“Candidatus Mycoplasma haemominutum” Infections in 21 Client-Owned Cats

Introduction

Mycoplasma haemofelis, “Candidatus Mycoplasma haemominutum,” and “Candidatus Mycoplasma turicensis” are gram-negative, epicellular parasites of feline erythrocytes. These organisms, formerly known as Haemobartonella felis, can cause clinical illness—a syndrome known as hemoplasmosis—in some infected cats.^1–4 In experimental studies, intravenous (IV) inoculation of blood with any of the three species can induce anemia and clinical disease in some cats; however, most cats experimentally inoculated with “Candidatus Mycoplasma haemominutum” have been minimally affected.^3,5–7 These reports suggest that “Candidatus Mycoplasma haemominutum” is only mildly pathogenic for immunocompetent cats. Only small numbers of strains have been inoculated into healthy cats, so it is possible that other pathogenic isolates of “Candidatus Mycoplasma haemominutum” exist.

Polymerase chain reaction (PCR) assays capable of differentiating the deoxyribonucleic acid (DNA) of Mycoplasma haemofelis and “Candidatus Mycoplasma haemominutum” are now available.^8–10 Based on the results of those assays, it has been ascertained that “Candidatus Mycoplasma haemominutum” infection is more common in cats than Mycoplasma haemofelis infection, but it is not known how many infected cats actually become ill. In previous studies, “Candidatus Mycoplasma haemominutum” prevalence rates were compared between healthy and clinically ill cats.^8–10 In a prevalence study from the United States, “Candidatus Mycoplasma haemominutum” DNA was detected in blood from 12.7% of tested cats, and there was no difference in prevalence between healthy (13.8%) and ill (11.0%) cats.⁸ In contrast, the prevalence rates of “Candidatus Mycoplasma haemominutum” infection in healthy and ill cats in the United Kingdom were 8.3% and 20.3%, respectively.⁹ In another study of cats inAustralia, those with “Candidatus Mycoplasma haemominutum” DNA in their blood had lower average hematocrits than negative cats.¹⁰ The United Kingdom and Australia studies suggested that some cats develop clinical illness following natural infection by “Candidatus Mycoplasma haemominutum.” It is possible that different strains vary in pathogenic potential or that other concurrent disease can activate hemoplasmosis. The purpose of this study was to evaluate the clinical findings of a group of cats residing in the United States that were naturally infected with “Candidatus Mycoplasma haemominutum.”

Materials and Methods

The laboratory records section of the Infectious Diseases Diagnostic Laboratory, Colorado State University, was searched for all client-owned cats that had blood samples assessed for the presence of hemoplasma species infection between January 2001 and June 2004. All samples were tested with a PCR assay capable of amplifying and differentiating DNA of Mycoplasma haemofelis and “Candidatus Mycoplasma haemominutum.”⁸ All samples were also tested for the DNA of Bartonella spp., Ehrlichia spp., and Anaplasma phagocytophilum by use of previously reported PCR assays, as well as for feline leukemia virus (FeLV) antigen and feline immunodeficiency virus (FIV) antibodies. ^a,11,12

The laboratory submission forms were retrieved for all cats that were PCR-positive for “Candidatus Mycoplasma haemominutum.” Healthy cats (e.g., blood donors) were excluded from the study. Cats that tested positive for both “Candidatus Mycoplasma haemominutum” and Mycoplasma haemofelis were also excluded, because it would not have been possible to determine which organism was responsible for any reported clinical signs.

Complete medical records were obtained from the referring veterinarians for all clinically ill cats that were positive for “Candidatus Mycoplasma haemominutum.” Data retrieved from the records included signalment, state of origin, housing environment, flea exposure, clinical signs, laboratory test results, concurrent diseases, treatments, and outcome. When no outcome was documented in the medical record, telephone contact with the owners was attempted.

Because anemia is the most consistent clinical syndrome associated with feline hemoplasmosis, cats were arbitrarily classified as those with and without anemia.^1,2 Cats with packed cell volumes (PCVs) ≤29% were considered anemic, and the anemia was considered regenerative if the mean cell volume was >50 femtoliters, the reticulocyte count was >50,000 per μL, or the reticulocyte percent or corrected reticulocyte percent was >1.0%. Cats with either regenerative or nonregenerative anemia were included in the study, because many of the cats were presented acutely ill and did not have time to develop a measurable regenerative response at the blood sampling point. Cats with no other diseases that could have activated the “Candidatus Mycoplasma haemominutum” infection or caused the anemia were considered to have a primary “Candidatus Mycoplasma haemominutum” infection. Determination was made as to whether treatment with antibiotics having known efficacy against hemoplasmas (i.e., tetracycline derivatives or fluoroquinolones) was administered (with or without glucocorticoids), and the responses to treatment were noted.^1,2,5,13

Results

Of 332 cats tested, 37 (11.1%) cats were positive for only the DNA of “Candidatus Mycoplasma haemominutum,” 20 (6.2%) cats were positive for only the DNA of Mycoplasma haemofelis, and five (1.5%) cats were positive for both hemoplasmas. Of the 37 cats positive for only “Candidatus Mycoplasma haemominutum” DNA, complete medical records were available for 21 clinically ill cats.

Gender and breed of the 21 cats are summarized in Table 1. Ages ranged from 1 to 16 years (median 10.0 years, mean 9.8±4.0 years). States of origin included Colorado (n=7), Connecticut (n=3), Massachusetts (n=2), Oklahoma (n=2), and one cat each from Arizona, Michigan, Maine, Virginia, California, New Jersey, and South Carolina. Housing history was known for 17 cats; one cat was exclusively housed indoors, and 16 cats had access to the outdoors. Of the 21 cats, a history of flea infestation was documented in two cats, and topical insecticides were dispensed to three other cats.

Thirteen cats had a PCV ≤29% (range 8% to 28%; reference range 30% to 46%), and eight had a PCV >29% (range 35% to 43%; reference range 30% to 46%) on initial evaluation. Seven (53.4%) anemic cats had clinical or laboratory evidence of other diseases that could have activated hemoplasmosis or been the primary cause of anemia. Six (46.6%) anemic cats had no other apparent cause [Table 1]. There was no statistical difference (P=0.067) between the initial, group mean PCV of cats with “Candidatus Mycoplasma haemominutum” infection alone (mean PCV 14.52%±7.44%) when compared to that of cats with concurrent diseases (mean PCV 21.43%±1.99%). Diseases that could have activated hemoplasmosis or been the primary causes of anemia in the seven cats included FeLV infection (n=2), feline infectious peritonitis (n=1), suspected immune-mediated myeloid and erythroid maturation arrest (n=1), lymphosarcoma (n=1), erythroid leukemia (n=1), and concurrent polyarthritis, thrombocytopenia, and Bartonella henselae infection (n=1).

Of the 13 anemic cats, six had evidence supporting a regenerative anemia. Mean corpuscular volume (MCV) was available for one cat (68 fL). Reticulocyte percents were reported for three cats as >4% (n=2) and 6% (n=1). One cat was classified as regenerative based on a corrected reticulocyte percent of 2.2%. One cat had a reticulocyte count of 178,130/μL and a nucleated red blood cell count of 177,500/μL; this cat was diagnosed with erythroid leukemia. For three cats, no MCVs or reticulocyte counts were available. Four cats were classified as nonregenerative. One cat had a reticulocyte count of 1%. Three cats had reticulocyte counts <50,000/μL (range 0 to 6860/μL; mean 3583.3±3440.3/μL); however, one cat had erythroid hyper-plasia on bone marrow cytology.

In the 13 cats with anemia, additional clinical abnormalities or diagnoses included lethargy (n=8), inappetence (n=7), vomiting (n=3), elevated liver enzyme activities (n=2; alkaline phosphatase range 22 to 77 IU/L [reference range 6 to 65 IU/L], alanine aminotransferase [ALT] range 187 to 422 IU/L [reference range 30 to 100 IU/L], aspartate aminotransferase [AST] range 82 to 378 IU/L [reference range 15 to 50 IU/L], gamma glutamyl transferase [GGT] range 3 to 5 IU/L [reference range 0 to 2 IU/L]), uveitis (n=2), ascites (n=2), fever (n=2), weight loss (n=2), oral ulcers (n=1), hyperglobulinemia (n=1), hepatomegaly (n=1), splenomegaly (n=1), polyuria (n=1), neutrophilia (n=1; segmented neutrophils 16,300/μL, reference range 2000 to 12,000/μL), and conjunctivitis (n=1). In the eight cats without anemia, additional clinical abnormalities or diagnoses included fever (n=4), lethargy (n=5), polyarthritis (n=1), uveitis and FIV infection (n=1), anorexia (n=1), and stomatitis (n=1).

Information concerning treatment and outcome was known for 19 of the 21 cats [Table 2]. Treatment protocols and recheck schedules varied greatly between cases, which precluded statistical comparison. Fourteen cats were administered enrofloxacin^c or doxycycline with or without prednisone; one cat was administered azithromycin^d and enrofloxacin; one cat was administered amoxicillin alone; and four cats were not treated. Of the four untreated cats, two (case nos. 12, 13) were euthanized, one (case no. 11) had clinical signs resolve, and one (case no. 21) developed no further clinical signs. Overall, 17 cats had anemia or fever, and 14 of these were treated with enrofloxacin or doxycycline (alone or in combination with prednisone). Of the 14 cats, six responded to therapy. Three cats with anemia and concurrent disease (case nos. 8, 9, and 10) and two cats with anemia (case nos. 1, 3) were euthanized after failing to respond to treatment. The outcomes of two cats (case nos. 14, 15) are unknown. None of the 12 surviving cats were assayed for hemoplasma DNA in blood on recheck examinations.

Discussion

The prevalence of “Candidatus Mycoplasma haemominutum” in cats of this study (11.1%) was similar to previous reports in the United States (12.7%);⁸ United Kingdom (16.9%);⁹ Saskatchewan and Alberta, Canada (11.7%);¹⁴ Australia (23.1%);¹⁰ and South Africa (32.1%).¹⁵ Infected cats in this study were from many different states; both sexes were represented; and the age range varied greatly. Similar findings were in previous reports.^{1,8–10,15–17} Many of the previous transmission and epidemiological studies were performed before it was determined that Mycoplasma haemofelis, “Candidatus Mycoplasma haemominutum,” and “Candidatus Mycoplasma turicensis” were distinct organisms; therefore, documented information about transmission of “Candidatus Mycoplasma haemominutum” is minimal. “Candidatus Mycoplasma haemominutum” may be transmitted by Ctenocephalides felis; the majority of the cats in this study were allowed outdoors, and they may have been commonly exposed to fleas. Eight of the cats resided in Colorado or Arizona at the time of illness; these states are thought to have a low prevalence of Ctenocephalides felis infestation. In addition, in two previous studies, the organism was not transmitted between cats by flea feeding or by feeding infected fleas to cats.^18,19 Further studies are needed to determine the routes of transmission of “Candidatus Mycoplasma haemominutum.”

While “Candidatus Mycoplasma haemominutum” is thought to be minimally pathogenic based on experimental studies, the organism has been associated with clinical illness in some client-owned cats.^9,10 Five cats in this study were clinically ill, had no other obvious causes of disease, and apparently responded to administration of doxycycline (alone or in combination with prednisone), which supported the hypothesis that some immunocompetent cats develop clinical illness after infection with “Candidatus Mycoplasma haemominutum.” It is also possible, however, that the infection spontaneously resolved in the five cats, or they had an undiagnosed immunosuppressive disease that activated the infection. In addition, because prednisone was administered to nine of the cats in the study, it could not be determined with certainty what therapy induced the clinical responses. Results of the study suggest that cats having DNA that is positive for “Candidatus Mycoplasma haemominutum” in PCR assays in addition to fever or anemia should be treated with an antihemoplasma drug regardless of the presence or absence of an obvious concurrent disease.

Five cats that were administered a presumed antihemoplasma drug (with or without prednisone) were euthanized because of apparent therapeutic failure. It is possible that “Candidatus Mycoplasma haemominutum” infection was the cause of the clinical illness and that the cats succumbed to other causes or that the coexisting illness led to immune suppression and overwhelming hemoplasmosis. Lastly, it is also possible that these cats were treated with an antihemoplasma drug for which the infecting strain of the organism was resistant. In vitro antimicrobial susceptibility testing cannot currently be performed, because “Candidatus Mycoplasma haemominutum” has not yet been grown in culture. In addition, most studies of hemoplasmosis treatment have used cats infected with Mycoplasma haemofelis, not “Candidatus Mycoplasma haemominutum.”^13,21 In vivo studies suggest that antimicrobial resistance occurs with hemoplasma species similar to other bacteria. For example, treatment of cats experimentally infected with “Candidatus Mycoplasma haemominutum” with azithromycin or marbofloxacin failed to eliminate infection.^6,20 In addition, some cats infected with a hemoplasma species that failed to respond to tetracycline derivatives or fluoroquinolones responded to administration of imidocarb diproprionate.²¹

The biggest limitation to this retrospective study was the lack of standardization regarding diagnoses, treatments, and timing of recheck evaluations between clinics, so it is possible that some cases were categorized incorrectly with respect to responses and outcomes. It is also unknown whether the owners administered medications as prescribed. It is also possible that other hemoplasma species may have infected these cats but were not amplified by the PCR assay utilized, which would have influenced the results of this study.

Conclusion

In this study, “Candidatus Mycoplasma haemominutum” DNA was detected in the blood of 37 of the 332 clinically ill cats tested. Of the 21 cats with medical records available for review, six anemic cats had no other obvious cause of anemia, and some cats responded to drugs with antihemoplasma activity. Results of this study suggest that “Candidatus Mycoplasma haemominutum” may be a primary pathogen in some naturally infected cats. Further prospective studies utilizing standardized diagnostic and therapeutic protocols are needed to validate these findings.