The efficacy and safety of granulocyte colony-stimulating factor in the treatment of acute-on-chronic liver failure: A systematic review and meta-analysis

RESEARCH ARTICLE The efficacy and safety of granulocyte colonystimulating factor in the treatment of acuteon-chronic liver failure: A systematic review and meta-analysis Bo QiuID 1 , Jia Xu Liang 1,2 , Manuel Romero Go ´mez 3 * 1Department of Gastroenterology, The Second Affiliated Hospital of Hainan Medical University, HaiNan, China, 2Department of Diagnostic Radiology, People’s Hospital of Zhengzhou, Zhengzhou, China, 3Department of Digestive, Hospital Universitario de Virgen Rocio, Seville, Spain *[email protected] Abstract Background and objectives The safety and efficacy of granulocyte-colony stimulating factor (G-CSF) for the treatment of acute-on-chronic liver failure (ACLF) remain controversial. This meta-analysis aimed to evaluate the effectiveness and safety of G-CSF in treating ACLF. Methods The estimated pooled risk ratio (RR) and 95% confidence interval (CI) assessed the treatment effects of G-CSF. Mean differences (MD) and 95% confidence intervals were used to analyze continuous data. Heterogeneity was explored by sensitivity analysis. Potential publication bias was assessed using Egger’s test. Results Ten studies, comprising a total of 603 participants, were included in the analysis. The G- CSF group showed significantly lower MELD scores at 7-day (MD = -2.39, 95%CI: -3.95 to -0.82), CTP scores at 7-day (MD = -0.77, 95%CI: -1.41 to -0.14), and MELD scores at 30- day (MD = -3.01, 95%CI: -5.36 to -0.67) compared to the control group. Furthermore, the G- CSF group was associated with a reduced risk of sepsis (RR = 0.53, 95%CI: 0.35–0.80). The 30-day survival (RR = 1.26, 95%CI:1.10–1.43), 60-day survival (RR = 1.47, 95% CI:1.17–1.84, and 90-day survival (RR = 1.73, 95%CI: 1.27–2.35) of patients with ACLF treated with G-CSF were significantly higher than those of the control group. Conclusions Our meta-analysis suggests that G-CSF therapy may be a promising treatment for ACLF, with significant improvements in liver function and survival rates, however, further studies are needed to verify this conclusion. PLOS ONE PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 1 / 14 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Qiu B, Liang JX, Romero Go ´mez M (2023) The efficacy and safety of granulocyte colony-stimulating factor in the treatment of acuteon-chronic liver failure: A systematic review and meta-analysis. PLoS ONE 18(11): e0294818. https://doi.org/10.1371/journal.pone.0294818 Editor: Sona Frankova, Institute for Clinical and Experimental Medicine, CZECH REPUBLIC Received: May 22, 2023 Accepted: November 9, 2023 Published: November 30, 2023 Copyright: ©2023 Qiu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: The author(s) received no specific funding for this work. Competing interests: The authors have declared that no competing interests exist. Abbreviations: G-CSF, granulocyte-colony stimulating factor; ACLF, acute-on-chronic liver failure; BMSCs, bone marrow stem cells; MELD, Introduction Acute-on-chronic liver failure (ACLF) is a severe and life-threatening condition characterized by an acute deterioration of liver function in patients with pre-existing chronic liver disease [1]. ACLF can result from various etiologies, including viral hepatitis, alcohol-related liver disease, non-alcoholic fatty liver disease, and autoimmune liver disease [2]. It is characterized by severe systemic inflammation, organ failure, and poor prognosis [3,4]. The mortality exceeded 30% at 28 days and 63% at 90 days [5,6]. The only recommended treatment option is liver transplantation, but the scarcity and high cost of donor organs limit its use. To overcome the challenges of treating advanced liver disease, researchers have proposed alternative methods such as bone marrow stem cell (BMSC) therapy and mobilization of granulocyte colony-stimulating factor (G-CSF). BMSCs are an important source of liver stem cells that possess the ability to differentiate into multiple cell types and self-renew. Studies have demonstrated that BMSCs can promote hepatocyte proliferation and liver regeneration, as well as regulate immune function [7,8]. G-CSF mobilizes hematopoietic stem cells in patients with advanced liver disease. It improves liver function and survival in ACLF patients. Compared to bone marrow stem cell transplantation, mobilization of autologous bone marrow stem cells to the damaged area of the liver is a simple, safe, effective, and non-invasive procedure that avoids the immune response associated with stem cell transplantation. This approach does not require stem cell extraction and expansion and has promising applications. G-CSF mobilizes hematopoietic stem cells and has been shown to improve liver function and increase hepatocyte regeneration in both animal models and clinical studies [9–11]. However, the effectiveness and safety of G-CSF treatment in ACLF are still being debated. A meta-analysis suggested that G-CSF treatment may only reduce short-term overall mortality but not mortality secondary to gastrointestinal bleeding [12]. Another meta-analysis showed no significant beneficial effects of G-CSF in ACLF [13]. Moreover, some studies had small sample sizes and insufficient statistical power to predict the outcomes of a comprehensive study [14,15]. Thus, the efficacy and safety of G-CSF in liver failure remain controversial. To address this issue, we conducted a large sample analysis in this meta-analysis to evaluate the safety and effectiveness of G-CSF in the treatment of ACLF. Method Literature search strategy Two independent investigators (B. Q and JX. L) conducted a comprehensive search of the PubMed, Cochrane Library, and Embase to identify relevant studies published up to March 2023. The search strategy involved using MeSH terms and keywords, including (“granulocytecolony stimulating factor” or “G-CSF”) and (“acute-on-chronic liver failure” or “ACLF" or “liver failure” or “Hepatic failure” or “severe hepatitis” or “Fulminant hepatitis”). No restrictions were placed on article type or additional filters during the search. A manual search was also conducted by reviewing the references of original articles and relevant review articles. The results were collected independently by the two investigators. Inclusion and exclusion criteria Studies were considered suitable for this meta-analysis if they met the following inclusion criteria: (1) Randomized clinical trials (RCT); (2) The experimental group received G-CSF therapy while the control group received conventional treatment; (3) Availability of clinical outcomes. Exclusion criteria were (1) Non-therapeutic use of G-CSF; (2) Animal studies; (3) Non-randomized controlled studies. PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 2 / 14 Model for end stage of liver disease; CTP-scores, Child–Turcotte–Pugh scores; INR, international normalized ratio; HE, hepatic encephalopathy; RCT, randomized clinical trials; RR, risk ratios; Cl, confidence intervals; MD, mean differences; SD, standard deviation. Data extraction Two reviewers (B. Q and JX. L) independently extracted uniformly reported data elements for each study meeting the inclusion criteria, which were then cross-verified by a third reviewer (Romero Gomez M). In cases where the same population was reported in multiple articles, only the most informative or complete study was retained to prevent duplication. The following information was extracted from each study: first author, year of publication, country, sample size, etiology, route and dosage of G-CSF administration, survival rates, and complications. The Cochrane risk of bias tool was used to assess each trial, with standard criteria including the following domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other biases [16]. Statistical analysis Pooled risk ratios (RR) and their corresponding 95% confidence intervals (CIs) were used to evaluate the therapeutic effect of G-CSF. Mean differences (MD) and 95% confidence intervals were used for continuous data analysis. Heterogeneity between trials was quantified using the I 2 test, with I 2 �50% indicating significant heterogeneity. A fixed-effects model was used when heterogeneity was not significant. Obvious clinical heterogeneity was evaluated by removing a single study and repeating the meta-analysis [17]. We evaluated potential publication bias using Egger’s test, with a P value greater than 0.05 indicating no publication bias. All statistical analyses were performed using RevMan 5.4 (The Cochrane Collaboration, Oxford, UK) and Stata version 15 (Stata Corp, College Station, Texas, USA). Results Characteristics of the included studies Out of the 563 articles identified through searches of the Cochrane Library, PubMed, Embase, and references, 10 studies including a total of 603 participants were ultimately included in our meta-analysis [10,11,14,15,18–23]. A flowchart detailing the study identification and selection process is presented in Fig 1. Of the 10 studies analyzed, five were conducted in India [11,15,20–22], two in China [10,23], two in Bangladesh [14,19], and one in Germany [18]. The main characteristics of each study are summarized in Table 1. Prior to data analysis and synthesis, the quality of the studies was evaluated using the Cochrane risk of bias tool, as shown in Fig 2. Changes in liver function after G-CSF therapy MELD and CTP score. Six studies provided detailed post-treatment scores for the Model for End-Stage Liver Disease (MELD) and Child-Turcotte-Pugh (CTP) [10,14,15,19,20,22]. The results of the meta-analysis indicate that, compared to the control group, the experimental group showed a significant improvement in MELD scores (MD = -2.39, 95%CI: -3.95 to -0.82, P<0.05) and CTP scores (MD = -0.77, 95%CI: -1.41 to -0.14, P<0.05) after 7 days of treatment, and a significant improvement in 30-day post-treatment MELD score (MD = -3.01, 95%CI: -5.36 to -0.67, P<0.05). However, there was no significant difference in the 30-day post-treat- ment CTP score between the control and experimental groups (Fig 3A and 3B). Laboratory results. Three studies that compared G-CSF with control therapy provided data on liver function parameters [15,21,22]. The meta-analysis results showed that G-CSF PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 3 / 14 treatment did not result in a significant improvement in serum albumin (MD = 0.07, 95%CI: -0.16 to 0.31, P = 0.54, I 2 = 0%) and total bilirubin (MD = -1.53, 95%CI: -5.47 to 2.40, P = 0.44, I 2 = 0%), but international normalized ratio (INR) was slightly improved (MD = -0.25, 95%CI: -0.47 to -0.04, P<0.05, I 2 = 12%) within one month. Prevention of serious complications and improvement of survival Survival rates. The intervention methods used in the ten eligible studies were roughly commensurate, but there was heterogeneity between the articles (P = 0.05). Therefore, a random effects model was used to evaluate the 90-day survival rates. The results indicate that G-CSF therapy was associated with a reduced risk of death for patients with ACLF, with pooled RR (95% CI, P) values of 1.26 (1.10–1.43, P<0.05) for 30-day survival, 1.47 (1.17–1.84, P<0.05) for 60-day survival, and 1.73 (1.27–2.35, P<0.05) for 90-day survival (Fig 4A–4C). We conducted a subgroup analysis to compare the effect of different doses (5 μg/kg/day vs. 10 μg/kg/day) of G-CSF in treating ACLF. The findings demonstrated that both doses improved the survival rates of ACLF patients at 30, 60, and 90 days (Table 2). Development of complications and analysis of the cause of death Sepsis was reported in 6 studies involving 554 ACLF patients, 224 were in the G-CSF treatment group and 230 were in the control group. The meta-analysis showed that patients treated with G-CSF had a significantly lower risk of sepsis compared to those treated with standard medical therapy (SMT) alone (RR = 0.53, 95% CI: 0.35–0.80, P<0.05, I 2 = 0%). Fig 1. Study identification and selection flowchart. https://doi.org/10.1371/journal.pone.0294818.g001 PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 4 / 14 Table 1. Study characteristics. Author (year) Country Sample size Etiology Drug route/dose CTP score at baseline MELD score at baseline Survival (%) at 30 days Survival (%) at 60 days Survival (%) at 90 days Complications Duan (2013) [10] China 55 G-CSF:27 Control:28 Hepatitis B 5μg/kg/day subcutaneously for 6 consecutive days G-CSF: 12.17 ±1.47 Control: 12.25 ±1.29 G-CSF: 25.11 ±3.30 Control: 26.30 ±4.12 ———— ———— G-CSF: 13 (48.1%) Control: 6 (21.4%) Hepatorenal syndrome, hepatic encephalopathy, sepsis, hemorrhages Engelmann (2021) [18] Germany 176 G-CSF:88 Control:88 SAH 5 μg/kg/day subcutaneously for 5 consecutive days followed by every third day until 1 month ———— G-CSF: 24.4 ±6.3 Control: 24.5 ±6.1 G-CSF: 42 (47.7%) Control: 43 (48.9%) G-CSF: 29 (33.0%) Control: 31(35.2%) G-CSF: 27 (30.7%) Control: 26 (29.5%) Hemorrhages, sepsis, organ failures Garg (2012) [11] India 47 G-CSF:23 Control:24 Mixed*5μg/kg/day subcutaneously for 5 consecutive days followed by every third day until 1 month G-CSF: 12 (11–14) Control: 12 (10–14) G-CSF: 29 (21–40) Control: 31.5 (20–40) G-CSF: 18 (78.3%) Control: 11 (45.8%) G-CSF: 16 (69%) Control: 7 (29%) ———— Hepatorenal syndrome, hepatic encephalopathy, sepsis, organ failures Haque (2020) [19] Bangladesh 39 G-CSF:22 Control:17 Mixed 5 μg/kg/day subcutaneously for 6 consecutive days G-CSF: 11.77 ±1.11 Control: 11.65 ±1.36 G-CSF: 27.64 ±4.60 Control: 29.47 ±5.53 G-CSF: 14 (63.6%) Control: 10 (58.8%) G-CSF: 11 (50%) Control: 5 (29.4%) G-CSF: 8 (36.4%) Control: 5 (29.4%) Hepatorenal syndrome, hemorrhages, hepatic encephalopathy Saha (2017) [14] Bangladesh 32 G-CSF:16 Control:16 Mixed 5 μg/kg/day subcutaneously for 6 consecutive days G-CSF: 12 (10–13) Control: 12 (10–14) G-CSF: 24.5 (21–32) Control: 25.5 (21–35) G-CSF: 14 (87.5%) Control: 13 (81.3%) G-CSF: 14 (87.5%) Control: 13 (81.3%) G-CSF: 14 (87.5%) Control: 8 (50%) Hepatorenal syndrome, electrolytic imbalance Sharma (2019) [20] India 31 G-CSF:15 Control:16 Mixed 5 μg/kg/day subcutaneously for 5 consecutive days G-CSF: 12 ±1.4 Control: 12.75 ±0.85 ———— G-CSF: 10 (66.7%) Control: 6 (37.5%) G-CSF: 8 (53.3%) Control: 6 (37.5%) ———— None Shasthry (2019) [15] India 28 G-CSF:14 Control:14 SAH 5 μg/kg/day to a maximum of 300 μg per day for 5 doses followed by every third day until 4 weeks G-CSF: 10.9 ±1.3 Control: 10.7 ±1.2 G-CSF: 24.6 ±3.9 Control: 27.6 ±4.4 G-CSF:11 (78.6%) Control: 10 (71.4%) G-CSF: 10 (71.4%) Control: 5 (35.7%) G-CSF: 9 (64.3%) Control: 4 (28.6%) Hemorrhages, sepsis Singh (2014) [21] India 46 G-CSF:23 Control:23 SAH 5 μg/kg subcutaneously every 12 h for 5 consecutive days G-CSF: 12 Control: 12 G-CSF: 27 Control: 30 G-CSF: 18 (78.3%) Control: 10 (43.5%) G-CSF:18 (78.3%) Control: 9 (39.1%) G-CSF: 18 (78.3%) Control: 5 (21.7%) Hepatorenal syndrome, hemorrhages, hepatic encephalopathy Singh (2018) [22] India, 38 G-CSF:18 Control:20 SAH 5 μg/kg subcutaneously every 12 h for 5 consecutive days G-CSF: 11 (8–13) Control: 10 (8–14) G-CSF: 26 (19–37) Control: 27.5 (19–41) G-CSF: 16 (88.9%) Control: 11 (55%) G-CSF: 16 (88.9%) Control: 8 (40%) G-CSF: 16 (88.9%) Control: 6 (30%) Sepsis, hemorrhages, organ failures Tong (2022) [23] China 111 G-CSF:54 Control:57 Hepatitis B 5μg/kg/day subcutaneously for 6 consecutive days ———— G-CSF: 22.8 (20.7–26.0) Control: 24.1 (21.6– 27.1) G-CSF: 46 (85.2%) Control: 38 (66.7%) G-CSF: 39 (72.2%) Control: 31 (54.4%) G-CSF: 36 (66.7%) Control: 30 (52.6%) Sepsis, hepatic encephalopathy, acute kidney injury NOTE. Data are presented as mean ±SD, n (%), or median (range). Mixed *: Viral hepatitis and alcoholic hepatitis related acute and chronic liver failure; SAH, severe alcoholic hepatitis; G-CSF, granulocyte colony-stimulating factor. https://doi.org/10.1371/journal.pone.0294818.t001 PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 5 / 14 Fig 2. a. Risk of bias graph; b. Risk of bias summary. https://doi.org/10.1371/journal.pone.0294818.g002 PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 6 / 14 Three studies involving 205 ACLF patients reported hepatic encephalopathy (HE), with 103 patients in the G-CSF treatment group and 102 in the control group. The meta-analysis indicated that there was no significant difference in the risk of HE between patients treated with G-CSF and those treated with SMT alone (RR = 0.64, 95% CI: 0.27–1.50, P = 0.30, I 2 = 0%). Bleeding was reported in 8 studies involving 461 ACLF patients, 231 were in the G-CSF treatment group and 230 were in the control group. The meta-analysis showed that compared with SMT, patients receiving G-CSF did not have a significantly reduced risk of bleeding (RR = 1.04, 95% CI: 0.58–1.87, P = 0.90, I 2 = 0%). Fig 3. Meta-analysis of MELD and CTP scores between G-CSF and control therapy. a). MELD scores, b) CTP scores. https://doi.org/10.1371/journal.pone.0294818.g003 PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 7 / 14 Five studies involving 283 ACLF patients reported Hepatorenal syndrome (HRS), with 142 patients in the G-CSF treatment group and 141 in the control group. The meta-analysis indicated that there was no significant difference in the risk of HRS between patients treated with G-CSF and those treated with SMT alone (RR = 0.58, 95% CI: 0.31–1.07, P = 0.08, I 2 = 0%). (S1 Fig). Fig 4. Survival rate between patients with ACLF, G-CSF groups vs. control group. a). 30-day, b) 60-day, c) 90-day. https://doi.org/10.1371/journal.pone.0294818.g004 PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 8 / 14 Sensitivity analysis We performed a sensitivity analysis in which one study was removed at a time and the pooled estimates were recalculated for the remaining studies. The results showed that the overall findings were not substantially influenced by any individual study (S2 Fig). Publication bias Asymmetry was observed in the visual inspection of the funnel plot, indicating the possibility of publication bias. Therefore, sensitivity analysis was conducted using the Egger’s test and trim and fill method [24]. The difference between the original effect size estimate and the corrected effect size estimate was not substantial, indicating that publication bias did not have a significant impact on the results (S1 Table). Discussion Acute-on-chronic liver failure (ACLF) is a severe and potentially fatal condition that occurs in patients with pre-existing chronic liver disease. It is characterized by acute decompensation and organ failure, which significantly increases morbidity and mortality rates. The treatment of ACLF remains a major challenge, and current therapeutic options are limited. In recent years, granulocyte colony-stimulating factor (G-CSF) has been proposed as a potential therapy for ACLF. G-CSF is a cytokine that stimulates the production and differentiation of neutrophils and has been shown to have immunomodulatory and anti-inflammatory effects [25]. In this meta-analysis, we evaluated the efficacy and safety of G-CSF therapy in the treatment of ACLF. According to our study, the Child-Turcotte-Pugh (CTP) and Model for End-Stage Liver Disease (MELD) scores were utilized to predict disease severity and mortality at 3 months among patients diagnosed with end-stage liver disease [26]. Our results indicate that treatment with G-CSF was significantly associated with improved liver function, as evidenced by improvements in both MELD and CTP scores. This improvement was observed after only 7 days of treatment and persisted for up to 30 days post-treatment. Despite these positive findings, there were no significant differences observed in serum albumin and total bilirubin levels between the G-CSF and control groups. However, there was a slight improvement in International Normalized Ratio (INR) levels within one month of treatment. These results suggest that G-CSF therapy may have a beneficial effect on liver function in patients with ACLF. Recently, several definitions and prognostic scores have been developed specifically for ACLF. In particular, the CLIF-C ACLF score and the COSH-ACLF score provide higher predictive performance for patients with ACLF. However, these scores were rarely used in the articles included so far. It is expected that more studies will use prognostic scoring systems with higher sensitivity and specificity, which will lead to more valuable results. Table 2. Subgroup analysis based on dose for assessment of survival rate. Category Subgroups RR 95%Cl P I 2 30-day survival rate 5 μg/kg/day 1.19 1.03–1.36 <0.05 0% 10 μg/kg/day 1.71 1.22–2.38 <0.05 0% 60-day survival rate 5 μg/kg/day 1.31 1.09–1.57 <0.05 31% 10 μg/kg/day 2.10 1.42–3.12 <0.05 0% 90-day survival rate 5 μg/kg/day 1.37 1.11–1.70 <0.05 3% 10 μg/kg/day 3.22 1.91–5.43 <0.05 0% https://doi.org/10.1371/journal.pone.0294818.t002 PLOS ONE G-CSF and acute-on-chronic liver failure PLOS ONE | https://doi.org/10.1371/journal.pone.0294818 November 30, 2023 9 / 14