A review of feline infectious peritonitis virus infection: 1963 …...ton.1,15 Therefore, its existence as a signiﬁcant disease entity prior to this time is questionable. A reference

Journal of Feline Medicine and Surgery (2009) 11, 225e258doi:10.1016/j.jfms.2008.09.008

REVIEW ARTICLEA review of feline infectious peritonitis virus infection:1963e2008

Niels C Pedersen DVM, PhD*

Department of Medicine andEpidemiology and Center forCompanion Animal Health, Schoolof Veterinary Medicine, Universityof California, Davis, CA 95616,USA
Date accepted: 3 September 2008 � 2009 Published by Elsevier Ltd on behalf of ESFM and AAFP.
Historical

Feline infectious peritonitis (FIP) was first de-scribed as an ‘important disorder of cats’ byHolzworth in 19631 at the Angell Memorial An-

imal Hospital, Boston and a clinico-pathologic confer-ence on this disorder was published in the followingyear.2 The disease was thought to be infectious butno specific etiologic agent was identified at the time.Wolfe and Griesemer3 were the first to propose thatFIP was caused by a virus. Zook et al4 observed virusparticles in the tissues of experimentally infected cats,but were unable to characterize the agent. Ward5 rec-ognized the close similarities of FIP virus (FIPV) in tis-sues to members of the family Coronaviridae. In 1972Montali and Strandberg6 were the first to report thatFIPV infection could be either granulomatous (dry,parenchymatous) or effusive (wet, non-parenchyma-tous). The close genetic relationship of FIPV to corona-viruses of dogs and swine was first reported byPedersen et al in 1978.7 Fully virulent FIPV was firstpropagated in vitro in autochthonous macrophagecultures from experimentally infected cats8 and laterin tissue culture.9,10 It was also replicated in the epi-thelium of intestinal ring cultures.11 A strain of FIPV(FIPV-UCD1) was first propagated in continuouslypasssaged Felis catus, whole fetus-4 (Fcwf-4) cellsand shown to be virulent when inoculated intocats.12 The Fcwf-4 cells were later found to be of

*Corresponding author. Present address: Center for CompanionAnimal Health, Room 213, CCAH Building, School of VeterinaryMedicine, University of California, Davis, CA 95616, USA.Te l : þ1 - 5 3 0 - 7 5 2 - 7 4 0 2 ; F a x : þ1 - 7 5 2 - 7 7 0 1 . E - m a i l :[email protected]

1098-612X/09/040225+34 $34.00/0 � 2009

a macrophage type.13 The existence of two serotypes,feline coronavirus (FCoV)-like and canine coronavirus(CCV)-like, of feline coronaviruses was first reportedin 1984.14

The sudden appearance of FIP in the late 1950’s wasdocumented by long-term and meticulous necropsyrecords at the Angell Memorial Animal Hospital, Bos-ton.1,15 Therefore, its existence as a significant diseaseentity prior to this time is questionable. A reference toa cat with a disease resembling FIP was publisheda half century earlier (Jakob16 as quoted in de Grootand Horzinek17), but whether this was FIP is uncer-tain given the absence of reports of a similar conditionin the intervening decades. There was a steady in-crease in the incidence of the disease in the 1960s on-ward, and it is currently one of the leading infectiouscauses of death among young cats from shelters andcatteries. The reason for the sudden emergence ofFIP is not known, but there are at least three possibleexplanations. First, coronaviruses may have speciatedinto cats within the last half century. It is noteworthythat FIP appeared within a decade of the initial de-scriptions of transmissible gastroenteritis (TGE) ofpigs in North America.18 The causative virus of FIPis closely related to the TGE virus of pigs and CCV,7

although they are still genetically distinguishable.19

However, recombinants between these three virusesare known to occur.14,20e22 At least one strain ofCCV can induce mild enteritis in cats and enhancea subsequent infection with FIPV, indicating a specialcloseness to feline coronaviruses.23 Therefore, CCVmay be a more likely parent of FECV in this scenario.Recombination events are favored by the ease withwhich transcription units (RNAs) can be gained orlost during the divergent evolution of coronaviruses.24

Published by Elsevier Ltd on behalf of ESFM and AAFP.

mailto:[email protected]

226 NC Pedersen

Secondly, the FIP mutation may be selective to a vari-ant FECV that appeared in the 1950s. This variantcould also have arisen because of the intra- andinter-species mutability of coronaviruses in generaland in this case, FCoV in particular. A third explana-tion may involve changes in how cats were viewedas pets and their husbandry in this modern era. Therewas a dramatic shift in the status, keeping, and breed-ing of cats as pets after World War II. The numbers ofpet cats greatly increased, purebreeding and catteryrearing became increasingly popular, and more cats,and in particular kittens, found themselves in shelters.These large multiple cat indoor environments areknown to favor feline enteric coronavirus (FECV) in-fection and FIP (reviewed by Pedersen et al 2008).25

Interestingly, feline leukemia virus (FeLV) infectionalso became rampant among indoor multi-cat house-holds during this period, and FeLV infection was a sig-nificant cofactor for FIP until it was pushed back intonature with testing, elimination/isolation, and even-tual vaccination in the 1970s and 1980s.26

Causative agentFIPV is a virulence form (biotype) of the FECV.27e29

The generic name FCoV has been loosely applied toall serotypes and biotypes of feline coronaviruses. Us-ing this nomenclature, FECV is technically an entericbiotype of FCoV, while FIPV is an FIP biotype. Unfor-tunately, the widespread use of the generic term FCoVhas led to confusion when reviewing the world’s liter-ature, because it is often difficult, and sometimes im-possible, to determine which virus biotype is beingdiscussed or even studied. To avoid confusion, this re-view will attempt to decipher the various authors’meaning when the generic term FCoV has beenused, and to use FECV to refer to the ubiquitous en-teric biotype of FCoV present in most healthy cat pop-ulations and FIPV when referencing the virulentbiotype that causes FIP in individual cats.

The FIPV is somewhat unique from other viruses,because it is infrequently spread from animal to ani-mal in a horizontal manner, yet it is highly infectiouswhen extracts of affected tissues or fluids are inocu-lated into na€ıve cats by a number of routes (see discus-sion on experimental infection). In this regard, FIPVresembles the infectious acute sarcoma-inducing ret-roviruses of cats, where the mutant form is foundonly in tumors, and is not horizontally transmittedin nature, but the parent FeLV is shed in various excre-tions and secretions and horizontally spread.30 There-fore, FIPVs are strongly cell and tissue bound andshedding in feces or urine would be possible only un-der special circumstances, such as if lesions efface therenal collecting ducts or intestinal wall and are shed inurine or feces. The infectivity of urine has beenreported at least once in the literature.31

The genomic structure of all feline coronaviruses issimilar. Their genomes are incorporated into a singleRNA strand made up of around 29,000 nucleotides.

There are 11 putative open reading frames (ORFs) orgenes, including a non-structural replicase with twolarge ORFs; four structural ORFs encoding the spike,envelope, membrane and nucleocapsid proteins; andfive accessory ORFs (3aec) and 7a,b. The 7a gene isnot essential for virulence, based on identification offield strains of both FECV and FIPV that lack a func-tional 7a gene.32 Likewise, small deletion mutationsin the 7b gene were found in 8/32 isolates associatedwith both enteric and infection and FIP.33 However,the functionality of the mutated 7b genes was notmentioned.

The FIPV, like FECV, also exists in two serotypesbased on virus neutralizing antibodies, type I andtype II.14,34 Serotype I virus has a distinctively felinespike protein, while the spike protein of type IIserotypes is a recombinant between feline and canineenteric coronaviruses.35 Type I FECVs/FIPVs predom-inate throughout the world,14,36,37 but type II strainsmake up more than 30% of isolates in certain countriessuch as Japan.38 Type II strains appear to be moreadaptable to tissue culture; however type I strainsmay be more likely to cause clinical FIP.37

In addition to the two serotypes, FECVs and theirFIPV biotypes exist in numerous strains defined bydistinct single nucleotide polymorphisms (SNPs)and insertion/deletions (INDELS). Differences aresmall between isolates within the same group ofcats, seldom more than 1e2%, but can be up to 20%between isolates from geographically distinct regions.Mutations are found in all genes to some extent or theother, but tend to be concentrated in the nine acces-sory and structural genes rather than in the two genescomprising the large replicase. Functional mutationsin several of the accessory genes, including 3c and7b, do not appear to prevent replication either in vi-tro39 or in vivo.29,40 There is no evidence that certainstrains of FECV are more likely to mutate to FIPVs,but this needs to be more thoroughly studied.

The mutation responsible for the FIPV biotype isconsistently found in the 3c gene, which encodesa small protein of unknown function (NC Pedersen,UC Davis, 2009, manuscript submitted).29 The muta-tions are usually SNPs causing premature stop codonsor deletion mutations that negate or significantly trun-cate the 3c gene product. The specific deleterious mu-tation in ORF 3c can differ even between affectedkittens in the same litter (NC Pedersen, UC Davis,2009, manuscript submitted), again supporting inter-nal mutation and auto-infection rather than cat-to-cat transmission as the primary route of exposure.Loss of 3c gene function does not prevent replicationin vivo or in vitro, but is thought to drastically altercell tropism by enhancing internalization and replica-tion of FIPVs in macrophages.41,42 In contrast, the par-ent FECV is more tropic for mature apical epitheliumof the bowel.27 However, this may be simplistic be-cause FECV infection does have a systemic as wellas intestinal phase and may be present in monocytesand phagocyte rich organs during primary infection.43

227A review of FIP virus infection

The question, however, is to what extent FECV repli-cates in these non-intestinal cells and organs.

The FECV / FIPV mutation is more likely to occurduring primary infection and in kittens, because bothconditions lead to a higher level of FECV replicationand the latter to a decreased resistance to the mutationonce it occurs.25 The FIPV mutants occurred in 20% ofprimary FECV-infected cats in one study.28 However,FECV infection is often recurrent and there is someevidence that FIP may also occur as a consequenceof these recurrent bouts of virus replication. Addieet al44 followed 56 cats deemed to have recurrentFECV infections based on decreasing and rising anti-body titers; 3/56 animals subsequently developed FIP.

The ‘internal mutation’ theory of the FECV / FIPVconversion has been recently questioned by Dye andSiddell.45 They compared the full length genomic se-quence of viruses found in two different tissues ofa cat with classical FIP. This study was based on ear-lier findings of Addie et al,46 which showed thatmany cats with FIP shed a coronavirus (presumablyFECV) in their feces. Their FIPV was from lesional tis-sues (liver), while their putative ‘FECV’ was obtainedfrom a segment of jejunum. They found no differencesin sequence between the two viruses and suggestedthat this called the internal mutation theory into ques-tion. However, the sequences from viral genomespresent in the two tissues showed extensive mutationsin the 3ce7b region, which occur in FIPVs but not intrue FECVs. The jejunum is a known site for FECVreplication during acute infection,27 which is in the il-eum, the colon and the rectum.47 Furthermore, wehave recently documented that a portion of coronavi-ruses isolated from feces of cats with FIPV are muchmore likely to resemble the FIPVs in the tissue, thanthe parent FECV. This means that only fecal coronavi-ruses that have an intact 3c gene can be designated asan FECV. Therefore, the two viruses being comparedby Dye and Siddell45 were identical FIPVs.

Several studies have implied that whether a cat de-velops FIP after FECV exposure is determined by howthe host responds to the virus and not from differ-ences in biotype.48e51 Another study suggested thatFIP results from the systemic spread of FECV fromthe gut into blood monocytes.43 However, thereshould be no doubt that FECV and FIPV are geneti-cally distinct in a minor but crucial manner, and thatthis difference is essential for causing FIP. Further-more, immunity to FECV infection does not confer im-munity to FIPV infection and differences in immunityto the two viruses are qualitative and not merelyquantitative. However, host and environmental fac-tors admittedly play a role in whether or not FIP isclinically manifested.28,52,53

EpizootiologyFIP is mainly a disease of domestic cats. It has also beenrecognized in the African lion, Mountain Lion, Leop-ard, Cheetah, Jaguar, Lynx, Serval, Caracal, European

Wild Cat, Sand Cat and Pallas Cat.54e63 An almost iden-tical infectious peritonitis has been described in inter-feron-gamma deficient mice64e66 and in ferrets67

infected with their respective species of coronavirus.FIP occurs wherever FECV is found27,44,53,58,68 and

is, therefore, worldwide and ubiquitous among virtu-ally all cat populations. Thus, the epizootiology of FIPis closely linked to that of FECV. Shedding of FECVcan be transient, recurrent, or chronic over periodsof months or years.25,69,259 The FECV is present in vir-tually all catteries and shelters with more than six catsand is shed by 60% or more of pet cats from multi-cathouseholds. One-third of older cats and 90% of kittensand juveniles presented to shelters in Sacramento, CA,USA were shedding FECV at the time of entry.70 A sig-nificant proportion of these cats were from the feralpopulation, so it is fair to say that FECV is enzooticamong both indoor and outdoor cat populations. Sim-ilar findings have been reported for most, but not all,other countries. Bell et al71,72 found the seroprevalenceof FECV to be 34% among pedigree cats in the Sydneyarea and non-existent in feral cats tested. They con-firmed that the number of cats per household hada significant influence on the infection rate, and thatcertain breeds had a higher incidence of seropositivitythan others.

Extensive studies have been undertaken on bothnatural27,73,74 and experimental FECV infections.25,27

The infection is spread by the fecaleoral route andviral shedding in feces occurs within a week of expo-sure.25 This shedding occurs from the ileum, the colonand the rectum.47 Shedding of FECV can be persistentfor up to 18 months or more, be persistent for 4e6months and intermittent for months thereafter, orcleared within 6e8 months (studied and reviewedby Pedersen et al 2008).25 However, immunity in re-covered cats is tenuous and reinfections resemblingprimary exposure often occur. The same strain tendsto persist in all cats within a given cattery or region;however, an occasional cat in a group may be infectedwith more than one strain or a different strain.73,259

Virus shedding is somewhat associated with coronavi-rus antibody titer levels; groups of cats with indirectfluorescent antibody titers of �1:100 or much morelikely to have coronavirus shedders than groups ofcats with titers of �1:25.25

Most deaths from FIP occur in cats 3e16 months ofage and are uncommon after 5 years.15,53 In one study,FIP was the most common single cause of disease incats younger than 2 years of age, while cancer was themost common single disorder in the 2e8 years agerange.75 The incidence of FIP in a veterinary medicalteaching hospital from 1986e1995 was 1:200 amongnew feline visits, 1:300 among total cat accessions, and1 of 100 accessions at diagnostic laboratories.76 The in-cidence of FIP is significantly higher among kittensand younger cats originating from purebreed catteriesand shelters. An average of 1e5% of young cattery orshelter cats in the US will die from FIP, with loses in cat-teries higher than from shelters. Enzootics with >10%

228 NC Pedersen

mortality were noted at least once in 5 years amongyoung cats and kittens originating from four typical cat-teries that were studied.53 Cave et al77 studied thecauses of death among 274 sheltered and privatelyowned pedigreed and non-pedigreed kittens from1986e2000. Twenty-five percent of all kitten mortalitywas due to feline parvovirus (FPV). The main viral in-fections in 15e38 day-olds were feline herpesvirusand feline calicivirus. FIP caused the death of 17/203kittens in the postweaning (35e112 days of age) period,with an incidence of 8.4%.

The FIP losses occur as enzootics or epizootics, withthe former being more common. The FIP losses aresporadic, unpredictable and infrequent in the enzooticform; catteries with enzootic FIP may not have anydeaths for years, followed by several cases in rapidsuccession. The disease may then disappear, only toreappear months or years later. Overall mortalityfrom enzootic disease over a 5-year or longer periodis usually 1e5%. However, epizootics of 3e49% havebeen observed in groups of kittens raised in one cat-tery over a 4-year period and similar explosive out-breaks have been seen in several other catteries.58,78

Epizootics of FIP seldom last for more than 6e12months before returning to an enzootic state. EnzooticFIP is usually associated with persistence of the sameor similar strains of enteric coronaviruses withina population, while epizootics are multifactorial andreflect increased population stresses, usually associ-ated with overcrowding and high kitten production,the unintentional use of genetically predisposedbreeding stock, possible bouts of horizontal transmis-sion, or the introduction of a new strain of FECV.

Several risk factors for FIP have been identified in cat-teries.53 In this study FIP was not significantly associatedwith particular catteries, mean cat number, mean age,gender, cattery median coronavirus antibody titer, hus-bandry and quarantine practices, caging and breedingpractices, or prevalence of concurrent diseases. How-ever, individual cat age, individual cat coronavirus titer,overall frequency of fecal coronarvirus shedding, andthe proportion of cats in the cattery that were chronicFECV shedders, were associated with increased risk ofFIP. Deaths from FIP were more frequent in fall and win-ter, and on the basis of analysis of cattery records, thenumber of deaths varied yearly. Epizootics (>10% mor-tality rate) were reported at least once in 5 years amongthe four catteries studied. One of the most significant fac-tors appeared to be genetic susceptibility, which ac-counts for up to 50% of the incidence.52 Multi-cathouseholds that have not personally experienced FIPloses, but which relocated or sold cats that later developFIP, were no more likely to experience FIP over a 6-yearperiod than households that had experienced no priorFIP loses or catteries that had lost cats to FIP.79 Thus, ev-ery cattery with enzootic FECV infection is at risk of los-ing cats from FIP if they breed enough kittens over a longenough time. The risk of developing FIP appears to de-crease over periods of 3 years in groups of cats, indicat-ing an increase in population resistance.79

The incidence of FIP in shelters seems to be directlycorrelated with the numbers and density of youngcats and the length of time that they are held beforeadoption; greater and longer being much worse(K Hurley, UC Davis, unpublished findings, 2008). Re-ports from the US and Europe indicate an increasedrisk for young cats, purebreds, and intact males anda decreased risk in spayed females.76 A report of 42confirmed FIP cases from Australia between 1990and 2002 has shown FIP to be over-represented in cer-tain pure breeds (Burmese, Australian Mist, BritishShorthair and Cornish Rex) and under-representedin other breeds (domestic shorthair, Persian) anda more even distribution across the age spectrum.80

The breed incidence of another veterinary teachinghospital was analyzed over a 16-year period and thebreed, gender and reproductive status of affectedcats were compared to the general cat populationand to mixed breed cats evaluated during the sameperiod.81 As with previous studies, sexually intactcats and purebreed cats were significantly more likelyto be diagnosed with FIP; males and young cats alsohad a higher prevalence of disease. Abyssinians, Ben-gals, Birmans, Himalayans, Ragdolls and Rexes hada significantly higher risk, whereas Burmese, ExoticShorthairs, Manxes, Persians, Russian Blues and Sia-mese cats were at decreased risk. Such studies indi-cate that the incidence of FIP among breeds can varygreatly between countries and regions and FIP losesare probably more related to bloodlines within a breedthan to breeds themselves.

Clinical featuresFIP refers to the more common effusive (wet, non-parenchymatous) form of the disease; a transmissibleinflammatory condition of the visceral serosa andomentum with exudation into the abdomen.3 A sec-ond form of the disease is characterized by granulo-matous involvement of parenchymatous organs suchas the kidneys, mesenteric lymph nodes, bowel wall,liver, central nervous system (CNS) and the eyes.6,15

Granulomatous FIP is called ‘dry’, parenchymatous,or non-effusive because there is no inflammatory exu-dation into body cavities.

The most common form of FIP is wet or effusive,although the proportion of cats with dry FIP appearsto be rising over the last few decades. Cats rarely man-ifest both forms of the disease at the same time, and ifthey do, it is usually a transition stage from wet to dryor dry to wet. Under experimental conditions, catsthat develop dry FIP often have a brief episode ofeffusive disease at the onset, while some cats withdry FIP may become more effusive in the terminalstages of their illness. However, these transitions aremuch less noticeable in nature.

The incubation period (time from infection to dis-ease) of effusive FIP is 2e14 days under experimentalconditions.12,82e84 The incubation period for experi-mentally induced dry FIP is several weeks longer.

Fig 1. Grossly distended abdomen of a kitten with effusiveFIP. Note the scrotal enlargement due to inflammation of thetunics.


However, the actual incubation period in naturalinfections is unknown, but there is evidence the infec-tion may smolder in a subclinical state for weeks,months and even years before overt signs are noticed.In retrospect, the clinical onset of FIP may be precededby a long history of vague ill health and stuntedgrowth. Co-infections with other feline pathogensmay lower resistance to FIPV or may complicate thedisease picture (see preceding discussion).

The earliest signs of overt FIP, besides a failure tothrive in young cats, include a progressively worsen-ing malaise, fluctuating fever, inappetence, andweight loss. Other signs of disease are superimposedon these basic disease signs, depending on the formand organ distribution of the inflammation (Table 1).Abdominal distension is the most common physicalfinding in wet FIP and ranks higher than cardiovascu-lar disease, neoplasia, hepatic or renal disease ascauses of ascites in cats85 (Fig 1). The abdomen, be-sides being greatly enlarged, is often doughy feelingand painless on palpation, and a fluid wave is easilyinduced on percussion. Upon opening, the abdomenis found to contain up to a liter of a yellow-tinged,slight to moderately cloudy, mucinous fluid (Fig 2).Dyspnea can be a feature of cats with pleural involve-ment and thoracic effusions (Table 1). Clinical signsdue to ocular and CNS involvement are seen in lessthan 9% of cat with the wet form of FIP (Table 1).

There are several uncommon features of effusiveFIP that warrant mention. Intact males frequently de-velop scrotal enlargement due to extension of the peri-tonitis to the tunics surrounding the testes and edema(Fig 1). A syndrome of hepatic lipidosis and extremeskin fragility has been described in one cat with wetFIP.86 In-utero FIPV infections have been observed inkittens born to queens that developed effusive FIPduring pregnancy; pneumonia, pleuritis and hepatitiswere the principal lesions in affected kittens.87 Manycats with FIP have a generalized synovitis, due to im-mune complexing or the migration of infected macro-phage/monocytes into the synovium. A cat with FIPmay present, therefore, with signs of fever and

Table 1. Variability in clinical signs of effusive(wet) FIP

Clinical signs referable toinvolvement of the:

% of affected cats

Peritoneal cavity 58.0Peritoneal and pleural cavity 22.0Pleural cavity 11.0Peritoneal cavity and eyes 2.8Peritoneal cavity and CNS 1.9Peritoneal and pleural cavity, CNS 0.9Peritoneal and pleural cavity, eyes 0.9Pleural cavity, CNS and eyes 0.9Peritoneal cavity, CNS, eyes 0.9

lameness. This can be mistaken for another type ofinfection or an immune-mediated polyarthritis. How-ever, the more classic signs of FIP usually developsoon after, making the cause of the lameness apparent.

Fig 2. Over 600 ml of a yellow, mucinous effusion was re-moved from the abdomen at necropsy. Note fibrin tags onliver and spleen and ground glass appearance of the serosa.

Table 2. Variability in clinical signs of non-effusive (dry) FIP

Clinical signs referable toinvolvement of:

% of affected cats

Peritoneal cavity 32.0CNS 23.0Eyes 15.0CNS and eyes 8.5Peritoneal cavity and eyes 7.4Peritoneal and pleural cavities 4.3Peritoneal and pleural cavities, CNS 3.2Peritoneal and pleural cavities, eyes 2.1Peritoneal cavity, CNS, eyes 2.1Pleural cavity 1.1Pleural cavity, CNS, eyes 1.1

230 NC Pedersen

As the name ‘dry FIP’ implies, thoracic and abdom-inal effusions are either absent or too scant to be de-tected other than at necropsy. Involvement of theeyes and/or CNS predominates in 60% of the catswith dry FIP (Table 2). Signs referable to abdominalinvolvement are seen in 40% of animals, either withor without ocular and CNS disease (Table 2). The

Fig 3. Cross section of a kidney from a cat with dry FIP.Numerous granulomatous lesions are seen on the capuslueof the kidney and extending downward into theparenchyma.

abdominal lesions of dry FIP are much larger, fewerin number and less widespread than the lesions ofwet FIP. Lesions of dry FIP tend to extend downwardfrom the serosal or pleural surfaces into underlyingparenchyma; hence the alternative name ‘parenchy-matous FIP’. Abdominal lesions are frequently foundin the kidneys (Fig 3) and mesenteric lymph nodes(Figs 4 and 5), and somewhat less frequently in theliver and hepatic lymph nodes (Fig 5). Involvementof the wall of the caecum and colon with associatedcaeco-colic lymphadenopathy is a specific form ofdry FIP (Fig 6) associated with signs typical of an ul-cerative colitis (ie, soft, blood and mucus ladenstools).88,89 Abdominal lesions are often evident onpalpation and sometimes associated with local pain.About 10% of cats with dry FIP will have thoracic in-volvement, but chest lesions are usually localized andonly one part of a more systemic infection. Small gran-ulomas may involve the pleura and underlying lungparenchyma (Fig 7). Involvement of the pericardiumhas been described, and can lead to fluid distentionof the pericardial sac, cardiac tamponade and heartfailure.90

CNS involvement is frequent in cats with dry FIP(Table 2). Over one-half of cats with inflammatory dis-ease of the CNS have FIP, as well as one sixth of thetotal number of cats showing CNS signs from anycause.91 FIP is also the most common disease of thespinal cord in cats less than 2 years of age and isone of the three leading causes, along with lympho-sarcoma and vertebral neoplasia, of spinal diseaseamong cats of all ages.75 Most cats with CNS FIP areless than 2 years of age and often originate from largemultiple cat households.92

Fig 4. Enlarged mesenteric lymph node in a cat with thedry form of FIP. Note the residual fibrinous plaque on thespleen. Such residual lesions support the concept thatmany cases of dry FIP began as a brief bout of wet FIP.

Fig 5. Mesenteric and hepatic lymph nodes and liver froma cat with non-effusive FIP. The lymph nodes are enlargedand involved with granulomatous adenitis. The liver cap-sule contains raised, whitish foci 0.5e1 cm in diameter,extending into the underlying parenchyma.

Fig 7. Lungs and heart of cat shown in Fig 5. A solitarypleural granuloma is noted along the edge of a craniallung lobe.


CNS involvement in cats with dry FIP is varied in itsclinical expression, depending on what other organsare involved (Table 2), its exact localization in the ner-vous system, and severity. Signs referable to spinalcord involvement, such as posterior paresis, incoordi-nation, hyperesthesia, seizures and palsy of the bra-chial, trigeminal, facial and sciatic nerves, have allbeen reported.15,75,93e98 Hydrocephalus, secondary todisease of the choroid and ependyma, has also beendocumented92,99e101 and can lead to dementia, person-ality changes (aggression, rage, hiding/withdrawal,etc) or convulsive disorders. Cerebellarevestibular

Fig 6. Gross appearance of the caecum, colon and ceco-colic lymph nodes of a cat with the intestinal form of dry FIP.

signs, such as nystagmus, head tilt or circling, havealso been caused by FIP.

Ocular involvement, like CNS disease, is much morelikely to occur in cats with dry than wet FIP (Tables 1and 2). Uveitis and chorioretinitis are the predominantocular manifestations of dry FIP96,102e106 (Figs 8 and 9).FIP is also the most frequent cause of uveitis/choriore-tinitis in cats, with less common causes being FeLV-as-sociated lymphosarcoma; trauma; toxoplasmosis; andlens-induced uveitis.107,108 Ocular disease in dry FIPoccurs solely or in association with lesions in the CNSor peritoneal cavity (Table 2). A change in the colora-tion of the iris is a frequent early sign of ocular FIP(Fig 8). Keratic precipitates on the caudal aspect ofthe cornea are characteristic of FIP and are due to accu-mulations of fibrin, macrophages, and other inflamma-tory cells (Fig 9). Focal lesions akin to the granulomas ofparenchymatous organs may be apparent in the irisand distort the shape of the pupil (Figs 8 and 9).

There are several uncommon, but interesting, man-ifestations of dry FIP. Granulomatous involvement ofthe peripheral tissues resulted in priapism in a cas-trated cat.109 Chronic fibrinous and necrotizing orchi-tis may cause the scrotum and testicles to appearenlarged110,111 and scrotal enlargement may be oneof the primary presenting complaints. Miscellaneoussites for lesions in non-effusive FIP include the nasal

Fig 8. Uveitis of the right eye in a cat with the dry form ofFIP. The color of the iris has changed, the anterior chamber issomewhat hazy, and there is a pigmented lesion in the centerof the cornea (a keratic precipitate). Note the irregularity inthe shape of the right pupil compared to the normal leftpupil.

232 NC Pedersen

passages, tongue and distal small intestine. Syringo-myelia has been caused by involvement of the fourthventricle in one cat.112 Cutaneous lesions of a granulo-matous type have been reported in a feline immuno-deficiency virus (FIV) infected cat with FIP; thelesions were thought to be due to a coronavirus

Fig 9. Keratic precipitates on the inner cornea of a cat withdry FIP. Note the reversed D-shape of pupil due to infiltra-tion of the iris.

induced vascultis.113 Toxic epidermal necrolysis hasalso been observed in a cat with dry FIP (NC Peder-sen, UC Davis, unpublished observation, 2008). Non-puritic, slightly raised intradermal papules over theneck and chest walls were observed in the terminalstages of a cat that had non-effusive, progressing to ef-fusive, FIP.114 The FIPV was identified in skin lesionsby immunohistochemistry.

The lesions of dry FIP have also been mistaken forcancer, in particular lymphoma.94,115 One FeLV nega-tive cat with FIP subsequently developed a myelopro-liferative disease.116 Monoclonal gammopathies havebeen observed in four cats with FIP117 and a fifth catwith FIP was observed to convert from a polyclonal-to monoclonal-gammopathy (Hurvitz, 1982 as quotedin Pedersen, 1987).119

FIP has many interesting interactions with other in-fectious agents. These agents may affect immunity toFIPV, such as FeLV infecton. Alternatively, FIPV mayinduce an immunosuppression that encourages op-portunistic type pathogens. Concurrent FeLV infectionwas seen in one-third to one-half of all FIP cases thatwere tested in the 1970s and 1980s.15,120 Infection withFeLV seems to have a specific interfering effect on theongoing immunity to FIPV. Cats that failed to developFIP after experimental infection developed FIP within6e16 weeks after becoming FeLV viremic, indicatingthat many FIP recovered cats maintain residual infec-tions.119 Mimicking the situation in the field, FIP alsoappears after natural exposure to both viruses.119,121

With the virtual elimination of FeLV from pet cat pop-ulations in westernized countries, this relationshiphas become much less common. Affected kittenswith subclinical or clinical FIP are more susceptibleto upper respiratory infections caused by myco-plasma, chlamydophila or herpesvirus, indicatingthat their resistance is not normal. It is also knownthat cats with advanced FIV infection are more sus-ceptible to FIP when exposed to FECV.28 Systemictoxoplasmosis, a relatively rare clinical infection ofcats, has been described in two cats with FIP. One re-port was from 1966122 and a second report from1995.123 The occurrence of systemic toxoplasmosis inthese two cats was probably associated with reactiva-tion of encysted organisms left from a much earlierself-limiting infection. FIP is frequently associatedwith a lymphopenia, which is a non-specific findingindicative of immunosuppression. We have also no-ticed an increase in bacterial infections in cats withFIP, some manifested as terminal sepsis or morechronic local infections of internal organs (NC Peder-sen and P Pesavento, UC Davis, unpublished observa-tions, 2008). One young cat had a chronic bacterialinfection of the reproductive tract that was manifestedearly as a purulent vaginal discharge and terminallyas a suppurative infection of the ovary and fallopiantube, while a second cat had a secondary bacterialpneumonia. The stresses of pregnancy and parturi-tion, surgical procedures such as spaying, neuteringand declawing have also increased the incidence of


FIP in young cats, suggesting that stresses occurring ata time when the young cats are fighting an FIPV infec-tion may tip the balance against the host.15

Table 3. Virulence (lowest to highest) of severaltissue culture-adapted strains of FIPV

FIPV strain Route infection Mortality

FIPV-UCD2 IN (intranasal) 0/5IP (intraperitoneal) 0/5

FIPV-UCD3 PO (by mouth) 0/4IP 2/5

FIPV-UCD4 PO 0/4IP 3/8

FIPV-UCD1 PO 3/15IT 7/10IP 4/4

Wellcome PO 3/4IP 4/4

FIPV-TN406 (Black) PO 4/5

FIPV-79-1146 (-Nor15) PO 25/27IP 7/7

PathogenesisThe FIPV enters target macrophage/monocytes byfirst binding to the cell surface and than being inter-nalized by a clathrin and caveolae independent anddynamin dependent endocytosis.124 Dewerchin andcoworkers125 added upon these findings and sug-gested that viral antigen-antibody complexes in FIPwere not internalized through any of the previouslydescribed pathways. The internalization process wasindependent from phosphatases and tyrosine kinasesbut did depend on serine/threonine kinases. Once in-ternalized, viral antigen-antibody complexes passthrough the early endosomes and accumulate in thelate endosomes. The complexes left the late endo-somes within 30e60 min but were not degraded inthe lysosomes.

Binding involves specific attachment sites on thespike protein and corresponding cell-surface recep-tors. The cellular receptor appears to be aminopepti-dase-N, which was first identified as the receptor forTGE coronavirus of swine.126 Aminopeptidase-N isa cell-surface metalloprotease on intestinal, lung andkidney epithelial cells. However, this interactionmay be host species specific, because non-permissivecell lines genetically engineered to express porcineaminopeptidase-N become permissive to transmissi-ble gastroenteritis virus (TGEV) but not to feline andcanine coronaviruses.126 The species specificity of theaminopeptidase-N receptor has also been shown bytesting intraspecies chimers of the protein.127 Determi-nants on aminopeptidase-N responsible for binding ofporcine, feline, and canine aminopeptidase-N to theirrespective host coronaviruses are found on its C-ter-minus.128 Hohdatsu et al129 noted differences in recep-tors for serotype I and II FIPVs using the blockingeffect of specific mouse monoclonal antibodies. Theirfindings suggested that feline aminopeptidase-N isa receptor for type I FIPV but not necessarily type IIFIPV. Interestingly, although neither porcine nor hu-man aminopeptidase-N will act as a receptor forFIPV, a chimer of both is able to act as a receptor forFIPV.130 All of these findings suggest that subtle dif-ferences in certain regions of aminopeptidase-N canradically change species and even serotype tropisms.

The acquisition of macrophage tropism appears tobe an essential step in the evolution/transformationof an FECV to an FIPV and from a largely non-patho-genic and localized enterocyte pathogen to a highly vir-ulent and systemic monocyte/macrophage pathogen.The relationship between virulence and macro-phage/monocyte tropism has been firmly establishedin the literature. Attachment and internalization ofFIPV by monocytes by endocytosis has been comparedfor two FIPV strains; FIPV-79-1146 being internalizedmuch more efficiently than FIPV-Black.131 Strains of

FIPV such as WSU-79-1146 are more macrophagetropic than tissue culture-adapted and non-virulentstrains such as WSU-79-1683. The former virus is as-sumed to be a prototypic FIPV, while the latter is as-sumed to be the prototype of an FECV.40 However,the assignment of these isolates as prototypic mayhave been premature. Both are serotype II viruses, un-like most field strains of FECVs and FIPVs; both are tis-sue culture-adapted to non-macrophage cell lines; andFIPV-WSU-79-1146, unlike a number of serotype IFIPVs, is exceptionally virulent by every commonroute of infection (Table 3). The same doubts apply toWSU-79-1683 as a prototypic FECV. It lacks most ofthe 7b gene, which is present in non-tissue culture-adapted strains of FECV132 and all other attempts toadapt additional field strains of FECV to tissue culturehave so far failed. We have also noted that WSU-79-1683 has a deleterious mutation in its 3c gene, indicat-ing that it may have originated from an FIPV (NCPedersen, UC Davis, personal observation, 2008).However, WSU-79-1683 does behave as an enteric bio-type, when compared to FIPV-WSU-79-1146, in its de-pendence on cathepsin B and cathepsin L activityand low pH for entry into host cells.133 It also causesan inapparent to mild enteritis when infected into lab-oratory kittens, targets appropriate tissues, and doesnot cause FIP.40 Nevertheless, without more bonifiedtissue culture-adapted FECV strains, grown on appro-priate host cells, comparisons of macrophage versusenterocyte tropisms of FECVs and FIPVs in vitro usingthese two FCoV biotypes, should be interpreted withsome caution. Studies with FECVs have been under-taken in nature, but they have dealt more with systemicspread than intestinal tropism. For instance, FECVsmay be detected in the blood during the course of nat-ural infection,43 but at much lower levels than seen

234 NC Pedersen

with FIPVs.134 Therefore, the affinity of the enteric bio-type for the intestinal epithelium (Fig 10) and the FIPbiotype for macrophages may be more quantitativethan qualitative.

The genetic basis for differences in macrophage tro-pism between FECVs and FIPVs is not fully under-stood. It would be logical to assume that thedifferences are due to characteristic functional muta-tions in the 3c gene. However, some research suggeststhat the acquisition of macrophage tropism by FIPVsis determined by mutations in the spike protein.42 Un-fortunately, this study also compared WSU-79-1683and WSU-79-1146, which do not appear to be typicalfor either of their respective biotypes. Mutant virusesof the FIPV phenotype have been detected in as highas 20% of primary FECV infections,28 indicating thatmacrophage tropic mutants occur with great fre-quency. However, only a fraction of mutants may ac-tually produce a progressive disease.28,52 Theseabortive infections are thought to be associated witha strong and rapid cellular immune response andare probably contained within the mesenteric lymphnodes. In the unfortunate cats, mutant viruses are al-lowed to proliferate unchecked within macrophagesand are carried by these cells to specific sites in thebody.135 Many of these tissues and organs appear tobe fed by the cranial mesenteric artery and are richin phagocytic cells. Targets are the mesenteric lymphnodes, serosal surfaces of the gut, and to a lesser ex-tent the pleura, and the omentum. Some virus also ap-pears to reach the meninges (especially on theposterior-ventral surfaces of the brain), the ependyma

Fig 10. Immunofluorescent antibody staining for feline co-ronavirus antigens on a section of small intestine from a catexperimentally infected with FIPV. Virus is concentrated inthe mature apical epithelium at the tips of the intestinal villi.

surrounding the ventricles, the dura mater of the spi-nal cord, and the uveal tract and retina of the eyes.

The issue of whether different variants of FIPV canexist in the same cat was raised by Battilani et al,136

who analyzed quasispecies composition and phyloge-netic relationships of coronaviruses in cats with natu-rally acquired FIP using single-strand conformationalpolymorphisms in the nucleocapsid and 7b genes.They found a high degree of heterogeneity, especiallyin the nucleocapsid gene, which correlated directlywith the seriousness of the clinical disease. Sequencesalso varied according to geographical origin. The exis-tence of quasispecies in cats with FIP has recentlybeen confirmed in the author’s laboratory. Two offour sibling cats that died of FIP were found to havetwo distinct variants in their tissues, and two ormore variants in their feces, at the same time (NC Ped-ersen and H-W Liu, UC Davis, personal observation,2008). A recent published immunohistochemicalstudy on brain lesions from 15 cats with the dryform of FIP used two different monoclonal antibodiesdirected against unrelated epitopes of CCV to searchfor variants.137 These authors also used a myeloid-histiocytic marker, MAC387, to identify infected mac-rophages. Double immunostaining demonstratedmacrophages that were positive for either the CCV-monoclonal antibody alone, or for both the CCV-and FIPV-monoclonal antibodies. The conclusionwas that there were two cohorts of phagocytes, bothcontaining virus with the CCV-epitope, but differingby the presence or absence of the FIPV epitope. Unfor-tunately, the strength of reactivity of the two monoclo-nal antibodies differed and the authors did notconsider that various macrophages contained differ-ent amounts of antigen, giving the impression thatthere was differential staining. Given the origin ofFIPVs, it is not surprising that a similar phenomenonwas observed in cats infected naturally and experi-mentally with FECVs.138 Most cats shed FECV in fecesthroughout the experiment without disease signs.Sequences from the nucleocapsid and 7b genes wereamplified by reverse transcriptase polymerase chainreaction (RT-PCR) from the cerebral cortex, dura ma-ter, pancreas, lungs, third eyelid, and the heart musclein four cats. Viruses differed between organs inwhether or not they expressed the 7b gene and in theirnucleocapsid gene sequences.

Virus replication in macrophages appears to bevery slow during the first 2 weeks or so. A dramaticincrease in the levels of macrophage infiltration, mac-rophage infection, virus replication, and viral spreadoccurs around 10e21 days after primary infection inexperimental infections, coincidental with the appear-ance of virus specific antibodies.139,140 Recovery, whenit occurs, probably takes place sometime during thisinitial quiescent period. Antibodies are virus neutral-izing in vitro, but appear to have a disease facilitatingeffect in vivo. The failure of the antibody response toneutralize or prevent virus spread in vivo can be ex-plained in at least two ways. First, antibody binding


to virus may block aminopeptidase-N receptors andprevent virus-to-cell attachment, but enhance virusuptake by macrophages through their complementbinding (Fc) receptors. This antibody-mediated en-hancement is specific for feline antibodies and felinemacrophages, and may involve antibodies to bothvirus spike and membrane proteins.141 Second, viralantigens may not be arrayed on the surface of infectedmacrophages, thus limiting virus spread to intimatecell-to-cell contact.142

The hallmark of the antibody phase of the infectionis a rapidly spreading serosal/omental centered in-flammation associated with expanding cycles of mac-rophage infection, virus replication, virus release fromdying infected macrophages, and yet more virusbound by yet more macrophages. This reaction is cen-tered on small venules and the result is a lesion calleda pyogranuloma. Pyogranulomas form around smallvenules in target tissues and are surrounded by an in-flux of protein-rich edema fluid, and a sprinkling oflymphocytes, neutrophils and plasma cells. Thisreaction has been likened to Arthus-type hypersensitivity.13,143,144 Cytokines and other inflammatorymediators released by macrophages and polymorpho-nuclear leukocytes undoubtedly add to the characterand intensity of the lesions.

Complement activation and vascular damage re-sults in a disseminated intravascular coagulopathy(DIC),145 which is not usually clinically appreciated.Cats, by their nature, are much less likely to developovert hemorrhages when in DIC, but DIC may con-tribute to the mucinous nature of the FIP exudate.The yellowish or bluish tinged fluid, serum, andurine, which are often found in cats with FIP, arealso due to microhemorrhage and red cell damage re-sulting from vasculitis, complement and clotting fac-tor activation, and DIC. Red blood cells that leakinto the tissues, or manage to make it to the spleenin a damaged state, are destroyed by phagocytes. Anattempt is made to recycle hemoglobin by breakingit down to biliverdin and bilirubin. The carrier pro-teins of both biliverdin and bilirubin are rapidly satu-rated in cats, allowing them to build up to detectablelevels in the exudate, serum, urine and tissues.

The occurrence of two forms of FIP, effusive (wet,non-parenchymatous) and non-effusive (dry, paren-chymatous), has long been intriguing. Observationswith experimentally and naturally infected catsshow that dry FIP is often preceded by a brief episodeof wet FIP. The lesions of dry FIP are therefore rem-nants of an initial surface orientated infection. DryFIP may also become effusive in the terminal stagesof disease when the immune system collapses. Thedry form of FIP is presumed to result from a state ofpartial protective immunity. Protective immunity toFIPV is largely cell mediated and involves a changein the way infected macrophages respond to virusand vice versa. If infected macrophages receive propersignals from the host, they gain the capacity to destroyvirus rather than to act as virus incubators and

disseminators. The host also comes to view infectedmacrophages as foreign and will attempt to destroythem as well. If the cellular immunity develops earlyin the course of infection and has sufficient strength,clinical signs of disease will not occur and virus repli-cation will be checked. Some virus may persist ina subclinical or latent state, most likely within macro-phages in organs such as the mesenteric lymph nodes,for months or even years. If humoral immunity oc-curs, but cellular immunity fails to develop, the result-ing disease is effusive or wet. However, anintermediate stage of immunity can also occur, withstrong humoral immunity and weak cellular immu-nity. Virus is contained to a much greater extentthan in the effusive form of the disease, but is not fullyeliminated. A tug of war between virus replicationand virus destruction follows, resulting in the non-effusive or dry form of FIP. This form of the diseaseis characterized by more classic granulomas. Unlikethe pyogranuloma, which is filled with virus ladenmacrophages and a few scattered neutrophils andlymphocytes, the granuloma contains small foci ofmacrophages in the center, many of which containno or small amounts of viral antigen, surrounded bybroad bands of lymphocytes and plasma cells. Granu-lomatous lesions of dry FIP are less common andmuch less widespread than pyogranulomas of wetFIP, and tend to extend down from the surfaces of or-gans into the underlying parenchyma. This surfaceorientation of granulomas is consistent with their ori-gin as pyogranulomas. The organ distribution of thelesions of dry FIP can also be explained by a state ofpartial or weak immunity. During the initial stage ofFIPV infection, virus reaches the brain and eyes via in-fected monocytes, but immunity has difficulty pene-trating these areas due to the blood-brain barrier.If protective immunity is strong, virus and virus-infected cells will be easily cleared from organs onboth sides of the blood-brain barrier. However, if pro-tective immunity is weak, it will be more effective inhandling virus and virus-infected cells in the core ofthe body than it will be in the eyes and brains. There-fore, lesions in the brain and eye will be more likely topersist than lesions in the abdomen or other sites, thusproviding an explanation as to why ocular and CNSlesions predominate in the dry form of FIP.

Hematologic, clinicopathologic andimaging featuresThe diagnosis of FIP should be relatively simple, givenits affinity for younger cats, its strong tendency to in-volve catteries and shelters, the typical physical andhistorical findings, and numerous characteristic labo-ratory abnormalities. Nonetheless, it somehow re-mains one of the most difficult of diagnoses for manyveterinarians. The truth is that veterinarians have littletrouble in placing FIP high, or at the top, of their diag-nostic list, but have great difficulty, and even reluc-tance, in confirming their diagnosis. However, it is

236 NC Pedersen

not only the veterinarian that has difficulties, but theowners as well are reluctant to give up without a defin-itive diagnosis. It must be remembered that a diagnosiscan be based on cumulative odds rather than a single,simple, definitive test result. Ayoung cat from a catteryor shelter with chronic uveitis and/or neurologic signs,high serum proteins, hyperglobulinemia and hypoal-buminemia, fluctuating antibiotic unresponsive fever,leukocytosis with a lymphopenia, and an anemia ofchronic disease can have no other disease than dryFIP based on odds alone. Likewise, the same cat withsimilar history and laboratory findings, but with a yel-low-tinged, mucinous, inflammatory ascites instead ofuveitis or neurologic signs is highly unlikely to haveany other disease than wet FIP. Results from a testthat is incorrectly touted as being definitive or highlyindicative of FIP will only confuse the issue and inducedoubt when the diagnosis should not have been indoubt. Ultimately, FIP must be diagnosed by applyinga workable knowledge of the disease with sensibleweighing of signalment, history, clinical signs, clinico-pathologic findings, serology and ante- or post-mortemexamination of affected tissues by histopathology andimmunohistochemistry.

Hemograms of cats with FIP often demonstrate sev-eral abnormalities, which tend to be similar regardlessof the form of disease. A low-grade to moderatelynon-responsive anemia with hypoalbuminemia is typ-ical of chronic disease. Other common hematologicchanges include a leukocytosis with an absolute lym-phopenia and neutrophilia, and hyperproteinae-mia.15,146e148 Serum total protein elevations areassociated with increased levels of globulin and de-creased levels of albumin. Cats with virus positivelymph nodes at necropsy were purportedly morelikely to have globulin elevations.147

Hypergammaglobulinemia is not directly related toFCoV antibody titers,146 suggesting that non-specificantibody globulins and/or complement factors are im-portant contributors to the globulin elevation. An albu-min globulin ratio <0.6 is considered diagnostic for aninflammatory process, and most commonly FIP.149 Themost consistent changes shown by serum protein elec-trophoresis were increases in a2- and g-globu-lins.15,146,148 Haptoglobin makes up most of theincrease in a2-globulins.150

Abdominal and pleural effusions of FIP have char-acteristics that are as definitive as any other singletest. Almost one-half of all effusions in cats are dueto FIP149; therefore, the odds of FIP are 50:50 by thepresence of an effusion alone. If the effusion havethe features of FIP the odds are even greater; if thecat is younger the odds are greater still, if the cat isfrom a shelter or cattery, the odds go up even more,etc. FIP effusions are usually a light to dark yellow,and occasionally green-tinged, due to elevated levelsof bilirubin or biliverdin, respectively. The fluid is mu-cinous in character and when a drop on a slide istouched with a needle tip and then pulled away,a string will often form. The fluid ranges from

somewhat clear to cloudy and often contains visible fi-brin tags. The protein content of the effusions rangesfrom 3.9 to 9.8 mg/ml with the globulins comprising50e82% of the total.148 Cell counts can range from1600 to 25,000 per ml. Macrophages, non-toxic neutro-phils and lymphocytes predominate. Fluid effusions,particularly from the pleural cavity, may be tingedwith red cells. Effusions will often form partial clotswhen allowed to sit in an anticoagulant free tube.The electrophoretic pattern of the effusions will besimilar to that of the corresponding serum.146 Anti-body titers higher than those of the corresponding se-rum have often been detected in the effusions,146 butlike comparative antibody titers in serum and cerebro-spinal fluid (CSF),92,151 the significance of this findinghas not been determined. A definitive diagnosis canbe made by direct immunohistochemical examinationof cells in the fluid for viral antigens.146,152 Immuno-fluorescence is several times more sensitive than im-munoperoxidase. Real-time RT-PCR analysis of cellsfrom FIP effusions will demonstrate high levels ofviral RNA.

Hyperbilirubinemia, and less commonly hyperbili-verdinemia, frequently without jaundice, are commonin cats with FIP, especially the effusive form. In fact,FIP is the most common cause of an icteric serum orplasma in cats under 3 years of age. The urine isalso yellow-tinged due to a hyperbilirubinuria. Agreenish hue to the urine and serum is less commonand due to high levels of biliverdin. The increasedlevel of bilirubin or biliverdin in the blood is usuallynot due to liver involvement per se, but rather to mi-crohemorrhage into tissues and extravascular destruc-tion of red blood cells by phagocytic cells, as part ofthe vasculitis and DIC. Heme from destroyed red cellsis then recycled as biliverdin and finally bilirubin.Cats are deficient in carrier systems to remove thesehemoglobin breakdown products from the blood,hence their tendency to accumulate in effusions,lesions, or serum.

Aqueous humor and CSF in cats with ocular orCNS disease also show similar, but more variable,increases in proteins and leukocytes. Typical CSFfindings in cats with FIP include a protein concentra-tion of greater than 200 mg/dl and a white cell countof over 100 cells/ml, consisting predominantly of neu-trophils.153 In contrast, the CSF of cats with other vi-ral disease usually has a protein concentration of lessthan 100 mg/dl and a total white cell count of lessthan 50 cells/ml. Singh et al154 defined an inflamma-tory CSF as having a total nucleated cell countgreater than 5 cells/ml, or one in which the totalcount was normal but the differential count was ab-normal. Gender, degree of CSF inflammation, neuro-anatomical location and systemic signs providedwere not helpful in determining whether the inflam-mation was due to FIP or other diseases such as cryp-tococcus, lymphoma or trauma. Synovial fluid fromcats with effusive FIP is frequently inflammatory incharacter.


Many serum antibody tests have been used for thediagnosis of FIP, including indirect immunofloures-cent antibody (IFA), virus neutralization, and en-zyme-linked immunosorbent assay (ELISA).155e157

Some feel that ELISA is the most sensitive proce-dure,157 but this author feels that most antibody testsare equivalent if done properly. The most commontest is an IFA assay using TGE virus37 or FIPV156 in-fected swine or cat cells as a substrate. Titers are usu-ally read in distinct multiples of serum dilutions (1:5,1:25, 1:100, 1:400, 1:1600, 1:6400, .). They were origi-nally reported as being helpful in diagnosing cats thathad FIP or were suffering from other illnesses.156 Un-fortunately, coronavirus antibodies do not differenti-ate between cats infected with FECV and FIPV, andalthough very high titers (�1:1600) are highly sugges-tive of FIP and negative titers tend to rule outFIP,149,156 the overlap in titers between healthy corona-virus-exposed and diseased cats is so great that it haslittle definitive diagnostic value in individual ani-mals.58,68,146,148 Titers appear to rise progressively inmany cats as they go from a subclinical to clinicalstage of disease.119 However, progressive monitoringof titers is rarely done and titers are usually measuredwhen disease signs appear and titers levels have al-ready plateaued. Moreover, titers may also fall dra-matically at the end, especially in cats withfulminating effusive FIP.

The sensitivity and specificity of so-called FIP serol-ogy has been repeatedly questioned. Hematology, an-tibody titers and serum protein electrophoresis from48 cats (34 effusive and 14 non-effusive forms) af-fected with FIP were studied and compared withthose of 20 healthy cats.148 In the effusive form, anti-body titers and protein electrophoresis in the effusionswere analyzed. Seropositive animals (antibodytiter> 1:100) were present among both the FIP in-fected (73%) and healthy cats (70%). Sparkes et al148

also compared serologic data from 65 cats in whichFIP was considered as a differential diagnosis, butended up having another disease. They found thatthe presence of multiple abnormalities compatiblewith FIP increased the specificity (ie, decreased falsepositives), but decreased the sensitivity of the diagno-sis (ie, increased false negatives). However, there is nodoubt that cats with very low or negative (�1:25)FCoV antibody titers are less likely to have FIP, whilecats with very high titers (�1:1600) are more likely.Given these limitations, this author recommends thatFIP serologic tests only be used as an aid in rulingin or out the possibility of FIP. The diagnosis of FIPshould never be made on antibody titers alone, a rec-ommendation made by others.72,158

There have been many attempts to improve thespecificity and sensitivity of antibody based tests forFIP. One commercial test measures antibody titers tothe 7b protein of FCoV based on the presumptionthat a prototypic FECV (WSU-79-1683) lacks a 7b pro-tein, while a prototypic FIPV (WSU-79-1146) has an in-tact 7b protein. In theory, cats with FIP should have

antibodies to the 7b protein, while cats exposed tothe common FECV would not have such antibodies.Unfortunately, the lack of 7b protein in WSU-79-1683is an artifact of that specific isolate and field strainsof FECV have an intact 7b gene.39 Therefore, the 7bantibody test is no more specific or sensitive thanthe indirect IFA test and this has been substantiatedby published studies.72,159

Because of the vagaries of FIP serology, FIP anti-body testing should also not be used as a means tocontrol or eliminate FIP from catteries. Vast amountsof money are spent each year by cattery owners onFIP testing. In almost all cases, the results are non-interpretable, even by the veterinarians ordering thetesting. Virtually all catteries having six to eight catsand an active breeding program will have FECV intheir cattery and 50e80% or more of their animalswill have titers of 1:100 or greater (usually1:25e1:1600). Cattery testing usually confirms whatis already known, that FECV is enzootic in the cattery.Antibody titers do not answer the critical questions:(1) do any of the cats actually have FIP, (2) are anyof the cats subclinically infected with FIPV, (3) willa particular cat develop FIP in the future, and (4)which cats are shedding FECV? Over- and mis-interpretation of various coronavirus antibody testsresult in considerable mortality from senseless eutha-nasia of healthy cats and undue emotional and finan-cial stress.

Serology has also been applied to the diagnosis ofthe neurologic form of FIP. Foley et al92 observedwhat they believed to be specific FCoV antibody pro-duction within the CNS. They found that the mostuseful ante-mortem indicators of neurologic FIPwere positive IgG anti-coronavirus antibody titer inthe CSF, high serum total protein concentration, andmagnetic resonance imaging (MRI) findings suggest-ing periventricular contrast enhancement, ventriculardilatation and hydrocephalus. The diagnostic valueof positive anti-coronavirus antibody titers in CSFwas later questioned by Boettcher et al.151 They col-lected CSF from four clinical groups: (1) cats withFIP involving the CNS (n¼ 10); (2) cats with FIP notinvolving the CNS (n¼ 13); (3) cats with CNS disor-ders caused by diseases other than FIP (n¼ 29); and(4) cats with diseases other than FIP and not involvingthe CNS (n¼ 15). CSF was evaluated for concentra-tions of erythrocytes, leukocytes, and total protein.Anti-coronavirus IgG was measured in CSF and se-rum by indirect IFA. CSF IgG (1:32e1:4096) was de-tected in 12 cats, including six cats with neurologicmanifestation of FIP, four cats with FIP not involvingthe CNS, and two cats with brain tumors. CSF IgGwas detected only in cats with correspondingly highserum IgG titers (1:4096e1:16,384) and significantlycorrelated with serum IgG titers. In another attemptto measure local CNS antibody production in catswith FIP, Steinberg et al160 used an albumin quotientand IgG index to determine whether proteins in theCSF were of blood or local origin. Neither the albumin

238 NC Pedersen

quotient nor IgG index identified a pattern consistentwith intrathecal IgG synthesis in cats with the CNSform of FIP. The conclusion of these various studiesis that coronavirus antibodies will enter the CSFwhen they are at very high levels in the serum; highserum titers are likely to be associated the dry formof FIP,15 most neurologic cases of FIP are of the drytype (Table 2); therefore, positive FCoV antibody titersin the CSF are likely to be associated with FIP.

Tests for the presence of FIPV RNA have been de-veloped, but suffer from many of the same weak-nesses as serology. Some of the more popular PCRbased tests are based on the lack of the 7b gene inthe FECV prototype WSU-79-1683 and its presencein the FIPV prototype WSU-79-1146. As mentionedabove, the lack of the 7b gene in this one isolate isan artifact, probably of tissue culture adaptation.Therefore, PCR tests based on genetic differences be-tween WSU-79-1146 and WSU-79-1683 are invalid.Moreover, PCR tests are notoriously susceptibility tolaboratory contamination with the DNA products ofamplification. Many commercial laboratories do theseprocedures in an uncontrolled manner. In an attemptto increase specificity and sensitivity, a PCR was de-veloped that would only detect forms of the viralRNA that were present during it replication stage.161

The rationale was that replicating forms of the viralRNA would only be found in the blood of cats withFIP. The test was designed to amplify subgenomicmRNA of the highly conserved M gene. The testwas applied to 424 healthy and 651 cats suspected ofhaving FIP. Almost one-half of the diseased catswere positive for the replicating form of FCoVmRNA in their peripheral blood cells, whereas only5% of healthy cats tested positive. Seventy-five of 81cats (93%) with post-mortem confirmed FIP testedpositive, whereas 17 non-FIP cases all tested negative.Such optimistic results were not found when the sametest was applied to another group of cats. In this study,26 blood samples were collected from 25 healthy catsand one cat with clinical signs suggestive of FIP,namely, fever, weight loss, enlarged abdomen, and as-cites.134 Blood samples were then tested for replicatingFCoV messenger RNA by the procedure of Simonset al.161 Fourteen (54%) of the cats were positive forFCoV including the cat with clinical disease, buta high rate of positivity was also observed amonghealthy cats, suggesting that feline coronavirusesmay be present in the blood samples from healthycats as well as cats with clinical FIP. This suggestionhas been confirmed for healthy cats by others.43,138

These conflicting findings call into question the valueof PCR for testing of blood. However, RT-PCR has ac-curately differentiated FIP effusions from effusions ofother causes.158 The problem of laboratory contamina-tion with PCR products can be virtually eliminated byusing a technique called real-time PCR (TaqMan).

Some clinicians and cattery owners will test forFCoV in feces by PCR in an attempt to identifyFECV carriers within groups of cats and eliminate

shedders from the environment as a means to preventFIP losses. The FECV is shed at high levels in the feces,but the carrier rate is 40e80% or higher in manymulti-cat populations and shedding is intermittentand infections recurrent (reviewed by Pedersen et al2008).25 PCR inhibitors have been described in feces,so it is important to purify viral RNA using certainmethods.162 Such testing is expensive, and onlya few laboratories can do it reliably on a commercialbasis. However, in the final analysis, it is virtually im-possible to maintain a group of cats free of the virus,even if the infection can be eliminated, without strictquarantine facilities and measures, as well restrictingthe movement of cats and people between cat popula-tions (see Treatment and prevention).

Rivalta’s test has touted for the diagnosis of wetFIP.158 A test tube is filled with distilled water andone drop of 98% acetic acid is added, followed bya drop of the peritoneal or pleural effusion. If thedrop dissipates in the solution, the test is negative,and if the drop retains its shape, the test is positive.A negative Rivalta’s test is reportedly 97% accuratein ruling out FIP, while a positive test is 86% accuratein ruling in FIP. This author sees no improvement indiagnostic value of this test over routine gross, micro-scopic, and biochemical analysis of the fluid, but it isa simple and cheap supportive test.

Several indirect tests, usually based on the levels ofcertain inflammatory proteins or byproducts (acutephase reactants), have been used to detect inflamma-tory conditions in cats, and in particular FIP(reviewed by Paltrinieri et al).50 Alpha-1-acidglycoprotein (AGP) is an acute phase protein that in-creases in concentration in infectious and inflamma-tory conditions. The serum and peritoneal fluidconcentrations of AGP has been found useful in thediagnosis of FIP.163,164 Amyloid A and AGP both in-crease a few hours after the inflammatory stimulusand remain elevated for as long as the inflammationpersists and have been evaluated in cats. SerumAGP levels have also been used to study FIP in groupsof cats.165 Serum AGP concentrations were observedto fluctuate over time in clinically healthy cats fromcatteries with the highest prevalence of FIP and signif-icantly increased just before an outbreak of FIP.Although increased levels of AGP and other inflam-matory proteins are particularly common in catswith FIP,164 they are not specific. Levels of AGP areusually high in cats with FIP and other inflammatorydiseases; moderate serum AGP levels (1.5e2 mg/ml)can discriminate cats with FIP from other diseasesthat have a low pretest probability, while high serumAGP levels (>1.5e3 mg/ml) are not usually seen incats with diseases other than FIP.164,166 However, thespecificity of increased AGP levels has been ques-tioned by others.166 The potential value of raisedlevels of the acute phase reactants, AGP and haptoglo-bin in the diagnosis of FIP was examined in cats withconfirmed FIP and in cats with other conditions.Levels of AGP greater than 1.5 g/l in serum, plasma


or effusion samples were found to be of value in dis-tinguishing field cases of FIP from cats with similarclinical signs and differentiated these two groups ofcats more effectively than the albumin:globulin ratio.The concentration of haptoglobin was higher in catswith FIP than in the group of healthy cats, but not suf-ficiently to be of diagnostic value. Serum samplesfrom FIV-infected cats also had significantly elevatedlevels of AGP and haptoglobin, illustrating that raisedlevels of these inflammatory proteins are not patho-gnomonic for FIP. Modifications of AGP have alsobeen tested to increase its specificity.167 AGP in hu-mans is heavily glycosylated and undergoes severalmodifications of its glycan moiety during acute andchronic inflammatory processes. Using human testmodifications, feline AGP had very little L-fucose res-idues on its surface and its branching degree was verylow in normal and in several pathological conditions.In contrast, feline AGP underwent several modifica-tions during acute FIP, including decreased expressionof both a (2e6)-linked and a (2e3)-linked sialic acid(76% and 44%, respectively) when compared to non-pathological feline AGP. The possible role of someacute phase (inflammatory) proteins and immuno-globulins in both the pathogenesis and diagnosis ofFIP was also reported by Giordano et al.168 Serum pro-tein electrophoresis and the concentration of hapto-globin (Hp), serum amyloid A (SAA), AGP, IgG andIgM were evaluated in healthy coronavirus-exposedcats and cats with FIP. The highest concentration ofacute phase proteins was detected in affected cats,confirming the role of these proteins in supportinga clinical diagnosis of FIP. Interestingly, healthy coro-navirus-exposed cats also had increased acute phaseproteins at the same time that members of the groupdeveloped FIP appeared in the group. However, thisincrease persisted only in cats that developed FIP. Itis apparent from these various studies that levels ofAGP, as well as other inflammatory proteins, changedramatically in FIP, but that none of these changesare definitive in their own right.

MRI has proven useful in confirming the presenceof inflammatory neurologic disease indicated by CSFanalysis.169 Fourteen cats with inflammatory diseasesaffecting the CNS were reviewed, including eightcats with FIP and two cats with toxoplasmosis. Abnor-malities affecting the CNS were observed in MR im-ages in 10/14 (71%) cats. Intracranial lesionsappeared as slightly hypointense foci in T1-weightedimages in two (14%) cats, as hyperintense foci inT2-weighted images in seven (50%) cats, and ashyperintense foci after intravenous administration ofa gadolinium-based contrast medium in 10 (71%)cats. In six cats with lesions in T1- and/or T2-weighted images, additional lesions were visible inT1-weighted images obtained after gadolinium-basedcontrast medium administration. In three cats, lesionswere visible only after contrast medium administra-tion. The MRI in this study did not detect all casesof CNS inflammation in a population of cats with

inflammatory CSF. However, it did add informationon the location of lesions, which can be important indifferentiating FIP from other inflammatory condi-tions. This substantiated the conclusions of an earlierMRI study on cats with neurologic FIP.92 Sixteen do-mestic cats with confirmed neurologic FIP and eightcontrol cats with non-neurologic FIP were studiedpre- and ante-mortem. The MRI imaging demon-strated periventricular contrast enhancement, ventric-ular dilatation, and hydrocephalus in cats withneurologic FIP.

The present gold standard for FIP diagnosis is im-munohistochemistry on effusions or lesions containinginfected macrophages. Monoclonal or polyclonal anti-bodies that are highly FCoV specific, and that will reactwell with formalin fixed tissues, have been used.170

Sections of lesions or cell pellets from ascetic or pleuralfluids can be directly examined for virus using fluores-cein or horseradish peroxidase polyclonal or monoclo-nal antibodies. Detection of coronavirus antigen in FIPeffusions is very specific, but less sensitive than detec-tion of viral antigens in characteristic FIP parenchymallesions. Hirschberger et al149 identified antigen in 34/49 confirmed FIP effusions, whereas 50 effusions dueto other causes were negative. The specificity of immu-nohistochemistry is a factor of the poly- or monoclonalantibodies used and the characteristic localization ofFIPV antigen within macrophages.146,170 The sensitiv-ity of the test depends on having infected macrophagesin the tissues or exudate cells on the slides. For this rea-son, random biopsy of liver or kidney of cats with FIPoften fail to yield FIPV antigen by immunohistochem-istry,171 and the same would be true of real-time PCR.Sensitivity is five to 10 times greater with fluoresceinthan horseradish peroxidase staining, but the latterhas the advantage of using formalin fixed and paraffinembedded tissues, while the former requires frozensections of tissues. Both can be used with equal easeon slides of cells harvested from effusions and acetonefixed.

Virus isolation in tissue culture has not been yetpossible for FECVs and is difficult for most fieldstrains of FIPV. Serotype II FIPVs seem to grow betterin tissue culture, and will often grow on both Crandellfeline kidney (CrFK) and Fcwf-4 cells. The latter cellline is of macrophage lineage13 and is the preferredcell line for the isolation of serotype I FIPVs.12,84 Thecost and low yield of cell culture isolation prohibitsits routine application to clinical diagnosis.

Following the introduction of tests for detection ofFeLV infection, one-third or more of cats with FIPwere found to have concomitant FeLV infections.120,172

With elimination of FeLV from many catteries and petcat households, and the steady decline in the inci-dence of FeLV in the entire cat population, the propor-tion of cats with FIP and concurrent FeLV infectionshas greatly decreased. At the present, virtually allcases of FIP in purebred cattery-bred cats are FeLVnegative, and FeLV infection is detected in 10% orless of domestic pet cats with FIP.

240 NC Pedersen

Pathologic featuresThe principle lesion of effusive FIP is the pyogranu-loma3,135,173 and the infection has been characterizedas a distinct form of vasculitis (ie, phlebitis).174 Thepyogranulomas of wet FIP are particularly prevalentin the abdomen and can literally cover the serosal sur-faces of abdominal organs (Fig 11). They range frommicroscopic to several millimeters in size, and cansometimes become confluent as a fibrinous sheet(Fig 12). The pyogranulomas of effusive FIP tend tofollow the course of the cranial mesenteric arteryand are, therefore, concentrated in the omentum andon serosal surfaces of the abdominal viscera (Table 1).The omentum is often thickened with edema, inflam-matory infiltrate and necrosis and retracted into a com-pact mass (Fig 12). Lesions are less common in thethorax and are centered on the pleura and pericar-dium (Table 1).

Pyogranulomas are made up of central aggregatesof macrophages adjacent to small venules and sur-rounded by a rich inflammatory exudate containingmainly neutrophils and macrophages with a scatteringof plasma cells and T-lymphocytes175 (Fig 13). Almostall macrophages within pyogranulomas of cats withwet FIP contain viral antigen at a high concentration(Fig 14). Considerable amounts of fibrin and protein-rich fluid are also deposited within and around thelesions and necrosis is often evident.135 Thoughthe pyogranulomatous process is usually surfaceoriented, focal lesions of phlebitis and a mixed inflam-matory-cell infiltrate may be seen deep in underlyingmuscle or organ parenchyma.

Fig 11. Abdominal viscera of a cat with effusive FIP. Theserosal surface of the intestines and spleen are coveredwith punctate, coalescing fibrinous plaques, the classicpyogranulomas of effusive FIP. Some peritoneal effusion isevident, although most has been removed.

Fig 12. Abdominal viscera of a cat with effusive FIP. Theserosal surface of the intestines and spleen is covered withpunctate, coalescing fibrinous plaques, the classic pyogranu-lomas of effusive FIP. Some peritoneal effusion remains,though most has been removed.

The lesions of dry FIP have been also referred to aspyogranulomas, but they more closely resemble clas-sical granulomas in their outward appearance, espe-cially those in the abdomen (Figs 3e6). Lesions ofdry FIP typically contain foci of macrophages aroundvessels, reminiscent of the lesions of effusive FIP.However, these foci are surrounded by dense infil-trates of lymphocytes (mainly B-cells) and plasmacells that extend into surrounding tissues, as seen inclassical granulomas (Fig 15). Edema, hyperemia, ne-crosis, fibrin deposition and protein exudation arenot as pronounced as in the pyogranulomatous le-sions of effusive FIP.175

Abdominal/pleural lesions of dry FIP have beendescribed as extraperitoneal,6 but probably originatedin the serosa or pleura and invaded along vessels intothe underlying parenchyma. Granulomas, especiallyin the abdomen, can vary in size from microscopicto several centimeters in diameter, and larger lesionscan be mistaken outwardly for cancers such as lym-phoma.6,115 Similar to the pyogranulomas of wet FIP,the granulomas of dry FIP are also most apt to occur

Fig 13. Abdominal viscera of a cat with effusive FIP. Theserosal surface of the intestines and spleen is covered withpunctate, coalescing fibrinous plaques, the classic pyogranu-lomas of effusive FIP. Some peritoneal effusion remains,though most has been removed.

Fig 14. Immunofluorescent antibody staining for felinecoronavirus antigen of a pyogranuloma in the serosa of thebladder. Antigen is concentrated in a high proportion ofthe macrophages within focal pyogranulomas.


in the abdomen (kidneys, mesenteric lymph nodes,liver, caecum/colon), and are uncommon in the chestcavity (Table 2) (Figs 3e7). Lesions in the eyes andCNS are much more frequently seen in cats with drythan wet FIP, and may occur in the absence or pres-ence of abdominal/thoracic lesions. Lesions withinthe brain, spinal cord and eyes tend to be smallerthan those found in the abdomen and are more similarin nature to the lesions of effusive FIP. CNS lesions arecentered in the meninges and ependyma and also ex-tend down into the underlying brain, spinal cord, orspinal nerves. Lesions are particularly common inthe posterior-ventral aspect of the brain.

Ocular lesions of FIP have been well studied.108 Themost common morphologic feature is a lymphocyticand plasmacytic anterior uveal infiltrate, which is ei-ther diffuse or nodular. This is manifested grossly byswelling and discoloration of the iris, with foci ofgranulomatous involvement sometimes observed.Keratic or mutton-fat precipitates are common onthe back sides of the cornea (Fig 9). These precipitatesare made up mainly of macrophages, other inflamma-tory cells, and fibrin that have exuded from the in-flamed uveal tract.

Lymphoid lesions are common in effusive and non-effusive FIP. Splenic enlargement may be due tohistiocytic and plasmacytic infltration of the red pulp,hyperplasia of lymphoid elements in the white pulp,necrotizing splenitis with fibrin deposition and poly-morphonuclear cell infltrates (more common in wetFIP), or by more organized pyogranulomatous reac-tions (more common in dry FIP). Gross lymph node en-largement is usually limited to thoracic and abdominalnodes with lesions resembling those described for thespleen. Kipar et al115,176 reported on lymphoid organlesions in cats with FIP and cats that had resisted exper-imental infection. In cats with FIP, the precardial medi-astinum (including site of the thymus) and mesenteric

lymph node parenchyma were often affected by gran-ulomatous-necrotizing processes. In general, lym-phoid tissues showed T- and B-cell depletion, oftenincluding massive to complete thymic involution or at-rophy. In some cases, the number of apoptotic lympho-cytes was increased in lymphoid follicles as well as inT-cell zones. The number of macrophages was in-creased in the splenic red pulp. In contrast, the FIPV-exposed cats without FIP generally showed lymphoidhyperplasia.

Fluorescent or immunoperoxidase antibody stain-ing of tissue sections from cats with both forms ofthe disease shows FIPV in the lesions.139,175 In effusiveFIP, a large amount of viral antigen is contained inphagocytic cells that make up a large part of the pyog-ranulomas (Fig 14).12,135,175,177 Less viral antigen ispresent in lesions of non-effusive FIP (Fig 16) andmore in lesions of effusive FIP (Fig 14).175 Lesions as-sociated with necrosis (mainly pyogranulomas), there-fore contain more virus than lesions without necrosis(mainly granulomas).175 Some plasma cells within le-sions are found to be producing coronavirus specificantibodies.175 Lesions, especially pyogranulomas ofeffusive FIP, also demonstrate large amounts of bound

Fig 15. Microscopic view of a granulomatous lesion froma cat with non-effusive FIP. Distinct focal accumulations ofmacrophages are seen just under the capsule of the kidney.These foci are surrounded by dense accumulations of lym-phocytes and plasma cells extending downward into theparenchyma.

Fig 17. Immunofluorescent antibody staining for feline IgG.Accumulations of IgG are seen both within and without macro-phages, mainly within and adjacent to pyrogranulomas. Someplasma cells also contain IgG.

242 NC Pedersen

complement and Ig, indicating their Arthus-likenature (Figs 17 and 18).

Experimental infectionExperimental infection studies were initially concen-trated on highly virulent strains of FIPV such as79-1146 (-Nor15 and -DF2 are identical to 79-1146).Studies with such strains give the impression that vir-tually all cats infected with FIPV will develop diseaseand die (Table 3). However, FIPV-79-1146 is atypical,not only in its virulence, but because of its ease ofgrowth in non-macrophage type cells such as CrFK fi-broblast cells and its CCV-like spike protein. Themost virulent serotype I isolates are FIPV-UCD1 andFIPV-TN406, but they are still less virulent than

Fig 16. Immunofluorescent staining of a foci from the kid-ney section shown in Fig 15. A much smaller porportion ofthe macrophages stain for feline coronavirus antigen thanin the pyrogranulomas of effusive FIP (Fig 14).

FIPV-79-1146. Interestingly, outbreaks of FIP with ex-tremely high morbidity and mortality are very uncom-mon in nature, suggesting that such highly virulentlaboratory strains as FIPV-79-1146, -UCD1 and-TN406 are atypical. This is borne out by experimentalinfection studies with yet other isolates of FIPV (Table3).84 Oral (PO) infection is least likely to cause disease;intratracheal/intranasal (IT/IN) infection is intermedi-ate, while intraperitoneal (IP) is the most efficient.

The first experimental evidence of the relationshipbetween FECVs and FIPVs came from studies of enzo-otic FECV infection in a colony of conventional catsused for FeLV research.119,121 The presence of anFCoV among this colony of cats did not become evi-dent until cats were horizontally exposed to FeLV bynatural contact and began to die of FIP.119,121 The pres-ence of two virtually indistinguishable coronavirusesamong these cats, one infecting the gut and causinglittle disease and another capable of causing FIP, waslater reported by Pedersen et al.27 This study resultedin the isolation of FECV-UCD. This relationship wasgiven further credence by an inadvertent introductionof an FECV, named FECV-Rogers-Morris or FECV-RM, into a barrier contained research colonies of spe-cific pathogen-free (SPF) cats.178 A few cases of FIPover several years among hundreds of kittens wasthe first indication that the colony was infected. TheFECV / FIPV mutation was further confirmed by in-fecting cats with long-term FIV infections and a cohortof non-FIV-infected siblings with FECV-RM.28 It washypothesized that an FIV immunocompromised hostwould replicate FECV at higher levels and lead toa greater chance of mutation, and that the mutant vi-rus would be more likely to manifest itself in an im-mune suppressed host. The FECV-RM did replicateto higher levels in the FIV-infected cats and about20% died of FIP within 8e10 weeks; no deaths fromFIP were seen in a cohort of non-FIV-infected siblings.This study established that the FIPV mutation is

Fig 18. Immunofluorescent antibody staining for feline C3.Complement staining of macrophages has a web-likeappearance.


frequent during primary FECV infection and that theimmune status of the cat is important in determiningthe clinical outcome of this mutation.

Conclusive studies on age susceptibility have notbeen undertaken. While most studies have used kit-tens or younger cats, there is no direct evidence thatkittens are more susceptible to infection. There is,however, indirect evidence. Cats vaccinated before16 weeks of age with an attenuated live FIP vaccinewere apparently more susceptible to challenge-exposure and less responsive to vaccination (NC Ped-ersen, UC Davis, personal observation, 2008) as de-duced from Christianson et al 179 and Gerber et al.180

Sixteen weeks is also a time when the immune systemof cats has reached a near-adult level of maturation tothis and other infections.181

Tissue culture-adapted strains of FIPV are oftenmixtures of viruses, and the predominant virus maynot be the most virulent. Variants may arise while inculture, or may be present in the cat prior to culturing(NC Pedersen, unpublished observation, 2008). ThreeFIPV strains from Japan, KUK-H, M91-266 andM91-267, were examined for biologic properties andvirulence.182 Both large and small plaque forming vi-ruses were identified in culture for each isolate.

Non-plaque purified virus of each strain inducedFIP in 50%, 67% and 89% of cats, respectively. Interest-ingly, only 29% of kittens inoculated with the largeplaques from one of these isolates (KUK-H) devel-oped FIP, while 80% of cats inoculated with the smallplaque form succumbed to FIP. The TN406 strain ofFIPV exists in a low and high passage form. The lowpassage form is very virulent, while the high passageform is non-virulent.83 This same reversion to non-vir-ulence occurred in a culture of FIPV-UCD1.183 The vir-ulent form in vivo may also not be the predominantform in vitro. For instance, it is difficult to re-isolateFIPV-UCD1 in a form that will grow in tissue culture,even though the cats were inoculated with tissue cul-ture-adapted virus (NC Pedersen, UC Davis, personalobservation, 2008). We have also studied cats with nat-urally acquired FIP that have had two genetic variantsin their tissues at the same time (NC Pedersen, UCDavis, personal observation, 2008).

The incubation period and clinical outcome of FIPVinfection depends both on the strain of the virus andimmunologic responsiveness of the host. Experimentalstudies using serotype I strains of FIPV, of low to mod-erate virulence, best reflect what happens in nature.Tissue culture strains of this type include FIPV-UCD1and FIPV-Black(low passage) (FIPV-TN406), which aremoderately virulent,12,83 and strains such as FIPV-UCD2, -UCD3, -UCD4 and -UCD8 that are even lessvirulent (Table 3).84,183 Even though low virulence lab-oratory strains of FIPVexist, most experimental studiesare with high virulence serotype I strains such as FIPV-UCD1, or serotype II strains such as FIPV-79-1146.Almost all cats inoculated with high virulence isolates,regardless of route, develop an initial fever spike at24e72 h, followed by a second and prolonged feverstarting 10e21 days or longer after inoculation, at thetime antibody appears. The subsequent disease courseis almost always of the effusive type and animals are ina terminal stage within 1e3 weeks. The sites of earlyreplication of FIPV were studied following oral inocu-lation of SPF cats with the Wellcome strain of FIPVgrown in cell cultures, a moderately virulent serotypeI virus.184 Viral antigen was first detected by immuno-fluorescence in the tonsils and small intestine within24 h of inoculation, and was later found in the caecum,colon, mesenteric lymph nodes and liver. Virus was re-covered from the oropharynx and the feces from asearly as the second day after inoculation, and sheddingcontinued until euthanasia. However, histologicalchanges in the gut did not appear until relatively latein the course of infection. Histological changes ap-peared to correlate with the appearance of outwardsigns of disease, which is often 2 weeks later than theappearance of virus in the feces and oropharynx.185

The appearance of a sustained fever usually coin-cides with the appearance of detectable levels ofFIPV antibodies by the indirect immunofluorescentantibody technique,135,139,177 which in na€ıve cats is10e16 days after infection. Early studies with FIPVstrains such as FIPV-UCD1 demonstrated

244 NC Pedersen

a phenomenon of immune enhancement. Cats previ-ously infected with an FECV, a non-disease causingFIPV, or cats passively given FCoV antibodies fromhealthy or FIP-diseased cats, developed a fever within48e72 h and rapidly died.139,140 The FIPV antigen in‘antibody primed’ cats was present in macrophagesand large mononuclear cells in the tracheobronchiallymph nodes, lungs, and trachea on postchallenge-exposure day (PCD) 2; in the liver and spleen onPCD 3; in the kidneys and omentum on PCD 4; andsubsequently in the nasal turbinates, thoracic and ab-dominal lymph nodes, thymus, bone marrow, parotidsalivary gland, eyes, and brain.135,177 Initial antiviralimmunofluorescence on PCD 2 coincided with the on-set of viremia and vascular lesions. Systemic lesionscharacterized by perivascular necrotizing pyogranu-lomatous inflammation, phlebitis and thrombosis, fi-brinous serositis, and generalized lymphoid necrosisdeveloped on PCD 3 and 4. Coronavirus-like particleswere observed by electron microscopy in the cytoplas-mic vacuoles or the smooth endoplasmic reticulum ofdegenerating macrophages in inflammatory lesions.The timing of viremia was delayed in kittens thatwere seronegative at the time of FIPV infection; antivi-ral immunofluorescence in tracheobronchial lymphnodes was first detected on PCD 5, and viremia oc-curred on PCD 6.135,177 Systemic necrotizing lesions,comparable with those observed in seropositive kit-tens on PCD 3 or 4, did not occur in seronegative kit-tens until PCD 13 or 16. In both groups of kittens,initial viral infection in regional lymphoreticular tis-sue was followed by viremia and infection of macro-phages in reticuloendothelial organs (liver, spleen,lymph nodes) and perivascular locations.

Experimental studies of FIPV infection have beenrecently advanced by the development of molecularclones of the virus. As in studies with other viruses,infectious clones allow genes to be altered, switchedand deleted to study their effects on infectivity, viru-lence and other disease propeties. Molecular cloneshave been constructed mainly by a process called re-verse genetics, and at least three cloned viruses pres-ently exist.186,187 Some of these cloned FIPVs, suchas one derived from the highly virulent FIPV-79-1146strain, retain their virulence when inoculated intoexperimental cats.186

Treatment and preventionIt must be stated at the onset that no treatment hasproven effective in curing cats of FIP, in spite of theclaims. Cats that develop FIP inevitably die of theirdisease in days, weeks or months. The reason for thesenumerous false claims is uncertain, but spontaneousremissions may account for at least some reportsand misdiagnosis for the remainder. Cats with ocularsigns and no other systemic manifestations of FIPhave occasionally gone into remission with just symp-tomatic treatment. Cats with chronic fever, enlargedmesenteric lymph nodes that were histologically

compatible with FIP, and high coronavirus titers,have gone into remission without treatment. Somecats without overt signs of FIP have demonstrated fi-brous lesions on the spleen and liver when necropsiedfor other reasons, indicating a previous bout of FIP.Small quiescent lesions in the spleen and mesentericlymph nodes have also been found in otherwisehealthy cats upon routine ovariohysterectomies.Therefore, spontaneous remissions occur and at leastsome of these natural responses may have fortuitouslycoincided with various treatments. Cures were first re-ported with tylosin and prednisolone,188 sparkinga decade of tylosin use for treatment of FIP. However,tylosin has no effect on FIP. Interestingly, a significantproportion of cats with FIP are still treated with an an-tibiotic of one type or the other. Some cats have goneinto remission after use of prednisolone and phenylal-anine mustard or cyclophosphamide.15 Another catwas successfully ‘treated’ with prednisolone and phe-nylalanine mustard.116 No immunosuppressive drugregimen has withstood the test of time.

A number of other equally dubious non-specifictreatments have been used for FIP, almost all with in-sufficient patient numbers, inadequate documentationof infection, or lack of essential placebo controls anddouble blinding.189 The FIPV is very sensitive to hu-man a and b interferons in vitro.190 Feline interferonomega also will inhibit FIPV in vitro and is commer-cially available in many countries.191 Feline interferonomega reportedly induced complete or partial remis-sions in two thirds of cats with FIP.192 However, ina larger and double blinded study, this treatmentwas found to be totally ineffective.193 Various immu-nosuppressants such as glucocorticoids and cyclo-phosphamide have been used, but these drugs mayprolong life but do not alter the fatal outcome.189 Im-munostimulants, megadoses of vitamins, and numer-ous nutriceuticals have also been advocated but foundto be without merit. Pentoxyfiline, a tumor necrosisfactor alpha (TNF-a) inhibitor,194 has been used oncats with FIP based on its benefit in treating sometypes of human and feline vasculitis.195 The TNF-ais upregulated in FIP183 and FIP is basically a vaculitis.However, this treatment has not proven beneficial onits own and has fallen from use. We have treatedone FIP cat with feline IFN-g and a TNF-a inhibitor(Pentoxyfiline) with no beneficial affect.

Effective vaccines have been as elusive as effectivetreatments. Pedersen196 hypothesized that the idealFIPV vaccine should contain a live virus that wouldpersist in the body in a subclinical state, inducinga state of premonition immunity. However, a success-ful vaccine against FIP has not been developed, eventhough most have been based on this hypothesis.Cats that survive infection with a progressively in-creasing dose of virus, starting with sublethal levels,appear to develop a type of immunity.83 However,this approach is not clinically applicable, because asmany cats die as become immune, and immunity ap-pears to be tenuous. Some cats that appear to have


resisted disease have developed FIP months or yearslater, indicating the persistence of subclinical infec-tions. This finding was supported by a subsequentstudy by Baldwin and Scott.197 They first immunizedcats intratracheally with a sublethal dose of virulentFIPV, followed by a high dose of temperature attenu-ated virus. Cats demonstrated immunity to an aerosolchallenge with highly virulent FIPV but residual le-sions were found upon necropsy examination sug-gesting that immunity was either partial or of thepremoniton type. Early attempts to use an attenuatedlive FIPV strain, FIPV-Blackhigh passage, to induce im-munity failed to provide protection and even causedimmune enhancement. A similar finding was reportedlater for a virulence attenuated strain of FIPV-UCD1.183 The phenomenon of FIPV vaccine inducedenhancement has been recently reviewed and is a com-mon theme in many vaccine approaches.198

A temperature sensitive mutant of FIPV-79-1146,administered intranasally, was later developed andcorporate studies showed a high degree of efficacyagainst challenge with highly virulent FIPV-79-1146.179,180 Immunity was ascribed to a local IgAresponse and systemic cellular response measuredby FIPV-induced lymphocyte proliferation. This vac-cine is given as two intranasal doses, 3 or more weeksapart, starting at 16 weeks or older. Presumably, effi-cacy could not be shown for cats vaccinated and/orchallenge-exposed at a younger age. This commercialvaccine was studied in the field in a large singlebuilding shelter housing a thousand or morecats.199,200 Five hundred FIV/FeLV negative, FCoV an-tibody negative cats were divided into two groupsprior to being put into this environment. One-halfgot the intranasal live vaccine and one-half wassham vaccinated. The cats were than followed for 16months. Overall deaths in both groups were thesame. However, two cats in the vaccinated group de-veloped FIP during the 16 months and eight non-vac-cinated cats succumbed to the disease. Protection inthis study was not convincing and the validity of us-ing the vaccine only on coronavirus negative cats andmeasuring protection over a limited period of timewere questioned.201 Hoskins et al202 vaccinated 20SPF cats intranasally at 16 and 19 weeks of age, andadded 20 non-vaccinated controls. All 40 cats werethen challenge-exposed orally at 23 weeks of agewith FIPV-DF2 (ie, FIPV-79-1146). One vaccinatedand five non-vaccinated cats developed effusive FIPand were euthanased. However, six of the remainingvaccinates and 12 of the remaining controls had histo-logic evidence of FIP when necropsied 8 weeks later.Wolf 201 criticized this study for using a homologouschallenge strain and for only following the cats for 8weeks. In a second field study, the vaccine failed toprevent FIP in cattery kittens immunized in the pres-ence of pre-existing FECV antibodies.203 The issues ofserotype specificity (ie, the vaccine is against serotypeII, while most field isolates are serotype I), interfer-ence with maternal antibodies, the need to withhold

vaccination until at least 16 weeks, and the relativelypoor efficacy are serious limitations. This vaccine isstill marketed, and is used by some veterinary practi-tioners. It has proven safe over a long period of useand does not appear to enhance disease.200 However,its degree of effectiveness in the general cat popula-tion, where the incidence of FIP is very low, and incatteries and multi-cat households where virtuallyall 16 week olds are already seropositive and there-fore non-responsive, is highly questionable. Serotypecross-reactivity is also an issue. Cross protection be-tween closely related viruses, such as TGE virus orCCV, and FIPV either does not occur or is weak.204e206

Cross protection between various isolates of the sameserotype may also vary.207 Cats infected with FIPV-UCD2 neither became ill nor developed protectiveimmunity against challenge-exposure with more viru-lent strains. The FIPV-UCD3 recovered cats tend to beimmune to challenge with FIPV-UCD1, while FIPV-UCD4 recovered cats do not have good protective im-munity to FIPV-UCD1. These are all serotype I FIPVs.

There have been a number of different potentialvaccines based on recombinant technology describedin the literature since 1990. A genetically engineeredvaccinia virus expressing the envelope protein ofFIPV has been tested208; it enhanced virulent FIPV in-fection rather than protecting cats. Wasmoen et al209

developed a recombinant raccoon poxvirus that ex-pressed the gene encoding the nucleocapsid proteinof FIPV and used it as a vaccine. Cats were parenter-ally or orally vaccinated twice, 3 weeks apart, andthen infected orally with FECV to sensitize them tovirulent FIPV. Three weeks later, the cats were orallychallenged with FIPV; 4/5 (80%) of the non-vaccinated controls succumbed to FIP, while 0/5 catsvaccinated subcutaneously with the raccoon recombi-nant poxvirus developed FIP. One of the five cats inthis latter group died following a second FIPV chal-lenge, suggesting immunity was of the premonitiontype. Vaccination induced serum antibody responsesto FIPV nucleocapsid protein, but they were not neu-tralizing. This suggested that the protective immuneresponse involved a mechanism other than neutraliz-ing antibodies to the spike glycoprotein or that FECVboosting was important for immunity. The subsequentfate of this vaccine approach is unknown. In anotherstudy, the spike gene of FECV-79-1683 was clonedinto vaccinia virus and used as an expression vectorvaccine.210 Vaccinated cats developed neutralizing an-tibodies, but kittens vaccinated with the recombinantvaccinia virus died earlier than animals vaccinatedwith native vaccinia virus alone. A plasmid vaccinecontaining the membrane (M) and nucleocapsid (N)genes failed to protect kittens against a challenge in-fection with FIPV,211 while a vaccinia virus expressingthe M protein of FIPV failed to protect cats againstFIPV-79-1146.212 A vaccine made up of a cell lysateof recombinant baculovirus-expressed FIPV-KU-2 nu-cleocapsid protein protected 6/8 vaccinates versus1/8 of cats immunized with a control antigen.213

246 NC Pedersen

Haijema et al214 reported successful vaccination ofcats against a homologous virus challenge of catswith a live FIPV that has been attenuated by deletionof the 3aec, and to lesser extent 7ab gene clusters. Nofollow-up research has been reported on these lattertwo vaccine approaches.

Antiviral drug therapy has become increasingly im-portant in a number of viral diseases such as human im-munodeficiency virus and human influenza virus.Ribavarin, which is strongly inhibitory of FIPVin vitro,215

is not effective in vivo.216 This dichotomy between in vitroand in vivo inhibition is mirrored by recent experiencewith the human severe acute respiratory syndrome(SARS) cornavirus.217 Sixteen different compoundswere subsequently tested for their ability to inhibitFCoV replication in vitro; six of the compounds werefound to reduce the virus titers by 0.401 to 0.833 log10(P< 0.05), using cytopathic effect as the endpoint.218

The 50% effective dose (ED50) levels for these six com-pounds was pyrazofuin> 6-azauridine> 3-deazaguano-sine> hygromycin B> fusidic acid> dipyridamole,respectively. Compounds that had no statistically signifi-cant effect on FIPVin the same assay were caffeic acid, car-bodine, 3-deazauridine, 5-fluoroorotic acid,5-fluorouracil, D(þ)glucosamine, indomethacin, D-peni-cillamine, rhodamine and taurine. No subsequent invivo studies were reported on the compounds that wereinhibitory in vitro. The more recent appearance of SARSin humans has greatly increased interest and knowledgeof other potential anti-coronaviral drugs. A wide varietyof pyridine N-oxide derivatives have been found to be in-hibitory against FCoV (FIPV strain) and human SARS-CoV (Frankfurt strain-1) in CrFK and simian kidney(Vero) cell cultures, respectively.219 The most selectivecompounds were active at a high microgram per literrange and non-toxic at 50e100 mg/l. One of the most ac-tive and selective compounds tested was shown to inhibitFCoV replication at the transcriptional level. Balzarini etal220 also demonstrated inhibition of FIP and humanSARS coronaviruses by semisynthetic derivatives of gly-copeptide antibiotics including vancomycin, eremomy-cin, teicoplanin, ristocetin A and DA-40926. The 50%effective concentrations (EC50) for the most active deriva-tives were in the lower micromolar range. The EC50

values differed between FIPV and SARS-CoV, indicatingconsiderable differences in species specificity. No invivo studies on any of these FCoV inhibitory compoundshave been reported to date. Indomethacin has recentlybeen found to have a potent inhibitory effect on replica-tion of both SARS coronavirus and CCV, but no studieswere done on feline coronaviruses by this group.221 How-ever, an earlier study found indomethacin to be ineffec-tive against FIPV.218

The incidence of FIP within catteries can be de-creased by proper management.222 Mortality tendsto increase as the population of animals, especially kit-tens, increases. The effect of overcrowding on FIP in-cidence has been particularly evident in animalshelters. The losses caused by FIP inevitably increasewhen shelters are overwhelmed with cats, usually

during kitten season, and decrease when overcrowd-ing is controlled; longer shelter stays compound thesituation (K Hurley, UC Davis, unpublished informa-tion, 2008). Losses from FIP are also proportional tothe severity of other kittenhood diseases, includingthose caused by herpesvirus, calicivirus, chlamydo-phila, mycoplasma, dermatophytes, and numerousother bacterial, viral, fungal, protozoal, and parasiticskin, upper respiratory and enteric infections. Kittenskept in crowded catteries or shelters with a large num-ber of other young animals suffer greatly from concur-rent diseases. These diseases stress the kittens’immune system and cause a decrease in growth rateand an increase in susceptibility to disease in general.FeLV infection, a bane of many catteries, shelters andother multi-cat environments in the past, was once thesingle most powerful potentiator of FIP in cats. Fortu-nately, FeLV infection has been pushed back into na-ture in most westernized countries by an aggressivecampaign of testing, eradication/isolation of sheddersand vaccination, and is no longer a problem in mostmulti-cat environments.26 Genetics also play an im-portant role in FIP. Genetic resistance/susceptibilityfactors also affect FIP incidence and obviously havetheir strongest influence in catteries. Death lossesfrom FIP can sometimes be traced to certain breed-ings, and further breeding of pairs that produced af-fected kittens should be avoided. Toms that producelitters that suffer from FIP should be especiallyavoided, because a given male affects far more littersand individuals than a queen.52 Breeding practices incatteries often result in an abundance of youngerbreeding animals. The incidence of FIP in a staticmulti-cat household is known to decrease over 3years.44 This may be associated with age resistanceto chronic or intermittent FECV shedding; an immu-nity of sorts develops over a period of many monthsor years.25 Age resistance over this period of time isseen with several chronic type infections, such as her-pesvirus, feline calicivirus FeLV and ringworm. This iswhy catteries or shelters that keep mainly older catshave much less infectious disease problems than cat-teries with younger breeding or sheltered animals.

The practice of isolating queens and early weaning ofkittens has been advocated for the control of FIP in cat-teries79,260 and is widely practiced. Cattery kittens arenot infected with FECV until around 9e10 weeks ofage.25 If queens are isolated just prior to parturition,kept in strict quarantine from other sources of FECVex-posure, removed from their kittens at the earliest possi-ble time (4e6 weeks after birth), and kittens maintainedthereafter in isolation, it is possible to prevent kittensfrom becoming infected with FECV and hence preventFIP. However, this practice works best on small catteriesand requires very elaborate quarantine facilities andstrict control of exposure. Gross and microscopic litterdust contains high levels of virus. Such litter dust is om-nipresent and easily spread through the air, caging,pans, accessories and on clothing. Therefore, strictquarantine involves separate quarters and care givers,


separate litter, food and water pans, separate air space,and change-in/change-out protective clothing includ-ing shoes/shoe covers/boots, foot baths, etc. Isolationfacilities cannot consist of just another room ina home, with no other special precautions. If specificallydesigned isolation rooms are not available, an alterna-tive would be to transfer kittens at early weaning to dis-tant facilities (or homes) devoid of other cats. Theproblems with the ‘isolation of pregnant queens andearly weaning program’ are several fold. First, the levelof quarantine required to prevent contamination re-quires expensive facilities and great costs to operatethem. Second, what should be done with coronavirusfree kittens once they are produced? Infection with co-ronavirus is almost assured once these kittens are rein-troduced into the normal environment and expected tocarry out normal activities, such as relocation to newhomes, showing and breeding. However, there remainsone advantage of a successful isolation and early wean-ing program, and that is to delay infection until after thekittens are 16 weeks of age or older. This is a time whenthe immune system began to attain adult levels of activ-ity and when kittens become more resistant to clinicaldisease of all sorts, including FIP.

Catteries that undertake isolation and early wean-ing programs should determine whether it is workinggiven their facilities and level of husbandry. If kittenscoming out an early weaning program are found to beseropositive at 16 weeks of age and/or sheddingFECV from their feces, the program has failed. Givencosts and logistical problems, this author does nothighly recommend isolation and early weaning pro-grams, especially for larger catteries (over six breed-ing cats). The most cost-effective means to controlFIP losses are fivefold: (1) eliminate overcrowding,and if possible, maintain no more than six breedinganimals; (2) maintain a larger proportion of oldercats (3 years of age and older); (3) manage feces tomouth spread by properly managing litter boxes, litterreplacement, and gross and microscopic spread of lit-ter and litter dust; (4) be highly selective in the breed-ing program and produce a minimum number oflitters, and (5) control genetics by not using any tom(at a minimum) that has produced kittens that havedeveloped FIP, or preferably, use neither toms norqueens that have produced kittens that died fromFIP. These steps are far more effective than improperlyconducted isolation and early weaning programs andcan be easily implemented.

Infection and immunityImmunity to FIPV infection is a fascinating topic fortwo reasons. First, it appears that humoral immunityis not important in protection, but may actually partic-ipate in the disease process. Second, protective immu-nity appears to be largely cell mediated and may beof an infection or premonition type. Third, the typeand strength of immunity appears to determine theform that FIPV infection will take. It is reasonable to

assume that strong humoral immunity with veryweak or non-existent cellular immunity will lead to ef-fusive FIP, humoral immunity with intermediate cellu-lar immunity will manifest as non-effusive FIP, andhumoral immunity with strong cellular immunitywill prevent the disease. Although these are mainlyworking hypotheses, there is considerable evidencefor each.

Antibodies to FIPV antigens participate in two dif-ferent immune processes, neither of which correlateswith immunity. The first process is an Arthus-type hy-persensitity reaction centered on small venules, whilethe second process is an antibody-mediated enhance-ment of viral uptake and replication by macrophages.All three components of an Arthus-type reaction, anti-gen, complement, and antibody are present at highlevels within lesions, especially in those cats withthe effusive FIP (Figs 14, 17, 18). Arthus reactionsare characterized by vasculitis, edema, inflamma-tory-cell migration, and necrosis. Antibody-mediatedenhancement of disease was first described by Peder-sen and Boyle139 and confirmed by Weiss and Scott.140

They noted that the appearance of antibodies and dis-ease signs were always linked. Seronegative (na€ıve)cats developed the earliest signs of FIP within 10e16days or more after experimental infection, and thetiming of disease always coincided with appearanceof antibodies. Healthy FCoV antibody positive (ie,FECV exposed) cats were then challenge-exposedwith FIPV; fever and disease signs occurred within24e72 h, instead of the usual 10e21 days or more. Ifcoronavirus seronegative cats were then passively im-munized with FECV induced antibodies in the form ofserum or immunoglobulin, they also manifested accel-erated disease when challenge-exposed with FIPV.The phenomenon of antibody-mediated enhancementwas later confirmed to involve antibodies to specificepitopes on the spike protein; these antibodies alsofunctioned as neutralizing antibodies in vitro.223,224

Macrophage infection was greatly enhanced in vitroby the addition of immune serum.225,226 This led toa plausible theory in which antibody and complementcoated viral particles were taken up through fc recep-tors on macrophages by a process of endocytosis.224

This placed the FIPV in its host cell of choice, andfrom then on the virus is spread to other places ofthe body by macrophage migration. Antibody-medi-ated enhancement of macrophage infection appearsto be serotype specific, ie, antibody to serotype IIFIPVs only enhance macrophage infection with sero-type II FIPVs.226

The antibody-mediated immunopathogenesis in ef-fusive FIP, and the role of macrophages in disease, ap-pears similar to that described for the Denguehemorrhagic shock syndrome.227e232 The hallmarkof Dengue hemorrhagic fever and effusive FIP is anArthus-type reaction. Viral laden macrophages, viralparticles, and viral proteins accumulate around smallvenules. These localized sources of antigen, as well asantigens in the blood, can react with complement to

248 NC Pedersen

form immune complexes. In turn, immune complexesdeposit around small venules and trigger the releaseof specific macrophage factors that cause tissuedamage.13,135,139,143,144,177,225 Paltrinieri et al146 alsodemonstrated that FIP lesions contained many vi-rus-infected macrophages and that extracellular viralantigens were also detectable in the foci along withnecrosis. Necrosis is one hallmark of Arthus-typevasculitis.

Although a great deal has been made of antibody-mediated enhancement in FIP, it may not be as impor-tant in nature as in the laboratory.44 Furthermore, thephenomenon has been described mainly between cer-tain FCoV isolates, such as FIPV-UCD1 and FECV-UCD,139 FIPV-UCD1 and unknown FECV strains,140

and FIPV-UCD8 and FECV-RM.183 We have not ob-served this phenomenon with other field strains ofFIPV and FECV-RM (NC Pedersen, UC Davis, per-sonal observation, 2008). In general, it appears thatthe more virulent the FIPV isolate, the more likely itis to respond to antibody enhancement.

Antibodies to the spike protein may also not alwaysbe harmful. Western blot antibody responses againstthe three major structural proteins have been mea-sured in cats naturally infected with FECV and com-pared to disease outcome, i.e., recovery, chronicasymptomatic carrier state, or FIP.233 The cats withchronic asymptomatic infections developed a spikeglycoprotein relative to anti-M glycoprotein antibodyresponse that was at least 30-fold higher than catswith chronic asymptomatic or symptomatic infec-tions. It was concluded that antibodies to the spikeprotein favored clearance of the virus after natural in-fection and were not a risk factor for the establishmentof a chronic infection.

Although many studies support the special role ofantibodies in FIP and in the form it takes, there are rel-atively few mechanistic studies on actual B-cell re-sponses in FIP. Takano et al234 demonstrated fivefactors associated with B-cell activation in FIP. Theyfound that the ratio of peripheral blood sIg(þ)CD21(�) B-cells was higher in cats with FIP than inSPF cats; that the albumin-to-globulin ratio has nega-tive correlation with the ratio of peripheral bloodsIg(þ) CD21(�) B-cell; that cells strongly expressingmRNA of the plasma cell master gene, B-lympho-cyte-induced maturation protein 1 (Blimp-1), were in-creased in peripheral blood in cats with FIP; thatmRNA expression of B-cell differentiation/survivalfactors, IL-6, CD40 ligand, and B-cell-activating factorbelonging to the TNF family (BAFF), were enhancedin macrophages in cats with FIP; and that mRNAsof these B-cell differentiation/survival factors wereover-expressed in antibody-dependent enhancement(ADE)-induced macrophages. In this scenario, virus-infected macrophages overproduce B-cell differentia-tion/survival factors, which in turn act to promoteB-cell differentiation into plasma cells. This samegroup235 also investigated key steps in ADE in FIPsuch as involvement of the virus receptor and the

requirement for acidification of the endosome for cel-lular invasion. They concluded that aminiopeptidatseN was not required for antibody enhancement. How-ever, acidification of the endosome was necessary forcellular invasion by FIPV, regardless of the presenceor absence of antibodies.

If humoral immunity does not protect cats againstvirulent strains of FIPV, what then is the nature ofFIPV immunity? It has been postulated that immu-nity to FIPV is largely cell mediated.83 Reasons forthis assumption include the following: (1) The non-effusive form of FIP resembles tuberculosis anddeep mycotic infections of humans and animals,and immunity to these infections is known to involvemainly cellular mechanisms; (2) the lesions of dry FIPresemble type IV hypersensitivity reactions withcentral macrophages containing relatively smallamounts of virus and surrounded by dense infiltratesof plasma cells and CD4þ lymphocytes,236 while thepyogranulomas of wet FIP are aggregates of macro-phages stuffed with virus and surrounded mainlyby neutrophils and edema; (3) the clinical incidenceof FIP can be increased greatly by concurrent FeLVinfection,119,121,172,237 and FeLV infection is a potentsuppressant of cellular immunity and T-cell mediatedhumoral immunity237; (4) immunity to FIP cannot betransferred passively with hyperimmune serum, re-gardless of whether the serum is taken from FECV-infected cats or cats that survived an FIPVchallenge83,139; (5) a delayed-type hypersensitivity re-action to FIP antigens can be evoked in the conjunc-tiva of FIP immune cats and peripheral bloodlymphocytes of recovered cats respond in vitro toFIPV antigen,119 and (6) cats are known to carryFIPV as a latent or sequestered infection, and this in-fection can be reactived by infecting such carrierswith FeLV, but not with methylprednisoloneacetate.119 A carrier state of the latter type is knownto exist in infections like tuberculosis, blastomycosis,histoplasmosis, and coccidoidomycosis and immu-nity is sustained in these situations by the persistenceof small numbers of organisms in mesenteric or bron-chial lymph nodes. This type of immunity, called pre-monition or infection immunity, persists only as longas intracellular pathogens persist in a reactivatableform.119 This same phenomenon may have been ob-served with cats immunized with a virulence modi-fied live FIPV.197 Immunized cats showed nooutward signs of disease when challenge-exposedwith virulent FIPV over 4 months later, but residuallesions of FIP were found histologically. Stoddart etal185 studied virus shedding and immune responsesin eight SPF cats inoculated orally or intraperitone-ally with cell culture-adapted FIPV. Virus was recov-ered from both feces and oropharyngeal swabs forapproximately 2 weeks after inoculation and beforeclinical signs of disease developed. All eight catsshowed a profound suppression in their peripheralblood lymphocyte blastogenic responses to Con A,which recovered to pre-inoculation levels in only


two of the cats, one of which survived. Both Con A re-active cats also had lymphocyte blastogenic responsesto FIPVantigen on the 21st day after infection, with thegreatest response being in the survivor. Cats dying ofFIP after 16e18 days developed no response to FIPVantigen. Antibody titers, measured by immunofluores-cence and by virus neutralization, rose rapidly to veryhigh levels in all cats, regardless of the route ofinoculation. This study also supported the role of cellmediated immunity, rather than antibodies in FIP.

DeGroot-Mijnes et al238 put forward a unified con-cept of T-cell responses in FIP. They postulated that vi-rus-induced T-cell depletion and the antiviral T-cellresponse are opposing forces and that the efficacy ofearly T-cell responses critically determines the outcomeof the infection. If the virus wins out, FIP will result,while if the host wins out no disease will develop.They observed a consistent rise in the levels of viralRNA in the blood of cats with end-stage FIP, indicatingfatal disease is directly related to a loss of immune con-trol and unchecked viral replication. Paltrinieri et al239

analyzed lymphocte subsets CD5, CD4, CD8, CD21markers by flow cytometry. Cats that were recently in-fected with FECV that did not develop FIP had a tran-sient increase in T-cells. The FECV-infected cats witha high prevalence of FIP had a moderate and persistentdecrease in T-cell subsets, while cats with FIP had se-vere decreases in all lymphocyte subsets. Of course, itcan be argued in a chicken or egg manner that it is thelevel of viremia that determines the outcome.

There are additional studies that support a role forcellular immunity, possibly of the premonition type,in FIP resistance and susceptibility. It is possible to im-munize cats against FIPV by starting with small sub-infectious doses of virus and gradually increasingthe dose over time.83 Cats immunized in this mannercan eventually resist infection with dosages of virusthat would be lethal to na€ıve cats. However, this im-munity can be overcome with time and by very largeamounts of virulent virus. Moreover, one cat immu-nized in this manner, and repeatedly immune toover 1000 cat infectious doses of virulent FIPV over3 or more years eventually succumbed to non-effusiveFIP (NC Pedersen, UC Davis, personal observation,2008). The author also observed a naturally infectedcat that responded to non-specific treatment andwent into a sustained remission, only to re-developFIP 2 years later and die. Premonition immunitymay also be passed from FIPV immune queens to theirkittens.119 Kittens born to such queens frequentlyshowed a typical antibody pattern of both maternalimmunoglobulin transfer and maternal infection. Ma-ternal antibodies to FCoV were noted in kitten’s bloodshortly after birth and declined over the next severalweeks, and than spontaneously increased. The firstpeak of antibodies corresponded to the acquisitionand loss of maternal antibodies, while the secondpeak resulted from an active immune response bythe kittens. Five of six kittens that were challengedwith FIPV at 8e10 weeks of age resisted FIPV-UCD1

challenge-exposure, while 3/5 of these surviving kit-tens succumbed to an accelerated form of effusiveFIP when infected at 22 weeks of age.119 FIP was reac-tivated by a secondary FeLV infection in almost allcats within the first 2 months after resisting diseasefrom a primary FIPV infection, but not after 4e6months.119 This situation resembles that seen in latentFeLV infections.240 Latency in FeLV infection is merelyan extension of the recovery process and usually re-solves within 6 months of the disappearance of vire-mia. This appears to be characteristic of manyinfectious diseases in which cellular immunity is im-portant for recovery; the longer the period after recov-ery, the more difficult it is to demonstrate persistenceof the agent. Immunity to many infections, includingFIP, must be a slow, ongoing process that takes weeks,months or years. In some individuals, the agent maypersist for a lifetime.

Several studies of FIP have attempted to link the oc-currence of disease to certain Th1 profiles of cytokine ex-pression. Kiss et al,183 in a small pilot type study,suggested that cats that developed either wet or dryFIP after experimental infection expressed high levelsof mRNA for TNF-a and low levels of mRNA for inter-feron gamma, while cats that resisted disease had the op-posite response. The negative role of TNF-a in FIP wassupported by studies that linked TNF-a released frommacrophages to apoptosis of CD4þ and CD8þ T-cellsand the resultant lymphopenia.241 The TNF-a may alsoupregulate the feline aminopeptidase-N receptor in fe-line macrophages,242 making them more susceptible toinfection. A similar type study on wet FIP measuredthe cellular expression of interferon gamma by varioustypes of lesional inflammatory cells.243 Cells within le-sions included B-cells and plasma cells as well asCD4þ and CD8þ T-cells. However, the macrophageseemed to be the key cellular element in both wet anddry forms of FIP. Interferon-gamma expression was up-regulated within macrophages in the inflammatory le-sions suggesting a local activation. Dean et al244

attempted to correlate histologic changes in lymphoidtissues, such as lymphoid hyperplasia, lymphoid deple-tion, histiocytosis, and granuloma formation, with cyto-kine expression. Viral RNA was found in 76% ofmediastinal lymph nodes, spleen and mesenteric lymphnode as compared to 27% of popliteal lymph node, cer-vical lymph node and femoral bone marrow. All tissuespositive for virus replication also demonstrated lym-phoid depletion. Affected tissues had lower levels ofIL-4 and IL-12-p40 mRNA and higher levels of IL-10mRNA, but no differences in IFN-g or TNF-a mRNA.The TNF-a protein expression was greater in affected tis-sues and associated with a shift in the source of TNF-a from macrophages to lymphocytes. Therefore, FIPVreplication, lymphocyte depletion in tissues, and alter-ations in cytokine transcription and translation wereco-localized to the same tissues. A possible role forTNF-a in the previously described FIPV-induced lym-phocyte apoptosis was also suggested. Cytokine re-sponses have been compared between cats with

250 NC Pedersen

neurological (focal) or generalized FIP, on the rationalethat localized neurologic disease is more homogeneous,ie, stereotypic, than generalized FIP.245 IL-1b IL-6, IL-12,IL-18, TNF-a, macrophage inhibitory protein (MIP)-1a,and RANTES showed no up-regulation in the brains ofcontrol cats, moderate up-regulation cats with neuro-logic FIP, and high up-regulation in cats with general-ized FIP. IFN-g gene transcription was upregulated incats with generalized FIP and slightly down-regulatedin cats with neurologic FIP; variance was high in catswith generalized FIP and much lower in cats with neuro-logic FIP. A large component of the variance (91.7%) wasaccounted for by IL-6, MIP-1a, and RANTES levels.

In another cytokine study, a group of 8-week-oldkittens were pre-immunized with two recombinantFIPV spike proteins and then challenged intranasallywith FIPV-79-1146.246 All of the kittens developed con-firmed FIP. The recombinant proteins induced little orno specific antibody response prior to challenge andfailed to enhance or inhibit the course of disease com-pared to controls. Mitogen stimulated PBMCs col-lected after 1 week showed small increases in theexpression of IL-6 and interferon-gamma mRNA,which correlated with an initial febrile response.Thereafter, the expression of IL-6 mRNA remainedunaltered until signs of FIP developed. At this point,mRNA levels of IL-2, IL-4, IL-10, IL-12 and interferongamma were markedly depressed.

Kipar et al51 compared cytokine mRNA levels in var-ious lymphoid tissues of cats with exposure to FECV,with or without developing FIP. They noted that lym-phoid tissues were normal or hyperplastic in healthyFECV exposed cats, but that both B- and T-cell deple-tion was present in cats that developed FIP. Therefore,they compared functional changes in bone marrowand mesenteric lymph nodes in healthy FECV exposedcats and cats that developed FIP. The FECV-infectedcats that did not develop FIP exhibited significantlyhigher IL-10 levels in the spleen and significantly lowerlevels of IL-6, G- and M-CSF in mesenteric lymph no-des. In contrast, IL-12-p40 levels were significantlylower in lymphatic tissues in cats with FIP comparedto FECV-infected cats without FIP. Cats with FIP hadsignificantly higher IL-1b levels and lower TNF-a levelsin mesenteric lymph nodes and lower M-CSF levels inthe spleen. They concluded that FECVexposed cats thatdo not develop FIP are able to make protective immu-nity and avoid excessive macrophage activation andFIP, possibly by up-regulation of IL-10. The develop-ment of FIP may also be associated with a lack ofIL-12 production, which inhibits an effective cellularimmune response and allows for monocyte/macro-phage activation. Rapid activation of the p38 MAPKpathway has been demonstrated in FIPVinfected bloodmononuclear cells from SPF cats.247 The MAPKactivation led to the production of pro-inflammatorycytokines such as TNF-a and IL-1b. The TNF-a produc-tion was inhibited by pyridinyl imidazole inhibitors.

Recent studies also measured the concentration of in-terferon gamma in FECV exposed cats that developed

FIP.49 They found that healthy FECVexposed cats livingin catteries with a high incidence of FIP had the highestserum interferon-gamma concentrations. However, se-rum concentrations of interferon gamma were not sig-nificantly different in cats that developed FIP than inhealthy FECVexposed cats in catteries with a low prev-alence of FIP. The concentration of interferon gammawas significantly higher in the effusions than in the se-rum of cats with FIP, probably due to interferon-gammaproduction within lesions. A similar study measuredcytokine mRNA248; when compared with SPF cats, cy-tokine production was upregulated in non-symptom-atic FECV exposed cats in catteries with no FIP. Theappearance FIP in the cattery was associated with an in-creased expression of cytokines, in particular there wasan increased production of IL-1b and interferongamma, suggesting that these cytokines helped protectagainst FIP. This hypothesis was supported by the lowlevels of interferon gamma recorded in blood from catswith FIP. The most compelling corroborative evidencefor the role of interferon gamma in protection againstFIP in cats comes from interferon-gamma gene knock-out mice. Kyuwa et al65,66 described granulomatousperitonitis and pleuritis virtually identical to that ofFIP in knockout mice infected with a laboratory strainof mouse hepatitis virus (a murine coronavirus). Anidentical disease was observed in this same strain ofknockout mice that were naturally exposed to a fieldstrain of murine coronavirus.64

Both genetic52 and host28 factors appear to havea strong influence on resistance of susceptibility toFIPV in nature. Studies of FIP in a number of purebredPersian catteries showed that susceptibility is heritableand accounts for about 50% of the disease incidence.52

Addie et al48 attempted to link disease susceptibilityto certain alleles within the DRB gene of the felineMHC (feline leucocyte antigen or FLA). Individualcats were shown to have between two and six FLA-DRB alleles, but no specific allele appeared to be associ-ated with either the development of FIP, resistance toFCoV in general, or to FECV carrier status. However,this was only a pilot study and lacked the numbers ofcats or breadth of study to conclude that genetic differ-ences within the FLA complex are not involved in FIP.Kipar et al249 reported that cats that developed FIP afternatural FCoVexposure had a significantly higher rate ofviral replication or a reduced capacity for virus clear-ance than cats that were exposed but did not developFIP, suggesting a host factor.

The FeLV is a potent potentiator of FIP. Thirty-fivecases of FIP were observed among 500 conventional(non-SPF except for FeLV) cats that had been experi-mentally infected with FeLV by natural contact or ex-perimental infection.119,121 Deaths due to FIP occurredwithin a relatively short time after they became FeLVviremic, and only one cat in a similar cohort of non-FeLV exposed cats died of FIP. Antibody titers toFIPV began to raise progressively after the cats be-came viremic for FeLV and reached high levels priorto death from FIP.119


Animal and public health considerationsFIPV is a naturally occurring infection of domestic andwild Felidae. People are not hosts for the virus. Dogsand swine can be experimentally infected with FIPV.One of the first commercial CCV vaccines for dogsmay have actually consisted of FIPV; hundreds ofdogs died of severe neurologic and pancreatic diseasebefore it was recalled.250 A mild to moderately severeTGE-like syndrome occurs in baby pigs.251 However,it is doubtful that FIPV is a cause of naturally occur-ring disease in these species.

The comparison of feline coronaviruses, in particu-lar FIPV, and the human SARS coronavirus is note-worthy. The coronavirus found in Palm civet cats (amustelidae and not a felidae) has over 99% homologyto the human SARS coronavirus, and is apparentlywell adapted to its host species.252 However, this viruswas able to geneticially adapt to humans252 with ef-fects that were potentially devastating. A coronaviruswith 92% or more homology to the SARS coronavirusexists in the horseshoe bat, 252,253 suggesting that coro-navirues have adapted species by species over rela-tively short to long periods of time. This ability tomutate through recombination, insertion/deletionmutations, and single nucleotide polymorphismsand change host species is reminiscent of the close re-lationships between canine, porcine and feline corona-viruses.7,14,18e23 The human SARS coronavirus alsohas an interesting 29 nucleotide deletion in one of itsaccessory genes (ORF 8a) that is not present in itshost species.254 Damaging mutations within an acces-sory gene of FECV (ORF 3c) are known to be associ-ated with the FECV / FIPV mutation in cats.29 Theparticipation of both host and virus factors are impor-tant for the development of SARS in a rat model, withincreased virulence in old compared to young rats, asin humans.255 Both host and virus factors are involvedin FIP in cats. Diseases closely resembling FIP of cats,and caused by species related coronaviruses, havebeen described in both ferrets,256e258 and mice.64e66

References1. Holzworth JE. Some important disorders of cats. Cornell

Vet 1963; 53: 157e60.2. Feldman BF, Jortner BS. Clinico-pathology conference.

J Am Vet Med Assoc 1965; 144: 1409e11.3. Wolfe LG, Griesemer RA. Feline infectious peritonitis.

Pathol Vet 1966; 3: 255e70.4. Zook BC, King NW, Robinson RL, McCombs HL. Ultra-

structural evidence for the viral etiology of feline infec-tious peritonitis. Pathol Vet 1968; 5: 91e5.

5. Ward J. Morphogenesis of a virus in cats with experi-mental feline infectious peritonitis. Virol 1970; 41:191e4.

6. Montali RJ, Strandberg JD. Extraperitoneal lesions infeline infectious peritonitis. Vet Pathol 1972; 9: 109e21.

7. Pedersen NC, Ward J, Mengeling WL. Antigenic rela-tionship of the feline infections peritonitis virus to coro-naviruses of other species. Arch Virol 1978; 58: 45e53.

8. Pedersen NC. Morphologic and physical characteristicsof feline infectious peritonitis virus and its growth inautochthonous peritoneal cell cultures. Am J Vet Res1976; 37: 567e72.

9. O’Reilly KJ, Fishman B, Hitchcock LM. Feline infectiousperitonitis: isolation ofa coronavirus. Vet Rec 1979; 104: 348.

10. Black JW. Recovery and in vitro cultivation of a corona-virus from laboratory-induced cases of feline infectiousperitonitis (FIP). Vet Med Small Anim Clin 1980; 75:811e4.

11. Hoshino Y, Scott FW. Immunofluorescent and electronmicroscopic studies of feline small intestine organ cul-tures infected with feline infectious peritonitis virus.Am J Vet Res 1980; 41: 672e81.

12. Pedersen NC, Boyle JF, Floyd K. Infection studies inkittens utilizing feline infectious peritonitis viruspropagated in cell culture. Am J Vet Res 1981; 42:363e7.

13. Jacobse-Geels HE, Horzinek MC. Expression of feline in-fectious peritonitis coronavirus antigens on the surfaceof feline macrophage-like cells. J Gen Virol 1983; 64:1859e66.

14. Pedersen NC, Black JW, Boyle JF, Evermann JF,McKeirnan AJ, Ott RL. Pathogenic differences betweenvarious feline coronavirus isolates. Coronaviruses; mo-lecular biology and pathogenesis. Adv Exp Med Biol1984; 173: 365e80.

15. Pedersen NC. Feline infectious peritonitis. Somethingold, something new. Feline Pract 1976; 6: 42e51.

16. Jakob H. Therapeutsiche, kasuisitische und statistischeMitteilungen aus der Klinik fur kleine Haustiere ander Reichstierarzneischule in Utrecht (Holland), Jahr-gang 1912/13. Z Tiermed Microbiol Immunol 1914; 18:193.

17. de Groot RJ, Horzinek MC. Feline infectious peritonitis.In: Siddell SG, ed. The Coronaviridae: a review ofcoronaviruses and toroviruses. New York and London:Plenum Press, 1995: 294.

18. Haelterman EO. Epidemiological studies of transmissi-ble gastroenteritis of swine. US Livestock Sanitation As-soc Proc 1962; 66: 305e15.

19. Motokawa K, Hohdatsu T, Hashimoto H, Koyama H.Comparison of the amino acid sequence and phyloge-netic analysis of the peplomer, integral membrane andnucleocapsid proteins of feline, canine and porcinecoronaviruses. Microbiol Immunol 1996; 40: 425e33.

20. Benetka V, Kolodziejek J, Walk K, Rennhofer M, MostlK. M gene analysis of atypical strains of feline andcanine coronavirus circulating in an Austrian animalshelter. Vet Rec 2006; 159: 170e4.

21. Herrewegh AA, Smeenk I, Horzinek MC, Rottier PJ, deGroot RJ. Feline coronavirus type II strains 79-1683 and79-1146 originate from a double recombination betweenfeline coronavirus type I and canine coronavirus. J Virol1998; 72: 4508e14.

22. Wesley RD. Th S gene of canine coronavirus, UCD-1, ismore closely related to the S gene of transmissible gas-troenteritis virus than to that of feline infectious perito-nitis virus. Virus Res 1999; 61: 145e52.

23. McArdle F, Bennett M, Gaskell RM, Tennant B, KellyDF, Gaskell CJ. Induction and enhancement of feline in-fectious peritonitis by canine coronavirus. Am J Vet Res1992; 53: 1500e6.

24. de Groot RJ, ter Haar RJ, Horzinek MC, van der ZeijstBA. Intracellular RNAs of the feline infectious peritoni-tis strain 79-1146. J Gen Virol 1987; 68: 995e1002.

252 NC Pedersen

25. Pedersen NC, Allen CE, Lyons LA. Pathogenesis of fe-line enteric coronavirus infection. J Feline Med Surg2008; 10: 529e41.

26. Weijer K, UijtdeHaag F, Osterhaus A. Control of felineleukaemia virus by a removal programme. Vet Rec1986; 119: 555e6.

27. Pedersen NC, Boyle JF, Floyd K, Fudge A, Barker J. Anenteric coronavirus infection of cats and its relationshipto feline infectious peritonitis. Am J Vet Res 1981; 42:368e77.

28. Poland AM, Vennema H, Foley JE, Pedersen NC. Two re-lated strains of feline infectious peritonitis virus isolatedfrom immunocompromised cats infected with a felineenteric coronavirus. J Clin Microbiol 1996; 34: 3180e4.

29. Vennema H, Poland A, Foley J, Pedersen NC. Feline in-fectious peritonitis viruses arise by mutation from en-demic feline enteric coronaviruses. Virol 1998; 243:150e7.

30. Pedersen NC, Johnson L, Theilen GH. Biological behav-ior of tumors and associated retroviremia in cats inocu-lated with SnydereTheilen fibrosarcoma virus and thephenomenon of tumor recurrence after primary regres-sion. Infect Immun 1984; 43: 631e6.

31. Hardy Jr WD, Hurvitz AI. Feline infectious peritonitis:experimental studies. J Am Vet Med Assoc 1971; 158:994e1002.

32. Kennedy M, Boedeker N, Gibbs P, Kania S. Deletions inthe 7a ORF of feline coronavirus associated with an ep-idemic of feline infectious peritonitis. Vet Microbiol 2001;81: 227e34.

33. Lin CN, Su BL, Huang HP, Lee JJ, Hsieh MW, ChuehLL. Field strain feline coronaviruses with small dele-tions in ORF7b associated with both enteric infectionand feline infectious peritonitis. J Feline Med Surg.[Epub ahead of print].

34. Shiba N, Maeda K, Kato H, Mochizuki M, Iwata H. Dif-ferentiation of feline coronavirus type I and II infectionsby virus neutralization test. Vet Microbiol 2007; 124:348e52.

35. Fiscus SA, Teramoto YA. Antigenic comparison of felinecoronavirus isolates: evidence for markedly differentpeplomer glycoproteins. J Virol 1987; 6: 2607e13.

36. Benetka V, Kubber-Heiss A, Kolodziejek J, Nowotny N,Hofmann-Parisot M, Mostl K. Prevalence of feline coro-navirus types I and II in cats with histopathologicallyverified feline infectious peritonitis. Vet Microbiol 2004;99: 31e42.

37. Kummrow M, Meli ML, Haessig M, et al. Feline corona-virus serotypes 1 and 2: seroprevalence and associationwith disease in Switzerland. Clin Diagn Lab Immunol2005; 12: 1209e15.

38. Hohdatsu T, Okada S, Ishizuka Y, Yamada H, KoyamaH. The prevalence of types I and II feline coronavirusinfections in cats. J Vet Med Sci 1992; 54: 557e62.

39. Herrewegh AA, Vennema H, Horzinek MC, Rottier PJ,de Groot RJ. The molecular genetics of feline coronavi-ruses: comparative sequence analysis of the ORF7a/7btranscription unit of different biotypes. Virol 1995; 212:622e31.

40. Pedersen NC, Meric SM, Ho E, Johnson L, Plucker S,Theilen GH. The clinical significance of latent feline leu-kemia virus infection in cats. Feline Pract 1984; 14:32e48.

41. Dewerchin HL, Cornelissen E, Nauwynck HJ. Replica-tion of feline coronaviruses in peripheral blood mono-cytes. Arch Virol 2005; 150: 2483e500.

42. Rottier PJ, Nakamura K, Schellen P, Volders H, HaijemaBJ. Acquisition of macrophage tropism during the path-ogenesis of feline infectious peritonitis is determined bymutations in the feline coronavirus spike protein. J Virol2005; 79: 14122e30.

43. Meli M, Kipar A, Muller C, et al. High viral loads de-spite absence of clinical and pathological findings incats experimentally infected with feline coronavirus(FCoV) type I and in naturally FCoV-infected cats.J Feline Med Surg 2004; 6: 69e81.

44. Addie DD, Toth S, Murray GD, Jarrett O. Risk of felineinfectious peritonitis in cats naturally infected withfeline coronavirus. Am J Vet Res 1995; 56: 429e34.

45. Dye C, Siddell SG. Genomic RNA sequence of felinecoronavirus strain FCoV C1Je. J Feline Med Surg 2007;9: 202e13.

46. Addie DD, Toth S, Herrewegh AA, Jarrett O. Feline co-ronavirus in the intestinal contents of cats with felineinfectious peritonitis. Vet Rec 1996; 139: 522e3.

47. Herrewegh AA, Mahler M, Hedrich HJ, et al. Persis-tence and evolution of feline coronavirus in a closedcat-breeding colony. Virol 1997; 4: 349e63.

48. Addie DD, Kennedy LJ, Ryvar R, et al. Feline leucocyteantigen class II polymorphism and susceptibility to fe-line infectious peritonitis. J Feline Med Surg 2004; 6:59e62.

49. Giordano A, Paltrinieri S. Interferon-gamma in the se-rum and effusions of cats with feline coronavirus infec-tion. Vet J 2008; 10: [Epub ahead of print].

50. Paltrinieri S, Gelain ME, Ceciliani F, Ribera AM, Batti-lani M. Association between faecal shedding of felinecoronavirus and serum alpha1-acid glycoprotein sialy-lation. J Feline Med Surg 2008; 10: 514e8.

51. Kipar A, Meli ML, Failing K, et al. Natural feline coro-navirus infection: differences in cytokine patterns in as-sociation with the outcome of infection. Vet ImmunolImmunopathol 2006; 112: 141e55.

52. Foley JE, Pedersen NC. Inheritance of susceptibility offeline infectious peritonitis in purebred catteries. FelinePract 1996; 24: 14e22.

53. Foley JE, Poland A, Carlson J, Pedersen NC. Risk fac-tors for feline infectious peritonitis among cats in mul-tiple-cat environments with endemic feline entericcoronavirus. J Am Vet Med Assoc 1997; 210: 1313e8.

54. Colby ED, Low RJ. Feline infectious peritonitis. Vet MedSmall Anim Clin 1970; 65: 783e6.

55. Colly LP. Feline infectious peritonitis. Vet Clin North Am1973; 3: 34.

56. Fowler ME. Zoo and wild animal medicine. Philadel-phia: WB Saunders, 1978: 660.

57. Juan-Salles C, Domingo M, Herraez P, Fernandez A,Segales J, Fernandez J. Feline infectious peritonitis inservals (Felis serval). Vet Rec 1998; 143: 535e6.

58. Pedersen NC. Feline infectious peritonitis and feline en-teric coronavirus infections. Part I: feline enteric corona-virus. Feline Pract 1983; 13: 13e9.

59. Pfeifer ML, Evermann JF, Roelkie ME, Gallina AM, OttRL, McKeirnan AJ. Feline infectious peritonitis in a cap-tive cheetah. J Am Vet Med Assoc 1983; 183: 1317e9.

60. Poelma FG, Peters JC, Mieog WHW, Zwart P. InfectisePeritonitiss bei Karakal (Felis caracal) and Nordluchs(Felis lynx lynx). Proceedings Erkrankungen der Zoo-tiere, 13th International Symposium of Zoo Veterinar-ians; Helsinki, Finland, 1974: 249e53.

61. Theobald J. In: Fowler ME, ed. Zoo and wild animalmedicine. Philadelphia: WB Saunders, 1978: 650e67.


62. Tuch K, Witte KH, Wuller H. Feststellung der felinen in-fecktisen Peritonitis (FIP) bei Hauskatzen and Leopar-den in Deutschland. Zenterblatt Veterinarmed B 1974;21: 426e41.

63. Watt NJ, MacIntyre NJ, McOrist S. An extended out-break of infectious peritonitis in a closed colony of Eu-ropean wildcats (Felis sylvestris). J Comp Pathol 1993;108: 73e9.

64. France MP, Smith AL, Stevenson R, Barthold SW. Gran-ulomatous peritonitis and pleuritis in interferon-gamma gene knockout mice naturally infected withmouse hepatitis virus. Aust Vet J 1999; 77: 600e4.

65. Kyuwa S, Tagawa Y, Machii K, et al. MHV-induced fatalperitonitis in mice lacking IFN-gamma. Adv Exp MedBiol 1988; 440: 445e50.

66. Kyuwa S, Tagawa Y, Shibata S, Doi K, Machii K, Iwa-kura Y. Murine coronavirus-induced subacute fatalperitonitis in C57BL/6 mice deficient in gamma inter-feron. J Virol 1998; 72: 9286e90.

67. Martınez J, Reinacher M, Perpinan D, Ramis A. Identi-fication of group 1 coronavirus antigen in multisyste-mic granulomatous lesions in ferrets (Mustela putoriusfuro). J Comp Pathol 2008; 138: 54e8.

68. Pedersen NC. Feline infectious peritonitis and felineenteric coronavirus infections. Part II: feline infectiousperitonitis. Feline Pract 1983; 13: 5e19.

69. Foley JE, Poland A, Carlson J, Pedersen NC. Patterns offeline coronavirus infection and fecal shedding fromcats in multiple-cat environments. J Am Vet Med Assoc1997; 210: 1307e12.

70. Pedersen NC, Sato R, Foley JE, Poland AM. Common vi-rus infections in cats, before and after being placed inshelters, with emphasis on feline enteric coronavirus.J Feline Med Surg 2004; 6: 83e8.

71. Bell ET, Toribio JA, White JD, Malik R, Norris JM. Sero-prevalence study of feline coronavirus in owned andferal cats in Sydney, Australia. Aust Vet J 2006; 84:74e81.

72. Bell ET, Malik R, Norris JM. The relationship betweenthe feline coronavirus antibody titre and the age, breed,gender and health status of Australian cats. Aust Vet J2006; 84: 2e7.

73. Addie DD, Jarrett O. Use of a reverse-transcriptasepolymerase chain reaction for monitoring the sheddingof feline coronavirus by healthy cats. Vet Rec 2001; 148:649e53.

74. Foley JE, Poland A, Carlson J, Pedersen NC. Patterns offeline coronavirus infection and fecal shedding fromcats in multiple-cat households. J Am Vet Med Assoc1997; 210: 1307e12.

75. Marioni-Henry K, Vite CH, Newton AL, Van Winkle TJ.Prevalence of diseases of the spinal cord of cats. J VetIntern Med 2004; 18: 851e8.

76. Rohrbach BW, Legendre AM, Baldwin CA, Lein DH,Reed WM, Wilson RB. Epidemiology of feline infectiousperitonitis among cats examined at veterinary medicalteaching hospitals. J Am Vet Med Assoc 2001; 218:1111e5.

77. Cave TA, Thompson H, Reid SW, Hodgson DR,Addie DD. Kitten mortality in the United Kingdom:a retrospective analysis of 274 histopathological ex-aminations (1986 to 2000). Vet Rec 2002; 151:497e501.

78. Potkay S, Bacher JD, Pitts TW. Feline infectious perito-nitis in a closed breeding colony. Lab Anim Sci 1974;24: 279e89.

79. Addie DD, Jarrett O. Control of feline coronavirus inbreeding catteries by serotesting, isolation, and earlyweaning. Feline Pract 1995; 23(3): 92e5.

80. Norris JM, Bosward KL, White JD, Baral RM, Catt MJ,Malik R. Clinicopathological findings associated withfeline infectious peritonitis in Sydney, Australia: 42cases (1990e2002). Aust Vet J 2005; 83: 666e73.

81. Pesteanu-Somogyi LD, Radzai C, Pressler BM. Preva-lene of feline infectious peritonitis in specific cat breeds.J Feline Med Surg 2006; 8: 1e5.

82. Evermann JF, Baumgartener L, Ott RL, Davis EV,McKiernan AJ. Characterization of a feline infectiousperitonitis virus isolate. Vet Path 1981; 18: 256e65.

83. Pedersen NC, Black JW. Attempted immunization ofcats against feline infectious peritonitis using eitheravirulent live virus or sublethal amounts of virulent vi-rus. Am J Vet Res 1983; 44: 229e34.

84. Pedersen NC, Floyd K. Experimental studies with threenew strains of feline infectious peritonitis virus FIPV-UCD2, FIPV-UCD3, and FIPV-UCD4. Compend ContEduc Pract Vet 1985; 7: 1001e11.

85. Wright KN, Gompf RE, DeNovo Jr RC. Peritoneal effu-sion in cats: 65 cases (1981e1997). J Am Vet Med Assoc1999; 214: 375e81.

86. Trotman TK, Mauldin E, Hoffmann V, Del Piero F, HessRS. Skin fragility syndrome in a cat with feline infec-tious peritonitis and hepatic lipidosis. Vet Dermatol2007; 18: 365e9.

87. McKiernan AJ, Evermann JF, Hargis A, Ott RL. Isola-tion of feline coronaviruses from two cats with diversedisease manifestations. Feline Pract 1981; 11(6): 16e20.

88. Harvey CJ, Lopez JW, Hendrick MJ. An uncommon in-testinal manifestation of feline infectious peritonitis: 26cases (1986e1993). J Am Vet Med Assoc 1996; 15:1117e20.

89. Van Kruiningen. The classifcation of feline colitis. J CompPathol 1983; 93: 275e94.

90. deMadron E. Pericarditis with cardiac tamponade sec-ondary to feline infectious peritonitis in a cat. J AmAnim Hosp Assoc 1986; 22: 65e9.

91. Bradshaw JM, Pearson GR, Gruffydd-Jones TJ. A retro-spective study of 286 cases of neurological disorders ofthe cat. J Comp Pathol 2004; 131: 112e20.

92. Foley JE, Lapointe JM, Koblik P, Poland A, PedersenNC. Diagnostic features of clinical neurologic felineinfectious peritonitis. J Vet Intern Med 1998; 12:415e23.

93. Holliday TA. Clinical aspects of some encephalopathiesof domestic cats. Vet Clin North Am Small Anim Clin1971; 1: 367e78.

94. Kornegay JN. Feline infectious peritonitis. J Am AnimHosp Assoc 1978; 14: 580e4.

95. Legendre AM, Whitenack DL. Feline infectious perito-nitis with spinal cord involvement in two cats. J AmVet Med Assoc 1975; 167: 931e2.

96. Slausen DO, Finn JP. Meningoencephalitis and panoph-thalmitis in feline infectious peritonitis. J Am Vet MedAssoc 1972; 160: 729e34.

97. Quesnel AD, Parent JM, McDonell W, Percy D, Lums-den JH. Diagnostic evaluation of cats with seizure dis-orders: 30 cases (1991e1993). J Am Vet Med Assoc1997; 210: 65e71.

98. Timmann D, Cizinauskas S, Tomek A, Doherr M, Van-develde M, Jaggy A. Retrospective analysis of seizuresassociated with feline infectious peritonitis in cats.J Feline Med Surg 2008; 10: 9e15.

254 NC Pedersen

99. Fankhauser R, Fatzer R. Meningitis and Chorioependymi-tis granulomatosa der Katze: mogliche Beziehungen zurfelinen infectisen Peritonitis (FIP). Kleintierpraxis 1977;22: 19e22.

100. Hayashi T, Utsumi F, Takahashi R, Fujiwara K. Pathologyof non-effusive-type feline infectious peritonitis and ex-perimental transmission. Jpn J Vet Sci 1980; 42: 197e210.

101. Krum S, Johnson K, Wilson J. Hydrocephalus associ-ated with the noneffusive form of feline infectious peri-tonitis. J Am Vet Med Assoc 1975; 167: 746e8.

102. Campbell LH, Reed C. Ocular signs associated with fe-line infectious peritonitis in two cats. Feline Pract 1975;5: 32e5.

103. Campbell LH, Schiessl MM. Ocular manifestations oftoxoplasmosis, infectious peritonitis and lymphosar-coma in cats. Mod Vet Pract 1978; 59: 761e4.

104. Doherty MJ. Ocular manifestations of feline infectiousperitonitis. J Am Vet Med Assoc 1971; 159: 417e24.

105. Gelatt KM. Iridocyclitis-panophthalmitis associatedwith feline infectious peritonitis. Vet Med Small AnimClin 1973; 68: 56e7.

106. Gillespie JH, Scott FW. Feline viral infections. Adv VetSci Comp Med 1973; 17: 163e200.

107. Goodhead AD. Uveitis in dogs and cats: guidelines forthe practitioner. J S Afr Vet Assoc 1996; 67: 12e9.

108. Peiffer Jr RL, Wilcock BP. Histopathologic study of uve-itis in cats: 139 cases (1978e1988). J Am Vet Med Assoc1991; 198: 135e8.

109. Rota A, Paltrinieri S, Jussich S, Ubertalli G, Appino S.Priapism in a castrated cat associated with feline infec-tious peritonitis. J Feline Med Surg [Epub ahead ofprint].

110. Foster RA, Caswell JL, Rinkardt N. Chronic fibrinousand necrotic orchitis in a cat. Can Vet J 1996; 37:681e2.

111. Sigurdardottir OG, Kolbjørnsen O, Lutz H. Orchitis ina cat associated with coronavirus infection. J CompPathol 2001; 124: 219e22.

112. Kitagawa M, Okada M, Sato T, Kanayama K, Sakai T. Afeline case of isolated fourth ventricle with syringomye-lia suspected to be related with feline infectious peritoni-tis. J Vet Med Sci 2007; 69: 759e62.

113. Cannon MJ, Silkstone MA, Kipar AM. Cutaneous le-sions associated with coronavirus-induced vasculitisin a cat with feline infectious peritonitis and concurrentfeline immunodeficiency virus infection. J Feline MedSurg 2005; 7: 233e6.

114. Declercq J, De Bosschere H, Schwarzkopf I, Declercq L.Papular cutaneous lesions in a cat associated with felineinfectious peritonitis. Vet Dermatol 2008; 19: 255e8.

115. Kipar A, Koehler K, Bellmann S, Reinacher M. Felineinfectious peritonitis presenting as a tumour in theabdominal cavity. Vet Rec 1999; 144: 118e22.

116. Madewell BR, Crow SE, Nickerson TR. Infectious peri-tonitis in a cat that subsequently developed a myelopro-liferative disorder. J Am Vet Med Assoc 1978; 172:169e72.

117. MacEwen EG, Hurvitz AI. Diagnosis and managementof monoclonal gammopathies. Vet Clin North Am 1977;7: 119e32.

119. Pedersen NC. Virologic and immunologic aspects of fe-line infectious peritonitis virus infection. Adv Exp MedBiol 1987; 218: 529e50.

120. Cotter SM, Gilmore CE, Rollins C. Multiple cases offeline leukemia and feline infectious peritonitis ina household. J Am Vet Med Assoc 1973; 162: 1054e8.

121. Pedersen NC, Theilen G, Keane MA, et al. Studies ofnaturally transmitted feline leukemia virus infection.Am J Vet Res 1977; 38: 1523e31.

122. Ward BC, Pedersen NC. Infectious peritonitis in cats.J Am Vet Med Assoc 1966; 154: 26e35.

123. Toomey JM, Carlisle-Nowak MM, Barr SC, et al. Con-current toxoplasmosis and feline infectious peritonitisin a cat. J Am Anim Hosp Assoc 1995; 31: 425e58.

124. Van Hamme E, Dewerchin HL, Cornelissen E, Verhas-selt B, Nauwynck HJ. Clathrin- and caveolae-indepen-dent entry of feline infectious peritonitis virus inmonocytes depends on dynamin. J Gen Virol 2008; 89:2147e56.

125. Dewerchin HL, Cornelissen E, Van Hamme E, Smits K,Verhasselt B, Nauwynck HJ. Surface-expressed viralproteins in feline infectious peritonitis virus-infectedmonocytes are internalized through a clathrin- and cav-eolae-independent pathway. J Gen Virol 2008; 89:2731e40.

126. Delmas B, Gelfi J, Sjostrom H, Noren O, Laude H. Fur-ther characterization of aminopeptidase-N as a recep-tor for coronaviruses. Adv Exp Med Biol 1993; 342:293e8.

127. Benbacer L, Kut E, Besnardeau L, Laude H, Delmas B.Interspecies aminopeptidase-N chimeras reveal spe-cies-specific receptor recognition by canine coronavi-rus, feline infectious peritonitis virus, andtransmissible gastroenteritis virus. J Virol 1997; 71:734e7.

128. Kolb AF, Hegyi A, Maile J, Heister A, Hagemann M,Siddell SG. Molecular analysis of the coronavirus-receptor function of aminopeptidase N. Adv Exp MedBiol 1998; 440: 61e7.

129. Hohdatsu T, Izumiya Y, Yokoyama Y, Kida K, KoyamaH. Differences in virus receptor for type I and type IIfeline infectious peritonitis virus. Arch Virol 1998; 143:839e50.

130. Hegyi A, Kolb AF. Characterization of determinants in-volved in the feline infectious peritonitis virus receptorfunction of feline aminopeptidase N. J Gen Virol 1998;79: 1387e91.

131. Van Hamme E, Dewerchin HL, Cornelissen E, Nau-wynck HJ. Attachment and internalization of feline in-fectious peritonitis virus in feline blood monocytes andCrandell feline kidney cells. J Gen Virol 2007; 88:2527e32.

132. Herrewegh AA, Vennema H, Horzinek MC, Rottier PJ, deGroot RJ. The molecular genetics of feline coronaviruses:Comparative sequence analysis for the ORF 7a/b tran-scription unit of different biotypes. Virology 1995; 212:622e31.

133. Regan AD, Shraybman R, Cohen RD, Whittaker GR. Dif-ferential role for low pH and cathepsin-mediated cleav-age of the viral spike protein during entry of serotype IIfeline coronaviruses. Vet Microbiol [Epub ahead of print].

134. Can-Sahna K, Soydal Ataseven V, Pinar D, OguzogluTC. The detection of feline coronaviruses in blood sam-ples from cats by mRNA RT-PCR. J Feline Med Surg2007; 9: 369e72.

135. Weiss RC, Scott FW. Pathogenesis of feline infetiousperitonitis: pathologic changes and immunofluores-cence. Am J Vet Res 1981; 42: 2036e48.

136. Battilani M, Coradin T, Scagliarini A, et al. Quasispeciescomposition and phylogenetic analysis of feline corona-viruses (FCoVs) in naturally infected cats. FEMS Immu-nol Med Microbiol 2003; 39: 141e7.


137. Poncelet L, Coppens A, Peeters D, Bianchi E, Grant CK,Kadhim H. Detection of antigenic heterogeneity in fe-line coronavirus nucleocapsid in feline pyogranuloma-tous meningoencephalitis. Vet Pathol 2008; 45: 140e53.

138. Kiss I, Kecskemeti S, Tanyi J, Klingeborn B, Belak S. Pre-liminary studies on feline coronavirus distribution innaturally and experimentally infected cats. Res Vet Sci2000; 68: 237e42.

139. Pedersen NC, Boyle JF. Immunologic phenomena in theeffusive form of feline infectious peritonitis. Am J VetRes 1980; 41: 868e76.

140. Weiss RC, Scott FW. Antibody-mediated enhancementof disease in feline infectious peritonitis: comparisonswith dengue hemorrhagic fever. Comp Immunol Micro-biol Infect Dis 1981; 4: 175e89.

141. Dewerchin HL, Cornelissen E, Nauwynck HJ. Feline in-fectious peritonitis virus-infected monocytes internal-ize viral membrane-bound proteins upon antibodyaddition. J Gen Virol 2006; 87: 1685e90.

142. Cornelissen E, Dewerchin HL, Van Hamme E, Nau-wynck HJ. Absence of surface expression of feline infec-tious peritonitis virus. Vet Microbiol 2007; 121: 131e7.

143. Jacobse-Geels HEL, Daha MR, Horzinek MC. Antibodyimmune complexes and complement activity fluctua-tions in kittens with experimentally induced feline in-fectious peritonitis. Am J Vet Res 1982; 43: 666e70.

144. Jacobse-Geels HEL, Daha MR, Horzinek MC. Isolationand characterization of feline C3 and evidence for the im-mune complex pathogenesis of feline infectious peritoni-tis. J Immunol 1980; 125: 1606e10.

145. Weiss RC, Dodds WJ, Scott FW. Disseminated intravas-cular coagulation in experimentally induced feline infec-tious peritonitis. Am J Vet Res 1980; 41: 663e71.

146. Paltrinieri S, Cammarata MP, Cammarata G, ComazziS. Some aspects of humoral and cellular immunityin naturally occurring feline infectious peritonitis. VetImmunol Immunopathol 1998; 65: 205e20.

147. Paltrinieri S, Grieco V, Comazzi S, Cammarata ParodiM. Laboratory profiles in cats with different pathologi-cal and immunohistochemical findings due to felineinfectious peritonitis (FIP). J Feline Med Surg 2001; 3:149e59.

148. Sparkes AH, Gruffydd-Jones TJ, Harbour DA. Feline in-fectious peritonitis: a review of clinicopathologicalchanges in 65 cases, and a critical assessment of theirdiagnostic value. Vet Rec 1991; 129: 209e12.

149. Hirschberger J, Hartmann K, Wilhelm N, Frost J, LutzH, Kraft W. Clinical symptoms and diagnosis of felineinfectious peritonitis. Tierarztl Prax 1995; 23: 92e9.

150. Harvey JW, Gaskin JM. Feline haptoglobin. Am J Vet Res1978; 39: 549e53.

151. Boettcher IC, Steinberg T, Matiasek K, Greene CE,Hartmann K, Fischer A. Use of anti-coronavirus anti-body testing of cerebrospinal fluid for diagnosis of fe-line infectious peritonitis involving the centralnervous system in cats. J Am Vet Med Assoc 2007;230: 199e205.

152. Paltrinieri S, Parodi MC, Cammarata G. In vivo diagno-sis of feline infectious peritonitis by comparison of pro-tein content, cytology, and direct immunofluorescencetest on peritoneal and pleural effusions. J Vet DiagnInvest 1999; 11: 358e61.

153. Rand JS, Parent J, Percy D, Jacobs R. Clinical, cerebro-spinal fluid, and histological data from twenty-sevencats with primary inflammatory disease of the centralnervous system. Can Vet J 1994; 35: 103e10.

154. Singh M, Foster DJ, Child G, Lamb WA. Inflammatorycerebrospinal fluid analysis in cats: clinical diagnosisand outcome. J Feline Med Surg 2005; 7: 77e93.

155. Barlough JE, Adsit JC, Scott FW. The worldwide occur-rence of feline infectious peritonitis. Feline Pract 1982;12(6): 26e30.

156. Pedersen NC. Serologic studies of naturally occurringfeline infectious peritonitis. Am J Vet Res 1976; 37:1449e53.

157. Pratelli A. Comparison of serologic techniques for thedetection of antibodies against feline coronaviruses.J Vet Diagn Invest 2008; 20(1): 45e50.

158. Hartmann K, Binder C, Hirschberger J, et al. Comparisonof different tests to diagnose feline infectious peritonitis.J Vet Intern Med 2003; 17: 781e90.

159. Kennedy MA, Abd-Eldaim M, Zika SE, Mankin JM, Ka-nia SA. Evaluation of antibodies against feline corona-virus 7b protein for diagnosis of feline infectiousperitonitis in cats. Am J Vet Res 2008; 69: 1179e82.

160. Steinberg TA, Boettcher IC, Matiasek K, et al. Use of al-bumin quotient and IgG index to differentiate blood- vsbrain-derived proteins in the cerebrospinal fluid of catswith feline infectious peritonitis. Vet Clin Pathol 2008;37: 207e16.

161. Simons AF, Vennema H, Rofina JE, et al. A mRNA PCRfor the diagnosis of feline infectious peritonitis. J VirolMethods 2005; 124: 111e6.

162. Dye C, Helps CR, Siddell SG. Evaluation of real-timeRT-PCR for the quantification of FCoV shedding inthe faeces of domestic cats. J Feline Med Surg 2008; 10:167e74.

163. Bence LM, Addie DD, Eckersall PD. An immunotrubi-dimetric assay for rapid quantitative measurement offeline alpha-1-acid glycoprotein in serum and perito-neal fluid. Vet Clin Pathol 2005; 34: 335e41.

164. Saverio P, Alessia G, Vito T, Stefano S. Critical assessmentof the diagnostic value of feline alpha1-acid glycoproteinfor feline infectious peritonitis using the likelihood ratiosapproach. J Vet Diagn Invest 2007; 19: 266e72.

165. Paltrinieri S, Metzger C, Battilani M, Gelain ME, Gior-dano A. Serum alpha1-acid glycoprotein (AGP) concen-tration in non-symptomatic cats with feline coronavirus(FCoV) infection. J Feline Med Surg 2007; 9: 271e7.

166. Duthie S, Eckersall PD, Addie DD, Lawrence CE, JarrettO. Value of alpha 1-acid glycoprotein in the diagnosisof feline infectious peritonitis. Vet Rec 1997; 141:299e303.

167. Ceciliani F, Grossi C, Giordano A, Pocacqua V, Paltri-nieri S. Decreased sialylation of the acute phase proteinalpha1-acid glycoprotein in feline infectious peritonitis(FIP). Vet Immunol Immunopathol 2004; 99: 229e36.

168. Giordano A, Spagnolo V, Colombo A, Paltrinieri S.Changes in some acute phase protein and immunoglob-ulin concentrations in cats affected by feline infectiousperitonitis or exposed to feline coronavirus infection.Vet J 2004; 167: 38e44.

169. Negrin A, Lamb CR, Cappello R, Cherubini GB. Resultsof magnetic resonance imaging in 14 cats with menin-goencephalitis. J Feline Med Surg 2007; 9: 109e16.

170. Tammer R, Evensen O, Lutz H, Reinacher M. Immuno-histological demonstration of feline infectious peritoni-tis virus antigen in paraffin-embedded tissues usingfeline ascites or murine monoclonal antibodies. VetImmunol Immunopathol 1995; 49: 177e82.

171. Giordano A, Paltrinieri S, Bertazzolo W, Milesi E,Parodi M. Sensitivity of Tru-cut and fine needle

256 NC Pedersen

aspiration biopsies of liver and kidney for diagnosisof feline infectious peritonitis. Vet Clin Pathol 2005;34: 368e74.

172. Cotter SM, Hardy Jr WD, Essex M. The association offeline leukemia virus with lymphosarcoma and otherdisorders. J Am Vet Med Assoc 1975; 168: 448e54.

173. Holmberg CA, Gribble DG. Feline infectious peritonitis:diagnostic gross and microscopic lesions. Feline Pract1973; 3: 11e4.

174. Kipar A, May H, Menger S, Weber M, Leukert W, Rein-acher M. Morphologic features and development ofgranulomatous vasculitis in feline infectious peritonitis.Vet Pathol 2005; 42: 321e30.

175. Kipar A, Bellmann S, Kremendahl J, Kohler K, Rein-acher M. Cellular composition, coronavirus antigenexpression and production of specific antibodies in le-sions in feline infectious peritonitis. Vet ImmunolImmunopathol 1998; 65: 243e57.

176. Kipar A, Kohler K, Leukert W, Reinacher M. A compar-ison of lymphatic tissues from cats with spontaneousfeline infectious peritonitis (FIP), cats with FIP virus in-fection but no FIP, and cats with no infection. J CompPathol 2001; 125: 182e91.

177. Weiss RC, Scott FW. Pathogenesis of feline infectiousperitonitis: nature and development of viremia. Am JVet Res 1981; 42: 382e90.

178. Hickman MA, Morris JG, Rogers QR, Pedersen NC. Elim-ination of feline coronavirus infection from a large exper-imental specific pathogen-free cat breeding colony byserologic testing and isolation. Feline Pract 1995; 23:96e102.

179. Christianson KK, Ingersoll JD, Landon RM, Pfeiffer NE,Gerber JD. Characterization of a temperature-sensitivefeline infectious peritonitis coronavirus. Arch Virol1989; 108: 185e96.

180. Gerber JD, Ingersoll JD, Gast AM, et al. Protectionagainst feline infectious peritonitis by intranasal inocu-lation of a temperature-sensitive FIPV vaccine. Vaccine1990; 8: 536e42.

181. Pedersen NC. Basic and clinical immunology. In: Holz-worth JE, ed. Diseases of the cat. Philadelphia: WB Sa-unders, 1987: 146e81.

182. Mochizuki M, Mitsutake Y, Miyanohara Y, HigashiharaT, Shimizu T, Hohdatsu T. Antigenic and plaque varia-tions of serotype II feline infectious peritonitis corona-viruses. J Vet Med Sci 1997; 59: 253e8.

183. Kiss I, Poland AM, Pedersen NC. Disease outcome andcytokine responses in cats immunized with an aviru-lent feline infectious peritonitis virus (FIPV)-UCD1and challenge-exposed with virulent FIPV-UCD8.J Feline Med Surg 2004; 6: 89e97.

184. Stoddart ME, Gaskell RM, Harbour DA, Pearson GR.The sites of early viral replication in feline infectiousperitonitis. Vet Microbiol 1988; 18: 259e71.

185. Stoddart ME, Gaskell RM, Harbour DA, Gaskell CJ. Vi-rus shedding and immune responses in cats inoculatedwith cell culture-adapted feline infectious peritonitisvirus. Vet Microbiol 1988; 16: 145e58.

186. Haijema BJ, Volders H, Rottier PJ. Switching speciestropism: an effective way to manipulate the feline coro-navirus genome. J Virol 2003; 77: 4528e38.

187. Tekes G, Hofmann-Lehmann R, Stallkamp I, Thiel V,Thiel HJ. Genome organization and reverse genetic anal-ysis of a type I feline coronavirus. J Virol 2008; 82: 1851e9.

188. Colgrove DJ, Parker AJ. Feline infectious peritonitis.J Small Anim Pract 1971; 12: 225e32.

189. Hartmann K, Ritz S. Treatment of cats with feline infec-tious peritonitis. Vet Immunol Immunopathol 2008; 123:172e5.

190. Weiss RC, Toivio-Kinnucan M. Inhibition of feline infec-tious peritonitis virus replication by recombinanthuman leukocyte (alpha) interferon and felinefibroblastic (beta) interferon. Am J Vet Res 1988; 49:1329e35.

191. Mochizuki M, Nakatani H, Yoshida M. Inhibitory ef-fects of recombinant feline interferon on the replicationof feline enteropathogenic viruses in vitro. Vet Microbiol1994; 39: 145e52.

192. Ishida T, Shibanai A, Tanaka S, Uchida K, Mochizuki M.Use of recombinant feline interferon and glucocorticoidin the treatment of feline infectious peritonitis. J FelineMed Surg 2004; 6: 107e9.

193. Ritz S, Egberink H, Hartmann K. Effect of feline inter-feron-omega on the survival time and quality of lifeof cats with feline infectious peritonitis. J Vet InternMed 2007; 21: 1193e7.

194. Zabel P, Schade FU, Schlaak M. Inhibition of endoge-nous TNF formation by pentoxifylline. Immunobiology1993; 187: 447e63.

195. Nichols PR, Morris DO, Beale KM. A retrospectivestudy of canine and feline cutaneous vasculitis. VetDermatol 2001; 12: 255e64.

196. Pedersen NC. Animal virus infections that defy vacci-nation: equine infectious anemia, caprine arthritis-encephalitis, maedi-visna, and feline infectious perito-nitis. Adv Vet Sci Comp Med 1989; 33: 413e28.

197. Baldwin CW, Scott FW. Attempted immunization ofcats with feline infectious peritonitis virus propagatedat reduced temperatures. Am J Vet Res 1997; 58: 251e6.

198. Huisman W, Martina BE, Rimmelzwaan GF, Gruters RA,Osterhaus AD. Vaccine-induced enhancement of viralinfections. Vaccine [Epub ahead of print].

199. Postorino-Reeves N, Coyne MJ, Herman JG. Evaluationof the field efficacy of Primucell-FIP vaccine. FelineHealth Top 1992; 7: 6e8.

200. Reeves NC, Pollock RV, Thurber ET. Long-term follow-up study of cats vaccinated with a temperature-sensitive feline infectious peritonitis vaccine. CornellVet 1992; 82: 117e23.

201. Wolf AM. Feline infectious peritonitis, part 2. FelinePract 1997; 25: 24e8.

202. Hoskins JD, Taylor HW, Lomax TL. Challenge trial ofan intranasal feline infectious peritonitis vaccine. FelinePract 1994; 22: 9e13.

203. Fehr D, Holznagel E, Bolla S, et al. Placebo-controlledevaluation of a modified life virus vaccine against felineinfectious peritonitis: safety and efficacy under fieldconditions. Vaccine 1997; 15: 1101e9.

204. Stoddart CA, Barlough JE, Baldwin CA, Scott FW.Attempted immunisation of cats against feline infec-tious peritonitis using canine coronavirus. Res Vet Sci1988; 45: 383e8.

205. Woods RD. Efficacy of vaccination of sows with sero-logically related coronaviruses for control of transmissi-ble gastroenteritis in nursing pigs. Am J Vet Res 1984; 45:1726e9.

206. Woods RD, Pedersen NC. Cross-protection studies be-tween feline infectious peritonitis and porcine transmis-sible gastroenteritis viruses. Vet Microbiol 1979; 4: 11e6.

207. Pedersen NC. Virologic and immunologic aspects offeline infectious peritonitis virus infection. Adv ExpMed Biol 1987; 218: 529e50.


208. Vennema H, de Groot RJ, Harbour DA, Horzinek MC,Spaan WJ. Early death after feline infectious peritonitisvirus challenge due to recombinant vaccinia virus im-munization. J Virol 1990; 64: 1407e9.

209. Wasmoen TL, Kadakia NP, Unfer RC, et al. Protectionof cats from infectious peritonitis by vaccination witha recombinant raccoon poxvirus expressing the nucleo-capsid gene of feline infectious peritonitis virus. AdvExp Med Biol 1995; 380: 221e8.

210. Klepfer S, Reed AP, Martinez M, Bhogal B, Jones E,Miller TJ. Cloning and expression of FECV spike genein vaccinia virus. Immunization with FECVS causesearly death after FIPV challenge. Adv Exp Med Biol1995; 380: 235e41.

211. Glansbeek HL, Haagmans BL, te Lintelo EG, et al. Ad-verse effects of feline IL-12 during DNA vaccinationagainst feline infectious peritonitis virus. J Gen Virol2002; 83: 1e10.

212. Hebben M, Duquesne V, Cronier J, Rossi B, Aubert A.Modified vaccinia virus Ankara as a vaccine againstfeline coronavirus: immunogenicity and efficacy. J Fe-line Med Surg 2004; 6: 111e8.

213. Hohdatsu T, Yamato H, Ohkawa T, et al. Vaccine effi-cacy of a cell lysate with recombinant baculovirus-ex-pressed feline infectious peritonitis (FIP) virusnucleocapsid protein against progression of FIP. VetMicrobiol 2003; 2: 31e44.

214. Haijema BJ, Volders H, Rottier PJ. Live, attenuated co-ronavirus vaccines through the directed deletion ofgroup-specific genes provide protection against felineinfectious peritonitis. J Virol 2004; 78: 3863e71.

215. Barlough JE, Scott FW. Effectiveness of three antiviralagents against FIP virus in vitro. Vet Rec 1990; 126: 556e8.

216. Weiss RC, Cox NR, Martinez ML. Evaluation of free orliposome-encapsulated ribavirin for antiviral therapyof experimentally induced feline infectious peritonitis.Res Vet Sci 1993; 55: 162e72.

217. Barnard DL, Day CW, Bailey K, et al. Enhancement ofthe infectivity of SARS-CoV in BALB/c mice by IMPdehydrogenase inhibitors, including ribavirin. AntiviralRes 2006; 71: 53e63.

218. Barlough JE, Shacklett BL. Antiviral studies of feline in-fectious peritonitis virus in vitro. Vet Rec 1992; 135:177e9.

219. Balzarini J, Keyaerts E, Vijgen L, et al. Pyridine N-oxidederivatives are inhibitory to the human SARS and fe-line infectious peritonitis coronavirus in cell culture.J Antimicrob Chemother 2006; 57: 472e81.

220. Balzarini J, Keyaerts E, Vijgen L, et al. Inhibitionof feline (FIPV) and human (SARS) coronavirus bysemisynthetic derivatives of glycopeptide antibiotics.Antiviral Res 2006; 72: 20e33.

221. Amici C, DiCoro A, Ciucci A, et al. Indomethacin hasa potent antiviral activity against SARS coronavirus.Antivir Ther 2006; 11: 1021e30.

222. Pedersen NC, Addie D, Wolf A. Recommendationsfrom working group of the international coronavirusand feline infectious peritonitis workshop. Feline Pract1995; 23: 108e11.

223. Corapi WV, Darteil RJ, Audonnet JC, Chappuis GE.Localization of antigenic sites of the S glycoprotein offeline infectious peritonitis virus involved in neutrali-zation and antibody-dependent enhancement. J Virol1995; 69: 2858e62.

224. Olsen CW, Corapi WV, Ngichabe CK, Baines JD, ScottFW. Monoclonal antibodies to the spike protein of feline

infectious peritonitis virus mediate antibody-depen-dent enhancement of infection of feline macrophages.J Virol 1992; 66: 956e65.

225. Stoddart CA, Scott FW. Intrinsic resistance of felineperitoneal macrophages to coronaviruses correlateswith in vivo virulence. J Virol 1989; 63: 436e40.

226. Hohdatsu T, Yamada M, Tominaga R, Makino K, KidaK, Koyama H. Antibody-dependent enhancement offeline infectious peritonitis virus infection in feline al-veolar macrophages and human monocyte cell lineU937 by serum of cats experimentally or naturally in-fected with feline coronavirus. J Vet Med Sci 1998; 60:49e55.

227. Halstead SB. In vivo enhancement of dengue infectionwith passively transferred antibody. J Infect Dis 1979;140: 527e33.

228. Halstead SB, Larsen K, Kliks S, Peiris JS, Cardosa J, Por-terfield JS. Comparison of P388D-1 mouse macrophagecell line and human monocytes for assay of dengue-2infection-enhancing antibodies. Am J Trop Med Hyg1983; 32: 157e63.

229. Halstead SM, Rojanasuphot S, Sangkawibha N. Origi-nal antigenic sin in dengue. Am J Trop Med Hyg 1983;32: 154e6.

230. Peiris JSM, Porterfeld JS. Antibody-mediated enhance-ment of flavivirus replications in macrophage-like celllines. Nature 1979; 28: 507e11.

231. Peiris JSM, Gordon S, Unkeless JC, Porterfield JS.Monoclonal anti-Fc receptor IgG blocks antibody en-hancement of viral replication in macrophages. Nature1981; 289: 189e91.

232. Porterfield JS. Immunological enhancement and thepathogenesis of dengue hemorrhagic fever. J Hyg1982; 89: 355e64.

233. Gonon V, Duquesne V, Klonjkowski B, Monteil M, Au-bert A, Eloit M. Clearance of infection in cats naturallyinfected with feline coronaviruses is associated with ananti-S glycoprotein antibody response. J Gen Virol 1999;80: 2315e7.

234. Takano T, Azuma N, Hashida Y, Satoh R, Hohdatsu T.B-cell activation in cats with feline infectious peritonitis(FIP) by FIP-virus-induced B-cell differentiation/sur-vival factors. Arch Virol [Epub ahead of print].

235. Takano T, Katada Y, Moritoh S, et al. Analysis of themechanism of antibody-dependent enhancement of fe-line infectious peritonitis virus infection: aminopepti-dase N is not important and a process of acidificationof the endosome is necessary. J Gen Virol 2008; 89: 1025e9.

236. Paltrinieri S, Cammarata Parodi M, Cammarata G,Mambretti M. Type IV hypersensitivity in the patho-genesis of FIPV-induced lesions. Zentralblatt Vet Med B1989; 45: 151e9.

237. Hardy Jr WD. Immunopathology induced by the felineleukemia virus. Springer Semin Immunopathol 1982; 5:75e106.

238. de Groot-Mijnes JD, van Dun JM, van der Most RG, deGroot RJ. Natural history of a recurrent feline coronavi-rus infection and the role of cellular immunity in sur-vival and disease. J Virol 2005; 79: 1036e44.

239. Paltrinieri S, Ponti W, Comazzi S, Giordano A, Poli G.Shifts in circulating lymphocyte subsets in cats with felineinfectious peritonitis (FIP): pathogenic role and diagnosticrelevance. Vet Immunol Immunopathol 2003; 15: 141e8.

240. Pedersen NC, Evermann JF, McKeirnan AJ, Ott RL.Pathogenicity studis of feline coronavirus isolates79-1146 and 79-1683. Am J Vet Res 1984; 45: 2580e5.

258 NC Pedersen

241. Takano T, Hohdatsu T, Hashida Y, Kaneko Y, Tanabe M,Koyama H. A ‘possible’ involvement of TNF-alpha inapoptosis induction in peripheral blood lymphocytesof cats with feline infectious peritonitis. Vet Microbiol2007; 119: 121e31.

242. Takano T, Hohdatsu T, Toda A, Tanabe M, Koyama H.TNF-alpha, produced by feline infectious peritonitis vi-rus (FIPV)-infected macrophages, upregulates expres-sion of type II FIPV receptor feline aminopeptidase Nin feline macrophages. Virol 2007; 364: 64e72.

243. Berg AL, Ekman K, Belak S, Berg M. Cellular composi-tion and interferon-gamma expression of the local in-flammatory response in feline infectious peritonitis(FIP). Vet Microbiol 2005; 111: 15e23.

244. Dean GA, Olivry T, Stanton C, Pedersen NC. In vivo cy-tokine response to experimental feline infectious perito-nitis virus infection. Vet Microbiol 2003; 97: 1e12.

245. Foley JE, Rand C, Leutenegger C. Inflammation andchanges in cytokine levels in neurological feline infec-tious peritonitis. J Feline Med Surg 2003; 5: 313e22.

246. Gunn-Moore DA, Caney SM, Gruffydd-Jones TJ, HelpsCR, Harbour DA. Antibody and cytokine responses inkittens during the development of feline infectious peri-tonitis (FIP). Vet Immunol Immunopathol 1988; 23: 221e42.

247. Regan AD, Cohen RD, Whittaker GR. Activation of p38MAPK by feline infectious peritonitis virus regulatespro-inflammatory cytokine production in primaryblood-derived feline mononuclear cells. Virol [Epubahead of print].

248. Gelain ME, Meli M, Paltrinieri S. Whole blood cytokineprofiles in cats infected by feline coronavirus andhealthy non-FCoV infected specific pathogen-free cats.J Feline Med Surg 2006; 8: 389e99.

249. Kipar A, Baptiste K, Barth A, Reinacher M. NaturalFCoV infection: cats with FIP exhibit significantlyhigher viral loads than healthy infected cats. J FelineMed Surg 2006; 8: 69e72 [Epub 2005].

250. Martin ML. Coronavirus enteritis and a recent outbreakfollowing modified live virus vaccination. CompendCont Educ Pract Vet 1985; 7: 1012e7.

251. Woods RD, Cheville NF, Gallagher JE. Lesions in the smallintestine of newborn pigs inoculated with porcine, feline,and canine coronaviruses. Am J Vet Res 1981; 42: 1163e9.

252. Shi Z, Hu Z. A review of studies on animal reservoirs of theSARS coronavirus. Virus Res 2008; 133: 74e87 [Epub 2007].

253. Wang LF, Eaton BT. Bats, civets and the emergence ofSARS. Curr Top Microbiol Immunol 2007; 315: 325e44.

254. Oostra M, de Haan CA, Rottier PJ. The 29-nucleotidedeletion present in human but not in animal severeacute respiratory syndrome coronaviruses disruptsthe functional expression of open reading frame 8.J Virol 2007; 81: 13876e88.

255. Nagata N, Iwata N, Hasegawa H, et al. Participation ofboth host and virus factors in induction of severe acuterespiratory syndrome (SARS) in F344 rats infected withSARS coronavirus. J Virol 2007; 81: 1848e57 [Epub 2006].

256. Garner MM, Ramsell K, Morera N, et al. Clinicopatho-logic features of a systemic coronavirus-associated dis-ease resembling feline infectious peritonitis in thedomestic ferret (Mustela putorius). Vet Pathol 2008; 45:236e46.

257. Perpinan D, Lopez C. Clinical aspects of systemic gran-ulomatous inflammatory syndrome in ferrets (Mustelaputorius furo). Vet Rec 2008; 162: 180e4.

258. Martınez J, Ramis AJ, Reinacher M, Perpinan D. Detec-tion of feline infectious peritonitis virus-like antigen inferrets. Vet Rec 2006; 158: 523.

259. Addie DD, Schaap IA, Nicolson L, Jarrett O. Persistenceand transmission of natural type I feline coronavirusinfection. J Gen Virol 2003; 84: 2735e44.

260. Addie DD, Paltrinieri S, Pedersen NC. Recommenda-tions of the second international feline coronavirus/feline infectous peritonitis symposium. J Feline MedSurg 2004; 6: 125e30.

Available online at www.sciencedirect.com

Documents

A review of feline infectious peritonitis virus infection: 1963 …...ton.1,15 Therefore, its existence as a signiﬁcant disease entity prior to this time is questionable. A reference