384 Variability in bovine liver abscess frequency, proportional mass, and location within abscess severity scores.

Abstract Liver abscesses (LA) are a multifactorial, polymicrobial disease that occur in all ages and classes of cattle. Abscesses vary in count, size, and shape. Bovine LA are associated with reduced feed intake, reduced weight gain, decreased feed efficiency, and decreased carcass yield. Liver abscess scoring systems are based upon visual appraisal of the external surfaces of the parenchyma. Limited research has been conducted on the location effect of LA within the bovine liver or the totality (frequency or mass percentage) of LA contained within the liver. The objective of this study was to document location effects of LA within finished bovine, and to compare visual evaluation of LA external on the liver to total number of LA. Abscessed livers (n = 294) were assigned an Elanco Liver Score by visual appraisal on-line in commercial harvest facilities; livers were removed offline for intensive evaluation. Diaphragmatic and visceral surfaces of each liver were photographed, and visual external LA number were recorded for both surfaces. Liver abscesses were individually excised, and the total count and mass percentages were determined for each liver. Relative location of visually external LA was determined by creating a frequency heatmap. Data were analyzed using GLIMMIX models; the fixed effect was visually appraised Elanco liver score, and means were separated using the LINES option. Mass percentage and frequency of abscesses were estimated from visual surface counts using REG procedures. Percentage of liver mass represented by LA was greatest (P < 0.01) in livers scored A+Open (13.4%), compared to A+Adhesion or A+ (10.5 and 8.3%, respectively), which were greater (P < 0.01) than livers score A or A- (3.1 and 1.6%, respectively). Similarly, frequency of LA was greater (P < 0.01) for livers scored A+Open (31.0) than all other scores (A+ = 14.1, A+Adhesion = 11.2, A = 3.5, A- = 2.5). Heatmap analysis indicated that LA were observed (P < 0.01) on the diaphragmatic (3.9) surface of the liver at twice the frequency of the visceral surface (1.9). Abscesses within the liver parenchyma and not visible on either surface were excised in the greatest (P < 0.01) frequency for livers scored A+Open (18.8) in comparison to those scored A+ (7.8), A+Adhesion (6.0), A (0.9), or A- (0.7). Regression outcomes indicate strong ability (R2 = 0.83 to 0.90) to estimate total frequency of LA via counting diaphragmatic and/or visceral surfaces. In contrast, linear regression suggests a weak ability (R2 = 0.25 to 0.27) to estimate the mass percentage of LA from diaphragmatic or visceral surface counts, due to extreme variability in absolute LA size (< 1 g to >11,500 g). These data support the creation of a new LA scoring system to include the wide variability in A+ severity.