Journal of Clinical Virology - Volume 82 - Supplement 1 - September 2016

16 / 152

S12 Abstracts / Journal of Clinical Virology 82S (2016) S1–S142
lial tissues. A RT-PCR was optimized to amplify complete EV-D68
genomes, which were then sequenced using Ion Torrent PGM.
Results: We detected a high prevalence of EV-D68 in 2012 and
2014, with respectively 55 (15.5%) out of 355 and 37 (11.1%) cases
detected out of 334 samples tested. In contrast, only six (out of
352) samples were tested positive for EV-D68 in 2010, only one
out of 242 samples in 2011, 0 out of 314 samples in 2013 and one
out of 417 in 2015. EV-D68 detected in 2012 belonged to both
clades A (34.3%) and B (66.7%) while EV-D68 detected in 2014
belonged mainly to clade B (76.5%). We were able to produce most
of the EV-D68 complete genomes. Phylogenetic analysis of these
sequences reveals the five-year molecular evolution of EV-D68. We
also correlate variations observed in the sequences with clinical
characteristics of the patients.
Discussion: This analysis may highlight viral polymorphisms
associated with mild or severe forms of EV-D68 infection. These
polymorphisms could be used as prognosis or diagnosis markers
and could be used in clinical management of patient infected with
EV-D68.
Abstract no: 169
Presentation at ESCV 2016: Oral 21
Norovirus molecular epidemiology in a
paediatric UK hospital: Unexpected diversity,
seasonality and sources of infection
J.R. Brown ∗, S. Roy , D. Shah , C. Ruis ,
R. Williams ara Romero reuer
1 Great Ormond Street Hospital for Children NHS
Foundation Trust, United Kingdom
2 UCL, United Kingdom
Norovirus is the leading cause of gastroenteritis worldwide, typ-
ically associatedwith seasonal outbreaks of diarrhoea and vomiting
inhealthcare institutions anddominatedby a single genotype, GII.4.
Using custom-designed norovirus baits for target enrichment
(SureSelect) and Illumina sequencing we have sequenced the com-
plete genomes of all norovirus episodes in a paediatric UK hospital
over a 19 month period (July 2014–February 2016), consisting of
193 episodes from 186 patients.
We identified a broad range of norovirus genotypes, consist-
ing of 34% GII.P21 GII.3, 22% GII.4, 16% GII.P4 GII.4, 1%
GII.P16 GII.4 and 27% a mixture of GI.P1 GI.1, GI.1, GI.3,
GI.3, GI.P3 GI.3, GI.P2 GI.2, GII.P2 GII.2, GII.P7 GII.6,
GII.P7 GII.7, GII.P16 GII.17 and GII.P16 GII.17. We did not see
the typical winter seasonality expected with norovirus; instead
there is a correlation between the number of norovirus episodes
and the number of hospital admissions per month (R = 0.703,
P = 0.011). We estimate that movement in and around the hospital
accounts for 50% of the variability in the number of norovirus
episodes per month. A broad range of genotypes and lack of season-
ality is analogous to norovirus epidemiology in the community,
rather than the epidemiology seen in non-paediatric healthcare
settings.
MaximumLikelihood phylogeny identified 19 sequence clusters
suggestive of transmission, involving median 3 patients per clus-
ter (range 2–17). This included clusters that were widely separated
in time or location, therefore could not be identified by conven-
tional epidemiological methods. Only 47% of cases were part of
a cluster while 53% of norovirus episodes, including some appar-
ently acquired as an inpatient, were sporadic cases with no onward
transmission.
We postulate that in this paediatric setting the majority of
norovirus episodes are caused by frequent introductions of infec-
tion from the community, possibly by infected visitors or staff. In
contrast fewer than 50% of cases are associated with nosocomial
transmission. These findings have implications for infection control
policy in paediatric facilities. Importantly, in addition to established
practice of preventing transmission from known in-patient cases,
efforts should be focused on identifying norovirus episodes in new
admissions and visitors. The diversity of genotypes identified in this
study has major implications for vaccine development.
Abstract no: 66
Presentation at ESCV 2016: Oral 22
Longer duration of viral shedding following
infection with a novel norovirus GII.4 strain
L. Gustavsson ∗, R. Nordén ,
V. Letelier Molnegren , M. Lindh , J. Westin ,
L.-M. Andersson
1 Department of Infectious Diseases, Institute of
Biomedicine, University of Gothenburg, Gothenburg,
Sweden
2 Department of Clinical Microbiology, Sahlgrenska
University Hospital, Gothenburg, Sweden
Background: Prolonged viral excretion is common following
norovirus gastroenteritis. A key determinant of the duration of
shedding is the host immune response. Major new strains of geno-
type GII.4 virus emerge every two to four years through antigenic
drift, and the appearance of antigenically novel strains is associated
withmore widespread norovirus epidemics. The aim of the present
study was to investigate if duration of viral shedding is longer in
patients infected with such emerging, epidemic GII.4 strains, as
compared to patients infected with previously circulating strains
Materials and methods: We conducted a prospective cohort
study during three consecutive norovirus seasons (2010–2013),
which included patients hospitalised with community-onset
norovirus gastroenteritis. Patients with concurrent bacterial
infection or immunosuppression were excluded. Norovirus was
diagnosed with real-time PCR and subtype strain was deter-
mined from capsid sequence (ORF2), via phylogenetic tree analysis.
Follow-up samples were obtained weekly until four weeks after
inclusion. Rapid clearance was defined as a negative sample for
norovirus by day 14, and slow clearance was defined as shedding
of norovirus RNA for more than fourteen days. Proportions were
compared with Fisher’s exact test.
Results: Novel epidemic strains appeared twice during the
study period; GII.4-2010 (New Orleans) in the first season, and
GII.4-2012 (Sydney) in the 2012-2013 season. We included 24
patients from these two seasons, in whom duration of shedding
could be determined and a GII.4 virus was found. Median age was
83 years (range 25–99) and 50% were women.
The majority of patients (n = 19) were infected with viruses that
clusteredwith the respective season’s epidemic strain, whereas five
patients were infected with previously circulating subtypes (GII.4-
2008 or older in 2010/11; GII.4-2010 or older in 2012/13). Slow
clearance was observed in 1/5 (20%) patients infected with previ-
ously circulating strains. This proportion was significantly higher
among patients infectedwith an epidemic strain, where 16/19 (84%,
p = 0.01) remained PCR-positive for more than two weeks.
Conclusion: In this limited pilot study, we found that a long
duration of viral shedding appears to be more common in patients
infected with novel, epidemic norovirus GII.4 strains. This may be