Feline coronaviral polyserositis, as it should be termed, is the fatal
'tip of the iceberg' to a common feline infection with a group of ubiquitous
viruses. Most of these coronaviruses are harmless and perfectly adapted
to growth in the gut. They have been named 'feline enteric coronaviruses',
to distinguish them from the killer viruses that replicate in the feline
macrophage. Persistently infected, healthy cats play the most important
epidemiologic role in FIP, because by harbouring feline coronaviruses
(FCoVs) in their intestines and blood, they act as a constant source
of infection. The virus is shed in the faeces, saliva and perhaps other
body fluids of infected cats. In addition to these 'pathotypes', coronaviruses
also occur in two serotypes, both of which can cause FIP after having
undergone subtle genetic changes. It is only for reasons of convenience
that we shall continue to use the term FIP virus (FIPV) - to denominate
those FCoV strains that carry the mutation(s) responsible for the increase
in virulence. The legitimacy of such a nomenclature is questionable,
however; it is like giving different names to a virus and its attenuated
vaccine strain.
Fig. 4 Genomic organization of and recombination
between carnivore coronaviruses. The boxes represent the genes responsible
for the 'structural' proteins building the virus particle, such as the
'peplomers' (see legend to Fig. 2) or spikes (S), two membrane proteins
(E and M) and the nucleocapsid protein (N), which wraps the genome.
POL stands for polymerase, and also the genes 3 and 7 code for non-structural
proteins; mutations in the 3c gene have been found in coronavirus infected
cats that developed FIP. The uppermost graph symbolizes the genome of
a canine coronavirus, the lowest that of a feline coronavirus. Recombinants
between both have been found in the field, and the varying cross-over
sites indicate that this event occurs regularly.
The FCoV carrier state
Epidemiological studies suggest that FCoVs may cause persistent infections,
that a carrier state exists, and that many infections are not cleared
by the cat's immune system - or only after a long time. It is common
knowledge that healthy cats with coronavirus antibodies may cause seroconversion
in contact animals within 2 to 10 weeks. The infection is spread presumably
via the fecal-oral route, and some of the contact animals will subsequently
succumb to FIP [1]. First solid evidence
for a carrier state came from an experiment in which cats were infected
with a sublethal dose of tissue culture grown FIPV and kept in isolation.
To induce FIP, the cats were superinfected with feline leukaemia virus,
which is known to be strongly T-cell immunosuppressive. From this work
it appeared that FIPV could persist in the experimentally infected host
for at least 4 months [11].
To identify asymptomatic FCoV carriers and to monitor virus shedding,
a group of scientists from Utrecht developed a nested RT-PCR assay targeted
to a highly conserved (invariant) region of the FCoV genome. Using this
assay, viral RNA had previously been detected in the faeces, tissues
and body fluids of cats with FIP [2,9].
Interestingly, however, FCoV RNA was also found in the faeces, and occasionally
in the serum, of perfectly healthy cats. The virus persistence and evolution
was then studied in a closed cat breeding facility in Hanover/Germany
with an endemic FCoV type I infection. Viral RNA was detected in the
faeces and/or plasma of 85% of the cats tested. Of five cats identified
as FCoV shedders during the initial survey, four had viral RNA in their
faeces when tested almost four months later. This could be due either
to repeated reinfections or to persistence of the virus in the cat's
organism. To distinguish between these possibilities, two cats were
placed in perfect isolation, and fecal virus shedding was monitored
every 2-4 days. In one cat, shedding continued for up to seven months.
The other animal was sacrificed after three months of continuous shedding
since we wanted to find the sites of viral replication. Viral genomic
RNA was found in almost any tissue tested, but messenger RNA (which
is synthesized only when a virus multiplies) was detected exclusively
in the ileum, colon and rectum; in these parts of the gut individual
FCoV-infected cells (i.e. not large parts of infected tissue) were also
spotted by immunohistochemistry. These findings provide first formal
evidence that FCoV causes chronic infections [10].
Fig. 5 The difference between recurrent and
persistent infections. Viruses may be maintained in a population either
by animal-to-animal transfer or by prolonged presence in an individual's
body. As we learn more about viruses, the persistent infection appears
to be the rule rather than the exception - and it occurs also in feline
coronaviruses.
Recently, a modification of the conventional RT-PCR procedure was
introduced - the so-called TaqMan technique. It allows testing of
many samples within a short period of time, and in addition permits
the reliable quantitation of FCoV genomes in a sample, e.g. in rectal
swabs taken from individual cats [7]
(for details of the TaqMan technique, look into the Tools section
of this VetSciTe issue). Using this technique, the coronavirus shedding
pattern was determined over a period of 24 weeks in 77 cats kept in
multi-cat household situations. We found a highly significant correlation
between the amount of FCoV shed in feaces and the frequency of shedding
[13]
Fig. 6 Correlation between shedding frequency
and shedding intensity in 77 cats followed for 24 weeks. It becomes
clear that with higher shedding frequency the amount shed is significantly
increased (r=0.9895, p<0.0001). The amount of FcoV shed in 1 g
faeces with a Ct value of 30 equals 10 exp.7 viral particles.
The shedding frequency in these cats was found to be low (<30%
of all samples positive) in 78%, intermediate (between 30% and 90%
of all samples positive) in 21% and high (>90% of all samples positive)
in 1% of the cats.
To study viral evolution during chronic infection, the FCoVs sampled
from individual cats were characterized. Phylogenetic comparisons
of sequences obtained for independent European and American isolates
indicated that the viruses in the breeding facility form a clade (a
closely related cluster) and are likely to have originated from a
single founder infection. Each cat harboured distinct FCoV quasispecies
with immune selection (antigenic drift) occurring during chronic infection.
These data support a model in which chronic carriers maintain the endemic
infections in cat societies. Virtually every kitten born in a breeding
facility becomes infected, probably from its queen[1],
as soon as its maternal protection wanes. Once infected, the cats appear
to resist superinfection by closely related FCoVs, every cat carrying
its private, harmless clan of variants.
Fig.7 Unrooted 'phylogenetic trees' showing genetic relationships
between feline coronaviruses, where branch length indicates evolutionary
distance. Graph A illustrates the relationship between coronaviruses
shed by cats in a closed breeding facility in Hanover, Germany (isolates
indicated by H) as compared with laboratory strains; field strains
from the Netherlands (C, Dahlberg) and the United States (RM) are
included. The non-structural gene 7b was used in this comparison.
Panel B shows a more detailed analysis - of the structural S gene
- with in the Hanover cattery alone. Clusters of more closely related
viruses can be distinguished, most of them from littermates. Amino
acid substitutions were not random but are linked to predicted epitopes,
indicating antigenic drift.
At the 1997 WSAVA/BSAVA Congress in Birmingham, workers from Bristol
University first presented data that confirmed this epidemiologic concept
using a different approach. These workers were indeed able to culture
FCoVs from the blood of healthy cats from seropositive catteries. Blood
samples obtained from healthy cats of nine different breeds from nine
separate catteries were examined, and growth of FCoV, demonstrable by
PCR, was obtained in most cases, some of which were FCoV antibody-negative.
The conclusion reached was again: most healthy cats living in catteries
with a past history of FIP are persistently infected with FcoVs [6].
The important finding of this biologically meaningful analysis was that
the isolated viruses were of the 'non-cultivable' subtype I.
From the FCoV carrier state to FIP
What leads from infection to disease, from the chronic FCoV carrier
state to FIP? This question will be asked by anybody who hears the classical
anamnesis of a kitten obtained from a bona-fide breeder, kept in isolation
from other cats that succumbed to FIP some weeks after the purchase.
The key pathogenic event in FIP is the infection of monocytes and macrophages.
We formerly thought that avirulent FCoV strains remain confined to the
digestive tract and would not spread beyond the intestinal epithelium
and regional lymph nodes while virulent strains would disseminate to
other organs via blood-borne monocytes. This idea can no longer be sustained,
in view of the PCR results in healthy cats quoted above - the difference
must rather be a quantitative one. In vitro, the virulence of FCoV strains
was indeed correlated with their ability to infect cultured peritoneal
macrophages. When strains were compared, however, the avirulent ones
infected fewer macrophages and produced lower virus titres than virulent
strains. Moreover, the avirulent strains were less able to sustain viral
replication and to spread to other macrophages. This is no black-and-white
phenomenon, rather a gradual transition, as the course of FIP is not
uniform.
There is ample evidence for an involvement of the immune system in the
pathogenesis of FIP. Humoral immunity is obviously not protective. FCoV-seropositive
cats that are experimentally infected with FIPV often develop an accelerated,
fulminating course of the disease, leading to the 'early death' phenomenon
mentioned above. Clinical signs and lesions develop earlier, and the
mean survival time is dramatically reduced as compared to seronegative
cats. Direct evidence for the involvement of antibodies was obtained
by transfusion of purified IgG from cat FCoV-antisera into cats, which
indeed developed accelerated FIP upon experimental challenge. We also
know, which antibodies are the killers: when vaccinia virus recombinants
expressing singe gene products were used to immunize cats, 'early death'
occurred only in the group that had seen the spike (S) protein before.
Fig.8 The 'early death' phenomenon seen in
cats that had been immunized with the S protein of feline coronavirus
expressed by a vaccinia virus recombinant; the control animal had
been vaccinated with the 'empty' vaccinia virus vector. Upon challenge
with an FIP-producing coronavirus, these cats show the typical biphasic
temperature rise and fatal course; the disease took more than two
weeks, whereas the animals with antibodies to S succumbed within a
week.
Most authors consider the vascular and perivascular lesions in FIP
to be immune-mediated, but there is uncertainty about the actual pathogenetic
mechanism. At least some vascular injury may be attributed to immune-mediated
lysis of infected cells: FIPV-infected white blood cells were detected
in the lumen, intima and wall of veins and in perivascular locations.
Furthermore, inflammatory mediators such as cytokines, leukotrienes
and prostaglandins that are released by infected macrophages could
play a role in the development of the perivascular pyogranulomata.
These products could induce vascular permeability changes and provide
additional chemotactic stimuli for neutrophils and monocytes. In response
to the inflammation, the attracted cells may release additional mediators
and cytotoxic substances; the monocytes would also serve as new targets
for FIPV. The end result would be enhanced local virus production
and increased tissue damage.
Other observations point towards an immune complex (ICX) pathogenesis.
Deposition of ICX and subsequent complement activation is thought
to cause an intense inflammatory response that may extend across blood
vessel walls. The resulting vascular damage would permit leakage of
fluid into the intercellular space and eventually lead to the accumulation
of thoracic and abdominal exudate. The morphologic features of the
vascular lesions (necrosis, polymorphonuclear cell infiltration associated
with small veins and venules) strongly indicate an Arthus type reaction.
The lesions contain focal deposits of virus, IgG and C3. Moreover,
complement depletion and circulating ICX were demonstrated in cats
with terminal FIP. In a horizontal study of experimentally infected
cats, first clinical signs were accompanied by increased C3 concentrations
in the plasma; subsequently antibody titres and circulating ICX increased
with a concomitant decrease of complement concentrations. At the time
of death, maximum ICX and minimum C3 concentrations were measured.
Fig.9 Relationship between the clinical course
of FIP, as represented by body temperature (upper curve), complement
levels (middle), and concentrations of antibodies and immune complexes
(lower curves) in the plasma. While the latter two parameters rise
steadily, C3 levels are plummeting before death, indicating complement
activation.
Although FIP viruses do not infect T-cells, depletion and programmed
cell death (apoptosis) was observed in lymphoid organs of infected cats.
Apoptosis was mediated by the ICX present in the serum and ascitic fluid
of diseased cats and affected only activated T-cells, including lymph
node cells, but not unstimulated T-cells. This hitherto unrecognised
mechanism of T-cell suppression may operate not only in FIPV infection
but also in other ICX diseases [5].
The fatal scenario thus may be as follows: a kitten is born, suckled
by its seropositive queen and protected by colostral antibody from infection
during the first few weeks. As the maternal antibodies wane, mucosal
protection ebbs away and during an episode of maternal FCoV shedding
the kitten is infected. A bout of diarrhoea and an occasional sneezing
may be the only signs this has happened. It now develops an active immunity,
but not a sterilizing one in most cases: virus and antibodies continue
to co-exist in the kitten's organism, and an efficient cell-mediated
immunity keeps infected macrophages and monocytes in check. In a small,
socially stable cat community this animal can live happily ever after.
Problems emerge when our kitten is experiencing any situation of stress,
which we want to equate with immune suppression. Infection with the
feline leukaemia or immunodeficiency viruses would be the most unmistakable
immunosuppressive event, but density (numbers of cats per surface unit),
geographic change (displacement into a new environment) and other territorial
factors (e.g. change in group hierarchy, dominance) are becoming more
and more important - in view of the declining prevalence of retrovirus
infections. The failing immune surveillance allows the coronaviral quasispecies
cloud of mutants to expand, and more macrophage-tropic mutants emerge
in this stochastic process. Amongst them are some that reach high titres
and outcrowd the moderate ones. This is the point when immune pathogenesis
starts.
Laboratory tests - are they useful?
The diagnosis of FIP has been covered in many textbooks and articles;
in our view, the algorithm developed in the Zurich laboratory is still
the best clinical guideline [10].
Although it includes titres in the decision tree, these constitute
only a minor factor. A negative serologic result would not invalidate
the diagnosis based on clinical and blood chemistry data.
Fig. 10 Algorithm used to facilitate the diagnosis
of FIP
In the absence of clinical signs, serology is of no use for the prognosis
in individual cats. A statistical correlation indeed exists between
antibody titres and post-mortem confirmation of FIP three months after
testing. However, about 40% of the animals with titres of <300 do develop
FIP, and of those with titres exceeding 1000 only about one half succumb;
in other words: about half of the tested animals that remained healthy
showed the same high titre values as the cats at risk. Tossing a coin
would have given a similar result.
How does the other face of that coin look? About 12% of the cats with
titres <100 still developed FIP in the observation period. Based
on these data, one in eight owners would have been sent home with the
erroneous information that nothing will happen to his cat. Serology
does not distinguish between harmless and FIP-inducing FCoV mutants,
it only shows past - and in many cases still ongoing - infection. Any
seropositive cat may succumb to FIP, irrespective of the titre. We now
can explain this statistical correlation between high titers and poor
prognosis: expansion of the coronavirus quasispecies cloud obviously
not only provides much genomic material with the increased probability
for FIP-inducing mutants to occur, it also provides the large antigenic
mass to induce high levels of antibody. However, FIP-inducing mutants
can always occur, also at low replication levels (with low antibody
production), though with a lesser likelihood.
On the other hand: an uninfected cat - which is not synonymous with
a seronegative cat - will not develop into a case of FIP. This may sound
as a truism, nevertheless some consideration is justified. In the Hanover
cattery study, a few seronegative kittens shed FCoV in both faeces and
plasma, some in faeces only and others in plasma only. In concreto:
86% positive animals were detected using PCR, while serology showed
only 71% positives. However, not a single seropositive animal tested
PCR-negative [10].
The question must be asked whether there is a place for diagnostic
and prognostic laboratory testing for FIP at all. There are presently
no diagnostic assays available - neither in-practice tests nor assays
performed in the research laboratory - that would distinguish between
virulent (FIPV) and avirulent FCoV variants. Also the 'novel' PCR formats
touted by some firms do not keep this promise, irrespective of the claims.
We have reasons to believe that discriminatory assays based on the molecular
properties of the variants will not be feasible, perhaps not even possible.
However, there is a future for tests based on the evidence of immunological
changes in an animal developing FIP.
Both serology and PCR are able to detect infected cats, with different
sensitivity, and are invaluable for the management of catteries. They
can be used for monitoring the success of the quarantine and 'early
weaning' programmes, for controlling the specified pathogen-free and
the coronavirus-free status of catteries. Especially PCR could be useful
for monitoring individual animals to be introduced into FCoV-free catteries.
A promising approach to controlling FIP - based on isolation of litters
after early weaning - has been developed by the Glasgow group [1]
- but it is laborious, requires the dedicated cooperation of cat owners
and has no veterinary appeal. Incidentally, other studies performed
under similar conditions showed only marginal effects [7].
Another possibility is the removal of strong shedders from a multi-cat
society. These can now be recognized by using the TaqMan technique:
for a reliable characterization of the shedding pattern it is sufficient
to test four feaces samples taken at weekly intervals. Strong shedders
can be identified under field conditions and separated from the group,
thereby decreasing infection pressure for the remaining cats. It remains
to be shown whether this approach will work. However, common sense suggests
that in conjunction with other measures (keeping cats in small groups,
without contact between groups, frequent cleaning of litter boxes, introduction
of new cats only after quarantine and PCR testing etc.) the elimination
of strong shedders might be useful.
The seronegative catteries established through any control programme
must of course be protected against re-introduction, and the live temperature-sensitive
vaccine could prove useful for this purpose - if it indeed did not induce
antibodies, thereby compromising a serology-based test-and-isolation
programme. Also, persistence and recrudescence of the vaccine virus
might then be studied. Still much must be learned about this most enigmatic
infectious condition in veterinary medicine, feline infectious peritonitis.
References
1. Addie, D. D. and Jarrett, J. O. (1992). A
study of naturally occurring feline coronavirus infections in kittens.
Vet. Rec. 130, 133-137.
2. Egberink, H. F., Herrewegh, A. P. M., Schuurman,
N. M. P., van der Linde-Sipman, J. S., Horzinek, M. C., and de Groot,
R. J. (1995). FIP,
easy to diagnose? Vet. Quart. 17, 24-25.
3. Eigen, M. (1993) Viral
quasispecies. Scientific American, July ;269(1):42-49
4. Fehr D., Holznagel E., Bolla S., Hauser B., Herrewegh
A.A., Horzinek M.C., Lutz H. (1997) Placebo-controlled
evaluation of a modified life virus vaccine against feline infectious
peritonitis: safety and efficacy under field conditions. Vaccine. 15:1101-1109.
5. Groot, R.J. de and Horzinek, M.C. (1995) Feline
infectious peritonitis. In: The Coronaviridae (Siddell, S.G. ed.)
pp. 293-309. Plenum Press, New York.
6. Gunn-Moore D.A., Gruffydd-Jones T.J., Harbour D.A.
(1998)
Detection of feline coronaviruses by culture and reverse transcriptase-polymerase
chain reaction of blood samples from healthy cats and cats with clinical
feline infectious peritonitis. Vet. Microbiol. 62:193-205.
7. Gut M., Leutenegger C, Huder J., Pedersen N., Lutz
H. (1999) One-tube
fluorogenic reverse transcription-polymerase chain reaction for the
quantitation of feline coronavirus. J. Virol. Methods 77, 37-46
8. Haagmans, B.L., Egberink, H., Schijns V.E.C.J. and
Horzinek, M.C. (1996)
Apoptosis and T-cell depletion during feline infectious peritonitis.
J.Virol. 70, 789-793.
9. Herrewegh, A.A.P.M., de Groot, R.J., Cepica, A.,
Egberink, H.F., Horzinek, M.C., and Rottier, P.J.M. (1995) Detection
of feline coronavirus RNA in feces, tissue, and body fluids of naturally
infected cats by reverse transcriptase PCR. J.Clin.Microbiol. 33, 684-689.
10. Herrewegh, A.A.P.M., Mähler M., Hedrich, H.J.,
Haagmans, B.L. Egberink, H., Horzinek, M.C., Rottier, P.J.M. and de
Groot, R.J. (1997). Persistence
and evolution of feline coronavirus in a closed cat breeding colony.
Virology 234, 349-363
11. Pedersen, N. C. (1987). Virologic
and immunologic aspects of feline infectious peritonitis virus infection.
Adv. Exp. Med. Biol. 218, 529-550.
12. Poland, A.M., Vennema, H., Foley, J.E. and Pedersen,
N.C. (1996) Two
related strains of feline peritonitis virus (FIPV) isolated from immunocompromised
cats infected with a feline enteric coronavirus (FECV). J.Clin.Microbiol.
35, 258-262.
13. Rohner. M. (1999) Bestimmung der Ausscheidungskinetik
von felinen Coronaviren unter Feldbedingungen. Inaugural-Dissertation
Vet.Med., Zürich (1999).
14. Rohrer, C., Suter, P.F. and Lutz, H. (1993) Die
Diagnostik der felinen infektiösen Peritonitis (FIP): retrospektive
und prospektive Untersuchungen. Kleintierpraxis 6, 379-381.
15. Vennema, H., Poland, A., Floyd Hawkins, K., and
Pedersen, N. C. (1995) A comparison of the genomes of FECVs and FIPVs
and what they tell us about the relationships between feline coronaviruses
and their evolution. Feline Pract. 23, 40-44.
© 2000 Veterinary Sciences Tomorrow
Feline
infectious peritonitis (FIP) is an important disease to the cat clinician
for several reasons; it is fatal in most (clinical) cases, its biology
is poorly understood and prevention is difficult, to say the least.
It is also an enigmatic disease: a sporadic viral condition is a contradiction
in terms. One expects epidemic expansion from virus diseases, or at
least a consistent pattern of spread in a cat society. Another mystifying
trait is the fact that antibodies - the molecules we associate with
immunity and protection - have no beneficial effect for the cat. Indeed,
under certain circumstances they may even precipitate disease, causing
the 'early death' phenomenon. The detection of antibodies is of no benefit
to the consulting veterinarian either, since titres are meaningless
for the purposes of diagnosis and prognosis in individual patients.
Nevertheless, there is a place for serology, as we will discuss later.
There is also a vaccine available in some countries, which has been
shown to provide some protection [4].
However, its efficacy is a matter of debate.
