Identification of blood-based biomarkers for detection of liver abscesses in beef × dairy heifers*

Abstract

Objective

The objective was to determine the effects of liver abscesses on complete blood counts (CBC) and liver function and to identify blood-based biomarkers for detection of liver abscesses in feedlot cattle.

Materials and Methods

Blood samples were collected on a subset of beef × dairy crossbred heifers with and without liver abscesses (n = 94; n = 113, respectively). From these blood samples, CBC and serum chemistry analytes related to liver health (alkaline phosphatase, gamma glutamyl transferase, bile acids, total bilirubin, and cholesterol) were analyzed to characterize animal health and model for liver abscesses. Abscess presence was predicted using logistic regression models and machine learning algorithms, with individual heifer serving as the experimental unit.

Results and Discussion

In a binary system (abscessed or not), heifers with liver abscesses had lower concentrations of lymphocyte (LYM) and total white blood cell (WBC) concentrations but elevated platelet-to-LYM ratio. Serum chemistry measures were not affected by abscess presence or liver score (severity). In univariate, predictive modeling of CBC measures, only total WBC concentration was accurate (65.85%) at predicting liver abscess presence; however, this was not deemed an acceptable biomarker due to its poor specificity (54.79%). Using total WBC, basophils, red blood cells, and procalcitonin as variables to predict liver abscess condition, the Logit- Boost model produced the greatest accuracy (80.65%) and specificity (85.71%). An accurate blood test that allows for the detection of liver abscesses during the finishing phase may allow for therapeutic management or alternative marketing of cattle; however, the marginal accuracy demonstrated in this study combined with the impracticality and cost of analyzing a blood sample may not be feasible. Regardless, the most explanatory CBC variables in cattle with abscessed livers were suggestive of an immune response.

Implications and Applications

Liver condition influenced the CBC of beef × dairy heifers but not serum chemistry. Multivariate predictive models showed potential for predicting liver abscess condition based on CBC. The results of this study warrant further investigation for biomarkers of the liver abscess condition. An antemortem indicator for liver abscesses could prove useful for making management decisions and in research applications, particularly in evaluating interventions.