Diagnosing Minimal Hepatic Encephalopathy: Comparison of 6 Frequently Used Tests

Introduction: Minimal hepatic encephalopathy (mHE) is associated with poor health-related quality of life (HRQL) and outcome in patients with liver cirrhosis. Treatment of mHE should improve at least the first. Therefore, screening for mHE in cirrhotic patients is recommended. By now, there is little data on the comparability of frequently used tests for diagnosing mHE. We compared 6 frequently used tools for diagnosing mHE. Methods: One hundred thirty-two patients with liver cirrhosis were recruited, considering the following exclusion criteria: underage, severe neurological comorbidity, impaired kidney function, malignant tumour, previous organ transplantation. All patients underwent a neurological examination and mHE assessment consisting of the PSE-Syndrome Test yielding the PHES (cut-off <-4), the Animal Naming Test (ANT, cut-off <23), the Critical Flicker Frequency (CFF, cut-off <39 Hz), the Inhibitory Control Test (ICT, cut-off >24 weighted lures), the EncephalApp (Stroop, cut-off >190 sec) and the Continuous Reaction Time Test (CRT, cut-off <1.9). The tests were applicated with changing order with each patient. Results: Thirty-one percent of the patients were female; the prevalent etiology was alcohol related and the median age was 57 years. Median MELD was 12 and Child-Pugh-Score 7. Twenty-3 (17.4%) patients presented with HE grade 1-2. Among the 109 neurologically unimpaired patients, mHE assessment yielded pathological results in 35.8% (PHES), 51.4% (ANT), 25.7% (CFF), 36.7% (ICT), 45.0% (Stroop) and 44.0% (CRT). Test results correlated significantly with each other (P<0.05), except of the CFF. The highest correlation was found between PHES and Stroop (r= -0.772, P<0.01). Patients with HE 1-2 (n=23) achieved abnormal test results in 100% (PHES), 91.3% (ANT), 39.1% (CFF), 78.3% (ICT), 91.3% (Stroop) and 82.6% (CRT). Conclusion: Thirty percent to 50% of the patients without clinical signs of HE had pathological test results, confirming the importance of screening for mHE in this group. The test results varied due to representation of different cognitive functions. Strong correlations between some test results indicate overlap of evaluated domains. PHES, ANT, Stroop and CRT appear most suitable for mHE screening. However, a larger group of patients with HE grade 1 – 2 needs to be studied with all tests to be able to evaluate the true positive rate (see Table 1 and Figure 1). Table 1. - Baseline Table All patients Neurologically unimpaired patients Patients with HE 1-2 P-value Number of patients N=132 N=109 N=23 Age 57 (51-65) 56 (50-65) 60 (55-64) 0.281 Elderly (>65 yrs) 31 (23.5%) 26 (23.9%) 5 (21.7%) 0.828 Sex female 41 (31.1%) 35 (32.1%) 6 (26.1%) 0.571 Etiology (multiple selection possible) Alcohol related 69 (52.3%) 55 (50.5%) 14 (60.9%) 0.364 NASH 21 (15.9%) 14 (12.8%) 7 (30.4%) 0.036 Viral 20 (15.2%) 19 (17.4%) 1 (4.3%) 0.112 PSC/AIH 16 (12.1%) 15 (13.8%) 1 (4.3%) 0.209 other 22 (16.7%) 18 (16.5%) 4 (17.4%) 0.918 TIPS 36 (27.3%) 33 (30.3%) 3 (13.0%) 0.092 Diabetes 35 (26.5%) 27 (24.8%) 8 (34.8%) 0.323 Previous oHE episodes 40 (30.3%) 30 (27.5%) 10 (43.5%) 0.130 Years of school education 10 (9-12) 10 (9-12) 9 (9-10) 0.036 PHES -4 (-8- -2) -3 (-5- -1) -12 (-13- -9) <0.001 PHES pathological 62 (47%) 39 (35.8%) 23 (100%) <0.001 CRT Index 1.8610 (1.4305-2.3160) 2.041 (1.526-2.435) 1.500 (1.098-1.719) 0.001 CRT pathological 67 (50.8%) 48 (44.0%) 19 (82.6%) <0.001 Stroop Off+OnTime (sec) 188.7 (162.8-218.7) 185.6 (160.3-208.3) 259.0 (204.7-355.45) 0.005 Stroop pathological 70 (53%) 49 (45.0%) 21 (91.3%) <0.001 ANT (animals/minute) 21 (17-27) 22 (19-28) 16 (14-19) <0.001 ANT pathological 77 (58.3%) 56 (51.4%) 21 (91.3%) <0.001 ICT weighted lures (=Lures/ Target Accuracy2) 18.6 (8.8-38.7) 14.3 (7.4-30.5) 50.9 (21.1-84.1) 0.020 ICT pathological 58 (43.9%) 40 (36.7%) 18 (78.3%) <0.001 CFF (corrected for standard derivation) 41.55 (38.62-46.22) 41.6 (38.9-46.0) 40.4 (33.6-47.8) 1.000 CFF pathological 37 (28%) 28 (25.7%) 9 (39.1%) 0.192 Sodium (mmol/l) 136 (134-139) 136 (134-139) 136 (131-138) 0.765 Creatinine (µmol/l) 86 (70-110) 85 (70-105) 104 (70-138) 0.359 CHE (kU/l) 3.15 (2.13-4.47) 3.33 (2.32-4.56) 2.35 (1.48-2.84) 0.006 Bilirubin (µmol/l) 25 (15-45) 24 (15-44) 37 (24-164) 0.066 Albumin (g/l) 33 (28-37) 33 (29-38) 29 (27-34) 0.019 Erythrocytes (mio/µl) 3.71 (3.02-4.26) 3.76 (3.13-4.36) 3.12 (2.68-3-82) 0.002 White blood cells (tsd/µl) 4.85 (3.73-7.38) 4.60 (3.40-7.05) 6.10 (4.60-8.00) 0.169 Platelets (tsd/µl) 99 (62-139) 101 (61-141) 84 (62-133) 0.646 Hemoglobin (g/dl) 11.1 (9.8-13.0) 11.2 (10.0-13.2) 10.3 (8.6-12.0) 0.026 INR 1.24 (1.11-1.38) 1.21 (1.11-1.33) 1.43 (1.25-1.78) 0.005 Child-Pugh-Score 7 (6-9) 7 (5-8) 10 (8-12) <0.001 MELD 12 (9-15) 11 (9-14) 16 (13-22) 0.009 Any HE prophylaxis 84 (63.6%) 64 (58.7%) 20 (87.0%) 0.011 Lactulose intake 70 (53%) 52 (47.7%) 18 (78.3%) 0.008 Rifaximine intake 46 (34.8%) 31 (28.4%) 15 (65.2%) <0.001 L-Ornithine L-Aspartate intake 21 (15.9%) 15 (13.8%) 6 (26.1%) 0.142 Note: All continuous variables are presented as median and IQR, nominal variables are presented as frequencies and percentages, t-test for normative distributed values, Mann-Whitney-U for not normative distributed values, Chi-Square for dichotomous values.Abbreviations: TIPS: transjugular intrahepatic portosystemic shunt, NASH: nonalcoholic steatohepatitis, HE: hepatic encephalopathy, mHE: minimal hepatic encephalopathy, PHES: portosystemic hepatic encephalopathy score, CRT: Continuous Reaction Time Test, ANT: animal naming test, ICT: Inhibitory Control Test, CFF: critical flicker frequency, CHE: cholinesterase, INR: international normalized ratio, MELD: model for end-stage liver disease. Figure 1.: Percentages of pathological test results.