Influence of non-pathological factors on ALT measurements

Alanine transaminase (ALT) is still the most sensitive marker for evaluating the damage of liver cells. ALT measurements may be affected by common non-pathological factors such as gender, age, pregnancy, different physiological conditions, circadian rhythm, diet, drinking, smoking, strenuous exercise, metabolic factors, drugs, blood collection methods, and specimen collection, transmission, and preservation. It is important to emphasize the definition of limits of normal ALT and the influence of the above-mentioned non-pathological factors on its test results.

Histologic evidence of active liver injury in chronic hepatitis B patients with normal range or minimally elevated alanine aminotransferase levels

GOALS

To evaluate the proportion of patients with histologic evidence of active liver disease (HEALD) who have chronic hepatitis B (CHB) and normal/minimally elevated serum alanine aminotransferase (ALT). Sub-analysis was performed to determine whether HEALD based upon liver biopsy better correlates with ALT using modified ALT (30 men/19 women) upper limit of normal (ULN) criteria compared with local conventional laboratory.

BACKGROUND

There are limited data on CHB with normal range ALT (NRALT). We designed a study to evaluate histologic damage in this cohort of patients.

STUDY

A retrospective, multicenter study evaluated CHB patients with normal/minimally elevated ALT [< or = 1.2 x ULN (hepatitis B e antigen positive) or < or = 1.5 x ULN (hepatitis B e antigen negative)]. Liver biopsy specimens were reviewed by an independent histopathologist. HEALD was defined as Knodell necroinflammatory score greater than 5 and Ishak fibrosis stage greater than 1.

RESULTS

Forty-five patients met criteria: median age of 40 years; 51% males; 73% Asian; and 67% hepatitis B e antigen negative. Median hepatitis B virus DNA was 6.04 log10 copies/mL, aspartate aminotransferase (AST) 30 IU/L, and ALT 42 IU/L; and 40% of the patients had ALT greater than ULN. Overall, 20% had HEALD and among patients with NRALT, 4 of 27 (15%) and 0 of 5 (0%) had HEALD through conventional or modified ALT ULN, respectively.

CONCLUSIONS

One fifth of patients with CHB and normal/minimally elevated ALT had HEALD. Among the subset of patients with NRALT, 15% (4 of 27) had HEALD when using conventional laboratory compared with 0% (0 of 5) patients by modified ALT ULN criteria. Use of the modified ALT ULN will likely improve accuracy in identifying patients who may have HEALD compared with conventional laboratory ULN.

Validity of a point of care device (Cholestech LDX) to monitor liver enzyme activity (aminotransferase measures) during a clinical trial

BACKGROUND

Serum transaminase activity is a common measure of liver injury used in clinical trials. The use of a point of care device to monitor serum transaminases would allow immediate evaluation of this safety endpoint and may be less expensive than standard laboratory testing.

PURPOSE

The objective of this study was to compare a point of care transaminase test to a standard laboratory measurement.

METHODS

Subjects were healthy adults participating in a clinical trial measuring the effects of therapeutic doses of acetaminophen on serum transaminase activity. For this study, serum transaminase activity was determined every 3days for 14days. At each measurement, a sample was sent to the clinical laboratory for measurement and also analyzed using a point of care device (Cholestech LDX, Hayward CA). The results were compared using a Bland-Altman plot to identify bias and we also measured the agreement between the techniques for categorizing samples as "low", "normal" or "elevated".

RESULTS

One hundred thirty three samples from 35 subjects were compared. The 95% confidence interval for the limits of agreement between the LDX and the clinical laboratory was -8.46 to 27.09IU/L. Agreement for classification as low, normal or elevated was moderate between the two methods (Kappa=0.37).

CONCLUSION

The point of care device provided moderate agreement with the laboratory transaminase measurement. However, use of the device would have resulted in misclassification of approximately 1/3 of samples. We do not recommend this point of care device for the measurement of serum transaminases in clinical trials.