Rhodococcus equi

Updated: Apr 13, 2015
  • Author: Indira Kedlaya, MD; Chief Editor: Burke A Cunha, MD  more...
  • Print


Rhodococcus equi primarily causes zoonotic infections in grazing animals, mainly horses and foals. [1, 2] Although R equi rarely infects immunocompetent humans, it is emerging as an important pathogen in immunocompromised persons.

R equi is a facultative, intracellular, nonmotile, non–spore-forming, gram-positive coccobacillus (an organism that has the ability to exist as a coccus or bacillus or intermediate form). Called Rhodococcus because of its ability to form a red (salmon-colored) pigment, R equi can be weakly acid-fast and bears a similarity to diphtheroids. R equi was previously called Corynebacterium equi and is currently grouped with the aerobic actinomycetes. Of the 40 genera in the actinomycetes group, the Rhodococcus genus is placed among the nocardioform bacteria, along with the genera Mycobacterium, Nocardia, Gordonia, Tsukamurella, and Corynebacterium.

R equi was first isolated in 1923 from foals as Corynebacterium equi. R equi infection in a human was first reported in 1967 in a 29-year-old man with plasma cell hepatitis receiving immunosuppressant medications. Since then, R equi has become an important opportunistic pathogen in immunocompromised patients, especially those with acquired immunodeficiency syndrome (AIDS). R equi infection is associated with significant mortality. The organism can be difficult to eradicate, making treatment challenging at times. Increased awareness of R equi infection may help with early diagnosis and timely treatment. Treatment may require prolonged combination antibiotic therapy, sometimes in combination with surgical therapy.



Necrotizing pneumonia is the most common manifestation of R equi infection . Extrapulmonary R equi infections have included wound infection, subcutaneous abscess, brain abscess, thyroid abscess, retroperitoneal abscess, peritonitis, meningitis, pericarditis, osteomyelitis, endophthalmitis, lymphadenitis, lymphangitis, septic arthritis, osteitis, bloody diarrhea, and fever of unknown origin, among others. Bacteremia and dissemination of infection follow from the primary infection site, which is usually the lung.

The primary source of infection may also be from the alimentary tract. This hypothesis arises from a few observations. First, in patients infected with human immunodeficiency virus (HIV), R equi has been isolated from the stool, with or without evidence of pneumonia. Second, R equi frequently colonizes the gastrointestinal tract in grazing animals. In addition, Verville and colleagues (1994) described an incidental finding of R equi mesenteric lymphadenitis in a woman undergoing laparotomy for cholelithiasis. [3] Hence, the possibility of asymptomatic carriage in the gastrointestinal tract in human beings, similar to grazing animals, has been proposed. Third, a case of cervical lymphadenitis has been attributed to a history of sucking raw carrots.

In experimental and natural animal infections, R equi acts as an intracellular bacterium, which survives within macrophages and eventually destroys them. Experimental data suggest that R equi is capable of inhibiting oxidative bactericidal functions of polymorphonuclear cells. Electron microscopy of R equi in equine macrophages demonstrates that the organisms appear to avoid being killed by interfering with phagosome-lysosome fusion.


Most of the information about the pathogenesis of R equi infections is derived from animal isolates. However, the infection in humans seems to differ from that in foals. Makrai et al demonstrated that a 15- to 17-kd virulence-associated protein antigen (VapA), which is highly virulent, may mediate about 88% of the isolates from foals. [4] Nearly all isolates from pigs are of 20-kd virulence-associated protein antigen (VapB) origin, which is of intermediate virulence. In human beings, only about 20-25% of isolates have been reported to express VapA. However, in a study performed by Takai et al in Thailand, about 75% of human isolates expressed VapB, and 25% were avirulent. Most of these patients were infected with HIV. [5]

The expression of VapB is known to vary by geographic location. These differences between human and animal R equi infections are important since most of the investigation has involved VapA isolates. Hence, the conclusion drawn from animal models may not be entirely applicable to the pathogenesis of R equi infections in humans.




United States

R equi infections have been reported in at least 28 states.


R equi infections have been reported on 5 continents. Thus far, a few hundred cases of R equi infections in immunocompromised persons have been reported in the literature. At least 19 cases of R equi infection have been described in immunocompetent patients.


Morbidity is related to complications and chronicity of the infection. Numerous complications are related to R equi infections. R equi pneumonia may be complicated by the following:

  • Abscess
  • Empyema
  • Hemoptysis
  • Direct chest wall involvement
  • Pneumothorax

Pericardial tamponade may result from purulent pericarditis. Bacteremia leading to overwhelming sepsis has been reported, more often in immunocompromised patients. In a review by Verville et al (1994), about 47% of patients infected with HIV and 17% of patients with non–HIV-associated immunocompromised conditions had chronic R equi infection. [3] Relapses are also common after discontinuation of antibiotics. An important site of extrapulmonary relapse is the central nervous system.

R equi infections carry an overall mortality rate of about 25%. In 2 different reports, by Cornish et al [6] (1999) and by Harvey and Sunstrum [7] (1991), the mortality rate was 50-55% in patients infected with HIV and 20-25% in patients with non–HIV-associated immunocompromised conditions. In contrast, the mortality rate is only about 11% in immunocompetent patients.

Since the advent of highly active antiretroviral therapy (HAART) therapy, mortality rates in HIV-infected patients have decreased. Torres-Tortosa et al conducted a multicenter observation of 67 HIV-infected patients in Spain. [8] The mortality rate related to R equi was 34.3%. In a univariate analysis in the same study, factors associated with worse prognosis included multilobar involvement, absence of HAART, and inappropriate antibiotic therapy. In multivariate analysis, absence of HAART was the only factor independently associated with R equi –related mortality. While the mortality rate is lower in immunocompetent patients, it is still significant. Lower mortality rates in this subgroup of patients may be due to the fact that localized infections represent about 50% of the cases reported. The high mortality rates associated with R equi infections are due to several factors.

R equi may be misidentified as diphtheroids, Mycobacterium species, or Nocardia species among both immunocompetent and immunocompromised populations.

Patients with R equi infection may receive inappropriate initial antibiotic therapy because of misdiagnosis. R equi pneumonia does not respond to standard empirical treatment with beta-lactams (other than imipenem and meropenem) and tetracyclines. On the other hand, some cases of R equi pneumonia may be susceptible to macrolides and the newer quinolones.

Simultaneous opportunistic infections are common, especially in patients infected with HIV. In this subgroup of patients, the mortality rate directly attributed to R equi infection alone may be less. Capdevila et al (1997) reported on a series of patients infected with HIV who had R equi pneumonia; in this group of patients, the mortality rate directly attributed to R equi infection was only 15.4%. [9] In another review of R equi infections in patients infected with HIV, 4 of 12 patients died of R equi infection, while 3 deaths were due to opportunistic infections.


R equi infections have no reported racial predilection.


In all R equi infections, the male-to-female ratio is about 3:1. The reason for this is not clear; however, among immunocompromised patients, the predilection in males may be explained by the higher prevalence of HIV infection among males.


R equi infections have been described in all age groups, from infants to elderly persons. Separate reviews have found a mean age of infection of 34-38 years.

Infection in children

R equi infections in children differ from those in adults. Immunocompromising conditions, including hematopoietic malignancies, immunosuppression associated with chemotherapy, and HIV infection, account for only about one third of reported R equi infections in children. Pediatric R equi infections account for approximately one third of all cases among immunocompetent individuals, perhaps because of the increased prevalence of trauma among children, predisposing them to localized R equi wound infections .R equi infections in immunocompetent children carry an extremely favorable prognosis.