Materials and Methods

Selection of farms and horses—Two Thoroughbred breeding farms located in central Kentucky were included in the study. Farm A had never had a confirmed case of EPE; at farm B, 5 severe cases of EPE (in which weanlings had clinical signs and required aggressive supportive care at the farm or hospitalization for treatment) had occurred from August to January during each of the preceding 2 years. Diagnosis of EPE in those horses was made on the basis of clinical signs as well as findings of hypoalbuminemia and hypoproteinemia combined with positive results for L intracellularis–specific IPMA or L intracellularis–specific PCR assay of fecal samples.

Fifty foals (26 colts and 24 fillies) were selected by use of a simple randomization method from a total of 75 foals at farm A at the start of the study. All foals from farm A were born at the farm and were kept at the same barn and pasture from birth until weaning. After weaning, each colt or filly was moved to a weanling and yearling farm located approximately 20 miles away from farm A. Each weanling was assigned to a specific barn, stall, and pasture at the weanling and yearling farm throughout the study period. The weanling and yearling farm was located approximately 15 miles away from farm B.

All foals (n = 53; 24 colts and 29 fillies) present at farm B at the start of the study were selected for participation. All of these foals were born in the same barn and then, depending on space availability, were moved along with the dam to another barn on the farm. During this interval, each mare and foal was assigned to a group that used the same barn and pasture. After weaning, mares were moved to one end of the farm and weanlings were moved to the other end. Weanlings were then assigned to a specific barn, stall, and pasture for the duration of the study.

Weanlings at farm A and farm B were segregated to specified areas of the respective farms on the basis of sex. Additionally, both farms were broodmare and foaling farms, with no stallions kept on the premises. Several dogs and cats, which would occasionally roam freely through the barns, were kept as pets at each farm.

The study began on August 14, 2008, and continued until January 14, 2009. At the start of the study, each farm had a group of weanlings and a group of foals that were still paired with mares, as foals were typically weaned at 4 to 5 months of age. All foals were weaned by the first week of October 2008. No horses (broodmares, foals, or weanlings) at either farm were vaccinated with the commercially available swine L intracellularis vaccine.^b Consent for this study was provided by the owners of both farms.

Daily evaluation—No changes, except for the collection of study materials, were made to the routines of either farm for purposes of the study reported here, including daily evaluations of the foals and weanlings. Management at the 2 farms was similar in that each foal or weanling was examined carefully at least twice daily for wounds, swelling (including dependent edema), changes in demeanor, or decreased appetite. Evaluations were made by the same workers at each farm; these individuals were familiar with each horse's typical behavior, demeanor, and appetite. Grain rations were given to each foal or weanling twice daily, and appetite was assessed via observation of the amount of grain consumed and the rapidity with which it was eaten. Any concerns about a horse were brought to the attention of an assistant farm manager, who contacted the farm veterinarian. Rectal temperatures were measured for each foal or weanling every morning, and any temperature > 39.2°C (102.5°F) was reported to the farm veterinarian.

Fecal sample collection and evaluation—Fecal samples were collected from each foal or weanling enrolled in the study every 2 weeks from August 14, 2008, through January 14, 2009. Fecal samples were collected into individually labeled resealable plastic bags, placed in insulated boxes with ice packs, and sent overnight to the University of California-Davis Lucy Whittier Molecular Laboratory for real-time PCR detection of the aspA gene of L intracellularis as previously described.¹⁷ If samples could not be sent the day of collection, samples were stored in a standard laboratory refrigerator at 4°C until overnight shipment was possible.

Blood sample collection and serum IPMA—A blood sample was collected from each foal or weanling in the study once monthly from August 14, 2008, through January 14, 2009. Blood samples (7 mL) were collected via jugular venipuncture into individual plain, sterile 7-mL collection tubes. Serum was separated via centrifugation at 800 × g. Serum samples for farm A were stored in a freezer at −80°C and sent in batches every 3 months to the University of Minnesota diagnostic laboratory for analysis. Refrigerated serum samples (4°C) from farm B were sent to the same laboratory overnight in insulated boxes with ice packs within 24 hours after collection. Serum antibody titers were measured in serial 2-fold dilutions of serum; L intracellularis–specific IgG antibodies were quantitated by use of IPMA as previously described.⁹ A titer of 1:60 was used as the minimum cutoff value for a foal or weanling to be considered seropositive, according to the experience of one of the authors (CJG).

Further diagnostic evaluation—An additional blood sample (8 mL; 4 mL in a heparinized collection tube and 4 mL in an collection tube containing EDTA) was obtained from any foal or weanling with a serum anti–L intracellularis antibody titer ≥ 1:60 or a positive result for the L intracellularis–specific fecal PCR assay. This sample was used to assess a CBC, serum fibrinogen concentration, and results of serum biochemical analysis in an effort to identify subclinical L intracellularis infections. For purposes of the present study, young horses were considered subclinically affected if they had positive results for L intracellularis–specific IPMA (ie, serum antibody titers ≥ 1:60) or for the L intracellularis–specific fecal PCR assay with low serum concentrations of total protein (< 6.0 g/dL) and albumin (< 3.4 g/dL) but without clinical signs of EPE. The CBC and serum fibrinogen concentrations were evaluated to assess the potential for other systemic derangements, including inflammatory processes. Foals and weanlings that had clinical signs compatible with EPE had the described analysis performed and were also examined via abdominal ultrasonography.

The additional blood samples were obtained via the described methods as soon as possible after IPMA or PCR assay results were received. Samples were submitted to the Hagyard Equine Medical Institute's laboratory for evaluation within 24 hours of collection. An automated analyzer^c was used for CBCs, and any anomalies were evaluated by certified laboratory personnel using manual techniques. Serum fibrinogen concentrations were determined by use of a plasma heat precipitation method at 56°C.¹⁹ Serum biochemical values were determined by use of a chemistry analyzer.^d Included in the biochemical analysis were concentrations of sodium, potassium, chloride, total carbon dioxide, glucose, BUN, creatinine, total calcium, phosphorus, total protein, albumin, and total bilirubin; activities of γ-glutamyl transpeptidase, aspartate aminotransferase, lactate dehydrogenase, and creatine kinase; and anion gap calculation.

Treatment and referral—All foals and weanlings that had positive results for L intracellularis–specific PCR assays were treated because of concern that they would develop clinical signs of EPE. For the remaining foals and weanlings, the decision to provide treatment often involved evaluation of initial clinicopathologic data, particularly serum concentrations of total protein and albumin. Treatment was initiated for a period of 1 week for any foal or weanling that had low serum concentrations of total protein (< 6.0 g/dL) and albumin (< 3.4 g/dL). Initial treatment consisted of an orally administered antimicrobial drug, typically doxycycline (10 mg/kg [4.5 mg/lb], PO, q 12 h). Prior to initiation of treatment that included the use of a tetracycline-related antimicrobial drug, serum BUN and creatinine concentrations were evaluated to ensure that renal function was adequate.

After the 1-week treatment period, an additional blood sample (4 mL) was collected and serum biochemical analysis was repeated. Treatment was stopped for foals or weanlings with total protein and albumin concentrations within reference limits, whereas treatment was continued for an additional week for those with persistently low values. Weanlings that had a decrease in serum total protein and albumin concentrations after 1 week of doxycycline administration had this drug discontinued and were administered chloramphenicol (50 mg/kg [22.7 mg/lb], PO, q 8 h). Any foal or weanling that had clinical signs of EPE received treatment for 2 weeks; those with severe clinical signs (complete anorexia, moderate or severe dependent edema indicating a need for colloidal support, persistently high rectal temperature > 39.4°C [103°F], or poor or no response to treatment at the farm) were referred to the Hagyard Equine Medical Institute's McGee Medicine Center for treatment and supportive care.

Results

All foals in the present study had been weaned by October 2008. Two weanlings and 1 foal (2 fillies and 1 colt) from farm A were excluded from the data set after the study ended. The 2 weanlings had been moved to another farm and were not exposed to the same environmental factors as the other weanlings. The third young horse from farm A for which data were excluded was 1 of 2 from that farm that died during the study. The death of this foal within the first 2 months of the study was attributed at necropsy to Rotavirus infection, causing it to be censored. A weanling at farm A died from acute cervical trauma after running into a fence during the last month of the study (data from this weanling were included in the analysis). One weanling at farm B that was hospitalized for EPE was euthanized; data from this weanling were also included in the analysis. Eight to 12 fecal samples were collected from each foal or weanling during the study for L intracellularis–specific PCR assays, and up to 6 blood samples (1 sample/mo) were collected from each foal or weanling in the study for L intracellularis–specific IPMA analysis; some samples were not obtained from horses that left farm B near the end of the study.

At farm A, a total of 9 fecal samples were collected from each foal or weanling, except for 2 foals for which 1 collection was unintentionally omitted. Collection of 10 fecal samples from each individual was planned; however, a Rotavirus outbreak precluded collection of samples during the second half of September. None of the 421 fecal samples collected at this farm during the study tested positive for L intracellularis via PCR assay. All foals and weanlings were determined to be seronegative via L intracellularis–specific IPMA at the onset of the study, and no foals seroconverted until after the time they were weaned. Eight of 47 (17.0%) weanlings at farm A had serum anti–L intracellularis antibody titers ≥ 1:60 detected via IPMA during the study; 2 had antibody titers of 1:240, and the other 6 had antibody titers of 1:60. Clinicopathologic data (results of CBC, serum fibrinogen concentration determination, and biochemical analysis) were assessed for these 8 weanlings, and no abnormalities were detected. The first positive results for L intracellularis–specific IPMA of serum samples from weanlings at farm A occurred in November 2008; the largest number of weanlings that tested positive for the first time was detected in December 2008 and January 2009 (Figure 1). During the study period, no weanlings at farm A developed clinical signs compatible with EPE and none received treatment for EPE.

View Full Size

Distribution of serum samples from 8 Thoroughbred weanlings that tested positive (antibody titers ≥ 1:60) via IPMA for anti–Lawsonia intracellularis antibodies for the first time at a farm in which L intracellularis was nonendemic (farm A). A small increase in the incidence of newly identified antibody titers above the cutoff value in weanlings is evident during the months of December and January. Bars represent serum antibody titers (white, 1:60; black, 1:240).

Citation: Journal of the American Veterinary Medical Association 238, 11; 10.2460/javma.238.11.1482

• Download Figure
• Download figure as PowerPoint slide

At farm B, collection of 12 fecal samples from each foal or weanling was planned; the larger number of samples obtained at this farm, relative to farm A, was the result of a difference in collection schedules between farms. Eighteen weanlings from farm B left the farm during the last 3 months of the study (2 in November, 4 in December, and 12 in January), which precluded collection of 25 fecal samples; however, the remaining data were valid and were included in the analysis. Eight of 611 (1.3%) fecal samples tested positive for L intracellularis via PCR assay during the study period. These 8 positive results were from 6 different weanlings; 2 weanlings had consecutive positive test results 1 week apart. The second test was performed on all horses 1 week early at the request of farm staff because of a perceived increase in incidence of EPE. Ages of the 6 weanlings at the time of positive PCR assay results ranged from 6 to 8 months. Of the 6 weanlings that had positive PCR assay results, 4 had positive IPMA results at the same time. The other 2 weanlings that had positive PCR assay results seroconverted before the next blood sample was collected. On the basis of results of PCR assays alone, the month with the greatest number of samples with positive results (5 samples) was October 2008.

None of the foals or weanlings at farm B had detectable serum anti–L intracellularis antibody titers during the month of August, and none developed measureable serum antibody titers before weaning. Of 243 blood samples collected from foals and weanlings at farm B between September 2008 and January 2009, 62 (25.5%) had positive results for serum anti–L intracellularis antibody titers via IPMA. The 62 serum samples with positive IPMA results were obtained from 32 of the 53 (60.4%) weanlings at farm B. The largest number of weanlings that tested positive for the first time was detected in November 2008, and the second largest number was detected in October 2008 (Figure 2). Analysis of initial (first-time positive) anti–L intracellularis antibody titers revealed 5 weanlings with titers of 1:60, 19 weanlings with titers of 1:120, and 8 weanlings with titers of 1:240. During the study period, the maximum serum antibody titers detected were as follows: 1:60 (5 weanlings), 1:120 (17 weanlings), 1:240 (9 weanlings), and 1:960 (1 weanling).

View Full Size

Distribution of serum samples from 32 Thoroughbred weanlings that tested positive for anti–L intracellularis antibodies for the first time at a farm in which L intracellularis was endemic (farm B). An increased incidence of newly identified antibody titers above the cutoff value in weanlings is evident during the months of October and November. Bars represent serum antibody titers (white, 1:60; gray, 1:120; black, 1:240).

Citation: Journal of the American Veterinary Medical Association 238, 11; 10.2460/javma.238.11.1482

• Download Figure
• Download figure as PowerPoint slide

Five weanlings at farm B had clinical signs of EPE during the study period. Of these, 4 had maximum serum antibody titers of ≥ 1:240 and 1 had a maximum serum antibody titer of 1:120; the 4 weanlings with maximum titers of ≥ 1:240 had serum total protein and albumin concentrations that were considered moderately to severely low (total protein < 5.4 g/dL and albumin < 3.0 g/dL). Nineteen of the 32 (59.4%) seropositive weanlings were treated for EPE during the study. The decision to treat was made primarily on the basis of serum total protein and albumin concentrations and, if present, clinical signs; a blood sample was inadvertently not obtained for this purpose from 1 weanling at farm B (Table 1). Examination of the initial clinicopathologic data for each weanling did not reveal any abnormalities in total WBC counts (reference range, 5,000 to 12,600 cells/μL), serum fibrinogen concentrations (reference range, 200 to 500 g/dL), or serum biochemical variables other than total protein and albumin concentrations.

Clinicopathologic and IPMA data were compared with the results of fecal PCR assays. At the time their feces tested positive via PCR assay, 3 of 6 weanlings had normal serum total protein concentrations (6.0 to 7.9 g/dL) and 2 of 6 weanlings had normal serum albumin concentrations (3.4 to 4.1 g/dL). Of the 6 weanlings that tested positive for L intracellularis via PCR assay, 3 had maximum serum anti–L intracellularis antibody titers of 1:120 via IPMA and the remaining 3 had maximum antibody titers of ≥ 1:240. Four of 5 weanlings that had positive PCR assay results and were examined via abdominal ultrasonography had thickened small intestinal segments (small intestinal wall thickness > 3 mm). A total of 5 weanlings developed clinical signs of EPE, and 3 of these weanlings concurrently had positive PCR assay results. Of the 14 weanlings without clinical signs of EPE that received treatment (13 because of clinicopathologic abnormalities and 1 without a serum sample), none were determined to develop clinical EPE. This assessment was based on the lack of clinical signs as well as improvement in serum total protein and albumin concentrations after appropriate antimicrobial treatment.

Of the 5 weanlings with clinical signs of EPE at farm B, 1 weanling was euthanized after prolonged hospitalization because of complications directly attributable to EPE. This weanling had tested positive for L intracellularis via IPMA and PCR assay. Treatments included administration of multiple antimicrobial drugs, including oxytetracycline (6.6 mg/kg [3.0 mg/lb], IV, q 12 h), chloramphenicol (50.0 mg/kg, PO, q 8 h), clarithromycin (7.5 mg/kg [3.4 mg/lb] PO, q 12 h) plus rifampin (5.0 mg/kg [2.3 mg/lb], PO, q 12 h), enrofloxacin (7.5 mg/kg, PO, q 24 h), and enrofloxacin (7.5 mg/kg, PO, q 24 h) plus potassium-penicillin (20,000 U/kg [9,090.9 U/lb], IV, q 6 h). Intravenous colloid and crystalloid fluids (10 mL of hetastarch^a/kg, q 2 to 3 d; 5 mL of equine plasma/kg, q 2 to 3 d; and 4 to 6 mL of an isotonic electrolyte replacement fluid^e/kg [1.8 to 2.7 mL/lb], q 1 h) and other supportive care, including flunixin meglumine (1.1 mg/kg [0.5 mg/lb], IV, q 12 h) and butorphanol tartrate (0.01 to 0.02 mg/kg [0.0045 to 0.009 mg/lb], IV, q 4 to 6 h), were given. Despite this aggressive treatment, the weanling developed salmonellosis, in addition to severe, persistent gastric reflux; at necropsy, a perforated gastroduodenal ulcer, fibrinous peritonitis, and peritoneal adhesions were found.

One other weanling was referred for hospitalization and supportive care, including antimicrobial treatment (6.6 mg of oxytetracycline/kg diluted in 1 L of saline [0.9% NaCl] solution, IV, q 12 h) and IV colloid administration (10 mL of hetastarch^a/kg, q 2 to 3 d, and 5 mL of equine plasma/kg, q 2 to 3 d); a third weanling was treated for EPE at the farm with an antimicrobial drug (6.6 mg of oxytetracycline/kg, IV, q 12 h) and synthetic colloid fluid administration (10 mL of hetastarch^a/kg, IV, once).

Discussion

In recent years, the interest in EPE caused by L intracellularis has increased from clinical and research perspectives. This is especially true in the Thoroughbred industry; results of a 2008 study³ revealed that prior L intracellularis infection can have a substantial impact on yearling sales prices.

During the study period, 5 weanlings from farm B each had a diagnosis of EPE on the basis of clinical signs and positive antemortem diagnostic test results. Of these 5 weanlings, 2 were referred to a hospital for treatment because of severe dependent edema requiring colloidal support as well as a lack of response to treatment at the farm. In addition, a third weanling from farm B was given antimicrobials IV and 1 colloidal fluid treatment because of moderate dependent edema that was not severe enough to warrant hospitalization. There were fewer weanlings (n = 2) with severe EPE at farm B during the study period than during the same period (from August to January) in each of the 2 preceding years, when 5 weanlings required hospitalization or aggressive treatment at the farm for EPE, suggesting that this method of EPE detection warrants further examination with respect to the possibility of decreasing the occurrence of EPE-associated hospitalizations or aggressive on-site care.

The 2 most commonly used commercially available diagnostic tests for L intracellularis infection are an L intracellularis–specific PCR assay for detection of the pathogen shed in feces and IPMA for detection of anti–L intracellularis antibodies in serum. In another study,⁹ IPMA of porcine serum samples for anti–L intracellularis antibodies was found to have high sensitivity and specificity. The use of IPMA for detection of antibodies against L intracellularis in horses has yet to be validated, although the test is widely used in the swine industry and by equine practitioners that suspect EPE in patients. The L intracellularis–specific PCR assay has high specificity¹ but variable sensitivity. Several potential reasons for false-negative results have been reported,^4,20 including the presence of inhibitory components in feces and variable bacterial shedding. It has been suggested that the antemortem diagnosis of EPE ideally includes the combination of serum antibody detection via IPMA and a positive PCR assay result for a fecal sample, in addition to detection of clinical signs and clinicopathologic changes.³ In the study reported here, the most common clinical signs were anorexia (partial or complete), depression, fever, and dependent edema. None of the weanlings at farm B with a diagnosis of EPE had diarrhea or signs of colic at the time of diagnosis. A previous study³ confirmed that the most consistent clinicopathologic change detected in horses with EPE is hypoalbuminemia, although it is not uncommon for clinically affected horses to also have other metabolic derangements.¹⁸ Four of 5 weanlings from farm B with clinical signs of EPE had serum total protein and albumin concentrations that were considered moderately to severely low (total protein concentration < 5.4 g/dL and albumin concentration < 3.0 g/dL). These data, with those of other studies,^13,21 support the determination of serum total protein and albumin concentrations as a reliable presumptive ancillary test for identification of EPE when combined with the presence of characteristic clinical signs.

Among weanlings for which CBC, serum fibrinogen concentration, and serum biochemical analysis results were obtained (n = 31 weanlings), WBC counts and fibrinogen concentrations were within the reported reference intervals, as were remaining serum biochemical variables other than total protein and albumin concentrations. This is different from the results of other studies,^3,21,22 in which various derangements in these values were revealed. The most likely explanation for this is that only 5 weanlings in the present study were clinically affected with EPE, whereas in previous studies, only clinically affected horses in a referral hospital setting were evaluated. This finding suggests that, although useful for assessment of other infectious and inflammatory processes, the inclusion of CBCs and serum fibrinogen concentration determinations may not be necessary for EPE screening of horses without clinical signs of disease.

Other clinicopathologic abnormalities in horses with EPE, including hyponatremia,^13,14 hypocalcemia,¹⁴ azotemia,^14,21 and increased creatine kinase activity,¹³ have been reported. Arguably, these findings are non-specific and likely reflect secondary effects of EPE. Azotemia can be caused by dehydration or administration of a potentially nephrotoxic drug, such as tetracycline-related antimicrobials, to a dehydrated patient. Hyponatremia and hypocalcemia can develop in patients with prolonged anorexia, whereas creatine kinase activity may increase during a catabolic state. Findings of the present study, in addition to the potential influences of these factors, suggest that the use of these biochemical variables with respect to EPE diagnosis may not be as useful as determination of serum total protein and albumin concentrations.

On the basis of results of fecal L intracellularis–specific PCR assays, 6 weanlings at farm B were found to shed L intracellularis during the study, whereas none of the foals or weanlings at farm A shed detectable amounts of the pathogen in their feces at any time during the study period. Because the PCR assay cannot be used to determine whether L intracellularis is viable or nonviable, it is possible that the presence of L intracellularis in the feces was not indicative of an active infection. However, each of the weanlings with positive PCR assay results either concurrently or subsequently developed a strong serum antibody titer (≥ 1:120), which confirmed exposure to L intracellularis.

Detection of anti–L intracellularis IgG by use of IPMA in the present study revealed a 60.4% serop-revalence among weanlings at farm B, whereas those at farm A had a seroprevalence of 17.0%. Another serop-revalence study²⁰ demonstrated a wide range of positive results (33.8% to 45.5% seroprevalence) in horses at farms in which L intracellularis was endemic. To the authors' knowledge, the study reported here was the first to reveal a high seroprevalence rate (60.4% at farm B) for L intracellularis in horses with a large sample size. Other than natural exposure, possible reasons for detection of serum antibodies via IPMA include the presence of maternal antibodies or administration of a vaccine. None of the foals or weanlings in the present study was seropositive during the first month of the study, and several weanlings were seropositive the following month. These findings, in conjunction with the use of a titer ≥ 1:60 as the cutoff value for an assay result to be considered positive, indicate that the concentration of any maternal antibodies present was less than the detection threshold for the IPMA. Although there is an effective, commercially available swine vaccine for the prevention of porcine proliferative enteropathy caused by L intracellularis, this vaccine was not used on either farm during the study period.

It is the opinion of the authors that the use of IPMA to evaluate serum anti–L intracellularis antibody titers in horses is appropriate for the antemortem diagnosis of EPE when paired with compatible clinical signs as well as total protein and albumin concentrations. Furthermore, antibody titers above the assay cutoff value should be considered indicative of exposure, regardless of the titer's endpoint. This assertion seems to be supported by recent work in which experimental challenge with L intracellularis resulted in seroconversion of all exposed weanlings but no seroconversion was detected in unchallenged controls.²³ Thirty-two weanlings in the present study had serum anti–L intracellularis antibody titers ≥ 1:60. The 5 weanlings at farm B that had clinical signs of EPE had or soon developed serum anti–L intracellularis antibody titers ≥ 1:60; 4 of these weanlings each had a maximum antibody titer ≥ 1:240, and the remaining weanling had a maximum antibody titer of 1:120. On the basis of these findings, clinicians should exercise caution when reaching a diagnosis of EPE in horses with serum anti–L intracellularis antibody titers of 1:60, because treatment is not without potential complications, including the risk of antimicrobial-induced colitis and potential renal adverse effects of tetracycline-related drugs.²⁴ In instances such as this, submission of an additional serum sample for IPMA analysis 2 to 4 weeks after a positive test result is reported can aid in the diagnosis of EPE, especially in the event that the farm has never had a confirmed case.

Interestingly, the first serum anti–L intracellularis antibody titer ≥ 1:60 at farm A was detected in November, whereas the first antibody titer above the cutoff value on farm B was detected in September. Furthermore, the 2 weanlings from farm A that developed antibody titers of 1:240 did not do so until December or January, whereas the first antibody titer of 1:240 at farm B occurred in October. This might suggest that the burden for exposure at farm A was low, compared with that of farm B, or that the timeframe for exposure was different between the farms. However, as indicated by the sero-prevalence of 17.0% among weanlings at farm A, there remains the potential for exposure to L intracellularis and therefore cases of EPE, despite no prior history of the disease on a farm.

Of the 32 weanlings on farm B that had positive results for L intracellularis–specific IPMA, 19 were treated for EPE during the study period (no foals on farm A were treated for EPE during this period). The treatment of choice in the present study was a tetracycline-related drug because of the cellular penetration (a function of the lipophilic nature) and reported success of these drugs in the treatment of EPE.^25,26 Treatment of all but 3 weanlings involved only orally administered antimicrobials, typically doxycycline, with a few weanlings receiving chloramphenicol. Of these weanlings, all responded favorably and required 2 weeks or less of treatment to respond with increased serum total protein and albumin concentrations. The remaining 3 weanlings were among those that had clinical signs of EPE, and 2 were treated with oxytetracycline IV, whereas the third received several different antimicrobials. Macrolides were not used in the present study because of the risk of hyperthermia in combination with sunlight exposure as well as the concern for possible antimicrobial-induced colitis.²⁴

The months of August to January were chosen for the present study because of the authors' past experience with EPE in Kentucky as well as published reports^3,27 that indicated the highest occurrence of EPE was detected at this time of the year. As stated previously, the first serum samples that had positive results for L intracellularis–specific IPMA and the first fecal samples that had positive PCR assay results were obtained at farm B in mid-September 2008. The month with the greatest number of fecal samples that tested positive via PCR assay was October (5 samples), whereas the month with the greatest number of weanlings that tested positive for the first time via IPMA of serum samples was November. By combining these 2 findings, one could surmise that the largest exposure of weanlings on farm B occurred in late September or early October. However, it is impossible to determine from these findings whether EPE has a temporal component because the study population was comprised solely of Thoroughbreds. In the northern hemisphere, the Thoroughbred breeding season extends from mid-February to mid-June with the goal to have foals born as early in the following year as possible. As a result, there is an inherent flaw in the use of Thoroughbreds for a study in which time of year can be a predisposing factor because most or all of the horses will be approximately the same age during the same time of year.

In the present study, the only weanlings with serum concentrations of total protein and albumin categorized as moderately (total protein, 4.0 to 5.0 g/dL; albumin, 2.0 to 2.9 g/dL) or severely low (total protein < 4.0 g/dL; albumin < 2.0 g/dL) were those that had maximum antibody titers ≥ 1:240. This would suggest that weanlings with antibody titers ≥ 1:240 are more likely to have clinical signs of EPE. Care should be taken when interpreting these results, however, because the serum antibody titers described here are the highest values recorded for each weanling. In a clinical setting, evaluation of a single serum antibody titer via IPMA is insufficient to determine whether the titer is increasing or decreasing. Analysis of a second sample taken 2 to 4 weeks after the first sample is required for accurate interpretation.

A potential confounding factor associated with use of farm A as the farm in which L intracellularis was nonendemic was that once weaned, the weanlings were moved to another farm approximately 20 miles away from the foaling farm. Some other differences between farms A and B included presence of a river which divided farm B into 2 parts and greater numbers of trees and bushes on farm B than on farm A. Farm A had an active trapping program for wild animals larger than rodents. Although these variations exist between the 2 farms, the differences between seroprevalence and number of clinical cases on each farm make for an interesting comparison. Further research comparing potential risk factors between farms in which L intracellularis is endemic and nonendemic is warranted.

To the authors' knowledge, only 1 study²⁸ has been conducted in which investigators examined the survival time of the L intracellularis organism. That study revealed that, under simulated environmental conditions ranging from 5° to 15°C, L intracellularis isolated from acute porcine proliferative enteropathy lesions was capable of colonizing the intestine of pigs after up to 2 weeks in culture. As a result, it is possible that feces contaminated with L intracellularis could be tracked by horse farm personnel from one area of the farm to another, acting as a potential source of infection. With respect to the 2 farms of the present study, every barn in which foals and weanlings enrolled in the study were housed was staffed by the same personnel with minimal cross-traffic, except for the farm managers, assistant farm managers, and farm veterinarians. Based on past reports¹³ of outbreaks of EPE, it is possible that humans act as a source of L intracellularis spread on farms. This possibility should be examined further, including whether strains of L intracellularis isolated from horses have the same survival time in the environment as those isolated from swine.

Additional focus in several different areas of research would benefit our understanding of L intracellularis and EPE. First, with respect to the objectives of this study, more frequent antibody titer measurements via IPMA (every 2 weeks for the duration of the region's EPE season) would allow for more rapid detection of newly developing antibody titers. Second, a reliable challenge method for EPE needs to be created to further the study of this infection and disease process under controlled conditions. In a recent small-scale study,²³ EPE was experimentally induced in challenged horses, although additional, larger scale studies are needed to validate this method. Lastly, there needs to be a concerted effort to examine the epidemiology of L intracellularis and EPE, including potential risk factors.

The screening techniques described in the study reported here are not applicable to every farm but may be feasible on certain farms where economic and labor resources justify screening for EPE. Results of the present study revealed that serial detection of anti–L intracellularis antibodies by use of IPMA is useful when combined with careful clinical observation and monitoring of serum total protein and albumin concentrations for identification of subclinical EPE in Thoroughbred weanlings.