4. Discussion
Due to pancreatic and extra-pancreatic necrosis, acute pancreatitis (AP) is largely secondary to infection and causes Multiple organ dysfunction syndrome (MODS), which is a inflammatory disease with high morbidity and mortality64. Researchs have shown that during the pathogenesis of AP, neutrophils, lymphocytes, monocytes, natural killer cells and endothelial cells can produce a variety of cytokines or inflammatory mediators such as TNF-α, IL-1 and IL-6, and then induce inflammatory response or reduce cellular immune response through various pathways65. Early treatment in intensive care units has been recognized to be of great benefit to patients with severe episodes. Patients with AP are generally divided into three subgroups: mild, moderate, and severe. It is necessary to adjust the treatment regimen according to their specific needs, but determining the severity of the disease remains a clinical challenge. Serum markers are generally regarded as important indicators to predict the severity of AP. Serum C-reactive protein (CRP) is an acute phase reactant synthesized by the liver. In response to inflammation and infection, its levels in the blood increase within hours. Especially in inflammatory diseases, it is often used in infection and inflammation follow-up due to its short half-life, easy measurement and close relationship with prognosis of the disease66. In many textbooks, CRP is still considered as a gold standard for disease severity assessment67. Studies have reported that low serum albumin is independently associated with an increased risk of persistent organ failure and death in acute pancreatitis and can be used to predict the severity of acute pancreatitis68.
Continuous renal replacement therapy (CRRT) is defined as a blood purification treatment technique that continuously and slowly removes water and solutes by means of extracorporeal circulation blood purification to replace renal function. Compared with common hemodialysis, CRRT can prolong the treatment time of blood purification and reduce the treatment efficiency per unit time, so as to minimize the impact of changes in the concentration and volume of solute in blood on the body. Meanwhile, it adopts a filter with high permeability and good biocompatibility. It provides an important homeostasis balance for the treatment of severe patients. With the continuous development of science and technology, CRRT has new functions in addition to regulating water and electrolyte, maintaining acid-base balance and removing metabolic wastes69. Its application scope is no longer limited to kidney disease, and began to be used in the treatment of non-renal failure diseases such as pancreatitis13.Early CRRT can reduce the fatality rate of AP patients, as early as 2006 JPN Guideline wrote CRRT into the treatment of AP70. Researches have shown that CRRT can effectively remove the components of damaged vascular endothelial cells, improve endothelial cell function, thus reducing the incidence of MODS, and can delay or even block the process of MODS71.
This meta-analysis, which was based on 53 RCTs including 3382 participants, found that CRRT may indeed more beneficial to AP patients than conventional treatment. The study showed that after CRRT treatment, the mortality rate of the CRRT group was significantly lower than that of the control group, and there was no obvious heterogeneity between the groups, and the difference was statistically significant. CRRT also significantly reduced the APACHEⅡscores and cleared serum amylase and markers of the patients, and was superior in inflammatory factor clearance rate, alleviate the liver and kidney injury and without significant adverse reactions. According to the data of each research scope is different, we carried out subgroup analysis for serum inflammatory markers, liver and kidney function, APACHE Ⅱ scores. We found that there was less heterogeneity within the subgroup (less than 50%) after grouping according to data range, but greater heterogeneity (90%) between groups as a whole. However, there were statistically significant differences in the results, shown the efficacy of CRRT treatment was better. Therefore, we believe that the inter-group heterogeneity is mainly due to clinical heterogeneity, that is, the severity of AP patients in different studies is different, so that the datas of outcome indicators is different.
The difference between the meta-analysis in our study and the previous meta-analysis are the following : First of all, our study covered a large number of RCT studies and was not limited by language. A total of 3,392 subjects were included, with a larger sample size. Secondly, despite the past research done on meta-analysis of CRRT treatment of acute pancreatitis, but does not involve that much serum markers of inflammation factors and discuss the effects on liver and kidney function. As an updated and more comprehensive meta-analysis, this study further strengthened previous meta-analysis results, focused on more representative and specific results, fully described the impact of CRRT on AP patients, and strengthened the persuadability of existing evidence. Third, we registered the agreement of this study with PROSPERO in order to enhance PROSPERO’s transparency and quality of this meta-analysis.
From the perspective of the included literature content, the original research has several limitations due to the defects in design, measurement and evaluation. First of all, the randomized controlled trials included in this meta-analysis were conducted in different patient groups and in different clinical Settings. Therefore, potential heterogeneity risk exists. Secondly, although baseline status was compared between groups in each study, due to the different degree of AP patients included in the meta-analysis, the baseline status varied widely from study to study, various outcome indicators in different studies may also be different, which is also considered as the main source of heterogeneity in some outcome indicators. Thirdly, since CRRT treatment is significantly different from conventional treatment, doctors and patients cannot be blinded, which may cause performance bias and observation bias. Fourthly, due to the different conditions of different patients and the different time of CRRT treatment, the experimental results may be affected. Finally, the causes of these AP patients are different, and the description of whether they have diabetes, hypertension and other underlying diseases is not detailed.