Association between EASIX and acute kidney injury in critically ill cancer patients

Abstract

Background

To analyze the relationship between endothelial activation and stress index (EASIX) and the occurrence of acute kidney injury (AKI) in critically ill cancer patients.

Methods

Critically ill cancer patients were selected from the Medical Information Mart for Intensive Care IV (MIMIC-IV). Multivariate logistic regression was used to analyze the association between EASIX and the occurrence of AKI in critically ill cancer patients.

Results

One thousand forty-one cancer patients were retrospectively included, including 607 men and 434 women with mean age of 64.86 ± 13.67 years. Univariate analysis showed that high EASIX levels were associated with an increased risk of AKI occurrence in intensive care unit (ICU) cancer patients (OR: 1.47,95% CI: 1.13–1.91, P < 0.05). After adjusting for other confounders, high EASIX levels remained an independent risk factor predicting the development of AKI (OR: 1.42,95% CI: 1.08–1.88, P < 0.05). Trends in effect sizes were generally consistent across all subgroups in the prespecified subgroup analyses.

Conclusion

EASIX is an independent risk factor for AKI in critically ill cancer patients. More prospective studies are needed to validate the effect of EASIX on the occurrence of AKI in critically ill cancer patients in the future.

Introduction

With the advancement and development of medical technology, the overall survival rate of cancer patients has increased significantly in recent years, becoming a growing group of intensive care unit (ICU) patients [1]. Acute kidney injury (AKI) has an incidence of 11–22% in cancer patients and is a common complication [1].The incidence of AKI in critically ill cancer patients in the ICU was more than 50%, which was significantly higher than in non-critically ill patients [2]. AKI appreciably increases the risk of short-term mortality in critically ill cancer patients in the ICU [3].

Endothelial injury is a risk factor for AKI and is detrimental to renal recovery [4, 5]. This may be due to the loss of endothelial cell function which affects peritubular microvascular blood flow and barrier function of the kidney [6]. Endothelial activation and dysfunction-related markers found to be associated with sepsis-associated AKI, risk of AKI after severe trauma [7, 8]. Tumor progression and metastasis are also strongly associated with endothelial dysfunction [9], and endothelial dysfunction is also a complication of anticancer therapy [10]. However, none of the studies have explored the relationship between endothelial functional status and the risk of AKI development in cancer patients. In addition, the selection and testing of endothelial function markers may limit their study in critically ill patients. Recently, a new index based on lactate dehydrogenase (LDH), creatinine, and platelet count calculations, the endothelial activation and stress index (EASIX), has been used to assess endothelial status and function as well as has been found to be associated with the risk of death in critically ill and cancer patients [11,12,13].

Therefore, this study intends to investigate the relationship between EASIX and the risk of AKI in critically ill cancer patients in ICU. To provide a convenient tool for identifying critically ill cancer patients with high AKI risk in ICU.

Materials and methods

Source of data and samples

We conducted a single-center retrospective cohort study. We collected all relevant data from the Medical Information Mart for Intensive Care (MIMIC-IV version 2.2) database [14]. The MIMIC database is an open and freely available clinical database developed by the Massachusetts Institute of Technology(MIT) and Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA. The database records clinical medical information on patients admitted to the ICU at Beth Israel Deaconess Medical Center from 2008 to 2019. Information recorded in the database includes basic information, vital signs, additional tests, medication status and diagnoses. The Institutional Review Boards at MIT and Beth Israel Deaconess Medical Center approved the database. The database is available to researchers who have received a certificate after completing an online course on the protection of human subjects organized by the National Institutes of Health. All patient information in the MIMIC-IV database is anonymous and does not require informed consent from the patient.

Selection criteria

This study focused on patients with cancer in the ICU. The diagnosis of cancer was confirmed based on International Classification of Diseases 9th edition (ICD-9) or ICD-10 codes.The exclusion criteria were as follows: 1) age < 18 years old; 2) ICU stay < 24 h, 3) renal replacement therapy (RRT) at admission; 4) AKI at baseline (first 24 h in ICU); 5) estimated glomerular filtration rate (eGFR) < 15 ml/min at baseline (first 24 h in ICU); 6) missing measurement of baseline LDH, creatinine and platelet count. The screening process is shown in Fig. 1.

Fig. 1

The flowchart of patient selection. Abbreviations: MIMIC-IV: Medical Information Mart for Intensive Care IV; ICU: intensive care unit; RRT, renal replacement therapy; AKI, Acute kidney injury; eGRF, estimated glomerular filtration rate; LDH, lactate dehydrogenase

Relevant definitions

EASIX was defined [12] as EASIX = LDH [U/L] × creatinine [mg/dl] / platelet count. Relevant values were taken as first LDH value, blood creatinine value, platelet count on admission to ICU.

AKI judgment: according to Kidney Disease: Improving Global Outcomes (KDIGO) Clinical Practice Guidelines [15]. The time cutoff was between 24 h admission to the ICU and discharge from the ICU.

Baseline variables

Clinical data of patients who matched the inclusion criteria were collected from the MIMIC-IV v2.2 database by two researchers using PostgreSQL 10 software, by applying the structured query language (SQL). The data collected included: demographic variables such as patients' gender, age, and ethnicity. As well as disease information such as tumor type, chronic kidney disease (CKD), sepsis, heart failure, hypertension, and diabetes. Data are obtained on all vital signs, laboratory findings, 24-h urine output, mechanical ventilation, angiotensin, angiotensin-converting enzyme inhibitors/angiotensin receptor inhibitor (ACEI/ARB) use, nephrotoxic antibiotics, and platelet transfusions. The quick sequence organ failure assessment (qSOFA) score and the charlson comorbidity index (CCI) were calculated at the time of ICU admission.

Statistical analysis

R version 4.3.2 was used to statistically analyze the extracted data. Missing values were filled in using multiple interpolation. The “maxstat” (maximum selective rank statistic) statistic was used to determine the optimal critical value for continuous variables, and the cohort was categorized into low (< 2.06) or high (≥ 2.06) EASIX groups to compare the characteristics of the two cohorts. Continuous variables were tested for normality, with information conforming to a normal distribution expressed as mean and standard deviation expressed as independent samples t-test for comparisons between groups, and information with a non-normal distribution expressed as M(Q1,Q3) for comparisons between groups using the Mann Whitney U test. Measurement data were expressed as n (%) using the chi-square test. To further assess the relationship between EASIX levels and outcomes, a univariate logistic regression analysis was performed. Subsequently, multifactorial logistic regression analyses were performed to calibrate subsequent significant covariates. Odds ratios (OR) were calculated using 95% confidence intervals (CI) to indicate efficacy. We also controlled for factors associated with outcome using multifactor analysis for confounders; in model 2, confounders were adjusted for sepsis, systolic blood pressure (SBP), diastolic blood pressure(DBP), respiratory rate, anion gap, 24 h urine output, CCI, qSOFA, whereas in model 3, confounders were adjusted for sepsis, SBP, DBP, respiratory rate, anion gap, 24 h urine output, CCI, qSOFA, mechanical ventilation, vasopressors, nephrotoxic antibiotics. in addition, a subgroup analysis to confirm the validity of our results. P < 0.05 was considered statistically significant difference.

Results

Patient characteristics

The optimal cutoff value for EASIX was determined using the maximum selective rank statistic with AKI occurrence as the outcome (Fig. 2). The cohort was divided into an EASIX < 2.06 group and an EASIX ≥ 2.06 group, and the baseline characteristics of the patients in both groups were compared (Table 1) The incidence of AKI was significantly higher in the high EASIX group than in the low EASIX group (37.06% vs. 28.55%, P < 0.05). In contrast to the low EASIX group, patients in the high EASIX group had more male members and a higher incidence of hematoma than the low EASIX group. Respiratory rate, body temperature, red cell distribution width (RDW), blood urea nitrogen (BUN), hemoglobin, hematocrit, anion gap, international normalized ratio (INR) and prothrombin time (PT) were higher in the high EASIX group than in the low EASIX group, while eGFR, White blood cell (WBC), glucose and calcium were lower than in the low EASIX group. Patients in the high EASIX group had higher qSOFA scores than those in the low EASIX group, while CCI scores were not significantly different. The use of antibiotics for kidney injury was significantly more in patients in the high EASIX group than in the low EASIX group.

Fig. 2

Maximum Selection Process Note: The black dashed line in the figure shows the maximum value of the standardized statistic. Abbreviations: EASIX, endothelial activation and stress index

Table 1 Comparison of baseline characteristics of patients grouped according to EASIX truncation values

Relationship between EASIX and AKI

After adjusting for potential confounding variables, we developed different models to assess the independent effect of EASIX on AKI in cancer patients (Table 2). Model 1 was a univariate logistic regression, which showed that the high(≥ 2.06) EASIX group contributed to a 1.47-fold risk of AKI occurrence compared with the low EASIX group (OR: 1.47,95% CI:1.13–1.91, P < 0.05). Model 2, after adjusting for the confounders of sepsis, SBP, DBP, respiratory rate, anion gap, 24 h urine output, CCI,and qSOFA, revealed that the high(≥ 2.06) EASIX group was still associated with the development of AKI (OR: 1.44,95%CI: 1.10–1.88, P < 0.05). Model 3 further adjusted for the confounders of mechanical ventilation, vasopressors, and nephrotoxic antibiotics based on model 2, and showed that high(≥ 2.06) EASIX was an independent risk factor for the development of AKI (OR: 1.42, 95% CI: 1.08–1.88, P < 0.05).

Table 2 Association of EASIX with the risk of AKI in cancer patients in the ICU

Subgroup analysis

Subgroup analyses were performed for age (< 60 or ≥ 60 years), tumor type (hematologic malignancy or solid tumor), heart failure (with or without), CCI score (< 6 or ≥ 6), qSOFA score (< 2 or ≥ 2), mechanical ventilation (with or without), vasopressin (received or not), and exposure to nephrotoxic antibiotics (yes or no) (Fig. 3). As shown in Fig. 3, high EASIX (≥ 2.06) is a risk factor for AKI in patients aged ≥ 60 years(OR 1.62,95% CI:1.16–2.27), patients without heart failure (OR 1.42,95% CI:1.04–1.94), patients with CCI ≥ 6(OR 1.70,95% CI:1.18–2.45), patients with qSOFA score ≥ 2 (OR 1.41,95% CI:1.03–1.91), patients who have not received mechanical ventilation (OR 2.02,95% CI:1.10–3.72), patients who have not received vasopressin (OR 1.61,95% CI:1.17–2.22), and patients who have not been exposed to nephrotoxic antibiotics (OR 1.66,95% CI:1.08–2.56). It is noteworthy that the significance of EASIX predicting the risk of AKI seems to be more prominent in solid tumor patients (P_interaction = 0.017).

Fig. 3

Subgroup analysis of EASIX and AKI. Abbreviations: AKI, Acute kidney injury; OR, odds ratio; CI, confidence interval; CCI, Charlson comorbidity index; qSOFA, Quick sequential organ failure assessment; EASIX, endothelial activation and stress index

Discussion

As far as we know, this is the first study that has explored the association between EASIX and the risk of AKI in a specific group of critically ill cancer patients in the ICU. We observed a strong association between EASIX and the occurrence of AKI in critically ill cancer patients in the ICU. Critically ill cancer patients with high EASIX values were significantly more likely to develop AKI than those with low levels of EASIX. And high EASIX level was an independent risk factor for AKI in ICU critical cancer patients.

AKI has a significant impact on cancer patients and may affect subsequent treatment and even lead to higher mortality. People are concerned about the occurrence of AKI in cancer patients mainly attributed to the potential manifestations of malignant tumors and the side effects of various therapies used to treat malignant tumors [2]. The kidney is a highly vascularized organ. The renal vascular system consists of different populations of endothelial cells, which have specific structures and functions according to their microenvironment, and are involved in filtration barriers and reabsorption [5]. Endothelial damage induced by toxins, antibodies, immune cells, or inflammatory cytokines, or defects in factors that provide endothelial protection (e.g., complement or angiogenesis modulators), can lead to acute or chronic kidney injury [5].Endothelial dysfunction is associated with the onset and development of AKI and is an early event in the development of AKI [16].

Measuring biomarkers of endothelial dysfunction is a costly procedure that cannot be performed at many institutions. Previous studies have shown that endothelial dysfunction is present in patients with COVID-19, and that high EASIX ratios are strongly associated with high levels of serum endothelial markers [angiopoietin-2 (ANG-2), soluble thrombomodulin (sTM), and suppressor of tumorigenicity-2 (ST-2)] [17]. Endothelial dysfunction of hepatic sinusoidal endothelial cells contributes to liver fibrosis and slow plus acute liver injury, and EASIX may be a potentially useful marker for evaluating the prognosis of patients with such end-stage liver disease [18]. The indicators used in EASIX are derived from the results of common laboratory tests, including LDH, creatinine, and platelet count, which are readily available and affordable. These markers are markers of endothelial pathology and are associated with inflammation, oxidative stress, and other processes [19]. Thus EASIX is an index to assess endothelial dysfunction and is associated with the development of AKI in critically ill cancer patients.

Based on the formulation of EASIX acquisition, the occurrence of AKI in critically ill cancer patients is associated with higher LDH and lower platelet levels, in addition to higher EASIX values due to elevated creatinine caused by renal injury as a result of endothelial dysfunction. Activation injury of endothelial cells can lead to elevated serum levels of the mesocytoplasmic enzyme LDH [20]. Endothelial cell damage and dysfunction can also be part of vascular structural abnormalities that strongly promote pro-inflammatory signaling within cancer cells, leading to activation of the NF-κB pathway in cancer cells to promote invasion and metastasis, resulting in a vicious cycle of abnormal perfusion, local hypoxia, and outflow of cancer cells from the tumor [21]. LDH regulates the conversion of glucose to lactate, and increased metabolism in rapidly dividing cancer cells can also lead to elevated LDH [22]. Platelets support the integrity of the endothelial barrier by stabilizing adhesion junctions between neighboring endothelial cells through the release of a large number of mediators, such as platelet factor-4 (PF-4), endothelialin, or thrombospondin-1 (TSP-1) [23]. Decreased platelet counts are the result of activation of platelets by constant interaction with the endothelial surface after severe acute endothelial injury, which can further induce thrombotic microangiopathy (TMA) leading to kidney injury [24]. Interestingly, the proportion of critically ill cancer patients in the high EASIX index group who were transfused with platelets was higher than that in the low EASIX group, but the incidence of AKI was still higher, suggesting that patients in the high EASIX group did not fully benefit from platelet therapy by transfusion for the amelioration of AKI (Table 1).

The subgroup analyses showed roughly consistent trends in effect sizes across all subgroups, suggesting that EASIX is strongly associated with the risk of developing AKI in patients with critically ill cancer. We also found a possible interaction between EASIX and tumor type. Patients with solid tumors with high EASIX had a higher risk of AKI than hematological malignancies. To our knowledge, this is the first study to explore these interactions. The underlying mechanisms are not unknown. Further studies are needed. In the other subgroups, no significant interaction was observed (P_interaction > 0.05).

Limitations of our study

First, the present study cohort was derived from the MIMIC-IV database, which is so limited in its representativeness that potential selection bias cannot be avoided, and further validation in populations from different regions is needed to broaden its applicability in the future. Second, this is a retrospective study that does not allow causal inference, and prospective studies are needed to address this issue. Third, EASIX results were obtained from a single blood test. Serial testing may be more beneficial than a single test performed on admission, and the impact of dynamic changes in EASIX on AKI in critically ill cancer patients needs to be further validated.

Conclusion

In conclusion, we demonstrated that EASIX is associated with an increased incidence of AKI in critically ill cancer patients. EASIX is a conveniently available and reliable index for assessing endothelial injury and can be used to predict the occurrence of AKI in critically ill cancer patients in the ICU. This finding needs to be confirmed by prospective studies in the future.