Graefe’s Arch Clin Exp Ophthalmol

DOI 10.1007/s00417-006-0283-9 LABORATORY INVESTIGATION

Roman Wölfel
Martin Pfeffer
Evaluation of sampling technique and transport
Sandra Essbauer
Sylke Nerkelun
media for the diagnostics of adenoviral
Gerhard Dobler eye infections
Adenovirus sampling and transport

Received: 23 August 2005

Revised: 23 December 2005
Accepted: 20 January 2006
# Springer-Verlag 2006
R. Wölfel (*) . M. Pfeffer .
S. Essbauer . G. Dobler
Bundeswehr Institute of Microbiology,
Neuherbergstrasse 11,
80937 Munich, Germany
e-mail: romanwoelfel@bundeswehr.org
Tel.: +49-89-31682637
Fax: +49-89-31683292
S. Nerkelun
Bundeswehr Medical Specialist Service
Section of Ophthalmology,
Munich, Germany
Abstract Background: Human ade-
noviruses (HAdV) may cause pha-
ryngoconjunctival fever, follicular
conjunctivitis or epidemic keratocon-
junctivitis (EKC). Especially, out-
breaks of the latter may lead to severe
economic losses when preventive
measures are implemented too late.
Thus, a safe sampling method, proper
specimen transport conditions and a
fast and sensitive diagnostic technique
is mandatory. Methods: Two com-
mercially available virus transport
systems (VTS) were compared with
two NaCl-moisturised sampling de-
vices, one of which comprises Da-
cron-tipped plastic-shafted swabs and
the other a cotton-tipped wood-
shafted swab, available in most
ophthalmologists’ offices. Down-
stream methods for specific detection
of HAdV included direct
immunofluorescence assay (IFA) of
conjunctival swabs, virus isolation by
cell culture and quantitative real-time
polymerase chain reaction (qPCR).
Furthermore, the influence of appli-
cation of local anaesthetics prior to
swabbing on subsequent detection of
HAdV was investigated. Results:
Application of local anaesthetics had
a positive influence on the amount of
swabbed cells, thus increasing the
chance of obtaining positive results
by IFA. Neither isolation of HAdV by
cell culture nor by qPCR was nega-
tively influenced by this pretreatment.
Surprisingly, both commercially
available VTS performed signifi-
cantly worse than the NaCl-mois-
turised swabs. This was shown with
regard to virus recovery rates in cell
culture as well as viral genome copy
numbers in the qPCR. Conclusions:
Based on our results, the following
recommendations are provided to
improve sampling, transport and di-
agnostic techniques regarding con-
junctival swabs for diagnosis of
human adenovirus infection:
(1) application of local anaesthetics,
(2) NaCl-moisturised VTS for ship-
ment of specimens, and (3) detection
of HAdV by qPCR. The latter method
proved to be superior to virus isola-
tion by cell culture, including subse-
quent identification by IFA, because it
is faster, more sensitive and allows
simultaneous handling of a number of
samples. Hence, countermeasures to
prevent further virus spread in an
outbreak situation can be implemen-
ted earlier, thus reducing the number
of subsequent adenoviral infections.
Keywords Viral transport . Clinical
samples . Adenovirus
keratoconjunctivitis

Introduction tivitis (EKC). The latter is commonly caused by species

Human adenoviruses (HAdV) are a leading cause of acute
conjunctivitis in clinical practice [7, 8, 15]. HAdV
infections of the eye may lead to pharyngoconjunctival
fever, follicular conjunctivitis or epidemic keratoconjunc-
HAdV-D, particularly by serotypes 8, 19 and 37, whereas
follicular conjunctivitis and pharyngoconjunctival fever are
most frequently caused by HAdV-B serotypes 3 or 7 and
HAdV-E4 [1, 12, 18]. Most EKC outbreaks are commu-
nity-based and transmitted from person to person by
fingers or have been reported to occur due to contaminated
ophthalmologic solutions and instruments [5, 8, 23].
Because of crowded living conditions, military personnel
are at particularly high risk for ocular adenoviral infections.
For example, in spring 2004, an outbreak of conjunctivitis
appeared at different locations of the German armed forces,
causing disruption of military training and even closure of
some military facilities. Finally, a total of 940 eye swab
samples of symptomatic patients were analysed, but only
12 adenovirus-positive samples were identified. In conclu-
sion, a majority of patients who where clinically diagnosed
as EKC contracted conjunctivitis of another, yet unidenti-
fied, aetiology [24]. This incidence stresses that during an
epidemic outbreak of conjunctivitis, rapid laboratory
diagnosis is required to adopt appropriate preventive
measures and to limit spread of the virus if necessary
[21]. Collection and preparation of patients’ specimens is
of fundamental importance for diagnosis of adenoviruses.
Because HAdV are cell associated, specimens have to be
collected in such a way that they comprise as many intact
cells as possible [9]. During transport and shipment of the
specimens for further investigation, samples should be kept
under conditions allowing later identification of HAdV
either by virological or molecular methods.
The gold standard for laboratory diagnosis of HAdV
conjunctivitis is virus isolation by cell culture, usually
requiring 1–2 weeks for completion [16]. More rapid
assays have been developed, including enzyme immuno-
assay, direct immunofluorescence, DNA hybridisation and
electron microscopy [4, 16, 17, 19, 22]. These techniques
can be complex, laborious and often inappropriate for
routine situations. In contrast, real-time polymerase chain
reaction (real-time PCR) assays are now widely used for
diagnosis of viruses in most microbiological laboratories.
Protocols have been developed for HAdV detection in
ocular swab samples, offering significant improvement in
speed and sensitivity [10].
In cooperation with the German National Reference
Laboratory for Adenoviruses, Hanover Medical School, we
were involved in the investigation of a suspected outbreak
of EKC in the German Armed Forces, as mentioned above.
We found that in the past, recommendations for sampling
techniques were partially based on popular or traditional
assumptions [9]. Several studies investigating the ability of
different viral transport systems (VTS) to maintain viability
of HAdV were carried out [2, 6, 11, 13, 14]. However, it is
difficult to compare these results since they were carried
out under different conditions, employing a variety of
methods for viral detection. However, none of these studies
compared rapid virus isolation and quantitative real-time
PCR (qPCR). The purpose of our study was to address the
following questions:
– Is application of local anaesthetics to the eye prior to
swabbing necessary, and does it affect the detection of
HAdV by cell culture or quantitative real-time PCR?
– Does the material of the swabs or the swab shaft
influence the results of qPCR?
– Are there significant differences in the performance of
commercially available VTS compared with a simple
NaCl-moisturised transport method regarding detec-
tion of viruses by cell culture and/or qPCR?
Based on our experiments and results, we provide
recommendations on adequate VTS for efficient investiga-
tion of EKC outbreaks.
Materials and methods
Conjunctival samples
A total of 73 conjunctival swabs were collected from adult
patients with suspected viral conjunctival infections by
physicians at the Ophthalmology Specialist Service Sec-
tion, Munich. The eyes of 51 patients were anaesthetised
topically with one drop of oxybuprocaine (4 mg/ml,
Conjuncain single dose, Mann-Pharma, Berlin, Germany)
1 min prior to sampling; controls were taken from 22
patients without topical anaesthesia. Using a Dacron-tipped
swab, both upper and lower conjunctival surfaces were
vigourously wiped, rotating the swab during the sampling
process to ensure that the entire conjunctival surface was
sampled. The specimen swab was rolled, using slight
pressure, onto the 6-mm well area of a microscope slide. It
was ensured that the whole swab tip was used to prepare
the slide. The smear was air-dried at room temperature and
then fixed in ice-cold acetone for 10 min. After immediate
transfer to the laboratory, the number of conjunctival cells
on the slide was counted using a phase-contrast microscope
prior to diagnostic treatment of the specimens.
Cells and media
Culture cells employed in this study were A549 cells
(human lung cell line) kindly provided by Frank Hufert,
University of Freiburg, Germany. Cell cultures were grown
in minimum essential medium (MEM; Invitrogen, Karls-
ruhe, Germany) containing 3% foetal calf serum (FCS;
Biochrom, Berlin, Germany), sodium bicarbonate, penicil-
lin, streptomycin and amphotericin B (all from Invitrogen,
Karlsruhe, Germany).
Viruses
HAdV type 8 from a clinical isolate was kindly provided by
Hermann Schätzl, Technical University of Munich,
Germany. A viral stock was obtained by inoculation of
A549 cells. Supernatant was harvested when a nearly
complete cytopathic effect was observed and was briefly
centrifuged (10 min, 1,200 g) to pellet cellular debris. The
titre of the viral stock was determined by immunofluores-
cence (IFA), as described below. Stock aliquots containing
6×105 focus-forming units (FFU)/ml were stored at −80°C
until use. The HAdV stock was serially diluted in MEM to
test for different initial inoculum concentrations.
Topical anaesthetics
Oxybuprocaine-containing eye drops (50 μl, 25 μl and
12.5 μl) (novesine, 4 mg/ml oxybuprocaine, chlorhexidine
diacetate 90 μg/ml; OmniVision, Puchheim, Germany)
were mixed in a reaction tube with 50 μl of the following
HAdV solutions: FFU 2×105/ml, 2×104/ml, 2×103/ml.
MEM was added up to a final volume of 1,000 μl. Reaction
tubes were incubated for 60 min until further cell-culture
and qPCR processing.
Viral transport systems (VTS)
Two different VTS were evaluated. One was the Copan
New Virus Transport System (Copan VTS, Copan,
Bovezzo, Italy) comprising a liquid virus transport medium
in a sponge carrier located at the bottom of the tube. The
Dacron swab tip is placed in direct contact with the
medium-soaked sponge. About 900 μl transport medium
could be easily squeezed out of the tube for further sample
processing. The second was the Viral Culturette (BD VTS,
Becton-Dickinson, Heidelberg, Germany), a swab-tube
system containing Hanks balanced salt solutions. Medium
is sealed in a reagent cradle, which has to be broken after
insertion of the rayon-tipped swab. There is no instruction
given by the manufacturer as to how to recover the sample
from the tube: We added 900 μl MEM, inserted the swab
again and vortexed the tube. Thus, about 900 μl medium
could be extracted. For comparative purposes, we took 15-
ml reaction tubes (VWR, Darmstadt, Germany) comprising
900 μl sterile 0.9% NaCl (Braun, Melsungen, Germany)
plus either Dacron-tipped plastic-shafted swabs or cotton-
tipped wood-shafted swabs (all from VWR, Darmstadt,
Germany). All experiments were performed at least twice
by different investigators.
Inoculation of transport systems
Concentration-dependent stability Both commercial VTS,
Dacron-tipped plastic-shafted swabs and cotton-tipped
wood-shafted swabs stored in NaCl, were evaluated
using the following virus inocula: FFU 6×105/ml,
6×104/ml, 6×103/ml, 6×102/ml, 6×101/ml and 6×100/ml.
All swabs were inoculated with 50 μl of the specified viral
stock concentration and stored for 24 h in the dark at
21°C. The diluted stock controls were stored at −80°C
until final analysis.
Time-dependent stability The swab of each transport type
was inoculated with 50 μl of a virus dilution (6×104
FFU/ml) and immediately placed into the respective
transport medium. Dacron-tipped plastic-shafted swabs
stored in NaCl were used for comparison. Tubes were
maintained in the dark at 21°C for 0 h, 24 h, 48 h, 72 h,
96 h and 120 h. Subsequent to the respective storage
periods, tubes were frozen at −80°C and finally analysed
by IFA and qPCR. As controls, 50 μl-aliquots of swab-free
virus dilutions stored at −80°C for the corresponding
inoculation period were used.
Centrifugation culture and detection of viruses by IFA
Titres of viral samples and stocks were determined using
96-well plates containing 80% confluent A549 cells.
Samples were vortexed again prior inoculation. Following
aspiration of the medium, each well was inoculated with
100 μl sample. The plates were sealed with an adhesive
film and centrifuged at 2,000 g for 30 min at 25°C. After an
additional incubation period of 30 min, the inoculate was
removed. To each well was added 200 μl MEM containing
3% FCS, and plates were further incubated at 37°C and 5%
CO2. After 48 h, the cell culture supernatant was carefully
removed; plates were air dried and fixed with fresh
methanol-acetone (1:1 v/v) at room temperature for 30
minutes. IFA staining was done using the Imagen adeno-
virus kit (DakoCytomation, Cambridgeshire, UK): The
monoclonal antibody solution was diluted 1:3 with phos-
phate-buffered saline (PBS) without any remarkable loss of
sensitivity (data not shown). To each well, 40 μl of diluted
antibody solution was added. Plates were then incubated
within a moist chamber for 30 min at 37°C. Excess reagent
was washed out with 100 μl PBS. PBS was then drained
off, and the plates were again air dried at room temperature.
Mounting fluid (DakoCytomation, Cambridgeshire, UK)
was added, and wells were examined using an inverted
fluorescence microscope at x100 magnification. Cells with
an intracellular, nuclear or cytoplasmic granular apple-
green fluorescence were considered HAdV-positive. The
average turn-around time for clinical samples examined by
this technique is 56–60 h.
Quantitative real-time PCR (qPCR)
DNA was extracted from samples using the Qiagen DNA
kit (Qiagen, Hilden, Germany), according to the manu-
facturer’s instructions. qPCR was performed using a
LightCycler (Roche, Mannheim, Germany) with generic
primers, as described previously [10]. Cycling conditions
were modified slightly by reducing the extension period at
65°C from 60 s to 15 s. The average turn-around time for
clinical sample preparation and qPCR analysis was 4–8 h.
The positive control standard, a HAdV-2 PCR amplicon
(nt. 18856–19137 of the HAdV-2 sequence) cloned into a
pGEM-T easy plasmid vector, was kindly provided by
Albert Heim, Medical University of Hanover, Germany.
Concentration of the plasmid was determined by photom-
etry and calculated as genome equivalents (DNA copies)
per millilitre. Serial dilution of the HAdV plasmid was used
to create a standard curve. DNA concentrations were
calculated considering the sample dilution factor using the
LightCycler software. The standard was stored at −20°C in
aliquots until use.

Fig. 1 On average, only 10±2 conjunctival cells were wiped off

Statistical analysis from nonanaesthetised patients (n=22). In contrast, 119±7 conjunc-
tival cells can be obtained from patients with locally anaesthetised
The Wilcoxon signed rank test was used to determine eyes (n=51, P<0.0001)
significant differences between the different sample
groups. Statistical analyses were performed by using
GraphPad prism 4.00 for Windows (GraphPad Software,
San Diego, CA, USA). Results are given as mean ±
standard error of the mean (SEM). P values are given when
appropriate. Significance was accepted at P<0.05.

Results

Benefit of local anaesthesia

We found significant differences in the number of

conjunctival cells obtained from patients with locally
anaesthetised eyes compared with nonanaesthetised control
patients (Fig. 1). Particularly, united cell structures, as
demonstrated in Fig. 2, were only obtained following
pretreating the patients’ eyes with oxybuprocaine. To
investigate possible influences of the local anaesthetics on Fig. 2 Conjunctival cells and united cell structure wiped off an
cell culture and qPCR, we added different amounts of anaesthetised eye (a 100-fold magnification, b 400-fold magnification)
chlorhexidine-conserved oxybuprocaine-containing eye
drops to three solutions with different HAdV concentra-
tions . Concentrations of local anaesthetics were deliber- of the highest concentration of oxybuprocaine, we
ately overrated, corresponding to a maximum of 1 drop observed cytotoxic-like cell damage in seven of 18 wells.
added directly to a specimen. Results of titrations of However, in all cases, the appropriate FFU titre was re-
centrifuged cultures are summarised in Table 1. In presence trieved. Differences between the groups were nonsignifi-

Table 1 Stability of human adenoviruses (HAdV) in presence of local anaesthetics (LA)

Inoculum Log-focus-forming units per millilitre
log FFU/ml Control 50 μl LA 25 μl LA 12.5 μl LA
Mean SEM Number Mean SEM Number Mean SEM Number Mean SEM Number

5.3 ∞a 6 ∞a 4 ∞a 6 ∞a 6
4.3 4.0 ± 0.01 6 3.9 ± 0.06 3 4.0 ± 0.02 6 4.0 ± 0.02 6
3.3 2.9 ± 0.08 6 2.9 ± 0.11 4 2.9 ± 0.16 6 3.0 ± 0.08 6
FFU focus-forming units, SEM standard error of the mean
a
More than 300 focus-forming units per well
cant (P >0.9). We did not see any comparable cell damage
in the diagnostic process of specimens from locally
anaesthetised patients. Likewise, by using DNA extraction
and qPCR protocols described above, we did not find any
significant influence of oxyprocain on real-time detection
of HAdV-DNA (data not shown), indicating that neither the
local anaesthetic nor the conservation material inhibited
detection of HAdV by qPCR.

Influences of VTS on cell culture and qPCR results

We compared the performance of NaCl-moisturised cotton-

tipped wood-shafted swabs and Dacron-tipped plastic-
shafted swabs to the Copan VTS and BD VTS: Titration of
several solutions with different virus concentrations
employing the culture centrifugation technique provided
no significant difference between controls and cotton-
tipped wood-shafted swabs. In strong contrast to these
results, almost 90% of the infectious virus was lost by
storage in Copan or BD VTS. (Table 2, Fig. 5). The number
of DNA copies of samples stored in Copan or BD VTS was
significantly lower compared with controls. The corre-
sponding qPCR results of the remaining two swab systems
revealed no sign of inhibition by cotton- and/or wood-
containing swabs after storage for 2 days (Fig. 3).
Stability of adenovirus storage in different VTS
Clinical HAdV specimens are commonly sent by postal
service for further virological investigation. Therefore, we
tested the stability of a 6×104 FFU/ml virus solution stored
in the dark at room temperature. At day 0, similar amounts
of HAdV were detected in all transport media when tested
by immunofluorescence (Table 3) However, the HAdV
titre decreased in the Copan VTS as well as the BD VTS as
early as 24 h in the storage period. After 3 days, no virus
count was detected for either commercial systems employ-
ing culture centrifugation technique or IFA. In the
Fig. 3 Concentration-dependent stability of human adenovirus
(HAdV) DNA in different transport media: The amount of viral
DNA that can be recovered from commercial virus transport systems
(VTS) is up to 2 logs lower when compared with a NaCl-
moisturised transport system (P=0.03). The quantitative real-time
polymerase chain reaction (qPCR) results of cotton- and wood-
containing swabs did not significantly differ from the Dacron-made
swabs
corresponding samples stored with NaCl-moisturised
Dacron-tipped plastic-shafted swabs, viable virus was
detected for up to 5 days using cell culture. With qPCR,
differences between the tested transport media were less
distinct (Fig. 4). Inoculated virus was detected over 5 days
in commercial VTS as well as Dacron/NaCl swabs.
Discussion
Specimen collection and transport is an important issue in
virological diagnostics. In case of a suspected outbreak of
EKC, reliable microbiological diagnosis is needed to
confirm clinical diagnosis or to avoid inappropriate
treatment and overreactions. The first objective of this

Table 2 Concentration-dependent stability of human adenoviruses (HAdV) after 24-h storage

Inoculum Log-focus-forming units per millilitre
log FFU/ml Control Copan VTS BD VTS Cotton/wood/NaCl
Mean SEM Mean SEM Mean SEM Mean SEM

5.8 ∞a 4.6 ± 0.08 4.5 ± 0.03 ∞a

4.8 4.7 ± 0.07 3.4 ± 0.03 3.3 ± 0.03 4.6 ± 0.02
3.8 3.8 ± 0.05 2.4 ± 0.06 2.4 ± 0.10 3.6 ± 0.09
2.8 2.7 ± 0.07 Øb Øb 2.5 ± 0.16
1.8 Øb Øb Øb Øb
0.8 Øb Øb Øb Øb
FFU focus-forming units, VTS viral transport system, SEM standard error of the mean
a
More than 300 focus-forming units per well
b
Less than one focus-forming unit per well
Table 3 Time-dependent stability of human adenovirus (HAdV) in virus transport systems (VTS) over 5 days
Log-focus-forming units per millimetre
Day Control Copan VTS BD VTS Dacron/NaCl
Mean SEM Mean SEM Mean SEM Mean SEM

0 4.7 ± 0.13 4.3 ± 0.04 4.2 ± 0.06 4.7 ± 0.01

1 4.7 ± 0.08 3.7 ± 0.05 3.5 ± 0.06 4.7 ± 0.04
2 4.7 ± 0.13 2.8 ± 0.06 2.6 ± 0.00 4.7 ± 0.02
3 4.7 ± 0.03 2.3 ± 0.00 Øa 4.6 ± 0.01
4 4.7 ± 0.03 Øa Øa 4.3 ± 0.03
5 4.6 ± 0.01 Øa Øa 3.8 ± 0.01
a
Less than one focus-forming unit per well

study was to determine if local anaesthesia of the eye Results of this study on HAdV infectivity in centrifu-
should be recommended for HAdV specimen collection. gation indicate a significant difference between the tested
Because HAdV replicates mainly within the nucleus [3], it VTS for virus preservation. In contrast to specific virus
is most important to collect as many conjunctival cells as
possible. We showed that topical application of an
anaesthetic is an appropriate way of collecting material
for diagnostics, which does not interfere with downstream
techniques such as qPCR, an important screening method,
or virus isolation by cell culture. The cytotoxic-like effects
that we observed in cell culture at the highest concentration
of local anaesthetics are most likely not relevant for the
investigation of clinical samples. The appropriate amount
of anaesthetic eye drops (50 μl) corresponds to the
complete dose normally administered to a patient’s eye.
Therefore, the amount of local anaesthetics in the specimen
will always be much lower.

Fig. 4 Time-dependent stability of human adenovirus (HAdV)

DNA in the investigated transport media stored at room temperature.
Curve progressions for Copan virus transport system (VTS) and BD
VTS differ significantly from controls (P=0.03). However, differ-
ences between control and Dacron/NaCl are nonsignificant (P=0.13)
Fig. 5a, b Representative details of immunofluorescence-stained
human adenovirus (HAdV)- infected cells in 96-well cell culture
(100-fold magnification). a Virus stored on an NaCl-moisturised
swab for 2 days. b Same virus sample stored in Copan VTS
transport media, sterile cotton-tipped wood-shafted swabs
are available in most general practitioners’ and ophthal-
mologists’ offices. However, no investigation is available
on inhibition of PCR or interference with cell culture by the
natural materials these swabs are made off [20]. In both
commercial systems (Copan VTS and BD VTS), HAdV
survival was reduced when compared with NaCl-mois-
turised Dacron-tipped plastic-shafted swabs. These results
are in contrast to previous studies [6, 13], which found
adenovirus to be rather stable in commercial VTS. This
discrepancy may partly be explained by differences in the
test protocols, such as incubation time (up to 17 days) or
virus quantification (TCID50). For example, high concen-
trations of HAdV type 2 were used to inoculate different
VTS, and this may be a further reason why no decrease in
virus titre was detected at 22°C over a period of 10 days
[13]. Nevertheless, our data demonstrate a loss of HAdV
viability depending on the VTS if a routine diagnostic test
(rapid cell culture detection method) is used as the read-out
system.
The surprising results of the well-preserving nature of
NaCl-moisturised swabs without any additives are also not
easy to interpret. One suggestion we can come up with is
that at least some of the ingredients of the commercial VTS
interfere with virus viability. This might be due to the
nonenveloped morphology of HAdV naturally having a
high tenacity. But we have no explanation as to why a
commercially available VTS causes more harm to a virus
than physiological NaCl moisture does. Similar experi-
ments with other nonenveloped viruses, e.g. enteroviruses,
may show whether this HAdV-specific phenomenon does
apply to other nonenveloped viruses as well.
Concerning recommendations for adenovirus viral
transport medium, HAdV DNA was fairly stable in all
media tested, but both commercially available systems
from Copan and Becton-Dickinson did not perform as well
as the NaCl-moisturised swabs in preserving HAdV
infectivity for rapid cell culture detection. Hence, neither
system can be recommended for specimen transport of
ophthalmologic samples for HAdV diagnosis, regardless of
the downstream methods used. Furthermore, both com-
mercial VTS were up to three times more expensive
compared with the swab/tube combinations, which were
used in this study. In conclusion, our recommendation for
an adenovirus transport medium is an NaCl-moisturised
Dacron-tipped plastic-shafted or cotton-tipped wood-
shafted swab stored and shipped in a standard reaction
tube. Although virus titres maintained the initial level for
almost 4 days, fast transport and shipment is crucial to
achieve diagnosis in due course. Because of its much lower
turn-around time and wider availability compared with
rapid cell culture detection, we strongly recommend qPCR
for a fast and sensitive screening of ophthalmologically
relevant HAdV.
Acknowledgements We acknowledge the excellent technical
support of Heike Prabel, Aileen Lorber, Peter Klein, Rahime
Terzioglu and Gudrun Zöller during the course of this study. We
thank Dr. Albert Heim, German National Reference Laboratory for
Adenoviruses, Hanover Medical School, Germany, for his helpful
support in setting up the quantitative HAdV real-time PCR. The
views expressed in this article are those of the authors and do not
necessarily reflect the official policy or position of the German
Ministry of Defence or the German Government.

References
1. Aoki K, Tagawa Y (2002) A twenty-
one year surveillance of adenoviral
conjunctivitis in Sapporo, Japan. Int
Ophthalmol Clin 42:49–54
2. Azevedo AM, Durigon EL, Okasima V,
Queiroz DA, de Moraes-Vasconcelos
D, Duarte AJ, Grumach AS (2003)
Detection of influenza, parainfluenza,
adenovirus and respiratory syncytial
virus during asthma attacks in children
older than 2 years old. Allergol
Immunopathol (Madr ) 31:311–317
3. Boniuk M, Phillips CA, Hines MJ, and
Friedman JB (1966) Adenovirus infec-
tions of the conjunctiva and cornea.
Trans Am Acad Ophthalmol
Otolaryngol 70:1016–1026
4. Darougar S, Walpita P, Thaker U,
Viswalingam N, Wishart MS (1984)
Rapid culture test for adenovirus iso-
lation. Br J Ophthalmol 68:405–408
5. Dawson C, Jawetz E, Hanna L, Winn
WE, and Thompson C (1960) A family
outbreak of adenovirus 8 infection
(epidemic keratoconjunctivitis). Am
J Hyg 72:279–283
6. Dunn JJ, Billetdeaux E, Skodack-Jones
L, and Carroll KC (2003) Evaluation of
three Copan viral transport systems for
the recovery of cultivatable, clinical
virus isolates. Diagn Microbiol Infect
Dis 45:191–197
7. Fitch CP, Rapoza PA, Owens S,
Murillo-Lopez F, Johnson RA, Quinn
TC, Pepose JS, Taylor HR (1989)
Epidemiology and diagnosis of acute
conjunctivitis at an inner-city hospital.
Ophthalmology 96:1215–1220
8. Ford E, Nelson KE, and Warren D
(1987) Epidemiology of epidemic
keratoconjunctivitis. Epidemiol Rev
9:244–261
9. Gardner PS, McQuillen J (1980) Ap-
plication of immunofluorescence. In:
Rapid virus diagnosis, 2nd edn.
Butterworth, London, pp 92–109
10. Heim A, Ebnet C, Harste G, Pring-
Akerblom P (2003) Rapid and quanti-
tative detection of human adenovirus
DNA by real-time PCR. J Med Virol
70:228–239
11. Huntoon CJ, House RF, Jr., and Smith
TF (1981) Recovery of viruses from
three transport media incorporated into
culturettes. Arch Pathol Lab Med
105:436–437
12. Jawetz E, Thygeson P, Hanna L,
Nicholas A, Kimura SJ (1957) The
etiology of epidemic keratoconjuncti-
vitis. Am J Ophthalmol 43:79–83
13. Jensen C, Johnson FB (1994) Com-
parison of various transport media for
viability maintenance of herpes simplex
virus, respiratory syncytial virus, and
adenovirus. Diagn Microbiol Infect Dis
19:137–142
14. Johnson FB (1990) Transport of viral
specimens. Clin Microbiol Rev
3:120–131
15. Kinchington PR, Turse SE, Kowalski
RP, and Gordon YJ (1994) Use of
polymerase chain amplification reac-
tion for the detection of adenoviruses in
ocular swab specimens. Invest
Ophthalmol Vis Sci 35:4126–4134
16. Kowalski RP and Gordon YJ (1989)
Comparison of direct rapid tests for the
detection of adenovirus antigen in
routine conjunctival specimens.
Ophthalmology 96:1106–1109
17. Lehtomaki K, Julkunen I, Sandelin K,
Salonen J, Virtanen M, Ranki M, Hovi
T (1986) Rapid diagnosis of respiratory
adenovirus infections in young adult
men. J Clin Microbiol 24:108–111
18. Mellman-Rubin TL, Kowalski RP,
Uhrin M, Gordon YJ (1995) Incidence
of adenoviral and chlamydial coinfec-
tion in acute follicular conjunctivitis.
Am J Ophthalmol 119:652–654
19. Pereira HG, Allison AC, Balfour B
(1959) Multiplication of adenovirus
type 5 studied by infectivity titrations
and by the fluorescent antibody tech-
nique. Virology 7:300–314
20. Radstrom P, Knutsson R, Wolffs P,
Lovenklev M, and Lofstrom C (2004)
Pre-PCR processing: strategies to gen-
erate PCR-compatible samples. Mol
Biotechnol 26:133–146
21. Richmond S, Burman R, Crosdale E,
Cropper L, Longson D, Enoch BE,
Dodd CL (1984) A large outbreak of
keratoconjunctivitis due to adenovirus
type 8. J Hyg (Lond) 93:285–291
22. Vastine DW, Schwartz HS,
Yamashiroya HM, Smith RF, and Guth
SB (1977) Cytologic diagnosis of
adenoviral epidemic keratoconjunctivi-
tis by direct immunofluorescence.
Invest Ophthalmol Vis Sci 16:195–200
23. Warren D, Nelson KE, Farrar JA,
Hurwitz E, Hierholzer J, Ford E,
Anderson LJ (1989) A large outbreak
of epidemic keratoconjunctivitis:
problems in controlling nosocomial
spread. J Infect Dis 160:938–943
24. Wolfel R, Pfeffer M, Dobler G, and
Finke EJ (2004) [Virological investi-
gation of Keratoconjunctivitis epidem-
ica]. Wehrmed Wehrpharm 2:36–41