Treatment of acute and chronic endometritis (Proceedings)

Correcting the defects in uterine defense, neutralizing virulent bacteria, and controlling post-breeding inflammation are the goals of successful therapy. This is accomplished by surgically correcting anatomical defects, improving physical drainage after insemination, reducing the length or modulating the inflammatory response to insemination and inhibiting bacterial growth. Post-breeding inflammation is most commonly treated by improving physical clearance of uterine fluid with uterine irrigation followed immediately by administration of either oxytocin (10-25 IU i.v. or i.m.) or cloprostenol (250 μg i.m.). In some cases, the uterus is infused with antibiotics post mating. Bacterial or fungal endometritis are routinely treated for 3 to 5 days during estrus with either intra-uterine or systemic antibiotics in combination with uterine irrigation. Emphasis here will be on the use of mucolytics, chelating agents and administration of steroids for modulating the inflammatory response.

Mucolytics

Not all infections respond to uterine irrigation and antibiotic treatment. Treatment failure may be due to continual contamination of the uterus because of anatomical abnormalities in the caudal tract, degradation of antibiotic in uterine exudates, or biofilm production by the micro-organism. Previous work indicates that mucus secretion increases during experimental uterine inflammation, and in mares with delayed uterine clearance and bacterial endometritis. Mares with chronic endometritis have an increase in the thickness of the mucus ribbon overlying the endometrium, increased staining intensity of both intracellular and extracellular mucus, and epithelial cell loss. Excessive mucus or exudate can interfere with antibiotic penetration, can render aminoglycosides chemically inert or may interfere with sperm transport to the oviduct. Treatment with a mucolytic agent may help clear mucus and increase effectiveness of intra-uterine antibiotics. Solvents and mucolytic agents have been added to uterine irrigation fluids in an attempt to clear exudate, mucus or biofilm. Agents used include DMSO, kerosene and N-acetylcysteine. Each compound appears to have some beneficial effects. Barren mares (n = 16) infused with a 30% solution of DMSO after breeding tended to have higher pregnancy rates than mares infused with saline. Intrauterine DMSO therapy also resulted in a significant improvement in endometrial biopsy classification in 18 of 27 mares; whereas only 2 of 18 barren mares improved following intrauterine saline treatment. In contrast, intrauterine infusion of 50 ml of commercially available kerosene in 26 mares with varying degrees of endometrial pathology induced diffuse moderate to severe endometritis, severe diffuse edema and production of a serum-like exudates. Half of the mares exhibited mild to severe necrosis of luminal epithelium. Mares were subsequently bred on the next cycle and surprisingly, 50% of the mares with Category II or III biopsy scores carried foals until term. Although kerosene was associated with significant inflammatory changes, pregnancy may have been established because mucus and exudate were removed via destruction and necrosis of uterine epithelium.

N-acetylcysteine (NAC) is a mucolytic agent that disrupts disulphide bonds between mucin polymers, thereby reducing the viscosity of mucus. In addition, NAC possesses antioxidant and possibly some antimicrobial properties. NAC has been used to treat respiratory diseases such as pneumonia, the pulmonary component of cystic fibrosis in humans, meconium impactions in both humans and equine neonates and meconium aspiration pneumonia in equine neonates. Multiple studies support its beneficial anti-oxidative properties especially in chronic inflammatory diseases. We have recently evaluated its effect on the endometrium and epithelium (Gores-Lindholm et al. 2009). Endometrial biopsies were obtained from 12 fertile and 10 barren mares before and after infusion of a 3.3% solution of N-acetylcysteine (day 1) and compared to biopsies obtained from mares infused with saline. The uterus of all mares was irrigated with 2 L of lactated Ringer's solution on days 2 and 3 and a second biopsy obtained. Endometrial biopsies were given a Kenney grade by a board certified veterinary pathologist and changes in epithelial architecture and mucus blanket were measured by image analysis. Data indicated that NAC was not harmful to the endometrium and that it may counteract the irritating effect of saline, as reflected through increased cell height in control mares. As further evidence that NAC does no harm and may be beneficial, 20 Thoroughbred mares each bred 2 to 5 times in 2007 or 2008 and with a history of endometritis were mated naturally to commercial stallions in Central Kentucky in late May and June 2008. Mares received a 0.6% solution of ACE either the treatment cycle before (n = 10) or in the 48 h before breeding (n = 10) in addition to conventional treatments. Infusion before breeding was associated with higher than expected pregnancy rates as 17 of 20 mares (85%) conceived and foaled in 2009. Before this study, the rationale for using NAC as a uterine infusion had been the removal of inspissated secretions, exudate and biofilm, (i.e. as a mucolytic). However, since increased vaginal mucus viscosity is documented to inhibit sperm forward progression in cows, it is also speculated that NAC may improve sperm-transport in mares with excessively viscous mucous secretions by breaking the cross-linking disulfide bridges between mucin polymers.

Bacterial and yeast biofilms

Antibiotic failure in chronic endometritis may be due to biofilm produced by some gram negative bacteria, yeast and fungi. Bacterial biofilms consist of a heterogeneous community of different bacterial species, surrounded by an extracellular matrix, that co-exist in a symbiotic relationship. Such biofilms are found throughout the human body, e.g. the oral cavity, the skin, the intestines and the vagina. In most cases, the inhabitants of this community are considered as normal flora and serve as a protective mechanism to prevent the colonization of frank and opportunistic pathogens. If the balance of this biofilm community is upset or disrupted, pathogens may colonize, proliferate, and cause disease. Biofilms confer antibiotic resistance and therefore contribute to treatment failure. A number of theories have been advanced to account for this increased resistance. One is simply that the antibiotic is unable to penetrate the extracellular matrix of the biofilm. Another is that antibiotics are less active on biofilms due to the lower rate of metabolism and growth. A currently popular theory is that there are "persister cells" within the biofilm community. Persister cells are defined as a small subpopulation of essentially invulnerable cells that neither grow or die in the presence of bactericidal agents and exhibit multi-drug tolerance or resistance to antibiotics.

Pseudomonas aeruginosa is a potent biofilm producer and is often cultured from the uterus of mares with chronic endometritis. Other equine pathogens that produce biofilm and can be isolated from the uterus include Staphylococcus epidermis, Escherichia coli, Enterobacter cloacae and a number of yeast and fungi. These organisms more commonly cause endometritis in older, pluriparous barren mares that have anatomical defects than in young, fertile mares, although uterine defenses can be broached in the latter resulting in chronic infection. Infections by these organisms can be difficult to treat, are often refractory to a 3 to 5 day course of antibiotics, and may result in a population of bacteria colonizing the uterus that is highly resistant to the drug initially used for treatment. Work in other species and in the mare has shown that buffered chelating agents (Tris-EDTA, Tricide®) may potentiate the actions of antimicrobials, dissolve exudate, and break up biofilm.

Buffered chelators such as first generation Tris-EDTA (ethylene-diamine tetra-acetic acid (3.5 M)-tromethamine 50mM;) and third generation Tricide® (8mM disodium EDTA dehydrate and 20 mM 2-amino-2-hydroxymethyl-1,2-propanediol; Medical Molecular Therapeutics, LLC Georgia Biobusiness Center, Athens, Georgia 30602; tricideinfo@yahoo.com) potentiate the actions of antimicrobials. They have been shown to enhance the bactericidal effects of antimicrobials in dogs with refractory otitis, pyoderma, osteomyelitis, multiple fistulas, rhinitis, and cystitis. Isolates of Pseudomonas collected from the uterus of mares with chronic endometritis exhibited decreased growth and/or death when exposed to Tris-EDTA solution. Others have shown that addition of Tris-EDTA to gentamicin in vitro improved killing of Pseudomonas aeruginosa by 1000 fold more than treatment with only gentamicin. Addition of Tris-EDTA to penicillin, ampicillin, oxytetracycline, neomycin, and amikacin has also been shown to be synergistic. A recent study showed that Tricide®, a third generation buffered chelating agent, increased in vitro activity of antifungal drugs against common fungal pathogens isolated from eyes of horses with mycotic keratitis. The mechanism of action of buffered chelating agents is not completely understood but it is speculated that the chelating agent (EDTA) chelates calcium and/or magnesium from the outer membrane of bacteria, thereby altering the integrity and permeability of the cell wall. Damage to the cell wall interferes with the effectiveness of the bacterial efflux pump and facilitates osmotic collapse. Unlike bacteria, fungal cell walls are composed mainly of polysaccharides (beta-glucans and chitin) and protein. It is hypothesized that removal of divalent cations in the cell wall by third generation chelating agents may alter membrane proteins that are important in maintaining the construction and maintenance of the polysaccharides in the wall.

Buffered chelating agents must come in direct contact with the bacterial cell wall in order to kill the organism so the volume of solution needed for infusion will vary with the size of the uterus. Doses ranging from 200 to 500 ml are recommended. The chelating agent binds to the bacteria within minutes resulting in cell death and accumulation of debris so the uterus should be lavaged within 12 hours to remove these by-products. Our current recommended therapy for gram negative bacteria and yeast repeatedly isolated from the uterus of a mare with endometritis is to infuse 250 to 500 ml of Tricide® or Tris-EDTA into the uterus on Day 1, lavage the solution out within 24 h and examine the efflux. If the efflux is cloudy or has mucus strains the chelating agent is infused into the uterus again on Day 2. Antibiotics are then begun on Day 3 following uterine irrigation and continued daily for a minimum of 5 days.

Modulation of the inflammatory response

Fluid may accumulate within the uterine lumen during estrus because it is not physically drained through the cervix, production is increased because of chronic inflammation, or because the mare is refluxing urine into her uterus. Degenerative uterine changes such as vascular elastosis may also contribute to fluid accumulation. Vascular elastosis appears to indirectly reduce fertility through a reduction in endometrial perfusion, and through disturbances in uterine drainage caused by reduced venous return in capillary beds. For the past 20 years, treatment of post mating induced endometritis has emphasized methods for improving physical drainage. However, modulation of the immune response with steroids given around the time of mating has been shown to increase pregnancy rates in mares with fluid accumulation or uterine inflammation. Immunomodulation may help restore homeostatic local inflammatory mechanisms through reducing pro-inflammatory cytokines. This may be especially helpful in older mares that may be suffering from inflamm-aging. Inflamm-aging is a low-grade, systemic inflammatory response associated with advanced age in humans and horses that is characterized by increased inflammatory cytokine production. Peripheral blood mononuclear cells collected from old horses have been shown to produce more inflammatory cytokines than young horses; moreover, fat old horses have even greater frequencies of lymphocytes and monocytes producing inflammatory cytokines than thin, old horses. Weight loss in old, fat mares reduced the percent of IFNγ and TNFα positive lymphocytes and monocytes and serum levels of TNF α protein. When weight and fat increased in these old horses, there was a significant increase in inflammatory cytokine production.

Single dose dexamethasone administered within one hour of mating and daily prednisolone administration given before and after mating have improved pregnancy rates in mares with uterine fluid. A single injection of dexamethasone administered within one hour of mating (50 mg, IV; approximately 0.1 mg/kg) combined with routine post breeding therapies (uterine irrigation, ecbolic drugs and in some cases intra-uterine antibiotics) resulted in increased pregnancy rates in mares with a history of fluid accumulation after ovulation and in mares with cervical incompetence. Treated mares exhibited decreased uterine edema, decreased intrauterine fluid and an increase in uterine fluid clarity. Although dexamethasone did not increase pregnancy rates in the general population, pregnancy rates were increased in mares that had 3 or more risk factors for susceptibility to endometritis. Risk factors included abnormal reproductive history, abnormal perineal conformation, vulvoplasty not repaired after foaling, an incompetent cervix, positive endometrial culture, ≥ 2 cm of endometrial fluid before breeding, endometrial fluid post mating between 1.5 and 2.0 cm, or a fluid volume ≥ 2cm, and endometrial fluid persisting more than 36 hours after mating. Increased pregnancy rates were also observed in mares with a history of intra-uterine fluid accumulation following oral administration of acetate 9-alpha-predinisolone (0.1 mg/kg) given at 12 h intervals for 4 days beginning 48 hrs before breeding. In contrast, administration of dexamethasone (10 or 20 mg, IM) 6 to 12 h after insemination did not improve pregnancy rates of warmblood mares (n=783 cycles) with a history of intra-uterine fluid retention. A plausible cause for the different results is that steroids block both the cyclooxygenase and 5-lipoxygenase pathways of inflammation. The 5- lipoxygenase pathway includes leukotriene B, a potent neutrophil chemotactic factor found in uterine fluids of susceptible mares after mating. Reducing neutrophil chemotaxis and the number of neutrophils recruited into the uterus post mating may diminish the severity and length of the inflammatory response. Candidates for steroid use should be chosen carefully as misuse in mares with bacterial endometritis may exacerbate the infection.