Hepatobiliary long-term consequences of COVID-19: dramatically increased rate of secondary sclerosing cholangitis in critically ill COVID-19 patients

Background Increasing evidence suggests that secondary sclerosing cholangitis (SSC), which can lead to cirrhosis or liver failure, may be a hepatobiliary long-term complication of COVID-19. The aim of this study was to estimate the frequency and outcome of this COVID-19 sequela and to identify possible risk factors. Methods This observational study, conducted at University Hospital Charité Berlin and Unfallkrankenhaus Berlin, Germany, involved hospitalized patients with COVID-19 pneumonia, including 1082 ventilated COVID-19 patients. We compared COVID-19 patients who developed SSC with a COVID-19 control group by univariate and multivariate analyses. Results SSC occurrence after COVID-19 was observed exclusively in critically ill patients with invasive ventilation, albeit with extreme clustering among them. One in every 43 invasively ventilated COVID-19 patients developed this complication. Risk factors preceding the development of secondary sclerosing cholangitis in critically ill COVID-19 patients (SSC-CIP) were signs of systemic reduced blood oxygen supply (e.g., low PaO2/FiO2, ischemic organ infarctions), multi-organ failure (high SOFA score) at admission, high fibrinogen levels and intravenous ketamine use. Multivariate analysis confirmed fibrinogen and increased plasma lactate dehydrogenase as independent risk factors associated with cholangiopathy onset. The 1-year transplant-free survival rate of COVID-19-associated SSC-CIP was 40%. Conclusions COVID-19 causes SSC-CIP in a substantial proportion of critically ill patients. SSC-CIP most likely develops due to severe tissue hypoxia and fibrinogen-associated circulatory disturbances. A significant increase of patients with SSC-CIP is to be expected in the post-COVID era. Supplementary Information The online version contains supplementary material available at 10.1007/s12072-023-10521-0.


Introduction
A novel coronavirus causing severe acute respiratory syndrome emerged in late 2019.This SARS-CoV-2 is a "classic" respiratory virus, which primarily affects the upper and lower respiratory system.More recent evidence has shown that coronavirus disease 2019 (COVID- 19) is not limited to the airways, but also involves other organ systems, such as skin, kidneys, eyes, and endocrine organs as well as the cardiovascular and central nervous system [1][2][3][4][5][6][7].Gastrointestinal symptoms are also common and liver involvement, in particular, is frequently observed.Liver function test (LFT) abnormalities are found in 14-65% of hospitalized patients with symptomatic COVID-19; they are more significant in severe cases and occur in up to 80% of fatal cases [8][9][10].Early data from the first wave of the pandemic suggested that hepatocellular and mixed type patterns of liver involvement were the predominant form, whereas cholestatic patterns were less common [11][12][13].
However, as new clinical data emerge about COVID-19, there has been a growing focus on cholestatic complications of COVID-19, which are primarily associated with a severe course of COVID-19 [14].Several experimental lines of evidence, including biliary organoid cultures, work with cell-lines, and analyses of the SARS-CoV-2 entry receptor ACE2 in human liver, indicate that SARS-CoV-2 can directly infect and replicate in liver cells, particularly cholangiocytes (i.e., biliary epithelial cells) [9].From a clinical perspective, various authors have reported pronounced biliary tract destruction with the development of secondary sclerosing cholangitis in association with COVID-19 pneumonia, also termed "COVID-19 cholangiopathy" [15][16][17][18][19][20].Previous experience with the occurrence of a secondary sclerosing cholangitis in conjunction with viral diseases was first gathered during the influenza A (H1N1) pandemic of 2009/2010 [21].Based on these results, severe acute respiratory distress syndrome (ARDS) was established as an underlying condition for the development of secondary sclerosing cholangitis in critically ill patients (SSC-CIP).
To date (March 2023), more than 759 million confirmed cases of coronavirus disease 2019 have been reported worldwide (WHO Dashboard), and awareness of the long-term health effects of this disease is growing.Nevertheless, the cholestatic long-term consequences of COVID-19 are not well reflected in the current literature [22].Acquired in the early course of intensive care, COVID-19 cholangiopathy persists from the acute phase of COVID-19 to the time of recovery.Patients with COVID-19 cholangiopathy remain seriously ill, even after complete pulmonary recovery.
The aim of this study was to determine whether COVID-19 promotes a cholangiopathy and if so, at what frequency.Under the hypothetical assumption that this cholangiopathy corresponds to SSC-CIP, we aimed to define risk factors for COVID-19-associated SSC-CIP by comparing these patients to a well-matched COVID-19 control group.To our knowledge, there is no such comparative analysis in the available literature.
SSC-CIP.SSC-CIP is a highly symptomatic disorder of critically ill patients, who present with severe progressive, gamma-GT-enhanced cholestasis.As part of our hospital's standardized treatment strategy for COVID-19 patients, cholestasis parameters were evaluated on admission and during follow-up of all of our COVID-19 patients, and every 1-2 days in ICU patients.Inclusion criteria for SSC-CIP were: A. COVID-19 patients with suspected SSC-CIP, defined as: A1.Increase in GGT > 3 times the upper limit of normal (ULN) and ALP > 1.5 × ULN early (4-9 days) after ICU admission (according to EASL definition "cholestasis").A2.Progressive elevation of cholestasis parameters after A1.A3.Cholestasis peaks approximately 24 to 36 days after intubation.A4.Peak of GGT reaches > 15 × ULN and ALP > 5 × ULN (Table 1).If SSC-CIP was suspected (or uncertainly assignable cases, borderline cholestasis values, etc.), non-invasive differential diagnosis was performed (according to the recommendations of specialist societies, including imaging studies, such as ultrasound, CT, MRI, and MRCP), followed by ERC, if needed.B. The diagnosis of SSC-CIP was confirmed if imaging and/or autopsy revealed the following findings in addition to the aforementioned laboratory chemistry abnormalities: B1.Intrahepatic bile duct destruction (100% of cases), cholangiographic signs of secondary stenoses (in the area of necrotic wall destruction) with prestenotic dilatation ("beaded appearance"), and onset of bile duct injury in the liver periphery.B2.Biliary cast formation (76% of cases).Only confirmed cases were included in the study (23 × by ERC, 1 × by MRCP and 1 × by autopsy).Control subjects consisted of patients from the PaCOVID-19 study cohort, who were retrospectively sampled in a nested case-control fashion.The control group was carefully selected for age (± 5 years) and gender to produce a comparable age and gender distribution of the SSC-CIP and control group.The inclusion criteria for controls were as follows: confirmed SARS-CoV-2 pneumonia, age ≥ 18 years, invasive ventilation, no or only mild cholestasis initially (GGT < 3 × ULN and ALP < 1.5 × ULN), no progression to severe cholestasis (GGT peak of controls were median: 3.9 × ULN; and ALP peak median: 1.3 × ULN, see Table 1), improvement of mild cholestasis during ICU stay, no occurrence of cholestatic liver disease during follow-up (cut-off date: 30 April 2022; after the end of follow-up, the controls or their relatives were contacted by telephone and interviewed).Exclusion criteria for controls were: nasal mask or high-flow oxygen therapy, progressive cholestasis during ICU treatment, increase in GGT > 14 × ULN AP > 4.5 × ULN during the ICU stay (even if this was only detectable temporarily), new cholestatic liver disease in the follow-up period, and unclear bile duct changes in MRCP or ERC.All included patients (SSC-CIP and control group) met the criteria for the Berlin definition of ARDS [24].
As a tertiary university hospital and ARDS center, Charité provides the highest level of critical care for COVID-19 patients, including extracorporeal membrane oxygenation (ECMO) support (level 1 hospital).Accordingly, Charité treated some of the most severely ill COVID-19 patients, and many patients in our cohort were transferred to Charité from other hospitals after their condition worsened.To detect a resulting selection bias and, if any, to mitigate it, patients from a second treatment center (Unfallkrankenhaus Berlin) were also included in our analysis.The cases treated at this level 2 hospital (high level of critical care without ECMO) were recorded anonymously and retrospectively.
After the relevant study results were available, we subsequently included an additional control group (with comparable age and gender distribution as the other groups).To estimate the impact of typical COVID-19 pathophysiology on our results, we selected a control group of ARDS patients due to non-COVID-19 pneumonia.Their data were collected anonymously and retrospectively.Only targeted variables (PaO 2 /FiO 2 , fibrinogen) were recorded.These patients were invasively ventilated for viral or bacterial pneumonia and met the ARDS criteria (all of these controls had tested negative for COVID-19).

Outcome measures
The primary outcome measure was the frequency of SSC-CIP in COVID-19 patients.For analysis of secondary outcome measures, we attempted to identify potential risk factors for SSC-CIP, especially those that might define this form of COVID-19 cholangiopathy as SSC-CIP.

Data collection
Demographic and clinical data (e.g., comorbidities, medications, ventilation data, organ replacement therapy, and laboratory tests) of all patients were extracted from the patient records (for exact list, see supplementary material, Table S1).Any biologically plausible cause of bile duct damage was considered a potential risk factor for SSC-CIP in critically ill COVID-19 patients.Fifty-two potential risk factors were recorded (Supplementary table 1).Only those parameters that manifested before the onset of cholestasis were included in the analysis.Day 1 was defined as the day of intubation.According to the recommendations of the European Association for the Study of the Liver (EASL), the onset of cholestasis was defined as serum alkaline phosphatase (ALP) levels higher than 1.5 times the upper limit of normal (> 1.5 × ULN), gamma-glutamyl transpeptidase (GGT) levels > 3 × ULN, and bilirubin levels > 2 × ULN [25].Lymphopenia was defined as a lymphocyte count of less than 1100 × 10 3 /mm 3 based on other studies and local laboratory reference ranges [26].Thrombocytopenia was defined as a platelet count of less than 150/nl.

Statistical analysis
We compared the characteristics of COVID-19-related ARDS patients with and without SSC-CIP (control group).Continuous variables were expressed as mean ± standard deviation (SD) or median with interquartile range (IQR).The Mann-Whitney U test was used to examine the differences between the two groups.Categorical variables were expressed as percentages (%) and compared across the groups using Fisher's exact test.All, statistical tests were two-sided, and a p value < 0.05 were considered statistically significant.Kruskal-Wallis test was used to compare three different ARDS groups (Table 5).A multivariate analysis by means of logistic regression with stepwise selection (significance level p < 0.05) was performed to identify risk factors for the development of SSC-CIP.Considering the two relevant pathogenic hypotheses for SSC-CIP development ("toxic bile" and "biliary ischemia"), and to avoid overloading the model, only selected variables identified in the univariate analysis were included in the multivariate analysis.
For survival analysis, we used the Kaplan-Meier method.

Frequency of SSC-CIP among COVID-19 patients
In a large COVID-

Baseline characteristics of COVID-19 patients with and without SSC-CIP (controls) were equivalent
To identify risk factors for the development of SSC-CIP, we included mechanically ventilated COVID-19 patients without evidence of SSC-CIP as the control group (comparable in age and gender distribution).COVID-19 patients with and without SSC-CIP had similar baseline characteristics (Table 1).Those with SSC-CIP were predominantly males (72%) with a median age of 59 years (range 26-71) and a predominance (48%) of obesity (BMI > 30 kg/m 2 ).Comorbidities were common in the SSC-CIP group (88%) and the control group (80%).However, there were no significant differences between the groups in terms of the prevalence of hypertension, chronic lung disease, or other comorbidities.

COVID-19 patients with SSC-CIP were more likely to receive ketamine
Associations between certain medications and cholangiopathies have been reported.Therefore, we compared the medications that the patients had been taking prior to hospital admission or in the ICU before onset of the first symptoms of SSC-CIP.The analysis revealed that the drugs taken by COVID-19 patients with and without SSC-CIP prior to hospital admission and in the initial ICU period were generally comparable (Table 2).However, ketamine use was significantly more common in the SSC-CIP group (n = 24; 96%) than in the control group (n = 17; 68%, P = 0.023, Fig. 1).

SSC-CIP development was associated with a significantly greater severity of hypoxemia and of COVID-19 disease
The SSC-CIP and control groups differed in terms of illness severity, as determined using Sequential Organ Failure Assessment (SOFA) score (

Cytokine storm and coagulation parameters were roughly comparable between the groups
We also sought evidence of differences in classical inflammation parameters between the two groups that would support the toxic bile concept (Table 4).Mechanically ventilated COVID-19 patients with and without SSC-CIP had comparable initial inflammation levels on admission (Table 4), as indicated by C-reactive protein (CRP), procalcitonin (PCT), interleukin-6 (IL-6), and ferritin.Clinical and laboratory parameters indicative of a hypercoagulable state include D-dimers, fibrinogen, prothrombin time (PT), platelet count, and the rate of thrombotic events.Median plasma fibrinogen levels on admission were significantly higher in the patients who developed SSC-CIP than in those who did not (8.0 g/L [IQR 7.1-9.0]vs. 6.7 g/L [IQR 6.2-7.6];p = 0.008).However, the other coagulation parameters mentioned above did not differ significantly between the groups.

Diminished blood and oxygen supply to the tissues precedes the development of SSC-CIP
Organ infarctions are considered a consequence of diminished blood or oxygen supply to the tissues.Interestingly, the number of organ infarctions (heart, testis, spleen, and intestinal ischemia) was significantly higher in the SSC group (24% vs. 0%, p = 0.022).We noted striking differences in hypoxia markers between the groups.For example, initial hypoxia, as measured using the oxygenation index (see above) was significantly more severe in SSC-CIP patients than in controls.Moreover, COVID-19 patients who developed SSC-CIP had significantly higher creatine kinase (CK) levels than the controls.CK elevation occurred in 96% of SSC-CIP-positive patients compared to 60% of SSC-CIP-negative controls.Likewise, median myoglobin values were significantly higher in SSC-CIP patients, and there was a trend toward higher LDH values at ICU admission (p = 0.055) in the SSC-CIP group.Fibrinogen is not only a parameter of coagulation; its effect goes beyond that.High fibrinogen levels are associated with increased plasma viscosity, which can impair organ perfusion, leading to ischemia and infarction.Our COVID-19 patients who developed SSC had significantly higher fibrinogen levels (Table 4).S2).No independent associations were found for oxygenation index (PaO 2 /FiO 2 ), SOFA score, or CK.Also, no significant associations were observed for inflammation markers like CRP, PCT, ferritin, IL-6, and lymphopenia.

Frequency and outcome
The data from this study draw attention to a clinically relevant hepatobiliary long-term consequence of COVID-19: COVID-19 increases the risk of secondary sclerosing cholangitis (SSC).SSC development was observed exclusively in invasively ventilated, critically ill COVID-19 patients (SSC-CIP).This cohort study revealed a dramatically increased rate of SSC-CIP in these patients and suggests that this complication is more likely due to established risk factors for SSC-CIP rather than to viral tropism or to a specific SARS-CoV-2-related etiology.Our observations confirm and extend the findings of recent case reports [15][16][17][18].We found that SSC-CIP occurs at a rate of one in every 43 ICU patients on invasive ventilation.Compared to recently published data from the pre-pandemic era (1 case of SSC per 1990 ventilated patients, incidence: 0.05/100; 95%CI: 0.001-0.280)[27], the incidence of SSC-CIP is more than 46 times higher in critically ill COVID-19 patients (IRR 46.0; 95%CI: 7.5-1887.8).This is of great relevance in view of the resulting burden on healthcare systems around the world after the COVID-19 pandemic.The long and often complicated course of the disease and limited therapeutic options pose major challenges to hepatologists.Liver transplantation is the only curative treatment option, and previous reports indicate that up to 22% of all SSC-CIP patients require transplantation [15][16][17][18]28].Thus, an increasing need for liver transplantation in these relatively young post-COVID-19-SSC patients (mean age 56 years) can be expected, posing new challenges for transplantation medicine.Without transplantation, patients with COVID-19-associated SSC-CIP have a poor prognosis: only 40.0% of our patients were alive 1 year after SSC-CIP onset.
The clinical features of COVID-19-related SSC-CIP closely resemble those of our previous non-COVID cohort [29].The available clinical data provide little-to-no evidence that post-COVID-19 SSC is a new disease entity originating primarily in the bile ducts, e.g., by direct SARS-CoV-2-mediated damage to cholangiocytes [16].Instead, the disease shows most clinical and endoscopic features of classical secondary sclerosing cholangitis in critically ill patients (supplementary Fig. 1).

Causes of SSC-CIP in COVID-19 patients
The extreme clustering of SSC-CIP among COVID-19 patients with ARDS raises the question of why these patients have a particular predisposition to SSC-CIP.Currently, there are two main concepts that might explain the pathophysiological mechanisms leading to SSC-CIP: The first theory is that hyperinflammation and a cytokine storm during critical illness may induce a dysregulation of the hepatobiliary transporter system, resulting in an altered bile composition (toxic bile concept).The second is that local ischemia of the biliary epithelium occurs during critical illness, leading to subsequent destruction of the bile ducts (ischemia concept) [27].

The "toxic bile concept" does not explain the increased incidence of SSC in critically ill COVID-19 patients
Inflammation biomarker levels in the present study were higher than those for critically ill COVID-19 patients in other studies [30].This applies to our control group as well as to the SSC-CIP group.SARS-COV-2 triggers a cytokine storm, characterized by excessive elevation of interleukin-1β, interleukin-6, IP-10, TNF, interferon-γ, macrophage inflammatory protein (MIP) 1α and 1β, and vascular endothelial growth factor (VEGF).In COVID-19, IL-6 is the most prominently elevated cytokine and elevated IL-6 levels are strongly associated with poorer outcomes [31,32].Cytokine storm-induced dysregulation of the hepatobiliary transporter system can lead to impaired bile flow at the level of the small bile ducts (ductular cholestasis) or impaired uptake, transport and excretion of bile acids and bilirubin by the hepatocytes (hepatocellular cholestasis) [33][34][35][36].Theoretically, the altered bile composition due to excessive cytokine release might jeopardize the integrity of bile duct epithelia.However, we found no evidence of significant differences in laboratory markers of cytokine storm (e.g., plasma levels of IL-6, CRP, PCT, and initial ferritin) between ventilated COVID-19 patients with and without SSC-CIP.This is in agreement with the other publications [14].Thus, our results do not support the theory of cytokinemediated bile duct destruction in SSC-CIP development in COVID-19 patients.
The "ischemia concept" can explain the increased incidence of SSC in critically ill COVID-19 patients Shock and systemic hypoxia as well as local circulatory disturbances in the supplying peribiliary plexus (PBP) can contribute to bile duct ischemia in critically ill patients.Vasculitis, microthrombi, and hyperviscosity can lead to impaired blood flow in the PBP.Our COVID-19 patients who developed SSC-CIP had significantly more severe initial hypoxia, as measured using PaO 2 /FiO 2 as a surrogate marker.As a consequence of severe hypoxia in COVID-19-related ARDS, the oxygen supply to the tissues is insufficient.Global tissue hypoxia may promote the development of multi-organ failure (MOF).As expected, SOFA scores on ICU admission, a measure of multi-organ failure, were significantly higher in SSC-CIP patients than in controls.The need for organ support as an expression of MOF differed accordingly.Renal replacement therapy due to acute kidney injury in the initial ICU course was significantly more common in COVID-19 patients with SSC-CIP than in those without it.To obtain further evidence of general tissue hypoxia, we compared the frequency of organ infarctions between the two groups.COVID-19 patients with SSC-CIP had significantly more organ infarctions (e.g., intestinal ischemia, splenic and testicular infarction) than the controls.Taken together, our results suggest that bile duct ischemia in the context of initial hypoxemic multi-organ failure may be a trigger mechanism precipitating SSC-CIP in COVID-19 patients.The multivariate analysis did not corroborate the results of the univariate analysis regarding the association of disease severity (PaO 2 /FiO 2 and SOFA score) with the occurrence of SSC-CIP.The large number of variables in combination with the small sample size may hamper the power of the multivariate analysis for this question.
The results of our univariate and multivariate analyses indicate that fibrinogen plays a special role in the development of SSC-CIP."Hyperfibrinogenemia" is associated with hyperviscosity, especially in COVID-19 patients [37].Increased susceptibility of the bile ducts to hyperviscosity is a well-known phenomenon in the context of liver transplantation.Hyperviscosity is associated with ischemic bile duct lesions [38] as it causes disturbances of the microvascular circulation and tissue hypo-perfusion, damages the endothelium, and poses a risk factor for thrombosis.Evidence regarding the role of the vascular component in the pathology of COVID-19 is compelling [39,40].The blood of critically ill COVID-19 patients often has a unique stickiness, and the characteristics of COVID-19-associated coagulopathy are distinctly different from those ever seen in other critical illnesses [41,42].Subsequent thrombotic events like thrombi in the microvasculature are a hallmark of COVID-19.Tsutsumi et al. observed an association between elevated fibrinogen levels and liver involvement in COVID-19 and concluded that liver dysfunction in COVID-19 patients might be induced by microvascular thrombosis [43].This hypothesis is supported by the results of a histopathological study by Kondo et al. [44].Although the exact mechanism has yet to be proven, hyperfibrinogenemia may contribute to hypoxemia of the bile ducts and is an independent risk factor for SSC-CIP development in critically ill patients.

Other possible mechanisms
Our COVID-19 patients who developed SSC-CIP received ketamine significantly more often before the onset of cholestasis (96%) than those in the control group (68%).These data are in line with our previous observations in non-SSC-CIP COVID-19 patients [27].A possible association between ketamine use and cholangiopathies in COVID-19 patients was also reported in a multicenter study by French researchers [18,45].The fact that ketamine can cause bile duct damage is known from numerous publications, especially from Asian countries, but the underlying mechanisms are unclear.Based on the results of our study, an unfavorable effect of ketamine on the development of SSC-CIP cannot be excluded.However, as one patient in our SSC-CIP group did not receive ketamine, it seems unlikely that ketamine was the sole cause of SSC-CIP development.It is also possible that the correlation is due to confounding factors, because COVID-19 patients with severe refractory ARDS tend to require deeper sedation, often in combination with ketamine, to enable prone positioning for escalation of treatment.However, the findings suggesting an association between ketamine and SSC-CIP need to be validated in future studies.Similarly, prone positioning was used significantly more often in patients with SSC.Prone positioning may further worsen the blood and oxygen supply to the bile ducts.On the other hand, it is notable that especially those patients with severe hypoxia who respond inadequately to conservative ventilation strategies are ventilated in prone position.Therefore, the presence of confounding is conceivable.These questions might be better addressed in studies involving patients who develop SSC-CIP as a result of prolonged shock rather than hypoxia.
According to our data, positive end-expiratory pressure (PEEP) and catecholamines do not appear to be independent risk factors for the development of SSC-CIP.Likewise, our results regarding the use of various medications do not support the hypothesis of drug-related cytotoxic effects on biliary epithelia.As the first controlled trial of SSC-CIP, our study provides important insights into the general pathogenetic mechanisms of SSC-CIP and evidence in support of the ischemic bile duct damage hypothesis favored by most authors.

Others
The fact that COVID-19-associated SSC-CIP has not been previously reported in Asia may suggest a genetic predisposition.However, the fact that our SSC-CIP cohort included one patient of Asian origin and one of African origin does not support this assumption.

Conclusion
Critically ill COVID-19 patients have an increased risk of developing secondary sclerosing cholangitis (SSC-CIP).This hepatobiliary long-term consequence of COVID-19 is common, affecting one in every 43 invasively ventilated patients.Our comparison of COVID-19 patients with non-COVID-19 ARDS controls suggests that the typical COVID-19 milieu of fibrinogen elevation in conjunction severe hypoxia may be conducive to the development of an SSC-CIP.This would also explain the jump in incidence of SSC-CIP due to COVID-19.Our study indicates that post-COVID-19 SSC-CIP is a hypoxic/ischemic event that may possibly be exacerbated by fibrinogen-associated hyperviscosity.In view of its high incidence and poor survival rate, COVID-19-associated SSC-CIP poses a major challenge for hepatologists.

Fig. 1
Fig. 1 Selected risk factors for secondary sclerosing cholangitis in critically ill COVID-19 patients (prior the onset of cholestasis).Critically ill COVID-19 patients who developed SSC-CIP (n = 25, red bar graphs; SSC-CIP group) versus those who did not (n = 25; control group).Bar graphs show the proportion of patients with the selected risk factor in the group.A Ketamine use, B Prone position, C Organ infarction, and D Renal replacement therapy.* p < 0.05, ** p < 0.01

Table 1
Baseline and clinical characteristics of COVID-19 patients with (cases) and without SSC-CIP (controls)

Table 2
Drug use prior to COVID-19 and medications during initial ICU course ‡ Incomplete data the Berlin definition occurred in thirteen SSC-CIP-positive vs. six SSC-CIP-negative cases.Hence, 13 SSC-CIP patients and 6 non-SSC-CIP controls required ECMO.

Table 3
Symptoms of organ failure, comparison of the two groups p-values <0.05 were considered as significant and highlighted in bold § Incomplete data from 2 patients ¶ Including acute and acute-on-chronic kidney failure ARDS Acute respiratory distress syndrome, ECMO Extracorporeal membrane oxygenation, IQR Interquartile range, MAP Mean arterial pressure, n.a.not applicable, SD Standard deviation, SOFA score Sequential organ failure assessment score, SSC-CIP Secondary sclerosing cholangitis in critically ill patients, PEEP Positive end-expiratory pressure, PT Prothrombin time, VTE Venous thromboembolism PEEP Positive end-expiratory pressure, PT Prothrombin time, VTE Venous thromboembolism 1 Fisher's exact test unless otherwise stated

Table 4
Selected risk factors for SSC-CIP, results of univariate analysis p-values <0.05 were considered as significant and highlighted in bold CI Confidence Interval, CK Creatine kinase, CRP C-reactive protein, IL-6 Interleukin-6, IQR Interquartile range, LDH Lactate dehydrogenase, OR Odds ratio, SOFA score Sequential organ failure assessment score, SSC-CIP Secondary sclerosing cholangitis in critically ill patients, PT Prothrombin time, PCT Procalcitonin 1 Mann-Whitney U test unless otherwise stated 2 Fisher´s exact test *