Molecular and Clinical Characterization of Human Adenovirus E4–Associated Conjunctivitis


To determine the characteristics of conjunctivitis associated with human adenovirus E4 (AdV E4).


Samples and outcomes from 500 patients with conjunctivitis were obtained from the NVC-422 randomized controlled clinical trial comparing auriclosene to placebo. Molecular typing identified 36 cases associated with AdV E4. Signs and symptoms at presentation and at the day 18 endpoint were compared with the larger cohort of 262 subjects with conjunctivitis caused by due to AdV D8. Full viral genomes of 22 AdV E4 isolates were reconstructed.


AdV E4 was the most frequently identified adenoviral type in conjunctivitis cases from the United States. Signs and symptoms at presentation were comparable to those associated with AdV D8. Viral load at presentation was comparable between groups but resolution was more rapid in the AdV E4 group. Clinical signs were fully resolved by day 18 in 26 of 36 (72%) patients with AdV E4. Subepithelial infiltrates developed in 12 of 36 (33%) patients with AdV E4 compared with 98 of 215 (45%) patients with AdV D8 ( P = .0001). One hundred twenty-four polymorphisms were observed among 22 whole viral genome sequences, which clustered into 3 clades. Patients in each clade developed subepithelial infiltrates. Neither single nucleotide polymorphism analysis nor machine learning approaches identified specific sequence features predictive of presenting signs or outcome.


AdV E4 conjunctivitis may be indistinguishable at presentation from AdV D8–associated disease. Resolution of viral load for AdV E4 appears more rapid than for AdV D8, and the risk for subepithelial infiltrates appears lower. Multiple substrains of AdV E4 are in circulation but all appeared equivalently pathogenic for conjunctivitis. NOTE: Publication of this article is sponsored by the American Ophthalmological Society.

C onjunctivitis is among the more common conditions in medicine, by some estimates accounting for between 1% and 2% of all visits to primary care providers in the United States. More than 4 million visits were made to U.S. physicians for conjunctivitis in 2005, and in excess of $430 million annually is attributed to the diagnosis and treatment of conjunctivitis. Conjunctivitis may be caused by bacteria, allergies, viruses, or parasites. Viral causes are the most common. The most severe form of viral keratoconjunctivitis is epidemic keratoconjunctivitis (EKC), which is typified by severe follicular conjunctivitis, preauricular lymphadenopathy, corneal involvement, and the development of subepithelial infiltrates. Because of its highly infectious nature, EKC-associated outbreaks can cause substantial morbidity. , The effect of EKC outbreaks in health care or community settings can last weeks to months. For example, from 2008 to 2010, 6 isolated EKC outbreaks were reported in 4 states: Florida, Illinois, Minnesota, and New Jersey. These outbreaks included >400 infected patients and forced a temporary closure of a neonatal intensive care unit and several clinics. In the military, EKC outbreaks can be catastrophic. An outbreak at a U.S. military base in the Philippines in 1989 resulted in 2603 cases and the loss of 9038 work days, or almost 36 work years in addition to incapacitating the base. Within the U.S. health care system, provider contraction of adenoviral conjunctivitis remains prevalent and is a significant threat to patient welfare. A recent study of the Johns Hopkins Medicine medical system demonstrated that, between 2011 and 2016, of 10,000 full-time employees, >1000 visited the occupational health clinic for “pink eye,” and 10% of these had polymerase chain reaction (PCR)–confirmed viral infection.

Viral conjunctivitis, including EKC, is most commonly caused by adenovirus (AdV) infection. AdVs are ∼90-nm nonenveloped viruses with an icosahedral nucleocapsid containing a ∼35-kb double-stranded DNA genome. AdVs are classically classified into 5 major species (A-E) based on serologic responses, and >60 types within those species. EKC has historically been associated with AdV types D8, D19 (recently reclassified as D64 9 ), and D37. Less frequently, types including E4, D53, D54, and D56 have also been reported in EKC. , Recently, the results of a large clinical trial involving 500 subjects on 3 continents testing the efficacy of auricloscene (NovaBay NVC-422) in the treatment of adenoviral conjunctivitis have been published. While this study confirmed that most cases of adenoviral conjunctivitis were caused by type D8, 9.4% of PCR-positive cases overall were attributed to type E4 (the sole type of E species). There was a strong geographic predilection for conjunctivitis associated with AdV E4, with this type being the most commonly encountered virus in the United States (accounting for 19/63 [30%] of PCR-positive cases ) and only rarely encountered in other countries (accounting for 17/327 [5.2%] of PCR-positive cases). Historically, AdV E4 has been primarily associated with pharyngoconjunctival fever and respiratory illness rather than EKC.

To better understand the clinical presentation of conjunctivitis associated with AdV E4, and to characterize the molecular genetics of this virus in cases of conjunctivitis, we analyzed outcomes from AdV E4–associated cases. In addition, we sequenced and reconstructed full viral genomes from 22 AdV E4 conjunctivitis cases from the BayNovation NVC-422 study and analyzed the influence of genetic variation on presenting signs, symptoms, and clinical outcome.



This prospective study was conducted with Goodwin University Institutional Review Board approval (Cincinnati, Ohio, USA). In addition, local approval at all sites was obtained (including Drug Controller General in India, Scientific and Ethical Review Board, Faculty of Medicine, University of Kelaniya, Colombo, Sri Lanka, and National Ethics Committee in Research review in Brazil). All research was conducted in accord with to the tenets of the Declaration of Helsinki, applicable privacy laws, and ARVO guidelines for vision research. The clinical trial was registered at (NCT01877694).

Participants were recruited from 58 ophthalmology practices in Brazil, India, Sri Lanka, and the United States. Written informed consent was obtained from all participants. Outside the United States, patients were ≥18 years of age, while in the United States patients could be as young as 1 year old. Inclusion criteria included: 1) a recent history of upper respiratory infection, contact with infected person, or recent visit to an eye care provider; 2) ≥2 of 9 clinical signs suggestive of conjunctivitis with onset <3 days before enrollment; and 3) positive point-of-service antigen AdV screening test (Adeno Plus; Rapid Pathogen Screening Inc, Saratosa, Florida, USA). Patients were randomized 1:1 to receive the study drug (auriclosene, NVC-422) or vehicle using 1 drop 8 times a day in both eyes for 10 days.

Investigators and subjects were masked to treatment assignment. Return visits were scheduled on days 3, 6, 11, and 18, with an optional follow-up visit on day 42. For the current study, day 18 was considered final. Copan swabs (Copan Diagnostics, Murrieta, California, USA) were collected from the lower conjunctival fornix of each eye at each visit. Ophthalmologic examinations were conducted at each visit. Visual acuity (VA) was measured with Early Treatment Diabetic Retinopathy Study (ETDRS) charts. The following clinical signs were assessed at each visit: lymphadenopathy, lid erythema, lid edema, conjunctival hyperemia, conjunctival discharge/lid crusting, tear meniscus, corneal fluorescein staining, tear break-up time, and the presence of subepithelial infiltrates. Six subjective ocular symptoms were assessed on every visit: blurred vision, itching, burning, tearing, foreign body sensation, and photophobia. All symptoms were scored as either present or absent, as were lymphadenopathy, normal/abnormal tear break-up time, and subepithelial infiltrates. The remainder of signs were scored as absent, mild, moderate, or severe using criteria supplied by the sponsor.


Polymerase chain reaction typing was performed by Viracor Eurofins Clinical Diagnostics (Lee’s Summit, Missouri, USA). Quantitative PCR was used to determine adenoviral load (normalized to copies/ml). Genomic DNA was purified using NucliSENS easyMag (bioMerieux, Inc, Durham, North Carolina, USA). TaqMan Fast Advanced PCR Master Mix (ThermoFisher Scientific, Waltham, Massachusetts, USA) containing 10 × Buffer, Taq polymerase, dNTP mix and the primers were used for the PCR reaction. For each reaction, 10 μL of genomic DNA was used. Primer/probe selection programs using the ABI Primer express. All primers targeted the 5′ end AdV hexon gene. The following primers (5′-TACATGCATCGCCGG-3′, 5′-TACATGCATATCGCCGG-3′, 5′-GGTCTTACAT-GCACATCTCGGG-3′, 5′-ACCCYGATGTGACCACSG-3′, 5′-TGCGTTTTGTCCCGTG-3′, 5′-TGCGTTTTGTGCCCGTG-3′, 5′-CCACCGATACCTACTTCAGCCT-3′, CACCGAGACGTACTTCAGCCT-3′) and the following probes were used: 5′-CCGGGT-CTGGTGCAGT-3′, 5′-CAGGACGCCTCGGAGTA-3′, 5′-CCACGGACACCTACTTCACCCTGGG-3′. Standards were generated from plasmid clones containing the target region of the hexon gene. Quantitative PCR was performed on an Applied Biosystems 7500 FAST instrument (ThermoFisher Scientific) using the single well method and stored standard curve. Cycling conditions were 2 minutes at 50 C, 2 cycles of 20 seconds of denaturation at 95 C, followed by 40 cycles of 3-second denaturation at 95 C and 30 second annealing at 60 C.

Specific viral species and types were determined by sequencing of 400 to 500 base pairs (bp) of the AdV hexon gene. The following primers were used: 5′-CTGATGTACTACAACAGCACTGGCAACATGGG-3′ and 5′-GCGTTGCGGTGGTGGTTAAATGGGTTTACGTTGTCCAT-3′. Cycling conditions for PCR were 2 minutes of denaturation at 95 C, 40 cycles of 30 seconds at 95 C, 50 seconds at 51 C, and 60 seconds of 72 C, followed by 3 minutes of 72 C.

Shotgun whole genome sequencing was performed in AdV E4–positive samples using the Illumina Nextera XT DNA library preparation kit (Illumina, San Diego, California, USA) using the manufacturer’s recommended procedure. One nanogram of genomic DNA was fragmented by the Nextera NT transposome at 55 C for 5 minutes and then held at 10 C. PCR amplification was performed using Nextera PCR Master Mix (NPM), index primers, and processed DNA. Cycling conditions were 3 minutes at 72 C, 30 seconds of denaturation at 95 C, followed by 12 cycles of 10-second denaturation at 95 C, 30-second annealing at 55 C, 30-second extension at 72 C, and the final step of 5 minutes of extension at 72 C. After PCR cleanup, the library was normalized. Approximately 2 ml to 4 ml of normalized library was pooled and sequenced on the MiSeq platform (Illumina). Sequencing was analyzed with the Scalable Metagenomics Alignment Research Tool. SPAdes was used for reconstruction of viral genomes. Variant calling was performed using Annovar.


Analysis was performed on a per-eye basis. The data from the worse-seeing eye on day 1 were included in the analysis of participants with bilateral involvement. If both eyes had equal VA, then the study eye was chosen at random using a random number generator. The Fisher exact or χ 2 tests were used for comparison of categorical variables. Parametric and nonparametic methods were used for the comparison of features between viral types with multiple comparator statistics used where indicated. All analyses were performed with XLSTAT with R software (R.



Overall demographics of patients in the NVC-422 clinical trial have been presented previously. In the current study, we focused on those patients with PCR evidence for AdV E4 compared with those with AdV D8 ( Table 1 ). A total of 36 subjects were PCR-positive for Adv E4 compared with 262 who tested positive for AdV D8. The mean age was similar between groups (33.6 ± 16.8 years for AdV E4, 34.2 ± 11.7 years for AdV D8, P = .81). Only 4 subjects among the 298 were <18 years of age, 3 of 36 in the AdV E4 group and 1 of 262 in the AdV D8 group. However, because only U.S. sites had approval to enroll children, and because AdV E4 was disproportionally frequent in the United States (see below), we cannot infer any age-based predilection of viral type from these data. There was no significant difference in the worse-presenting eye, with both groups showing approximately equal numbers of right and left eyes as worse. A trend toward higher male predominance was noted for AdV D8 compared with AdV E4 (60% vs 44%, P = .07 by Fisher exact test). Country of origin, however, did differ significantly between groups. Globally, among the 500 enrolled subjects, 390 (78%) were PCR-positive for any AdV. Of these, 36 (9%) were positive for AdV E4, compared with 262 (67%) who were PCR-positive for AdV D8. A total of 113 patients with conjunctivitis were enrolled in the NVC-422 trial from the United States; 63 (55%) of these subjects were PCR-positive for any AdV. Of these 63 subjects, 19 (30%) were PCR-positive for AdV E4. The U.S. cases of E4 were distributed among 6 geographically diverse centers, suggesting that this result was not caused by a single epidemic outbreak. Outside the United States, AdV E4 accounted for 17 of 390 (4%) PCR-positive cases ( P < .001 by χ 2 ). Although the United States accounted for 113 of 500 (23%) total subjects in the study, 19 of 36 (53%) patients with AdV E4 were from the United States. Conversely, outside the United States, AdV D8 accounted for 259 of 390 (66%) cases. Only 3 of 63 (5%) PCR-positive subjects from the United States were AdV D8 cases ( P < .001 by χ 2 ).


Patient Demographics

Adeno E4 Adeno D8 P value
N 36 262
Worse Eye 20L/16R 122L/140R 0.31
Age 33.6 +/- 16.8 34.2 +/- 11.7 0.81
Gender 20 F/ 16 M 103 F/
159 M
Treatment 16 NVC-422
20 Vehicle
141 NVC-422
121 Vehicle
Country of origin <0.00001
Brazil 4 (11 %)
4 Centers
62 (24%)
6 Centers
India 10 (28 %)
6 Centers
133 (51%)
16 Centers
Sri Lanka 3 (8 %)
3 Centers
64 (24%)
5 Centers
USA 19 (53%)
6 Centers
3 (1%)
3 Centers

CMT = central macular thickness; MV = macular volume; NPR = nonperfusion region;

R = Spearman correlation coefficient.

a Calculated using linear Spearman correlation. Two-tailed P values < .05 were considered statistically significant.

Of 13 measured signs and symptoms at presentation, 5 (tear breakup time, bulbar conjunctival injection, subjective burning, foreign body sensation, and photophobia) showed significant differences in distribution at presentation between AdV E4 and AdV D8 groups, with P < .05 in each case after Bonferroni correction ( Table 2 ). In each case, severity of signs and symptoms showed worse presentation in the AdV E4 group. The remaining 8 signs and symptoms showed no significant difference in pattern between groups. VA at presentation in the AdV E4 group was logarithm of minimal angle of resolution (logMAR) 0.14 ± 0.18 (Snellen equivalent 20/27), compared with 0.16 ± 0.16 (Snellen equivalent 20/29) for the AdV D8 group (nonsignificant, P = .61).


Signs and Symptoms of Adenoviral Conjunctivitis at Presentation

AdV E4 AdV D8 P value Bonferroni corrected
Tear breakup
Normal 26 237 0.0039 0.050
Abnormal 10 25
Corneal staining
None 30 234 0.27 >1
Present 6 28
Bulbar conjunctival injection
Absent 0 44 0.0033 0.042
Mild 5 87 Chi square
Moderate 18 90
Severe 8 41
Conjunctival edema
Absent 17 161 0.04679 0.59
Mild 10 74 Chi square
Moderate 7 24
Severe 2 3
Lid edema
Absent 12 89 0.2857 >1
Mild 16 111 Chi square
Moderate 8 51
Severe 0 11
Preauricular node
None 24 203 0.21 >1
Present 12 59
Excessive meniscus
None 21 138 0.59 >1
Present 15 123
Blurred vision
None 22 177 0.45 >1
Present 14 85
Absent 17 140 0.000074 0.00094
Mild 6 86 Chi square
Moderate 12 21
Severe 1 15
Foreign body sensation
Absent 9 103 0.00001 0.00013
Mild 8 101 Chi square
Moderate 17 40
Severe 2 18
Absent 13 138 0.022 0.28
Mild 9 78 Chi square
Moderate 10 29
Severe 4 17
Absent 20 183 0.0017 0.022
Mild 8 53 Chi square
Moderate 7 10
Severe 1 16
Absent 7 64 0.32 4.064
Mild 7 103 Chi square
Moderate 14 57
Severe 8 38

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jan 3, 2022 | Posted by in OPHTHALMOLOGY | Comments Off on Molecular and Clinical Characterization of Human Adenovirus E4–Associated Conjunctivitis

Full access? Get Clinical Tree

Get Clinical Tree app for offline access