Background: An outbreak of epidemic keratoconjunctivitis (EKC) due to adenovirus (Ad) type 8 and involving 14 members of the hospital staff and 33 neonates admitted to the Neonatal Intensive Care Unit of the local University Hospital occurred between September and December 2000 in Pavia, Italy. The outbreak was preceded by an outbreak of EKC within the community.
Objective: To compare the performance of conventional virus isolation on cell cultures, direct detection of Ad antigens in conjunctival cells by a direct fluorescent assay (DFA) and Ad DNA detection in conjunctival swabs by polymerase chain reaction (PCR) for diagnosis of adenoviral conjunctivitis.
Study design: Of conjunctival swabs collected from 47 patients, all were tested by virus isolation, 43 by direct Ad antigen detection, and 37 by Ad DNA detection. Direct Ad antigen detection was carried out by DFA using a group-specific monoclonal antibody. Detection and subgrouping of Ad DNA by nested PCR was performed using two sets of primers complementary to hexon and fiber genes, respectively.
Results: Ad was detected in 24/47 (51.1%), 21/43 (48.8%), and 23/37 (62.1%) samples by virus isolation, direct antigen detection and PCR, respectively. Overall, 30/47 (63.8%) samples were Ad-positive. Of 37 specimens tested in parallel by all three methods, Ad was detected by at least one of the three techniques in 26/37 (70.3%). All Ad isolates were identified as serotype 8 by neutralization, while all PCR-positive samples were identified as belonging to subgroup D. No other virus was isolated from any conjunctival swab. Time required for test completion was 9.6 (4-20) days for virus isolation, 1-2 h for DFA and 24 h for PCR.
Conclusions: DFA was a sensitive and rapid assay but results depend on the quality of sample and the expertise of the observer. PCR was the most sensitive assay, although it takes longer to perform and requires dedicated facilities; thus, it could be restricted to DFA-negative samples. Virus isolation is still useful from an epidemiological point of view.