Abstract:The article by Chand and Sanyal1 addresses an important issue and is very interesting and informative. Although the association between hyperbilirubinemia and sepsis is known to many clinicians, patte...The article by Chand and Sanyal1 addresses an important issue and is very interesting and informative. Although the association between hyperbilirubinemia and sepsis is known to many clinicians, patterns and mechanisms of hyperbilirubinemia are not so well characterized. Furthermore, in settings in which hyperbilirubinemia may easily be explained by other factors (for instance by transient liver insufficiency after major liver resection) the responsibility of sepsis may be overlooked with a dangerous impact on outcome. Indeed, this is our setting. In an ongoing prospective study on liver resection patients, we have assessed a cumulative impact of sepsis, inadequate residual parenchymal volume, and duration of intraoperative normothermic ischemia in determining maximum postoperative plasma bilirubin (up to greater than 33.0 mg/dL, or 564 μmol/L, in extreme cases). The effect of sepsis was to increase bilirubin or to cause an acute and transient peak overimposed on the pattern generated by the other factors. A striking observation was the tendency for a fairly constant relationship between conjugated and total bilirubin (regression slope = 0.67, r2 = 0.98, P ≪ 0.001; n = 263) which was independent of the prevailing factor (major resection, prolonged ischemia, sepsis). Alkaline phosphatase (AP) and gamma glutamyl transpeptidase (GGT) were normal or very slowly increasing, which was consistent with liver regeneration, and continued to slowly increase many weeks after surgery, when bilirubin was decreasing. The obvious exceptions were some larger increases in AP and GGT observed in patients with biliary sepsis or fistula. With regard to the article by Chand and Sanyal1, we wonder if the apparently similar and synergistic impact of inadequate liver volume, prolonged ischemia, and sepsis, with maintenance of a rather constant conjugated/total bilirubin ratio (≈2/3) may be consistent with a common and synergistic effect of these factors on hepatocellular bilirubin excretion (for instance, on multidrug resistance–associated protein activity), and whether this may have a clinical relevance, also for other treatments independent of that of sepsis. Very rare exceptions to this main pattern were observed outside the prospective period, and consisted of almost totally conjugated hyperbilirubinemia, larger increases in AP and GGT, and variable increases in aspartate aminotransferase and alanine aminotransferase. These were basically unrelated to sepsis, although sepsis could aggravate them, and had more complex or unfavorable outcomes. One of these patients had ischemic cholangitis after arterial thrombosis, other cases remained unexplained, and whether a small-for-size residual liver could have been a factor remains speculative. In conclusion, the authors1 should be congratulated for their effort, because many patterns of changes in plasma bilirubin are far from obvious in reality, and the relevance of their characterization extends far beyond the diagnosis of sepsis. The issue of sepsis is of utmost clinical importance in our setting, because the biochemical synergism of sepsis superimposed on liver insufficiency in increasing bilirubin is paralleled by a powerful synergism in worsening outcome. Finally there is a very minor inaccuracy in the article1 concerning the daily rate of production of bilirubin: although it is obvious that 4 mg per day should mean 4 mg/kg/day, this might require an "erratum". Ivo Giovannini*, Carlo Chiarla*, Felice Giuliante*, Maria Vellone*, Francesco Ardito*, Gennaro Nuzzo*, * Department of Surgery,, Hepatobiliary Unit and CNR-IASI Shock Center, Catholic University of the Sacred Heart, Rome, Italy.Read More
Abstract: The article by Chand and Sanyal1 addresses an important issue and is very interesting and informative. Although the association between hyperbilirubinemia and sepsis is known to many clinicians, patterns and mechanisms of hyperbilirubinemia are not so well characterized. Furthermore, in settings in which hyperbilirubinemia may easily be explained by other factors (for instance by transient liver insufficiency after major liver resection) the responsibility of sepsis may be overlooked with a dangerous impact on outcome. Indeed, this is our setting. In an ongoing prospective study on liver resection patients, we have assessed a cumulative impact of sepsis, inadequate residual parenchymal volume, and duration of intraoperative normothermic ischemia in determining maximum postoperative plasma bilirubin (up to greater than 33.0 mg/dL, or 564 μmol/L, in extreme cases). The effect of sepsis was to increase bilirubin or to cause an acute and transient peak overimposed on the pattern generated by the other factors. A striking observation was the tendency for a fairly constant relationship between conjugated and total bilirubin (regression slope = 0.67, r2 = 0.98, P ≪ 0.001; n = 263) which was independent of the prevailing factor (major resection, prolonged ischemia, sepsis). Alkaline phosphatase (AP) and gamma glutamyl transpeptidase (GGT) were normal or very slowly increasing, which was consistent with liver regeneration, and continued to slowly increase many weeks after surgery, when bilirubin was decreasing. The obvious exceptions were some larger increases in AP and GGT observed in patients with biliary sepsis or fistula. With regard to the article by Chand and Sanyal1, we wonder if the apparently similar and synergistic impact of inadequate liver volume, prolonged ischemia, and sepsis, with maintenance of a rather constant conjugated/total bilirubin ratio (≈2/3) may be consistent with a common and synergistic effect of these factors on hepatocellular bilirubin excretion (for instance, on multidrug resistance–associated protein activity), and whether this may have a clinical relevance, also for other treatments independent of that of sepsis. Very rare exceptions to this main pattern were observed outside the prospective period, and consisted of almost totally conjugated hyperbilirubinemia, larger increases in AP and GGT, and variable increases in aspartate aminotransferase and alanine aminotransferase. These were basically unrelated to sepsis, although sepsis could aggravate them, and had more complex or unfavorable outcomes. One of these patients had ischemic cholangitis after arterial thrombosis, other cases remained unexplained, and whether a small-for-size residual liver could have been a factor remains speculative. In conclusion, the authors1 should be congratulated for their effort, because many patterns of changes in plasma bilirubin are far from obvious in reality, and the relevance of their characterization extends far beyond the diagnosis of sepsis. The issue of sepsis is of utmost clinical importance in our setting, because the biochemical synergism of sepsis superimposed on liver insufficiency in increasing bilirubin is paralleled by a powerful synergism in worsening outcome. Finally there is a very minor inaccuracy in the article1 concerning the daily rate of production of bilirubin: although it is obvious that 4 mg per day should mean 4 mg/kg/day, this might require an "erratum". Ivo Giovannini*, Carlo Chiarla*, Felice Giuliante*, Maria Vellone*, Francesco Ardito*, Gennaro Nuzzo*, * Department of Surgery,, Hepatobiliary Unit and CNR-IASI Shock Center, Catholic University of the Sacred Heart, Rome, Italy.