In the present study, we constructed LRM: a new prediction model to evaluate the prognosis of pregnant women with liver failure. We found that age, INR, and |Na-135| were independent risk factors for poor prognosis of pregnant women with liver failure. The sensitivity of the LRM was higher than that of the classical MELD model. Furthermore, there was a statistically significant difference between them (P < 0.05).
The severity of coagulation disorders is positively correlated with the prognosis of liver failure 8. Prothrombin time (PT), INR, and PTA can reflect blood clotting function. The higher the INR, the worse the outcome
It is very important to maintain the electrolyte balance which is essential for the normal functioning of cells and organs. Many studies have reported that Na is a useful predictor of mortality in end-stage liver disease 18–20. Hyponatremia can increase intracranial pressure (ICP), and lead to cell damage. In the present study, we found that when Na decreased from 135mmol/L to 126.3mmol/L, the risk of intractable cure increased gradually. When the Na value was below 123.6mmo/L or beyond 146.4mmol/L, the risk increased significantly. Taken together, it is better to maintain Na at 135-146.4mmo/L which is also recommended by the EASL guidelines 21. The kidneys can regulate electrolytes, such as K, Na, and PH. Renal dysfunction causes electrolyte disorders and imbalance of the acid base, which might be led to deterioration of the disease13,22,23,but in our study, the Scr was not independent risk factors. The blood volume increases during pregnancy, and compared with the non-pregnant state, the physiological increase of glomerular filtration rate leads to the decrease of Scr level, therefore, the degree of renal injury is often underestimated in pregnancy. It may be inaccurate to evaluate the degree of renal injury by Scr alone, we should explore more tools to evaluate renal injury during pregnancy.
Advanced maternal age is associated with various complications during pregnancy 24,25. A large-scale search showed that the age of pregnant women older than 30 years was related to numerous adverse pregnancy outcomes 26. Our results were in accordance with those of previous reports, when a pregnant woman was beyond 30.5 years old, the possibility of poor prognosis increased significantly. Advanced age might be earlier than traditional age (≥ 35 years).
In recent years, ALSS has been used to treat liver failure. In some reports, ALSS was helpful for women with AFLP 27–29, HELLP syndrome 10,30, or HSV 10. However, it did not significantly affect the short-term prognosis of pregnant women with liver failure in our study. This finding might be related to the type of liver failure. AFLP played a major role, whereas ALF was the dominant type in our research. However, AGA does not believe that ALSS can improve survival in patients with ALF4. If AFLP leads to liver failure, ALSS is unlikely to work.
Hepatitis viruses are the most common cause of liver failure in China, the Indian subcontinent, and some African regions6,9,31–35. In our study, AFLP was the most common cause of liver failure, while hepatitis viruses were the second most common cause of liver failure, which is similar to another study36. We speculated that this change might be related to measurement in China. In China, the harmful effects of HBV had been widely publicized, and pregnant women had been provided free pre-pregnancy check-ups and regular antenatal care for decades. The increasing number of mothers were receiving early medical treatment and regular prenatal examinations. Therefore, the risk of liver failure or MTCT (mother-to-child transmission) is significantly reduced9,37.
Although some scholars have proposed different models 14,15,38,39 to predict the prognosis of liver disease during pregnancy, MELD remain the most classical model. In the present study, we compared LRM with MELD and verified it in another group of patients. We concluded that the sensitivity of LRM was higher than that of MELD in predictive ability. LRM is easier to obtain and calculate than MELD, and it might be a new predictive model which can be considered for liver failure during pregnancy.
Our study had the following limitations that must be taken into account. Firstly, we were unable to include a large number of cases because of its rare incidence which might affect the results. Secondly, it was a retrospective study, and we need to include more cases to further verify the model. Finally, this was a cross-sectional study that could not reflect dynamic changes in the disease. Whatever, we enrolled a lot of cases in this field, and the data were obtained from multiple centers. Taken together, our results might provide some significance for evaluating the prognosis of pregnant women with liver failure.