Zoonosis por Mycobacterium
Publicado: Jue, 02 Oct 2014, 01:38
Hola amigos, les cuento que soy estudiante de Medicina Veterinaria de la universidad Andres Bello, donde en nuestra malla curricular tenemos muchos ramos de peces, lo que es genial para mi que me gustan pero un infierno para mis compañeros dog y cat lovers jeje, bueno, les cuento que algo que siempre en nuestras clases es tema, son las zoonosis, para el que no conosca el termino, son las enfermedades que se pueden transmitir de los animales a los humanos. justamente ahora me encuentro haciendo un analisis de un paper en ingles sobre una infeccion aparentemente por mycobacterium en un pez koi de pileta, para nosotros los acuaristas es importante saber un poco sobre estas enfermedades ya que la prevencion puede ser tan facil y simple como no meter las manos con heridas en agua posiblemente contaminada. ya que no se como adjuntar un documento por acá copiaré la informacion del paper, saludos!!
Acid-fast bacteria found in granulomatous lesions in a
koi carp (Cyprinus carpio)
W.H. Wildgoose
655 High Road, Leyton, London E10 6RA
Abstract
This report describes the clinical investigation of a granulomatous lesion on
the pectoral fin of a koi (Cyprinus carpio) and demonstrates the practical
benefit of histological examination of a biopsy. In view of the chronic and
potentially infectious nature of the disease, the fish was killed humanely and
a detailed post mortem performed. The difficulties in isolating the acid-fast
bacteria by culture, despite their presence in the granulomata, is highlighted.
The epidemiology, control and zoonotic aspects are also discussed.
Introduction
Granulomatous diseases in fish are chronic progressive diseases which
produce various clinical signs depending on the organs affected. The
commonest organisms involved in these diseases of fish are acid-fast
bacteria, Mycobacterium and Nocardia species. The infection may take
several weeks or years to progress from an asymtomatic state to a clinical
illness. Both bacteria produce similar clinical characteristics and
identification of the pathogen by histological methods alone has resulted in
confusion (Austin and Austin 1993). Attempts to isolate the aetiological
agent by bacterial culture often fail, highlighting the fastidious nature of the
organisms (Austin and Austin 1993).
Mycobacteria from fish are zoonotic and can cause granulomatous nodules
in Man by contamination of wounds on the hands and fingers (Savin 1992).
The first report of acid-fast bacteria in freshwater fish was published by
Bataillon and others (1897) who described a tuberculous lesion in a
common carp (Cyprinus carpio) from a lake polluted by a hospital for
patients with tuberculosis. Nocardia species have occasionally been isolated
17 Fish Veterinary Journal • Number 4 • 1999
SCIENTIFIC PAPER Fish Veterinary Journal (1999) (4) 17–32
from lesions in man. Detailed reports of granulomatous diseases in
ornamental fish are rare (Lansdell and others 1993) and only a few
photographs of confirmed cases in koi (Cyprinus carpio) have been
published (Butcher 1992).
This report describes the clinical and post mortem findings in a koi with
systemic bacterial granulomata due to Gram-positive acid-fast bacteria and
discusses the management of the case.
Case history
The fish, an orange ogon variety of koi, was 490 mm long and weighed
1,585 grams. It had been owned for 2½ years and was kept in a formal
outdoor koi pond containing 13,500 litres with 19 other adult koi. It was
bought from a local fish dealer in north west London and was thought to be
about 5½ years old.
Clinical examination
The fish was presented on 14 July 1995 with small multiple swellings on
the left pectoral muscle (Fig 1). The owner noticed that the koi had not been
using the left pectoral fin for several weeks although the dark coloured
swellings had only been present for 3 weeks and had increased in size
during this time.
The koi was anaesthetised with tricaine methane sulphonate (MS-222,
Thomson & Joseph) by immersion to allow closer inspection and for a
biopsy to be taken for histological examination. On the dorsal aspect of the
pectoral muscle, multiple dark cystic swellings measuring between 3 mm
and 5 mm in diameter had developed beneath the skin. Some smaller cysts
on the ventral aspect also appeared to originate from within the muscle
mass. A biopsy containing a few cysts was taken from the dorsal surface and
fixed in 10% formal saline. Many small light brown nodules were seen in
the dermal tissues at the biopsy site, revealing the extent of the disease. It
was considered to be impractical to remove the entire lesion without
amputating the whole limb and so the wound was treated conservatively to
restore integrity of the skin. Povidone-iodine (Pevidine® Surgical Scrub,
Vericore) was applied to the surgical wound which was then packed with a
waterproof protective paste (Orabase®, ConvaTec). The fish was given
62·5 mg sulfadoxine and 12·5 mg trimethoprim (Borgal® 7·5%, Hoechst)
Fish Veterinary Journal • Number 4 • 1999 18
W.H. Wildgoose
by injection into the body muscle below the dorsal fin to prevent secondary
bacterial infection.
Microscopic examination of a skin scraping of body mucus revealed a few
non-motile trichodinids but these were not considered to be significant. No
further treatment was considered necessary and the fish was returned to the
pond.
Initial laboratory results
Histological examination revealed multifocal granulomatous lesions
throughout the sampled tissue (Fig 2). The level of inflammatory cell
infiltration and the classic appearance of epithelioid tissue surrounding the
granulomata (Fig 3) suggested that the pathology was related to a chronic
infectious agent. There was no neoplastic change. It was not possible to
identify any causative agent in the lesions with the haematoxylin and eosin
stain therefore differential staining with Ziehl-Neelsen (ZN) and Gram’s
stains were performed but still failed to reveal any bacterial organisms. A
second opinion also suggested that the lesions were bacterial granulomas.
Although the biopsy site had healed completely after 10 days, the fish was
lethargic and continued to swim with the left pectoral fin clamped against
its body. In view of the chronic and potentially infectious nature of this
disease the owner requested the fish to be killed.
Post mortem examination
The fish was killed with an overdose of anaesthetic on 30 July 1995. The
koi’s condition and body weight were good. The biopsy site had healed and
locally there was some epidermal hyperplasia. The pectoral fin was resected
at the ‘shoulder’ but the granulomata were only found at the distal end of
the muscle mass (Fig 4).
Internally there was a good amount of body fat in the abdomen but no free
ascitic fluid. This male fish had continued to eat well and the bowel was full
of food. There were several light brown granulomata measuring between
2 mm and 4 mm in the stroma of the liver (Fig 5). These were firm, well
circumscribed and could be easily expressed from the tissue (Fig 6). One
granuloma was found on the posterior pole of the left kidney but no lesions
were visible on the other organs. Routine samples were taken from the gill,
heart, spleen, liver, skin and, anterior and posterior kidney for histological
19 Fish Veterinary Journal • Number 4 • 1999
Granulomatous lesions in koi
Fish Veterinary Journal • Number 4 • 1999 20
FIG 1: Posterior aspect of the left pectoral fin at first presentation.
FIG 2: Histological section of the biopsy from site in Fig 1 revealing
numerous granulomata with inflammatory cell infiltration in the
dermis. (H&E) ×25
W.H. Wildgoose
21 Fish Veterinary Journal • Number 4 • 1999
FIG 3: The periphery of a granuloma showing classical structure with central
necrotic material (✽) surrounded by fibroblasts and inflammatory
cells. (H&E) ×400
FIG 4: Extensive granulomatous formation (arrows) found at the biopsy site.
Granulomatous lesions in koi
✽
Fish Veterinary Journal • Number 4 • 1999 22
FIG 5: Granulomata (arrowed) in the liver and mesenteric tissues.
FIG 6: Granulomata removed from the liver (scale in millimetres).
W.H. Wildgoose
examination. Further samples were taken from the left pectoral fin.
Bacteriological samples were taken from a crushed granuloma with a
standard swab and charcoal transport agar for culture on selective media
and air-dried slides were prepared for differential staining. Entire
granulomata were also sent for direct culture on to selective agar.
Final laboratory results
The results of differential staining of sections of the liver revealed masses of
rod-shaped bacteria in a small number of granulomata. The ZN stain
revealed weakly-stained acid-fast bacteria which were Gram-positive.
Grocott silver and periodic acid-Schiff (PAS) staining techniques failed to
demonstrate the presence of fungal elements in any of the sections.
Bacterial culture was unsuccessful despite the range of samples. Culture
was attempted on Dorset egg medium and Löwenstein-Jensen slopes but
these showed no growth after several weeks at 15° and 30°C. Differential
staining of the granuloma smears confirmed the presence of acid-fast
bacteria with ZN stain.
Recently, Gómez and others (1993) described the use of an immunocytochemical
staining method using an avidin biotin complex (ABC) to
reveal mycobacterial antigens in sections where bacilli were not seen when
ZN stain was used. A sample of the granulomatous lesion on the fin was
sent to Dr Gómez but no mycobacterial antigen was found.
Discussion
Granulomatous disease in fish can be due to infection with bacteria, fungi,
protozoan or myxozoan parasites. Histological examination with differential
staining methods revealed the presence of acid-fast bacteria in this case.
Both mycobacteria and Nocardia species are acid-fast bacteria which
produce granulomata and similar clinical signs. Nocardia species usually
produce an abundance of bacterial filaments in the centre of lesions,
irrespective of the stain used, whereas Mycobacterium species appear as
rodshaped
bacteria as seen in this fish.
The most common species of mycobacteria pathogenic to fish are
M marinum, M fortuitum and M chelonae. Of these, M marinum is the most
commonly isolated species and has been found in tropical freshwater and
Fish Veterinary Journal • Number 4 • 1999 24
W.H. Wildgoose
marine species (van Duijn 1981). Mycobacterium fortuitum is isolated less
frequently and has been reported mainly from fish in freshwater tropical and
temperate waters (Frerichs 1993). Mycobacterium chelonae has so far been
identified only in coldwater salmonids (Frerichs 1993).
Epidemiology
Mycobacteria and Nocardia species are free-living and found in the soil and
aquatic environment. In practice, mycobacteriosis is most frequently
recognised in aquarium fish probably because they are maintained under a
degree of captivity stress for long periods of time (Frerichs 1993). In the
absence of natural predation, this allows the slow progressive infection to
develop into a clinically diagnostic condition. Ambient temperatures may
also play a role. The relatively low temperature of outdoor ponds in the
United Kingdom by contrast to indoor room temperatures may account for a
lower incidence of mycobacteriosis in pond fish. The degree of infection
may therefore depend on the adaptability of the bacterial species to multiply
inside coldwater fish where water temperatures are rarely above 20°C .
Mycobacterial infections of fish are probably transmitted naturally by
ingestion of contaminated food or aquatic debris, although bacterial
invasion through damaged skin or gill tissue may also be possible (Frerichs
1993). There had been no visible damage to the affected fin during the
previous 2½ years of ownership but periodic infestation with ectoparasites
had been suspected. Koi frequently scavenge the pond surfaces for algae
and other food matter, and in this way may ingest various pathogenic
organisms. However, the filter system appeared to be efficient at removing
waste matter and debris from the bottom of this pond and should have
reduced this risk.
A wide range of amphibians including European frogs (Rana species) are
known to be susceptible to mycobacterial disease (Nigrelli and Vogel 1963).
Frogs are often found in garden ponds particularly during their mating
season in spring, and the owner reported that he had seen up to 30 frogs in
the pond on some occasions. Here, the water is artificially heated in winter
and this also attracts frogs into the pond. Occasionally, frogs would get
caught in the pump and be killed, resulting in the disintegration of their
body tissues and potential dissemination of infective organisms throughout
the pond system. However, following investigation of unusual mortality in
common frogs (R temporaria) in the UK, detailed post mortem and histo-
25 Fish Veterinary Journal • Number 4 • 1999
Granulomatous lesions in koi
logical examinations were carried out on 59 frogs and no acid-fast
organisms were found (Cunningham and others 1996).
The feeding of infected trash fish is known to spread disease to other fish
(Ross and Johnson 1962, Chinabut and others 1990) but this practice is not
relevant in koi ponds where the fish are usually fed on commercial pelleted
foods. Although these foods usually contain 30% to 40% fish protein as
herring meal, it is sterilised in the manufacturing process.
Ross and Johnson (1962) investigated trans-ovarian infection in chinook
salmon (Oncorhynchus tshawytscha) and although their results did not
demonstrate this under controlled conditions, it was not possible to exclude
this route as a means of transmission. However, other researchers have
confirmed that trans-ovarian infection occurs in viviparous platyfish
Xiphophorus maculatus (Conroy 1966), guppies Lebistes reticulatus
(Conroy and Conroy 1999) and in the eggs of Siamese fighting fish Betta
splendens (Chinabut and others 1994). The origin and breeding history of
this koi is unknown and the route of transmission cannot be confirmed here.
Clinical signs
The clinical signs of mycobacteriosis in fish depend on the species involved
and if vital organs are affected. These signs may include some of the
following: listlessness, anorexia, emaciation, dyspnoea, nervous disorders,
unusual behaviour, exophthalmos, skeletal abnormalities, skin discoloration
and external lesions such as ulceration and fin necrosis (van Duijn 1981,
Frerichs 1993). Internally the lesions are similar in tropical and coldwater
fish, and are seen as grey-white miliary granulomata which develop in the
tissues of the spleen, liver and kidney in particular. The significance of the
light brown colour of the granulomata found in this case is not known but
may be related to the species of fish.
Laboratory methods
In the literature much diagnosis of mycobacterial infection is based on
histopathology alone with little attempt to isolate and identify the organism.
This results in confusion since both Mycobacterium and Nocardia species
produce similar pathology.
Whereas the finding of acid-fast bacteria in granulomata on histological
sections is strongly suggestive of mycobacterial infection, failure to find
Fish Veterinary Journal • Number 4 • 1999 26
W.H. Wildgoose
them is not proof to the contrary (van Duijn 1981). In human lesions, when
M marinum is present, examination of ten or more sections may be required
before acid-fast bacteria are found (Chow and others 1983). Similar
difficulties have been experienced in piscine samples (A. Holliman personal
communication) and in older granulomata these bacteria are generally not
visible (van Duijn 1981). This may account for the failure to find bacteria in
most tissue sections in the present case.
Dorset egg and Löwenstein-Jensen agar slopes are now the standard
recommended media for the isolation of Mycobacterium and Nocardia
species (Austin and Austin 1993). All mycobacteria pathogenic to fish grow
when incubated between 20° and 30°C although M marinum shows slow
growth, often requiring up to 3 weeks. Cultures are not always obtained
despite evidence of infection, and sometimes a large inoculum is required
(Frerichs 1993). This may account for the failure to isolate any acid-fast
bacteria in the present case.
Failure to find mycobacterial antigen using the immuno-cytochemical
staining method may be due to the lack of organisms in the granulomatous
sample from the fin since most were found in histological sections of the
liver. It may also be due to the specificity of the antibodies since the
antiserum was raised against M paratuberculosis and M bovis.
Recently, other researchers have developed monoclonal and polyclonal
antibody probes to various strains of mycobacteria and these have been used
in the development of enzyme-linked immunosorbent assay (ELISA),
immunohistochemistry and immunofluorescence tests (Adams and others
1995, Adams and others 1996, Chen and others 1997). A polymerase chain
reaction (PCR) test has also been developed (Knibb and others 1993).
Treatment
Due to the risk to human health, the treatment of fish with bacterial
granulomata is questionable. Frerichs and Roberts (1989) unequivocally
advise against treatment and advocate a slaughter and disinfection policy.
However, the considerable value of the other koi precluded this approach.
Various protocols and medications have been suggested but the prolonged
use of antibiotic baths, up to three weeks, were impractical in this case
(van Duijn 1981, Conroy and Conroy 1999). The bacteria are encapsulated
within the granulomata and since the lesions have a poor blood supply, it is
27 Fish Veterinary Journal • Number 4 • 1999
Granulomatous lesions in koi
difficult to achieve therapeutic drug levels at the site of infection.
Since the visible lesion on the pectoral fin appeared to be localised, the
practicality of surgical amputation of the fin at the ‘shoulder’ joint was
considered. However, this would have created a large cavity in the body
wall making recovery and healing unpredictable.
Control
Prior to biopsy, the skin lesions on the affected koi had not ulcerated,
theoretically limiting the spread of infection. The duration of the disease in
this case and the extent of infection in the other koi is unknown. At present
there are no suitable tests which may identify infected fish although the
collection and concentration of faecal matter has been used to identify the
presence of carrier fish (Post 1987).
The use of environmental disinfectants has been suggested by van Duijn
(1981) and the owner has used chloramine-T on several occasions in the
past to treat ectoparasites, the last treatment being given several months
earlier. This chemical has some effect on external bacterial infections and
may have some disinfectant activity at the dosages used here.
Vaccines against mycobacterial infections in fish are not available at
present. However, experimental vaccines using M tuberculosis have been
shown to produce a cell-mediated immune response to mycobacterial
antigens in rainbow trout Oncorhynchus mykiss (Bartos and Sommer 1981).
More recently, Chen and others (1996) have reported elevation of both the
nonspecific and specific immune responses of rainbow trout to vaccination
with extracellular products from Mycobacterium species. Although there are
no reports of their use in practice, this suggests that Bacille Calmette-Guérin
(BCG)-like vaccines for fish could be developed in the future.
The owner decided to wait for further cases to develop and have routine
post mortem examinations performed on any fish dying in the pond in an
attempt to assess the prevalence of the disease. At the time of writing there
have been no further cases nor any deaths.
Zoonotic aspects
All species of mycobacteria from fish are capable of infecting man, of
which M marinum is the most common. This produces cutaneous nodules,
Fish Veterinary Journal • Number 4 • 1999 28
W.H. Wildgoose
commonly called ‘swimming pool granuloma’, usually on the hands,
elbows or knees which are slow to heal, often taking many months (Black
and Eykyn 1977). Although the owner himself has not had any such lesions,
this aspect of the case has been fully discussed with him. The current
medical approach to this disease in humans involves a six week course of
therapy with cotrimoxazole,
tetracycline (in particular minocycline) or
rifampicin often in combination with ethambutol (Savin 1992).
Rare, and sometimes fatal, cases of disseminated systemic infections with
M marinum have been reported in humans who are immunocompromised,
such as those infected with human immunodeficiency virus (HIV)
(Tchornobay and others 1992, Hanau and others 1994), receiving
chemotherapy or following organ transplant (Gombert and others 1981).
Conclusion
This report has shown the practical benefit of successfully taking a biopsy
of a lesion from a valuable fish for histological examination. In view of the
potential of a systemic infection being present, euthanasia was considered to
be more practical than local amputation of the limb. The difficulties in
isolating acid-fast bacteria by culture, despite their presence in the
granulomata have also been highlighted.
Acknowledgements: I am grateful to Peter Southgate MRCVS who
performed the histological examination. The second opinion was given by
Andrew Holliman MRCVS at the Veterinary Laboratories Agency, Penrith.
Bacteriological samples were sent for examination and culture to Dr Gavin
Barker at CEFAS, Weymouth and immuno-cytochemistry investigation was
performed by Dr S. Gómez at the University of Murcia in Spain. Andrew
Cunningham at London Zoo provided advice on common frogs. The histological
photographs (Figs 2 and 3) were taken by Dr Steve Feist at CEFAS,
Weymouth.
Acid-fast bacteria found in granulomatous lesions in a
koi carp (Cyprinus carpio)
W.H. Wildgoose
655 High Road, Leyton, London E10 6RA
Abstract
This report describes the clinical investigation of a granulomatous lesion on
the pectoral fin of a koi (Cyprinus carpio) and demonstrates the practical
benefit of histological examination of a biopsy. In view of the chronic and
potentially infectious nature of the disease, the fish was killed humanely and
a detailed post mortem performed. The difficulties in isolating the acid-fast
bacteria by culture, despite their presence in the granulomata, is highlighted.
The epidemiology, control and zoonotic aspects are also discussed.
Introduction
Granulomatous diseases in fish are chronic progressive diseases which
produce various clinical signs depending on the organs affected. The
commonest organisms involved in these diseases of fish are acid-fast
bacteria, Mycobacterium and Nocardia species. The infection may take
several weeks or years to progress from an asymtomatic state to a clinical
illness. Both bacteria produce similar clinical characteristics and
identification of the pathogen by histological methods alone has resulted in
confusion (Austin and Austin 1993). Attempts to isolate the aetiological
agent by bacterial culture often fail, highlighting the fastidious nature of the
organisms (Austin and Austin 1993).
Mycobacteria from fish are zoonotic and can cause granulomatous nodules
in Man by contamination of wounds on the hands and fingers (Savin 1992).
The first report of acid-fast bacteria in freshwater fish was published by
Bataillon and others (1897) who described a tuberculous lesion in a
common carp (Cyprinus carpio) from a lake polluted by a hospital for
patients with tuberculosis. Nocardia species have occasionally been isolated
17 Fish Veterinary Journal • Number 4 • 1999
SCIENTIFIC PAPER Fish Veterinary Journal (1999) (4) 17–32
from lesions in man. Detailed reports of granulomatous diseases in
ornamental fish are rare (Lansdell and others 1993) and only a few
photographs of confirmed cases in koi (Cyprinus carpio) have been
published (Butcher 1992).
This report describes the clinical and post mortem findings in a koi with
systemic bacterial granulomata due to Gram-positive acid-fast bacteria and
discusses the management of the case.
Case history
The fish, an orange ogon variety of koi, was 490 mm long and weighed
1,585 grams. It had been owned for 2½ years and was kept in a formal
outdoor koi pond containing 13,500 litres with 19 other adult koi. It was
bought from a local fish dealer in north west London and was thought to be
about 5½ years old.
Clinical examination
The fish was presented on 14 July 1995 with small multiple swellings on
the left pectoral muscle (Fig 1). The owner noticed that the koi had not been
using the left pectoral fin for several weeks although the dark coloured
swellings had only been present for 3 weeks and had increased in size
during this time.
The koi was anaesthetised with tricaine methane sulphonate (MS-222,
Thomson & Joseph) by immersion to allow closer inspection and for a
biopsy to be taken for histological examination. On the dorsal aspect of the
pectoral muscle, multiple dark cystic swellings measuring between 3 mm
and 5 mm in diameter had developed beneath the skin. Some smaller cysts
on the ventral aspect also appeared to originate from within the muscle
mass. A biopsy containing a few cysts was taken from the dorsal surface and
fixed in 10% formal saline. Many small light brown nodules were seen in
the dermal tissues at the biopsy site, revealing the extent of the disease. It
was considered to be impractical to remove the entire lesion without
amputating the whole limb and so the wound was treated conservatively to
restore integrity of the skin. Povidone-iodine (Pevidine® Surgical Scrub,
Vericore) was applied to the surgical wound which was then packed with a
waterproof protective paste (Orabase®, ConvaTec). The fish was given
62·5 mg sulfadoxine and 12·5 mg trimethoprim (Borgal® 7·5%, Hoechst)
Fish Veterinary Journal • Number 4 • 1999 18
W.H. Wildgoose
by injection into the body muscle below the dorsal fin to prevent secondary
bacterial infection.
Microscopic examination of a skin scraping of body mucus revealed a few
non-motile trichodinids but these were not considered to be significant. No
further treatment was considered necessary and the fish was returned to the
pond.
Initial laboratory results
Histological examination revealed multifocal granulomatous lesions
throughout the sampled tissue (Fig 2). The level of inflammatory cell
infiltration and the classic appearance of epithelioid tissue surrounding the
granulomata (Fig 3) suggested that the pathology was related to a chronic
infectious agent. There was no neoplastic change. It was not possible to
identify any causative agent in the lesions with the haematoxylin and eosin
stain therefore differential staining with Ziehl-Neelsen (ZN) and Gram’s
stains were performed but still failed to reveal any bacterial organisms. A
second opinion also suggested that the lesions were bacterial granulomas.
Although the biopsy site had healed completely after 10 days, the fish was
lethargic and continued to swim with the left pectoral fin clamped against
its body. In view of the chronic and potentially infectious nature of this
disease the owner requested the fish to be killed.
Post mortem examination
The fish was killed with an overdose of anaesthetic on 30 July 1995. The
koi’s condition and body weight were good. The biopsy site had healed and
locally there was some epidermal hyperplasia. The pectoral fin was resected
at the ‘shoulder’ but the granulomata were only found at the distal end of
the muscle mass (Fig 4).
Internally there was a good amount of body fat in the abdomen but no free
ascitic fluid. This male fish had continued to eat well and the bowel was full
of food. There were several light brown granulomata measuring between
2 mm and 4 mm in the stroma of the liver (Fig 5). These were firm, well
circumscribed and could be easily expressed from the tissue (Fig 6). One
granuloma was found on the posterior pole of the left kidney but no lesions
were visible on the other organs. Routine samples were taken from the gill,
heart, spleen, liver, skin and, anterior and posterior kidney for histological
19 Fish Veterinary Journal • Number 4 • 1999
Granulomatous lesions in koi
Fish Veterinary Journal • Number 4 • 1999 20
FIG 1: Posterior aspect of the left pectoral fin at first presentation.
FIG 2: Histological section of the biopsy from site in Fig 1 revealing
numerous granulomata with inflammatory cell infiltration in the
dermis. (H&E) ×25
W.H. Wildgoose
21 Fish Veterinary Journal • Number 4 • 1999
FIG 3: The periphery of a granuloma showing classical structure with central
necrotic material (✽) surrounded by fibroblasts and inflammatory
cells. (H&E) ×400
FIG 4: Extensive granulomatous formation (arrows) found at the biopsy site.
Granulomatous lesions in koi
✽
Fish Veterinary Journal • Number 4 • 1999 22
FIG 5: Granulomata (arrowed) in the liver and mesenteric tissues.
FIG 6: Granulomata removed from the liver (scale in millimetres).
W.H. Wildgoose
examination. Further samples were taken from the left pectoral fin.
Bacteriological samples were taken from a crushed granuloma with a
standard swab and charcoal transport agar for culture on selective media
and air-dried slides were prepared for differential staining. Entire
granulomata were also sent for direct culture on to selective agar.
Final laboratory results
The results of differential staining of sections of the liver revealed masses of
rod-shaped bacteria in a small number of granulomata. The ZN stain
revealed weakly-stained acid-fast bacteria which were Gram-positive.
Grocott silver and periodic acid-Schiff (PAS) staining techniques failed to
demonstrate the presence of fungal elements in any of the sections.
Bacterial culture was unsuccessful despite the range of samples. Culture
was attempted on Dorset egg medium and Löwenstein-Jensen slopes but
these showed no growth after several weeks at 15° and 30°C. Differential
staining of the granuloma smears confirmed the presence of acid-fast
bacteria with ZN stain.
Recently, Gómez and others (1993) described the use of an immunocytochemical
staining method using an avidin biotin complex (ABC) to
reveal mycobacterial antigens in sections where bacilli were not seen when
ZN stain was used. A sample of the granulomatous lesion on the fin was
sent to Dr Gómez but no mycobacterial antigen was found.
Discussion
Granulomatous disease in fish can be due to infection with bacteria, fungi,
protozoan or myxozoan parasites. Histological examination with differential
staining methods revealed the presence of acid-fast bacteria in this case.
Both mycobacteria and Nocardia species are acid-fast bacteria which
produce granulomata and similar clinical signs. Nocardia species usually
produce an abundance of bacterial filaments in the centre of lesions,
irrespective of the stain used, whereas Mycobacterium species appear as
rodshaped
bacteria as seen in this fish.
The most common species of mycobacteria pathogenic to fish are
M marinum, M fortuitum and M chelonae. Of these, M marinum is the most
commonly isolated species and has been found in tropical freshwater and
Fish Veterinary Journal • Number 4 • 1999 24
W.H. Wildgoose
marine species (van Duijn 1981). Mycobacterium fortuitum is isolated less
frequently and has been reported mainly from fish in freshwater tropical and
temperate waters (Frerichs 1993). Mycobacterium chelonae has so far been
identified only in coldwater salmonids (Frerichs 1993).
Epidemiology
Mycobacteria and Nocardia species are free-living and found in the soil and
aquatic environment. In practice, mycobacteriosis is most frequently
recognised in aquarium fish probably because they are maintained under a
degree of captivity stress for long periods of time (Frerichs 1993). In the
absence of natural predation, this allows the slow progressive infection to
develop into a clinically diagnostic condition. Ambient temperatures may
also play a role. The relatively low temperature of outdoor ponds in the
United Kingdom by contrast to indoor room temperatures may account for a
lower incidence of mycobacteriosis in pond fish. The degree of infection
may therefore depend on the adaptability of the bacterial species to multiply
inside coldwater fish where water temperatures are rarely above 20°C .
Mycobacterial infections of fish are probably transmitted naturally by
ingestion of contaminated food or aquatic debris, although bacterial
invasion through damaged skin or gill tissue may also be possible (Frerichs
1993). There had been no visible damage to the affected fin during the
previous 2½ years of ownership but periodic infestation with ectoparasites
had been suspected. Koi frequently scavenge the pond surfaces for algae
and other food matter, and in this way may ingest various pathogenic
organisms. However, the filter system appeared to be efficient at removing
waste matter and debris from the bottom of this pond and should have
reduced this risk.
A wide range of amphibians including European frogs (Rana species) are
known to be susceptible to mycobacterial disease (Nigrelli and Vogel 1963).
Frogs are often found in garden ponds particularly during their mating
season in spring, and the owner reported that he had seen up to 30 frogs in
the pond on some occasions. Here, the water is artificially heated in winter
and this also attracts frogs into the pond. Occasionally, frogs would get
caught in the pump and be killed, resulting in the disintegration of their
body tissues and potential dissemination of infective organisms throughout
the pond system. However, following investigation of unusual mortality in
common frogs (R temporaria) in the UK, detailed post mortem and histo-
25 Fish Veterinary Journal • Number 4 • 1999
Granulomatous lesions in koi
logical examinations were carried out on 59 frogs and no acid-fast
organisms were found (Cunningham and others 1996).
The feeding of infected trash fish is known to spread disease to other fish
(Ross and Johnson 1962, Chinabut and others 1990) but this practice is not
relevant in koi ponds where the fish are usually fed on commercial pelleted
foods. Although these foods usually contain 30% to 40% fish protein as
herring meal, it is sterilised in the manufacturing process.
Ross and Johnson (1962) investigated trans-ovarian infection in chinook
salmon (Oncorhynchus tshawytscha) and although their results did not
demonstrate this under controlled conditions, it was not possible to exclude
this route as a means of transmission. However, other researchers have
confirmed that trans-ovarian infection occurs in viviparous platyfish
Xiphophorus maculatus (Conroy 1966), guppies Lebistes reticulatus
(Conroy and Conroy 1999) and in the eggs of Siamese fighting fish Betta
splendens (Chinabut and others 1994). The origin and breeding history of
this koi is unknown and the route of transmission cannot be confirmed here.
Clinical signs
The clinical signs of mycobacteriosis in fish depend on the species involved
and if vital organs are affected. These signs may include some of the
following: listlessness, anorexia, emaciation, dyspnoea, nervous disorders,
unusual behaviour, exophthalmos, skeletal abnormalities, skin discoloration
and external lesions such as ulceration and fin necrosis (van Duijn 1981,
Frerichs 1993). Internally the lesions are similar in tropical and coldwater
fish, and are seen as grey-white miliary granulomata which develop in the
tissues of the spleen, liver and kidney in particular. The significance of the
light brown colour of the granulomata found in this case is not known but
may be related to the species of fish.
Laboratory methods
In the literature much diagnosis of mycobacterial infection is based on
histopathology alone with little attempt to isolate and identify the organism.
This results in confusion since both Mycobacterium and Nocardia species
produce similar pathology.
Whereas the finding of acid-fast bacteria in granulomata on histological
sections is strongly suggestive of mycobacterial infection, failure to find
Fish Veterinary Journal • Number 4 • 1999 26
W.H. Wildgoose
them is not proof to the contrary (van Duijn 1981). In human lesions, when
M marinum is present, examination of ten or more sections may be required
before acid-fast bacteria are found (Chow and others 1983). Similar
difficulties have been experienced in piscine samples (A. Holliman personal
communication) and in older granulomata these bacteria are generally not
visible (van Duijn 1981). This may account for the failure to find bacteria in
most tissue sections in the present case.
Dorset egg and Löwenstein-Jensen agar slopes are now the standard
recommended media for the isolation of Mycobacterium and Nocardia
species (Austin and Austin 1993). All mycobacteria pathogenic to fish grow
when incubated between 20° and 30°C although M marinum shows slow
growth, often requiring up to 3 weeks. Cultures are not always obtained
despite evidence of infection, and sometimes a large inoculum is required
(Frerichs 1993). This may account for the failure to isolate any acid-fast
bacteria in the present case.
Failure to find mycobacterial antigen using the immuno-cytochemical
staining method may be due to the lack of organisms in the granulomatous
sample from the fin since most were found in histological sections of the
liver. It may also be due to the specificity of the antibodies since the
antiserum was raised against M paratuberculosis and M bovis.
Recently, other researchers have developed monoclonal and polyclonal
antibody probes to various strains of mycobacteria and these have been used
in the development of enzyme-linked immunosorbent assay (ELISA),
immunohistochemistry and immunofluorescence tests (Adams and others
1995, Adams and others 1996, Chen and others 1997). A polymerase chain
reaction (PCR) test has also been developed (Knibb and others 1993).
Treatment
Due to the risk to human health, the treatment of fish with bacterial
granulomata is questionable. Frerichs and Roberts (1989) unequivocally
advise against treatment and advocate a slaughter and disinfection policy.
However, the considerable value of the other koi precluded this approach.
Various protocols and medications have been suggested but the prolonged
use of antibiotic baths, up to three weeks, were impractical in this case
(van Duijn 1981, Conroy and Conroy 1999). The bacteria are encapsulated
within the granulomata and since the lesions have a poor blood supply, it is
27 Fish Veterinary Journal • Number 4 • 1999
Granulomatous lesions in koi
difficult to achieve therapeutic drug levels at the site of infection.
Since the visible lesion on the pectoral fin appeared to be localised, the
practicality of surgical amputation of the fin at the ‘shoulder’ joint was
considered. However, this would have created a large cavity in the body
wall making recovery and healing unpredictable.
Control
Prior to biopsy, the skin lesions on the affected koi had not ulcerated,
theoretically limiting the spread of infection. The duration of the disease in
this case and the extent of infection in the other koi is unknown. At present
there are no suitable tests which may identify infected fish although the
collection and concentration of faecal matter has been used to identify the
presence of carrier fish (Post 1987).
The use of environmental disinfectants has been suggested by van Duijn
(1981) and the owner has used chloramine-T on several occasions in the
past to treat ectoparasites, the last treatment being given several months
earlier. This chemical has some effect on external bacterial infections and
may have some disinfectant activity at the dosages used here.
Vaccines against mycobacterial infections in fish are not available at
present. However, experimental vaccines using M tuberculosis have been
shown to produce a cell-mediated immune response to mycobacterial
antigens in rainbow trout Oncorhynchus mykiss (Bartos and Sommer 1981).
More recently, Chen and others (1996) have reported elevation of both the
nonspecific and specific immune responses of rainbow trout to vaccination
with extracellular products from Mycobacterium species. Although there are
no reports of their use in practice, this suggests that Bacille Calmette-Guérin
(BCG)-like vaccines for fish could be developed in the future.
The owner decided to wait for further cases to develop and have routine
post mortem examinations performed on any fish dying in the pond in an
attempt to assess the prevalence of the disease. At the time of writing there
have been no further cases nor any deaths.
Zoonotic aspects
All species of mycobacteria from fish are capable of infecting man, of
which M marinum is the most common. This produces cutaneous nodules,
Fish Veterinary Journal • Number 4 • 1999 28
W.H. Wildgoose
commonly called ‘swimming pool granuloma’, usually on the hands,
elbows or knees which are slow to heal, often taking many months (Black
and Eykyn 1977). Although the owner himself has not had any such lesions,
this aspect of the case has been fully discussed with him. The current
medical approach to this disease in humans involves a six week course of
therapy with cotrimoxazole,
tetracycline (in particular minocycline) or
rifampicin often in combination with ethambutol (Savin 1992).
Rare, and sometimes fatal, cases of disseminated systemic infections with
M marinum have been reported in humans who are immunocompromised,
such as those infected with human immunodeficiency virus (HIV)
(Tchornobay and others 1992, Hanau and others 1994), receiving
chemotherapy or following organ transplant (Gombert and others 1981).
Conclusion
This report has shown the practical benefit of successfully taking a biopsy
of a lesion from a valuable fish for histological examination. In view of the
potential of a systemic infection being present, euthanasia was considered to
be more practical than local amputation of the limb. The difficulties in
isolating acid-fast bacteria by culture, despite their presence in the
granulomata have also been highlighted.
Acknowledgements: I am grateful to Peter Southgate MRCVS who
performed the histological examination. The second opinion was given by
Andrew Holliman MRCVS at the Veterinary Laboratories Agency, Penrith.
Bacteriological samples were sent for examination and culture to Dr Gavin
Barker at CEFAS, Weymouth and immuno-cytochemistry investigation was
performed by Dr S. Gómez at the University of Murcia in Spain. Andrew
Cunningham at London Zoo provided advice on common frogs. The histological
photographs (Figs 2 and 3) were taken by Dr Steve Feist at CEFAS,
Weymouth.