Since the report of Groeneweg et al. (1) in 1998, there has been interest in the effect of subclinical (now called minimal) hepatic encephalopathy (HE) in cirrhosis subjects on the measures of quality of life (QOL). Marchesini et al. (2) also noted a major reduction in the QOL of cirrhotic patients. The factors most associated with perceived poor health in the study were severity of liver disease and muscle cramping. Interestingly, HE was not identified by logistic regression as a major factor in poor QOL, but lactulose therapy was noted to protect against reduction in QOL. Subsequently, a number of other papers noted an association of reduced QOL in cirrhotic patients diagnosed with minimal HE using a variety of techniques. The paper of Prasad et al. (3) demonstrated for the first time that QOL in cirrhotics with well-documented minimal HE significantly improved with 3 months of lactulose therapy. In the current issue of the journal, we have a study of the relationship of QoL using two measurement techniques for health-related QOL; Short Form Health Survey (SF-36) and Chronic Liver Disease Questionnaire (CLDQ) with minimal HE in cirrhotic patients. This well-conducted study showed no relationship between these health-related QOL measures and the presence or absence of minimal HE (4). Why might this study be in contrast to the others previously published?
The gold standard at present as noted by the authors for the diagnosis of minimal HE is the psychometric hepatic encephalopathy score (PHES) (5). However, this particular system was validated in a non-alcoholic liver disease population. The current study features a fairly high percentage of patients with alcoholic liver disease (over 40%). The impact of the inclusion of this population of cirrhotics is difficult to gauge. However, another study by Bao et al. (6) demonstrated a relationship of minimal HE with the same QOL measures used in this study adapted for the Chinese population. Patients with alcoholic cirrhosis were excluded in this study, but it must be added that minimal HE in this study was determined by only two psychometric tests and/or electroencephalographic changes. Many countries now have validated health-related QOL measures for their own population but as the authors state standardization of measures of minimal HE have not been consistent in most studies on QOL published to date.
Could the PHES be more sensitive than the other techniques put forward to diagnose minimal HE? If so, when applied the PHES may be identifying patients very early in the spectrum of neurological impairment in cirrhosis (7). Conversely, some of the prior studies on the relationship of QOL of life measures to minimal HE may have included patients with more advanced minimal HE or very mild overt HE. This might explain the differences in this study compared with others but needs to be examined in future studies. Certainly, if these findings are reproduced in another study, we may have to reconsider the relationship of QOL to minimal HE as determined by PHES. However, ultimately QOL is a subjective perspective of patients. As such, perhaps the response of patients to HE treatment may be the best judge of whether patients have a reduced QOL due to minimal HE as noted in the RIME study (8).
The recommendation that PHES be the gold standard for diagnosis of minimal HE has stimulated a lot of research investigation. However, in many studies the tests used are not strictly those originally proposed (see Table 1). The paper and pencil tests like the PHES are difficult to score accurately and are not widely available especially in the USA (9). We need a totally standardized testing system so that studies are comparable. Based on this study, one might surmise that perhaps PHES is not the optimum testing system for the detection of minimal HE. Other psychometric test batteries most of which are computerized are being developed to diagnose minimal HE. Once these are validated against our arbitrary gold standard, we can begin to re-examine their relationship to measures of QOL. In an ideal world, we would eventually have a global system that is cost-effective and is available to all clinicians, is not too time consuming and can give immediate indication of whether minimal HE is present or not. However, we have not reached that point yet and we are currently using different techniques to measure health-related QOL and minimal HE. If we could understand why no correlation could be found between QOL and MHE in this study, we may be able to answer a lot of questions pertaining to this area.
|Author||Year||Number of MHE patients/total number of cirrhosis patients||Inclusion of alcoholic-related liver disease in the study||Criteria for defining minimal hepatic encephalopathy||Quality of life instruments used in the study||Therapeutic trial||Outcome|
|Groeneweg et al. (1)||1998||48/179 (26.8%)||Yes (21%)||Abnormal EEG and/or abnormality in one of the two psychometric tests (Number Connection test A and Digit Symbol test)||SIP Scale||No||Subclinical (or minimal) HE is independently related to deterioration in mean total SIP score. This was reproduced when SIP score was repeated after 3 months|
|Prasad et al. (3)||2007||61/90 (67.7%)||Yes (63%)||Abnormality in two tests as per Ferenci et al. (5) Number Connection test A and B; Figure Connection test A and B; Picture Completion test; Block Design test||SIP Scale||Lactulose for 3 months||MHE was associated with deterioration in 11 of the 12 scales and total SIP score. Significant improvement was seen in the mean total SIP score following 3 months of lactulose therapy compared with the non-treatment arm|
|Bao et al. (6)||2007||33/106 (31.1%)||Excluded||Abnormal performance in at least one of the two psychometric tests (NCT A and DST) or if the EEG was abnormal||SF-36 CLDQ||No||Significant difference in SF-36 score was seen in MHE positive patients compared with non-MHE patients. No difference seen in the CLDQ score except for abdominal symptoms|
|Sidhu et al. (8)||2011||94/263 (35.7%)||Yes (48%)||Abnormality in at least two psychometric tests as per Ferenci and colleagues Tests conducted were Number Connection test A, Figure Connection test A, Digit Symbol test and Picture Completion test and Block Design test||SIP Scale||Rifaximin (1200 mg/day) for 8 weeks||MHE patients had significantly worse SIP scores compared with non-MHE patients. Significant improvement in SIP score in the treatment group compared to placebo arm|
|Wunsch et al. (4)||2011||29/77 (37.7%)||Yes (42.8%)||Abnormal performance in psychometric test with a composite score of −5 or less. Psychometric tests used were Number Connection Test A and B, Digit Symbol Test, Serial dotting test and Line Tracing test||SF-36 CLDQ||No||No significant difference seen in patients with MHE in both HRQOL instruments SF-36 and CLDQ|