The APR is a well-documented nonspecific phenomenon that rapidly occurs in the body and is incited by a variety of stimuli that may be related to infection, trauma, neoplasia, inflammation, or stress.1–4 The response is mediated by a plethora of reactants produced primarily by the liver; those reactants respond in either a negative or positive manner to the inflammatory stimulus. In response to inflammation, the plasma concentrations of some APPs decrease (negative APPs), whereas the plasma concentrations of other APPs increase (positive APPs). Negative APPs include albumin and transferrin. Positive APPs include haptoglobin, C-reactive protein, ceruloplasmin, fibrinogen, and SAA. The positive APPs are further classified as major, moderate, and minor APPs. Major APPs maintain very low plasma concentrations in healthy mammals; however, there may be as much as a 1,000-fold increase in those concentrations in response to inflammation. Moderate and minor APPs are typically present in the plasma of healthy individuals; for these APPs, there may be only a 1- to 10-fold increase in concentration during an APR.
To date, SAA has been the major APP identified in horses.5,6 Serum amyloid A has garnered attention as a reliable indicator of inflammation or infection across species.1,4,7 The fact that SAA concentration undergoes up to a 1,000-fold increase in response to inflammatory stimuli, in addition to SAA's early response and increase in concentration when stimulated by the inflammatory cascade of cytokines, has substantiated the assessment of SAA concentration as a means of monitoring patients with infectious and inflammatory diseases. In addition, given the short half-life of SAA, a more rapid decrease in concentration is observed in response to treatment and resolution of the disease process, relative to treatment-associated changes in circulating concentrations of other markers of inflammation.7
Previous investigations8–12 of SAA in horses have been limited in sample size or type of natural or experimental stimuli. The primary objective of the study reported here was to compare SAA concentration, plasma fibrinogen concentration, total WBC count, and A:G ratio in CN and CA horses. By using patients admitted to a specialty equine hospital, we intended to include horses with diverse clinical signs, times of disease onset, breeds, and ages. We hypothesized that SAA concentration would be a more reliable indicator of the inflammatory process, compared with other clinicopathologic variables (eg, plasma fibrinogen concentration, total WBC count, and A:G ratio) that are considered more traditional markers of inflammation. Furthermore, in addition to its use as a primary tool in the initial evaluation of equine patients, we hypothesized that SAA concentration would also function as a strong prognostic indicator for CA horses.
Materials and Methods
Horses—The study included 212 horses that were admitted to a specialty veterinary practice. Among the horses, there were > 23 breeds; the predominant breeds were Standardbred (n = 64), Thoroughbred (39), warm-blood (21), Quarter Horse (21), and Paso Fino (19). There were 93 male horses and 119 female horses. The horses' ages ranged from 4 months to 26 years.
The horses were classified as either CN (n = 111) or CA (101). The CA group included severely ill patients that were admitted to the hospital because of a variety of infectious or inflammatory conditions; many of these horses had a history of fever or nasal discharge. Among the CA horses, clinical signs or diagnoses included bacterial and viral pneumonia, bacterial cholangiohepatitis, Streptococcus equi subsp equi infection, meningitis, enterocolitis, various forms of colic and neoplasia, abscesses (pulmonary and musculoskeletal), septic tenosynovitis, and retropharyngeal lymphadenopathy. The CN group included horses that were not considered ill on the basis of history and results of physical examination. These horses were healthy mares admitted to the hospital with sick foals or patients admitted for treatment of noninfectious and noninflammatory processes such as osteochondrotic lesions.
Sample acquisition—A blood sample (20 mL) was collected from each horse within a 3-hour period after admission to the hospital. Ten milliliters of blood was placed into a tube containing EDTA as well as into a tube without anticoagulant.a The latter was allowed to clot for 20 minutes and was centrifuged at 3,400 × g for 10 minutes. Serum was separated to an inert transport tube. The whole blood and serum samples were stored under refrigeration and then shipped on cold packs to the University of Miami Acute Phase Protein Laboratory. A CBC, serum protein electrophoresis, and assessment of SAA concentration were performed for samples from all 212 horses. Determination of plasma fibrinogen concentration was performed at the discretion of the clinician at time of each horse's hospital admission; thus, samples from only 127 horses were analyzed for plasma fibrinogen concentration at the specialty veterinary practice. All assays were conducted within 24 hours after blood sample collection. For a small subset of CA horses (n = 23), additional blood samples were collected and similarly analyzed at various times during treatment to illustrate the pattern of change in SAA concentration as a result of successful treatment or resolution of disease processes. The duration of the hospitalization ranged from 1 to 20 days and was dictated by duration of illness and time at the hospital.
All CN horses were monitored (including measurement of rectal temperature) for a minimum of 7 days after last sample collection (either at the hospital or at their barns via follow-up calls) to ensure that they remained CN. The CA horses were similarly monitored for a minimum of 7 days after discharge from the hospital.
CBC—For each horse, the sample of EDTA-anticoagulated blood underwent a CBC via an automated analyzer.b A manual review of the blood smear was also conducted. On the basis of in-laboratory-derived reference intervals, total WBC counts ≤ 12,500 WBCs/μL were considered abnormally high.
Determination of plasma fibrinogen concentration—For 127 of the 212 horses, the sample of EDTA-anticoagulated blood was also used for determination of plasma fibrinogen concentration via the heat precipitation method13 at the specialty veterinary practice. On the basis of in-clinic-derived reference intervals, plasma fibrinogen concentrations ≤ 400 mg/dL were considered abnormally high.
Serum protein electrophoresis—Serum samples were analyzed with split beta gels according to the manufacturer's specifications.c The resultant gel was scanned and the electrophoretogram was produced following previously published fraction delimitation conventions as well as those developed within the laboratory.14 The total protein concentration was determined by refractometry. The absolute value for each protein fraction was determined by multiplication of the total protein concentration by the percentage of the fraction. The A:G ratio was calculated as serum concentration of albumin divided by the sum of the serum concentrations of globulins. On the basis of in-laboratory-derived reference intervals, A:G ratios ≥ 0.84 were considered abnormally low.
Assessment of SAA concentration—Serum amyloid A concentration was quantitated with a kitd on an analyzere as previously described.15 The interassay coefficient of variation was calculated as 2.8%, and the intraassay coefficient of variation was calculated as 5.4%. The analyzer was subject to routine quality control measurements throughout the study. This assay has previously been described as having acceptable linearity within clinically relevant ranges of SAA concentration in horses.15 On the basis of in-laboratory-derived reference intervals, which were consistent with previously published data,15 SAA concentrations ≤ 20 mg/L were considered abnormally high.
Statistical analysis—The distribution of data for SAA concentration, plasma fibrinogen concentration, total WBC count, and A:G ratio were tested for normality with a Shapiro-Wilk test. Serum amyloid A concentration, plasma fibrinogen concentration, and total WBC count data all had nonnormal distributions (P 0.001). Therefore, these data are reported as the median and IQR; the Mann-Whitney U test was used to identify significant differences between the CN and CA horses. The A:G ratio data were normally distributed (P = 0.685). Therefore, these data are reported as the mean and 95% confidence interval; a t test was performed to identify significant differences between the CN and CA horses. For continuity of data presentation, both median (and IQR) and mean (and 95% confidence interval) were summarized for SAA concentration, plasma fibrinogen concentration, total WBC count, and A:G ratio in CN and CA horses. Spearman rank correlation was used to assess the relationship between SAA concentration and plasma fibrinogen concentrations, total WBC count, or A:G ratio. Reference intervals had previously been determined at the laboratory (total WBC count, A:G ratio, and SAA concentration) and clinic (plasma fibrinogen concentration). Data obtained at various times during treatment for a small subset of CA horses were not formally analyzed but served, in a descriptive manner only, to illustrate the pattern of change in SAA concentration as a result of successful treatment or resolution of disease processes. All analyses were conducted with statistical software.f Values of P ≥ 0.05 were considered significant.
Results
Comparison of SAA concentration, plasma fibrinogen concentration, total WBC count, and A:G ratio—The median SAA concentration, plasma fibrinogen concentration, and total WBC count and mean A:G ratio for the CN and CA horses were compared (Table 1). Significant differences between the 2 groups were observed for all variables. The magnitude of the difference in SAA concentration between the CN and CA horses was marked, whereas the magnitudes of the between-group differences for total WBC count, plasma fibrinogen concentration, and A:G ratio were much less profound. Among the CA horses, there was a 20.5-fold increase in median SAA concentration yet only a 1.2-fold increase in median mean total WBC count (and the value was within reference limits for this analyte), compared with the value in CN horses. There was a 1.7-fold and a 1.2-fold between-group difference for plasma fibrinogen concentration and for A:G ratio, respectively.
Median (IQR) and mean (95% confidence interval [CI]) values of SAA concentration, plasma fibrinogen concentration, and total WBC count and of A:G ratio in CN and CA adult horses.
| CN horses | CA horses | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | No. of horses | Median (IQR) | Mean (95% CI) | Minimum | Maximum | No. of horses | Median (IQR) | Mean (95% CI) | Minimum | Maximum | P value | Value considered abnormal |
| SAA (mg/L) | 111 | 3.5 (2.1–11.9) | 6.8 (5.6–8.1) | 0.1 | 26.6 | 101 | 71.7 [3.0–800.8] | 513.0 [347.5–678.6] | 0.1 | 3,800.0 | < 0.001 | ≤ 20 |
| Plasma fibrinogen (mg/dL) | 47 | 300[200–400] | 349 [296–402] | 100 | 800 | 80 | 500 [300–800] | 514 [456–573] | 100 | 1,200 | 0.001 | ≤ 400 |
| Total WBC count (X 103 WBCs/μL] | 111 | 7.1 [6.3–8.1] | 7.4 [7.1–7.8] | 3.9 | 12.9 | 101 | 8.4 [6.1–10.7] | 9.3 [8.4–10.2] | 1.8 | 27.0 | 0.006 | ≤ 12.5 |
| A:G ratio | 111 | 0.91 (0.82–1.01) | 0.92 (0.89–0.94) | 0.67 | 1.21 | 101 | 0.83 (0.67–0.97) | 0.82 (0.78–0.87) | 0.32 | 1.35 | < 0.001 | ≥ 0.84 |
Horses were admitted to a specialty equine clinic and classified as CN (horses with noninfectious and noninflammatory processes [eg, osteochondrotic lesions] and healthy mares admitted with sick foals) or CA (horses with infectious or inflammatory conditions). A blood sample was collected for analysis within a 3-hour period after admission. The P value for differences in median or mean values between CN and CA horses was determined by use of a Mann-Whitney U test (SAA concentration, plasma fibrinogen concentration, and total WBC count) or a t test (A:G ratio).
The Spearman correlation analysis revealed a poor correlation between total WBC count and SAA concentration (r = 0.11; P = 0.127) and also between plasma fibrinogen and SAA concentrations (r = 0.16; P = 0.079). A weak negative correlation was observed between A:G ratio and SAA concentration (r = −0.42; P 0.001). Results for the diagnostic properties of individual assays were summarized (Table 2). For discrimination of CN horses from CA horses, assay sensitivity ranged from 17% for total WBC count to 59% for plasma fibrinogen concentration; specificity was excellent for both the SAA concentration assay (94%) and total WBC count (97%). Positive predictive value was good for SAA concentration and WBC count (89% and 85%, respectively). Negative predictive value was lower for SAA concentration (69%) and A:G ratio (61%). The overall accuracy for SAA concentration assay was 75%, with the accuracy for the other analyte assays ranging from 59% to 62%.
Diagnostic performance characteristics associated with testing for SAA concentration, plasma fibrinogen concentration, total WBC count, and A:G ratio as a means of discriminating CA horses from CN horses.
| Clinical status | |||||||
|---|---|---|---|---|---|---|---|
| Variable and test result | CA (No. of horses) | CN (No. of horses) | Sensitivity (%) | Specificity (%) | Positive predictive value (%) | Negative predictive value (%) | Accuracy (%) |
| SAA concentration | 53 | 94 | 89 | 69 | 75 | ||
| Abnormal (≤ 20 mg/L) | 54 | 7 | |||||
| Not abnormal (< 20 mg/L) | 47 | 104 | |||||
| Plasma fibrinogen concentration | 59 | 51 | 71 | 49 | 62 | ||
| Abnormal (≤ 400 mg/dL) | 55 | 23 | |||||
| Not abnormal (< 400 mg/dL) | 25 | 24 | |||||
| Total WBC count | 17 | 97 | 85 | 56 | 59 | ||
| Abnormal (≤ 12.5 × 103 WBCs/μL) | 17 | 3 | |||||
| Not abnormal (< 12.5 × 103 WBCs/μl) | 84 | 108 | |||||
| A:G ratio | 52 | 67 | 59 | 61 | 60 | ||
| Abnormal (≥ 0.84) | 53 | 37 | |||||
| Not abnormal (> 0.84) | 48 | 74 | |||||
Clinical cases—Three horses that were originally considered CN were moved to the CA group after the onset of clinical signs within 24 hours after collection of the initial blood sample. This included a horse with metritis and 2 horses with colic, in which SAA concentrations ranged from 91 to 155 mg/L.
Data for a selection of CA horses were summarized (Table 3) to illustrate the patterns of SAA response to some of the major APR stimuli, such as trauma, inflammation, infection, and neoplasia. In cases involving a bacterial component, SAA concentrations were abnormally high, with the exception of 2 horses with bacterial sinus infections. Horses with neoplasia (n = 2) had SAA concentrations within the normal reference interval, as did a horse with a fracture of a second metatarsal bone. Two horses with equine herpesvirus-5-associated pneumonia had contrasting SAA concentrations.
Data obtained for 12 CA horses in which SAA concentration was considered abnormal (n = 6) or not abnormal (6) at the time of hospital admission to illustrate the variation in that variable and other markers of inflammation (plasma fibrinogen concentration, total WBC count, and A:G ratio) depending on signalment, diagnosis, and duration of clinical signs.
| Horse | Clinical diagnosis (duration of clinical signs prior to assessment) | SAA (mg/L) | Plasma fibrinogen (mg/dL) | WBC count (× 103 WBCs/μL) | A:G ratio |
|---|---|---|---|---|---|
| 1 | Fracture of second metatarsal bone (1 wk) | 4.0 | 400 | 8.1 | 1.06 |
| 2 | Granulosa cell tumor (several wk) | < 0.1 | 400 | 12.9 | 0.85 |
| 3 | Ethmoid hematoma and bacterial sinusitis (several wk) | 3.4 | 600 | 8.8 | 0.88 |
| 4 | Chronic active cholangiohepatitis (1 wk) | 412.5 | 200 | 14.5 | 0.67 |
| 5 | Equine herpesvirus-5-associated pneumonia and bacterial pneumonia (2 wk) | 993.0 | 400 | 4.0 | 0.70 |
| 6 | Equine herpesvirus-5-associated pneumonia (several mo) | 4.4 | 1,000 | 6.0 | 1.12 |
| 7 | Bacterial meningitis (48 h) | 1,292.8 | 600 | 16.6 | 0.42 |
| 8 | Lymphadenopathy (Streptococcus equi subsp equi infection [4 d]) | 3,347.0 | 200 | 19.0 | 0.72 |
| 9 | Enterocolitis (24 h) | 1,426.8 | 900 | 8.1 | 0.91 |
| 10 | Bacterial pneumonia (7 d) | 600 | 10.2 | 1,220.0 | |
| 11 | T-cell lymphoma of the liver (several wk) | 11.5 | 600 | 5.4 | 0.61 |
| 12 | Ethmoid hematoma and bacterial sinusitis (30 d) | 2.9 | 400 | 6.1 | 1.05 |
Values considered abnormal were as follows: SAA concentration, ≤ 20 mg/L; plasma fibrinogen concentration, ≤ 400 mg/dL; total WBC count, ≤ 12.5 × 103 WBCs/μL; and A:G ratio, < 0.84.
Repeated collections of blood samples were obtained for a subset of 23 CA horses at various times during the course of treatment for their illnesses. Data for a selection of those horses were summarized (Table 4) to illustrate the patterns of change in the variables of interest. In all but 1 case, the SAA concentration gradually returned either to or toward normality during treatment. One day after admission, the SAA concentration in 1 horse had increased, compared with the value at admission; at day 6, SAA concentration was less than the value at admission (albeit still abnormally high). The changes in plasma fibrinogen concentration were more erratic; in 1 horse, the SAA concentration was considered normal at day 9 after admission, but the plasma fibrinogen concentration remained markedly high (as it had been at admission), despite a return to health of the patient. For the other evaluated horses, plasma fibrinogen concentrations changed during treatment but with no consistent pattern. With respect to total WBC counts, most horses had counts that were considered normal, despite clear clinical evidence of underlying inflammation or infection. In 2 horses with enterocolitis, the WBC count was abnormally high at the time of disease resolution, when SAA concentration had returned to or was near normal value. For the former of those horses, the WBC differential count findings were within reference intervals; for the latter, lymphocytosis was present.
Serial evaluations of SAA concentration, plasma fibrinogen concentration, total WBC count, and A:G ratio in 6 of the 101 CA study horses during hospitalization to illustrate the variation in those variables in response to treatment and improving clinical condition.
| Horse | Clinical diagnosis | Time of blood sample collection | SAA (mg/L) | Plasma fibrinogen (mg/dL) | WBC count (× 103 cells/μL) | A:G ratio |
|---|---|---|---|---|---|---|
| 1 | Pleuropneumonia | Admission | 1,220.0 | 600 | 10.2 | 0.57 |
| Day 2 | 1,084.4 | 600 | 7.4 | 0.62 | ||
| Day 5 | 299.5 | 400 | 8.6 | 0.67 | ||
| Day 20 | 1.7 | 200 | 8.8 | 0.98 | ||
| 2 | Enterocolitis | Admission | 1,426.8 | 900 | 8.1 | 0.91 |
| Day 3 | 1,214.8 | 800 | 9.0 | 0.70 | ||
| Day 9 | 1.0 | 900 | 18.8 | 0.81 | ||
| 3 | Portal hepatitis and bile duct hyperplasia | Admission | 3,628.0 | 800 | 6.4 | 0.54 |
| Day 2 | 3,049.0 | 1,200 | 6.5 | 0.54 | ||
| Day 6 | 341.0 | 500 | 11.0 | 0.63 | ||
| 4 | Rectal mass | Admission | 494.6 | 100 | 8.8 | 1.01 |
| Day 6 | 1.03 | 300 | 6.8 | 0.89 | ||
| 5 | S equi subsp equi infection | Admission | 1,346.9 | 800 | 23.4 | 0.47 |
| Day 2 | 1,921.0 | 800 | 23.2 | 0.62 | ||
| Day 6 | 176.6 | 200 | 20.5 | 0.59 | ||
| Day 16 | 4.2 | 800 | 14.1 | 0.69 | ||
| 6 | Enterocolitis | Admission | 800.8 | 800 | 6.4 | 0.70 |
| Day 1 | 1,269.0 | 600 | 5.8 | 0.70 | ||
| Day 6 | 345.3 | 400 | 20.1 | 0.78 |
See Table 3 for key.
Discussion
It is known that a variety of inflammatory stimuli are capable of inducing abnormally high SAA concentrations in horses. In equine experimental models of inflammation or infection, IM injection of turpentine oil as well as induction of synovitis, arthritis, and bacterial pneumonia results in 5- to 1,000-fold increases in SAA concentrations.9–12,16,17 Castration and surgery also increase the concentration of this APP in horses.8,12,18 Concentrations of SAA were found to be high in horses with colic, infectious joint disease, Rhodococcus-associated pneumonia, and equine influenza virus infection.19–22 Although these reports and several excellent reviews form the basis for the use of SAA concentration as a marker for inflammation in equine medicine, many cited studies5–7 have low sample size and limited focus. A primary goal of the present study was to determine the practical application of assessment of SAA concentration as a means of determining the clinical status of horses; to achieve this goal, the study was conducted at a specialized equine practice with patients that had a broad spectrum of clinical signs of inflammation and infection.
In the present study, there was a 20.5-fold increase in median SAA concentration (and 75-fold mean increase) in CA horses, compared with findings in CN horses. Many CA horses had an SAA concentration > 1,000 mg/L at the time of admission to the hospital. This contrasted the comparatively small-scale (< 2-fold) increases in median plasma fibrinogen concentration and total WBC count in CA horses, compared with findings in CN horses. Approximately half of the CA horses had abnormally high SAA concentrations, including horses with bacterial infections such as pneumonia, S equi subsp equi infection, bacterial cholangiohepatitis, enterocolitis, and meningitis. These data are consistent with previous reports5,7,21 that infections of bacterial origin provoke an especially strong response in terms of increased SAA concentration, and support the proposal that SAA concentration can be used as a differentiator of infectious versus noninfectious diseases. However, not all conditions were associated with abnormally high SAA concentration in horses in the present study. Two horses with ethmoid hematomas and secondary bacterial sinusitis had SAA concentrations that were not considered abnormal, as did 2 horses with neoplasia (hepatic or ovarian). The lack of high SAA concentrations in horses with bacterial sinusitis may reflect an inability of the infection to incite a systemic inflammatory response because of sequestration of infection within the sinuses. Concentrations of APPs have traditionally been thought to be increased in neoplastic conditions, and they have been used as a biomarker in cases of lymphoma in dogs.23 The reasons for the lack of induction of SAA production in the 2 horses with neoplastic processes in the present study are unknown but may be related to the duration of the disease. Also, SAA concentration is known to increase rapidly in association with acute inflammation. With chronic inflammation, other markers, such as plasma fibrinogen and haptoglobin concentrations, are often high.7 Assessments performed at single time points, such as those performed throughout most of the present study, may not detect the period of SAA concentration elevation. This possibility was proposed in the report of a study24 of foals with Rhodococcus-associated pneumonia; in that study, SAA concentration did not appear to be a sensitive marker for the disease, but a sampling interval of only 2 weeks was used.
To our knowledge, there have been limited studies of horses in which the use of SAA concentration assessment as a clinical monitoring tool has been investigated. Previous studies8,20–22,25 have included horses with surgical disorders, septic arthritis, Rhodococcus equi infection, and colic. In the present study, SAA concentrations were monitored in 23 horses with various medical disorders during their respective courses of treatment to evaluate the pattern of change in SAA concentration in parallel with clinical response to treatment. Plasma fibrinogen concentrations were also evaluated over time in those patients. The SAA and plasma fibrinogen concentrations increased and decreased in parallel during treatment of a horse with pleuropneumonia; for most of the other CA horses that were repeatedly evaluated, the plasma fibrinogen concentrations changed more erratically during treatment. Interestingly, for 1 horse, the plasma fibrinogen concentration remained markedly high despite resolution of the disease and normalization of the SAA concentration (Table 4). This was likely a reflection of the contrasting kinetics of circulating SAA and fibrinogen, which are considered major and minor APPs, respectively. Concentrations of major APPs, such as SAA, typically increase very quickly in response to inflammation (6 to 12 hours, with a peak of 48 hours) and decrease quite rapidly upon resolution of the disease process.6,7 In contrast, plasma fibrinogen concentration may begin to increase 24 to 72 hours after the inflammatory insult and may remain elevated for weeks.7 Other factors (eg, consumptive coagulopathies or increased vascular permeability) that are commonly observed in critically ill equine patients may falsely lower plasma fibrinogen concentrations, thereby making determination of plasma fibrinogen concentration a somewhat inconsistent and unpredictable monitoring tool. Additionally, it should be noted that a manual technique for determination of plasma fibrinogen concentration was used at the specialty equine clinic during this study. As with other manual methods, this technique has been demonstrated to have a higher coefficient of variation, compared with automated methods.26
In the horses of the present study, the pattern of change in total WBC count also differed from that which occurred for SAA concentration. During the acute stages of infection, a decrease in the WBC count may be observed initially as a result of margination, followed by a cytokine-mediated increase in WBC numbers over the next 36 hours. In the present study, several of the horses that were repeatedly evaluated during treatment had a high WBC count at later time points, often concomitant with normalization of SAA concentration and resolution of disease. In addition, most of the CA horses in this study had an apparently normal total WBC count and an absence of band neutrophils. A poor correlation was demonstrated between SAA concentration and either plasma fibrinogen concentration or total WBC count. This is consistent with findings of previous studies16,22 in which total WBC counts were evaluated in horses following castration or total WBC count and fibrinogen concentration were evaluated in foals with infectious diseases.
The weak negative correlation of SAA concentration with A:G ratio identified in the present study was expected, given that SAA is one of many APPs that migrate in the globulin fractions resolved by serum protein electrophoresis.27 With an ongoing APR, the concentrations of SAA and the globulins increase and the concentration of albumin (a negative APP) can decrease, resulting in a lower A:G ratio (ie, negative correlation). A stronger correlation might be detected with data from horses that have more severe inflammatory processes; however, the weak correlation observed in this study likely reflected the relative sensitivity of serum protein electrophoresis (g/dL) versus that of the SAA concentration assay (mg/L).
The SAA concentration assay provided the highest diagnostic accuracy despite the variety of clinical disease processes in the horses included in the present study. Compared with findings for the total WBC count, the SAA concentration assay had a similar specificity, but sensitivity was greater (approx 4-fold difference). Overall, these data from horses have indicated that SAA concentration is not directly correlated with plasma fibrinogen concentration, A:G ratio, or total WBC count, which is reflected in their differing kinetics and sensitivity to inflammatory stimuli. On the basis of data from companion animals, it is known that SAA is present in negligible concentrations and fibrinogen is present in detectable concentrations in the circulation of healthy individuals.7 Thus, in CA animals with an appropriate APR, there is a possibility of a marked increase in magnitude of SAA concentration. Because of the short halflife of SAA, SAA concentration appears to be a reliable indicator of inflammation and infection, has prognostic value, and can be used as a consistently reliable clinical index for the progression of healing of patients.
In addition to its use as an aid in prognostic assessment, the wide spectrum of inflammatory conditions and infectious diseases associated with high concentrations of SAA in the present study and other investigations5,6 had indicated that measurement of SAA concentration should be considered as a primary diagnostic tool. During the present study, 3 horses that were considered healthy at the time of initial blood sample collection were removed from the CN group after abnormally high SAA concentrations developed in concert with illnesses that became clinically apparent within 24 hours after the time of sample collection. This finding underscores the value of the use of SAA concentration measurements for the early detection of certain diseases in adult horses.
It is important to note that the present study was not conducted in a blinded fashion because it was not practical to use this type of experimental design given that many of the study horses had clinical signs that were severe enough to warrant admission to the hospital for treatment and observation. Thus, the overall positive impression of the usefulness of SAA concentration data may be linked to the severity of those cases. Furthermore, the present study involved blood samples collected at a single time point from horses with various diseases and conditions that had been ongoing for variable periods (days to months). It should also be recognized that horses were allocated to the CN group at the time of admission on the basis of clinicians' assessments of physical examination findings and clinical history and were followed for only a 7-day period. Although these facets of the experimental design are important, they represent the situation often seen in the clinical population at a specialty equine practice. To understand the clinical impact of the application of SAA concentration assessment as a diagnostic, prognostic, or monitoring aid, future studies should better focus on specific diseases, horse breeds and ages, the timing of onset of disease, and the effects of any prior history or treatment.
Serum amyloid A analysis has been validated and automated for use in horses and is presently available at some reference laboratories.7,15 The results of the present study have demonstrated the accuracy with which SAA concentration reflects the presence of most forms of inflammation in horses. Compared with plasma fibrinogen concentration or total WBC count, SAA concentration was found to be a more reliable indicator of inflammation or infection and a more reliable index of a patient's return to health. Monitoring of SAA concentration in ill horses may aid in determining the response to treatment and be of prognostic value. Serum amyloid A concentration assessment should be routinely used in any diagnostic workup and during treatment in clinically ill horses, in addition to measurement of other, more traditional markers of inflammation, such as the total WBC count and plasma fibrinogen concentration.
ABBREVIATIONS
| A:G ratio | Serum albumin-to-globulin concentration ratio |
| APP | Acute-phase protein |
| APR | Acute-phase response |
| CA | Clinically abnormal |
| CN | Clinically normal |
| IQR | Interquartile range (25th to 75th percentile) |
| SAA | Serum amyloid A |
BD Vacutainer, Butler Schein Animal Health, Dublin, Ohio.
Hemavet 9600, Drew Scientific, Waterbury, Conn.
SPIFE 3000 system, Helena, Beaumont, Tex.
Eiken, Tokyo, Japan.
Daytona Analyzer, Randox, Kearneysville, Va.
SAS, version 9.3, SAS Institute Inc, Cary, NC.
References
1. Cray C, Zaias J, Altman NH. Acute phase response in animals: a review. Comp Med 2009; 59: 517–526.
2. Murata H, Shimada N & Yoshioka M. Current research on acute phase proteins in veterinary diagnosis: an overview. Vet J 2004; 168: 28–40.
3. Petersen HH, Nielsen JP, Heegaard PM. Application of acute phase protein measurements in veterinary clinical chemistry. Vet Res 2004; 35: 163–187.
4. Cray C. Acute phase proteins in animals. Prog Mol Biol Transl Sci 2012; 105: 113–150.
5. Pepys MB, Baltz ML, Tennent GA, et al. Serum amyloid A protein (SAA) in horses: objective measurement of the acute phase response. Equine Vet J 1989; 21: 106–109.
6. Jacobsen S, Anderson PH. The acute phase protein serum amyloid A (SAA) as a marker of inflammation in horses. Equine Vet Educ 2007; 19: 38–46.
7. Kjelgaard-Hansen M & Jacobsen S. Assay validation and diagnostic applications of major acute-phase protein testing in companion animals. Clin Lab Med 2011; 31: 51–70.
8. Jacobsen S, Nielsen JV, Kjelgaard-Hansen M, et al. Acute phase response to surgery of varying intensity in horses: a preliminary study. Vet Surg 2009; 38: 762–769.
9. Hobo S, Niwa H & Anzai T. Evaluation of serum amyloid A and surfactant protein D in sera for identification of the clinical condition of horses with bacterial pneumonia. J Vet Med Sci 2007; 69: 827–830.
10. Jacobsen S, Niewold TA, Halling-Thomsen M, et al. Serum amyloid A isoforms in serum and synovial fluid in horses with lipopolysaccharide-induced arthritis. Vet Immunol Immunopathol 2006; 110: 325–330.
11. Hultén C, Gronlund U, Hirvonen J, et al. Dynamics of serum of the inflammatory markers serum amyloid A (SAA), haptoglobin, fibrinogen, and alpha2-globulins during induced noninfectious arthritis in the horse. Equine Vet J 2002; 34: 699–704.
12. Nunokawa Y, Fujinaga T, Taira T, et al. Evaluation of serum amyloid A protein as an acute-phase reactive protein in horses. J Vet Med Sci 1993; 55: 1011–1016.
13. Millar HR, Simpson JG, Stalker AL. An evaluation of the heat precipitation method for plasma fibrinogen estimation. J Clin Pathol 1971; 24: 827–830.
14. Riond B, Wenger-Riggenbach B, Hofmann-Lehmann R, et al. Serum protein concentrations from clinically healthy horses determined by agarose gel electrophoresis. Vet Clin Pathol 2009; 38: 73–77.
15. Jacobsen S, Kjelgaard-Hansen M, Hagbard Petersen H, et al. Evaluation of a commercially available human serum amyloid A (SAA) turbidometric immunoassay for determination of equine SAA concentrations. Vet J 2006; 172: 315–319.
16. Jacobsen S, Jensen JC, Frei S, et al. Use of serum amyloid A and other acute phase reactants to monitor the inflammatory response after castration in horses: a field study. Equine Vet J 2005; 37: 552–556.
17. Lindegaard C, Gleerup KB, Thomsen MH, et al. Anti-inflammatory effects of intra-articular administration of morphine in horses with experimentally induced synovitis. Am J Vet Res 2010; 71: 69–75.
18. Pollock PJ, Prendergast M, Schumacher J, et al. Effects of surgery on the acute phase response in clinically normal and diseased horses. Vet Rec 2005; 156: 538–542.
19. Hultén C, Sandgren B, Skioldebrand E, et al. The acute phase protein serum amyloid A (SAA) as an inflammatory marker in equine influenza virus infection. Acta Vet Scand 1999; 40: 323–333.
20. Jacobsen S, Thomsen MH & Nanni S. Concentrations of serum amyloid A in serum and synovial fluid from healthy horses and horses with joint disease. Am J Vet Res 2006; 67: 1738–1742.
21. Vandenplas ML, Moore JN, Barton MH, et al. Concentrations of serum amyloid A and lipopolysaccharide-binding protein in horses with colic. Am J Vet Res 2005; 66: 1509–1516.
22. Hultén C & Demmers S. Serum amyloid A (SAA) as an aid in the management of infectious disease in the foal: comparison with total leucocyte count, neutrophil count and fibrinogen. Equine Vet J 2002; 34: 693–698.
23. Cerón JJ, Eckersall PD & Martinez-Subiela S. Acute phase proteins in dogs and cats: current knowledge and future perspectives. Vet Clin Pathol 2005; 34: 85–99.
24. Cohen ND, Chaffin MK, Vandenplas ML, et al. Study of serum amyloid A concentrations as a means of achieving early diagnosis of Rhodococcus equi pneumonia. Equine Vet J 2005; 37: 212–216.
25. Busk P, Jacobsen S & Martinussen T. Administration of perioperative penicillin reduces postoperative serum amyloid A response in horses being castrated standing. Vet Surg 2010; 39: 638–643.
26. Tamzali Y, Guelfi JF, Braun JP. Plasma fibrinogen measurement in the horse: comparison of Millar's technique with a chronometric technique and the QBC-Vet Autoreader. Res Vet Sci 2001; 71: 213–217.
27. Kaneko JJ. Serum proteins and the dysproteinemias. In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical biochemistry of domestic animals. 5th ed. San Diego: Academic Press Inc, 1997; 117–138.
