-intensive are now widely used in many clinical practice and research settings.11,37 To assure diagnostic accuracy, ocular specimens obtained from the everted eyelid should contain conjunctival cells and not simply eye exudate. Thus, best practice is to obtain additional specimens from the nasopharynx. 1
Treatment
There is still inconclusive evidence as to the most effective treatment of chlamydial ophthalmia neonatorum if prophylaxis fails. 38 The US CDC recommends erythromycin 50 mg/kg body weight/day orally, divided into 4 doses daily for 14 days, 26 while the current WHO guidelines (2016) prefer 20 mg/kg oral azithromycin (once daily for 3 days) 39 over erythromycin due to the increased risk of pyloric stenosis reported with erythromycin. However, risk of pyloric stenosis with azithromycin exposure has also been reported. 40 The efficacy for erythromycin is more characterized and has been estimated to be approximately 80%.26,41 With low-certainty, Zikic et al. 38 report in their meta-analysis that the efficacy could be as high as 96% for erythromycin and 86% for azithromycin. Since etiologic STI screening is not routine, Botswana guidelines do not provide treatment recommendations for individual infections. However, they currently recommend erythromycin for the syndromic management of suspected chlamydial ophthalmia neonatorum. 42
Prevention
The World Health Organization (WHO) currently provides guidelines regarding the prevention of ophthalmia neonatorum. The WHO recommends universal prophylaxis with 1% tetracycline hydrochloride or 0.5% erythromycin or 2.5% povidone iodine or 1% silver nitrate or 1% chloramphenicol administered into both eyes of all newborns immediately after birth for prevention of both gonococcal and chlamydial conjunctivitis. 39
However, while those agents have been shown to be effective for preventing ophthalmia neonatorum from N. gonorrhoeae, there is no antimicrobial agent that has been shown to reliably prevent ophthalmia neonatorum resulting from C. trachomatis.12,43
In a recently published Cochrane review to determine optimal prophylactic regimens, 44 the authors summarized evidence from 30 trials, only 4 of which were carried out in African countries (Angola, Kenya, and Zaire). Using blindness (or other adverse visual outcomes) or conjunctivitis at 1 month, the authors reported that compared to no prophylaxis, any of tetracycline, erythromycin, povidone iodine, or silver nitrate had very limited effect (0.96, 95% CI 0.57–1.61) on the risk of chlamydial ophthalmia neonatorum. That effect was still minimal for tetracycline only (RR 0.82, 95% CI 0.42–1.63). None of these findings were statistically significant and the authors judged the overall evidence as insufficient and of low certainty. 44 Smith-Norowitz et al., 45 from a similarly themed systematic review, analyzed evidence from 3 eligible studies, all carried out in the United States, and concluded that neither erythromycin nor tetracycline were efficacious in preventing chlamydial ophthalmia neonatorum.
Discussion
The cases and evidence-review presented highlight the clinical and public health significance of chlamydial ophthalmia neonatorum, especially for lower income counties.
In a longitudinal study of sexually transmitted infections in pregnant women and their neonates, we describe 12 clinical cases of chlamydial ophthalmia neonatorum. Eight cases were confirmed with a positive Cepheid GeneXpert® NAAT test and treated with erythromycin as per study protocol, while four were diagnosed as probable cases based on history and high clinical suspicion and their mothers were encouraged by the study clinician to visit their local health facilities if eye symptoms persisted.
Including both confirmed and probable cases of infection, the diagnosis of chlamydia infection occurred in 41% (12/29) of exposed infants in our cohort. That value is similar to the 43% reported by Peters et al., 46 wherein authors measured transmission rates for C. trachomatis in infants in South Africa. However, unlike that referenced study where all reported infections were asymptomatic, most cases in our study (75%, 9/12) presented with symptoms. Thus, this report provides novel data which suggests deficiencies in current prophylactic and therapeutic options for chlamydial conjunctivitis in the newborn.
Of 24 infants that received tetracycline prophylaxis, 11 tested positive for chlamydial ophthalmia neonatorum, including both confirmed and probable cases of infection. The cases reported in our study reinforce that tetracycline prophylaxis is not effective at preventing chlamydial ophthalmia neonatorum, and helps clarify findings from a recent Cochrane review 44 which had suggested, with low certainty, some efficacy for tetracycline. All but one infant had received the standard ocular tetracycline prophylaxis at birth. In another review by Zuppa et al., 23 authors suggested that two other therapies, povidone-iodine 2.5% and fusidic acid, may be effective at preventing both chlamydial and gonococcal ophthalmia neonatorum. However, existing data on those two agents are limited and further investigation into their efficacy and risk of secondary chemical conjunctivitis is necessary before widespread implementation. This paucity of evidence leaves healthcare providers with few pharmaceutical options in the prevention of chlamydial ophthalmia neonatorum.
Finally, our study still leaves the question of the appropriate treatment for ophthalmia neonatorum unanswered. Only three out of the five symptomatic cases who reported completing their treatment course reported complete resolution of symptoms. We interpret these limited data with caution, as they suggest an efficacy of 60% (3/5), which is considerably lower than the expected 80%. 26 Notably, four infants had presented to another health facility with conjunctivitis or pneumonia prior to their study visit. Each of those four infants presented at their study visit with either a positive test result (for three confirmed cases) or purulent eye discharge (for one probable case) despite having received prior care at another facility. One of those infants was prescribed erythromycin at their visit to another healthcare facility prior to their study visit, but we could not reliably ascertain if the erythromycin treatment course was completed prior to the study visit. We were not able to ascertain what treatment was provided to the other three infants who visited another healthcare facility prior to their study visit. As noted earlier, the efficacy of erythromycin for the treatment of chlamydial ophthalmia neonatorum is relatively well-characterized compared to azithromycin 26 which is currently preferred in the WHO recommendations. 39 However, the existing data comparing erythromycin and azithromycin for treatment of chlamydial ophthalmia neonatorum is limited and the risk of developing pyloric stenosis has been reported for both medications.38,40
Conclusion and Relevance
Chlamydial ophthalmia neonatorum is a serious condition that may progress to permanent vision loss if left untreated. The success of C. trachomatis screening programs in minimizing the incidence of chlamydial ophthalmia neonatorum in the United States and other high-income countries has resulted in a relative lack of investigation into antibiotic prophylaxis and treatment of this condition that are suitable for other healthcare settings. Existing evidence and the cases presented in our study underscore those deficits in our current understanding of how to better prevent and treat ophthalmia neonatorum. Given the success of routine antenatal screening in high income countries at reducing the incidence of ophthalmia neonatorum, a similar approach which is contextually adapted and optimized to the resources available may be of benefit in low- and middle-income countries.
Footnotes
Acknowledgements
We thank Cepheid for loaning the GeneXpert system and donating diagnostic test kits for this study.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) [R21 HD100821-01].
Data Availability Statement
Partial data for all participants (29 infants with C. trachomatis exposure) is already available with publication. Full deidentified data will be made available upon request.
