In a retrospective investigation evaluating neonatal piglets receiving exclusive total parenteral nutrition, systemic bloodstream infection emerged as a potent and independent contributor to severe hepatobiliary dysfunction. Researchers examining experimental data collected across fourteen years determined that confirmed bacteremia significantly suppressed bile flow, elevated serum total bilirubin concentrations, and substantially increased animal mortality. Preclinical animal models have served for decades as the foundational testing ground for therapies targeting intestinal failure-associated liver disease, particularly in assessing novel intravenous lipid emulsions designed for human infants. However, experimental protocols in laboratory animals frequently fail to monitor, account for, or report intercurrent systemic infections. The new findings demonstrate that sepsis is not merely an occasional procedural complication but an influential confounding variable that independently alters primary liver outcomes, potentially distorting how scientists and clinicians interpret the biological efficacy of intravenous lipid formulations.
Intestinal failure represents a catastrophic gastrointestinal condition defined by the physiological inability of the gut to absorb adequate macronutrients, fluids, and electrolytes to sustain growth and development. Despite major therapeutic gains achieved through multidisciplinary intestinal rehabilitation teams, prolonged parenteral nutrition dependency exposes fragile pediatric patients to life-threatening complications. Although historical clinical literature frequently attributed hepatic pathology to the hepatotoxic properties of soybean-derived intravenous lipid emulsions, contemporary pediatric epidemiology demonstrates that central line-associated bloodstream infections represent an equally critical and independent driving factor, conferring a threefold increased risk for liver disease and dramatically driving up mortality rates.
Nevertheless, pre-existing animal studies have routinely neglected to measure, report, or statistically adjust for systemic bloodstream complications. In a comprehensive review of thirty-eight published animal studies examining total parenteral nutrition where liver disease served as the primary experimental outcome, investigators found that a mere 21 percent of published reports documented sepsis incidence, while only 47 percent reported animal mortality rates. To rigorously establish whether sepsis acts as an unmeasured confounding influence in preclinical hepatobiliary research, investigators at a single laboratory systematically re-evaluated their experimental records from 14-day total parenteral nutrition trials conducted in neonatal piglets between 2011 and 2025.
The retrospective cohort comprised 86 eligible neonatal piglets aged between two and five days, drawn from four distinct experimental protocols evaluating intravenous lipid formulations. Among the total cohort, 44 animals, or 51 percent, were clinically suspected of developing systemic infection based on standardized evaluations of acute physical signs. Following blood culture analyses—excluding four animals with missing culture data—confirmed bacteremia was verified in 28 out of 40 tested piglets, yielding a 70 percent microbiological positivity rate. Consequently, 82 piglets were incorporated into the definitive statistical analysis, comparing 54 non-septic piglets against 28 culture-confirmed septic animals. Baseline physical characteristics revealed that non-septic piglets were significantly older and exhibited higher starting body weights than their septic counterparts. Over the 14-day experimental timeframe, non-septic animals demonstrated superior somatic growth, finishing the trials with significantly higher final body weights and greater cumulative weight gains.
All included animal trials followed uniform baseline surgical and nutritional procedures. On experimental day zero, neonatal piglets underwent general anesthesia to place a 5-French jugular central venous catheter designed for continuous, round-the-clock parenteral delivery. Intravenous nutrition commenced immediately following catheterization and advanced to full target volumes according to institutional protocols, completely withholding enteral feeding for the entire two-week experimental window. The nutritional regimens maintained consistent macronutrient and micronutrient profiles apart from the specific lipid sources administered. Experimental groups received either a pure soybean oil-based emulsion, a pure fish oil-based emulsion, or a mixed-oil formulation, delivered at dosages of either 5 grams per kilogram daily or 10 grams per kilogram daily. In translational terms, these preclinical dosages correspond directly to standard clinical human infant regimens of 1 gram per kilogram daily and 2 grams per kilogram daily, respectively. To isolate lipid and infectious effects, trials involving supplemental vitamin E or trophic hormonal interventions were strictly excluded.
Piglets were housed individually within dedicated temperature-controlled metabolic cages and received comprehensive twice-daily health evaluations from trained laboratory personnel who assessed catheter patency, overall behavior, and clinical markers of sepsis. Febrile status was defined by a rectal temperature exceeding 39.5 degrees Celsius measured via calibrated digital rectal thermometers. Because fever was not treated as an isolated diagnostic indicator, it was evaluated alongside clinical lethargy—defined as reduced spontaneous physical activity tracked via standardized scoring charts with visual prompts—and active vomiting. Manifestations of acute illness immediately initiated a formal rescue protocol comprising diagnostic blood culture collection directly through the central venous catheter followed by protocolized systemic antimicrobial administration using enrofloxacin and gentamicin. In cases where bacteremic animals proved unresponsive to therapeutic antibiotic regimens, predefined humane intervention endpoints were strictly enforced.
Over the fourteen-year period analyzed, procedural protocols within the facility underwent important infection-control refinements. Prior to 2020, intravenous catheters were maintained using twice-daily flushes of 1.5 milliliters of standard heparin solution, while perioperative antibiotic prophylaxis consisted of ampicillin and trimethoprim-sulfadoxine administered postoperatively on days zero through four and repeated on days eight through twelve. Following protocol modernization in 2020, maintenance procedures integrated a 1-milliliter catheter lock consisting of 4 percent tetrasodium ethylenediaminetetraacetic acid instilled daily for two hours beginning on day two. Surgical prophylaxis was simultaneously restructured to deliver intramuscular florfenicol combined with intravenous ampicillin exclusively from days zero to four, completely eliminating the secondary routine prophylactic antibiotic course. The retrospective comparison confirmed that piglets maintained exclusively with heparin flushes experienced a significantly higher rate of culture-proven sepsis compared to animals receiving tetrasodium ethylenediaminetetraacetic acid locks, exhibiting infection rates of 40.6 percent versus 11.1 percent.
Detailed microbiological profiling of the 28 blood-culture-positive animals demonstrated a overwhelming dominance of gram-positive pathogens, which represented 79 percent of total microbial isolations, whereas gram-negative bacteria accounted for 18 percent and polymicrobial infections accounted for 3 percent. Among gram-positive isolates, Enterococcus faecium predominated with twelve confirmed cases, followed by four cases of Enterococcus faecalis, three of Leuconostoc citreum, two of Staphylococcus epidermidis, and single identifications of Staphylococcus simulans, Staphylococcus hyicus, Leuconostoc mesenteroides, Clostridium perfringens, and an unspecified Enterococcus species. Gram-negative infections included three cases of Klebsiella pneumoniae alongside single identifications of Enterobacter cloacae, Escherichia coli, and Bacteroides fragilis. Sepsis severely compromised animal viability, driving mortality to 43 percent in septic piglets compared to 9 percent among non-septic animals. While piglets infected with gram-negative bacteria experienced a nominally higher mortality rate of 60 percent compared to 41 percent among those with gram-positive infections, this microbial disparity did not achieve statistical significance.
Direct physiological and automated biochemical assays conducted at the conclusion of the 14-day trials on experimental day fourteen revealed profound biliary impairment in septic piglets. Direct surgical cannulation of the common bile duct demonstrated that bile flow was significantly lower in bacteremic animals relative to non-septic controls. Serum biochemical profiles processed through automated analyzers revealed that septic animals experienced marked elevations in total serum bilirubin and gamma-glutamyl transferase concentrations. In contrast, serum total bile acids and alanine aminotransferase levels did not demonstrate statistically significant divergence between the septic and non-septic cohorts, with each parameter yielding a marginal p-value of 0.07. These specific laboratory aberrations closely parallel human clinical manifestations of intestinal failure-associated liver disease, in which bacteremia-induced cholestasis impairs hepatic bile secretion and triggers progressive conjugated hyperbilirubinemia.
Univariate logistic regression analysis revealed that the odds of contracting systemic bloodstream infection were significantly reduced in piglets that were older at baseline, weighed more on day zero or day fourteen, or received the tetrasodium ethylenediaminetetraacetic acid catheter lock solution. Conversely, piglets administered a pure soybean oil-based lipid emulsion exhibited a dramatically higher sepsis rate of 59.4 percent compared to 23.1 percent in pure fish oil cohorts and 16.2 percent in mixed-oil groups. Piglet breed, biological sex, and lipid dosage levels showed no significant association with infection vulnerability. Subsequent multivariable logistic regression modeling confirmed that both baseline age and dietary lipid source remained powerful independent predictors of bacteremia. Advancing baseline piglet age conferred substantial protection against systemic infection, yielding an odds ratio of 0.42. In contrast, total parenteral feeding with pure soybean oil emulsion increased the odds of developing sepsis more than tenfold relative to pure fish oil, presenting an odds ratio of 10.77.
Multivariable linear regression modeling established that sepsis functioned as an autonomous driver of primary hepatic dysfunction independent of competing physiological variables. In the multivariable model for bile secretion, confirmed sepsis exerted a significant negative effect while baseline piglet body weight exerted a positive predictive effect, together accounting for 42 percent of the total variance in bile flow rates. For total serum bilirubin, both bacteremia and the administration of soybean-derived lipid emulsions emerged as independent positive predictors, jointly explaining 34 percent of the observed variance. Multivariable modeling further confirmed sepsis as the sole independent predictor of elevated serum bile acids, accounting for 12 percent of variance. Conversely, final day-fourteen body weight and the preventive utilization of tetrasodium ethylenediaminetetraacetic acid locks independently predicted lower circulating gamma-glutamyl transferase levels, explaining 25 percent of variance. Alanine aminotransferase lacked significant univariate associations and was consequently omitted from multivariate modeling.
Pathophysiological mechanisms linking sepsis to cholestatic injury involve both direct immunological signaling and disruptions across the gut-liver axis. Sepsis directly suppresses bile production by triggering hepatic synthesis of proinflammatory cytokines, including tumor necrosis factor-alpha and interleukin-1, alongside nitric oxide, which downregulate essential hepatobiliary bile acid transporters. The strong association between soybean-based emulsions and heightened sepsis risk likely stems from their heavy concentration of omega-6 fatty acids and phytosterols, which exert proinflammatory effects that can impair neonatal immune competence. Translating these animal findings to clinical care underscores existing controversies surrounding lipid administration during neonatal infection.
The authors outlined several methodological constraints inherent to their retrospective design, which spanned more than a decade of laboratory investigations. The overall representation of female piglets and animals managed with tetrasodium ethylenediaminetetraacetic acid catheter locks remained relatively small, while breeding herd genetics shifted over time following the introduction of Duroc genetics after 2012. Furthermore, gradual evolutions in laboratory staff, antimicrobial protocols, and hygiene standards could have introduced unmeasured shifts in the porcine microbiome. Diagnostic blood cultures may have yielded occasional false negatives due to perioperative antibiotic exposure, or false positives from skin contamination during sampling. Nonetheless, because sepsis independently suppresses bile flow, elevates bilirubin, and multiplies mortality, researchers must systematically implement refined antimicrobial line locks, routinely monitor infectious morbidity, and transparently publish sepsis data to prevent distorted conclusions in preclinical liver disease research.
Subject of Research: Sepsis as an independent confounding factor in preclinical total parenteral nutrition models of liver disease
Article Title: Sepsis as a confounding factor in assessing liver dysfunction in parenterally fed piglets: a model for human infants with intestinal failure
Article References: Alam, M., Wizzard, P. R., Nation, P. N., Zaugg, M., Wales, P. W., & Turner, J. M. (2026). Sepsis as a confounding factor in assessing liver dysfunction in parenterally fed piglets: a model for human infants with intestinal failure. Laboratory Animal Research, 42(1), Article 18. https://doi.org/10.1186/s42826-026-00279-5
Image Credits: AI Generated
DOI: 10.1186/s42826-026-00279-5
Keywords: Sepsis, Total Parenteral Nutrition, Intestinal Failure-Associated Liver Disease, Bile Flow, Total Bilirubin, Neonatal Piglet Model, Intravenous Lipid Emulsions, Central Line-Associated Bloodstream Infection, Tetrasodium Ethylenediaminetetraacetic Acid, Soybean Oil Emulsion, Gamma-Glutamyl Transferase, Biliary Cholestasis, Preclinical Animal Research, Catheter Lock Solutions, Enterococcus faecium, Microbial Dysbiosis
News Source: Drew Townsend. (November 18, 2026). Sepsis Impairs Bile Flow and Exacerbates Liver Dysfunction in Parenterally Fed Piglets. Scienmag.



