Since January 2020 Elsevier has created a COVID-19 resource centre with

free information in English and Mandarin on the novel coronavirus COVID-

19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website.

Elsevier hereby grants permission to make all its COVID-19-related

research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means
with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active.
[Link] 11/10/05 4:50 PM Page 132

▼ Chapter

10 Feline Infectious Peritonitis

(Feline Coronavirus)
Robert G. Sherding

Feline infectious peritonitis (FIP) is a progressive and Pathogenesis of Feline Infectious Peritonitis
highly fatal systemic disease of cats caused by feline
coronavirus. Feline coronavirus most frequently causes ▼ Key Point FIP-causing coronaviruses are genetic
inapparent enteric infection with fecal shedding mutants of harmless enteric FCoV.
of virus. Mild enteritis and diarrhea are seen rarely
(see Chapter 14). A mutation of feline coronavirus • It was previously thought that cats were infected by
during intestinal replication enables it to infect two similar but distinct coronaviruses, the innocuous
macrophages and cause FIP. Despite its name, the feline enteric coronavirus and the deadly FIP virus,
lesions of FIP are widespread and not restricted to but these are now considered phenotypic variants
the peritoneum. Effusive and non-effusive forms of (biotypes) of the same virus. The non-mutated
FIP occur. Since its recognition in the 1950s, FIP enterotropic FCoV typically causes a clinically inap-
has been one of the most studied diseases of cats, parent infection of intestinal epithelial cells with fecal
yet a definitive diagnostic test, an effective treatment, shedding of virus. During replication in the cat’s
and a reliable vaccine are lacking. With the decline intestinal tract, FCoV mutates frequently, especially in
in prevalence of feline leukemia virus from vaccination, kittens, and this sporadically results in critical genetic
FIP has become the deadliest infectious disease of mutations that enable FCoV to infect and replicate in
cats. macrophages.

ETIOLOGY ▼ Key Point Any cat with inapparent FCoV infection

has the potential to develop FIP if the virus mutates
Feline Coronavirus during replication to allow the mutated FCoV
variant to infect macrophages. Viral replication in
• Feline coronavirus (FCoV) is a single-stranded macrophages is the defining event in FIP.
enveloped RNA virus with distinctive petal-shaped
peplomers projecting from the surface. There are
two serotypes of FCoV that differ in cell culture • Macrophages then replicate the mutated coronavirus
characteristics. Serotype I predominates in North and carry it to target tissues such as the peritoneum,
America and Europe, whereas Serotype II is more pleura, kidney, uvea, and nervous system, resulting
closely related to canine coronavirus (CCV) and pre- in widespread immune-mediated vasculitis, dissemi-
dominates in Japan. nated perivascular pyogranulomatous inflammation,
• Coronaviruses are found in many animals and are and exudative fibrinous polyserositis. These are the
generally adapted for infecting epithelial cells of the characteristic lesions of FIP.
respiratory or gastrointestinal tract. FCoV is closely • The pathogenesis involves circulating immune com-
related to CCV, swine transmissible gastroenteritis plexes, complement fixation, cytokine release, apop-
virus (TGEV), porcine respiratory virus, and some tosis of activated T lymphocytes, and vascular damage
human coronaviruses. The CCV that causes enteritis with necrosis and increased permeability.
and diarrhea in dogs (see Chapter 14) can infect cats
and cause antibodies that crossreact with FCoV. ▼ Key Point It is not the virus that causes widespread
Experimentally, CCV can cause enteritis, diarrhea, damage in FIP. The disease is a consequence of the
and even FIP in cats. cat’s immune reaction to the virus.

132
[Link] 11/10/05 4:50 PM Page 133

Chapter 10 / Feline Infectious Peritonitis (Feline Coronavirus) 133

• Natural immunity to FCoV is poorly understood but ▼ Key Point Most cats with FIP come from catteries
is presumed to be cell mediated rather than antibody
and shelters.
mediated. Circulating FCoV antibodies can actually
enhance progression of the disease.
• Any factor that increases FCoV replication in the
intestines of an infected cat will increase the proba-
bility of the virus mutating to a form that can cause
EPIDEMIOLOGY FIP. Thus, viral load, stress, immune impairment, cor-
ticosteroids, surgery, and concurrent disease (e.g.,
Prevalence feline leukemia virus or feline immunodeficiency
virus) can be risk factors for FIP.
• FCoV is ubiquitous in cats worldwide. In many • Inherited genetic susceptibility to FIP is a factor in
regions, 50% of cats are positive for coronaviral anti-
some purebred cats and in cheetahs.
bodies (i.e., seroprevalence). The majority of these
seropositive cats represent current or past inapparent
infection with non-mutated FCoV. Only some of these Transmission
develop into mutated FIP-causing infections; thus, • FCoV is primarily excreted in feces from cats with
the prevalence of FIP is much lower. inapparent enteric infection. Healthy carriers often
• The prevalence of FCoV infection is highest in cats shed FCoV in their feces for at least 10 months, and
confined in crowded groups, such as catteries, shel- some cats shed persistently for many years, possibly
ters, and multiple cat households, where the sero- for life. One-third of healthy FCoV-seropositive cats
prevalence ranges from 50% to 90%. are actively shedding virus. In high-density endemic
• In endemic catteries where the seroprevalence catteries, up to 60% of healthy cats may be shedding
approaches 90%, most FCoV-infected cats remain virus at any given time.
healthy and only 5% develop FIP. • FCoV is usually inactivated in 24 to 48 hours at room
• The seroprevalence of FCoV in free-roaming feral temperature, but the virus can survive up to 7 weeks
and stray cats is 12% to 15%. The lower prevalence is in dried fecal debris; thus, environmental contami-
presumed to relate to less social interaction and less nation with small particles of used litter is an impor-
exposure to fecally excreted virus than cattery cats. tant source of infection. Contaminated surfaces, food
• The seroprevalence in single-cat households is 15% and water dishes, and human clothing, shoes, and
or less. hands can act as fomites. Most disinfectants and
• In a large survey of North American veterinary teach- detergents easily destroy FCoV.
ing hospitals, FIP was diagnosed in 1 of every 200 new • Transmission to uninfected cats most frequently
feline accessions. This population mostly represents occurs through oronasal contact with virus-
sick cats seen by veterinarians. containing feces or contaminated material from the
• FCoV can infect most wild felids, including the lion, environment. Contaminated litter and dust particles
cougar, cheetah, jaguar, leopard, bobcat, sand cat, deposited on the fur are ingested during normal
caracal, serval, and lynx. Cheetahs are especially sus- grooming activity.
ceptible to developing FIP.

Risk Factors ▼ Key Point Indoor confinement in crowded groups

increases exposure to large doses of infectious
▼ Key Point Whenever FCoV exists in a cat, so does virus in feces in shared litter boxes.
the potential for developing FIP.
• In catteries, kittens are most frequently infected as
• FCoV infection occurs most often in young kittens they lose maternal-derived immunity after 6 weeks of
after maternal antibodies dissipate, between 6 and age through contact with feces from virus-shedding
16 weeks of age; thus, the infection rate is highest in adult cats. Removing kittens from contact with adult
catteries where kittens are raised in association with cats at 5 to 6 weeks of age prevents infection (see
virus-excreting adult cats. “Prevention”).
• Young cats have increased risk for developing FIP. • FCoV can also be excreted in saliva, respiratory secre-
The peak incidence for FIP is between 6 months tions, and urine, but these are unlikely to be impor-
and 3 years of age, although cats of any age can be tant sources of infection.
affected. • Transplacental transmission is possible but
• Crowded group confinement (multicat households, uncommon.
purebred catteries, shelters) increases risk. • Cats with FIP shed mostly the avirulent non-mutated
• Factors intrinsic to the shelter experience increase FCoV, not the virulent mutated virus that causes FIP;
fecal shedding of FCoV up to 1 million-fold after 1 thus, FIP itself is not directly contagious. Cats with FIP
week. also excrete less FCoV than healthy carriers.
[Link] 11/10/05 4:50 PM Page 134

134 Section 2 / Infectious Disease

greatest risk for developing FIP within the first 6 to

CLINICAL SIGNS 18 months after initial infection with FCoV.
• This encompasses the unpredictable but often pro-
Cats with non-mutated FCoV infection of the intestinal longed period of time it takes a carrier of the harm-
tract infrequently develop clinical signs (see Chapter less, non-mutated FCoV to develop the critical viral
14). This section describes the clinical manifestations mutation and then the time it takes the mutated virus
of mutated, FIP-producing FCoV infection. Cats with to produce clinical disease.
FIP often present initially with nonspecific and non- • The onset of clinical signs is often insidious, but occa-
localizing signs, such as fever, anorexia, inactivity, sionally it is sudden, especially in young kittens.
weight loss, vomiting, diarrhea, dehydration, and pallor
(anemia). As the disease progresses, these signs worsen ▼ Key Point Chronic fluctuating fever that is unre-
and additional clinical signs develop that indicate either sponsive to antibiotics is a frequent early sign of
body cavity effusions in the “wet” form of the disease or FIP.
organ-specific abnormalities in the non-effusive or
“dry” form (Table 10-1). Approximately 75% are effu- Clinical Course
sive and 25% are non-effusive. Some cats manifest • Once viral dissemination occurs and clinical illness
features of both effusive and non-effusive disease or develops, FIP is virtually always progressive and fatal.
change over time from one form to the other. However, there is considerable variation in the dura-
tion of clinical illness before death; 3 to 6 weeks is
Incubation and Clinical Course typical, but prolonged illness exceeding 6 months can
Incubation and Onset of Feline Infectious Peritonitis occur, as can intermittent illness punctuated by
periods of remission.
• The natural incubation period is extremely variable, • For effusive FIP, the clinical course is usually acute
ranging from a few weeks to several years. Cats are at
(days to weeks).
• For non-effusive FIP, the course is often chronic and
insidious (weeks to months).
Table 10-1. CLINICAL SIGNS AND • Cats with only ocular involvement sometimes survive
LABORATORY ABNORMALITIES IN FELINE for a year or more.
INFECTIOUS PERITONITIS
Effusive (Wet) Form of Feline
Nonspecific Signs
Chronic unresponsive fever of unknown origin
Infectious Peritonitis
Unexplained anorexia, lethargy, and weight loss
Effusion Signs ▼ Key Point In cats with the effusive (wet) form of FIP,
Fluid distension of the abdomen the predominant location of the inflammatory fluid
Dyspnea due to pleural effusion is the abdominal cavity in 62%, the thoracic cavity
Muffled heart sounds due to pericardial effusion in 17%, or both cavities in 21%.
Scrotal swelling
Organ-Specific Signs Abdominal Effusion (Peritonitis)
Abdominal disease
Enlarged, firm, irregular kidneys • Effusive FIP involving the peritoneal cavity causes
Icterus and hepatomegaly progressive, non-painful, fluid distension of the
Intestinal pyogranulomatous mass abdomen.
Splenomegaly
Pancreatitis • Effusion is detected by palpation and percussion of a
Mesenteric lymphadenopathy fluid wave. In the early stages, a small amount of
Omental adhesions and mass abdominal fluid may be detected by palpation of
Neurologic signs (multifocal and progressive) intestinal loops that feel excessively slippery, as
Uveitis (iridocyclitis; chorioretinitis) though the serosal surfaces are highly lubricated.
Pyogranulomatous interstitial pneumonia
Testicular enlargement (orchitis) Pain on palpation is infrequent.
Laboratory Abnormalities
• Extension of the peritoneal inflammation may
Anemia (nonregenerative) involve the gastrointestinal tract (vomiting, diar-
Neutrophilic leukocytosis or leukopenia; lymphopenia rhea), hepatobiliary system (jaundice), or pancreas
Elevated serum protein (hyperglobulinemia) (vomiting due to pancreatitis).
Elevated serum liver enzymes and bilirubin (also bilirubinuria) • Scrotal swelling may occur in intact males as a direct
Azotemia of primary renal origin
Proteinuria of renal origin
extension of the peritoneal inflammation and effu-
Pyogranulomatous or fibrinous body cavity effusion sion into the testicular tunics.
Elevated CSF protein and leukocytes (neutrophils) • Adhesions may organize the mesentery, omentum,
and viscera into an irregular, firm mass that is palpa-
CSF, cerebrospinal fluid; FIP, feline infectious peritonitis. ble in the cranioventral abdomen.
[Link] 11/10/05 4:50 PM Page 135

Chapter 10 / Feline Infectious Peritonitis (Feline Coronavirus) 135

• Abdominal effusion is confirmed by radiography, over the surface and infiltrating throughout the renal
ultrasonography, or abdominocentesis, and the cortex.
results of fluid analysis are highly indicative of FIP • Extensive renal involvement occasionally causes
(see “Diagnosis”). renal failure with polyuria-polydipsia and azotemia
(increased blood urea nitrogen [BUN] and serum
Thoracic Effusion (Pleuritis) creatinine).
• Proteinuria is a frequent laboratory finding in renal
• Dyspnea, tachypnea, and exercise intolerance are the FIP. In addition, large quantities of circulating
major presenting signs because lung expansion is immune complexes may lead to subclinical glomeru-
restricted by compression from fluid in the pleural lonephritis with any of the other forms of effusive and
space. non-effusive FIP.
• Cats with pleural effusion may prefer a sitting or
sternal recumbent posture to facilitate breathing. Liver Disease
Increased respiratory distress may occur with exercise,
with physical restraint in the hospital, or with reposi- • Pyogranulomatous hepatitis (hepatomegaly, jaun-
tioning in lateral recumbency (i.e., orthopnea). dice, and signs of hepatic failure) may occur in FIP.
• Thoracic effusion may cause muffled heart and lung • The most consistent laboratory abnormalities are
sounds on auscultation and hyporesonance and a bilirubinuria and hyperbilirubinemia. Mild to mod-
horizontal fluid line on thoracic percussion. erate elevations of serum liver enzymes (alanine
• Thoracic effusion is confirmed by radiography or tho- aminotransferase, alkaline phosphatase) and serum
racocentesis (see Chapter 164). The results of fluid bile acids also may occur.
analysis are highly indicative of FIP (see “Diagnosis”).
Disease in Other Abdominal Organs
Pericardial Effusion • Pyogranulomatous lesions may cause palpable
• Pericardial effusion due to fibrinous pericarditis may enlargement of the visceral lymph nodes, spleen, or
occur in FIP, with or without other effusions. In one omentum.
survey, FIP was the second most frequent cause of • Pyogranulomatous enterocolitis may cause diarrhea
feline pericardial effusion, accounting for 14% of and diffuse or masslike intestinal thickening, espe-
cases. cially in the ileocecocolonic region.
• Pericardial effusion in FIP does not usually cause • Pancreatic involvement can occasionally cause pan-
overt clinical signs. It may be suspected from auscul- creatitis and, rarely, diabetes mellitus.
tation (muffled heart sounds), thoracic radiography,
or electrocardiography, and it is confirmed by Ocular Disease
echocardiography (see Chapter 151). • Ocular lesions of FIP are usually bilateral and affect
the vascular tunic or uvea (uveitis). Lesions may
Non-effusive (Dry) Form of Feline sometimes cause blindness.
Infectious Peritonitis • Manifestations of exudative anterior uveitis (iridocy-
clitis) may include miosis, aqueous flare, keratic fi-
▼ Key Point The non-effusive (dry) form of FIP is brinocellular precipitates, hypopyon (“mutton-fat”
characterized by multifocal pyogranulomatous deposits of cells and fibrin), hyphema, anterior
inflammation and necrotizing vasculitis in various chamber adhesions (synechia), corneal edema, and
organs, such as the abdominal viscera (e.g., liver, deep neovascularization of the cornea (see Chapter
spleen, kidneys, pancreas, and intestines), eyes, 136).
central nervous system (CNS), and lungs. • Chorioretinitis from posterior uveal involvement
is detected by ophthalmoscopic examination and
Pyogranulomas appear as multiple discreet or coa- may include perivascular cuffing, exudative retinal
lescing gray-white nodular masses of variable size on the detachment, and retinal hemorrhages (see Chapter
surface and within the parenchyma of affected organs. 138).
These are often mistaken for tumors. Effusion is often
minimal or absent. The specific organs affected and the Neurologic Disease
degree of resulting organ failure determine the pre-
senting clinical signs.
• Multifocal pyogranulomatous meningoencephalitis
and myelitis are frequent in FIP. Inflammatory lesions
are perivascular and often involve the meningeal and
Kidney Disease ependymal layers. In one report, 29% of cats with FIP
• Pyogranulomatous nephritis causes the kidneys to developed neurologic signs. In a retrospective survey
become palpably enlarged, firm, and irregular of 286 cats with neurologic disease, lesions indicating
(“lumpy”), associated with pyogranulomas scattered FIP of the CNS were found in 16%. In another large
[Link] 11/10/05 4:50 PM Page 136

136 Section 2 / Infectious Disease

survey, FIP was the most common spinal disease in Intestinal Biopsy
cats.
• The neuroanatomic distribution of the lesions deter-
• Intestinal biopsy shows nonspecific lesions of villous
tip injury, with stunting and fusion of villi.
mines clinical signs; some of the most common are
ataxia, tremors, vestibular dysfunction, seizures,
• Immunofluorescence or immunohistochemical stain-
ing for viral antigen in intestinal biopsies is confir-
posterior paresis, hyperesthesia, and behavioral
matory.
changes. The relentless progression and multifocal
nature of the signs are characteristic features of
neural FIP. DIAGNOSIS OF FELINE INFECTIOUS
• Neuropathies occasionally involve the cranial nerves
PERITONITIS
(e.g., trigeminal or facial) or peripheral nerves (e.g.,
brachial or sciatic).
The diagnosis of FIP is usually suspected from clinical
• Cats with neural FIP often develop secondary hydro-
signs and the results of routine laboratory evaluations
cephalus when the inflammatory process obstructs
(see Table 10-1). There is no single reliable confirma-
the flow of cerebrospinal fluid (CSF). In one study of
tory test for FIP; thus, base the clinical diagnosis of FIP
24 cats with neural FIP, 75% had hydrocephalus.
on the combined results of well-chosen laboratory
Dilatation of the ventricular and central canal system
evaluations (e.g., hematology, serum chemistry, cyto-
is identified on computed tomography (CT) and
logy, serology, and virology), diagnostic imaging, and
magnetic resonance imaging (MRI) scans. Meningeal
histopathology (Table 10-2).
enhancement also is seen on MRI.
• The diagnosis of neural FIP depends on CSF analysis ▼ Key Point Clinical signs and laboratory abnormali-
(see the section on diagnosis). ties in cats with FIP are not specific for the disease;
however, collectively these may provide strong cir-
Pulmonary Disease cumstantial evidence for a presumptive diagnosis
• Pyogranulomatous pneumonia can be found in cats of FIP.
with FIP on thoracic radiographs or at necropsy, but
in most cases this is clinically silent or only causes a
Routine Laboratory Evaluations
mild cough. • Hematology often reveals nonspecific abnormalities
• This appears radiographically as a diffuse, poorly reflecting the chronic inflammatory response, such
defined, patchy or nodular interstitial pulmonary as nonregenerative anemia, neutrophilic leukocytosis
infiltrate. or leukopenia, and stress lymphopenia.
• Total serum protein and serum globulins (especially
Reproductive Disease gamma and alpha2) are increased by the chronic
immune stimulation in 70% of non-effusive cases and
• Testicular enlargement caused by pyogranulomatous 50% of effusive cases, whereas serum albumin is often
orchitis has been reported in FIP. decreased. One study showed that a decreased serum
• Contrary to what has been speculated in the past, albumin-to-globulin ratio (A/G < 0.8) indicates a
FCoV is not directly associated with cattery repro- high probability of FIP (92% positive predictive
ductive problems, neonatal deaths, or birth of weak value); while an A/G > 0.8 suggests that FIP is unlikely
or “fading” kittens. (61% negative predictive value). Serum protein elec-
trophoresis is usually not necessary.
• Serum chemistries may detect involvement of abdom-
DIAGNOSIS OF ENTERIC inal organs such as the liver (increased serum liver
FELINE CORONAVIRUS enzymes, bilirubin, and bile acids), kidneys
(increased creatinine and BUN), or pancreas
Fecal Virus Detection (increased pancreatic lipase immunoreactivity).
• Non-mutated FCoV infection is characterized by per- • Urinalysis findings may include proteinuria or
sistent viral replication in enterocytes and fecal shed- bilirubinuria.
ding of virus. Active fecal shedding of FCoV can be • Disseminated intravascular coagulation sometimes
confirmed by reverse transcription polymerase chain develops in FIP, resulting in prolonged coagulation
reaction (RT-PCR) assay of feces (see the later section times, decreased platelets, and increased fibrin
on RT-PCR assay) or by electron microscopy of feces. degradation products (see Chapter 23).
False negatives occur with both of these diagnostic
techniques. Diagnostic Imaging
• The quantity of fecal virus can fluctuate, so ideally Diagnostic imaging is useful for identifying organ sites
feces should be checked daily for 4 to 5 consecutive of involvement in FIP. Imaging also facilitates procure-
days before determining a cat is a non-shedder. ment of diagnostic fluid or biopsy specimens.
[Link] 11/10/05 4:50 PM Page 137

Chapter 10 / Feline Infectious Peritonitis (Feline Coronavirus) 137

Table 10-2. DIAGNOSTIC FEATURES OF FELINE

INFECTIOUS PERITONITIS
Parameter or Procedure Findings Suggestive of FIP

Age 6 mo to 3 yr (but all ages affected)

Habitat Cattery or multicat household
Signs (see Table 10-1) Fever (unresponsive)
Effusion (abdominal, thoracic)
Liver disease ( jaundice, etc.)
Renal disease (renomegaly)
Intestinal disease (GI signs)
Ocular disease (uveitis)
CNS disease (multifocal)
Pulmonary disease (cough)
Clinical course Progressive
Complete blood count Anemia (nonregenerative)
Neutrophilia or neutropenia; left shift
Lymphopenia
Neutrophil inclusions (immune complexes?)
Plasma proteins Increased total serum protein
Hyperglobulinemia (gamma, alpha2)
A/G ratio < 0.8
Hyperfibrinogenemia (400–700 mg/dl)
Serum chemistries Abnormal liver tests (increased ALT, ALP, bile acids,
bilirubin)
Azotemia (increased BUN, creatinine)
Increased PLI assay (pancreatitis)
Urinalysis Proteinuria, bilirubinuria
Radiography and ultrasonography Effusions (abdominal, thoracic)
Organomegaly (liver, kidney, intestine, etc.)
Organ infiltration (lung)
CT and MRI brain imaging Hydrocephalus, meningeal enhancement
Fluid analysis of effusion: Yellow, clear, sticky, foamy, fibrinous
Protein 4–10 g/dl (A/G ratio < 0.8, globulin > 50%, gamma
globulin > 32%)
Leukocytes 1,000–20,000 cells/ml
Cytology Pyogranulomatous exudate
Cerebrospinal fluid analysis:
Protein 50–350 mg/dl
Leukocytes 100–10,000 cells/ml
Cytology Neutrophils > mononuclear
Serology High FCoV antibody titer (see text)
RT-PCR assay FCoV nucleic acid (in blood, fluid, or tissue)
Histopathology Vasculitis and pyogranulomatous inflammation
Positive immunofluorescence and immunohistochemistry

A/G, albumin-to-globulin ratio; ALP, alkaline phosphatase; ALT, alanine transaminase; BUN, blood urea
nitrogen; CNS, central nervous system; CT, computed tomography; FCoV, feline coronavirus; FIP, feline
infectious peritonitis; GI, gastrointestinal; MRI, magnetic resonance imaging; PLI, pancreatic lipase
immunoreactivity; RT-PCR, reverse transcription polymerase chain reaction.

Radiography and Ultrasonography scans. Contrast enhancement of the meninges and

ependyma also may be seen on MRI.
Radiography is useful for confirming abdominal or tho-
racic effusion, abdominal organ enlargement (e.g., Fluid Analysis of Effusions
kidney and liver), or pulmonary infiltration. Affected
abdominal organs (liver, kidney, spleen, pancreas,
▼ Key Point Fluid analysis is usually sufficient for the
intestines, omentum, and lymph nodes) also can be clinical diagnosis of effusive FIP. The typical fluid
imaged, aspirated, and biopsied using ultrasonography. in effusive FIP is a highly proteinaceous pyogran-
ulomatous exudate.
CT and MRI of the CNS
In cats with neurologic FIP, secondary obstructive • FIP fluid appears clear, viscous, and straw yellow or
hydrocephalus is a common finding on CT and MRI golden. It may be tenacious or sticky and contain
[Link] 11/10/05 4:50 PM Page 138

138 Section 2 / Infectious Disease

flecks or strands of fibrin. The fluid gets frothy when Coronaviral Antibody Tests
shaken because of its high protein concentration,
and it may clot when refrigerated. Testing for coronaviral serum antibodies is informative
• FIP fluid has a high protein concentration, approach- as an epidemiological screening tool and as a diagnos-
ing that of plasma, ranging from 4 to 10 g/dl. Effu- tic aid for FCoV and FIP if a reliable laboratory is used
sive FIP is highly likely if the total fluid protein is and the results are interpreted properly.
more than 3.5 g/dl and the globulin portion is
greater than 50%. The A/G is usually less than 0.8. A Principles of Feline Coronavirus-Antibody Testing
gamma globulin percentage of greater than 32% by
fluid protein electrophoresis is also highly indicative ▼ Key Point A positive coronaviral antibody test
of FIP. An A/G ratio of greater than 0.8, or an means only that a cat has been exposed to some
albumin percentage greater than 50%, indicates that coronavirus at some time.
a disease other than FIP is highly likely.
• FIP fluid usually has a nucleated cell count ranging • Serology does not provide a definitive diagnosis of
from 1,000 to 20,000 cells/ml, which is low compared FIP because the antibodies in cats infected with harm-
with other exudates. less non-mutated FCoV are not distinguishable from
• The cytologic pattern of FIP fluid is pyogran- the antibodies in cats with FIP. A large percentage of
ulomatous exudate. The predominant cells are the healthy cat population is seropositive for FCoV
well-preserved (non-degenerate) neutrophils and antibodies, and most of these cats never develop FIP.
macrophages with variable numbers of plasma cells In addition, seropositivity does not distinguish active
and lymphocytes. from past infection.
• Additional evaluations on effusions can include assay • Antibodies to the related non-FCoVs (e.g., CCV and
for anti-FCoV antibodies (see “Coronaviral Antibody TGEV) crossreact with FCoV antibodies, lowering
Tests”) and RT-PCR assay to detect coronaviral specificity further.
nucleic acid (see the later section on PCR assay). • Seroconversion after initial exposure to FCoV takes 1
Antibody and RT-PCR assays on effusions may have to 3 weeks.
higher diagnostic value than serum testing.
• Staining macrophages in effusions for intracellular ▼ Key Point A positive coronaviral antibody test does
FCoV antigen using either immunofluorescence or not confirm a diagnosis of FIP, and the absence of
immunohistochemistry is the most definitive con- FCoV antibodies does not rule out a diagnosis of
firmatory test for effusive FIP, with 100% positive FIP.
predictive value (see “Immunofluorescence and
Immunohistochemistry”). • Many commercial diagnostic labs measure FCoV-anti-
body titers; however, methodologies are variable;
Cerebrospinal Fluid and Aqueous of the Eye thus, results cannot be compared between labs. For
example, whether a feline or non-feline viral antigen
• Analyses of CSF for neural FIP and aqueous fluid is used in the test procedure will influence the
from the anterior chamber of the eye for ocular FIP titer values that indicate the lowest and highest titer
have high diagnostic value. Both the protein con- levels.
centration and the nucleated cell count (neutrophils, • Use a reliable commercial diagnostic laboratory
macrophages, lymphocytes) are increased in the CSF that measures quantitative FCoV-antibody titer
of most cats with neural FIP (see Table 10-2) and in levels. Avoid using rapid in-office enzyme-linked
the aqueous humor of cats with intraocular FIP. immunosorbent assay (ELISA) tests that merely
• CSF and anterior chamber fluid can also be evaluated report positive or negative results without titer
for the presence of anti-FCoV antibodies (see the fol- quantification—this is less useful than titer informa-
lowing section). In FIP, the ratio of CSF or ocular anti- tion, and these tests give less consistent results than
bodies to serum antibodies is generally much higher conventional serologic tests.
than the ratio of CSF or ocular total protein to serum • Commercially available ELISA tests for detection of
total protein. antibody to the 7b gene have been marketed as “FIP-
• A PCR assay of CSF and anterior chamber fluid for specific tests.” However, mutation of this gene is not
coronaviral nucleic acid is not as useful (see the specific for FIP, and this test has no advantage over
section on PCR assay). conventional serologic assays.
• Serum or plasma samples for FCoV antibody testing
▼ Key Point Neurologic FIP is highly likely in cats that can be refrigerated or stored at -20∞C without affect-
have the combination of progressive neurologic ing the test.
signs (especially if multifocal), increased CSF • FCoV antibodies also can be measured in effusions,
protein and leukocytes (especially neutrophils), CSF, or anterior chamber fluid (aqueous). Some
and hydrocephalus on imaging. studies suggest that these are more diagnostically
[Link] 11/10/05 4:50 PM Page 139

Chapter 10 / Feline Infectious Peritonitis (Feline Coronavirus) 139

useful than serum testing but have similar inter-

pretation pitfalls. ▼ Key Point A healthy cat that is negative for anti-
bodies is likely to be free of FCoV and thus is not
shedding virus, is not infectious to other cats, and
Indications
is not at risk for FIP.
▼ Key Point Use the FCoV-antibody titer as a diag-
nostic aid for FIP rather than as a definitive diag- • False-negative test results are infrequent but occur in
nostic test. Interpret a positive titer to indicate that
up to 10% of confirmed FIP cases; thus, a negative
FIP is possible (low titer) or probable (high titer),
titer does not entirely rule out FIP. False-negative
when accompanied by supportive clinical findings.
titers can result from low antibody levels seen with
Interpret a negative titer to indicate that FIP is
peracute infection (less than 10 days after exposure),
unlikely.
the terminal stages of infection, or with consumption
of antibody in immune complexes. A laboratory error
or an insensitive assay system also explains some false
• Diagnostic aid in sick cats suspected of FIP negatives.
• Diagnostic aid in cats suspected of FCoV enteritis
• Healthy cats that have been exposed to cats with FIP • The evaluation of FCoV antibodies in effusions was
or FCoV found to have good predictive value for FIP (90%
positive, 79% negative), suggesting that fluid testing
• Screening catteries for the presence of FCoV may be more diagnostically useful than serum testing.
• Aid the process of creating an FCoV-free cattery
• Screening new cats before entering FCoV-free
catteries Polymerase Chain Reaction

Positive Antibody Test Results ▼ Key Point Interpret the RT-PCR test with other clin-
ical findings, and do not use this as the sole basis
A positive FCoV-antibody titer in a cat can indicate any for diagnosis of FIP.
of the following:
• Clinical FIP caused by a mutant variant of FCoV • The RT-PCR viral assay is used to detect coronaviral
• Healthy carrier of non-mutant FCoV nucleic acid in blood, fluids, tissue, or feces.
• Recovered from previous FCoV infection • A positive result indicates the presence of FCoV, but
• Seroconversion to other non-FCoVs it does not distinguish between the mutated, FIP-
• False-positive as a result of recent vaccination producing variants and non-mutated FCoV. Viremia
occurs not only in FIP but also in healthy FCoV
Guidelines for Interpretation carriers. One study found that up to 80% of cats
in endemic catteries can be viremic and RT-PCR–
▼ Key Point Coronaviral-antibody titers are not suffi- positive, and this was not predictive of FIP. Thus, RT-
ciently specific to be used as a definitive diagnos- PCR by itself is not a reliable confirmatory diagnostic
tic test. A positive titer, no matter how high, does test.
not confirm a diagnosis of FIP.
▼ Key Point PCR assays do not distinguish between
• In healthy cats, the height of the antibody titer cor- mutated and non-mutated FCoV.
relates with the virus replication rate in the intestine,
the likelihood of fecal shedding, and the amount of • In general, effusions and tissue specimens are more
virus shed in the feces. likely than blood to be positive in cats with FIP.
• In general, low titers have the least diagnostic value. Plasma is more sensitive than serum. The lowest yield
• The highest measurable titer level for the assay used is in CSF and urine, so these are not recommended
by the lab is highly suggestive of FIP but is still not for routine PCR testing.
confirmatory by itself. One study found 94% proba- • Fecal RT-PCR can be used to document fecal shed-
bility of FIP at the highest titer level. ding of FCoV, especially in healthy cats as a compo-
• Rising antibody titers are not helpful because they are nent of a cattery control program. Evaluate feces
seen in healthy cats with non-mutated FCoV, and daily for 4 to 5 consecutive days. Retesting over
titers fluctuate unpredictably in both FIP and aviru- several months is required to identify chronic persis-
lent infections. tent shedders.
• Healthy carriers of non-mutated FCoV living in • A false-negative RT-PCR results from degradation of
endemic cattery and multicat environments tend to sample RNA and poor laboratory technique. The
have higher titers than individual pet cats; thus, high primers used also may not detect all strains of FCoV.
titers might be less meaningful in cattery cats. • Samples for PCR require careful handling to avoid
Less than 10% of cattery cats with titers ever develop invalid results. Keep samples frozen and assay them
FIP. as soon as possible for optimal results.
[Link] 11/10/05 4:50 PM Page 140

140 Section 2 / Infectious Disease

Histopathology • Interferons in high doses have both immunomodu-

lating and antiviral activity (see below).
• Histopathology is considered the “gold standard” for • New anticoronaviral drugs under development for
confirming FIP. Thus, biopsy of affected tissues is a treatment of human severe acute respiratory syn-
valuable diagnostic procedure for identifying the dis- drome coronavirus might have future application in
tinctive FIP lesions of vasculitis and pyogranuloma- cats.
tous inflammation.
• Coronavirus can be identified in tissue specimens by
Immunomodulator Therapy
PCR, immunofluorescent antibody, and immunohis-
tochemistry techniques (see the next section). • Immunomodulators are intended to stimulate com-
promised immune function. Anecdotal reports of
Immunofluorescence and Immunohistochemistry clinical improvement with these agents have not been
substantiated by controlled studies; thus, conclusive
• Staining of macrophages in tissue specimens and evidence of efficacy in FIP is lacking.
effusions for intracellular FCoV antigen using either
immunofluorescence or immunohistochemistry is
• Nonspecific immune stimulation can theoretically
potentiate the immune-mediated consequences of
the most definitive confirmatory test for effusive FIP,
FIP.
with 100% positive predictive value (i.e., virtually no
false positives).
• Immunomodulator agents that have been used
unsuccessfully to treat FIP include Propionibacterium
• Do not rule FIP out when these tests are negative.
acnes (ImmunoRegulin), thioproline (Promodulin),
False negatives occur frequently when there are in-
acemannan (Carrisyn), levamisole, and cyclosporine.
sufficient infected macrophages in the specimen
or when the viral antigen is masked by competitive
• Human and feline interferons have been used to
treat FIP (see below).
binding with FCoV antibodies.
Interferon Therapy
TREATMENT Recombinant forms of interferon given parenterally in
high doses can have both antiviral and immunomodu-
▼ Key Point No treatment has been proved to reduce latory effects. Feline interferon-omega and human
the risk of developing FIP in healthy antibody- interferon-alpha have both been used to treat FIP.
positive FCoV carriers.
Feline Interferon-Omega
Treatment has not been proved to lower the high
mortality rate of FIP (>95%) or to slow progression of • Recombinant feline interferon-omega (rFeIFN-w) is
the disease. Various antiviral, immunomodulating, and available in Europe and Japan. To treat FIP, rFeIFN-
immunosuppressive drugs have been used to treat FIP, w (Virbagen Omega, Virbac) has been given at
but efficacy is highly questionable. Some cats show tem- 1,000,000 U/kg SC every other day until clinical
porary improvement of clinical signs with supportive remission then once or twice weekly. Expense will be
care and anti-inflammatory therapy using cortico- prohibitive for many owners.
steroids. The best candidates for palliative medical • In a preliminary uncontrolled study of 12 FIP cats
therapy are cats that are eating, active, and in good body using rFeIFN-w with prednisone (2 mg/kg PO q24h,
condition. Spontaneous remissions occur but are tapered to 0.5 mg/kg q48h), 33.3% of the cats
extremely rare. Euthanasia is appropriate in severely achieved complete remission for more than 2 years,
affected cats with poor quality of life. 33.3% achieved partial remission but died after 2 to
5 months, and 33.3% failed to respond and died.
▼ Key Point Nearly all cats with confirmed clinical FIP
eventually die, regardless of treatment. Human Interferon-Alpha

Antiviral Therapy • Parenteral high-dose human interferon-alpha

(rHuIFN-a) (100,000–1,000,000 U/kg IM) does not
• Some antiviral drugs (e.g., acyclovir and zidovudine appear to be effective in FIP, and it is unsuitable for
[AZT]) have no activity against FCoV. Other antiviral long-term treatment because cats develop antibodies
drugs (e.g., ribavirin) show in vitro activity against against the human protein after 3 to 7 weeks that
FCoV but are either too toxic for cats or ineffective inhibit the drug’s activity.
when used clinically to treat FIP. • Oral low-dose human interferon-alpha (Roferon,
• Ribavirin-treated kittens, for example, had more Hoffman-LaRoche; Intron-A, Schering-Plough) diluted
severe clinical signs and a shortened survival time to 30 U/ml and given at 30 units PO, daily for 7
compared with untreated kittens. Ribavirin also days on alternate weeks, has reportedly improved
causes serious side effects of hemolysis, bone marrow appetite and well-being with minimal side effects.
toxicity, and liver toxicity in cats. The rHuIFN-a given orally is destroyed by gastric acid
[Link] 11/10/05 4:50 PM Page 141

Chapter 10 / Feline Infectious Peritonitis (Feline Coronavirus) 141

and does not achieve systemic levels, but it may exert • Persistent drug-induced anorexia is a frequent
immunomodulating activity on oropharyngeal lym- problem when these drugs are given daily. An alter-
phoid tissue leading to cytokine release. native is pulse administration, using a large dose of
the drug once every 2 to 3 weeks. In this way, a few
Palliative Medical Therapy days after each dose the appetite usually rebounds
and is maintained between treatment cycles.
Some cats transiently improve with supportive care com- • Regardless of the regimen chosen, if no response is
bined with palliative medical therapy using a high dose noted within the first 2 to 4 weeks, consider the
of corticosteroid, with or without a cytotoxic alkylating therapy ineffective and either modify or discontinue
agent such as chlorambucil or cyclophosphamide it. If a positive response occurs, continue the treat-
(Table 10-3). ment indefinitely.
• Corticosteroids and alkylating drugs have no effect
on the virus, but by virtue of their anti-inflammatory Supportive Treatment
and immunosuppressive effects, they are aimed at These measures may improve quality of life and possi-
controlling the widespread immune-mediated in- bly survival time.
flammatory reaction that occurs in FIP.
• These drugs may adversely affect cellular immunity • Minimize stress as an exacerbating factor.
mediated by T lymphocytes and macrophages and • Perform intermittent body cavity drainage of effusion
thus have the potential to promote the viral infection. as needed to relieve dyspnea.
• Give parenteral fluid therapy.
• Give nutritional support (via tube-feeding tech-
▼ Key Point Only use immunosuppressive drugs to niques; see Chapter 3).
treat overt FIP. In healthy FCoV carriers and • Give aspirin (10 mg/kg q72h) to inhibit platelet
seropositive cats, these drugs could have the aggregation caused by vasculitis.
unwanted effect of promoting the onset of FIP. • Give antibiotics as needed to control complicating
bacterial infections.
• The principal side effects of alkylating drugs are • Treat anterior uveitis with topical corticosteroids and
anorexia and bone marrow suppression; thus, atropine (see Chapter 136).
monitor a complete blood count periodically (see • Give blood transfusions (for severe nonregenerative
Chapter 26 for more details on the use of these drugs). anemia).

Table 10-3. TREATMENT FOR FELINE INFECTIOUS PERITONITIS

Mechanism of Action Drug or Treatment* Dosage

Immunomodulator and antiviral Feline interferon-omega (Virbagen 1 million U/kg SC q48h until remission, then weekly
Omega)
Immunomodulator Human interferon-alpha (Roferon, 30 units PO q24h for 7 days on alternating weeks
Intron-A)
Anti-inflammatory and Prednisone 2–4 mg/kg, q24h, PO
immunosuppressive
Immunosuppressive† Chlorambucil (Leukeran) or ‡ 20 mg/m2, every 2–3 wks, PO
Cyclophosphamide (Cytoxan) 2–4 mg/kg, 4 days each week, PO, or 200–300 mg/m2,
every 2–3 wks, PO
Platelet aggregation inhibitor Aspirin 10 mg/kg, q72h, PO
Ozagrel HCl 5 mg/kg, q12h, SC
Topical ophthalmic for uveitis Prednisone acetate (1%) 2–3 drops/eye q6h
Atropine (1%) 1–3 drops/eye up to q6h for mydriasis
Supportive treatment Minimize “stress” —
Parenteral fluid therapy As needed to maintain hydration
Nutritional therapy via tube feeding See Chapter 3
Body cavity drainage (thoracentesis) As needed to relieve dyspnea
Blood transfusion As needed for severe anemia
Antibiotics for complicating infections Dosage based on the drug

*Cats most likely to respond are in good physical condition, are eating, and are free of neurologic signs, severe anemia, and feline leukemia virus or feline
immunodeficiency virus infections.
†For conversion of body weight to body surface area (m2), refer to conversion tables in Chapter 26.
‡Choose only one of these two alkylating agents, and combine with a corticosteroid.
[Link] 11/10/05 4:50 PM Page 142

142 Section 2 / Infectious Disease

• Use good husbandry practices (e.g., good sanitation,

PREVENTION ventilation, and feeding practices) and limit feco-oral
contamination.
Vaccination • Minimize overcrowding and stress. High-density
housing allows a high level of fecal contamination
▼ Key Point The currently available FIP vaccine of the environment, facilitating feco-oral spread of
does not appear to be effective and is not infection.
recommended. • Control feline leukemia virus in the cattery with vac-
cination, testing, and removal (see Chapter 8).
• A modified-live, temperature-sensitive strain of FIP • Do not breed male and female cats that have a history
coronavirus became available in 1991 as an intranasal of producing kittens that later developed FIP,
vaccine for use in cats 16 weeks of age and older because they potentially may pass on a genetic sus-
(Primucell FIP, Pfizer). ceptibility to FIP.
• The vaccine virus replicates locally in the nasophar- • The risk of infection from cat shows and breeding
ynx and intestines but not systemically because of exchanges is considered low; however, in these situa-
temperature sensitivity. It is supposed to stimulate tions do not allow sharing of food, water, or litter.
local nasal and gut mucosal immunity, salivary • Only allow healthy seronegative cats to enter a
immunoglobulin A antibody, and cell-mediated coronavirus-free cattery. Ideally, before a cat can be
immunity. It does not protect against enteric FCoV safely mixed with other cats, confirm the absence of
infection. fecal shedding based on at least 4 consecutive
• The lack of proven efficacy in kittens younger than negative fecal RT-PCR tests (although this may not be
16 weeks of age is a fundamental pitfall, because practical in many situations).
under most circumstances kittens first become
infected with FCoV between 6 and 16 weeks of Control of Viral Exposure
age.
• The 2000 Report of the American Association of The greatest source of viral contamination of the envi-
Feline Practioners Panel on Feline Vaccines states: ronment is from small particles of fecal debris and litter
“At this time there is no evidence that the vaccine that are carried throughout the facility by movement of
induces clinically relevant protection and its use is air, animals, and people. Any contaminated material
not recommended.” can end up on the fur and thus be ingested by the cats.
• Various genetically engineered recombinant vaccines Consider the following control measures to reduce the
are under development and may become available in environmental virus load and to minimize exposure:
the future. • Use at least one litter box for every two cats. Locate
litter boxes away from food and water bowls to avoid
Control of Feline Infectious cross-contamination. Ensure that the area is easy to
Peritonitis in Catteries clean and disinfect.
Control of FCoV infection in catteries, shelters, and
• Scoop feces from litter boxes daily, and replace litter
and disinfect litter boxes as often as possible (at least
multicat households is aimed at limiting virus spread, weekly).
minimizing exposure, and reducing stress. Infection
with FCoV is ubiquitous, so complete eradication is
• Dispose of used litter in sealed plastic bags.
rarely feasible, and even if eradication is successful, rein-
• Use dedicated food and water bowls for each animal,
and clean and disinfect bowls regularly.
fection frequently occurs. Test and removal based on
FCoV antibody testing is not practical since it is typical
• Brush the fur of cats regularly to remove contami-
nated fecal particles and litter that could be ingested
for 80% to 90% of cats in endemic catteries to be during grooming.
seropositive. Depopulation and starting over would be
required.
• Reduce cross-contamination between cats by housing
cats individually or in stable small groups of four or less.
• Avoid having too many kittens because they shed the
General Principles greatest amount of virus.
• Ideally, identify and eliminate persistent carriers that
▼ Key Point Do not consider healthy FCoV antibody- continuously shed large amounts of virus by using
positive cats to be harmless. Seropositive cats fre- fecal RT-PCR tests repeated over several months.
quently shed FCoV in their feces that contaminates
the environment and infects other cats, and FCoV
always has the potential to mutate and cause FIP.
Control of Viral Spread in Kittens
Isolation and early weaning can be effective for con-
• Be familiar with risk factors and transmission of FCoV trolling the spread of FCoV in kittens, as well as her-
(see “Epidemiology”). pesvirus and calicivirus.
[Link] 11/10/05 4:50 PM Page 143

Chapter 10 / Feline Infectious Peritonitis (Feline Coronavirus) 143

• Isolate queens 1 to 2 weeks pre-partum so that queens Bradshaw JM, Pearson GR, Gruffydd-Jones TJ: A retrospective study
can give birth and nurse kittens in isolation from of 286 cases of neurological disorders of the cat. J Comp Pathol
131:112–120, 2004.
other cats in the cattery. Elston T, Rodan I, Flemming D, et al: 2000 Report of the American
• Wean the kittens early and remove them from the Association of Feline Practioners and Academy of Feline Medicine
mother at 5 to 6 weeks of age, coinciding with the Advisory Panel on Feline Vaccines. Available at [Link]
time when maternal-derived immunity dissipates, [Link].
Gunn-Moore DA, Gruffydd-Jones TH, Harbour DA: A reverse
then keep the litter of kittens isolated from all other transcriptase-polymerase chain reaction (RT-PCR) of blood
cats until at least 16 weeks of age. samples from healthy cats and cats with clinical feline infectious
• Maintain strict quarantine procedures to prevent peritonitis. Vet Microbiol 62:193–205, 1998.
environmental or fomite transfer of FCoV. Haijema BJ, Volders H, Rottier PJ: Live, attenuated coronavirus
• Confirm seronegativity at 12 to 16 weeks of age or vaccines through the directed deletion of group-specific genes pro-
vide protection against feline infectious peritonitis. J Virol
prior to moving kittens to a new home to document 78:3863–3871, 2004.
effectiveness. Hartmann K: Feline infectious peritonitis. Vet Clin North Am Small
• The disadvantages of this approach are that it Anim Pract 35:39–79, 2005.
requires facilities for isolation and quarantine and Hartmann K, Binder C, Hirschberger J, et al: Comparison of differ-
ent tests to diagnose feline infectious peritonitis. J Vet Intern Med
that early weaning may adversely affect social devel- 17:781–790, 2003.
opment of the kittens. Pedersen NC, Addie D, Wolf A: Recommendations from working
groups of the International Feline Enteric Coronavirus and Feline
Infectious Peritonitis Workshop. Feline Pract 23:108–111, 1995.
SUPPLEMENTAL READING Rohrbach BW, Legendre AM, Baldwin CA, et al: Epidemiology of
feline infectious peritonitis among cats examined at veterinary
medical teaching hospitals. J Am Vet Med Assoc 218:1111–1115,
Addie DD, Jarrett O: Use of a reverse-transcriptase polymerase chain 2001.
reaction for monitoring the shedding of feline coronavirus by Shelley SM, Scarlett-Kranz J, Blue JT: Protein electrophoresis on effu-
healthy cats. Vet Record 148:649–653, 2001. sions from cats as a diagnostic test for feline infectious peritonitis.
Addie DD, Jarrett O: Feline coronavirus infections. In Greene CE J Am Anim Hosp Assoc 24:495, 1988.
(ed): Infectious Diseases of the Dog and Cat, 3rd ed. St. Louis: Vennema H, Poland A, Foley J, Pedersen NC: Feline infectious peri-
Elsevier, 2006, pp 89–103. tonitis virus arises by mutation from endemic feline enteric coro-
Addie DD, Paltrinieri S, Pedersen NC: Second international feline navirus. Virology 243:150–157, 1998.
coronavirus/feline infectious peritonitis symposium: Recommen- Weiss RC: Feline infectious peritonitis and other coronaviruses. In
dations from workshops of the second international feline coro- Sherding RG (ed): The Cat, Diseases and Clinical Management,
navirus/feline infectious peritonitis symposium. J Feline Med Surg 2nd ed. New York: Churchill Livingstone, 1994, pp 449–477.
6:125–130, 2004.