Differential susceptibility of adenovirus clinical isolates to cidofovir and ribavirin is not related to species alone.

Abstract

We have previously reported that human adenovirus (HAdV) reference strains clearly show species-dependent resistance to ribavirin, whereas different species of HAdV are equally sensitive to cidofovir. All the serotypes tested were susceptible to cidofovir, whereas only serotypes from species C were sensitive to ribavirin. Here, we aimed to extend these investigations to clinical isolates.
In vitro, we tested 126 isolates obtained from 65 patients included in a European survey of HAdV infection.
Among the 126 isolates tested, all presented cidofovir 50% inhibitory concentration (IC50) in the same range as the HAdV 5 reference strain. Regarding ribavirin, all isolates from species C (79 tested) showed an IC50 comparable with previously reported results for reference strains; however, 24/32, 2/6 and 3/3 tested isolates from species A, B and D, respectively, were shown to have a ribavirin IC50 comparable with the HAdV 5 reference strain (species C), contrary to previous observations for reference strains of the same species. Among patients who were treated with cidofovir for disseminated HAdV infection, > or = 4 sequential isolates could be obtained from 9 patients; no variation in cidofovir susceptibility could be detected.
Cidofovir is active in vitro in all HAdV clinical isolates. Ribavirin was revealed to be active on most HAdV isolates from species A, B and D, and in all isolates from species C. Finally, no resistance to cidofovir became apparent in sequential isolates obtained from treated patients.

Full text

Page 1

© 2009 International Medical Press 1359-653555
Antiviral Therapy 14:55–61
Background: We have previously reported that human
adenovirus (HAdV) reference strains clearly show species-
dependent resistance to ribavirin, whereas different species
of HAdV are equally sensitive to cidofovir. All the serotypes
tested were susceptible to cidofovir, whereas only serotypes
from species C were sensitive to ribavirin. Here, we aimed
to extend these investigations to clinical isolates.
Methods: In vitro, we tested 126 isolates obtained from
65 patients included in a European survey of HAdV
infection.
Results: Among the 126 isolates tested, all presented
cidofovir 50% inhibitory concentration (IC50) in the same
range as the HAdV 5 reference strain. Regarding ribavi-
rin, all isolates from species C (79 tested) showed an IC50
comparable with previously reported results for reference
strains; however, 24/32, 2/6 and 3/3 tested isolates from
species A, B and D, respectively, were shown to have a
ribavirin IC50 comparable with the HAdV 5 reference strain
(species C), contrary to previous observations for reference
strains of the same species. Among patients who were
treated with cidofovir for disseminated HAdV infection,
≥4 sequential isolates could be obtained from 9 patients;
no variation in cidofovir susceptibility could be detected.
Conclusions: Cidofovir is active in vitro in all HAdV clinical
isolates. Ribavirin was revealed to be active on most HAdV
isolates from species A, B and D, and in all isolates from
species C. Finally, no resistance to cidofovir became appar-
ent in sequential isolates obtained from treated patients.
Human adenoviruses (HAdVs) are DNA viruses com-
prising 51 serotypes that are classified into 6 spe-
cies (A–F). HAdV is associated with a wide range
of diseases, such as respiratory, gastrointestinal and
conjunctival diseases that are usually self-limiting in
immunocompetent patients. For many years, HAdVs
have frequently been reported as an emerging patho-
gen causing significant morbidity and mortality after
haematopoietic stem cell transplantation (HSCT),
especially in children [1–3]. Infections could remain
local and asymptomatic or become disseminated
before becoming associated with a potentially fatal
clinical course warranting antiviral treatment [4].
Although serotypes from each of the six HAdV spe-
cies could be involved, serotypes from species C (1,
2 and 5) and B (7, 14 and 35) are most frequently
responsible for HAdV infections in HSCT patients [2].
Recently, HAdV 31 (species A) infections have also
Original article
Differential susceptibility of adenovirus clinical
isolates to cidofovir and ribavirin is not related to
species alone
Florence Morfin1,2*, Sophie Dupuis-Girod3,4, Emilie Frobert1, Stéphanie Mundweiler1,2, David Carrington5,
Petr Sedlacek6, Marc Bierings7, Petr Cetkovsky8, Aloys CM Kroes9, Maarten JD van Tol10 and
Danielle Thouvenot1,2
1Laboratoire de Virologie et Pathogénèse Humaine, Université Lyon 1, CNRS FRE 3011, Lyon, France
2Laboratoire de Virologie, Centre de Biologie et Pathologie Est, Hospices Civils de Lyon, Lyon, France
3Hematology and Bone Marrow Transplantation Unit, Hôpital Debrousse, Hospices Civils de Lyon, Lyon, France
4Genetics Unit, Hôpital de l’Hôtel Dieu, Hospices Civils de Lyon, Lyon, France
5Regional Virus Laboratory, Bristol, United Kingdom
6Department of Pediatrics, Motol University Hospital, Prague, Czech Republic
7Wilhelmina Kinderziekenhuis Hematology, Utrecht, the Netherlands
8Institute of Hematology and Blood Transfusion, UHKT, Prague, Czech Republic
9Department of Medical Microbiology, Leiden University Medical Center, Leiden, the Netherlands
10Department of Pediatrics, Leiden University Medical Center, Leiden, the Netherlands
*Corresponding author: E-mail: florence.morfin-sherpa@chu-lyon.fr
Introduction

Page 2

F Morfin et al.
© 2009 International Medical Press
56
been recognized in HSCT patients, including many
cases of nosocomial transmission [5].
There are two main antiviral drugs used in the treat-
ment of HAdV infections: cidofovir [6] and ribavirin [7].
Both drugs induce a variety of clinical outcomes (see [4]
for review). Cidofovir is often associated with clinical
improvement [8–13], although failures have also been
reported [14]. The effect of ribavirin on HAdV infections
is much more controversial: many failures have been
reported [15–17], as well as a few successes [7,18,19].
These discrepancies in the reported data could be associ-
ated with a wide range of factors, including the timing
of treatment initiation with respect to the progress of the
infection, the degree of immunocompetence in patients
during the treatment period and the susceptibility of the
virus to antiviral drugs. By focusing on reference strains
of HAdV, we have previously reported on the species-
dependent susceptibility of HAdV to ribavirin and the
broader activity for cidofovir on the different HAdV
serotypes [20].
The aim of the present study was to extend these
in vitro susceptibility results, which were mainly
obtained from laboratory reference strains, to a large
panel of clinical isolates collected from HSCT patients
followed-up for HAdV infections within a prospective
collaborative European study (unpublished data).
Methods
European HAdV infection follow-up protocol
Patients were followed-up for 6 months after HSCT.
Stool, urine and throat samples for viral culture and
plasma samples for real-time quantitative PCR were
taken prior to transplantation, once weekly until
60 days after HSCT and every 4 weeks thereafter until
6 months after HSCT.
Cell lines and virus
Antiviral assays were performed in human larynx
epidermoid carcinoma cells (HEp-2; American type
culture collection number CCL-23). The clinical iso-
lates tested were collected via a collaborative Euro-
pean study, in which HSCT centres in France (Lyon),
the United Kingdom (Bristol), the Netherlands (Leiden
and Utrecht) and the Czech Republic (Prague) took
part. Viruses were first isolated from clinical samples
in different cell lines after one or two passages. Virus
stocks used to perform antiviral susceptibility assays
were then prepared on HEp-2 cells with one addi-
tional passage. All strains were propagated using the
cell-associated viruses as the inoculum.
In vitro susceptibility assays
We evaluated the sensitivity of HAdV to two antiviral
agents: cidofovir ((S)-1-(3-hydroxy-2-phosphonylmeth-
oxypropyl) cytosine; HPMPC; Pharmacia & Upjohn,
St Quentin-Yvelines, France) and ribavirin (1-β-d-
ribofuranosyl-1,2,4-triazole-3-carboxamide;
Aldrich, St Quentin-Fallavier, France). Assays were
performed as previously described [20].
In summary, we used a chess board titration of each
virus strain, making possible simultaneous titration
of the virus against a range of antiviral drug concen-
trations. Cells were infected with 10-fold dilutions
(10-1–10-5) of cell-associated virus. Each dilution was
incubated with a series of five concentrations of each
antiviral drug. After 5 days at 36°C with 5% CO2, viral
multiplication was measured using an ELISA. For each
antiviral drug concentration, optical densities were rep-
resented in graph form in relation to virus dilution. By
using logistic regression analyses, this graph was used
to calculate the concentration of drug resulting in 50%
inhibition of viral replication (IC50) as already described
[21]. For each isolate and each antiviral drug, the IC50
ratios were determined as the ratio between the IC50
value of the clinical isolate and the IC50 value of an
HAdV 5 reference strain tested in the same assay. A
virus was considered to be resistant when both the IC50
value and the IC50 ratio were high (>5 for the ratio).
Sigma–
Serotyping and genotyping
Serotyping of clinical HAdV isolates was performed by
conventional haemagglutination inhibition as previously
described [22]. Strains that could not be serotyped were
genotyped for species using the Adenovirus Consensus
kit (Argène, Varhiles, France) on the basis of hybridiza-
tion of PCR products with species-specific probes.
Cytotoxicity assays
The cytotoxicity of cidofovir and ribavirin was evalu-
ated on confluent cell layers of HEp-2 cells. We used
five concentrations of each antiviral drug ranging from
50 to 5,000 µM; dilutions were performed in Eagle’s
Minimum Essential Medium supplemented with 2%
fetal calf serum. After 5 days incubation at 36°C with
5% CO2, cells were detached by trypsinization and
counted in trypan blue and the concentration of drug
cytotoxic for 50% of the cells (CC50) was calculated
using regression analyses.
Results
A total of 126 HAdV isolates obtained from 65 patients
were analysed (Table 1). Viruses were mainly from spe-
cies C (63%) with serotypes 1, 2, 5 and 6, and from
species A (25%) with serotype 31. A few isolates from
species B (serotype 3) and D (serotype 9) were also
tested. In total, 41 isolates could not be serotyped using
classical methods and so were genotyped for HAdV
species: 21 from species A, 15 from species C, 3 from

Page 3

HAdV clinical isolates in vitro susceptibility
Antiviral Therapy 14.1
57
species B and 2 from species D. Six isolates could not be
serotyped or genotyped.
For the HAdV 5 reference strain included in each
assay, mean IC50 values were 15 µM for cidofovir
(±sd=9, range 5–35, second quartile =12.5 and third
quartile =24) and 57 µM for ribavirin (±sd=21, range
28–114, second quartile =54 and third quartile =67).
For clinical isolates, the mean IC50 value for cido-
fovir was 24 µM (range 2–82). Mean susceptibility
ratio (isolate IC50/HAdV 5 reference strain IC50) was
2 (range 0.14–9.40) for cidofovir. Eight isolates pre-
sented a susceptibility ratio of >5 (3 isolates with a
ratio of >8), but all showed IC50 values within the nor-
mal range. All the isolates tested presented susceptible
levels that were similar to cidofovir as the HAdV 5
reference strain.
With ribavirin treatment, two groups emerged
clearly. For the first group (n=114), mean IC50 was
86 µM (range 4–269). Mean susceptibility ratio was
1.60 (range 0.06–5.93). One isolate presented with a
susceptibility ratio of >5, but the ribavirin IC50 was
166 µM and the virus was thus not considered to be
resistant to this drug. For the second group (n=12),
IC50 was >500 µM, except for one isolate with a riba-
virin IC50 of 292 µM, but this isolate also had a ratio
of 6.21.
For the 79 HAdV species C isolates tested, all pre-
sented the same susceptibility range as the HAdV 5 ref-
erence strain to both cidofovir and ribavirin. For the 32
species A isolates, all were susceptible to cidofovir as
was the HAdV 5 reference strain; with ribavirin treat-
ment, 24 isolates presented IC50 in the same range as the
HAdV 5 reference strain (first group) and 8 presented
increased IC50 (second group). For the species B isolates,
all were susceptible to cidofovir as was the HAdV 5 ref-
erence strain; with ribavirin treatment, 2 presented IC50
in the same range as the HAdV5 reference strain (first
group) and 4 presented increased IC50 (second group).
The three species D isolates tested presented IC50 com-
parable to the HAdV 5 reference strain for both cidofo-
vir and ribavirin. The IC50 results of the HAdV isolates
from species A, B and D that presented IC50 comparable
with the HAdV 5 reference strain (species C), which
represent a sensitivity profile that is different from pro-
files previously observed on reference strains from the
same species [20], are shown in Table 2.
For nine patients, infectious HAdV strains could
be isolated on ≥4 occasions and were tested for their
susceptibility to cidofovir and ribavirin. The results
concerning these isolates are shown in Table 3. Two
patients (R and W) had never been treated with either
cidofovir or ribavirin: one patient (W) presented with
faecal excretion of the same serotype (C1) over a
2-month period and one patient (patient R) had fae-
cal excretion of species C over a 1-month period. No
difference in the IC50 values for the two drugs tested
was observed in the species C isolates from either of
these two patients. For five patients, HAdV could be
recovered after a short period (<15 days) of cidofovir
therapy (patients N, S, U, V and X); no changes in
cidofovir IC50 were detected in pre-therapy and post-
therapy isolates or among isolates during the treat-
ment period. For two patients (T and Y), the HAdV
isolates were obtained during a long period (2 and 6
months, respectively) of cidofovir treatment and again
no change in cidofovir IC50 was noted. With regards to
ribavirin treatment, all sequential isolates from these
patients were susceptible to this drug and most of them
were from species C.
The CC50 of cidofovir was 1,000 µM in HEp-2 cells,
which is ≥40× higher than the observed IC50 values.
Ribavirin tested on HEp-2 cells had a CC50 at 400 µM,
which is 5× higher than the mean ribavirin IC50.
Discussion
The results presented here were obtained from 126
HAdV clinical isolates from 65 patients, all included
in a prospective follow-up of HAdV infections after
HSCT as part of a collaborative European study.
Of the serotypes tested from species A, B, C and D,
no difference in cidofovir IC50 values could be observed
regardless of the serotype. The IC50 values were within
the same range for these viruses and no resistant HAdV
strains could be detected among patients treated with
cidofovir and from whom sequential isolates were
tested. This is in line with the conditions required for
in vitro selection of a laboratory cidofovir-resistant
virus in cell culture under the selective pressure of
cidofovir [23]. The resistant strain, named R3, was
selected after 28 passages in the presence of increasing
Species Patient, n (%) Isolate, n (%) Serotype Isolate, n
A

B

C




D

NT
Total
19 (29)

4 (6)

36 (55)




3 (5)

3 (5)
65
32 (25)

6 (5)

79 (63)




3 (2)

6 (5)
126
31
NT
3
NT
1
2
5
6
NT
9
NT
NT
–
11
21
3
3
14
35
14
1
15
1
2
6
–
Table 1. Number of patients and adenovirus isolates tested
NT, non-typed.

Page 4

F Morfin et al.
© 2009 International Medical Press
58
concentrations of cidofovir. In comparison, resistance
of herpes simplex virus to acyclovir can be established
in vitro after only one or two passages in cell culture
in the presence of acyclovir [24]. These in vitro consid-
erations suggest that the selection of cidofovir-resistant
HAdV is very difficult and this concept is supported by
our results on the HAdV clinical isolates obtained from
treated patients.
Regarding the susceptibility of HAdV clinical isolates
to ribavirin, all the viruses from species C presented
IC50 comparable to the HAdV 5 reference strain for
this drug, thus confirming previously reported results
obtained on reference strains for serotypes 1, 2, 5 and
6 [20]. For the HAdV clinical isolates from species
A, B and D, a different pattern of ribavirin IC50 was
observed in comparison with the laboratory reference
strains. Reference strains from these species (serotypes
12 and 31 from species A, serotypes 3, 7, 11 and 14
from species B and serotype 9 from species D) presented
high IC50 for this drug (>500 µM) [20]. However, of the
clinical isolates, 75% of the species A viruses (24 out of
32), 33% of the species B viruses (2 out of 6) and 100%
of the species D viruses (3 out of 3) showed a lower IC50
with values in the same range as the HAdV 5 reference
strain (species C). These HAdV species thus included
viruses with high or low IC50 to ribavirin. This differs
from previously reported results on in vitro ribavirin
susceptibility between clinical isolates and reference
strains [20,25], but this might be associated with the
low number of clinical isolates tested in these two stud-
ies (14 and 3, respectively). This observation does not
seem to be related to the serotype as we detected HAdV
31 isolates (species A) with low IC50 for ribavirin and
other isolates with high IC50 for this drug. An attempt
was made to investigate the genetic basis of these differ-
ences in ribavirin susceptibility. We first focused on the
viral DNA polymerase gene of the HAdV 31 isolates
that either had low or high IC50 for ribavirin. Sequenc-
ing results showed no specific differences between these
strains (data not shown). Nevertheless, according to the

Patient/isolate

Sample

Date of sample

Species

Serotype
CDV
IC50, µM
CDV
ratio
RBV
IC50, µM

RBV ratio
Previous
antiviral treatment
A/1
A/2
B/1
B/2
B/3
C
D
E
F
G
H
I
J
K/1
K/2
L/1
L/2
M/1
M/2
N/1
N/2
N/3
N/4
N/5
O/1
O/2
P
Q
R
Stools
Stools
Stools
Stools
Urine
Stools
Stools
Stools
Stools
Stools
Stools
Stools
Stools
Stools
Throat
Stools
Stools
Stools
Stools
Stools
Stools
Stools
Throat
Stools
Stools
Throat
Stools
Stools
Stools
11/08/2003
19/08/2003
18/07/2003
23/07/2003
30/07/2003
27/08/2003
23/09/2003
11/02/2003
02/09/2003
26/06/2003
01/09/2003
18/02/2003
10/10/2003
23/01/2002
23/01/2002
04/08/2004
03/09/2004
17/12/2001
06/01/2003
12/11/2003
17/11/2003
24/11/2003
01/12/2003
03/12/2003
16/01/2004
13/02/2004
06/07/2004
28/06/2004
31/03/2005
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
D
B
D
B
D
NT
NT
NT
NT
NT
NT
NT
31
NT
31
31
31
NT
31
31
NT
31
31
31
NT
NT
NT
NT
NT
9
3
NT
3
NT
20
43
20
32
23
12
36
2
29
26
23
37
30
37
29
32
9
25
28
27
41
31
36
47
48
7
31
11
7
2.50
1.79
0.83
2.67
2.88
1.71
1.50
0.18
4.14
2.00
2.09
5;29
4.29
1.09
0.85
1.88
1.29
0.74
0.82
1.00
1.52
1.15
1.33
1.74
4.36
1.40
5.17
1.83
1.00
100
157
147
224
245
251
9
4
135
166
56
20
146
42
100
115
100
75
154
152
140
175
225
113
190
39
229
269
41
1.67
2.62
2.45
3.73
4.08
4.18
0.32
0.06
1.69
5.93
1.04
0.30
1.83
0.89
2.13
2.80
1.23
1.60
3.28
1.45
1.33
1.67
2.14
1.08
2.84
1.11
3.75
4.41
0.95
0
CDV
0
0
0
0
0
0
0
0
0
0
0
0
0
0
CDV
0
0
0
0
CDV
CDV
CDV
0
0
0
0
0
Table 2. IC50 of clinical adenovirus isolates from patients presenting with ribavirin IC50 results different from reference strainsa
aReference strains from species A, B and D were susceptible to cidofovir (CDV; calculated drug concentration inhibiting 50% of viral replication [IC50] ranging from 32
to 81 µM) and resistant to ribavirin (RBV; IC50>500 µM) [20]. NT, non-typed.

Page 5

HAdV clinical isolates in vitro susceptibility
Antiviral Therapy 14.1
59
complicated mechanism of action of ribavirin, many
other viral genes could also be involved. Ribavirin is
a purine nucleoside analogue that is converted to its
triphosphate form by cellular kinases. Antiviral activ-
ity has been reported against a broad range of viruses
including hepatitis C virus, Lassa virus, influenza
virus, parainfluenza virus, respiratory syncytial virus
and HAdV [26]. Five distinct mechanisms have been
proposed to explain the antiviral properties of ribavi-
rin. These include both indirect mechanisms (for exam-
ple, inosine monophosphate dehydrogenase inhibition
inducing depletion of the cellular guanosine triphos-
phate pool and immunomodulatory effects on the basis
of T-cell response induction) and direct mechanisms

Patient/isolate

Sample

Date of sample

Species

Serotype
CDV
IC50, µM
CDV
ratio
RBV
IC50, µM
RBV
ratio
Previous
antiviral treatment
N/1
N/2
N/3
N/4
N/5
R/1
R/2
R/3
S/1
S/2
S/3
S/4
T/1
T/2
T/3
T/4
U/1
U/2
U/3
V/1
V/2
V/3
V/4
V/5
V/6
V/7
V/8
V/9
W/1
W/2
W/3
W/4
X/1
X/2
X/3
X/4
X/5
Y/1
Y/2
Y/3
Y/4
Y/5
Y/6
Y/7
Stools
Stools
Stools
Throat
Stools
Stools
Stools
Stools
Stools
Stools
Stools
Urine
Stools
Urine
Stools
Stools
Stools
Stools
Nose
Throat
Blood
Urine
Throat
Stools
Stools
Blood
Stools
Blood
Stools
Stools
Stools
Stools
Urine
Throat
Stools
Throat
Stools
Stools
Stools
Stools
Stools
Urine
Stools
Stools
12/11/2003
17/11/2003
24/11/2003
01/12/2003
03/12/2003
07/02/2005
11/02/2005
14/03/2005
05/09/2003
11/09/2003
16/09/2003
16/09/2003
18/03/2003
10/04/2003
16/12/2003
12/02/2004
03/01/2005
07/01/2005
10/01/2005
14/04/2003
25/04/2003
28/04/2003
28/04/2003
28/04/2003
05/05/2003
05/05/2003
12/05/2003
12/05/2003
30/05/2003
23/06/2003
07/07/2003
04/08/2003
04/05/2004
10/05/2004
10/05/2004
16/08/2004
23/08/2004
03/01/2003
25/02/2003
08/04/2003
13/05/2003
17/06/2003
15/07/2003
11/08/2003
A
A
A
A
A
C
NT
NT
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
NT
NT
NT
NT
NT
1
NT
NT
NT
NT
2
NT
2
2
2
2
NT
1
NT
2
2
2
2
2
2
2
2
2
1
1
1
NT
5
5
5
5
5
2
2
2
2
2
2
2
27
41
31
36
47
10
20
7
27
40
50
42
20
9
20
10
45
20
15
76
51
47
62
13
20
9
40
11
10
12
19
19
7
5
11
6
11
11
9
11
5
6
8
8
1.00
1.52
1.15
1.33
1.74
1.43
2.86
1.00
1.13
5.00
6.25
5.25
1.18
1.29
2.86
0.59
1.88
1.43
1.07
3.80
2.55
9.40
3.10
0.65
4.00
3.00
8.00
3.67
1.67
2.40
1.46
1.46
0.20
0.14
0.31
0.17
0.31
2.20
1.80
0.61
0.42
0.50
0.67
0.67
152
140
175
225
113
8
127
125
41
20
100
50
65
100
132
100
69
100
63
115
158
70
138
83
77
49
82
45
22
46
134
100
53
100
60
41
43
58
55
132
65
67
55
62
1.45
1.33
1.67
2.14
1.08
0.12
1.90
1.87
0.68
0.49
1.67
0.83
1.38
2.13
2.81
2.13
1.64
1.23
0.78
2.13
2.93
1.56
2.56
1.54
1.71
1.04
1.82
0.96
0.76
1.59
1.34
1.00
1.29
2.44
1.46
1.00
1.05
1.71
1.62
2.54
0.82
0.85
0.70
0.78
0
0
CDV
CDV
CDV
0
0
0
CDV
CDV
CDV
CDV
0
0
CDV
CDV
0
0
CDV
0
0
0
0
0
CDV
CDV
CDV
CDV
0
0
0
0
0
0
0
CDV
CDV
0
CDV
CDV
CDV
CDV
CDV
CDV
Table 3. In vitro susceptibility to cidofovir and ribavirin of sequential clinical adenovirus isolates
CDV, cidofovir; IC50, calculated drug concentration inhibiting 50% of viral replication; NT, non-typed; RBV, ribavirin.