Jane E. Sykes and Séverine Tasker

Etiology and Epidemiology

Hemotropic mycoplasmas (hemoplasmas) are small (0.3-0.8 µm), unculturable mycoplasmas that reside on the surface of erythrocytes and can cause variable degrees of hemolytic anemia in infected hosts. Hemoplasmas infect a wide variety of mammalian species, including humans, and have a worldwide distribution. Although previously classified as Haemobartonella and Eperythrozoon spp., sequence analysis of the 16S rRNA genes of these organisms has shown that they are closely related to the pneumoniae group of mycoplasmas, which includes the human mycoplasmal pathogens Mycoplasma pneumoniae and Mycoplasma genitalium.

At least three hemoplasma species infect domestic as well as wild cats, Mycoplasma haemofelis, ‘Candidatus Mycoplasma haemominutum,’ and ‘Candidatus Mycoplasma turicensis.’ The ‘Candidatus’ prefix is applied to newly discovered hemoplasmas until more information is available to support their classification. This is because hemoplasmas cannot be cultured in the laboratory, which limits full characterization of these organisms. M. haemofelis (previously the Ohio strain, or large form of Haemobartonella felis) is the most pathogenic organism and can cause moderate to severe hemolytic anemia in immunocompetent cats. The resulting disease has been referred to as feline infectious anemia. Using cytologic evaluation of blood smears, M. haemofelis organisms are cocci that sometimes form short chains of three to six organisms (Figure 41-1). M. haemofelis is the least prevalent of the three feline hemoplasmas. It has been found using PCR in 0.5% to 5% of sick cats at veterinary hospitals. The whole genome sequences of M. haemofelis and M. haemominutum have been determined.^3–5

‘Ca. M. haemominutum’ (previously the California strain or small form of H. felis) is generally smaller than M. haemofelis and has not clearly been associated with disease in immunocompetent cats. Using cytologic evaluation of blood smears, ‘Ca. M. haemominutum’ are small cocci, 0.3 to 0.6 µm in diameter, although M. haemofelis and ‘Ca. M. haemominutum’ cannot always reliably be distinguished by this method. ‘Ca. M. haemominutum’ is common in the cat population worldwide, infecting as many as 20% of cats that visit veterinary hospitals.6–8 Infection of cats with ‘Ca. M. haemominutum’ results in a mild decrease in hematocrit. There is some evidence that ‘Ca. M. haemominutum’ may play a role in disease. For example, cats that are co-infected with FeLV and ‘Ca. M. haemominutum’ develop more significant anemia than cats infected with ‘Ca. M. haemominutum’ alone, and progression to FeLV-induced myeloproliferative disease occurred more rapidly.⁹ There are also descriptions of cats with hemolytic anemia in which the only recognized cause of anemia was ‘Ca. M. haemominutum.’¹⁰ ‘Ca. M. haemominutum’ is commonly found in co-infections with ‘Ca. M. turicensis’ or M. haemofelis. Mixed infections with all three feline hemoplasma species also have been described.

‘Ca. M. turicensis’ was first described in a cat from Switzerland that had severe intravascular hemolysis (turicensis pertains to Turicum, the Latin name of Zurich).¹¹ Infections with ‘Ca. M. turicensis’ have since been detected worldwide.^8,12–14 ‘Ca. M. turicensis’ has never been seen using light microscopic examination of blood smears, and organism loads in infected cats are typically low. Infection with ‘Ca. M. turicensis’ is slightly more prevalent in the cat population than infection with M. haemofelis. Most studies show a prevalence of 0.5% to 10% in sick cats that visit veterinary hospitals. The pathogenic potential of this organism is not fully understood. Inoculation of an immunosuppressed cat with ‘Ca. M. turicensis’ resulted in severe anemia,¹¹ but little or no anemia occurs in immunocompetent cats after inoculation with ‘Ca. M. turicensis.’ Cofactors, such as co-infection with other hemoplasmas or concurrent immunosuppression, may influence the development of anemia in cats infected with ‘Ca. M. turicensis.’

Infection of cats by hemoplasmas is strongly associated with male sex, nonpedigree status, and outdoor access (Box 41-1).^8,12,15,16 Infection with ‘Ca. M. haemominutum’ is more prevalent in older cats, presumably because the chance of acquiring persistent subclinical infection increases over time. In contrast, young cats may be more likely to develop disease after infection by M. haemofelis. Some studies, but not others, have shown an association between retrovirus and hemoplasma infections. Cats infected with M. haemofelis in the United States were 6 times more likely to be FIV infected than cats negative for hemoplasmas.¹⁵

Several hemoplasma species also infect dogs. Infection with Mycoplasma haemocanis (previously Haemobartonella canis) has been associated with hemolytic anemia in splenectomized dogs, and rarely in dogs with other immunosuppressive disease or concurrent infections. The 16S rRNA gene of M. haemocanis has the same sequence as M. haemofelis, but the whole genome sequence of M. haemocanis distinguishes it as a different species.¹⁷ M. haemocanis is a coccoid organism that often forms long chains of organisms (Figure 41-2). The prevalence of this infection is particularly high in kennel-raised dogs, which are often infected subclinically.¹⁸ In the southwestern United States, infection was also prevalent among coyotes.¹⁹

Three additional hemoplasma species have been detected in dogs. ‘Candidatus Mycoplasma haematoparvum’ is a small (0.3 µm) coccoid organism that resembles ‘Ca. Mycoplasma haemominutum’ morphologically as well as genetically (Figure 41-3). ‘Ca. M. haemominutum’ has also been detected in several dogs using PCR assays, and organisms that resemble ‘Ca. M. haematoparvum’ and ‘Ca. M. haemominutum’ have been detected in European wolves and bush dogs from Brazil.²⁰ The ovine hemoplasma Mycoplasma ovis was detected in the spleens of a small number of dogs from the southeastern United States, and the bovine hemoplasma ‘Ca. Mycoplasma haemobos’ was detected in a dog from northern Australia.^21,22 The clinical importance of these hemoplasma species in dogs remains unclear.

The mode of transmission of hemoplasmas remains unclear. To some extent, fleas and other arthropod vectors may be capable of transmitting feline hemoplasmas,²³ but experimental evidence for vector-borne transmission of feline hemoplasmas is weak. Transmission of M. haemocanis by the brown dog tick, Rhipicephalus sanguineus, has been demonstrated experimentally, although this was before PCR assays were available to confirm infection.²⁴ Geographic variation in the prevalence of hemoplasma infection in dogs and cats supports a role for arthropod vectors in transmission. For example, in Europe, infection with M. haemocanis is more prevalent in Mediterranean countries, which follows the distribution of Rh. sanguineus. Vertical (e.g., transplacental) spread has also been hypothesized and has been documented for bovine hemoplasmas. Biting has been suggested as a means of transmission of feline hemoplasmas, and the strong male sex predilection, recent history of cat bite abscessation in some cats, and association with FIV infection in some studies supports this mode. Additionally, studies in Switzerland found that subcutaneous inoculation of blood that contained ‘Ca. M. turicensis’ resulted in transmission, whereas inoculation of saliva that contained ‘Ca. M. turicensis’ did not.²⁵ This suggests that hemoplasma transmission by social contact (saliva via mutual grooming etc.) is less likely than transmission by aggressive interaction (blood transmission during a cat bite incident). Because infection can also be transmitted by ingestion of blood, it may be that the biting cat (rather than the bitten cat) is most at risk for acquisition of infection. Transmission of M. haemocanis through ingestion of infected blood has also been described,²⁶ so aggressive interactions between dogs may also have the potential to transmit hemoplasmas; however, this mode of transmission remains to be proven in field circumstances. Transmission can also occur following blood transfusion.

Clinical Features