In horses, colic is an economically important disease that causes substantial morbidity and death. Approximately 50% of horses referred to tertiary veterinary hospitals for colic require emergency abdominal surgery.1 Early, accurate recognition of ischemic lesions is essential for expediting surgical intervention and thereby decreasing the incidence of complications and increasing the likelihood of patient survival.2 Substantial changes in the treatment and management of horses with small intestinal strangulating lesions have been made to improve patient survival. Results of a 1989 study3 indicate that the short-term survival rate for horses with small intestinal strangulating lesions was only 49%, whereas more recent studies4–6 indicate that the short-term survival rate for such horses ranges between 76% and 84%. Clinical and cytologic evaluation of peritoneal fluid obtained during preoperative assessment of horses with signs of colic and intestinal ischemia expedites diagnosis and treatment of horses with strangulating intestinal lesions.2,7 Results of another study8 indicate that the color and protein concentration of peritoneal fluid are the most useful variables for differentiating lesions that require medical versus surgical management. Peritoneal fluid with a serosanguineous appearance is strongly associated with the presence of a small intestinal strangulating lesion.8,9
One of the greatest advancements for the diagnosis of strangulating intestinal lesions in horses has been evaluation of both systemic and peritoneal fluid L-lactate (lactate) concentrations. In 1 study,2 the mean ± SD peritoneal fluid lactate concentration for horses with signs of colic (4.00 ± 4.63 mmol/L) was significantly greater than that for healthy control horses (0.6 ± 0.19 mmol/L). Moreover, the mean ± SD peritoneal fluid lactate concentration for horses with strangulating intestinal lesions (8.45 ± 5.52 mmol/L) was significantly greater than that for horses with nonstrangulating intestinal lesions (2.09 ± 2.09 mmol/L).2 Conversely, in another study,9 the blood and peritoneal fluid lactate concentrations did not differ between horses with strangulating and nonstrangulating small intestinal lesions, although the mean ratio of peritoneal fluid lactate concentration to blood lactate concentration for horses with strangulating small intestinal lesions was significantly greater than that for horses with nonstrangulating small intestinal lesions. Additionally, serial measurement of blood and peritoneal fluid lactate concentrations is a more sensitive method than 1-time measurement of those variables for identification of strangulating intestinal lesions, but it requires more time and typically delays surgical intervention.7
Procalcitonin is an expedient inflammatory biomarker. Results of multiple studies in the human medical literature suggest that procalcitonin is a pivotal and reliable early biomarker for many conditions such as infections,10 sepsis,10 SIRS,11,12 inflammatory bowel disease,13 and pancreatitis.14 Research also suggests that procalcitonin is a sensitive and specific biomarker for SIRS in horses.15–17
Procalcitonin is the peptide precursor of the hormone calcitonin. It is produced by the parafollicular cells of the thyroid gland and is secreted from neuroendocrine cells of the intestine, which makes it an interesting potential biomarker for intestinal ischemia.18 Circulating procalcitonin concentration is considered a useful biomarker for the diagnosis of intestinal ischemia in human patients.18–20 To our knowledge, measurement of peritoneal fluid procalcitonin concentration as an aid in disease diagnosis has not been evaluated in veterinary or human patients.
The purpose of the study reported here was to assess the diagnostic value of plasma and peritoneal fluid procalcitonin concentrations for identification of horses with strangulating intestinal lesions. We hypothesized that the plasma and peritoneal fluid procalcitonin concentrations for horses with strangulating intestinal lesions would be greater than those for healthy horses as well as horses with nonstrangulating lesions and that peritoneal fluid procalcitonin concentration would be a sensitive biomarker for distinguishing between horses with strangulating and nonstrangulating intestinal lesions.
Materials and Methods
Animals
The study involved 10 healthy university-owned horses (controls) and 65 client-owned horses with signs of colic of intestinal origin (cases) that were examined at the University of California-Davis William R. Pritchard Veterinary Medical Teaching Hospital from August 2016 to April 2017. The 10 control horses had no history of colic, were determined to be healthy on the basis of results of a physical examination, and underwent blood sample collection and abdominocentesis for teaching purposes. The 65 case horses underwent blood sample collection and abdominocentesis as part of the diagnostic workup for signs of colic with the owners' consent. Horses < 1 year old were excluded from the study. Also excluded from the study were horses with concomitant primary kidney or liver disease and horses with any 2 of the following 3 conditions: marked leukopenia (WBC count, < 2,500 WBCs/μL), diarrhea (2 episodes of defecation of feces with a liquid consistency that precluded removal with a shovel), and fever (rectal temperature, > 38.6°C [101.5°F]). All study procedures were reviewed and approved by the University of California-Davis School of Veterinary Medicine Institutional Animal Care and Use Committee.
Sample collection and processing
For each horse, blood (approx 6 mL) was collected by jugular venipuncture directly into an evacuated glass tube containing EDTA (EDTA tube) and another tube containing lithium heparin (heparin tube) as anticoagulants. Blood samples from horses with colic were obtained during the initial examination at the teaching hospital before any calcium was administered to avoid interference with measurement of plasma procalcitonin concentration. Peritoneal fluid (approx 3 to 5 mL) was aseptically obtained by abdominocentesis. Briefly, hair was clipped from a 5 × 5-cm area at the most ventral aspect of the abdomen just to the right of the midline. The skin was aseptically prepared in a routine manner, a stab incision was made in the middle of the clipped area, and a stainless steel teat cannula was inserted through the stab incision and into the abdominal cavity. Freely flowing peritoneal fluid was collected into an EDTA tube and a heparin tube for cytologic and biochemical analyses.
All sample analyses were performed by the Hematological and Biochemical Laboratory at the William R. Pritchard Veterinary Medical Teaching Hospital, and the assay protocols remained unchanged throughout the study. A CBC was performed on all EDTA-anticoagulated blood samples immediately after laboratory arrival (ie, within 1 minute after collection). Immediately after laboratory arrival, heparin-anticoagulated blood and peritoneal fluid samples were centrifuged at 4,400 × g for 2 minutes. Plasma or supernatant was harvested from each sample and analyzed for l-lactate (lactate) concentration by use of a benchtop blood gas analyzera at room temperature (20°C). Thus, the lactate concentration of each sample was determined within 5 minutes after collection. The remaining harvested plasma and peritoneal fluid from the heparin-anticoagulated samples were immediately frozen and stored at −80°C until analysis for procalcitonin concentration. The procalcitonin concentration of all plasma and peritoneal fluid samples was determined at the same time by use of a commercial ELISA kitb designed for horses as described.15
Data collection
For each horse, data extracted from the medical record included signalment (breed, age, and sex), body weight, and pertinent physical examination findings (eg, heart rate, respiratory rate, rectal temperature, and presence or absence of borborygmi). Additional data recorded for case horses included duration of colic signs (which was categorized as < 6 hours, 6 to 12 hours, or > 12 hours), whether nasogastric reflux (defined as the retrieval of > 2 L of reflux following nasogastric intubation) was present, all abdominal examination per rectum and abdominal ultrasonographic and radiographic findings, diagnosis (when available), whether the lesion was strangulating or nonstrangulating (as determined during surgery or necropsy), whether the horse underwent surgical or medical treatment, and outcome. Horses that responded to medical treatment and had peritoneal fluid with a normal gross appearance and total protein concentration were considered to have nonstrangulating lesions. Outcome was defined as the patient status at hospital discharge (alive or dead; ie, short-term survival).
Statistical analysis
Descriptive statistics were generated for the control and case horses. The respective data distributions for plasma and peritoneal fluid procalcitonin and lactate concentrations were assessed for skewness and kurtosis. Results indicated that the plasma procalcitonin concentration data were not normally distributed; thus, those data underwent a logarithmic transformation so that the distribution approximated normality prior to analysis. The Pearson correlation coefficient (r) was calculated to assess the strength of the association between plasma and peritoneal fluid procalcitonin concentrations for both groups of horses. t Tests were used to compare plasma and peritoneal fluid procalcitonin concentrations between case and control horses. Plasma and peritoneal fluid procalcitonin concentrations were compared among control horses and case horses with and without strangulating lesions by means of a 1-way ANOVA, followed by the Bonferroni test when post hoc pairwise comparisons were necessary. Select clinical, hematologic, and biochemical variables were compared between case horses with strangulating and nonstrangulating lesions by use of the t test or χ2 test as appropriate. Receiver operating characteristic analyses were performed to identify the optimal cutoff values for plasma and peritoneal fluid procalcitonin and lactate concentrations that maximized sensitivity and specificity for differentiating between case horses with strangulating and nonstrangulating lesions. For all ROC analyses, the concentration that yielded the highest value for k was selected as the optimal cutoff value. Positive and negative predictive values were calculated for each optimal cutoff value. All analyses were performed by use of a commercially available statistical software program,c and values of P < 0.05 were considered significant.
Results
Horses
Control horses consisted of 3 mares and 7 geldings with a mean ± SD age of 14.5 ± 3.4 years and body weight of 524.7 ± 65.9 kg (1,154.3 ± 145.0 lb). There were 8 Thoroughbreds, 1 Standardbred, and 1 warmblood. No abnormalities were identified during physical examination for any of the control horses.
The 65 case horses consisted of 23 (35.4%) mares, 37 (57.0%) geldings, and 5 (7.7%) stallions with a mean ± SD age of 13.1 ± 6.2 years and body weight of 505 ± 96.5 kg (1,111 ± 212.3 lb). There were 18 (27.7%) American Quarter Horses, 16 (24.6%) warmbloods, 12 (18.5%) Thoroughbreds, 6 (9.2%) Arabians, 2 (3%) American Paint Horses, 2 (3%) Icelandic Horses, 2 (3%) mustangs, 2 (3%) ponies, 2 (3%) Percherons, and 1 (1.5%) each of American Miniature Horse, Lipizzaner, and Friesian. The sex distribution and mean age and body weight did not differ between case and control horses.
Of the 65 case horses, 44 (67.7%) had nonstrangulating intestinal lesions (nonstrangulating lesions), and 21 (32.3%) had strangulating intestinal lesions (strangulating lesions). The sex distribution and mean ± SD age (P = 0.65), body weight (P = 0.55), rectal temperature at initial examination (P = 0.77), duration of colic (P = 0.59), total nucleated cell count (P = 0.17), and PCV (P = 0.20) did not differ significantly between case horses with strangulating and nonstrangulating lesions (Table 1). Horses with strangulating lesions had a significantly greater heart rate (P = 0.047), plasma lactate concentration (P = 0.009), and peritoneal fluid lactate concentration (P < 0.001), compared with horses with nonstrangulating lesions.
Select clinicopathologic variables at the time of initial examination for 65 horses with signs of colic of intestinal origin that were examined at a veterinary teaching hospital from August 2016 to April 2017 and were subsequently determined to have nonstrangulating (n = 44) or strangulating (21) intestinal lesions.
| Variable | Horses with nonstrangulating lesions | Horses with strangulating lesions | P value* |
|---|---|---|---|
| Age (y) | 13.4 ± 5.5 | 12.6 ± 7.7 | 0.65 |
| Body weight (kg) | 507 ± 108.7 | 49 ± 60 | 0.55 |
| Heart rate (beats/min) | 52 ± 17.3 | 62.1 ± 16.9 | 0.047 |
| Respiratory rate (breaths/min) | 20.7 ± 7.4 | 27.7 ± 11.5 | 0.005 |
| Rectal temperature (°C) | 37.5 ± 0.7 | 37.5 ± 0.7 | 0.77 |
| Duration of colic (h) | 11.8 ± 7.3 | 10.8 ± 6.5 | 0.59 |
| Total nucleated cell count (cells/mL) | 7,584 ± 2,651 | 8,544 ± 2,353 | 0.17 |
| PCV (%) | 38.9 ± 8.8 | 42.1 ± 10.5 | 0.2 |
| Plasma lactate (mmol/L) | 2.2 ± 2.8 | 4.7 ± 4.6 | 0.009 |
| Plasma procalcitonin (pg/mL) | 257.8 ± 130.6 | 310.8 ± 184.7 | 0.491 |
| Peritoneal fluid lactate (mmol/L) | 2.8 ± 3.7 | 8 ± 5.5 | < 0.001 |
| Peritoneal fluid procalcitonin (pg/mL) | 262.3 ± 42.6 | 308.7 ± 52.5 | 0.001 |
| Peritoneal fluid total protein (g/dL) | 1.5 (0.2–4.4) | 3.7 (1.2–5.8) | < 0.001 |
Values represent mean ± SD or median (range) unless otherwise indicated.
Values of P < 0.05 were considered significant.
Procalcitonin concentrations
The mean ± SD plasma (175.5 ± 46 pg/mL) and peritoneal fluid (218.8 ± 48.7 pg/mL) procalcitonin concentrations for the control horses were significantly lower than the mean ± SD plasma (274.9 ± 150.8 pg/mL) and peritoneal fluid (277 ± 50.6 pg/mL) procalcitonin concentrations for the case horses. The mean ± SD plasma procalcitonin concentration for horses with strangulating lesions (310.8 ± 184.7 pg/mL) did not differ significantly from that for horses with nonstrangulating lesions (257.8 ± 130.6 pg/mL). However, the mean ± SD peritoneal fluid procalcitonin concentration differed significantly (P = 0.001) between horses with strangulating lesions (308.7 ± 52.5 pg/mL) and horses with nonstrangulating lesions (262.3 ± 42.6 pg/mL). There was a significant (P = 0.001) and moderate positive correlation (r = 0.557) between plasma and peritoneal fluid procalcitonin concentrations for all horses.
Patient outcome
Eleven of the 44 (25%) horses with nonstrangulating lesions underwent surgery. Diagnoses made during surgery included large colon displacement (n = 5), large colon impaction (2), idiopathic focal eosinophilic enteritis (1), jejunal impaction (1), and colitis (1). No clinically relevant abnormalities were observed during surgery for 1 horse. Surgery was indicated and recommended for an additional horse, but surgery was declined by the owner. That horse was euthanized, and an enterolith was discovered in the small colon during necropsy. The horse with colitis was euthanized during surgery, and bacteriologic culture of the colonic contents yielded growth of Clostridium perfringens type A. One horse with a nonstrangulating lesion was euthanized because it sustained an extensive muscle tear in a hind limb prior to arrival at the veterinary teaching hospital. The remaining 31 horses with nonstrangulating lesions were medically managed and survived to hospital discharge.
Three of 4 horses with enteritis, colitis, or small intestinal impactions underwent surgery. The mean ± SD plasma (237.6 ± 74.5 pg/mL) and peritoneal fluid (241.0 ± 59.5 pg/mL) procalcitonin concentrations for those 3 horses were lower than the mean plasma and peritoneal fluid procalcitonin concentrations for the other horses with nonstrangulating lesions.
Of the 21 horses with strangulating lesions, 4 had a large colon volvulus and 17 had strangulating lesions that involved the small intestine, including a lipoma (n = 6), epiploic foramen entrapment (4), gastrosplenic ligament entrapment (2), inguinal hernia (1), jejunal volvulus (1), Meckel diverticulum (1), and adhesions (1). The cause of small intestinal strangulation could not be definitively determined during necropsy for 1 horse. Ten horses with strangulating lesions were euthanized without undergoing surgery owing to a poor prognosis or financial constraints. Of the 11 horses that underwent surgery, 5 underwent a resection and anastomosis to remove the strangulated portion of the intestine. One horse was euthanized during surgery because an extensive proportion of its ileum was devitalized, and the horse required a jejunocecostomy, which the owner declined. The remaining 5 horses did not require intestinal resection. Nine of the 10 horses that recovered from surgery survived to hospital discharge. One horse that had approximately 90% of the small intestine entrapped within the gastrosplenic ligament was euthanized after surgery because of ileus and failure of the small
intestine to regain function. The remaining horses that underwent reduction of a small intestinal entrapment and did not require intestinal resection of the affected segment survived to hospital discharge.
Diagnostic performance of procalcitonin and lactate concentrations
The mean plasma and peritoneal fluid procalcitonin and lactate concentrations for horses with strangulating and nonstrangulating lesions were used as cutoffs to determine the sensitivity and specificity for detection of ischemic intestinal lesions (Table 2). The ROC curves for plasma and peritoneal fluid procalcitonin concentrations for distinguishing ischemic intestinal lesions were plotted (Figure 1). The AUC for plasma procalcitonin concentration was 0.610 (95% CI, 0.461 to 0.759), and the AUC for peritoneal fluid procalcitonin concentration was 0.783 (95% CI, 0.654 to 0.913). For prediction of ischemic intestinal lesions, a plasma procalcitonin concentration cutoff of 231.7 pg/mL had a diagnostic sensitivity of 72% and specificity of 51%, with a positive predictive value of 44% and a negative predictive value of 78%, whereas a peritoneal fluid procalcitonin concentration cutoff of 281.7 pg/mL had a diagnostic sensitivity of 81% and specificity of 69%, with a positive predictive value of 57% and a negative predictive value of 88%.
Receiver operating characteristic curves that depict the discriminatory ability of plasma (solid line) and peritoneal fluid (dashed line) procalcitonin concentrations for identification of horses with ischemic or strangulating intestinal lesions. The ROC analysis was performed by the use of plasma and peritoneal fluid procalcitonin concentration data for 65 horses with signs of colic of intestinal origin that were examined at a veterinary teaching hospital from August 2016 to April 2017 and were subsequently determined to have nonstrangulating (n = 44) or strangulating (21) intestinal lesions. The AUC for plasma procalcitonin concentration was 0.610 (95% CI, 0.461 to 0.759), and a plasma procalcitonin concentration cutoff of 231.7 pg/mL had a sensitivity of 72%, specificity of 51%, positive predictive value of 44.4%, and negative predictive value of 78% for detection of intestinal ischemia. The AUC for peritoneal fluid procalcitonin concentration was 0.783 (95% CI, 0.654 to 0.913), and a peritoneal fluid procalcitonin concentration cutoff of 281.7 pg/mL had a sensitivity of 81%, specificity of 69%, positive predictive value of 56.7%, and negative predictive value of 87.9% for detection of intestinal ischemia. The identity (solid diagonal) line, which is indicative of no discriminatory ability, is also plotted.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.927
Receiver operating characteristic curves that depict the discriminatory ability of plasma (solid line) and peritoneal fluid (dashed line) procalcitonin concentrations for identification of horses with ischemic or strangulating intestinal lesions. The ROC analysis was performed by the use of plasma and peritoneal fluid procalcitonin concentration data for 65 horses with signs of colic of intestinal origin that were examined at a veterinary teaching hospital from August 2016 to April 2017 and were subsequently determined to have nonstrangulating (n = 44) or strangulating (21) intestinal lesions. The AUC for plasma procalcitonin concentration was 0.610 (95% CI, 0.461 to 0.759), and a plasma procalcitonin concentration cutoff of 231.7 pg/mL had a sensitivity of 72%, specificity of 51%, positive predictive value of 44.4%, and negative predictive value of 78% for detection of intestinal ischemia. The AUC for peritoneal fluid procalcitonin concentration was 0.783 (95% CI, 0.654 to 0.913), and a peritoneal fluid procalcitonin concentration cutoff of 281.7 pg/mL had a sensitivity of 81%, specificity of 69%, positive predictive value of 56.7%, and negative predictive value of 87.9% for detection of intestinal ischemia. The identity (solid diagonal) line, which is indicative of no discriminatory ability, is also plotted.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.927
Receiver operating characteristic curves that depict the discriminatory ability of plasma (solid line) and peritoneal fluid (dashed line) procalcitonin concentrations for identification of horses with ischemic or strangulating intestinal lesions. The ROC analysis was performed by the use of plasma and peritoneal fluid procalcitonin concentration data for 65 horses with signs of colic of intestinal origin that were examined at a veterinary teaching hospital from August 2016 to April 2017 and were subsequently determined to have nonstrangulating (n = 44) or strangulating (21) intestinal lesions. The AUC for plasma procalcitonin concentration was 0.610 (95% CI, 0.461 to 0.759), and a plasma procalcitonin concentration cutoff of 231.7 pg/mL had a sensitivity of 72%, specificity of 51%, positive predictive value of 44.4%, and negative predictive value of 78% for detection of intestinal ischemia. The AUC for peritoneal fluid procalcitonin concentration was 0.783 (95% CI, 0.654 to 0.913), and a peritoneal fluid procalcitonin concentration cutoff of 281.7 pg/mL had a sensitivity of 81%, specificity of 69%, positive predictive value of 56.7%, and negative predictive value of 87.9% for detection of intestinal ischemia. The identity (solid diagonal) line, which is indicative of no discriminatory ability, is also plotted.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.927
Diagnostic sensitivity and specificity of select plasma and peritoneal fluid procalcitonin and lactate concentration cutoffs calculated for detection of horses with intestinal ischemia.
| Variable | Intestinal lesion | Concentration cutoff | Sensitivity (%) | Specificity (%) |
|---|---|---|---|---|
| Plasma procalcitonin (pg/mL) | Nonstrangulating | 257.8 | 52 | 67 |
| Strangulating | 310.8 | 29 | 83 | |
| Plasma lactate (mmol/L) | Nonstrangulating | 2.2 | 67 | 74 |
| Strangulating | 4.7 | 33 | 95 | |
| Peritoneal fluid procalcitonin (pg/mL) | Nonstrangulating | 262.3 | 81 | 48 |
| Strangulating | 308.7 | 52 | 95 | |
| Peritoneal fluid lactate (mmol/L) | Nonstrangulating | 2.8 | 81 | 81 |
| Strangulating | 8 | 38 | 95 |
The concentration cutoffs represent the mean values calculated for each variable for the horses with nonstrangulating and strangulating intestinal lesions described in Table 1. For each variable, sensitivity and specificity were calculated assuming a concentration greater than or equal to the given cutoff.
The ROC curves for plasma and peritoneal fluid lactate concentrations for distinguishing ischemic intestinal lesions were likewise plotted (Figure 2). The AUC for plasma lactate concentration was 0.780 (95% CI, 0.661 to 0.907), and the AUC for peritoneal fluid lactate concentration was 0.910 (95% CI, 0.037 to 0.982). For prediction of ischemic intestinal lesions, a plasma lactate concentration cutoff of 2 mmol/L had a diagnostic sensitivity of 71% and specificity of 71%, with a positive predictive value of 56% and a negative predictive value of 83%, whereas a peritoneal fluid lactate concentration cutoff of 3.7 mmol/L had a diagnostic sensitivity of 81% and specificity of 88%, with a positive predictive value of 77% and a negative predictive value of 90%.
Receiver operating characteristic curves that depict the discriminatory ability of plasma (solid line) and peritoneal fluid (dashed line) lactate concentrations for identification of horses with ischemic or strangulating intestinal lesions. The ROC analysis was performed by the use of plasma and peritoneal fluid lactate concentration data for the 65 horses described in Figure 1. The AUC for the plasma lactate concentration curve was 0.780 (95% CI, 0.661 to 0.907), and a plasma lactate concentration cutoff of 2 mmol/L had a sensitivity of 71.4%, specificity of 71.4%, positive predictive value of 56%, and negative predictive value of 83.3% for detection of intestinal ischemia. The AUC for the peritoneal fluid lactate concentration value was 0.910 (95% CI, 0.037 to 0.982), and a peritoneal fluid lactate concentration cutoff of 3.7 mmol/L had a sensitivity of 81%, specificity of 88%, positive predictive value of 77.3%, and negative predictive value of 90% for detection of intestinal ischemia. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.927
Receiver operating characteristic curves that depict the discriminatory ability of plasma (solid line) and peritoneal fluid (dashed line) lactate concentrations for identification of horses with ischemic or strangulating intestinal lesions. The ROC analysis was performed by the use of plasma and peritoneal fluid lactate concentration data for the 65 horses described in Figure 1. The AUC for the plasma lactate concentration curve was 0.780 (95% CI, 0.661 to 0.907), and a plasma lactate concentration cutoff of 2 mmol/L had a sensitivity of 71.4%, specificity of 71.4%, positive predictive value of 56%, and negative predictive value of 83.3% for detection of intestinal ischemia. The AUC for the peritoneal fluid lactate concentration value was 0.910 (95% CI, 0.037 to 0.982), and a peritoneal fluid lactate concentration cutoff of 3.7 mmol/L had a sensitivity of 81%, specificity of 88%, positive predictive value of 77.3%, and negative predictive value of 90% for detection of intestinal ischemia. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.927
Receiver operating characteristic curves that depict the discriminatory ability of plasma (solid line) and peritoneal fluid (dashed line) lactate concentrations for identification of horses with ischemic or strangulating intestinal lesions. The ROC analysis was performed by the use of plasma and peritoneal fluid lactate concentration data for the 65 horses described in Figure 1. The AUC for the plasma lactate concentration curve was 0.780 (95% CI, 0.661 to 0.907), and a plasma lactate concentration cutoff of 2 mmol/L had a sensitivity of 71.4%, specificity of 71.4%, positive predictive value of 56%, and negative predictive value of 83.3% for detection of intestinal ischemia. The AUC for the peritoneal fluid lactate concentration value was 0.910 (95% CI, 0.037 to 0.982), and a peritoneal fluid lactate concentration cutoff of 3.7 mmol/L had a sensitivity of 81%, specificity of 88%, positive predictive value of 77.3%, and negative predictive value of 90% for detection of intestinal ischemia. See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 256, 8; 10.2460/javma.256.8.927
Discussion
To our knowledge, the study reported here was the first to evaluate measurement of procalcitonin concentration in the peritoneal fluid of horses. Results of the present study indicated that procalcitonin is detectable in the peritoneal fluid of horses with signs of colic. Findings of the present study also suggested that peritoneal fluid procalcitonin concentration, when evaluated in conjunction with other clinicopathologic results, might be a sensitive indicator for ischemic intestinal lesions and facilitate the early identification of horses that require surgery to address the intestinal problem.
When the diagnostic performance of plasma and peritoneal fluid procalcitonin concentrations was compared with that for plasma and peritoneal fluid lactate concentrations, the sensitivity was similar but the specificity was lower for detection of ischemic intestinal lesions in horses. The lower specificity of plasma and peritoneal fluid procalcitonin concentrations relative to plasma and peritoneal fluid lactate concentrations for detection of ischemic intestinal lesions indicated that use of procalcitonin concentrations alone to predict the presence of a strangulating lesion could potentially lead to a false-positive diagnosis. Thus, it is important that procalcitonin concentrations be interpreted in conjunction with other clinicopathologic data.
An overall indication of the diagnostic accuracy of a binary classifier system (eg, calcitonin or lactate concentration for detection of intestinal ischemia) is the AUC. In the present study, the AUCs of the ROC curves for plasma and peritoneal fluid lactate concentrations were greater than the corresponding AUCs of the ROC curves for plasma and peritoneal fluid procalcitonin concentrations. This suggested that plasma and peritoneal fluid lactate concentrations were more accurate indicators of intestinal ischemia in horses than plasma and peritoneal fluid procalcitonin concentrations.
Results of other studies15,21 indicate that the plasma procalcitonin concentration in horses with SIRS is significantly greater than that in healthy horses. In one of those studies,15 the mean plasma procalcitonin concentration of healthy horses was significantly lower than that for horses with colic but did not differ significantly between horses with strangulating and nonstrangulating intestinal lesions, which is similar to results of the present study. However, of the 48 horses with SIRS evaluated in that study,15 only 18 and 13 had strangulating and nonstrangulating lesions, respectively, and the low number of horses with intestinal lesions might have impaired the investigators' ability to detect a significant difference in the plasma procalcitonin concentration between the 2 groups.
The human medical literature suggests that procalcitonin might be a useful biomarker for intestinal ischemia, with a reported diagnostic sensitivity that ranges between 72% and 100%, specificity that ranges between 68% and 91%, positive predictive value that ranges between 27% and 90%, and negative predictive value that ranges between 81% and 100%.18–20,22–25
Procalcitonin is the peptide precursor of the hormone calcitonin. It is produced by the parafollicular cells of the thyroid gland and is secreted from the neuroendocrine cells of the lungs and intestine. The clinical usefulness of procalcitonin as a biomarker for disease is dependent on its secretion from the neuroendocrine cells of the lungs and intestine, which increases in response to a proinflammatory stimulus, particularly when that stimulus is of bacterial origin.26 Injection of lipopolysaccharide into healthy horses causes a significant increase in plasma procalcitonin concentration. Likewise, in humans, blood procalcitonin concentrations increase > 400 times from baseline concentrations when endotoxin is experimentally injected IV into healthy subjects.27 Moreover, the human medical literature indicates that the plasma procalcitonin concentration response to inflammation does not differ significantly between subjects that have and have not undergone a thyroidectomy,28 which suggests that the inflammation-induced increase in procalcitonin concentration is not dependent on its production by the thyroid gland.
We theorized that procalcitonin concentrations in the peritoneal cavity would increase in response to local ischemia and that peritoneal fluid procalcitonin concentration would be a more sensitive indicator of intestinal ischemia than plasma procalcitonin concentration in a manner similar to that observed for lactate. Other studies2,7 have documented a significant disparity between plasma and peritoneal fluid lactate concentrations in horses with colic. During the initial stages of intestinal ischemia, peritoneal fluid lactate concentration is similar to or lower than plasma lactate concentration. However, as ischemia progresses, peritoneal fluid lactate concentrations increase. That increase is rapid in hemodynamically stable patients, which exacerbates the disparity between plasma and peritoneal fluid lactate concentrations likely owing to the large surface area of the peritoneal cavity.2,7 In the present study, although plasma procalcitonin concentrations did not differ significantly between horses with strangulating and nonstrangulating lesions, peritoneal fluid procalcitonin levels did differ significantly, and results of the ROC analysis indicated that peritoneal fluid procalcitonin concentration was a more sensitive indicator of intestinal ischemia than was plasma procalcitonin concentration.
In the present study, heart rate, peritoneal fluid total protein concentration, and plasma and peritoneal fluid lactate concentrations differed significantly between horses with strangulating and nonstrangulating lesions. Those findings were consistent with results of other studies2,9,29 and reflected differences in the extent of intestinal damage between the 2 groups.
Clinicopathologic findings for horses with nonstrangulating lesions can be indistinguishable from those for horses with strangulating lesions, which can make determining the need for surgical intervention difficult. For example, in the present study, 3 of the 4 horses with enteritis, colitis, or small intestinal impactions underwent surgery, and the mean plasma and peritoneal fluid procalcitonin concentrations for those 3 horses were significantly lower than those for the other horses with nonstrangulating lesions. It seems reasonable to speculate that any type of inflammation would upregulate secretion of procalcitonin from the neuroendocrine cells of the intestine. Although results of the present study suggested that peritoneal fluid procalcitonin concentration was not clinically useful for differentiating horses with colic that did and did not require surgical intervention, further research involving a larger number of horses with colitis, enteritis, and peritonitis than evaluated in the present study is warranted to fully elucidate the effects of those diseases on peritoneal fluid procalcitonin concentration.
A major limitation of the present study was the small study population. Evaluation of a larger number of horses with strangulating and nonstrangulating lesions would have increased the power of the study, particularly in regard to assessing the discriminatory ability of plasma procalcitonin concentrations.
Results of the present study suggested that, in horses, intestinal ischemia and the consequent systemic effects can lead to a significant increase in peritoneal fluid procalcitonin concentration. Thus, evaluation of peritoneal fluid procalcitonin concentration in conjunction with other clinicopathologic data may facilitate early identification of horses with strangulating intestinal lesions and thereby expedite surgical intervention and decrease morbidity and risk of death. However, further research is necessary to validate peritoneal fluid procalcitonin concentration for detection of intestinal ischemia before its use is recommended in clinical settings.
Acknowledgments
Supported by the Center for Equine Health, University of California-Davis, with funds provided by contributions from private donors.
ABBREVIATIONS
| AUC | Area under the curve |
| CI | Confidence interval |
| ROC | Receiver operating characteristic |
| SIRS | Systemic inflammatory response syndrome |
Footnotes
ABL 705 blood gas analyzer, Radiometer America Inc, Westlake, Ohio.
Horse Procalcitonin ELISA kit, MyBioSource Inc, San Diego, Calif.
SPSS, version 10.0, SPSS Inc, Chicago, Ill.
References
1. Hardy J. Design and organization of an equine intensive care unit. Vet Clin North Am Equine Pract 2004;20:1–10.
2. Latson KM, Nieto JE, Beldomenico PM, et al. Evaluation of peritoneal fluid lactate as a marker of intestinal ischaemia in equine colic. Equine Vet J 2005;37:342–346.
3. MacDonald MH, Pascoe JR, Stover SM, et al. Survival after small intestine resection and anastomosis in horses. Vet Surg 1989;18:415–423.
4. Espinosa P, Le Jeune SS, Cenani A, et al. Investigation of perioperative and anesthetic variables affecting short-term survival of horses with small intestinal strangulating lesions. Vet Surg 2017;46:345–353.
5. Freeman DE, Hammock P, Baker GJ, et al. Short- and long-term survival and prevalence of postoperative ileus after small intestinal surgery in the horse. Equine Vet J Suppl 2000;32:42–51.
6. Kilcoyne I, Dechant JE, Nieto JE. Comparison of clinical findings and short-term survival between horses with intestinal entrapment in the gastrosplenic ligament and horses with intestinal entrapment in the epiploic foramen. J Am Vet Med Assoc 2016;249:660–667.
7. Peloso JG, Cohen ND. Use of serial measurements of peritoneal fluid lactate concentration to identify strangulating intestinal lesions in referred horses with signs of colic. J Am Vet Med Assoc 2012;240:1208–1217.
8. Matthews S, Dart AJ, Reid SW, et al. Predictive values, sensitivity and specificity of abdominal fluid variables in determining the need for surgery in horses with an acute abdominal crisis. Aust Vet J 2002;80:132–136.
9. Shearer TR, Norby B, Carr EA. Peritoneal fluid lactate evaluation in horses with nonstrangulating versus strangulating small intestinal disease. J Equine Vet Sci 2018;61:18–21.
10. Mokart D, Merlin M, Sannini A, et al. Procalcitonin, interleukin 6 and systemic inflammatory response syndrome (SIRS): early markers of postoperative sepsis after major surgery. Br J Anaesth 2005;94:767–773.
11. Bell K, Wattie M, Byth K, et al. Procalcitonin: a marker of bacteraemia in SIRS. Anaesth Intensive Care 2003;31:629–636.
12. Boeken U, Feindt P, Petzold T, et al. Diagnostic value of procalcitonin: the influence of cardiopulmonary bypass, aprotinin, SIRS, and sepsis. Thorac Cardiovasc Surg 1998;46:348–351.
13. Herrlinger KR, Dittmann R, Weitz G, et al. Serum procalcitonin differentiates inflammatory bowel disease and self-limited colitis. Inflamm Bowel Dis 2004;10:229–233.
14. Rau BM, Kemppainen EA, Gumbs AA, et al. Early assessment of pancreatic infections and overall prognosis in severe acute pancreatitis by procalcitonin (PCT): a prospective international multicenter study. Ann Surg 2007;245:745–754.
15. Bonelli F, Meucci V, Divers TJ, et al. Plasma procalcitonin concentration in healthy horses and horses affected by systemic inflammatory response syndrome. J Vet Intern Med 2015;29:1689–1691.
16. Bonelli F, Meucci V, Divers T, et al. Evaluation of plasma procalcitonin concentrations in healthy foals and foals affected by septic systemic inflammatory response syndrome. J Equine Vet Sci 2015;35:645–649.
17. Bonelli F, Meucci V, Divers TJ, et al. Kinetics of plasma procalcitonin, soluble CD14, CCL2 and IL-10 after a sublethal infusion of lipopolysaccharide in horses. Vet Immunol Immunopathol 2017;184:29–35.
18. Cosse C, Sabbagh C, Kamel S, et al. Procalcitonin and intestinal ischemia: a review of the literature. World J Gastroenterol 2014;20:17773–17778.
19. Cosse C, Sabbagh C, Browet F, et al. Serum value of procalcitonin as a marker of intestinal damages: type, extension, and prognosis. Surg Endosc 2015;29:3132–3139.
20. Markogiannakis H, Memos N, Messaris E, et al. Predictive value of procalcitonin for bowel ischemia and necrosis in bowel obstruction. Surgery 2011;149:394–403.
21. Rieger M, Kochleus C, Teschner D, et al. A new ELISA for the quantification of equine procalcitonin in plasma as potential inflammation biomarker in horses. Anal Bioanal Chem 2014;406:5507–5512.
22. Cosse C, Regimbeau JM, Fuks D, et al. Serum procalcitonin for predicting the failure of conservative management and the need for bowel resection in patients with small bowel obstruction. J Am Coll Surg 2013;216:997–1004.
23. Ayten R, Dogru O, Camci C, et al. Predictive value of procalcitonin for the diagnosis of bowel strangulation. World J Surg 2005;29:187–189.
24. Nagata J, Kobayashi M, Nishikimi N, et al. Serum procalcitonin (PCT) as a negative screening test for colonic ischemia after open abdominal aortic surgery. Eur J Vasc Endovasc Surg 2008;35:694–697.
25. Karabulut K, Gül M, Dündar ZD, et al. Diagnostic and prognostic value of procalcitonin and phosphorus in acute mesenteric ischemia. Ulus Travma Acil Cerrahi Derg 2011;17:193–198.
26. Maruna P, Nedelniková K, Gürlich R. Physiology and genetics of procalcitonin. Physiol Res 2000;49(suppl 1):S57–S61.
27. Dandona P, Nix D, Wilson MF, et al. Procalcitonin increase after endotoxin injection in normal subjects. J Clin Endocrinol Metab 1994;79:1605–1608.
28. Russwurm S, Wiederhold M, Oberhoffer M, et al. Molecular aspects and natural source of procalcitonin. Clin Chem Lab Med 1999;37:789–797.
29. Nieto JE, Aldridge BM, Beldomenico PM, et al. Characterization of equine intestinal fatty acid binding protein and its use in managing horses with colic. Am J Vet Res 2005;66:223–232.
