|
摘要：
|
|
摘要：目的 探讨应用双重血浆分子吸附系统（double plasma molecular adsorption system，
DPMAS）治疗慢加急性肝衰竭（acute-on-chronic liver failure，ACLF）患者胆红素吸附
最佳时间以及影响人工肝疗效的因素，建立新的预测模型，优化人工肝治疗策略。方
法 以大连医科大学附属第二医院感染科2021年5月至2023年10月收治的行DPMAS治疗
的ACLF患者为研究对象。动态监测DPMAS治疗前、治疗1 h、2 h、3 h、4 h、治疗次日
的总胆红素（total bilirubin，TBil）水平，计算治疗各个时间点的总胆红素吸附率（total
bilirubin adsorption rate，TBAR）。根据患者90 d转归分为生存组和死亡组，收集患者入
院首次化验指标，并计算首次行DPMAS治疗后总胆红素反弹率（total bilirubin rebound
rate，TBRR）、Child-Turcotte-Pugh（CTP）评分、终末期肝病模型（model for end-stage
liver disease，MELD）评分、MELD-Na评分、iMELD评分、年龄-胆红素-国际标准化比
值-肌酐（age-bilirubin-international normalized ratio-creatinine，ABIC）评分、白蛋白-胆红
素（albumin-bilirubin，ALBI）评分、胆红素残存比（ratio of pre-treatment to post-treatment
total bilirubin，RPTB）等，应用多因素Logistic回归分析影响人工肝疗效的独立危险因
素并建立新的预测模型。绘制受试者工作特征（receiver operator characteristic，ROC）
曲线分析不同指标对患者预后的预测价值。结果 共纳入73例患者，治疗1 h、2 h、3 h、
4 h时TBAR分别为24.30%（17.69%，28.72%）、30.02%（24.13%，36.06%）、33.48%
（27.07%，39.43%）、36.01%（29.33%，41.89%），组间差异有统计学意义（H = 42.415，
P ＜ 0.001）。其中，治疗2 h、3 h、4 h的TBAR均显著高于治疗1 h（P均＜ 0.001），治疗
3 h、4 h的TBAR均显著高于2 h（P = 0.006，P ＜ 0.001），治疗3 h和4 h的TBAR差异无统
计学意义（P = 0.099）。90 d时36例患者生存，37例死亡。死亡组患者年龄（中位数：62岁比
51.5岁）、白细胞（中位数：6.39 × 109/L比7.53 × 109/L）、中性粒细胞绝对值（中位数：
4.52 × 109/L比6.48 × 109/L）、CTP评分（中位数：10.86分比12.59分）、MELD评分（中
位数：14.11分比17.83分）、MELD-Na评分（中位数：14.37分比21.16分）、iMELD评分
[（32.87 ± 1.83）分比（40.48 ± 1.66）分]、TBRR（中位数：13.52%比27.36%）、ABIC
评分（中位数：6.85分比8.32分）、MLR（中位数：0.49比0.72）及NLR（中位数：4.78
比8.61）显著高于生存组，血小板与白细胞比值（中位数：20.90比11.41）和总胆红素清
除率（total bilirubin clearance rate，TBCR）（中位数：26.36%比19.36%）显著低于生存
组，差异均有统计学意义（P均＜ 0.05）。多因素Logistic回归分析表明CTP评分（OR =
2.565，95%CI：1.201～5.476，P = 0.015）和TBRR（OR = 1.056，95%CI：1.002～1.114，
P = 0.043）是患者死亡的独立危险因素，通过系数法推导出新的预测模型为CTP-TBRR =
－13.729 + 0.942 × CTP + 0.055 × TBRR。CTP评分、TBRR及CTP-TBRR的ROC曲线下面
积分别为0.803、0.772、0.944，CTP-TBRR新模型较单一评分模型预测效能更高（z值分
别为3.158、3.423，P值分别为0.0016、0.0006），敏感度为91.9%，特异度为86.1%。结论
DPMAS治疗ACLF患者最佳胆红素吸附时间为3 h，CTP评分和TBRR可有效预测DPMAS
对ACLF患者的疗效，CTP-TBRR预测模型可进一步优化人工肝治疗策略。
|
|
Abstract: Objective To investigate the optimal duration of bilirubin adsorption in patients
with acute-on-chronic liver failure (ACLF) who were treated with the double plasma molecular
adsorption system (DPMAS), and to identify risk factors influencing the efficacy of artificial liver
therapy and to develop a new predictive model for optimizing artificial liver treatment strategies.
Methods Total of 73 ACLF patients treated with DPMAS at the Second Affiliated Hospital of
Dalian Medical University from May 2021 to October 2023 were enrolled. Total bilirubin (TBil)
levels were dynamically monitored at pre-treatment, 1 h, 2 h, 3 h, 4 h and the next day posttreatment
and total bilirubin adsorption rate (TBAR) was calculated for each time interval. Patients
were divided into survival group and death group based on 90-day outcomes. Initial laboratory
parameters were collected and clinical scores including the total bilirubin rebound rate (TBRR),
Child-Turcotte-Pugh (CTP) score, model for end-stage liver disease (MELD) score, MELD-Na
score, iMELD score, age-bilirubin-international normalized ratio-creatinine (ABIC) score, albuminbilirubin
(ALBI) score and the ratio of pre-treatment to post-treatment total bilirubin (RPTB)
were calculated. Independent risk factors influencing therapeutic efficacy were identified through
Multivariate Logistic regression analysis and a novel CTP-TBRR predictive model was established.
The predictive performance was evaluated by receiver operating characteristic (ROC) curves.
Results Total of 73 cases were enrolled. The TBAR values at 1 h, 2 h, 3 h and 4 h of DPMAS
treatment were 24.30% (17.69%, 28.72%), 30.02% (24.13%, 36.06%), 33.48% (27.07%, 39.43%)
and 36.01% (29.33%, 41.89%), respectively, the difference was statistically significant (H = 42.415,
P ＜ 0.001). TBAR at 2 h, 3 h and 4 h were significantly higher than that at 1 h (all P ＜ 0.001)
and TBAR of 3 h and 4 h were significantly higher than that of 2 h (P = 0.006, P ＜ 0.001), while
no difference was observed between 3 h and 4 h (P = 0.099). At 90 days, 36 patients survived and
37 died. Compared to those of the survival group, patients in death group exhibited significantly
higher values in age (median: 62 years vs. 51.5 years), white blood cell count (median: 6.39 × 109/L
vs. 7.53 × 109/L), neutrophil count (median: 4.52 × 109/L vs. 6.48 × 109/L), CTP score (median:
10.86 points vs. 12.59 points), MELD score (median: 14.11 points vs. 17.83 points), MELDNa
score (median: 14.37 points vs. 21.16 points), iMELD score [(32.87 ± 1.83) points vs. (40.48 ±
1.66) points], TBRR (median: 13.52% vs. 27.36%), ABIC score (median: 6.85 points vs.
8.32 points), MLR (median: 0.49 vs. 0.72) and NLR (median: 4.78 points vs. 8.61 points) (all P ＜ 0.05).
Conversely, platelet-to-white blood cell ratio (median: 20.90 vs. 11.41) and total bilirubin clearance rate
(TBCR; median: 26.36% vs. 19.36%) were significantly lower in death group. Multivariate Logistic
regression analysis showed that CTP score (OR = 2.565, 95%CI: 1.201～5.476, P = 0.015) and TBRR
(OR = 1.056, 95%CI: 1.002～1.114, P = 0.043) were independent risk factors for mortality. The
prognostic model was formulated as: CTP-TBRR = －13.729 + 0.942 × CTP + 0.055 × TBRR. The
area under the ROC curve of CTP score, TBRR and the CTP-TBRR model were 0.803, 0.772 and
0.944, respectively. The combined model demonstrated superior predictive performance compared
to individual parameters (z = 3.158, 3.423, P = 0.0016, 0.0006), with sensitivity and specificity of
91.9% and 86.1%. Conclusions The optimal adsorption duration for DPMAS therapy in ACLF
patients was 3 hours. CTP score and TBRR can effectively predict the efficacy of DPMAS in
patients with ACLF, and the CTP-TBRR prediction model can further optimize the artificial liver
treatment strategy.
|
|
基金项目：
|
|
|
|
作者简介：
|
|
|
|
参考文献：
|
|
|
|
服务与反馈：
|
|
【文章下载】【加入收藏】
|
|
|
|