Scoring Method Predicts Death, Transplant in Liver Failure

A scoring method that combines bilirubin and lactate values with the specific etiology of acute liver failure better predicts death or the need for transplant than do existing methods, Dr. Johannes Hadem and his colleagues reported.

The new scoring method, known as the bilirubin-lactate-etiology (BiLE) measure, has a sensitivity of 79% and a specificity of 84% for predicting death or the need for transplant if the patient's score is above 6.9. These sensitivity and specificity values are significantly better than are other measures including the Model for End-Stage Liver Disease (MELD) and the Simplified Acute Physiology Score III (SAPS-III), according to Dr. Hadem of Hannover (Germany) Medical School, and his colleagues (Clin. Gastroenterol. Hepatol. 2008;6:339–45).

The BiLE score is simple to calculate and is especially suited to bedside use immediately after ICU admission, the investigators wrote. But because of the diversity of acute liver failure etiologies in different regions of the world, the BiLE score will need to be validated in other centers and with other patient cohorts.

To develop the BiLE score, the investigators conducted a retrospective analysis of 102 ICU patients from a single institution who fulfilled the diagnostic criteria for acute liver failure. Of those, 39 survived for at least 8 weeks without the need for orthotopic liver transplant (OLT), 18 died without OLT, 5 died following OLT, and 40 survived following OLT. In all, 79 of the patients (77%) survived to week 8.

For the purposes of the study, patients with hepatic dysfunction were diagnosed with acute liver failure if they had hepatic encephalopathy, acute-onset increase of the International Normalized Ratio (INR) above 1.5, and the absence of signs of chronic liver disease during the clinical and ultrasound examinations.

There was no predominant etiology for acute liver failure among the patients in the study. Cryptogenic acute liver failure was the etiology in 21 patients, acute hepatitis B in 18, acetaminophen ingestion in 16, Budd-Chiari syndrome in 9, phenprocoumon toxicity in 7, idiosyncratic drug reactions in 5, Amanita phalloides ingestion in 5, Wilson's disease in 5, hepatitis A in 4, ischemic hepatitis in 4, and halothane reaction in 3. The remaining 5 patients had etiologies classified as “other.”

In comparing patients who survived with those who died or required OLT, the investigators found that 15 different laboratory values and other characteristics showed statistically significant differences. Multivariate linear regression revealed that bilirubin and lactate levels were the most predictive of survival.

Patients who survived without transplantation had a mean bilirubin level of 103 micromol/L, vs. 263 micromol/L in the liver transplantation or death group. Similarly, patients who survived without transplantation had a mean lactate level of 2.9 mmol/L, vs. 4.7 mmol/L in the liver transplantation or death group.

Patients with cryptogenic acute liver failure, Budd-Chiari syndrome, or phenprocoumon toxicity were more likely to die or require transplant, while those with acetaminophen toxicity were more likely to survive without the need for transplant.

The investigators designed the BiLE score empirically. To bring bilirubin and lactate into the same range of values, the equation calls for dividing bilirubin concentrations in micromol/L by 100. To this figure, one adds the lactate concentration in mmol/L and then adds or subtracts a value depending on the etiology. (See box.)

Using a cutoff value of 6.9 to predict death or the need for transplant, the BiLE score achieved a sensitivity of 79%, a specificity of 84%, a positive predictive value of 89%, and a negative predictive value of 71%. Sensitivity was 100% in patients with cryptogenic acute liver failure.

In contrast, lactate alone with a cutoff of 3.5 mmol/L achieved a sensitivity and specificity of 59% and 66%, respectively. The MELD score with a cutoff of 32 achieved a sensitivity and specificity of 65% and 69%, respectively, and the King's College Criteria achieved a sensitivity and specificity of 58% and 82%, respectively.

The investigators stated that they had no conflicts of interest to report.

Calculation of the BiLE Score

The calculation method for the BiLE score is as follows:

Bilirubin (micromol/L)/100 + lactate (mmol/L)

+ 4 (in the case of cryptogenic acute liver failure, Budd-Chiari syndrome, or phenprocoumon toxicity)

− 2 (in the case of acetaminophen toxicity)

+ 0 (in the case of other etiologies)

BiLE scores above 6.9 are predictive of death or liver transplantation.

A scoring method that combines bilirubin and lactate values with the specific etiology of acute liver failure better predicts death or the need for transplant than do existing methods, Dr. Johannes Hadem and his colleagues reported.

The new scoring method, known as the bilirubin-lactate-etiology (BiLE) measure, has a sensitivity of 79% and a specificity of 84% for predicting death or the need for transplant if the patient's score is above 6.9. These sensitivity and specificity values are significantly better than are other measures including the Model for End-Stage Liver Disease (MELD) and the Simplified Acute Physiology Score III (SAPS-III), according to Dr. Hadem of Hannover (Germany) Medical School, and his colleagues (Clin. Gastroenterol. Hepatol. 2008;6:339–45).

The BiLE score is simple to calculate and is especially suited to bedside use immediately after ICU admission, the investigators wrote. But because of the diversity of acute liver failure etiologies in different regions of the world, the BiLE score will need to be validated in other centers and with other patient cohorts.

To develop the BiLE score, the investigators conducted a retrospective analysis of 102 ICU patients from a single institution who fulfilled the diagnostic criteria for acute liver failure. Of those, 39 survived for at least 8 weeks without the need for orthotopic liver transplant (OLT), 18 died without OLT, 5 died following OLT, and 40 survived following OLT. In all, 79 of the patients (77%) survived to week 8.

For the purposes of the study, patients with hepatic dysfunction were diagnosed with acute liver failure if they had hepatic encephalopathy, acute-onset increase of the International Normalized Ratio (INR) above 1.5, and the absence of signs of chronic liver disease during the clinical and ultrasound examinations.

There was no predominant etiology for acute liver failure among the patients in the study. Cryptogenic acute liver failure was the etiology in 21 patients, acute hepatitis B in 18, acetaminophen ingestion in 16, Budd-Chiari syndrome in 9, phenprocoumon toxicity in 7, idiosyncratic drug reactions in 5, Amanita phalloides ingestion in 5, Wilson's disease in 5, hepatitis A in 4, ischemic hepatitis in 4, and halothane reaction in 3. The remaining 5 patients had etiologies classified as “other.”

In comparing patients who survived with those who died or required OLT, the investigators found that 15 different laboratory values and other characteristics showed statistically significant differences. Multivariate linear regression revealed that bilirubin and lactate levels were the most predictive of survival.

Patients who survived without transplantation had a mean bilirubin level of 103 micromol/L, vs. 263 micromol/L in the liver transplantation or death group. Similarly, patients who survived without transplantation had a mean lactate level of 2.9 mmol/L, vs. 4.7 mmol/L in the liver transplantation or death group.

Patients with cryptogenic acute liver failure, Budd-Chiari syndrome, or phenprocoumon toxicity were more likely to die or require transplant, while those with acetaminophen toxicity were more likely to survive without the need for transplant.

The investigators designed the BiLE score empirically. To bring bilirubin and lactate into the same range of values, the equation calls for dividing bilirubin concentrations in micromol/L by 100. To this figure, one adds the lactate concentration in mmol/L and then adds or subtracts a value depending on the etiology. (See box.)

Using a cutoff value of 6.9 to predict death or the need for transplant, the BiLE score achieved a sensitivity of 79%, a specificity of 84%, a positive predictive value of 89%, and a negative predictive value of 71%. Sensitivity was 100% in patients with cryptogenic acute liver failure.

In contrast, lactate alone with a cutoff of 3.5 mmol/L achieved a sensitivity and specificity of 59% and 66%, respectively. The MELD score with a cutoff of 32 achieved a sensitivity and specificity of 65% and 69%, respectively, and the King's College Criteria achieved a sensitivity and specificity of 58% and 82%, respectively.

The investigators stated that they had no conflicts of interest to report.

Calculation of the BiLE Score

The calculation method for the BiLE score is as follows:

Bilirubin (micromol/L)/100 + lactate (mmol/L)

+ 4 (in the case of cryptogenic acute liver failure, Budd-Chiari syndrome, or phenprocoumon toxicity)

− 2 (in the case of acetaminophen toxicity)

+ 0 (in the case of other etiologies)

BiLE scores above 6.9 are predictive of death or liver transplantation.

A scoring method that combines bilirubin and lactate values with the specific etiology of acute liver failure better predicts death or the need for transplant than do existing methods, Dr. Johannes Hadem and his colleagues reported.

The new scoring method, known as the bilirubin-lactate-etiology (BiLE) measure, has a sensitivity of 79% and a specificity of 84% for predicting death or the need for transplant if the patient's score is above 6.9. These sensitivity and specificity values are significantly better than are other measures including the Model for End-Stage Liver Disease (MELD) and the Simplified Acute Physiology Score III (SAPS-III), according to Dr. Hadem of Hannover (Germany) Medical School, and his colleagues (Clin. Gastroenterol. Hepatol. 2008;6:339–45).

The BiLE score is simple to calculate and is especially suited to bedside use immediately after ICU admission, the investigators wrote. But because of the diversity of acute liver failure etiologies in different regions of the world, the BiLE score will need to be validated in other centers and with other patient cohorts.

To develop the BiLE score, the investigators conducted a retrospective analysis of 102 ICU patients from a single institution who fulfilled the diagnostic criteria for acute liver failure. Of those, 39 survived for at least 8 weeks without the need for orthotopic liver transplant (OLT), 18 died without OLT, 5 died following OLT, and 40 survived following OLT. In all, 79 of the patients (77%) survived to week 8.

For the purposes of the study, patients with hepatic dysfunction were diagnosed with acute liver failure if they had hepatic encephalopathy, acute-onset increase of the International Normalized Ratio (INR) above 1.5, and the absence of signs of chronic liver disease during the clinical and ultrasound examinations.

There was no predominant etiology for acute liver failure among the patients in the study. Cryptogenic acute liver failure was the etiology in 21 patients, acute hepatitis B in 18, acetaminophen ingestion in 16, Budd-Chiari syndrome in 9, phenprocoumon toxicity in 7, idiosyncratic drug reactions in 5, Amanita phalloides ingestion in 5, Wilson's disease in 5, hepatitis A in 4, ischemic hepatitis in 4, and halothane reaction in 3. The remaining 5 patients had etiologies classified as “other.”

In comparing patients who survived with those who died or required OLT, the investigators found that 15 different laboratory values and other characteristics showed statistically significant differences. Multivariate linear regression revealed that bilirubin and lactate levels were the most predictive of survival.

Patients who survived without transplantation had a mean bilirubin level of 103 micromol/L, vs. 263 micromol/L in the liver transplantation or death group. Similarly, patients who survived without transplantation had a mean lactate level of 2.9 mmol/L, vs. 4.7 mmol/L in the liver transplantation or death group.

Patients with cryptogenic acute liver failure, Budd-Chiari syndrome, or phenprocoumon toxicity were more likely to die or require transplant, while those with acetaminophen toxicity were more likely to survive without the need for transplant.

The investigators designed the BiLE score empirically. To bring bilirubin and lactate into the same range of values, the equation calls for dividing bilirubin concentrations in micromol/L by 100. To this figure, one adds the lactate concentration in mmol/L and then adds or subtracts a value depending on the etiology. (See box.)

Using a cutoff value of 6.9 to predict death or the need for transplant, the BiLE score achieved a sensitivity of 79%, a specificity of 84%, a positive predictive value of 89%, and a negative predictive value of 71%. Sensitivity was 100% in patients with cryptogenic acute liver failure.

In contrast, lactate alone with a cutoff of 3.5 mmol/L achieved a sensitivity and specificity of 59% and 66%, respectively. The MELD score with a cutoff of 32 achieved a sensitivity and specificity of 65% and 69%, respectively, and the King's College Criteria achieved a sensitivity and specificity of 58% and 82%, respectively.

The investigators stated that they had no conflicts of interest to report.

Calculation of the BiLE Score

The calculation method for the BiLE score is as follows:

Bilirubin (micromol/L)/100 + lactate (mmol/L)

+ 4 (in the case of cryptogenic acute liver failure, Budd-Chiari syndrome, or phenprocoumon toxicity)

− 2 (in the case of acetaminophen toxicity)

+ 0 (in the case of other etiologies)

BiLE scores above 6.9 are predictive of death or liver transplantation.