Methicillin-resistant Staphylococcus aureus in horses and horse personnel, 2000-2002

Continued from page 1.
Clinical infections developed in 13 (16%) horses at 1 or more body sites. Incision or wound infections (n = 6), infection from intravenous catheter (n = 2), bacteremia (n = 2), pneumonia (n = 1), infection from surgical implant (n = 1), septic arthritis (n = 1), omphalophlebitis (n = 1), gluteal abscess (n = 1), and osteomyelitis (n = 1) were identified. In 2 neonatal foals, MRSA was isolated from the nares, blood, and intravenous (jugular) catheter; septic arthritis developed in 1 of the foals. The intravenous catheter was thought to be the original site of infection in both cases. Ten (85%) of 13 clinically affected horses had a history of contact with colonized persons; 2 of these also had contact with infected or colonized horses. No history of known contact with colonized humans or horses was present in 3 (15%) cases. Although many of the clinically infected horses were seriously ill and required prolonged hospitalization, MRSA was implicated directly in the death of 1 horse as a result of severe osteomyelitis.

The clinical significance of 3 isolates was unclear. One was from a routine prebreeding uterine swab from a mare without any apparent clinical abnormalities or history of reproductive disease. Two were from uterine swabs from 2 mares following abortions. Whether MRSA was the cause of abortion in either case was unclear. The remaining 63 (80%) horses had subclinical infection and were nasal carriers.

Typing was performed on all 27 OVC-VTH isolates, 37 (90%) of 41 farm A isolates, and all 11 isolates from other Ontario farms. The remaining 4 isolates, all from colonized horses, were lost before being typed. Of the 75 tested isolates, including all 13 isolates from clinical infections, 72 (96%) were identified as CMRSA 5. This strain is closely related to an international epidemic strain designated the "archaic clone" (ATCC BAA-38) (11). Nine different subtypes of CMRSA-5 were identified from horses. All 12 tested isolates from horses with clinical infections contained SCCmecIV, were spa type 7, and did not contain PVL genes. The 2 isolates from postabortion uterine swab specimens also contained SCCmecIV and were negative for PVL genes; however, 1 was spa type 7 and the other spa type 235. Five isolates from colonized horses were tested, and all also contained SCCmecIV and were spa type 7. All 3 non-CMRSA-5 isolates were obtained from farm A and were similar to, but distinguishable from, CMRSA-2, spa type 2 and contained SCCmecIV.

MRSA infection or colonization was suspected of being nosocomial in 17 (63%) of 27 OVC-VTH cases and community-acquired in 3 (11%) cases. Origin of infection was unclear in the remaining 7 (26%) cases. In these cases, the nasal swab specimen was not collected on admission; MRSA was isolated from the first sample collected >72 hours after admission, and horses were admitted from farm A during a time when numerous colonized horses were on the farm.

Sixty-eight of the equine isolates were obtained during periods of organized screening at OVC-VTH or on horse farms. The remaining 11 isolates were from clinical specimens submitted directly by primary care veterinarians to a diagnostic laboratory and are excluded from prevalence calculations. In 2000, MRSA was isolated from the nasal passages of 2 (4%) of 57 horses, including the 2 initial clinical cases. In 2002, MRSA was isolated from 25 (8%) of 320 horses at OVC-VTH and 41 (13%) of 321 horses on farm A. Of the 9 other farms evaluated after identification of an infected or colonized horse at OVC-VTH, MRSA was only identified on 1 farm, where 3 (5%) of 64 of horses were colonized. MRSA was not isolated from any of 277 horses from 8 other Ontario farms.

MRSA was isolated from 27 persons; 17 (14%) of 125 of tested OVC-VTH personnel, 8 (12%) of 67 of farm A personnel, 1 owner of a horse with an MRSA wound infection and the spouse of a colonized OVC-VTH clinician. Three human isolates were obtained in 2000, and 24 were obtained in 2002. Only 1 (4%) was from a source with clinical infection, an OVC-VTH veterinarian with a tattoo site infection. That person was infected with CMRSA-5 subtype H12, a strain that contained SCCmecIV, was spa type 7, and was PVL negative, and that strain was isolated from 2 horses that had been under that person's care for a week before the wound infection developed. All but 1 colonized person (96%) had previous contact with 1 or more MRSA-positive horses; in 24 (89%) of 27 persons, recent contact with a horse infected with an indistinguishable subtype was documented. The colonized spouse of the colonized OVC-VTH clinician reported no contact with horses; however, isolates from both of these persons were indistinguishable from an isolate recovered from a horse under that clinician's care. CMRSA-5 was isolated from 26 (96%) of 27 persons. One person harbored both CMRSA-5 and the CMRSA-2--like isolate in his nose at the same time. This person was a veterinarian from farm A, which was the origin of the 3 horses colonized with this strain. Nine different subtypes of CMRSA-5 were identified among human isolates.

All but 2 of the human isolates were obtained during organized screening of personnel from OVC-VTH or selected Ontario horse farms. In 2000, MRSA was isolated from 2 (10%) of 21 humans at the OVC-VTH. In 2002, MRSA was isolated from 15 (12%) of 127 persons at OVC-VTH and 8 (12%) of 68 from farm A.

Antimicrobial susceptibility testing was performed on 67 of the 72 equine and all 26 human CMRSA-5 isolates, and the 5 CMRSA-2-related strains. Five of the equine CMRSA-5 isolates were unavailable for testing. All 101 MRSA isolates tested were susceptible to ciprofloxacin, clindamycin, fusidic acid, linezolid, mupirocin, quinupristin-dalfopristin, and vancomycin. The range of oxacillin MIC was 1->32 [micro]g/mL, and although 21.8% of strains had oxacillin MIC of [less than or equal to] 2 [micro]g/mL at 24 h, all such strains grew on the NCCLS oxacillin screen agar and produced the PBP2a protein. Isolates of both CMRSA-5 and CMRSA-2--related strains that were erythromycin-resistant were found to be inducibly resistant to clindamycin when challenged by using the double disk approximation test. The remaining susceptibility test results are presented in the Table.


This study has identified the largest number of reported cases of clinical MRSA infection in horses and horse personnel. It also identified extensive nasal colonization in horses and horse personnel from a veterinary hospital and horse farm, nosocomial infection in a veterinary hospital setting with clinical illness in horses, and for the first time, clinical infection in 1 person working with infected horses. The subtyping information and timing of isolation provide solid evidence supporting both human-to-horse and horse-to-human transmission.

The prevalence of MRSA colonization in horses at the OVC-VTH was 4% in 2000 and 8% in 2002; however, care must be taken when interpreting these data because screening was performed during 2 periods that followed identification of clinical MRSA infection in horses at the facility. Similar limitations are present with the prevalence data regarding MRSA colonization on breeding farms, which ranged from 0% to 13% and were based on screening after identification of MRSA infection or colonization at OVCVTH in horses from these farms. The prevalence of human colonization at OVC-VTH and 1 Ontario horse farm is of concern, particularly because of the likelihood of transmission between horses and humans on these farms. As with horses, the prevalence data in humans must be interpreted with care because of the nature of sampling. Further studies are required to determine the prevalence of MRSA infection and colonization in horses and humans at veterinary hospitals and equine farms. One well-recognized human S. aureus strain, CMRSA-5, a relatively uncommon isolate in Canada (11), has the ability to colonize the nose of horses and to spread between horses and between horses and persons on farms and within a veterinary hospital setting. The PFGE pattern of the CMRSA-5 isolates was similar to the PFGE patterns published in a previous report of MRSA infection in horses (13). This finding, along with the isolation of CMRSA-5 from horses in Prince Edward Island, Canada (J.S. Weese et al., unpub, data) and Colorado, USA (P. Morley, pers. comm.) that did not have any contact with colonized Ontario horses or horse personnel, further suggests that CMRSA-5 may be more disseminated in the horse population beyond Ontario.

Article Pages

Page 1
Page 2
Page 3
Page 4