Additional Risk Factors for Acute Kidney Injury Defined in Patients with Cirrhosis

Although acute kidney injury is common in patients with cirrhosis, those who are admitted to the hospital are at increased risk, especially if they are of older age, Child-Pugh Class C, or have ascites or sepsis, according to the results of a new study.

Research indicates that the prevalence of acute kidney injury in people with cirrhosis can range between 14% and 50%. Mortality is also high in these patients, particularly among those who are admitted to the hospital and with the highest stage of acute kidney injury, but firmer estimates of prevalence are not known. As such, investigators led by Raseen Tariq, MD, resident in internal medicine at Rochester General Hospital in Rochester, New York, performed a meta-analysis to determine the prevalence and mortality outcomes in patients with cirrhosis who subsequently developed acute kidney injury and further evaluation of the predictors of acute kidney injury in these patients.

“As a medical resident, I see a lot of patients with acute kidney injury who have cirrhosis,” Tariq said in an interview with MD Magazine®. “Whenever these patients are admitted to the hospital, we don’t know what else we can do for them to prevent liver disease progression, and what the long-term outcomes are going to be for these patients. The questions that we were hoping to answer with this study were, ‘Why did they develop acute kidney injury?’ and ‘What are their long-term outcomes going to be?”

Studies to address this have been completed in the past, but most are retrospective studies and there are no clinical trials. Furthermore, the data are scant in terms of pooled analyses, particularly regarding trends in acute kidney injury in this population.

Tariq and her team performed a literature search up to March 2018 and identified 30 studies that assessed the prevalence or mortality of acute kidney injury in patients with cirrhosis. The Newcastle-Ottawa scale was used to assess the quality of each study and meta-analyses were performed using random effects models, according to the study abstract. P values of <.05 were deemed as significant when assessing heterogeneity and publication bias.

Tariq explained why the team did not find more studies to include in the analysis.

“We excluded a lot of studies because we wanted to include only those studies that actually assessed the prevalence of acute kidney injury,” Tariq said. “We wanted to ensure that every study had specific criteria to diagnose these patients. Most of the studies determined it by serum creatinine at least more than 1.5 times baseline. Our strict inclusion criteria might be the reason why we only found 30 studies to include in the analysis, which is actually more than we were expecting.”

A total of 18,645 patients were included in the studies, of which, 5,744 developed acute kidney injury (pooled prevalence 37.4%, 95% CI, 30.7%-44%). Data from 8 studies revealed that among patients with cirrhosis and acute kidney injury, in-hospital mortality was 6 times in those with acute kidney injury than those without (odds ratio [OR] 6.72, 95% CI 3.47-13, P<.0001; I² ,70%). Furthermore, 7 studies reported higher 30-day mortality in those with acute kidney injury than those without (OR 3.37, 95% CI 2.35-4.84). Mortality at 90 days to a year after the development of acute kidney injury was also high at (OR 4.43, 95% CI 2.93-6.70) and (OR 5.37, 95% CI 2.45-11.79) respectively. For those patients with acute kidney injury who were admitted to the intensive care unit (ICU), mortality was 6 times higher than those without acute kidney injury (OR 5.90, 95% CI 3.21-10.85).

The long-term follow-up data from those studies in terms of mortality was surprising, according to Tariq. “In those patients who developed acute kidney injury, even when they are discharged from the hospital, and the renal dysfunction had resolved, the mortality remained high. The clinicians had treated their renal dysfunction and their underlying etiology (because of infection or ATN, or something else); however, the patients still had increased mortality, even at 1 year follow up.”

Older age (mean difference 0.16, P<.001), Child Pugh Class C (OR 2.51, 95% CI 1.83-3.44), the presence of ascites (OR 2.06, 95% CI 1.25-3.41) and the presence of severe sepsis/septic shock (OR 2.72, 95% CI 1.05-7.06) were all associated with the risk of acute kidney injury. These are not surprising, but Tariq explained that it was important that we now had the data to support it.

“We know these are the risk factors, but it was good to have the data on it,” Tariq explained. “If patients come into the hospital with these risk factors, we now have the data to support that they should be monitored very closely. Not only in terms of bloodwork, but also in terms of urine output.

Variables not associated with a risk of acute kidney injury included model for end-stage liver disease (MELD) score, diabetes, being male, hepatic encephalopathy or bacterial infection on admission, or etiology of cirrhosis (alcoholic vs viral).

Furthermore, investigators found a trend of decreased risk of acute kidney injury in patients with a history of, or current variceal bleed (OR 0.69, 95% CI 0.48-0.99).

Overall, about 37% of the patients with cirrhosis who were admitted to the hospital developed acute kidney injury. Furthermore, among admitted patients who in ICU or non-ICU floor, the risk of in-hospital mortality was 6 times greater in patients with acute kidney injury than those without.

Some of the limitations of this study include that all studies included in the analysis were retrospective studies. Furthermore, in all studies monitored serum creatinine levels as an indicator of acute kidney injury; but, according to Tariq, monitoring urine output is just as important.

“Creatinine is a marker that is read from the muscle and sometimes in patients with cirrhosis, they are so malnourished that their muscle mass is reduced and so their serum creatinine may be falsely low,” Tariq said. “They might have a higher creatinine level than we are seeing. In those cases, urine output may tell us they have much more renal dysfunction or renal injury than we are seeing from the serum creatinine levels.”

According to Tariq, additional prospective studies are warranted to aid in the development of strategies to improve long-term outcomes for patients with cirrhosis who develop acute kidney injury to lower the risk of in-hospital mortality.

“For future studies—and we are planning to do one&mdash;our goal would be to include both the serum creatinine levels and urine output as a marker of renal dysfunction just to ensure that we are not characterizing something that might not be of much significance, or might be of more significance. The other thing we want to consider from this study is, ‘Is there anything that we can do prophylactically to prevent this from happening?’ Our study can be a base for a prospective study to see if patients can receive and albumin infusion or Transjugular Intrahepatic Portosystemic Shunt (TIPS) procedure that reduces the hypo to hypertension to see if we can prevent the development of acute kidney injury and underlying mortality.”

The study, “Prevalence, Outcomes and Predictors of Acute Kidney Injury in Patients with Cirrhosis: A Systematic Review and Meta-Analysis,” was presented at the 2018 American Association for the Study of Liver Diseases (AASLD) Liver Meeting, November 9-13, 2018, in San Francisco, California.