The Renaissance of Cholemic Nephropathy

Supported by the Andrew K. Burroughs short‐term training fellowship of the European Association for the Study of the Liver (to M.M.). Potential conflict of interest: Nothing to report. See Article on Page 2107. In this issue of hepatology, Bräsen and Mederacke et al.1 present an analysis of all kidney biopsies performed at the gastroenterology and hepatology department at Hannover Medical School over a 16‐year period. The authors identified 79 patients with both liver disease and impaired renal function, 57% of whom had acute kidney injury (AKI) or AKI overlapping chronic kidney disease (CKD), while 43% had CKD, as defined by the recently updated European clinical practice guidelines. Interestingly, 8 patients were diagnosed with cholemic nephropathy (CN), all of whom presented as AKI. The term “cholemic nephropathy” refers to renal dysfunction and characteristic histological changes, i.e., tubular epithelial injury in distal nephron segments accompanied by intraluminal cast formation leading to obstruction and dilatation of tubules, in patients with jaundice. Nevertheless, the histological criteria for diagnosing CN are poorly defined.2 CN has been described in a variety of acute or chronic conditions leading to jaundice, including advanced chronic liver disease (ACLD)/cirrhosis.3 Besides a prospective study on the reversibility of CN restricted to patients with acute obstructive jaundice4 and a clinicopathologic analysis of a highly heterogeneous cohort comprising both patients with obstructive and nonobstructive jaundice (mostly autopsies),3 the available clinical evidence on CN is anecdotal,2 which highlights the importance of the comprehensive present study by Bräsen and Mederacke et al.1 Patients with CN had higher serum bilirubin levels, and bilirubin was detectable in the urine of all patients with CN. Proteinuria and hematuria were observed in about half of patients, which was similar to AKI patients without CN. The high prevalence of urinalysis abnormalities might be explained by the study setting (i.e., patients with a clinical indication for kidney biopsy). The majority of patients with CN therefore would have been correctly identified as AKI not meeting the criteria of hepatorenal syndrome (non‐HRS‐AKI) by simple urinalysis, as recommended for the diagnostic workup of AKI in patients with liver disease. Unfortunately, in the absence of noninvasive diagnostics, histological analysis of kidney biopsy specimens is the only established diagnostic test for CN. From the nephrologist’s perspective, performing a kidney biopsy requires that the course of the suspected disease would likely be improvable with therapy and that the risks of the procedure are acceptable, considering the individual patient’s clinical context. In CKD patients, kidney biopsy is suggested for the further investigation of hematuria and/or severely increased proteinuria. Kidney biopsies in patients with suspected “intrinsic” AKI have recently been encouraged, especially in research settings or in case of high diagnostic uncertainty and—similar to CKD—therapeutic relevance.5 In chronic liver disease, however, kidney biopsies are accompanied by severe safety concerns, as indicated by previous series2 and the data of Bräsen and Mederacke et al.1: 5% (4/79) of patients had serious complications, with 2 potentially life‐threatening bleeding events requiring coiling or even surgery. Of note, a considerably better safety profile was obtained using a transjugular approach.2 The clinical context in which CN occurred in the Bräsen and Mederacke et al.1 study should be discussed. Only 2 patients had been diagnosed with ACLD/cirrhosis (and acute‐on‐chronic liver failure [ACLF]), while 3 other patients would currently be classified as ACLF in the absence of cirrhosis. The other 3 patients had drug‐induced liver injury, recurrent cholangitis, and benign recurrent intrahepatic cholestasis. The clinical course of renal dysfunction was highly dependent on the etiology of the underlying liver disease/cause of cholestasis: while 3 patients who underwent specific therapies for liver disease showed a decline in bilirubin and resolved AKI, all other patients required renal replacement therapy. Finally, Bräsen and Mederacke et al.1 also confirmed the loss of aquaporin 2 (AQP2) expression in the collecting ducts in patients with CN, a finding which has been observed in mice undergoing common bile duct ligation (CBDL).6 However, this observation was not restricted to patients with CN but was also observed in patients with AKI and hyperbilirubinemia, without histological proof for CN. Features specific for CN (i.e., intratubular bile casts) are therefore not a condicio sine qua non because they may be easily missed on histology, or CN may be only the tip (i.e., an extreme phenotype) of a cholestasis‐induced renal dysfunction iceberg. The findings of the present study have important implications for future practice and research. When patients with ACLD/ACLF are diagnosed with AKI stage 1B or higher, they should receive albumin and the conventional diagnostic workup aiming to exclude non‐HRS‐AKI. In the absence of an improvement in renal function and evidence for non‐HRS‐AKI, patients should further be treated with terlipressin (or other vasopressors). Even in patients with evidence suggestive of non‐HRS‐AKI who are nonresponsive to basic measures, many physicians will consider a trial of vasopressors to rule out an additional functional component. This is based on the experience that, despite clear diagnostic criteria, renal dysfunction in ACLD/ACLF is not dichotomous but rather a continuum, which might be explained by the limited ability of currently available diagnostic methods to differentiate between the types of AKI. Importantly, bile acid (BA) toxicity could be an important factor underlying nonresponsiveness to terlipressin. In a study by Nazar and colleagues,7 hyperbilirubinemia ≥10 mg/dL was the only independent predictor of nonresponse in patients with hepatorenal syndrome type 1 (i.e., AKI meeting the criteria of hepatorenal syndrome [HRS‐AKI]), besides the change in mean arterial pressure. However, it is difficult to assess whether the association was causal as hyperbilirubinemia ≥12 mg/dL denotes an additional organ failure, according to the Chronic Liver Failure Consortium Sequential Organ Failure Assessment criteria, and response rates to terlipressin gradually decrease with increasing ACLF severity.8 Accordingly, the link between hyperbilirubinemia and nonresponsiveness to terlipressin may also be interpreted as a consequence of multiple organ dysfunction syndrome not amendable by medical treatment. In addition, most patients with CN might have been excluded from clinical trials on HRS‐AKI due to urinalysis abnormalities. CN, however, may be common in non‐HRS‐AKI patients with evidence of acute tubular necrosis (ATN), the most common form of intrarenal AKI without specific treatment options and a particularly bad prognosis. Although urinary markers such as neutrophil gelatinase‐associated lipocalin show some promise in differentiating between the (presumably) “functional” HRS‐AKI and “structural” non‐HRS‐AKI/ATN, it seems unlikely that these unspecific markers are clinically helpful for differentiating between CN and other forms of tubular epithelial injury. The development of noninvasive diagnostics for CN is particularly important in light of emerging therapies for this long neglected condition. Studies in CBDL mice indicated a pivotal role of urinary excreted toxic BAs in the pathogenesis of CN.6 Accordingly, modulation of the BA pool toward a hydrophilic and less toxic BA profile, e.g., by norursodeoxycholic acid administration, prevented the worsening of renal function in experimental cholestasis.6 The role of BA signaling through the farnesoid X receptor and G protein–coupled BA receptor 1, also known as Takeda G protein–coupled receptor 5, in renal (patho)physiology is now increasingly recognized. Interestingly, the respective agonists have been shown to ameliorate diabetic nephropathy and to up‐regulate AQP2 expression in rodents.10 Finally, plasma exchange, a promising treatment option for both acute and ACLF, might also lower serum BA levels. In conclusion, there is increasing evidence for CN as an important cause of renal dysfunction in patients with liver disease. Future studies should aim to reveal the full clinical spectrum of cholestasis‐induced renal dysfunction and investigate potential noninvasive markers to identify patients who may benefit from specific therapies.