Abstract
Objectives
Information regarding the clinical course of natural infection with feline haemotropic mycoplasmas (haemoplasmas) is limited. The objective of the study was to describe the clinical findings and course of disease in naturally infected cats with haemoplasmosis and anaemia.
Methods
A retrospective analysis was performed on patient data from cats presenting with anaemia and haemoplasma infection regarding signalment, clinical signs, laboratory data and course of infection. The diagnosis was confirmed by conventional haemoplasma PCR analysis.
Results
Haemoplasma infection was found in 22 anaemic (haematocrit 5–25% [median 17%]; reference interval 30–44%) cats (‘Candidatus Mycoplasma haemominutum’, n = 12; Mycoplasma haemofelis, n = 3; ‘Candidatus Mycoplasma turicensis’, n = 2; species not determined, n = 4; coinfection with all three species, n = 1) between 2005 and 2014. Thirteen of the cats had concurrent diseases. All cats underwent antibiotic treatment; 15 cats received blood products. Six cats were euthanased within 11 weeks owing to concurrent disease, persistent severe anaemia or financial constraints. Ten cats underwent follow-up for a period of 14–199 weeks (median 26 weeks). Haemoplasma PCR was negative in 5/7 cases after 3–23 weeks. PCR remained positive in two cases after 18 and 199 weeks, respectively. Reactivation of the haemoplasma infection occurred in two cats, once and three times, respectively, up to 177 weeks after initial presentation. Reactivation was suspected in two further cases. Owing to concurrent disease, four of the 10 follow-up cats were euthanased 14–180 weeks after initial presentation.
Conclusion and relevance
Infection with haemoplasma species is often chronic, can reactivate months later and is rarely a reason for euthanasia.
Introduction
The prevalence of PCR-confirmed haemotropic mycoplasmosis (haemoplasmosis) in cats in Germany ranges from 7–22%.1–3 A recent study found 19/265 cats (7.2%) to be PCR-positive for infection with one or more haemoplasma species: ‘Candidatus Mycoplasma haemominutum’ (CMhm; 5.3%), Mycoplasma haemofelis (Mhf; 1.5%) and ‘Candidatus Mycoplasma turicensis’ (CMt; 1.1%). 3 ‘Candidatus Mycoplasma haematoparvum’, which is closely related to CMhm, has not been detected in Germany. 4 Several studies have documented a positive correlation between haemoplasmas and outdoor access, sex (male) and feline immunodeficiency virus (FIV) infection.3,5,6 Feline haemoplasmosis has a very variable course of disease, which is influenced by the immune status of the host, concurrent infection, and disease and medication (eg, immunosuppressants).4,7 The haemoplasma species also determines the disease process, with Mhf being the most pathogenic, while CMhm and CMt rarely result in clinical disease. 8
Clinical signs and the laboratory data of cats experimentally infected with haemoplasmas have been documented.9,10 However, there are few published data on the natural course of infection.11–15 No long-term follow-up studies of natural infections have been published. The following study describes the clinical signs, laboratory data and course of infection in 22 anaemic cats with haemoplasmosis.
Materials and methods
A retrospective analysis of patient data from all cats presenting to the Small Animal Clinic of the Freie Universität Berlin with anaemia and haemoplasma infection between January 2005 and September 2014 was performed. In all cases, diagnosis of haemoplasma infection was confirmed by a conventional generic haemoplasma PCR assay (Institute of Comparative Tropical Medicine and Parasitology, LMU, Munich; and Laboklin, Bad Kissingen, Germany). If a positive PCR was generated, the haemoplasma species was then determined in most cases by a species-specific PCR. Haematology (Cell Dyn 3500 [Abbott Diagnostika]; XT-2000i [Sysmex]) was conducted on potassium EDTA (K-EDTA)-anticoagulated whole blood, and clinical chemistry (Konelab 30 i/60 I; Fisher Scientific) on heparin plasma. Anaemia was defined as a haematocrit (Hct) of <30% (reference interval [RI] 30–44%). Red discolouration of the plasma was considered to be diagnostic for intravascular haemolysis. The number of aggregate and punctate reticulocytes were microscopically counted and expressed as a percentage. The absolute number of reticulocytes (/μl) was calculated as follows: (aggregate reticulocytes [%] × number of erythrocytes)/100. The baseline value for a regenerative anaemia was set at an aggregate reticulocyte count of >40,000/μl. 16 Most cats were tested for erythrocytic agglutination by mixing one drop of K-EDTA whole blood with one drop of 0.9% NaCl on a microscope slide. Positive samples underwent repeat washing of K-EDTA whole blood three times with 0.9% NaCl solution. The persistence of agglutination (gross or microscopic) after the washing procedure was considered to be autoagglutination. In addition to these tests, in the majority of cases a direct Coombs’ test (Department of Immunology, Tierärztliche Hochschule, Hannover, Germany), feline leukemia virus (FeLV) antigen/FIV antibody test (SNAP Combo Plus [IDEXX Laboratories]; ELISA [Laboklin]), FeLV-provirus PCR test (Laboklin), thoracic and abdominal radiographs, abdominal ultrasonography, and cytology of a bone marrow aspirate were performed. Medical records were reviewed with regard to flea and tick infestation of the cats, concurrent diseases, treatment and previous evidence of haemoplasmosis. In cases without follow-up in the clinic the owner was interviewed by telephone. Descriptive statistics were used to report the results of age, haematology, plasma biochemistry and drug dosages. Statistical analyses were performed with Microsoft Excel (2010).
Results
Haemoplasma infection was found in 22 anaemic cats. With species-specific primers CMhm (n = 12), Mhf (n = 3) and CMt (n = 2) were detected. One cat was infected with all three species. In four cats the haemoplasma species was not determined.
Signalment
Signalment details are shown in Table 1. At the time of the first presentation the cats ranged in age from 7 months to 16 years (median 7 years).
Signalment, haemoplasma species and concurrent disease in 22 cats with natural haemoplasma infection. When a concurrent diagnosis is stated (eg, case 1), this indicates the diseases found at initial presentation, before the subsequent diagnosis of haemoplasma infection was made
BSH = British shorthair; DSH = domestic shorthair; MN = male neutered; M = male; FN = female neutered; F = female; CMhm = ‘Candidatus Mycoplasma haemominutum’; Mhf = Mycoplasma haemofelis; CMt = ‘Candidatus Mycoplasma turicensis’; ND = not done; URTD = upper respiratory tract disease; HCM = hypertrophic cardiomyopathy; IMT = immune-mediated thrombocytopenia; FIV = feline immunodeficiency virus infection; DIC = disseminated intravascular coagulation
Concurrent diseases
Thirteen of the 22 cats presented with additional diseases concurrent to anaemia before the subsequent diagnosis of haemoplasma infection was made (Table 1).
Medical history
The nine cats not suffering from concurrent disease exhibited lethargy (n = 8), inappetence (n = 2), tachypnoea (n = 2) and weight loss (n = 1). The 13 cats with concurrent diseases presented with lethargy (n = 7), inappetence (n = 7), weight loss (n = 3), tachypnoea (n = 1), tenesmus (n = 1), trauma (high-rise syndrome; n = 1) and a bite wound (n = 1). Seventeen cats had outdoor access and three a previous flea infestation. Four cats were pretreated with immunosuppressive agents. Two cats (both infected with CMhm) had received a blood transfusion: case 5 for immune-mediated thrombocytopenia (IMT; 3 years previously) and case 2 for treatment of anaemia (12 h previously).
Physical examination
Common findings in nine cats without concurrent disease were pale mucous membranes (n = 8) and tachycardia (n = 5). Rectal temperature ranged from 33.1–39.1°C; four cats exhibited hypothermia (33.1–37.1°C) and two cats a mildly increased temperature (39.1°C). Case 14 presented with a collapse; in one cat (case 6) flea infestation was detected.
Thirteen cats with concurrent diseases presented with pale mucous membranes (n = 7) and tachycardia (n = 2). The rectal temperature ranged from 37.1–39.9°C; two cats had hypothermia (37.1–37.8°C) and three cats were febrile (39.7–39.9°C). One cat each presented with icterus (case 8), sneezing and petechiae (case 6), vomiting (case 15), stomatitis (case 13), infected bite wound (case 16), constipation (case 1), lymphadenopathy (case 10), ocular and nasal discharge (case 21), and flea infestation (case 1).
Clinical pathology
Hct ranged from 5–25% (median 17%) on initial presentation. Cats with and without concurrent diseases had a median Hct of 20% (range 6–25%) and of 11% (range 5–21%), respectively. Regenerative anaemia was initially diagnosed in 10/20 cases (five cats with and five cats without concurrent diseases: absolute aggregate reticulocytes 60,690–437,500/μl [median 143,340/μl]; punctate reticulocytes 25,060–525,500/μl [median 104,610/μl]). In 6/10 cats with a non-regenerative anaemia the reticulocyte counts could be monitored between 5 and 90 days, and the anaemia became regenerative in three of these six cats, after 5 and 7 days in two cats with concurrent disease (cases 5 and 1), and after 10 days in one cat without concurrent disease (case 3); the other three cats suffered from FIV infection (n = 2) and pure red cell aplasia (PRCA). Large numbers of haemoplasmas were found in the blood smear of one patient (case 18). In all cats the number of Heinz bodies was <5%. Autoagglutination of erythrocytes was detected in 10 cases. In 7/15 cats the Coombs’ test was positive; these cats were infected with CMhm (n = 3), Mhf (n = 1), all three species (n = 1) and in two cases the species was not determined. Two of these seven cats suffered from concurrent diseases (cases 4 and 21). Three of eight cats with a negative Coombs’ test were pretreated either with short-acting methylprednisolone (n = 2, 4 and 10 days before) or prednisolone/ciclosporin (n = 1, prednisolone for 3 years, ciclosporin for 4 weeks). In seven cats a Coombs’ test was not performed. Intravascular haemolysis was documented in one cat (case 18). Two cats (cases 2 and 17, CMhm) were diagnosed with PRCA. One cat (case 3, CMhm) was diagnosed with pancytopenia. Six cats had leukopenia (800–5000/µl [median 2600/µl]; RI 6000–11,000/µl) and seven cats leukocytosis (11,820–23,300/µl [median 16,900]). The automated platelet count ranged from 8 to 535 × 103/μl (median 131 × 103/μl; RI 180–550 × 103/μl). Two cats with and three cats without concurrent disease had a severe thrombocytopenia (<50,000/µl); these low platelet counts were confirmed by manual counting in a counting chamber. The median plasma bilirubin concentration in 22 cats was 0.32 mg/dl (range 0.07–3.20; RI 0–0.30 mg/dl); 11 cats had hyperbilirubinaemia (four cats with and seven without concurrent disease).Thirteen cats had mildly increased liver enzyme activities (alanine aminotransferase [ALT] 77–426 U/l [median 157 U/l; RI 0–70 U/l]; aspartate aminotransferase [AST] 35–529 U/l [median 82 U/l; RI 0–30]). One cat with a trauma-induced diaphragmatic rupture showed severe increases in ALT (1019 U/l), alkaline phosphatase (88 U/l [RI 0–76 U/l]) and AST (549 U/l). Thirteen of 21 cats were hyperglobulinaemic (4.1–13.5 g/dl [median 4.9 g/dl]; RI 3.1–4.0); six of them had no concurrent disease. The cat with severe hyperglobulinaemia (13.5 g/dl) suffered from multiple myeloma. All cats tested negative for FeLV antigen. PCR testing for FeLV provirus in the three cats with pancytopenia and PRCA was negative. Three of the 22 cats (CMhm [n = 2], CMt [n = 1]) tested positive for FIV antibodies.
Imaging
Radiological examination demonstrated cardiomegaly (n = 3), splenomegaly (n = 3), femoral fracture (n = 1) and moderate constipation (n = 1). In cases with splenomegaly, the spleen was sonographically homogeneous, with the exception of one cat with multiple myeloma, which had multiple hypoechoic nodules in the spleen (case 19).
Therapy
Nineteen cats were treated as inpatients over a period of 2–22 days (median 6 days), and three cats as outpatients. Six cats received doxycycline for 7–31 days (median 21 days) and 15 cats enrofloxacin for 3–28 days (median 21). One cat was initially treated with amoxicillin-clavulanic acid for 11 days for pancreatitis/cholangitis. Haemoplasmosis was diagnosed 11 days after initial presentation and the cat was euthanased without treatment for haemoplasmosis. Fifteen cats received blood products: whole blood (n = 5), Oxyglobin (n = 5; Biopure), whole blood and Oxyglobin (n = 4), and two packed red blood cell transfusions (n = 1). The cats’ Hct ranged from 5–25% (median 12%) at the time of transfusion (the cat with the Hct of 25% was transfused in order to treat haemorrhage due to disseminated intravascular coagulation [case 16]. Cats without a concurrent disease (7/9 cats) received more blood products than cats with concurrent diseases (6/13 cats). Prednisolone was given for suspected immune-mediated haemolytic anaemia (IMHA; n = 8), PRCA (n = 2) or IMT (n = 1) at an initial dose of 0.9–2.1 mg/kg (median 1) PO q12h and was then tapered off over 8–183 days (median 47 days). Additional immunosuppressants were given to one cat with IMT (case 5) and one with PRCA (case 17).
Course of disease
The course of disease of all cats is summarised in Table 2.
PCR results, treatment and outcome in 22 cats with natural haemoplasma infection
Enro = enrofloxacin; Doxy = doxycycline; Marbo = marbofloxacin; Prado = pradofloxacin; Amoxi = amoxicillin–clavulanic acid; IVFT = intravenous fluid therapy; BT = whole blood transfusion; PRBC = packed red blood cell transfusion; IMT = immune-mediated thrombocytopenia; FIV = feline immunodeficiency virus
Owing to concurrent diseases (n = 8) or financial constraints (n = 1), nine cats were euthanased between 3 days and 180 weeks after diagnosis. Only one cat was euthanased owing to persistent severe anaemia (case 17, CMhm). Twelve cats were discharged. Six of these cats were lost to follow-up. At the time of discharge, three of these six cats had a Hct within the reference interval, while the other cats had a Hct that was still subnormal but increased compared with presentation.
Ten cats were followed up for a period of 14–199 weeks. Six of these cats were diagnosed with concurrent diseases and were infected with CMhm (n = 4), Mhf (n = 1) and all three species (n = 1). CMhm (n = 3) and Mhf (n = 1) were detected in the remaining four cats. PCR analysis was repeated in 7/10 follow-up cats. In five cases (cases 1–4 and 10) the first negative PCR occurred after 3 (CMhm, during doxycycline treatment), 7 (CMhm, 3 weeks after enrofloxacin treatment), 7 (infected with all species, after 4 weeks of treatment with enrofloxacin and 3 weeks of treatment with doxycycline), 10 (CMhm, during enrofloxacin treatment) and 23 (Mhf, 19 weeks after doxycycline therapy) weeks.
In 2/7 cats with repeated PCR analysis (cases 5 and 6, both CMhm) remained PCR positive. Case 5 had suffered from IMT for 2 years (initially haemoplasma PCR was negative). Owing to a relapse of the IMT, the haemoplasma PCR was repeated, which was positive for CMhm. During the following 28 weeks the cat was under antibiotic treatment (doxycycline, marbofloxacin, pradofloxacin) but PCR remained positive (4, 13 and 18 weeks after starting treatment). In case 6 the Hct normalised within 7 days of starting enrofloxacin treatment. At week 199 the cat was presented owing to lethargy; PCR was positive but no anaemia was present.
Two cats (cases 1 and 2) that responded to therapy with a negative PCR result presented with three and one reactivations, respectively. Case 1 (CMhm) had eight episodes of constipation. The first negative PCR occurred in week 7 during doxycycline treatment. In 4/8 new episodes of constipation PCR was performed, and in three episodes (in weeks 13, 29 and 177) a concurrent reactivation of haemoplasmosis occurred. The Hct at t he time of positive PCR results ranged between 17–28%. In case 2, with PRCA (CMhm), reactivation occurred in week 7 (positive PCR, Hct 26%).
Owing to the recurrence of haemolytic anaemia, reactivation was suspected in two further cases after 10 days (case 4, Mhf) and 13 weeks (case 7, CMhm). However, PCR testing was not conducted at the times of relapse. After treatment with antibiotics the Hct was in the reference interval in both cases.
Four of the 10 follow-up cats (cases 1, 3, 5, 7) were euthanased after 14, 24 and 180 weeks owing to concurrent disease.
Discussion
The current study describes the natural course of haemoplasma infection in 22 anaemic cats. Haemoplasma infection is rarely a cause of anaemia. In a recent study, 4/194 anaemic cats presented to the Small Animal Clinic of the Freie Universität Berlin between October 2009 and March 2011 were haemoplasma PCR-positive (Merten et al, unpublished data). Ten of the cats in the current study were followed for a period of 14–199 weeks. Immunosuppression, stress and concurrent disease are predisposing factors for the clinical manifestation of feline haemoplasmosis and influence the course of infection. 12 Of the 13 cats with concurrent diseases, six were infected with CMhm, two with CMt, one with Mhf, one cat was infected with all three species and in three cats the species was not determined. Nine cats displayed no concurrent disease. Five of these nine cats were infected with CMhm, two with Mhf and in two cats the species was not determined.
PCR analysis is the diagnostic method of choice for haemoplasma infection. In the current study, the negative PCR results obtained after treatment were unlikely to have indicated elimination of infection as relapses were subsequently seen in some cats; instead, the negative PCR results were likely reflecting haemoplasma organism numbers in the blood of less than the PCR detection threshold. 10 In this study, quantitative PCR (qPCR) was not used as it was not available. Use of qPCR to determine organism numbers in the blood can help monitor the efficacy of antibiotic therapy.17–19 Fluctuations in the number of haemoplasmas in the blood can lead to alternate positive and negative PCR results; such cycling could be due to organ sequestration of organisms, but no conclusive evidence for this has yet been found. 20 Serological testing may be more useful to detect chronic infection, 21 although further studies on the use of serology for routine diagnostics are required, and commercial testing is not currently available.
Haemoplasma species vary in pathogenicity. 12 Experimental studies have confirmed that Mhf infections are more pathogenic than CMhm infections,22,23 whereas CMhm infections have not clearly been associated with disease in immunocompetent cats.13,21 In a study of 21 CMhm-positive cats, in six of the cases, CMhm infection was the only recognisable cause of the anaemia found, 14 suggesting that CMhm can act as a primary pathogen. The Hct of the cats in that study ranged from 8–28%. This is in agreement with the results of the current study: 5/9 cats without concurrent disease were infected with CMhm.
Some of the cats in the present study were prescribed immunosuppressive medications. In experimental studies methylprednisolone has been used to induce haemoplasma reactivation,10,24 but the doses used were higher (10–20 mg/kg IM) compared with the dosages given in the current study. Three of the four cats with haemoplasma reactivation, or suspicion for reactivation, received no immunosuppressive agents. The influence of other factors for induction of reactivation (eg, concurrent diseases and the haemoplasma species) remains unclear.
Infection with haemoplasma species can manifest as haemolytic anaemia.13,14,20 The pathophysiology is multifactorial, 19 but most commonly comprises extravascular haemolysis. 25 Intravascular haemolysis is occasionally described with haemoplasma infections, 7 but was uncommon in the current study. Seven of 15 cats had a positive direct Coombs’ test. This demonstrates the importance of testing for haemoplasma species in cats suspected of having IMHA. 26 Ten of 20 cats in the current study initially presented with non-regenerative anaemia.
Anaemia due to inflammatory disease, bone marrow disease or an only mildly reduced Hct can all be responsible for a failure to regenerate. 27 In the current study, PRCA was diagnosed in two cats. There is no current literature discussing the association between PRCA and feline haemoplasmosis. One of these two cats had a positive direct Coombs’ test. One explanation might be that the haemoplasma species induced a secondary immune reaction against erythrocytic precursors.
Two cats in the current study presented with one and three haemoplasma reactivations, respectively, and reactivations were suspected in two further cats. However, a new infection cannot be excluded in these cats. Exact mechanisms of reactivation are unclear. Haemoplasma organisms have been detected in various tissues such as liver, spleen, salivary glands and lungs,9,10,20 and it is possible that very small numbers can remain in such organs and can then seed organisms into the blood at a later time point. One study successfully used electron microscopy to document macrophage phagocytosis of haemoplasma species in the spleen. 25 Another study found that the administration of glucocorticoids produced positive PCR results for CMt in various tissues (liver and salivary gland tissues collected by fine-needle aspiration) in 2/5 cats. 10 Similar results have been documented with experimental CMhm infections. 24 The cats in the current study presented with reactivations 7–177 weeks after initial presentation. One cat (case 1), which suffered from recurrent bouts of constipation, had three reactivations of CMhm infection at times of constipation. We are not aware of a published description of such frequent recurrences of reactivation.
Multiple studies have demonstrated the success of various antibiotic therapies against haemoplasma species.24,28–30 However, these studies all describe experimental infections of healthy cats. Doxycycline and pradofloxacin have both been shown to reduce clinical signs and Mhf blood loads in experimentally infected cats. 30 However, confirmation of clearance is hard to prove and, indeed, some of the cats in the latter study that tested PCR-negative consequently tested PCR-positive following administration of methylprednisolone. 30 In one study with naturally infected cats (CMhm) treated with doxycycline or enrofloxacin, 43% (n = 6) responded to treatment. 14 The haemoplasma infection was probably not eliminated in most of the cases in the current study but was probably below the PCR detection threshold,31,32 thus yielding a negative PCR result. There have been individual case reports of complete elimination of the infectious agents (CMhm, CMt), 12 although proving the complete elimination of infection via blood PCR tests only remains difficult. The duration of therapy for cats in this study ranged from 7 days to >3 months, and depended on the attending clinician. However, PCR did not become consistently negative. A study showed that a 4 week treatment with marbofloxacin was effective in cats with CMhm infection. 29 However, the fall in CMhm copy number was not as dramatic as that observed in response to doxycycline or enrofloxacin. 29 Owing to a lack of standardised treatments, the effectiveness of antibiotic treatment in the cats of the present study was difficult to interpret.
Ten of the 22 cats in the current study were euthanased: nine cats owing to severity of concurrent disease or financial constraints. Only one cat (case 17) was euthanased owing to persistent anaemia after 11 weeks; this cat had PRCA (CMhm). Feline haemplasmosis is rarely lethal. In a study of 21 CMhm-infected cats only one cat was euthanased because of anaemia. 14
One limitation of the current study is its retrospective nature. Data were not available for all the measured parameters in each case. Another limitation is the lack of qPCR utilisation. The performance of a qPCR may have changed the treatment protocol. Furthermore, case management (therapy, PCR testing) differed among the clinicians, and there was no standardised treatment protocol. Some cats were lost to follow-up because they were treated by the referring veterinarian. The low number of cases is a further limitation, which could only be improved on by multicentre studies.
Conclusions
As demonstrated in this study, a change in immune status and stress might lead to reactivation of haemoplasma infection after long asymptomatic intervals. Clinicians ought to remain aware of this, as early detection and therapeutic intervention might reduce the risk of development of severe anaemia and the need for blood transfusions.
Footnotes
Parts of this work were presented as a poster at the 24th European College of Veterinary Internal Medicine Congress, Mainz, Germany.
Conflict of interest
The authors do not have any potential conflicts of interest to declare.
Funding
The authors received no specific grant from any funding agency in the public, commercial or not-for-profit sectors for the preparation of this article.