doi:10.1111/j.1365-2761.2009.01131.x
Journal of Fish Diseases 2010, 33, 391–402
Salmonid alphavirus (SAV) and pancreas disease (PD) in
Atlantic salmon, Salmo salar L., in freshwater and seawater
sites in Norway from 2006 to 2008
M D Jansen1, T Taksdal2, M A Wasmuth2,*, B Gjerset2,*, E Brun2, A B Olsen3, O Breck4 and
M Sandberg1
1
2
3
4
Norwegian School of Veterinary Science, Oslo, Norway
National Veterinary Institute Oslo, Norway
National Veterinary Institute Bergen, Norway
Marine Harvest Norway, Bergen, Norway
Abstract
A cohort study was initiated in the spring of 2006
to investigate epidemiological aspects and pathogenesis of salmonid alphavirus (SAV) subtype 3
infections and pancreas disease (PD). The aims
were to assess involvement of the freshwater production phase, the extent and frequency of subclinical infections and to follow PD-affected
populations throughout the entire seawater production cycle, as well as investigate possible risk
factors for PD outbreaks. Fish groups from 46
different Atlantic salmon freshwater sites in six
counties were sampled once prior to seawater
transfer and followed onto their seawater sites. A
total of 51 Atlantic salmon seawater sites were included, and fish groups were sampled three times
during the seawater production phase. SAV subtype
3 was not identified by real-time RT-PCR from
samples collected in the freshwater phase, nor were
any SAV-neutralizing antibodies or histopathological changes consistent with PD. In the seawater
phase, SAV was detected in samples from 23 of 36
(63.9%) studied sites located within the endemic
region. No SAV subtype 3 was detected in samples
from seawater sites located outside the endemic
region. The cumulative incidence of PD during the
production cycle amongst sites with SAV detected
Correspondence M D Jansen, Norwegian School of Veterinary
Science, Centre for Epidemiology and Biostatistics, PO Box 8146
Dep., N-0033 Oslo, Norway
(e-mail: Mona.Dverdal.Jansen@nvh.no)
*These authors contributed equally to the work.
2010
Blackwell Publishing Ltd
391
was 87% (20 of 23 sites). Average fish weight at
time of PD diagnosis ranged from 461 to 5978 g,
because of a wide variation in the timing of disease
occurrence throughout the production cycle. Mortality levels following a PD diagnosis varied greatly
between populations. The mean percentage mortality was 6.9% (7.06) (range 0.7–26.9), while the
mean duration of increased mortality following PD
diagnosis was 2.8 months (1.11) (range 1–6).
Keywords: Atlantic salmon, cohort study, epidemiology, pancreas disease, salmonid alphavirus.
Introduction
Pancreas disease (PD) is a viral fish disease with
significant impact on Norwegian salmonid aquaculture. Following the first description of PD in
farmed Atlantic salmon, Salmo salar L., in Scotland
in 1976 (Munro, Ellis, McVicar, McLay & Needham 1984), the disease has been diagnosed in
North America (Kent & Elston 1987), Norway
(Poppe, Rimstad & Hyllseth 1989), Ireland (Murphy, Rodger, Drinan, Gannon, Kruse & Korting
1992), France and Spain (Raynard, Houghton &
Munro 1992). In Norway, the disease affects both
farmed Atlantic salmon and rainbow trout, Oncorhynchus mykiss (Walbaum), in sea water (Olsen
& Wangel 1997; Christie, Fyrand, Holtet &
Rowley 1998; Hodneland, Bratland, Christie,
Endresen & Nylund 2005; Taksdal, Olsen, Bjerkås,
Hjortaas, Dannevig, Graham & McLoughlin
2007), contrary to experience in Ireland and
Journal of Fish Diseases 2010, 33, 391–402
Scotland where only marine Atlantic salmon production has been affected.
The aetiological agent was first isolated in Ireland
and named salmon pancreas disease virus (SPDV)
(Nelson, McLoughlin, Rowley, Platten & McCormick 1995), and later identified as an alphavirus
(Weston, Welsh, McLoughlin & Todd 1999). The
species name salmonid alphavirus (SAV) was suggested later (Weston, Villoing, Bremont, Castric,
Pfeffer, Jewhurst, McLoughlin, Rodseth, Christie,
Koumans & Todd 2002), and several subtypes
exist. In Ireland, SAV subtypes 1, 4 and 6 have been
isolated from fish affected by PD, while Scottish
outbreaks have been caused by SAV subtypes 1, 2, 4
and 5 (Weston, Graham, Branson, Rowley, Walker,
Jewhurst, Jewhurst & Todd 2005; Fringuelli,
Rowley, Wilson, Hunter, Rodger & Graham
2008; Ruane, Graham & Rodger 2008). Only
SAV subtype 3 has been isolated from Norwegian
PD outbreaks (Hodneland et al. 2005; Weston
et al. 2005; Karlsen, Hodneland, Endresen &
Nylund 2006; Fringuelli et al. 2008).
The number of seawater sites diagnosed with PD
in Norway increased from five in 1995 to 15 in
2001, with a further increase to 109 sites in 2008.
The endemic region of the central south-western
coast gradually extended southwards in 2004 and
northwards in 2006, resulting in the entire southwestern coast being considered an endemic region
by the end of 2006. In addition, sporadic outbreaks
of PD have been recorded in northern Norway
since 2003, with a cluster of repeated outbreaks
over several production cycles seen in the northernmost county between 2005 and 2008. A new
regulation was introduced in Norway in November
2007 making PD a notifiable disease and resulting
in the setting up of a restriction zone encompassing
the endemic region. The regulation aims to avoid
further spread of PD out of the endemic region, as
well as to mitigate disease within this region.
Epidemiological observations and investigations
into cases of PD in Norway, Ireland and Scotland
have been performed previously. The initial observations described the disease occurring only in the
seawater phase and that subsequent inputs of fish in
affected sites were diseased (McVicar 1987). Fallowing for at least 3 weeks, single generation
rearing, slaughtering away from site and no staff
movement between sites were found to reduce losses
because of PD (Wheatley, McLoughlin, Menzies &
Goodall 1995). More recent studies in Ireland
identified varying smolt strain susceptibility in the
2010
Blackwell Publishing Ltd
392
M D Jansen et al. SAV and PD in Norway 2006–2008
field (McLoughlin, Peeler, Foyle, Rodger, OÕCeallachain & Geoghegan 2003), and later in experimental studies (McLoughlin, Graham, Norris,
Matthews, Foyle, Rowley, Jewhurst, MacPhee &
Todd 2006). Movement during the production
cycle, high feeding rate prior to an outbreak, a PDinfected farm in the same water body, stocking
above 250 000 fish, previous PD status at the site,
high lice burden and geographical location have all
been found to have an association with PD at a site
(Rodger & Mitchell 2007). In Norway, well boat
movements and lack of fallowing have been found
to be associated with PD (Brun, Olsen & Rørvik
2005), as has infection pressure from nearby farms
with PD (Kristoffersen, Viljugrein, Kongtorp, Brun
& Jansen 2009).
No evidence was available on the importance of
subclinical SAV infections in Norway, although
their occurrence had been documented from
Ireland and Scotland (McLoughlin, Rowley &
Doherty 1998; Graham, Jewhurst, McLoughlin,
Sourd, Rowley, Taylor & Todd 2006). Although
there were no reports of clinical outbreaks of PD in
fry and fingerlings, a Norwegian SAV PCR-product
had been sequenced from a freshwater site (Karlsen
et al. 2006). As a result, an investigation into the
possible involvement of freshwater sites as SAV
reservoirs because of vertical transmission was
required. To investigate knowledge gaps, a cohort
study was initiated in the spring of 2006 to study
the prevalence, cumulative incidence and geographical distribution of SAV and PD in Norway, and
any involvement of the freshwater phase on the
SAV and PD status of the studied sites. Additionally, an investigation into possible risk factors for
PD at the studied sites was undertaken. Molecular
studies of isolates from SAV-positive sites presented
in this article will be published separately.
Materials and methods
Study design
A prospective, longitudinal cohort study was
designed to allow groups of Atlantic salmon to be
sampled at their freshwater sites and subsequently
followed throughout the seawater production cycle.
A total of 46 freshwater sites rearing Atlantic
salmon for commercial production were selected
(Fig. 1). The number of sites to be included was
based on an expected incidence of PD of approximately 30% among seawater sites in areas where
Journal of Fish Diseases 2010, 33, 391–402
M D Jansen et al. SAV and PD in Norway 2006–2008
Figure 1 Maps showing the locations of freshwater study sites (grey triangles, left hand map) and seawater study sites (black triangles,
centre map) in the cohort study in Norway 2006–2008. The locations of the seawater study sites found salmonid alphavirus/pancreas
disease positive during the study period (black dots, right hand map) are shown relative to the endemic region (indicated by the
rectangle) as defined at the end of the cohort study period.
PD had previously been diagnosed. From the
endemic region, 27 freshwater sites (Fig. 1) were
randomly selected by the Norwegian Food Safety
Authority (NFSA). An additional 19 sites from the
non-endemic region (Fig. 1) were selected from the
Norwegian Directorate of FisheriesÕ Aquaculture
Licence Register, by a random procedure. Of these
freshwater sites, 25 were hatcheries for their own
stock. The remaining 21 freshwater sites represented fish from an additional 15 different hatcheries. The stocks were hatched in the winter and
spring of 2005 for sea transfer the following spring
as 1-year-old smolts. At each freshwater site, a single
tank with increased smolt mortality or morbidity
was selected by the fish health personnel to
represent the freshwater study group. If no tank
with increased mortality or morbidity was available,
one tank was randomly chosen. Approximately
3 weeks prior to the expected time of seawater
transfer, 60 randomly selected smolts, including five
moribund smolts where possible, were sampled
from each group (Table 1).
Thirty-eight freshwater study groups were followed to their seawater phase at 42 seawater sites.
2010
Blackwell Publishing Ltd
393
Table 1 The number of Atlantic salmon sampled at each site at
each sampling point in the cohort study in Norway 2006–2008
Sampling point
Freshwater
Sea water
Sampled
group
3 weeks
prior to
transfer
2 months
posttransfer
8 months
posttransfer
At
slaughter
Study group
Control group
60
–
30
30
30
30
30
–
–, not sampled.
Three study groups were distributed to more than
one seawater site. One study group was distributed
to two seawater sites, while two study groups were
each distributed to three seawater sites. Additionally, one seawater site received smolts from two
study groups. From the reserve list of potential
study sites, an additional nine seawater sites stocked
with previously un-sampled smolt groups were
included in the study, to ensure an adequate sample
size and a more complete coverage of the Atlantic
salmon production area in Norway. As a result, a
Journal of Fish Diseases 2010, 33, 391–402
total of 51 sites were included in the seawater phase
(Fig. 1). At each seawater site, one cage containing
stock from the freshwater study group was selected
and termed the seawater study group. A second
cage, with stock originating from a different
freshwater site, was selected as an on-site control
group where possible. The nine sites without a
freshwater study group had two cages randomly
selected. The inclusion of an on-site control group
was done to allow detection of SAV status differences with regard to freshwater origin and assessment of agent spread within a site. At
approximately 2 and 8 months post-seawater transfer, 30 salmon, if possible including five moribund
fish, were sampled from each group, resulting in a
total of 60 salmon per site per sampling point (90
salmon for the site stocking two sampled freshwater
groups). At slaughter, 30 fish from the study groups
were sampled (Table 1). If the study group was
unavailable, the on-site control group or, as a last
resort, a previously un-sampled group was sampled.
Sites that submitted a reduced sample number
included three sites from the freshwater phase, as
well as three, 11 and six seawater sites at 2 and
8 months post-transfer and slaughter, respectively.
Additionally, one site in the endemic region failed
to submit any samples at 2 months post-transfer.
Eight months post-transfer, two seawater sites in the
endemic region with a diagnosis of PD close to the
sampling due date did not submit samples, while
three sites in the non-endemic region failed to
submit samples. At slaughter, six sites with an
existing PD diagnosis, five SAV-negative sites in the
endemic region and six sites in the non-endemic
region failed to submit samples (site data not
shown).
M D Jansen et al. SAV and PD in Norway 2006–2008
Tissue sampling
All salmon were euthanized prior to sampling,
either by a blow to the head or by an overdose of
anaesthetic. Sampling was performed by the individual companiesÕ fish health personnel or by staff
from the NFSA.
Heart ventricle and mid-kidney were sampled in
RNAlater (Ambion) for examination by real-time
RT-PCR. Gill, heart ventricle and atrium, pancreas,
liver, mid-kidney and red and white skeletal muscle
were fixed in 10% buffered phosphate formalin for
histopathological examination. Additionally, heparinised blood samples for serological analyses were
collected (Table 2).
Samples were shipped the same day on ice with
overnight delivery, or, if necessary, refrigerated
overnight prior to shipping to the NVI (Oslo). All
samples, except formalin-fixed tissues, were stored
at )80 C until analysis was performed.
Virology. The SAV status was determined by realtime RT-PCR analyses. A total of 20-mg mixed
tissue (heart ventricle and mid-kidney) was added to
700 lL NucliSens lysisbuffer (Biomérieux) together
with a tungsten carbide bead (QIAGEN Nordic),
and homogenized using the Mixer Mill MM 300
(Retch).
Ribonucleic
acid
(RNA)
from
homogenized (individual or pooled) samples was
extracted using the automated easyMAG protocol
(Biomérieux), eluted in 25 lL volume, and the
quantity and purity of the extracted RNA measured
by optical density using NanoDrop ND-1000
(NanoDrop
Technologies).
The
real-time
Stratagene PCR system (Stratagene) using 1-lg
template RNA with a QIAGEN OneStep RT-PCR
kit (QIAGEN Nordic) and primers PDF1 (5¢-CCC
Records of PD
Independent of the present study, all study sites
were monitored by regular fish health services.
Standard diagnostic procedures were followed,
including submission of samples for laboratory
assistance when needed. A PD diagnosis required
the detection of tissue changes characteristic for PD
by histopathology combined with the detection of
SAV (Taksdal et al. 2007). The majority of study
sites that were diagnosed with PD were examined at
the National Veterinary Institute (NVI), while two
were examined by other laboratories. All study site
PD diagnoses were recorded together with the
remaining site information.
2010
Blackwell Publishing Ltd
394
Table 2 The number of samples collected from 46 freshwater
sites and 51 seawater sites during the cohort study in Norway
2006–2008
Sampling point
Freshwater Sea water
Sample type
3 weeks
prior to
transfer
Tissues on
60
RNAlater
Blood samples
30
Tissues on
10
buffered formalin
2 months 8 months
postpostAt
transfer
transfer
slaughter
2 · 30
2 · 30
30
2 · 15
2·5
2 · 15
2·5
30
5
Journal of Fish Diseases 2010, 33, 391–402
CAT TCG ATC GCA AAG TA-3¢) and PDR1 (5CGT ACA GAT CAT TGG GCT TGA C-3)
covering a 141nt part of the E1 fragment
complementary to bases 10906–11046 of the
Norwegian SAV SF21/03 isolate (AY604238) and
PDE1-MGB TaqMan probe (FAM-AAG AGG
TCT ATA ACT ACG ACT GG-NFQ-MGB) was
used. A one-step RT-PCR protocol was carried out
using 0.5 lm of each primer and 0.3 lm probe in a
25-lL reaction, with cDNA synthesis at 50 C for
30 min and inactivation of the reaction by 15 min
at 95 C, followed by 45–50 cycles of 94 C for
30 s, 55 C for 30 s and 72 C for 45 s. The
results were interpreted on the basis of both curve
shape and obtained cycle threshold (Ct) values. Any
samples showing inconclusive results, giving Ct
values ‡40 or an anomalous curve shape, were reanalysed. When re-analysing, both standard primers
and probe as well as alternative primers and a probe
in the nsP1 region (Hodneland & Endresen 2006)
were used, with interpretation criteria as detailed
earlier.
The best practice guidelines were followed to
avoid cross-contamination between test samples.
Negative and positive controls were included for
each run, with results outside the accepted Ct value
range of positive controls considered invalid. Analysis performed on tenfold dilutions of cell culturegrown SAV isolates showed the E1 and nsP1
primers/probe sets to have similar analytical sensitivity.
Pooled samples were created by adding together
150 lL from each of five homogenized individual
samples to a final volume of 750 lL. For each
freshwater site, ten samples, including moribund
smolts if present, were analysed individually. The
remaining 50 samples were analysed as 10 pooled
samples. Seawater sites had six pooled samples
created per study- and control group, resulting in
twelve pooled samples per site per sampling point.
Where re-analysis of pooled samples was required,
new RNA was extracted from the individual
samples constituting the pool and analysed individually. All samples were analysed according to the
protocol described earlier.
Prior to laboratory analyses of the study samples,
a pilot study was performed to evaluate the dilution
effect of pooling. SAV-positive samples already at
the NVI were combined with four volumes of
samples known to be SAV-negative to create a 1/5
pool. Both the individual positive samples and the
pooled samples were analysed according to the real 2010
Blackwell Publishing Ltd
395
M D Jansen et al. SAV and PD in Norway 2006–2008
time RT-PCR procedure described earlier. The Ct
values obtained from both the individual and
pooled samples were recorded.
Serology. Plasma samples were examined by a virus
neutralization test for the detection of antibodies
against SAV (Graham, Jewhurst, Rowley,
McLoughlin & Todd 2003) with some
modifications (Taksdal et al. 2007).
Histopathology. After fixation, the samples were
processed and embedded in paraffin wax according
to standard procedures. Sections (4–6 lm) were
stained with haematoxylin and eosin and examined
by light microscopy, to detect any changes in
accordance with PD (Taksdal et al. 2007).
Data collation and statistical methods
Site identification, together with laboratory results,
was entered into Microsoft Excel and descriptive
statistics performed by the use of Stata SE 10
(StataCorp). Site data were collected through
questionnaires, with open ended and closed questions, which were distributed to the fish health
personnel of each site. Follow-up phone calls or email contact was used where required. The questionnaires collected information relating to the fish
stock (smolt strain), sea transfer data (peak stocking
numbers, well boat information) and site parameters [previous PD diagnosis (2004 or later), fallowing duration prior to stock transfer, stock
movement during production cycle, cage type].
Additional monthly production data on stock
numbers, average weights and mortalities and losses
were acquired through a national production
database (confidential data, not openly accessible).
Seaway distances to nearest slaughter plant and
vessel fairways were calculated as described by
Kristoffersen et al. (2009), as were the infection
pressure values. Analysis of mortality levels and
duration of elevated mortality following PD was
performed. Data were analysed using a Cox
proportional hazards model to attempt detection
of possible risk factors associated with a positive PD
status of the population at the site. The time
variable was defined as the number of weeks
between seawater transfer and the detection of
SAV or PD (grouped and termed PD-positive sites)
or from seawater transfer until the time of collection
of the last negative sample (PD-negative sites). The
failure variable was the classification of the site as
M D Jansen et al. SAV and PD in Norway 2006–2008
Journal of Fish Diseases 2010, 33, 391–402
Table 3 Potential predictor variables analysed for association with sites positive for salmonid alphavirus (SAV) and pancreas disease
(PD) in the endemic PD area
Predictor variable
Variable type
Description
Smolt strain
Categorical
Peak stocking numbers
Dichotomous
Well boat transfersa
Previous PDa
Fallowing durationa
Dichotomous
Dichotomous
Dichotomous
Stock movement
Cage typea
Vessel fairways
Slaughter planta
Infection pressure
Dichotomous
Dichotomous
Continuous
Continuous
Continuous
Smolt strain stocked. Three common smolt strains and two minor strains, or mixed
populations stocked. Re-categorized into three common strains and a minor/
mixed group.
Peak number of fish stocked at the site. Stocking below or above 600 000
salmon at peak stock numbers. Re-categorized from continuous on data
distribution.
Stock transferred with £2 or ‡3 well boat shipments.
Previous PD diagnosis on site from 2004
Fallowing duration prior to stock transfer, fallowing period £12 weeks or
>12 weeks.
Stock moved between sea water sites during production cycle
Use of compact or polar cages at site
Distance between site and main vessel fairways
Distance between site and nearest salmon slaughter plant
Infection pressure based on distance to nearest site(s) with PD
a
Factors with P £ 0.2 in univariable analyses, but without significant association in multivariable analysis (P £ 0.05).
PD positive. The categorical predictor variables
smolt strain and number of well boat transfers were
re-categorized into a reduced number of possible
categories to enable statistical evaluation. One
continuous predictor variable (fallowing duration)
was converted to a dichotomous variable for the
same reason. Categorisation was performed based
on data distribution and biological meaningfulness.
The complete set of assessed predictor variables is
shown in Table 3. An initial univariable analysis,
using P £ 0.2 as criteria for further inclusion, was
performed. For the multivariable Cox proportional
hazards model, the inclusion criterion was
P £ 0.05.
Maps
The data set was joined together with a file of all
geographical coordinates for aquaculture sites in
Norway (available at http://fiskeridir.no/register/
akvareg) in Arc View 9.3 (ESRI) and maps showing
site locations and SAV/PD status generated.
Results
Evaluation of sample pooling
The Ct values obtained from pooled samples were
compared to those from the individual analyses of
the same sample. The effect of sample pooling
showed that individual samples with Ct values
below 35 (Ct range individual samples 30.1–35.0)
had a pooled sample viral RNA load that remained
within the detection limit of the real-time RT-PCR
2010
Blackwell Publishing Ltd
396
in all samples. For individual samples with a Ct
value above 35 (Ct range individual samples 35.5–
40.0), half of the samples (seven of 14) failed to give
a Ct value in 1/5 pools.
Freshwater sites
All samples from 44 freshwater sites were found
negative for SAV when analysed by real-time RTPCR using the described E1 primer. For the two
remaining sites, the initial analyses yielded inconclusive results on one and three pooled samples,
respectively. Following repeated analyses of all
individual samples, using both E1 and nsP1
primers, no SAV was detected. Blood samples and
tissue samples from four freshwater sites were
analysed. These freshwater sites supplied smolts to
four seawater sites with SAV detected or PD
diagnosed approximately 2 months post-seawater
transfer. These four freshwater sites were not the
same as those yielding inconclusive results on the
real-time RT-PCR analyses. No antibodies against
SAV were detected in the blood samples from these
four freshwater sites. Histopathology performed on
tissue samples from the same four freshwater sites
revealed no changes in accordance with PD.
Seawater sites
In total, SAV was detected by real-time RT-PCR in
samples collected from 23 of the 36 study sites
within the endemic region, equalling 63.9% [95%
confidence interval (CI) 46.2–78.7] of these sites
(Fig. 1). No SAV was detected in samples from the
2010
Blackwell Publishing Ltd
397
PD
Dx*
6/12
10/12
1/12, 1a
12/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
Real-time
RT-PCR
(positive/
examined)
0/45
5/30
0/30
6/30
–
–
–
0/30
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0/30
Serology
(positive/
examined)
Cohort samples 2 months
post-transfer
0/10
2/9
0/10
9/10
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Histopathology
(positive/
examined)
PD
PDb
Dx*
Dx*
7/18
9/9
0/12
12/12
No samples
12/12
No samples
8/8
12/12
0/12
0/12
No samples
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
0/12
–
No samples
0/30
–
No samples
–
No samples
No samples
–
–
–
No samples
–
–
No samples
–
–
–
–
–
–
No samples
0/15
0/10
No samples
0/10
1/10
No samples
–
No samples
No samples
–
–
–
No samples
–
–
No samples
–
–
–
–
–
–
No samples
–
6/6
39/40
No samples
6/6
5/6
6/6
5/6
3/6
4/12
No samples
4/4
PD
PD
PD
PD
PD
PDd
PD
No samples
3/6
6/6
No samples
11/12
No samples
5/6
4/6
0/5
4/6
PD
PD
PD
PD
PD
PD
PD
3/20
0/5
1/5
0/5
0/5
0/5
0/5
20/30
28/30
16/30
25/30
0/30
0/30
No sample
0/5
4/40
0/5
2/5
1/5
1/5
1/5
3/10
Histopathology
(positive/
examined)
23/30
35/40
23/30
30/30
21/30
22/30
11/30
53/60
Serology
(positive/
examined)
Real-time
RT-PCR
(positive/
examined)
Histopathology
(positive/
examined)
Real-time
RT-PCR
(positive/
examined)
Serology
(positive/
examined)
Cohort samples at slaughter
Cohort samples 8 months
post-transfer
PD
PD
PD
PD
PD
SAVc
PD
SAVc
SAVc
PD
PD
PD
PD
PD
PD
PD
PD
PD
PD
PD
PD
Final
site
status
PD, pancreas disease; Dx*, diagnosis of PD, diagnosed independently of the scheduled project samples by regular diagnostic procedures at the fish farm, with the timing shown relative to the cohort sample collections; –, analysis
not performed.
a
SAV detected in a single individual sample.
b
PD diagnosis at 8 months post-transfer.
c
SAV only detected, no diagnosis of PD.
d
PD diagnosis on slaughter samples.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Analysis/
site
number
Sampling point
Table 4 Laboratory results for 23 seawater sites found salmonid alphavirus (SAV)-positive in the cohort study in Norway 2006–2008
Journal of Fish Diseases 2010, 33, 391–402
M D Jansen et al. SAV and PD in Norway 2006–2008
M D Jansen et al. SAV and PD in Norway 2006–2008
Journal of Fish Diseases 2010, 33, 391–402
15 sites located outside the endemic region. The
proportion of SAV-positive pooled samples varied
between sites and sampling points. The cumulative
incidence of PD throughout the production cycle
for SAV-positive sites was 87% (95% CI 70.5–
98.5) or 20 of the 23 sites.
Based on the three predefined sampling points
and recorded diagnosis of PD, the first time point
for detection of SAV varied from 11 to 81 weeks
post-sea transfer (n = 23). In four sites (sites 1, 3, 8,
9, Table 4), SAV was detected in study samples
prior to a PD diagnosis. The elapsed time between
SAV detection in study samples and diagnosis of
PD ranged from 18 to 71 weeks. PD was diagnosed
from 18 to 84 weeks post-seawater transfer [mean
53.3 (16.7) n = 20]. Three sites (sites 18, 20, 21,
Table 4) that were SAV-positive without a subsequent diagnosis of PD had SAV detected on
samples collected at slaughter, 66, 71 and 81 weeks
post-sea transfer, respectively.
The four sites found SAV-positive by real-time
RT-PCR 2 months post-sea transfer (sites 1, 2, 3, 4,
Table 4) had their blood samples analysed for
specific antibodies against SAV, with two sites (sites
2, 4, Table 4) showing SAV neutralization. Two
additional sites with samples assessed (sites 8 and
23, Table 4) showed no neutralization. All slaughter blood samples were assessed, and all sites with
SAV detected by real-time RT-PCR, except two
(sites 20, 21, Table 4), had samples showing SAV
neutralization. One site with no SAV detected
throughout the study showed SAV neutralization in
a single sample at slaughter (data not shown).
From the four seawater sites found SAV-positive
2 months after seawater transfer, histopathological
examination revealed two sites with severe loss of
exocrine pancreatic tissue, cardiomyocytic necrosis
and/or heart inflammation and inflammation in red
skeletal muscle. Such changes were also detected in
all the following samplings, although in fewer
individuals (sites 2, 4, Table 4). For site four, it is
worth recording that although PD was diagnosed in
our study samples collected 2 months after seawater
transfer, the clinical outbreak of PD at the farm was
first discovered by the regular fish health service in
May 2007, 1 year after seawater transfer. This
contrasts with site two where PD was diagnosed
independently of the present study already prior to
our first scheduled sampling. In slaughter samples,
changes characteristic for PD were detected in
samples from a total of eight of 16 assessed sites
(Sites 1, 3, 4, 6, 10, 14, 17, 23, Table 4).
2010
Blackwell Publishing Ltd
398
1.00
SAV/PD negative sites
SAV/PD positive sites
0.75
0.50
0.25
2
0.00
0
20
1
40
112111
11 1
60
80
100
Weeks post seawater transfer
Figure 2 Kaplan–Meier failure graph showing the proportion of
seawater sites diagnosed with salmonid alphavirus (SAV) infection or pancreas disease (thick line with stepwise increase) versus
the time post-seawater transfer. The flat lower line shows the
censoring times and numbers censored (indicated by a hash and
the number) for the remaining, SAV-negative sites within the
endemic region.
Epidemiological observations
The gradual increase in the number of seawater sites
with positive SAV or PD status in the endemic area
is shown graphically in a Kaplan–Meier failure
graph (Fig. 2). Mortality levels following a PD
diagnosis varied greatly between populations. The
mean percentage PD-related mortality was 6.9%
(7.06) (range 0.7–26.9), while the mean duration
of increased mortality rates (from elevation of
mortality from background level until return to
background level) following PD diagnosis was
2.8 months (1.11) (range 1–6). Average fish
weight at time of PD diagnosis ranged from 461
to 5978 g.
Several potential risk factors were tested for
association to PD outbreak. Fallowing duration,
number of well boat transfers, cage type, distance to
nearest slaughter plant and previous PD at the site
all showed P £ 0.2 in univariable analyses, but did
not show any significant association to PD outbreaks in a multivariable model (P < 0.05)
(Table 3).
Discussion
Salmonid alphavirus was not detected in any of the
smolt tissue samples from 46 freshwater sites
throughout six counties along the Norwegian
coastline. Additional sample evaluation of antibody
presence and histopathological changes from four
selected freshwater sites (which supplied stock to
four seawater sites with SAV detected by real-time
RT-PCR approximately 2 months post-seawater
Journal of Fish Diseases 2010, 33, 391–402
transfer) was negative. This is in accordance with
results obtained from the testing of 1223 randomly
selected smolts from 15 different freshwater sites,
conducted by a commercial salmon farming company using real-time RT-PCR (analysed by PatoGen AS, Norway), with three samples found
positive for the vaccine strain SAV1 only (O.
Breck, personal communication). Throughout the
study period, no SAV was detected in samples from
study sites outside the endemic region. No seawater
sites were detected with SAV or PD in the nonendemic area, underlining the fact that the risk of
contracting SAV infection is far lower in this region
than in the endemic region where SAV was detected
in samples from 64% of the studied sites. All SAVpositive seawater sites stocked fish that had been
sampled in the freshwater phase, except one which
was diagnosed with PD on diagnostic samples at
slaughter (site 22, Table 4). Two sites found SAVpositive at the first sampling showed no detectable
seroconversion (sites 1, 3, Table 4) indicating a
recent infection, as antibodies against SAV can be
found from days 10–16 post-infection (McLoughlin, Nelson, Rowley, Cox & Grant 1996). These
results indicate no freshwater phase involvement in
the SAV and PD status of the studied seawater sites.
The observed time range of 18–84 weeks from
sea transfer until the development of clinical disease
was more prolonged than previously reported (9–
43 weeks) (Munro et al. 1984; Poppe et al. 1989;
Murphy et al. 1992; Crockford, Menzies,
McLoughlin, Wheatley & Goodall 1999;
McLoughlin et al. 2003; Brun et al. 2005; Seim
2007). For the early outbreaks, this may be because
of an improved effect of preventive measures, while
for the later outbreaks it was likely influenced by the
ability of this study to follow several sites throughout the entire production cycle, including sampling
and evaluation at slaughter. The gradually increasing number of studied seawater sites found SAVpositive with increasing time at sea, suggests the
infection occurs in seawater. In this study, no
association between the study sitesÕ SAV/PD status
and the measured risk factors were detected, despite
previous studies having found a varying degree of
association between PD and a range of risk factors
(Wheatley et al. 1995; McLoughlin et al. 2003;
Brun et al. 2005; Rodger & Mitchell 2007; Kristoffersen et al. 2009). The lack of association in this
study between site status and infection pressure,
contrary to the findings of Kristoffersen et al.
(2009), is likely related to the smaller sample size
2010
Blackwell Publishing Ltd
399
M D Jansen et al. SAV and PD in Norway 2006–2008
in this study, as the sites used in this study were
effectively a subset of those used by Kristoffersen
et al. (2009). The studied sites constitute only a
small proportion of the number of sites in operation
in Norway and diagnosed with PD during the
2006–2008 production cycle. This, together with
the possible lack of ability to measure explanatory
variables at a more detailed level, may explain the
lack of association. The observed mortality levels
and duration of elevated mortality were within the
ranges previously reported (Wheatley et al. 1995;
Crockford et al. 1999; McLoughlin et al. 2003;
Brun et al. 2005; Rodger & Mitchell 2007). The
large variation in average fish weights at time of PD
diagnosis was a result of the wide range in the
timing of disease occurrence throughout the production cycle.
All SAV-positive seawater sites developed PD,
with the exception of three sites which had SAV
detected by real-time RT-PCR at slaughter. For
these sites, no PD was suspected by the site
management. These results indicate that once the
population has become SAV infected, disease is
highly likely to develop if the fish are given
adequate time at sea post-infection. Although no
permanent subclinical status was seen, as reported
for a Scottish salmon population infected by SAV
(Graham et al. 2006), the prolonged time sometimes elapsing between SAV detection and observed
clinical disease is noteworthy. It may be related to
fish-, management- or site factors, or a combination
of these. It could also be attributed to a delay in the
diagnosis because of a late request of diagnostic
visits. This latent infection and the variation in
incubation time pose a real risk for transmission of
virus to uninfected populations. Such risk can be
reduced by following strict hygiene measures, for
example related to well boat transport and stock
movement, as well as through close monitoring of
sites within high-risk locations.
One blood sample of 90 from one seawater site
was found to neutralize SAV, while tissue samples
failed to give a Ct value on real-time RT-PCR
analysis, nor were any histopathological changes
consistent with PD detected. The blood plasma
from this single fish could have contained unspecific
viral inhibitor(s) against SAV as has previously been
shown to occur for infectious pancreatic necrosis
virus (IPNV) (Park & Reno 2005). This is,
however, assumed to be rare in salmonid blood,
following a lack of such findings for other sites and
absence of reports of this phenomenon from other
Journal of Fish Diseases 2010, 33, 391–402
serological studies of SAV (Graham et al. 2003;
Graham, Jewhurst, Rowley, McLoughlin, Rodger
& Todd 2005).
A prospective, longitudinal cohort study design
was chosen to allow the entire production cycle to be
studied, ensuring an accurate determination of
exposure status for both the case- and the control
sites. In this study, no difference was observed in the
SAV status at each sampling point between the study
group and the on-site control group, with the
exception of two sites (sites 1, 3, Table 4). The lack
of seroconversion at the initial sampling at these two
sites indicates recent infections, and likely accounts
for the difference in SAV status between the groups
at the same site. The overall finding, that multiple
cages are being affected within a short period of time,
corresponds to the results obtained from longitudinal studies of other sites in Norway (M.A.Wasmuth
& A.B. Olsen, personal communication).
Our data were compared to the NVI diagnostic
database and reports from each site with respect to
diagnosed or suspected PD. No sites recorded SAVnegative in this study had diagnosed or suspected
PD. With only three scheduled sampling points
throughout the seawater phase, the time of infection
could not be calculated more accurately. Each site
could have been infected at any time between the
last negative sample and the sample collection in
which SAV was first detected. However, the
inclusion of any results from diagnostic investigations for each study site increased the reliability of
the estimates, by taking into account farmer and
fish health personnel observations, and in some
cases adding an additional sample collection point.
The most evident effect of the pooling strategy
on the results of the real-time RT-PCR analyses was
observed in samples with the lowest RNA viral load.
The decreased sensitivity of 1/5 pooling of samples
does not affect the real-time RT-PCR result when
using samples with individual Ct values lower than
35. Our pooling strategy may have given false
negative results if individual samples with viral
RNA load close to the detection limit of the realtime RT-PCR were examined. In our pilot study,
half the samples with an individual Ct value above
35 had a viral RNA load that failed to be detected
when analysed in pools. This decrease in sensitivity
has been noted previously in other studies evaluating the effect of pooling sample material (Vandenbussche, Vanbinst, Vandemeulebroucke, Goris,
Sailleau, Zientara & De Clercq 2008). In this
study, any pooled samples with inconclusive results
2010
Blackwell Publishing Ltd
400
M D Jansen et al. SAV and PD in Norway 2006–2008
(giving Ct values ‡40 or an anomalous curve shape)
were re-analysed individually with two different
primers, in an attempt to determine their true SAV
status. After the onset of our study, others have
shown the heart ventricle to be more optimal for
SAV detection than mid-kidney (Andersen, Bratland, Hodneland & Nylund 2007). However, we
believe that the decreased sensitivity and risk of false
negative results because of our method has partly
been compensated for by the testing of some
samples individually, as well as the inclusion of a
large number of samples overall from each site.
The value of real-time RT-PCR technology for
SAV detection as a population screening tool has
been described previously (Graham et al. 2006),
and its consistency was further demonstrated in this
study. Only two sites had differing SAV status of
the sampled groups (one without SAV detected at
the second seawater sampling, the other with
differing cages being detected SAV-positive on
repeated samples), which may suggest the sample
size used was occasionally too small to detect the
agent. This study was designed with a sample size to
allow, with 95% probability, detection of SAV in a
population with an expected prevalence of the agent
of 5%. It is possible that the SAV prevalence at
these sites was below this level at the time of
sampling, as the sample submissions for these sites
were complete. The overall very good consistency
found between real-time RT-PCR, serology and
histopathology, as well as results from routine
diagnostic work, in this study indicates method
reliability, and possibly reduces the impact of
borderline sample sizes.
In our study, SAV was not detected in freshwater
or in seawater sites outside the region considered
endemic for PD. No evidence was found for
subclinical SAV infections without subsequent
development of PD. The pattern of an increasing
number of seawater sites with SAV detected and PD
diagnosed with increased time at sea, together with
the epidemiological observations, suggests that
infection most likely occur after seawater transfer
through horizontal transmission.
Acknowledgements
This work was funded by The Norwegian Research
Council, The Fishery and Aquaculture Industry
Research Fund and Marine Harvest Norway AS.
We are very grateful to fish farmers, fish health
services and the Norwegian Food Safety Authority
Journal of Fish Diseases 2010, 33, 391–402
for sampling assistance. Thanks to I. Modahl, M.
Heum and R. Terland for technical assistance and
to other colleagues for helpful scientific discussions.
Kristoffersen A.B., Viljugrein H., Kongtorp R.T., Brun E. &
Jansen P.A. (2009) Risk factors associated with pancreas disease (PD) outbreaks in farmed Atlantic salmon and rainbow
trout in Norway during 2003–2007. Preventive Veterinary
Medicine 90, 127–135.
References
McLoughlin M.F., Nelson R.N., Rowley H.M., Cox D.I. & Grant
A.N. (1996) Experimental pancreas disease in Atlantic salmon
Salmo salar post-smolts induced by salmon pancreas disease
virus (SPDV). Diseases of Aquatic Organisms 26, 117–124.
Andersen L., Bratland A., Hodneland K. & Nylund A. (2007)
Tissue tropism of salmonid alphavirus (subtypes SAV1 and
SAV3) in experimentally challenged Atlantic salmon (Salmo
salar L.). Archives of Virology 152, 1871–1883.
Brun E., Olsen A.-B. & Rørvik L. (2005) Factors associated with
outbreaks of pancreas disease in farmed Atlantic salmon. Abstract O-147, 12th International Conference of the European
Association of Fish Pathologists, September 2005.
Christie K.E., Fyrand K., Holtet L. & Rowley H.M. (1998)
Isolation of pancreas disease virus from farmed Atlantic salmon, Salmo salar L., in Norway. Journal of Fish Diseases 21,
391–394.
Crockford T., Menzies F.D., McLoughlin M.F., Wheatley S.B.
& Goodall E.A. (1999) Aspects of the epizootiology of pancreas disease in farmed Atlantic salmon Salmo salar in Ireland.
Diseases of Aquatic Organisms 36, 113–119.
Fringuelli E., Rowley H.M., Wilson J.C., Hunter R., Rodger H.
& Graham D.A. (2008) Phylogenetic analyses and molecular
epidemiology of European salmonid alphaviruses (SAV) based
on partial E2 and nsP3 gene nucleotide sequence. Journal of
Fish Diseases 31, 811–823.
Graham D.A., Jewhurst V.A., Rowley H.M., McLoughlin M.F.
& Todd D. (2003) A rapid immunoperoxidase-based virus
neutralization assay for salmonid alphavirus used for a serological survey in Northern Ireland. Journal of Fish Diseases 26,
407–413.
Graham D.A., Jewhurst V.A., Rowley H.M., McLoughlin M.F.,
Rodger H. & Todd D. (2005) Longitudinal serological surveys of Atlantic salmon, Salmo salar L., using a rapid immunoperoxidase-based neutralization assay for salmonid
alphavirus. Journal of Fish Diseases 28, 373–379.
Graham D.A., Jewhurst H., McLoughlin M.F., Sourd P.,
Rowley H.M., Taylor C. & Todd D. (2006) Sub-clinical
infection of farmed Atlantic salmon Salmo salar with salmonid
alphavirus - a prospective longitudinal study. Diseases of
Aquatic Organisms 72, 193–199.
Hodneland K. & Endresen C. (2006) Sensitive and specific
detection of salmonid alphavirus using -time PCR (TaqMan).
Journal of Virological Methods 131, 184–192.
Hodneland K., Bratland A., Christie K.E., Endresen C. &
Nylund A. (2005) New subtype of salmonid alphavirus (SAV),
Togaviridae, from Atlantic salmon Salmo salar and rainbow
trout Oncorhynchus mykiss in Norway. Diseases of Aquatic
Organisms 66, 113–120.
Karlsen M., Hodneland K., Endresen C. & Nylund A. (2006)
Genetic stability within the Norwegian subtype of salmonid
alphavirus (family Togaviridae). Archives of Virology 151, 861–
874.
Kent M.L. & Elston R.A. (1987) Pancreas disease in pen-reared
Atlantic salmon in North America. Bulletin of the European
Association of Fish Pathologists 7, 29–31.
2010
Blackwell Publishing Ltd
M D Jansen et al. SAV and PD in Norway 2006–2008
401
McLoughlin M.F., Rowley H.M. & Doherty M.D. (1998) A
serological survey of salmon pancreas disease virus (SPDV)
antibodies in farmed Atlantic salmon, Salmo salar L. Journal of
Fish Diseases 21, 305–307.
McLoughlin M.F., Peeler E., Foyle K.L., Rodger H.D.,
OÕCeallachain D. & Geoghegan F. (2003) An epidemiological
investigation of the re-emergence of pancreas disease in Irish
farmed Atlantic salmon (Salmo salar L.) in 2002. Marine
Environment and Health Series, No. 14. Marine Institute,
Galway, Ireland.
McLoughlin M.F., Graham D.A., Norris A., Matthews D., Foyle
L., Rowley H.M., Jewhurst H., MacPhee J. & Todd D.
(2006) Virological, serological and histopathological evaluation of fish strain susceptibility to experimental infection with
salmonid alphavirus. Diseases of Aquatic Organisms 72, 133.
McVicar A.H. (1987) Pancreas disease of farmed Atlantic salmon, Salmo salar, in Scotland; epidemiology and early
pathology. Aquaculture 67, 71–78.
Munro A.L.S., Ellis A.E., McVicar A.H., McLay H.A. &
Needham E.A. (1984) An exocrine pancreas disease of farmed
Atlantic salmon in Scotland. Helgolander Meeresuntersuchungen 37, 571–586.
Murphy T.M., Rodger H.D., Drinan E.M., Gannon F., Kruse P.
& Korting W. (1992) The sequential pathology of pancreas
disease in Atlantic salmon farms in Ireland. Journal of Fish
Diseases 15, 401–408.
Nelson R.T., McLoughlin M.F., Rowley H.M., Platten H.M. &
McCormick J.I. (1995) Isolation of a toga-like virus from
farmed Atlantic salmon Salmo salar with pancreas disease.
Diseases of Aquatic Organisms 22, 25–32.
Olsen A.B. & Wangel C. (1997) An outbreak of PD-like disease
in rainbow trout (Oncorhynchus mykiss) in Norway. European
Association of Fish Pathologists 8th International Conference
‘‘Diseases of Fish and Shellfish’’, Edinburgh, Poster-062.
Park K.C. & Reno P.W. (2005) Molecular size, pH, temperature
stability and ontogeny of inhibitor(s) of infectious pancreatic
necrosis virus (IPNV) in normal rainbow trout serum. Journal
of Aquatic Animal Health 17, 177–190.
Poppe T., Rimstad E. & Hyllseth B. (1989) Pancreas disease of
Atlantic salmon (Salmo salar L.) post-smolts infected with
infectious pancreatic necrosis virus (IPNV). Bulletin of the
European Association of Fish Pathologists 9, 83–85.
Raynard R., Houghton G. & Munro A.L.S. (1992) Pancreas
disease of Atlantic salmon: proceedings of a European Commission workshop. Scottish Office Aquaculture Report 1, 2–4.
Rodger H. & Mitchell S. (2007) Epidemiological observations of
pancreas disease of farmed Atlantic salmon, Salmo salar L., in
Ireland. Journal of Fish Diseases 30, 157–167.
Journal of Fish Diseases 2010, 33, 391–402
Ruane N., Graham D. & Rodger H. (2008) Site investigations
and disease management of the pancreas disease virus in Irish
farmed salmon. Marine Environment and Health Series,
No. 32. Marine Institute, Galway, Ireland.
Seim R.R. (2007) Risk Factors and Distributions of Pancreas
Disease (PD). An Epidemiological Study of PD Outbreaks in
Norway from the Years 2002 to 2005. Master thesis, University
of Bergen, Norway.
Taksdal T., Olsen A.B., Bjerkås I., Hjortaas M.J., Dannevig
B.H., Graham D.A. & McLoughlin M.F. (2007) Pancreas
disease in farmed Atlantic salmon, Salmo salar L., and rainbow
trout, Oncorhynchus mykiss (Walbaum), in Norway. Journal of
Fish Diseases 30, 545–558.
Vandenbussche F., Vanbinst T., Vandemeulebroucke E., Goris
N., Sailleau C., Zientara S. & De Clercq K. (2008) Effect of
pooling and multiplexing on the detection of bluetongue virus
RNA by real-time RT-PCR. Journal of Virological Methods
152, 13–17.
Weston J.H., Welsh M.D., McLoughlin M.F. & Todd D.
(1999) Salmon pancreas disease virus, an alphavirus infecting
farmed Atlantic salmon, Salmo salar L. Virology 256, 188–195.
2010
Blackwell Publishing Ltd
402
M D Jansen et al. SAV and PD in Norway 2006–2008
Weston J., Villoing S., Bremont M., Castric J., Pfeffer M.,
Jewhurst V., McLoughlin M., Rodseth O., Christie K.E.,
Koumans J. & Todd D. (2002) Comparison of two aquatic
alphaviruses, salmon pancreas disease virus and sleeping disease virus, by using genome sequence analysis, monoclonal
reactivity, and cross-infection. Journal of Virology 76, 6155–
6163.
Weston J.H., Graham D.A., Branson E., Rowley H.M., Walker
I.W., Jewhurst V.A., Jewhurst H.L. & Todd D. (2005)
Nucleotide sequence variation in salmonid alphaviruses from
outbreaks of salmon pancreas disease and sleeping disease.
Diseases of Aquatic Organisms 66, 105–111.
Wheatley S.B., McLoughlin M.F., Menzies F.D. & Goodall E.A.
(1995) Site management factors influencing mortality rates in
Atlantic salmon (Salmo salar L.) during marine production.
Aquaculture 136, 195–207.
Received: 10 July 2009
Revision received: 16 September 2009
Accepted: 24 September 2009