中国血液净化

中国血液净化 ›› 2022, Vol. 21 ›› Issue (07): 473-477.doi: 10.3969/j.issn.1671-4091.2022.07.003

• 临床研究 • 上一篇    下一篇

纳入血清胱抑素C的非糖尿病腹膜透析患者技术生存率列线图建立及验证

张锁建   蔡小琴   叶 慧   李海涛   

  1. 225400 泰州，1泰兴市人民医院肾内科

  • 收稿日期:2022-01-13 修回日期:2022-03-16 出版日期:2022-07-12 发布日期:2022-07-12
  • 通讯作者: 李海涛
  • 基金资助:
    泰兴市人民医院院级课题(try2101)； 江苏大学临床医学科技发展基金项目(JLY2021193)； 蚌埠医学院科技项目(2020byzd272)

Establishment and validation of the Nomogram incorporating serum cystatin C for the technical survival of non-diabetic peritoneal dialysis patients

  1. Department of Nephrology, Taixing People's Hospital, Taizhou 225400, China
  • Received:2022-01-13 Revised:2022-03-16 Online:2022-07-12 Published:2022-07-12

PDF

81

摘要： 探索建立纳入血清胱抑素C的非糖尿病腹膜透析(peritoneal dialysis，PD)患者技
术生存率列线图，并进行验证。 方法 收集2010年1月～2019年1月在泰兴市人民医院新进入PD治疗
的非糖尿病患者，收集患者的一般人口学资料和实验室指标；开始规律PD治疗后1个月计算尿素清除指
数(urea clearance index，Kt/V)和 4h 时 PD 液/血液肌酐比值(4h dialysate to plasma creatinine
ratio，4h D/PCr)。记录患者第1年中发生腹膜炎次数。采用COX回归分析PD患者退出的危险因素。纳
入有意义指标，使用R语言绘制列线图，个体化预测PD患者1、3、5年技术生存率的情况，并采用校准曲线
图、Harrell一致性指数(concordance index，C-index)进一步验证列线图预测PD患者技术生存率的准
确性。 结果 共 163 名患者入组，男性 90 名(55.21%)，女性 73 名(44.78%)，平均年龄(48.75±13.23)
岁。血清胱抑素 C 值为(5.56±1.08)mg/L。单因素 COX 分析显示：较高的血清胱抑素 C(HR=1.275，
95% CI：1.031～1.577,P=0.026)、男性(HR=0.582，95% CI：0.350～0.968，P=0.037)、第 1 年发生腹膜炎
次数(HR=1.564，95% CI：1.002～2.440,P=0.049)、较高的 4h D/PCr(HR=1.217,95% CI：0.967～1.531,
P=0.095
)是PD治疗失败的危险因素，较高的血清白蛋白(HR=0.928,95% CI：0.876～0.982,P=0.010)、残
余肾功能(residual renal function，RRF)Kt/V(HR=0.174,95% CI：0.078～0.388,P＜0.001)和总 Kt/V
(HR=0.276,95% CI：0.137～0.554,P＜0.001)是PD治疗技术存活的保护因素。因RRF Kt/V和总Kt/V与
血清胱抑素C存在共线性，故未纳入COX多因素回归。多因素COX回归模型发现，校正性别、第1年发生腹
膜炎次数、血清白蛋白、4h D/PCr后，较高的血清胱抑素 C 是 PD 治疗失败的独立危险因素(HR=1.288，
95% CI：1.032～1.603,P=0.025)。绘制的纳入血清胱抑素C的列线图预测PD患者的技术生存率C-index
为0.712(95% CI：0.652～0.772)，校准曲线图显示预后列线图模型和实际观察值之间一致性较好。 结
论 透析前较高的血清胱抑素C水平的非糖尿病PD患者技术生存率较低。纳入血清胱抑素C的列线图对
非糖尿病PD患者的技术生存率有一定的预测价值。

关键词: 胱抑素C, 列线图, 腹膜透析

Abstract: Objective To establish and verify a Nomogram incorporating serum cystatin C for the technical survival of non- diabetic peritoneal dialysis (PD) patients. Methods We included non- diabetic patients
who newly admitted to PD treatment in our Hospital between January 2010 to January 2019. Demographic
and biochemical data were collected. Urea clearance index and 4h dialysate to plasma creatinine ratio (4h
D/PCr) were recorded 1 month after starting regular PD treatment. Episodes of peritonitis in the first year of
PD were also recorded. Risk factors for end-point events were analyzed by Cox regression model. The Nomogram was used to evaluate the prediction of the 1,3 and 5-year technical survival rate of PD patients by meaningful indicators. The calibration curve and Harrell concordance index (C-index) were used to verify the accuracy of the Nomogram in predicting the technical survival rates of PD patients. Results We enrolled 163 patients, including 90 males (55.21%) and 73 females (44.78%), with an average age of (48.75±13.23) years.
The serum cystatin C level was (5.56±1.08)mg/L. Univariate Cox regression showed that higher serum cys-tatin C (HR=1.275, 95% CI：1.031～1.577, P=0.026), male (HR=0.582, 95% CI：0.350～0.968, P=0.037),
peritonitis episodes in the first year (HR=1.564, 95% CI：1.002～2.440, P=0.049), and higher 4h D/PCr (HR=
1.217, 95% CI：0.967～1.531, P=0.095) were risk factors for PD failure, while higher serum albumin (HR=
0.928, 95% CI：0.876～0.982, P=0.010), residual renal function (RRF), urea clearance index (Kt/V) (HR=
0.174, 95% CI：0.078～0.388, P＜0.001) and total Kt/V (HR=0.276, 95% CI：0.137～0.554, P＜0.001) were
protective factors for PD technical survival. Because serum cystatin C had a collinearity with RRF Kt/V and
total Kt/V, the Kt/V measurements were excluded. After incorporating the covariates with P ＜0.1 into a multivariate Cox regression, serum cystatin C was an independent risk factor for PD failure (HR=1.288, 95% CI：
1.032～1.603, P=0.025). The C-index of Nomogram predicting the technical survival rate of PD patients was
0.712(95% CI：0.652～0.772). The calibration curve also showed a good consistency between the prognostic
Nomogram model and the actual observed values. Conclusion Technical survival was lower in non-diabetic
PD patients with higher serum cystatin C levels before dialysis. The Nomogram incorporating serum cystatin
C has a certain predictive value for the technical survival of PD patients

Key words: Cystatin C, Nomogram, Peritoneal dialysis

中图分类号: