Materials and Methods

A total of 310 blood samples that had been submitted to the IDEXX laboratory in Sacramento, Calif, between January 2006 and June 2006 were used in the study. All blood samples had been anticoagulated with EDTA and had been obtained from cats suspected to have hemoplasmosis because hemoplasma-like organisms had been observed during cytologic examination of a blood smear, the cat had acute or regenerative anemia (Hct < 29%) without any known history of blood loss, or both. For most cats, historical information was not available. However, laboratory records were searched to identify previous submissions of blood samples from the same cats, and cats were excluded if there was any previous laboratory history of anemia. For all blood samples, smears were made within 6 hours of sample collection and stained with Wright-Giemsa stain. Samples were stored at 4°C until processed; all samples were processed within 2 weeks of sample collection.

For each cat, information regarding signalment, zip code of the submitting veterinary practice, whether hemoplasma organisms were seen during cytologic evaluation of the blood smear, and results of a CBC and ELISAs for FIV and FeLV infection was recorded. If a serum biochemistry panel was performed at the time of sample submission, serum total protein, albumin, urea nitrogen, creatinine, and total bilirubin concentrations and serum alkaline phosphatase, alanine aminotransferase, and aspartate aminotransferase activities were also recorded. Complete blood counts^a and serum biochemistry panels^b were performed with automated equipment. The ELISAs for FIV and FeLV infection consisted of 96well assays for FeLV antigen and FIV antibody that were performed in accordance with the manufacturer's directions.^c Reported sensitivity and specificity of the FeLV assay were 97.6% and 99.1%, respectively, and reported sensitivity and specificity of the FIV assay were 100% and 99.5%, respectively.¹⁰ Confirmatory testing with an immunofluorescent antibody test for FeLV infection or a Western blot test for FIV infection was not routinely performed. For cats with positive FIV assay results, the submitting veterinary practice was telephoned to determine whether the cat had a history of having been vaccinated against FIV.

Blood samples were tested by means of a broadspectrum conventional PCR assay and 3 separate species-specific real-time PCR assays for hemoplasma DNA. The conventional PCR assay was used in addition to the species-specific real-time PCR assays to allow detection of hemoplasma strains and species that might not be detected with the real-time PCR assays. For these assays, DNA was extracted from a 160-μL aliquot of each sample with a robotic extractor,^d with a final elution volume of 100 μL.

The conventional PCR assay was performed as described.² Positive control samples consisted of blood samples from cats that had been experimentally infected with Mhf and Mhm. Negative control samples included samples in which the DNA in the reaction mixture was replaced with purified DNaseand RNase-free water and blood samples from a healthy adult dog that had previously tested negative for hemoplasmas by means of the conventional PCR assay. Samples were analyzed on a 2.5% agarose gel. Samples yielding products of approximately 170 base pairs were considered positive for Mhf or Mtc; samples yielding products of approximately 193 base pairs were considered positive for Mhm or “Candidatus Mycoplasma haematoparvum.”

Real-time PCR assays for detection of the 16S rRNA gene of Mhm, Mhf,⁸ and Mtc⁵ were performed as described. A real-time PCR assay for feline glyceraldehyde-3-phosphate dehydrogenase was used as a control to indicate the successful isolation of DNA and the absence of PCR inhibitors in samples for which real-time assay results were negative, as described.⁸ Validation of these assays has been reported previously.^5,8 Amplification was performed with the default cycling parameters of the manufacturer's real-time PCR assay platform^d and consisted of 2 minutes at 50°C, 10 minutes at 95°C, and 40 cycles of 15 seconds at 95°C and 60 seconds at 60°C. Data were collected and processed with sequence detection system software.^e

For each blood sample, results of the real-time PCR assays were compared with results of the conventional PCR assays. When results did not agree, assays were repeated. Amplification products from samples with repeatedly discordant results were sequenced by use of dye-terminator chemistry procedures and an automated DNA sequencer.^f

Statistical analysis—The D'Agostino and Pearson omnibus normality test was used to test for normality amongst data. For data following a Gaussian distribution, the unpaired t test was used to compare continuous variables between cats testing positive for each hemoplasma species and cats testing negative. The Mann-Whitney U test was used to compare continuous variables between these groups when data were not normally distributed. The 2-sided Fisher exact test was used to test for associations between categoric variables and positive test results for each hemoplasma species. Where appropriate, results are expressed as mean ± SD. All analyses were performed with standard statistical software^g; values of P < 0.05 were considered significant.

Results

Cats—Information on age was available for 289 of the 310 cats; mean ± SD age was 11.3 ± 5 years. There were 136 females and 171 males (sex of 3 cats was not reported). Thirty-two cats were purebred, and 269 cats were of mixed breeding (breed of 9 cats was not reported). All samples originated from northern California (n = 302) or Hawaii (8). Hemoplasmas were identified during cytologic examination of blood smears from 9 cats; 309 cats had acute or regenerative anemia. Eighteen of 298 (6%) cats were seropositive for FIV antibodies, and 34 of 298 (11%) cats were seropositive for FeLV antigen. All cats that were seropositive for FIV antibodies were > 5 years old, and none had any history of being vaccinated against FIV.

PCR assay results—Results of the real-time PCR assay for Mhm DNA were positive for 67 of the 310 (22%) cats, results of the real-time PCR assay for Mhf DNA were positive for 14 (4.5%) cats, and results of the real-time PCR assay for Mtc DNA were positive for 20 (6.5%) cats. All samples for which results of 1 or more of the real-time PCR assays were positive also had positive results for the conventional PCR assay. For 6 samples, although results of the conventional PCR assay were repeatedly positive, results of the real-time PCR assays were repeatedly negative. Sequencing of DNA amplified from these samples revealed that 5 cats were infected with a variant strain of Mhm not expected to be detected with the real-time PCR assay (GenBank accession No. HFU88564)⁸ and 1 cat was infected with Mhf. However, repeated attempts to amplify and sequence the full-length 16S rRNA gene of the Mhf isolate to determine whether sequence variation might explain the false-negative real-time PCR assay results were unrewarding. Molecular evidence of infection with “Candidatus Mycoplasma haematoparvum” was not detected in any sample. Overall prevalences of Mhm, Mhf, and Mtc infection were 23.2% (72/310), 4.8% (15/310), and 6.5% (20/310), respectively. Mixed infections were detected in 20 of the 310 (6.5%) cats, including 5 cats infected with Mhm and Mhf, 10 cats infected with Mhm and Mtc, 3 cats infected with Mhf and Mtc, and 2 cats infected with all 3 species. Thus, a total of 85 of the 310 (27%) cats were positive for hemoplasmas and 225 (73%) were negative.

Of the 9 cats for which organisms were seen during cytologic evaluation of a stained blood smear, 6 were infected with Mhf and 2 were infected with the variant Mhm strain. For the remaining cat, results of the conventional and real-time PCR assays were repeatedly negative.

Because of the low numbers of cats infected with Mhf or Mtc and the high rate of coinfection in these cats, coinfected cats were included when analyzing variables associated with Mhf or Mtc infection. In contrast, only cats infected with Mhm alone were included when analyzing variables associated with Mhm infection.

For all 3 hemoplasma species, infected cats were significantly more likely to be male than were uninfected cats (Table 1). Cats infected with Mhm were significantly less likely to be purebred than were uninfected cats. There was a significant association between FIV seropositivity and positive test results for Mhf but not between FIV seropositivity and positive test results for Mtc alone or Mhm alone. There was a significant association between FeLV seropositivity and positive test results for Mhf.

Table 1—

Potential risk factors for infection with various hemoplasma species among cats with possible hemoplasmosis.

View Table

Of the cats positive for Mhf, 13 had been tested for evidence of retroviral (FeLV or FIV) infection, and 7 of these 13 cats had positive retroviral test results. By comparison, only 32 of the 216 (15%) cats negative for hemoplasmas had positive retroviral test results. These proportions were significantly (P = 0.002; odds ratio, 6.7; 95% confidence interval, 2.1 to 21.2) different. In contrast, there were no significant associations between overall retrovirus seropositivity and positive test results for Mtc or Mhm.

Cats positive for Mhf were significantly younger than cats negative for hemoplasmas (median age, 10 years vs 12 years, respectively; P = 0.02), but the ages of cats positive for Mhm or Mtc were not significantly different from age of cats negative for hemoplasmas. Six of the 15 cats positive for Mhf were ≤ 3 years old, and 3 of the 6 were coinfected with FeLV. The remaining cats were ≥ 9 years old, and 3 of these 9 cats were coinfected with FIV.

There were no significant differences in the values for Hct, MCHC, neutrophil count, band neutrophil count, eosinophil count, basophil count, or platelet count between cats positive for each hemoplasma species and cats negative for hemoplasmas. Cats positive for hemoplasmas had significantly higher MCV values than cats negative for hemoplasmas (Figure 1). Of the 6 cats positive for hemoplasmas with an MCV > 70, 4 were coinfected with FeLV and 2 were infected with the variant Mhm strain. Of the 3 cats negative for hemoplasmas with an MCV > 70, 2 were infected with FeLV. Reticulocyte counts and nucleated RBC counts were significantly higher in cats positive for Mhf than in cats negative for hemoplasmas; a significant difference was still found when cats positive for both Mhf and retroviruses were excluded from this analysis. Lymphocyte counts were also higher in cats positive for Mhf, compared with cats negative for hemoplasmas. Cats positive for Mhf or Mhm had significantly higher values for MCH and MCV than cats negative for hemoplasmas. Cats positive for Mtc had higher WBC and monocyte counts than did cats negative for hemoplasmas. Compared with cats positive for Mtc, cats positive for Mhf had higher reticulocyte counts (P = 0.046).

View Full Size

Box plots of Hct (A), MCV (B), MCH (C), MCHC (D), reticulocyte count (E), lymphocyte count (F), monocyte count (G), and eosinophil count (H) in cats with possible hemoplasmosis grouped on the basis of results of PCR assays for hemoplasma DNA. For each plot, the box represents the interquartile (ie, 25th to 75th percentile) range, the line within the box represents the median, and the whiskers represent the range. The shaded area in each graph represents the reference range. Some cats infected with Mtc or Mhf were coinfected with other hemoplasma species; Mhm represents only those cats infected with Mhm alone. Neg = Cats for which results of the PCR assays for hemoplasma DNA were negative (ie, uninfected cats). The number in parentheses indicates the number of cats in each group. P values represent the results of statistical analysis (Mann Whitney U test) comparing the distribution of results between cats infected with each species and cats negative for hemoplasma DNA; a value of P < 0.05 was considered significant.

Citation: Journal of the American Veterinary Medical Association 232, 3; 10.2460/javma.232.3.372

• Download Figure
• Download figure as PowerPoint slide

Cats positive for Mhf were significantly less likely to have lymphopenia (2/14 vs 146/232, respectively; P < 0.001) and more likely to have reticulocytosis (6/19 vs 50/232, respectively; P < 0.001), a high MCH (7/14 vs 26/232, respectively; P < 0.001), a high MCHC (3/15 vs 6/225, respectively; P = 0.01), and a high MCV (5/14 vs 23/231, respectively; P = 0.01) than were cats negative for hemoplasmas (Figure 1). Cats positive for Mhf were also significantly more likely to have normoblastosis than were cats negative for hemoplasmas (8/14 vs 18/232, respectively; P < 0.001). Cats positive for Mtc were significantly more likely to have eosinophilia (3/20 vs 8/232, respectively; P = 0.047), a high MCH (7/20 vs 26/232, respectively; P = 0.01), and a high MCV (6/19 vs 23/231, respectively; P = 0.01) than were cats negative for hemoplasmas. Cats positive for Mhm alone were significantly more likely to have monocytosis than were cats negative for hemoplasmas (14/50 vs 31/232, respectively; P = 0.02).

View Full Size

Box plots of SUN (A), serum creatinine (B), and total protein (C) concentration in cats with possible hemoplasmosis grouped on the basis of results of PCR assays for hemoplasma DNA. See Figure 1 for key.

Citation: Journal of the American Veterinary Medical Association 232, 3; 10.2460/javma.232.3.372

• Download Figure
• Download figure as PowerPoint slide

There were no significant differences in alkaline phosphatase, alanine aminotransferase, and aspartate aminotransferase activities or total bilirubin concentration between cats positive for each hemoplasma species and cats negative for hemoplasmas. Serum creatinine concentrations were significantly lower in cats positive for Mtc or Mhf than in cats negative for hemoplasmas, and SUN concentrations were significantly lower in cats positive for Mhf than in cats negative for hemoplasmas. The mean serum albumin concentration was significantly higher in cats positive for Mhf (2.8 ± 0.3 mg/dL; reference range 2.3 to 3.3 mg/dL) than in cats negative for hemoplasmas (2.5 ± 0.5 mg/dL; P < 0.02). Total protein concentrations were significantly higher in cats positive for Mtc or Mhf than in cats negative for hemoplasmas (Figure 2).

Cats positive for Mhf were significantly less likely to have high SUN (2/12 vs 127/186, respectively; P < 0.001) and creatinine (1/12 vs 110/187, respectively; P < 0.001) concentrations than were cats negative for hemoplasmas (Figure 2), and cats positive for Mtc were significantly less likely to have high creatinine concentrations (5/18 vs 110/187, respectively; P = 0.01). Cats positive for any hemoplasma species were significantly less likely to have hypoproteinemia than were cats negative for hemoplasmas (6/71 vs 37/186, respectively; P = 0.04).

Discussion

In the present study, hemoplasma DNA was identified in 85 of 310 (27%) cats suspected to have hemoplasmosis, with Mhm being most common (72/310 [23.2%]), followed by Mtc (20/310 [6.5%]) and Mhf (15/310 [4.8%]). Previous studies have also found that Mhm was the most common hemoplasma species infecting cats. In a study⁶ of 713 cats from Switzerland, for instance, Mhm DNA was detected by means of a real-time PCR assay in 7% of healthy cats and 8.7% of ill cats, Mtc DNA was not detected in any of the healthy cats and was detected in only 1.1% of the ill cats, and Mhf DNA was detected in 2.3% of healthy cats and 0.2% of ill cats. In another study,⁷ prevalences of Mhm, Mtc, and Mhf in 426 sick and healthy cats from the United Kingdom were 17%, 2.3%, and 1.6%, respectively; prevalences in 147 Australian cats were 24%, 10%, and 4.8%, respectively; and prevalences in 69 South African cats suspected to have hemoplasmosis were 38%, 26%, and 15%, respectively. In a recent study⁸ performed by our laboratory with similar assays, prevalences of these species in sick cats examined at the University of California Veterinary Medical Teaching Hospital that were not necessarily suspected to have hemoplasmosis were significantly lower than in the present study (Mhm, 16% [P = 0.02]; Mtc, 0.5% [P < 0.001]; and Mhf, 0.5% [P = 0.001]), suggesting that all 3 species may be associated with disease. Coinfection with Mhm was common among cats in the present study that were infected with Mtc or Mhf, which has also been reported in previous studies.^6,7

The reason for the failure of the real-time PCR assay to detect Mhf DNA in 1 infected sample in the present study was unclear. The intensity of the band obtained with the conventional PCR assay was strong, but repeated attempts to detect Mhf DNA with the real-time PCR assay failed. Given the high sensitivity of real-time PCR assays,⁸ lack of sensitivity of the real-time PCR assay seemed an unlikely explanation. More likely, sequence divergence in the region targeted by the real-time PCR assay prevented successful detection of the organism, as was the case for the variant Mhm strain found in 5 cats. Attempts to amplify and sequence the full-length 16S rRNA gene of the Mhf isolate to determine whether this indeed was the case are warranted.

The finding in the present study that for all hemoplasma species, infected cats were more likely to be male was consistent with results of previous studies.^6,7,11-13 An association with FIV seropositivity was found in cats infected with Mhf but not in cats infected with Mtc or Mhm alone. In our recent study,⁸ Mhm infection was associated with FIV seropositivity, but in the Swiss study,⁶ no association between infection with Mhm and retrovirus seropositivity was identified. An association was also found between FeLV antigenemia and Mhf infection in the present study, and retroviral seropositivity was strongly associated with Mhf infection. An association between FIV and hemoplasma infection has been reported for feral cats in Florida, along with an association between FeLV and Mhm infection.¹³ In a study¹⁴ from Israel, prevalences of FeLV and FIV infection were higher in cats with clinical hemoplasmosis than in cats in the general population. As suggested previously,^8,13 the association between hemoplasma infection and male cats and between hemoplasma infection and FIV infection supports biting as a possible mode of transmission of these organisms.

In the present study, cats infected with Mhf were significantly younger than uninfected cats. The study population as a whole was skewed towards geriatric cats, but 6 of the 15 cats infected with Mhf were ≤ 3 years old, whereas the remaining 9 cats were ≥ 9 years old. Coinfection with FeLV was common in the younger cats, and coinfection with FIV was common in the older cats. Previous studies^8,11,12 of cats not necessarily suspected to have hemoplasmosis have shown an association between Mhm infection and older age. Because disease caused by infection with Mhm tends to be mild or subclinical, and the organism persists in infected cats, this may reflect the increased likelihood of becoming infected as a result of prolonged exposure over time. Older studies^14,15,16 that relied on detection of hemoplasma organisms in blood smears for diagnosis of hemoplasmosis, which often occurs in association with hemolytic disease caused by Mhf, also reported an association between infection and younger age. Thus, younger cats may be predisposed to clinical hemoplasmosis.

Cytologic evidence of hemoplasma infection was detected in only 9 of the 310 cats in the present study. Because hemoplasma organisms may detach from erythrocytes in the presence of EDTA,¹⁷ the delay in preparation of blood smears may have contributed to the low number of samples with positive cytology results. However, cytology has been shown to be much less sensitive for the diagnosis of hemoplasmosis than PCR assays.^3,11 Six of the 9 cats with positive cytology results were infected with Mhf, but 2 were infected with Mhm. Interestingly, these latter 2 cats were infected with the variant strain of Mhm that was not detected by the species-specific real-time PCR assay. Severe macrocytosis in the absence of concurrent FeLV infection was identified in 2 cats infected solely with the variant Mhm strain, both of which were also neutropenic, including 1 cat with organisms apparent on blood smears. The other cat with organisms apparent on blood smears had evidence of autoagglutination. Because of the low numbers of infected cats in this study, whether this strain is more pathogenic than other Mhm strains requires further investigation. To date, Mtc has never been detected during cytologic examination of blood smears.

Experimental inoculation of cats with Mhf results in development of moderate to severe anemia, and cats infected with Mhf demonstrate marked fluctuations in copy numbers of hemoplasma DNA.^9,18 Inoculation of an immunosuppressed cat with Mtc resulted in severe anemia,⁵ but only mild anemia was documented in an immunocompetent cat inoculated with Mtc, and significant fluctuations in copy number were not documented. Inoculation of cats with Mhm results in only a mild decrease in Hct; hemoplasma copy number gradually increases, then eventually reaches a plateau.^1,9 The results of the present study suggest that in naturally infected cats, Mtc may have a lower hemolytic capacity than Mhf because cats positive for Mhf had higher reticulocyte counts than did cats positive for Mtc. Unlike cats infected with Mtc, cats infected with Mhf were also more likely to have reticulocytosis and normoblastosis than were cats negative for hemoplasmas. Because 5 of the 20 cats infected with Mtc were coinfected with Mhf, our findings may slightly overestimate the hemolytic capacity of Mtc, as some of the hematologic changes seen in these coinfected cats may have been due to Mhf infection. The fact that macrocytosis, reticulocytosis, and normoblastosis were not more likely in Mhm-infected cats than in cats negative for hemoplasmas may reflect the relative inability of this pathogen to induce hemolysis. Infection with Mhm alone was nevertheless associated with higher MCV values than in uninfected cats, suggesting some effect of infection by this organism on erythrocyte production or turnover. The MCH was more often high than was MCHC in hemoplasmainfected cats, most likely because calculation of the MCHC corrects for cell volume, whereas high MCH can reflect macrocytosis.

Cats infected with Mhf or Mtc in the present study had lower creatinine concentrations than did cats negative for hemoplasmas, and SUN concentrations were also lower in cats positive for Mhf. Serum creatinine concentration was high in 109 of the 185 (59%) cats negative for hemoplasmas for which results of serum biochemistry testing were available. In some of these cats, azotemia may have been due to prerenal causes, but given the older age of this population and the high prevalence of chronic kidney disease in geriatric cats, many likely had renal azotemia. The higher SUN and creatinine concentrations in cats negative for hemoplasmas may reflect the contributions of renal failure to anemia through decreased erythrocyte survival and blood loss, with or without decreased erythrocyte production. Similarly, the lower serum protein concentrations in cats negative for hemoplasmas may reflect a higher prevalence of blood loss as a cause of anemia in these cats, as opposed to increased erythrocyte destruction in hemoplasma-infected cats.

In summary, our data support the suggestion that all 3 hemoplasma species commonly infect cats in the United States. Infection with Mhf was associated with positive retrovirus status, with 7 of 13 cats infected with Mhf also positive for FeLV or FIV infection. The pathogenicity of the 3 hemoplasma species appeared to vary, with Mhf having the greatest hemolytic capacity. Limitations of the present study included the low numbers of samples from cats infected with Mhf or Mtc, the fact that historical data and final diagnoses were not available for each cat, the fact that serum biochemistry panels were not performed on all cats, the lack of confirmatory testing in cats seropositive for retrovirus infection, and the fact that only univariate analyses could be performed because of low sample numbers. Cats in the present study were largely older, and results of serum biochemistry analyses suggested that many likely had underlying renal or hepatic disease. Nevertheless, evidence for variable pathogenicity of the hemoplasma organisms was still apparent, emphasizing the need for accurate diagnosis to the species level in cats infected with hemoplasmas.