Predictive Role of the D-Dimer Level in Acute Kidney Injury in Living Donor Liver Transplantation: A Retrospective Observational Cohort Study

This study aimed to determine the association between serum D-dimer levels and the risk of acute kidney injury (AKI) in patients undergoing living donor liver transplantation (LDLT). Clinical data of 675 patients undergoing LDLT were retrospectively analyzed. The exclusion criteria included a history of kidney dysfunction, emergency cases, and missing data. The final study population of 617 patients was divided into the normal and high D-dimer groups (cutoff: 0.5 mg/L). After LDLT, 145 patients (23.5%) developed AKI. A high D-dimer level (>0.5 mg/L) was an independent predictor of postoperative development of AKI in the multivariate analysis when combined with diabetes mellitus [DM], platelet count, and hourly urine output. AKI was significantly higher in the high D-dimer group than in the normal D-dimer group (odds ratio [OR], 2.792; 95% confidence interval [CI], 1.227–6.353). Patients with a high D-dimer exhibited a higher incidence of early allograft dysfunction, longer intensive care unit stay, and a higher mortality rate. These results could improve the risk stratification of postoperative AKI development by encouraging the determination of preoperative D-dimer levels in patients undergoing LDLT.

D-dimer is a fibrin breakdown product that usually increases in patients with pulmonary embolism (PE), deep vein thrombosis (DVT), or disseminated intravascular coagulation (DIC). A high plasma D-dimer level has been identified as a prognostic factor in critically ill patients. In addition, the D-dimer level increases in patients with liver cirrhosis and is correlated with the severity of liver dysfunction [13,14]. D-dimer is emerging as an early marker of AKI. The D-dimer level significantly increases in patients with AKI and is associated with an increased risk of mortality in pregnant women [15]. In addition, a high D-dimer level has been associated with the development of contrast-induced AKI in patients undergoing percutaneous coronary intervention (PCI) [16]. However, the effects of a high D-dimer level on the development of AKI in patients undergoing LDLT have not been reported.
We investigated the correlation between the D-dimer level and the severity of postoperative AKI, and our results suggest that determining the D-dimer level along with other preoperative factors helps estimate the risk of AKI. In addition, we evaluated the prognostic value of the D-dimer level for postoperative morbidity and mortality in patients undergoing LDLT.

Ethical Considerations
This study was performed in accordance with the ethical standards of the Declaration of Helsinki. Ethical approval was obtained from the Institutional Review Board of Seoul St. Mary's Hospital (KC20RISI0176; 6 April 2020). A waiver of informed consent was granted for this retrospective study.

Study Population
The medical records were retrospectively collected between January 2009 and February 2020 for 657 patients who underwent LDLT. The exclusion criteria were cadaveric LT, history of kidney dysfunction (i.e., hepatorenal syndrome, CKD, or history of dialysis [17]), emergency cases, age < 19 years, or missing laboratory data. Ultimately, 58 patients were excluded and the remaining 617 patients were included in the analyses.

Living Donor Liver Transplantation
The transplant surgery and general anesthesia were performed by expert surgeons and anesthesiologists, respectively. The surgical techniques and anesthetic management were performed as described previously [18]. Briefly, the piggyback technique was applied, and an inflow modification (i.e., splenic artery ligation, portocaval shunt, or splenectomy) was performed at the surgeon's discretion. Balanced anesthesia with proper hemodynamic management was supplied. Packed red blood cells (PRBCs) were transfused to maintain hematocrit >25%, and fresh-frozen plasma (FFP) or single-donor platelets (SDPs) were transfused under the guidance of thromboelastography or according to the laboratory results [9]. Immunosuppressants were administered and were tapered after surgery in accordance with our hospital's protocol.

Acute Kidney Injury
Postoperative AKI was diagnosed based on the Kidney Disease Improving Global Outcomes criteria [19]. AKI severity was determined as follows: Stage 1, increase in serum creatinine (sCr) ≥ 0.3 mg/dL (in 48 h) or 1.5-1.9 multiplied by the baseline (in 7 days); stage 2, sCr 2.0-2.9 multiplied by the baseline; and stage 3, sCr ≥ 3.0 multiplied by the baseline or an increase in sCr ≥ 4.0 mg/dL at the beginning of renal replacement therapy [20]. The study population was divided into non-AKI and AKI groups.

Measurement of Laboratory Data
Laboratory parameters were evaluated preoperatively. Blood samples were collected into a Clot Activator tube (Becton Dickinson and Co., Franklin Lakes, NJ, USA) and analyzed using an automated chemistry analyzer (Hitachi 7600; Hitachi, Tokyo, Japan) to obtain the laboratory data on the day before surgery. A Sodium Citrate Tube (Becton Dickinson and Co.) and automated coagulation analyzer (CS-5100; SysmexCrop., Kobe, Japan) were used to measure D-dimer levels. A high D-dimer level was defined as >0.5 mg/L fibrinogen equivalent units, as described previously [21]. If samples were collected several times, the one collected nearest the time of surgery was analyzed.

Clinical Postoperative Outcomes
Postoperative outcomes included the duration of hospital stay, the duration of intensive care unit (ICU) stay, the incidence of early allograft dysfunction (EAD), graft rejection, and overall mortality. EAD was clinically determined by the presence of at least one of the following: total bilirubin ≥10 mg/dL or international normalized ratio (INR) ≥ 1.6 on postoperative day 7 and alanine transaminase or aspartate transaminase ≥ 2000 IU/mL during the first week after surgery [25].

Statistical Analysis
The Mann-Whitney U test, χ 2 test, or Fisher's exact test were used to compare perioperative recipient and donor graft factors between the AKI and non-AKI groups. Tests for trends were conducted using a linear-by-linear association method. The values were expressed as the median (interquartile range [IQR]) and number (proportion). The associations between perioperative factors and the development of postoperative AKI were analyzed using logistic regression. Potentially significant factors (p < 0.1) in the univariate analysis were included in multivariate forward and backward logistic regression analyses. We conducted two separate multivariate analyses using the dichotomous or continuous D-dimer levels. Dichotomous D-dimer was included in the predictive multivariate model. The predictive performance of the models was estimated by the area under the receiver operating characteristic curve analysis. The relationships between the D-dimer level and inflammatory factors or the MELD were detected using Spearman's rank correlation coefficient analysis. A p-value < 0.05 was considered statistically significant. Statistical analyses were performed using MedCalc (ver. 11.0; MedCalc Software, Ostend, Belgium) and SPSS Statistics software (ver. 24.0; SPSS Inc., Chicago, IL, USA).

Analysis of Pre-and Intraoperative Clinical Data According to the Development of AKI
Intergroup differences in the preoperative recipient factors (i.e., MELD score, the incidence of ascites, DM, hemoglobin level, albumin level, platelet count, INR, and D-dimer level) were detected (Table 1). Differences in intraoperative recipient factors (i.e., total PRBC, FFP, and platelet transfusion amounts; average HR; average MBP; and hourly urine output) and donor graft factors (i.e., graft ischemic time) were observed between the groups ( Table 2).

Associations of Pre-and Intraoperative Factors with the Development of AKI
The D-dimer level was significantly associated with the development of AKI in the multivariate logistic regression (Table 3), when combined with the presence of DM, platelet count, and intraoperative hourly urine output (AUC: 0.672; 95% confidence interval (CI): 0.634-0.709; sensitivity: 72.4%; specificity: 55.9%; p < 0.001). Furthermore, the probability that patients with a high D-dimer level (>0.5 mg/L) would develop AKI was 2.79-fold higher than in those with a normal D-dimer level in the multivariate analysis using dichotomous D-dimer (odds ratio: 2.792; 95% CI: 1.227-6.353; p = 0.014) (Table S1).

Comparison of the Prevalence of AKI Stages between the Normal and High D-Dimer Groups
The prevalence of patients with postoperative AKI was higher in the high D-dimer group than in the normal group for every stage (stage 1 and stages 2-3) (p < 0.001; Table 4).

Relationship between the D-Dimer Level and the MELD Score
The D-dimer level was significantly correlated with the MELD score (Rho coefficient = 0.483) (p < 0.001).

Subanalysis of Patients with DM, HBV, or Heart Disease
The prevalence of AKI did not significantly differ between the normal and high D-dimer patients with DM when using the normal (0.5 mg/L) cutoff level. However, a significant difference (p = 0.008) was detected when using the alternative (1.1 mg/L) D-dimer cutoff level (Table S2). A significant difference in D-dimer level was observed between the AKI and non-AKI groups of patients with DM (Table S3). There were differences between patients with HBV and heart disease (diastolic dysfunction) with AKI vs. non-AKI and between such patients with a normal D-dimer level vs. a high D-dimer level (Tables S4-S7).

Postoperative Outcomes
Patients with a high D-dimer level had a longer ICU stay, higher incidence of EAD, and a higher mortality rate than did patients with a normal D-dimer level (p = 0.036, p = 0.005, and p = 0.02, respectively) ( Table 5). Patients with AKI had a longer duration of hospitalization, longer ICU stay, higher incidence of EAD, and higher overall patient mortality rate than did patients without AKI (p < 0.001, p = 0.002, p = 0.002, and p = 0.013, respectively) ( Table 6).

Discussion
The main finding of this study was that a high D-dimer level (>0.5 mg/L) was an independent predictor of postoperative development of AKI when combined with platelet count, DM, and hourly urine output. The prevalence of AKI was significantly higher in the high D-dimer group than in the normal D-dimer group. The D-dimer level and the proportion of patients with a high D-dimer level significantly increased with more severe AKI. In addition, patients with a high D-dimer level or AKI had more severe morbidities and higher mortality rates.
AKI is a common postoperative complication in LT recipients [3]. Although the etiology has not been fully elucidated, renal hypoperfusion, inflammation, hypovolemia, and the use of nephrotoxic drugs are possible contributing factors [26,27]. Recent studies have shown that the systemic inflammatory response plays a critical role in the development of AKI [28,29]. Systemic inflammation causes AKI by targeting tubular epithelial cells (TECs). Inflammatory mediators may contribute to reducing the blood flow in the outer medulla of the kidneys [30]. Serum CRP impairs G1/S-dependent TEC regeneration, while IL-6 and tumor necrosis factor-alpha (TNF-α) interact with TECs, causing renal injury [31,32]. Immune responses also contribute to renal damage after reperfusion injury [33]. The inflammatory reaction of the innate and adaptive immune systems in the post-ischemic kidney is an important factor in the pathogenesis of ischemia-reperfusion injury (IRI). Postoperative AKI is a major risk factor for morbidity and mortality after LDLT and is important for predicting the development of AKI [34]. In the present study, the AKI group had a higher incidence of EAD, longer hospital and ICU stays, and lower overall survival than the non-AKI group.
D-dimer is a fibrin degradation product that is widely measured to aid in the diagnosis of PE. Although D-dimer is a well-known marker for clotting disorders, such as PE, DIC, and DVT, recent studies have shown that a high D-dimer level is also associated with the prognosis of critically ill patients [35,36]. A high D-dimer level is an emerging marker for kidney dysfunction. In patients with ST-elevation myocardial infarction undergoing PCI, D-dimer levels >0.69 µg/mL are an independent predictor of contrast-induced AKI [16]. A retrospective study revealed that the D-dimer level was significantly higher in pregnant women with AKI and was associated with higher mortality risk [15]. In a study of patients undergoing hematopoietic cell transplantation, a high D-dimer level caused by activation of the coagulation system was associated with the development of AKI [37]. Although the mechanism underlying the relationship between D-dimer and AKI is unclear, a high D-dimer level may exacerbate inflammatory coagulation and fibrinolysis (via inflammatory mediator surges, thrombin generation, fibrin formation/degradation, and platelet aggregation), thus contributing to the development of AKI in critically ill patients [38][39][40]. Therefore, the close relationship between the D-dimer level and inflammation may play a role in kidney damage. In the present study, a high D-dimer level was correlated with more severe liver dysfunction [13,14,41], suggesting its use as an early surrogate marker for hepatic decompensation that increases the risk of AKI in patients scheduled for LDLT.
Our multivariate logistic analysis suggested that other risk factors for postoperative AKI include preoperative DM, platelet count, and intraoperative hourly urine output. A diabetic kidney has decreased resilience to renal perfusion after ischemia. Habib et al. [42] suggested that delayed reperfusion in the renal cortex and increased apoptosis of proximal tubular cells were possible mechanisms of renal IRI in the diabetic kidney. Additionally, microvascular dysfunction may inhibit renal perfusion after ischemia [43]. The preoperative platelet count was a preoperative risk factor for the development of AKI in the present study. In a study of elderly patients, thrombocytopenia was associated with an increased risk of AKI [44]. The probability of AKI significantly increased in patients with severe thrombocytopenia after LDLT [45]. Considering the close relationship between portal hypertension and thrombocytopenia, a low platelet count may be associated with portal hypertension-related renal injury in patients with ESLD. A decrease in the hourly urine output during surgery was also significantly associated with postoperative AKI. Low urine output usually indicates hypovolemia or hypotension, both of which are related to decreased perfusion pressure in the afferent arterioles [46]. Intraoperative oliguria is associated with the development of postoperative AKI in patients undergoing major abdominal surgery [47]. Reduced urine output was associated with an increased risk of AKI in an ICU-based study [48].
Several limitations of this study should be discussed. First, the pathophysiological processes underlying the association between an elevated D-dimer level and the development of AKI are unknown. Although a high D-dimer level is associated with systemic inflammation, further studies are required. Second, we used the D-dimer reference range based on clotting disorders, such as PE. Further studies are required to establish the Ddimer range in the AKI setting. Third, hidden biases could have been present because this study constituted a retrospective analysis. Fourth, liver grafts from deceased donors are more closely associated with postoperative AKI than grafts from living donors [49]. Therefore, the association between the D-dimer level and the development of AKI could differ according to the donor type. Fifth, we could not include the length of waiting time for LT or the liver disease duration. These periods would provide important information because a patient's condition could become worse with a longer waiting time and longer duration of cirrhosis. However, all donors were relative volunteers (parents, sister, and brother); this aspect may have contributed to the shorter wait time and cirrhotic duration than in LDLT with other types of donations or DDLT [50,51]. The MELD score played a key clinical role in assessing the severity of cirrhosis [52,53].

Conclusions
The findings in this study suggest that a high D-dimer level (>0.5 mg/L) constitutes an independent predictor of postoperative development of AKI when combined with DM, platelet count, and hourly urine output. Because its odds ratio (2.792) was highest among those risk factors, the D-dimer level could provide critical information regarding vulnerability to AKI. Our findings suggest that a high D-dimer level is an early and promising marker for the development of AKI; it could provide useful information for understanding a patient's condition. Risk factors for the development of AKI, including the D-dimer level, must be assessed before surgery; patients with those risk factors should be carefully monitored.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jcm11020450/s1, Table S1. Associations of pre-and intraoperative factors with the occurrence of AKI in living donor liver transplantation. Table S2. Comparison of the prevalences of AKI between the normal and high D-dimer groups of DM patients. Table S3. Comparison of the D-dimer levels between the non-AKI and AKI groups of DM patients. Table S4. Comparison of the prevalences of AKI between the normal and high D-dimer groups of patients with hepatitis B virus. Table S5. Comparison of the D-dimer levels between the non-AKI and AKI groups of patients with hepatitis B virus. Table S6. Comparison of the prevalences of AKI between the normal and high D-dimer groups of patients with heart disease (diastolic dysfunction). Table S7. Comparison of the D-dimer levels between the non-AKI and AKI groups of patients with heart disease (diastolic dysfunction). Informed Consent Statement: Patient consent was waived due to the retrospective study design.

Data Availability Statement:
The data presented in this study are available on reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.