Dierenfaculteitnr04_augustus2012_dierenfaculteit

224 Case report
Coinfection with Mycoplasma haemofelis and ‘Candidatus Mycoplasma
haemominutum’ in a cat with immune-mediated hemolytic anemia in Belgium
Co-infectie met Mycoplasma haemofelis en ‘Candidatus Mycoplasma haemominutum’ bij een kat met immuungemedieerde hemolytische anemie in België C. van Geffen
ABSTRACT
A young male domestic Shorthair was presented with weakness and anorexia of two days’ du-
ration. Physical examination showed pale mucous membranes, caused by severe, regenerative,
Coombs’ positive, hemolytic anemia. A blood smear revealed epicellular organisms compatible with
Mycoplasma spp. Real-time polymerase chain reaction (RT-PCR) on EDTA blood identified these or-
ganisms as Mycoplasma haemofelis and ‘Candidatus Mycoplasma haemominutum’. Despite the
lack of clearance of the organism from the blood, the cat responded well to antibiotic treatment with
doxycycline, together with immunosuppressive doses of corticosteroids.
Een jonge, mannelijke huiskat werd aangeboden voor zwakte en anorexia die reeds twee dagen aan- hielden. Op klinisch onderzoek werden bleke slijmvliezen gezien, veroorzaakt door erge, regeneratieve,Coombs’ positieve, hemolytische anemie. Een bloeduitstrijkje toonde epicellulaire organismen aan, com-patibel met Mycoplasma (vroeger bekend als Haemobartonella felis). Real-time polymerase chain reac-tion (RT-PCR) op EDTA-bloed identificeerde deze organismen als Mycoplasma haemofelis en ‘Candida-tus Mycoplasma haemominutum’. Ondanks de blijvende aanwezigheid van de organismen in het bloed,reageerde de kat goed op antibioticatherapie met doxycycline, samen met een immunosuppressieve dosiscorticosteroïden.
this infection to be present in cats worldwide (Duin etal., 2009, Fujihara et al., 2007; Gentilini et al., 2009; Hemobartonellosis or feline infectious anemia is caused by gram-negative bacteria that attach to the The incubation period after infection with Myco - outer surface of host erythrocytes. Previously, they plasma haemofelis varies from weeks to months, and were classified into genus Haemobartonella, family is followed by cycles of bacteremia which may last for Anaplasmataceae, order Rickettsiales. Recently, mo- months. Infected erythrocytes are less deformable in lecular techniques have resulted in a reclassification of circulation and elicit an immune response with subse- these organisms. Phylogenetically, they seem to be quent phagocytosis in lymphoid organs. Massive in- more closely related to organisms of the genus My- fection or severe anemia may result in death. Other an- coplasma. Two closely related species have been de- imals will recover but remain carriers despite their tected, and they have recently been renamed again.
immune response to the organism (Giger, 2005). Treat- Haemobartonella felis ‘Ohio strain’ or ‘large form’ is ment includes antibiotics, such as doxycycline, sup- currently called Mycoplasma haemofelis (Neimark et portive treatment with blood products in severely ane- al., 2001). Haemobartonella felis ‘California strain’ or mic animals, and possibly corticosteroids to halt ‘small form’, which seems to be a low-virulence bac- immune-mediated destruction of erythrocytes (Har- terium, has been given a candidate species name ‘Can- didatus Mycoplasma haemominutum’ (Foley and Ped-ersen, 2001). Some years ago, a third feline hemoplasma species was identified in Switzerland,and was named Candidatus Mycoplasma turicencis A 1.5-year-old, male, castrated domestic Shorthair (Willi et al., 2005). Recent studies have documented was presented with weakness and anorexia of 2 days’ Vlaams Diergeneeskundig Tijdschrift, 2012, 81 Figure 1. Blood smear from a cat infected with Myco -
Figure 2. Blood smear from a cat infected with Myco -
plasma organisms (Diff-Quick, magnification 1000 x).
plasma organisms (Diff-Quick, magnification 1000x).
Several Mycoplasma organisms are attached in an epi-
Howell-Jolly bodies (black arrow), nuclear remnants,
cellular location on erythrocytes (white arrows). A nu-
should not be confused with red blood cell parasites.
cleated red blood cell or normoblast (black arrow) is also
Epicellular Mycoplasma organisms are indicated by
present.
white arrows.
duration. The cat was mainly kept indoors and was cur- PCR, and revealed the presence of both Mycoplasma rently on vaccinations. There was no history of recent haemofelis (organism load of 2.4 x 106/mL of blood) drug administration or ectoparasite control. Recently, and ‘Candidatus Mycoplasma haemominutum’ (or- a young stray cat has also been adopted in the house- ganism load of 1.5 x 106/mL of blood).
hold. Physical examination revealed a lethargic cat The cat tested negative for feline leukemia virus with pale and mildly icteric mucous membranes, mild (FeLV) antigen (ELISA) and feline immunodeficiency tachycardia (180 beats/minute) and a low-grade sys- virus (FIV) antibodies (immunofluorescence antibody tolic murmur. Femoral pulses were bounding. Respi- test). Lateral and ventrodorsal chest radiographs were ratory rate and rectal temperature were normal (36 within normal limits. Abdominal ultrasound revealed breaths/minute and 38.6°C, respectively). The right mild diffuse splenomegaly with normal echogenicity.
prescapular lymph node was mildly enlarged. Ab- Fine needle aspirates of the right prescapular lymph node had a normal cellular distribution with the pre- A complete blood count (CBC) revealed a severe normochromic (mean corpuscular hemoglobin con- Intravenous fluid therapy (Hartmann®) was given at centration / MCHC 36.0 g/dL [reference 30-36 g/dL]), maintenance rate. A fresh blood transfusion was con- macrocytic (mean corpuscular volume / MCV 62 fl sidered but postponed until absolutely necessary, be- [reference 37-55 fl]), regenerative anemia (hematocrit cause of financial concerns of the owner. Despite the 8.1% [reference 24-45%] with an absolute reticulocyte severe anemia, the clinical and cardiorespiratory pa- count of 338000/µL [reference < 40000/µL]). Leuko- rameters remained stable during the following days.
cytes were within normal limits. The biochemistry Medical treatment included doxycycline (Vi- profile showed normal total protein (7.5 g/L [reference bramycine® 5 mg/kg twice daily) combined with an im- 5.4-8.2 g/L]) and albumin concentrations (3.6 g/L [ref- munosuppressive dose of parenteral dexamethasone erence 2.7-4.5 g/L]), as well as normal renal parame- (Rapidexon® 0.4 mg/kg once daily) initially, and oral ters, liver enzymes and electrolytes. However, the prednisolone (Prednisolone® 1 mg/kg twice daily) af- bilirubin concentration was increased (2.8 mg/dL [ref- ter the cat regained his appetite. A repeat blood smear erence 0.1-0.6 mg/dL]). Urinalysis was normal except on the second day of hospitalization no longer re- for bilirubinuria. A direct polyvalent Coombs’ test vealed Mycoplasma organisms. Signs of regeneration (Nordic, Tilburg, the Netherlands) was found strongly became more prominent (polychromasia, anisocytosis positive (titer 1:4096). A fresh blood smear was stained together with the presence of many normoblasts). A with Diff-Quick. A regenerative response was clearly CBC on the 6th day revealed a rising hematocrit demonstrated by the presence of polychromasia, aniso- (22.9%), with a MCV of 76.6 fl, the latter indicating the cytosis and a moderate number of normoblasts and presence of larger sized reticulocytes and thus an on- Howell-Jolly bodies (Figures 1 and 2). Heinz bodies going regenerative response. At that time, the systolic were absent. Several erythrocytes showed one or more murmur was no longer audible, suggesting that it was epicellular organisms resembling Mycoplasma spp. most likely caused by altered blood viscosity during the (Figures 1 and 2). An EDTA-blood sample was sent by regular mail to a veterinary laboratory specialized in The cat was sent home with oral doxycycline (Vi- molecular diagnostics for confirmation, characteriza- bramycine®) 5 mg/kg twice daily for three weeks and tion and quantification of the organisms by real-time tapering doses of oral prednisolone (Prednisolone®, 1 Vlaams Diergeneeskundig Tijdschrift, 2012, 81 mg/kg twice daily starting dose). Furthermore, a result in exposure of hidden antigens or changes in ex- monthly flea prevention with fipronil (Frontline®) was isting antigens, thereby activating the production of prescribed. Four weeks later (and five days after dis- host anti-erythrocyte antibodies. The anemia mostly re- continuation of prednisolone), the physical examina- sults from extravascular erythrophagocytosis by tion, blood work, fresh blood smear, Coombs’ test and macrophages in the spleen, liver, lungs and bone mar- PCR for both Mycoplasma organisms were repeated.
row (Harvey, 2006). Other conditions, such as inflam- According to the owner, the cat made full recovery and matory, infectious and neoplastic diseases, may result in was again alert and playful. The mucous membranes the immune-mediated destruction of erythrocytes in cir- were pink and the cardiac auscultation was normal, as culation or lymphoid organs. A thorough work-up, in- was the rest of the physical examination. The hemat- cluding retroviral testing, chest radiographs and ab- ocrit and red cell indices had returned within reference dominal ultrasound, revealed no significant limits (hematocrit 30%, MCV 45 fl, MCHC 36 g/dL).
abnormalities in the present case. The mild uniform The bilirubin level was normalized (0,2 mg/dL) and splenomegaly at initial presentation was believed to re- bilirubinuria was no longer present. Other parameters sult from sequestration of parasitized erythrocytes or were still within normal limits. The blood smear re- from extramedullary hematopoiesis. However, no fur- vealed normal red cell morphology and the absence of ther tests (fine-needle aspiration) were performed to visible epicellular organisms. The Coombs’ test re- sulted in a titer of 1:32. Real-time PCR revealed the There is still debate on the exact natural route of persistence of Mycoplasma haemofelis (hemoplasma transmission of Mycoplasma spp. A role of Cteno- load of 3.6 x 105/mL of blood) and ‘Candidatus My- cephalides felis has been proposed but is still unclear.
coplasma haemominutum’ (hemoplasma load of 3.2 x Recently, DNA of Mycoplasma haemofelis and ‘Can- 106/mL of blood). Another three weeks of doxycy- didatus Mycoplasma haemominutum’ has been am- cline were advised but declined by the owner because plified from blood of infected cats and from fleas col- the cat had been clinically healthy since discontinua- lected from client-owned cats by PCR (Lappin et al., tion and remained so during the following months.
2006). Woods et al. (2005) assessed the ability of fleasto transmit Mycoplasma spp. through hematophageous activity and found that both species were ingested byC. felis when allowed to feed on experimentally in- Anemia is a commonly encountered laboratory ab- fected cats. Only Mycoplasma haemofelis was found to normality in cats. It can be broadly classified into (1) be transmitted from C. felis to cats by hematogenous reduced hematopoiesis, such as anemia of inflamma- route. However, the efficiency of infection was poor tory disease, most FeLV infections and due to chronic and clinical or hematological signs of anemia did not renal failure, (2) external blood loss due to ectopara- develop. Another study failed to detect the transmission sites, tumors and hemostatic defects and (3) hemolysis.
of Mycoplasma spp. by oral ingestion of infected fleas In the present case, the macrocytic regenerative ane- or flea by-products (Woods et al., 2006). Recent stud- mia, supportive of the increased presence of larger ies also failed to show the successful transmission of sized reticulocytes in the peripheral circulation), to- ‘Candidatus Mycoplasma turicensis’ infection between gether with the normal serum protein concentration, cats by the oral or subcutaneous inoculation of ‘Can- hyperbilirubinemia and bilirubinuria, which is always didatus Mycoplasma turicensis’-infected saliva an abnormal finding in cats, were indicative of hemo - (Museux et al., 2009). Aggressive cat interaction could however result in transmission if a recipient cat is ex- Hemolytic anemia in this species is mostly caused by posed to blood of an infected cat (Museux et al., 2009). acquired disorders, such as hemobartonellosis, feline in- Despite the fact that many questions remain unans - fectious peritonitis (FIP), and other infections (e.g.
wered, most reports as well as the present case support FeLV subtype A), hypophosphatemia, oxidative agents the recommendation to maintain flea control.
(e.g. onions, zinc and drugs) and primary and secondary In the present case, it was not possible to determine immune-mediated processes (Adams et al., 1993; which of the two organisms was responsible for the Robertson et al., 1998; Sykes, 2003; Norris et al., 2005; clinical signs. Experimental infection with the large Kohn et al., 2006). Less often, hemolysis is a conse- variant, Mycoplasma haemofelis, resulted in more se- quence of certain hereditary defects (Giger, 2005).
vere clinical signs than with the small variant, ‘Can- Immune-mediated hemolytic anemia is considered didatus Mycoplasma haemominutum’ (Westfall et al., to occur relatively rarely in cats compared to dogs, and 2001). The same observations were made in client- is more often associated with secondary causes (Giger, owned cats with naturally occurring hemobartonel- 2005; Kohn et al., 2006). In the present case, the direct losis (Jensen et al., 2001). In the study of Westfall et al.
Coombs’ test was positive, indicating the presence of (2001), cats with dual infections had the most severe erythrocyte-bound antibodies and thus an immune- clinical abnormalities. Proposed explanations included mediated component to the hemolysis. Usually, most dose-dependent effects or predisposition to immune- damage to the red blood cell is not directly caused by mediated disease in ‘Candidatus Mycoplasma the parasite but rather because of immune-mediated in- haemominutum’-infected cats that were subsequently jury, especially in the acute phase of infection. The at- exposed to the more pathogenic Mycoplasma haemofe- tachment of the organisms to erythrocytes is thought to Vlaams Diergeneeskundig Tijdschrift, 2012, 81 In the present case, the clinical course was rapid, rofloxacin or marbofloxacin) would have resulted in with only a two-day history of clinical signs before ad- decreasing numbers of detectable microorganisms mission with severe anemia. Another (retrospective) (Dowers et al., 2002; Ishak et al., 2008). Cats may be- study suggested that ‘Candidatus Mycoplasma come PCR-positive again once the antibiotic treatment haemominutum’ might be a primary pathogen in some is stopped, or they can remain PCR-positive for months naturally infected, immunocompetent cats (Reynolds or years following infection, especially in the case of and Lappin, 2007). Concurrent diseases may predis- ‘Candidatus Mycoplasma haemominutum’ infection pose cats to develop more severe signs of anemia, as (Tasker, 2006). Complete recovery with negative PCR for example the presence of lymphoma (de Lorimier results has recently been described in a Mycoplasma and Messick, 2004). Pre-existing FeLV infection or haemofelis infected cat, after a 6-week course of doxy- FeLV-FIV co-infection also seems to potentiate the cycline (Braddock et al., 2004). Alternatively, in a severity of anemia caused by Mycoplasma haemofelis study by Dowers et al. (2009), treatment with prad- (George et al., 2002). However, the cat in this report ofloxacin resulted in a more effective clearance of or- tested negative for both retroviruses.
ganisms than with doxycycline. Another hypothesis for The diagnosis is based on the detection of the or- the persistent high bacterial load may be the concurrent ganism in an epicellular location on feline erythro- use of prednisolone, which might cause immunosup- cytes on a fresh blood smear. However, this technique pression with subsequent persistence of the organism.
has low sensitivity, especially in chronically infected However, in a study by Dowers et al. (2002), treatment animals, in cats with low parasite burden, or because with immunosuppressive doses of methylprednisolone of the cyclical parasitemia (Harvey, 2006; Tasker, acetate did not result in the recurrence of organisms in 2006). Organisms rapidly detach from red blood cells three cats after the successful treatment with doxycy- in vitro, probably reflecting organism death a few hours after blood collection (Tasker, 2006). Specificity In this case, immunosuppressive doses of corticos- may also be low, as organisms may be confused with teroids were added to the treatment protocol because Howell-Jolly bodies or staining artefacts, and it is not the positive Coombs’ test strongly suggested an im- possible to distinguish the two forms morphologically mune-mediated component to the hemolysis (Harvey, based on light microscopy alone (Jensen et al., 2001) 2006). This may indicate that one or both Mycoplasma
(Figure 2). Molecular techniques have greatly im- species may have acted as haptens and triggered an im- proved the ability to diagnose infections. The use of mune-mediated destruction of affected erythrocytes. real-time PCR offers several advantages over conven- In conclusion, this case documents that cats in Bel- tional PCR because it may both diagnose Mycoplasma gium may be co-infected with Mycoplasma haemofe- spp. and quantify infectious burden (Tasker et al., lis and ‘Candidatus Mycoplasma haemominutum’, 2003). This is important for determining the signifi- and that these organisms should be considered as pos- cance of a positive PCR result in infections with My- coplasma haemofelis and ‘Candidatus Mycoplasmahaemominutum’, the latter not always resulting in clin- ical disease in the host (Tasker et al., 2003). Further-more, quantification can help monitor the response to Adams L.G., Hardy R.M., Weiss D.J., Bartges J.W. (1993).
treatment (Tasker et al., 2003). On the other hand, a Hypophosphatemia and haemolytic anemia associated positive cytology or PCR does not necessarily correlate with diabetes mellitus and hepatic lipidosis in cats. Jour- with clinical illness, as clinically normal cats can have nal of Veterinary Internal Medicine 7, 266-271.
positive test results (Jensen et al., 2001). However, Braddock J.A., Tasker S., Malik R. (2004). The use of real- the more pathogenic form (Mycoplasma haemofelis) is time PCR in the diagnosis and monitoring of Mycoplasma rarely detected in cats without anemia (Jensen et al., haemofelis copy number in a naturally infected cat. Jour-nal of Feline Medicine and Surgery 6, 161-165.
de Lorimier L.-P., Messick J.B. (2004). Anemia associated In the present case, the combination of doxycycline with ‘Candidatus Mycoplasma haemominutum’ in a feline and prednisolone resulted in rapid clinical and hemato- leukemia virus-negative cat with lymphoma. Journal of logical recovery, despite the lack of clearance of mi- the American Animal Hospital Association 40, 423-427.
croorganisms as shown by the follow-up real-time PCR.
Dowers K.L., Olver C, Radecki S.V., Lappin M.R. (2002).
Total clearance of organisms, as confirmed by PCR, Use of enrofloxacin for treatment of large-form Haemo- seems to be rare, and most animals remain carriers bartonella felis in experimentally infected cats. Journal of (Dowers et al., 2002). Despite the persistence of both or- the American Veterinary Medical Association 2, 250-253.
ganisms after antibiotic treatment in the present case, a Dowers K.L., Tasker S., Radecki S.V., Lappin M.R. (2009).
tenfold decrease in Mycoplasma haemofelis load was Use of pradofloxacin to treat experimentally induced My- found by RT-PCR, and there was no longer evidence of coplasma hemofelis infection in cats. American Journal of ongoing hemolysis (normal hematocrit, MCV and biliru- Veterinary Research 70, 105-111.
bin level). This may indicate that Myco plasma haemofe- Duin M.B., Moyaert H., Pasmans F., Boyen F. (2009). He- lis was responsible for the hemolytic anemia. After motrofe mycoplasmen bij katten. Deel 1: literatuurover- treatment, only the Coombs’ test was still borderline zicht. Vlaams Diergeneeskundig Tijdschrift 78, 143-154.
positive. The question remains whether longer courses Foley J.E., Pedersen N.C. (2001). ‘Candidatus Mycoplasma haemominutum’, a low-virulence epierythrocytic para- of doxycycline or an alternative antibiotic (e.g. en- Vlaams Diergeneeskundig Tijdschrift, 2012, 81 site of cats. International Journal of Systematic and Evo- Journal of the American Animal Hospital Association 43, lutionary Microbiology 51, 815-817.
Fujihara M., Watanabe M., Yamada T., Harasawa R. (2007).
Robertson J.E., Christopher M.M., Rogers Q.R. (1998). Heinz Occurrence of ‘Candidatus Mycoplasma turicensis’ in- body formation in cats fed baby food containing onion fection in domestic cats in Japan. Journal of Veterinary powder. Journal of the American Veterinary Medical As- Medical Science 69, 1061-1063.
Gentilini F., Novacco M., Turba M.E., Willi B., Bacci M.L., Sykes J.E. (2003). Feline hemotropic mycoplasmosis (feline Hofmann-Lehmann R. (2009). Use of combined conven- hemobartonellosis). Veterinary Clinics of North America: tional and real-time PCR to determine the epidemiology Small Animal Practice 33, 773-789.
of feline haemoplasma infections in northern Italy. Jour- Tasker S., Helps C.R., Day M.J., Gruffydd-Jones T.J., Harbour nal of Feline Medicine and Surgery 11, 277-285.
D.A. (2003). Use of Real-Time PCR to detect and quantify George J.W., Rideout B.A., Griffey S.M., Pedersen N.C.
Mycoplasma haemofelis and ‘Candidatus Mycoplasma (2002). Effect of preexisting FeLV or FeLV and feline im- haemominutum’ DNA. Journal of Clinical Microbiology munodeficiency virus coinfection on pathogenicity of the small variant of Haemobartonella felis in cats. American Tasker S. (2006). Current concepts in feline haemobar- Journal of Veterinary Research 63, 1172-1178.
tonellosis. Companion Animal Practice 28, 136-141.
Giger U. (2005). Regenerative anemias caused by blood Westfall D.S., Jensen W.A., Reagan W.J., Radecki S.V., loss or hemolysis. In: Ettinger S.J., Feldman E.C. (editors).
Lappin M.R. (2001). Inoculation of two genotypes of He- Textbook of Veterinary Internal Medicine – Diseases of the mobartonella felis (California and Ohio variants) to induce Dog and Cat. 6th Edition, Elsevier Saunders, Philadel- infection in cats and the response to treatment with azithromycin. American Journal of Veterinary Research Harvey J.W. (2006). Hemotrophic Mycoplasmosis (Hemo- bartonellosis). In: Greene C.E. (editor). Infectious Dis- Willi B., Boretti F.S., Cattori V., Tasker S., Meli M.L., eases of the Dog and Cat. 3rd Edition. Elsevier Saunders, Reusch C., Lutz H., Hofmann-Lehmann R. (2005). Iden- tification, molecular characterization, and experimental Ishak A.M., Dowers K.L., Cavanaugh M.T., Powell C.C., transmission of a new hemoplasma isolate from a cat with Hawley J.R., Radecki S.V., Lappin M.R. (2008). Mar- hemolytic anemia in Switzerland. Journal of Clinical Mi- bofloxacin for the treatment of experimentally induced Mycoplasma haemofelis infection in cats. Journal of Vet- Willi B., Tasker S. Boretti F.S., Doherr M.G., Cattori V., Meli erinary Internal Medicine 22, 288-292.
M.L., Lobetti R.G., Malik R., Reush C.E., Lutz H., Hof- Jensen W.A., Lappin M.R., Kamkar S., Reagan W.J. (2001).
mann-Lehmann R. (2006). Phylogenetic analysis of ‘Can- Use of a polymerase chain reaction assay to detect and dif- didatus Mycoplasma turicensis’ isolates from pet cats in ferentiate two strains of Haemobartonella felis in naturally the United Kingdom, Australia, and South Africa, with infected cats. American Journal of Veterinary Research 62, analysis of risk factors for infection. Journal of Clinical Kohn B., Weingart C., Eckmann V., Ottenjann M., Leibold Woods J.E., Brewer M.M., Hawley J.R., Wisnewski N., W. (2006). Primary immune-mediated hemolytic anemia Lappin M.R. (2005). Evaluation of experimental trans- in 19 cats. Journal of Veterinary Internal Medicine 20, mission of ‘Candidatus Mycoplasma haemominutum’ and Mycoplasma haemofelis by Ctenocephalides felis to cats.
Lappin M.R., Griffin B., Brunt J., Riley A., Burney D., Haw- American Journal of Veterinary Research 66, 1008-1012.
ley J., Brewer M.M., Jensen W.A. (2006). Prevalence of Woods J.E., Wisnewski N., Lappin M.R. (2006). Attempted Bartonella species, haemoplasma species, Ehrlichia species, transmission of ‘Candidatus Mycoplasma haemominu- Anaplasma phagocytophilum, and Neorickettsia risticii tum’ and Mycoplasma haemofelis by feeding cats infected DNA in the blood of cats and their fleas in the United Ctenocephalides felis. American Journal of Veterinary States. Journal of Feline Medicine and Surgery 8, 85- 90.
Museux K., Boretti F.S., Willi B., Riond B., Hoelzle L.E., Wittenbrink M.M., Tasker S., Wengi N., Reusch C.E.,Lutz H., Hofmann-Lehmann R. (2009). In vivo transmis-sion studies of ‘Candidatus Mycoplasma turicensis’ in adomestic cat. Veterinary Research 40, 45.
Neimark H., Johansson K.E., Rikihisa Y, Tully J.G. (2001).
Proposal to transfer some members of the genera Haemo-bartonella and Eperythrozoon to the genus Mycoplasmawith descriptions of ‘Candidatus Mycoplasma haemofe-lis’, ‘Candidatus Mycoplasma haemomuris’, ‘CandidatusMycoplasma haemosuis’ and ‘Candidatus Mycoplasmawenyonii. International Journal of Systematic and Evo-lutionary Microbiology 51, 891-899.
Norris J.M., Bosward K.L., White J.D., Baral R.M., Catt M.J., Malik R. (2005). Clinicopathological findings asso-ciated with feline infectious peritonitis in Sydney, Aus-tralia: 42 cases (1990-2002). Australian Veterinary Jour-nal 83, 666- 673.
Reynolds C.A., Lappin M.R. (2007). ‘Candidatus Myco - plasma haemominutum’ infections in 21 client-owned cats.

Source: http://vdt.ugent.be/sites/default/files/art81405.pdf