The predictive values of monocyte-lymphocyte ratio in postoperative acute kidney injury and prognosis of patients with Stanford type A aortic dissection

Abstract

Objectives: Postoperative acute kidney injury (pAKI) is a serious complication of Stanford type A aortic dissection (TAAD) surgery, which is significantly associated with the inflammatory response. This study aimed to explore the relationship between blood count-derived inflammatory markers (BCDIMs) and pAKI and to construct a predictive model for pAKI.

Methods: Patients who underwent TAAD surgery were obtained from our center and the Medical Information Mart for Intensive Care (MIMIC)-IV database. The differences in preoperative BCDIMs and clinical outcomes of patients with and without pAKI were analyzed. Logistic regression was used to construct predictive models based on preoperative BCDIMs or white cell counts (WCCs). The performance of the BCDIMs and WCCs models was evaluated and compared using the receiver operating characteristic (ROC) curve, area under the ROC curve (AUC), Hosmer-Lemeshow test, calibration plot, net reclassification index (NRI), integrated discrimination improvement index (IDI), and decision curve analysis (DCA). The Kaplan-Meier curves were applied to compare the survival rate between different groups.

Results: The overall incidence of pAKI in patients who underwent TAAD surgery from our center was 48.63% (124/255). The presence of pAKI was associated with longer ventilation time, higher incidence of cerebral complications and postoperative hepatic dysfunction, and higher in-hospital mortality. The results of the logistic regression indicated that the monocyte-lymphocyte ratio (MLR) was an independent risk factor for pAKI. The BCDIMs model had good discriminating ability, predictive ability, and clinical utility. In addition, the performance of the BCDIMs model was significantly better than that of the WCCs model. Analysis of data from the MIMIC-IV database validated that MLR was an independent risk factor for pAKI and had predictive value for pAKI. Finally, data from the MIMIC-IV database demonstrated that patients with a high MLR had a significantly poor 28-day survival rate when compared to patients with a low MLR.

Conclusion: Our study suggested that the MLR is an independent risk factor for pAKI. A predictive model based on BCDIMs had good discriminating ability, predictive ability, and clinical utility. Moreover, the performance of the BCDIMs model was significantly better than that of the WCCs model. Finally, a high MLR was significantly associated with poor short-term survival of patients who underwent TAAD surgery.

Keywords: monocyte-lymphocyte ratio; postoperative acute kidney injury; predictive model; risk factor; type A aortic dissection.

Figure 1

BCDIMs values in patients with or without pAKI from our center. (A–E) The values of SIRI, SII, PLR, MLR, and NLR in patients with or without pAKI. BCDIMs, blood count-derived inflammatory markers; pAKI, postoperative acute kidney injury; SIRI, systemic inflammation response index; SII, systemic inflammation index; PLR, platelet–lymphocyte ratio; MLR, monocyte–lymphocyte ratio; NLR, neutrophil–lymphocyte ratio. **p < 0.01, ***p < 0.001.

Figure 2

The relationship between pAKI and clinical outcomes of patients who underwent TAAD surgery from the MIMIC-IV database. (A) The mean postoperative ventilation time of patients with or without pAKI. (B) The 1-year Kaplan–Meier curve of patients with or without pAKI. pAKI, postoperative acute kidney injury; TAAD, type A aortic dissection; MIMIC-IV database, the Medical Information Mart for Intensive Care (MIMIC)-IV database. *p < 0.05.

Figure 3

The nomogram of the predictive model based on BCDIMs. In the nomogram, gender 0 represents female, gender 1 represents male, smoking 0 represents non-smoker, and smoking 1 represents smoker. BCDIMs, blood count-derived inflammatory markers; BMI, body mass index; NLR, neutrophil–lymphocyte ratio; MLR, monocyte–lymphocyte ratio; AKI, acute kidney injury.

Figure 4

Comparison of the performance of BCDIMs model and WCCs model. (A) ROC curves and the AUC of BCDIMs model and WCCs model. (B) The calibration plot of BCDIMs model and WCCs model. (C) The reclassification plot of BCDIMs model and WCCs model. The black circles represent patients without pAKI, and red circles represent patients with pAKI. (D) The decision curve of BCDIMs model and WCCs model. BCDIMs, blood count-derived inflammatory markers; WCCs, white cell counts; ROC, receiver operating characteristic; AUC, area under ROC curve; pAKI, postoperative acute kidney injury.

Figure 5

Comparison of performance of BCDIMs-MIMIC model and WCCs-MIMIC model. (A) ROC curve and the AUC of BCDIMs-MIMIC model and WCCs-MIMIC model. (B) The calibration plot of BCDIMs-MIMIC model and WCCs-MIMIC model. (C) The reclassification plot of BCDIMs-MIMIC model and WCCs-MIMIC model. The black circles represent patients without pAKI, and red circles represent patients with pAKI. (D) The decision curve of BCDIMs-MIMIC model and WCCs-MIMIC model. BCDIMs, blood count-derived inflammatory markers; WCCs, white cell counts; MIMIC, the Medical Information Mart for Intensive Care; ROC, receiver operating characteristic; AUC, area under ROC curve; pAKI, postoperative acute kidney injury.

Figure 6

The relationship between MLR and survival of patients who underwent TAAD surgery from the MIMIC-IV database. (A) The 1-year Kaplan–Meier curves of patients with a high or low MLR. (B) The 28-day Kaplan–Meier curves of patients with a high or low MLR. MLR, monocyte–lymphocyte ratio; TAAD, type A aortic dissection; MIMIC-IV database, the Medical Information Mart for Intensive Care (MIMIC)-IV database.