Predicting Liver Disease Risk Using a Combination of Common Clinical Markers: A Screening Model from Routine Health Check-Up

Abstract

Background: Early detection is crucial for the prognosis of patients with autoimmune liver disease (AILD). Due to the relatively low incidence, developing screening tools for AILD remain a challenge.

Aims: To analyze clinical characteristics of AILD patients at initial presentation and identify clinical markers, which could be useful for disease screening and early detection.

Methods: We performed observational retrospective study and analyzed 581 AILD patients who were hospitalized in the gastroenterology department and 1000 healthy controls who were collected from health management center. Baseline characteristics at initial presentation were used to build regression models. The model was validated on an independent cohort of 56 patients with AILD and 100 patients with other liver disorders.

Results: Asymptomatic AILD individuals identified by the health check-up are increased yearly (from 31.6% to 68.0%, p < 0.001). The cirrhotic rates at an initial presentation are decreased in the past 18 years (from 52.6% to 20.0%, p < 0.001). Eight indicators, which are common in the health check-up, are independent risk factors of AILD. Among them, abdominal lymph node enlargement (LN) positive is the most significant different (OR 8.85, 95% CI 2.73-28.69, p < 0.001). The combination of these indicators shows high predictive power (AUC = 0.98, sensitivity 89.0% and specificity 96.4%) for disease screening. Except two liver or cholangetic injury makers, the combination of AGE, GENDER, GLB, LN, concomitant extrahepatic autoimmune diseases, and familial history also shows a high predictive power for AILD in other liver disorders (AUC = 0.91).

Conclusion: Screening for AILD with described parameters can detect AILD in routine health check-up early, effectively and economically. Eight variables in routine health check-up are associated with AILD and the combination of them shows good ability of identifying high-risk individuals.

Figure 1

Proportion of AILD patients diagnosed through regular health check-up. The proportion increased year by year (all: χ2 = 44.32, p < 0.001; AIH: χ² = 14.11, p < 0.001; PBC: χ² = 18.97, p < 0.001; PBC-AIH OS: χ² = 10.99, p < 0.001). p values were calculated by the Trend Chi-square test. ^∗∗∗ represents p value lower than 0.001, ns. represents no statistically significance.

Figure 2

The rate of liver cirrhosis in AILD patients at initial admission to the hospital. Except for PBC-AIH OS patients, the proportion of liver cirrhosis was reduced year by year (all: χ² = 19.36, p < 0.001; AIH: χ² = 16.00, p < 0.001; PBC: χ² = 6.95, p < 0.001; PBC-AIH OS: χ² = 3.33, p > 0.05). p values were calculated by the Trend Chi-square test. ^∗∗∗ represents p value lower than 0.001, ns. represents no statistically significance.

Figure 3

Performance of the logistic regression models in clinical parameters measured from routine health check-up. (a) Receiver operating characteristic curve (ROC) shows Model 1 had a good ability in predicting high-risk individuals with AILD in 8 variables, (b) Model 2 also showed good prediction ability without liver- or cholangetic injury parameters, (c) performance of two models in the test group. AUC: area under the receiver operating characteristic curve; PPV: positive predictive value; NPV: negative predictive value.

Figure 4

Performance of external validation cohort in other liver injuries with abnormal LFTs. Validating the two logistic models in external cohorts, (a) ROC shows a comparison of two models in the independent validation group, (b) performance of two models in the external validation group, which shows Model 2 with six variables has better ability to identify AILD from other liver diseases. ROC: receiver operating characteristic curve; AUC: area under the receiver operating characteristic curve; LFTs: liver function tests; PPV: positive predictive value; NPV: negative predictive value.

Figure 5

Classification trees (CART) for predicting high-risk individuals of AILD. (a) CART tree for prediction visualization. The two sets of numbers underneath each terminal node represent the proportions of patients or control subjects. The subgroups were marked with green and blue according to prediction outcomes, e.g., subgroup 2 represents the group with the predictive pattern for the disease (probability of AILD was 87%, dark green). The actual split values (thresholds) were indicated in the branches of the tree. HC: healthy controls; AILD: autoimmune liver disease patients. (b) Receiver Operating Characteristic Curve. The CART model worked well in predicting high-risk individuals with AILD in other liver injuries with an AUC of 0.91.