Department of Gastroenterology, G B Pant Hospital, New Delhi, India.
Eur J Gastroenterol Hepatol. 2010 May;22(5):526-31. doi: 10.1097/MEG.0b013e3283341b7d.
Lactulose is commonly used in the treatment of hepatic encephalopathy (HE). However, all patients do not respond to lactulose. We evaluated predictors of nonresponse to lactulose in patients with cirrhosis and HE.
Consecutive cirrhotic patients with HE were enrolled. HE was diagnosed by West Haven criteria. Patients were treated with lactulose and correction of any associated precipitating factors. Nonresponse was defined if patient remained in HE even after 10 days of treatment or died while in HE.
Of 300 patients with cirrhosis and HE, 231 (77%) patients met the inclusion criteria. The majority (95%) of the patients had Grade 2 or 3 HE. Of 231 patients, 180 (78%) responded to lactulose. Fifty-one (22%) did not respond to lactulose, 34 (15%) died without any improvement in HE and HE did not improve in 17 (7%) patients after 10 days of therapy. On comparing baseline parameters between nonresponders versus responders there was significant difference between baseline age (42.0+/-11.9 vs. 46.6+/-12.7 year, P=0.02), total leukocyte count (median, 9300 vs. 7300 cells/mm3, P=0.001), serum sodium level (129.9+/-6.2 vs. 133.7+/-7.1 mmol/l, P=0.001), model for end stage liver disease (MELD) score (22.9+/-3.8 vs. 19.9+/-4.2, P=0.001), mean arterial pressure (MAP, 77.9+/-10.0 vs. 86.3+/-8.7 mmHg, P=0.001), serum AST (median, 114 vs. 76 IU/l, P=0.01), serum ALT (median, 84 vs. 48.5 IU/l, P=0.001), spontaneous bacterial peritonitis [18 (35%) vs. 37 (21%), P=0.02] and hepatocellular carcinoma [HCC, 17 (33%) vs. 14 (7%), P=0.001]. On multivariate analysis baseline total leukocyte count, MELD, MAP, and HCC were independent predictors of nonresponse to lactulose (P=0.001). Combination of low MAP, high MELD, and presence of HCC had diagnostic accuracy of 81% in predicting nonresponse to lactulose.
Of 78% patients with chronic liver disease with HE (majority with Grade 2 and 3) responded to lactulose. High baseline MELD, high total leukocyte count, low serum sodium, low MAP, and presence of hepatocellular carcinoma were predictors of nonresponse to lactulose.
乳果糖常用于治疗肝性脑病（HE）。然而，并非所有患者对乳果糖都有反应。我们评估了肝硬化和 HE 患者对乳果糖无反应的预测因素。
连续纳入患有 HE 的肝硬化患者。HE 通过 West Haven 标准诊断。给予乳果糖治疗，并纠正任何相关的诱发因素。如果患者在治疗 10 天后仍处于 HE 状态或在 HE 期间死亡，则定义为无反应。
在 300 例患有 HE 的肝硬化患者中，231 例（77%）符合纳入标准。大多数（95%）患者为 2 级或 3 级 HE。在 231 例患者中，180 例（78%）对乳果糖有反应。51 例（22%）对乳果糖无反应，34 例（15%）在 HE 期间死亡，HE 无改善，17 例（7%）患者在治疗 10 天后 HE 无改善。在比较无反应者与反应者的基线参数时，基线年龄（42.0±11.9 岁 vs. 46.6±12.7 岁，P=0.02）、总白细胞计数（中位数，9300 与 7300 细胞/mm3，P=0.001）、血清钠水平（129.9±6.2 与 133.7±7.1 mmol/l，P=0.001）、终末期肝病模型（MELD）评分（22.9±3.8 与 19.9±4.2，P=0.001）、平均动脉压（MAP，77.9±10.0 与 86.3±8.7 mmHg，P=0.001）、血清天门冬氨酸氨基转移酶（AST，中位数，114 与 76 IU/l，P=0.01）、血清丙氨酸氨基转移酶（ALT，中位数，84 与 48.5 IU/l，P=0.001）、自发性细菌性腹膜炎[18 例（35%）与 37 例（21%），P=0.02]和肝细胞癌[HCC，17 例（33%）与 14 例（7%），P=0.001]之间存在显著差异。多变量分析显示，基线总白细胞计数、MELD、MAP 和 HCC 是对乳果糖无反应的独立预测因素（P=0.001）。低 MAP、高 MELD 和 HCC 并存的组合对预测对乳果糖无反应具有 81%的诊断准确性。
在患有 HE 的慢性肝病患者中（大多数为 2 级和 3 级），78%的患者对乳果糖有反应。基线 MELD、总白细胞计数、血清钠水平、MAP 和 HCC 较高是对乳果糖无反应的预测因素。
