Dear Editor, In response to the letter to the editors by Dr Deep et al. [1]: the aim of our retrospective analysis in a single center [2] was to study dialysis-dependent acute kidney injury (dAKI) and the feasibility of dialysis in critically ill children with severely impaired liver function. We agree that the pathophysiology of acute liver failure is distinct from acute-on-chronic liver failure and may influence the development and course of acute kidney injury (AKI), even though in our cohort, the clinical presentation was quite comparable (see Table 3 in [1]). Fortunately, almost a decade ago the term pediatric risk, injury, failure, loss, end-stage renal disease (pRIFLE) has replaced more than 30 different definitions of AKI that hampered the comparability of various older studies. In comparison with various biomarkers, a 50 % increase in serum creatinine appears to be superior in detecting AKI judged by its sensitivity and specificity [3]. Indeed, we agree that the definition of AKI in children with severely impaired liver function is difficult, particularly because of reduced muscle mass and low serum creatinine levels. In any case, dialysis-dependent AKI is substantially defined by the patient’s need for renal replacement therapy due to a rapid decline in kidney function. Technically, all the children in our cohort met the pRIFLE Bfailure^ criterion of estimated glomerular filtration rate (eGFR) below 35 ml/min/1.73 m (see page 2202). We applied the revised Schwartz formula to estimate GFR. This takes into account serum creatinine, urea, cystatin c, length and sex, and seems less susceptible to overestimating eGFR in children with low muscle mass. The definition of hepatorenal syndrome (HRS) in children remains a challenge of which we are well aware: the pediatric definition of HRS by Van Roey and Moore [4] is insufficient and the definitions of the International Club of Ascites (ICoA) are based on adult data, as are the two references quoted by Dr Deep. Adult data cannot be compared with those of children, especially when referring to cause, pathophysiology, and course of liver cirrhosis. For instance, we consider it problematic to apply a score to children that was developed in adults older than 50 years with an etiology of alcoholic liver cirrhosis in more than 50% of patients. Unfortunately, pediatric data on HRS are scarce to non-existent. Eventually, we used the definition of the ICoA. We highlight the need for studies in the pediatric population, ideally prospective, multicenter studies.
Read full abstract