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

9 / 152

Abstracts / Journal of Clinical Virology 82S (2016) S1–S142 S5
ANOVA were used to determine the significance of the relation-
ships between agent incidence and the meteorological factors.
Significant relationships with temperature, dew point, relative
humidity and fluctuation in humidity (humidity-range) were found
in many of the infectious agents, with HPIV-3, RSV and Influenza
viruses A and B showing the strongest correlations. Influenza
viruses A, B and RSV preferred a low temperature, dew point and
humidity-range, whilst also preferring a high humidity level.
Notably, HPIV-3 showed the opposite relationship. This is the first
time, an association between fluctuation in humidity and viral inci-
dence has been described.
A generalised linear model was constructed for each agent
to establish a statistically rigorous representation of its seasonal
pattern and the relationship with temperature, allowing for the
seasonality of all the agents to be confirmed.
The identification of each agent’s seasonal peaks allowed for the
prediction of the order inwhich each agent will arise in a given year,
starting with HPIV-3 in April, followed by rhinovirus, adenovirus,
HPIV-1, HPIV-2, M. pneumoniae, RSV, Influenza virus A, Influenza
virus B and ending with HMPV in March of the following year. A
change either in temperature or humidity or both was associated
with transition from one agent to another through the year.
Meteorological factors like temperature and humidity have a
significant effect on the incidence of the causative agents of the
common cold. This information could now possibly be used to pre-
dict the transition from one virus to another. By tracking changes
in meteorological factors, medical professionals could now be fore-
warned of an oncoming rise in infections from a particular agent,
allowing them to take appropriate preventative measurers.
Abstract no: 285
Presentation at ESCV 2016: Oral 9
Metagenomic analysis of the respiratory virome
associated with acute respiratory illness of
unknown etiology in infants
A. Ba ∗ . Pichon icard Billaud
J.S. Casalegno , M. Bouscambert-Duchamp ,
V. Escuret chuffenecker Valette
B. Lina , F. Morfin , L. Josset
1 Virology Department, University Hospital of Lyon,
France
2 Virpath, Inserm U1111 – CNRS UMR 5308, Lyon,
France
3 Internation Center for Research in Infectiology,
INSERM1111, Lyon University, Lyon France
Introduction: Acute respiratory infections (ARIs) are the lead-
ing cause of illness and death in children under five years old, who
experience three to six episodes per year. Viral infections are the
main etiology of ARIs but etiologic agents are often not identified.
Recent studies have suggested that virome has important effects
on human health. Therefore, respiratory virome characterization
may help to understand these ARIs of unknown etiology. The aim
of the study is to characterize the respiratory virome of children
under five with an ARI of undiagnosed etiology in order to iden-
tify potential novel respiratory viruses or variants not detected by
routine tests. For this purpose, a method for metagenomic analy-
sis of respiratory virome was optimized by implementing a quality
control to validate each steps of the process.
Methods: A retrospective study was conducted on samples
received at the virology laboratory of the University Hospital of
Lyon, France, between 2010 and 2015. Upper respiratory samples
from children under five were first selected by an epidemiolog-
ical approach based on respiratory viruses circulation patterns.
As an undiagnosed pathogen could be suspected in numerous
infectious diseases with negative biological investigation, “undi-
agnosed events”, defined as periods with proportion of negative
samples > 75% of the total number of samples received in the
laboratory, University Hospital of Lyon, were targeted. Samples
with a high probability of viral infections were selected accord-
ing to the following criteria: absence of documented infection
with routine techniques used at the time of diagnostic on at least
2 negative respiratory samples collected within 14 days of their
admission (hospitalization time > 24 h). These samples were then
controlled by a sensitive multiplex nested Polymerase Chain Reac-
tion (FilmArray® Respiratory Panel (FA RP), bioMérieux, Lyon,
France). After exclusion of positive samples with this technique,
a quality control (viral strain) was added to each sample before
metagenomic analysis.
Results: 223 patients were identified by targeting “undiagnosed
events” as described. Twenty-two patients with high probability
of viral infections were selected, 13/22 (59.1%) were under one
years old and 14/22 patients (63.6%) had comorbidities (mainly
respiratory chronic diseases). Patients developed mostly signs of
upper respiratory tract infection, such as cough and rhinorrhea,
but two patients developed severe respiratory distress with the
need of ventilation. Viruses were detected in most of the samples
with FA RP (17/22) (77.3%) mainly Picornaviridae viruses (13/22)
(59.1%). Ametagenomic analysis with a quality control process was
developed. An optimized metagenomic protocol was successfully
used for five negative patients. Sequencing analysis are currently
in progress.
Conclusion: No prior studies performed metagenomics anal-
ysis to characterize the respiratory virome involved in ARIs of
unknown etiology in infants. Identification of a novel respiratory
virus could have a strong impact on ARIs diagnosis and man-
agement. In absence of new virus identification, this study has
produced useful results describing the respiratory virome of chil-
dren with ARIs. Characterization of the whole viral communities
detectable in the human respiratory tract is key for understanding
the role of the virome in respiratory disease.
Abstract no: 37
Presentation at ESCV 2016: Oral 10
Assumption-free improvement of the maxRatio
algorithm increases the accuracy of qPCR assays
targeting viruses
Luigi Marongiu ∗, Eric Shain , Lydia Drumright ,
Reidun Lillestøl , Donald Somasunderam ,
Martin D. Curran
1 University of Cambridge, United Kingdom
2 Grove Street Technology, United Kingdom
3 Public Health England, Clinical Microbiology and
Public Health Laboratory, Addenbrooke’s Hospital,
United Kingdom
Introduction: Quantitative PCR (qPCR) is widely applied in
Laboratories of Virology worldwide for screening, diagnostic
and research purposes. Analysis of qPCR data is typically per-
formed with the cycle-threshold method (CT), which requires
the assignment of the cut-off and baseline range by the opera-
tor. These assumptions might impair the reproducibility of the
results between laboratories and could underestimate the impact
of inhibition of the reaction of amplification. The maxRatio method