Impact of nutritional index on contrast-associated acute kidney injury and mortality after percutaneous coronary intervention

The risk of malnutrition in acute kidney injury and mortality in coronary artery disease patients has not been studied. This study aimed to evaluate whether nutritional status assessed by Onodera’s prognostic nutritional index (PNI) was related to percutaneous coronary intervention (PCI) outcomes. A total of 3731 patients who received PCI between January 2010 and December 2018 were included. The relationship between PNI at the time of PCI and the occurrence of contrast-associated acute kidney injury (AKI) and all-cause death was evaluated using logistic regression and Cox proportional hazards models, respectively. AKI occurred in 271 patients (7.3%). A low PNI was independently associated with an increased risk of AKI on multivariate logistic regression analysis (OR 0.96, 95% CI 0.94–0.98, P = 0.001). During the median follow-up of 4.3 years, Kaplan–Meier analysis showed that patients with AKI/low PNI < 47.8 had a higher death rate. After adjusting for various risk factors, a low PNI was a significant risk factor for mortality (HR 0.98, CI 0.96–0.99, P = 0.003). A low level of PNI was associated with increased mortality, especially in the group aged over 70 years and female sex. PNI was closely associated with acute kidney outcomes and patient mortality after PCI.

. Clinical characteristics of subjects between no AKI and AKI. AKI acute kidney injury, BP blood pressure, RDW-CV red cell distribution width coefficient of variation, eGFR estimated glomerular filtration rate, CRP C-reactive protein, PNI prognostic nutritional index.  (Table 4).
We further analyzed the effects of PNI on various subgroups (Fig. 2). Low PNI levels were associated with increased mortality in the group over 70 years, female sex, no AKI, eGFR > 30 ml/min/1.73m 2 , and with or without diabetes. AKI had interaction with PNI (P for interaction = 0.033), which means PNI had different effect on mortality according to AKI or no AKI.

Discussion
PNI, as a clinical nutritional index, was closely associated with acute kidney outcome and patient mortality after PCI. Low PNI increases the risk of CA-AKI and mortality after PCI. Patients with CA-AKI and low PNI had the highest mortality rates. Low PNI increases mortality, especially in patients over 70 years or female, and is particularly related to cardiac death. Kidney dysfunction and cardiovascular disease are interconnected and can lead to long-term outcomes, such as mortality. The present study was the first to identify the impact of nutrition on CA-AKI and mortality after PCI. We defined CA-AKI according to the KDIGO guidelines 8 . Traditionally, contrast-induced AKI was defined as a rise in SCr of ≥ 0.5 mg/dL (≥ 44 μmol/L) or a 25% increase from the baseline value occurring two to five days after contrast exposure. However, this definition was limited in that it missed minor increases in creatinine that correlated with adverse outcomes 6 . Furthermore, other factors, such as medication or hypotension, may precipitate AKI after exposure to contrast media, so the term 'CA-AKI' is preferred 7 . The incidence of CA-AKI after PCI is varied due to different definitions of CA-AKI and study population characteristics. The incidence of CA-AKI after PCI in this study was 7.3%, and 226 (83.4%) of the cases belonged to the AKI stage I.
We confirmed that a low PNI increased the risk of CA-AKI. Similarly, previous studies have reported the effects of nutrition on AKI. Epidemiology study revealed that patients with NRS-2002 scores ≥ 3 had 1.8 times higher incidence of AKI in a 5-year retrospective cohort study 10 . Experimental studies demonstrated that several Table 3. Prognostic factors related to patients' mortality. HR hazard ratio, CI confidence interval, BP blood pressure, RDW-CV red cell distribution width coefficient of variation, eGFR estimated glomerular filtration rate, CRP C-reactive protein, PNI prognostic nutritional index, AKI acute kidney injury, ARB angiotensin II receptor blocker, ACEI angiotensin converting enzyme inhibitor.  www.nature.com/scientificreports/ nutrients serve as renoprotective agents in AKI. For instance, glutamine administration prevented AKI by downregulating oxidative stress in sepsis-induced AKI rat models 21 . Omega 3 fatty acids decrease polymorphonuclear leukocyte recruitment and cytokine levels in the kidney in ischemic renal injury model 22 . The current study was focused on PNI as a composite nutritional index; PNI is the summation of serum albumin and total lymphocyte counts. Albumin does not only represents the nutritional status 23 but also acts as an important circulating antioxidant 24 . Albumin acts through multiple binding sites and has free radical-trapping properties. The main mechanisms of contrast-induced nephropathy are vasoconstriction and oxidative stress, followed by renal medullary hypoxia and direct tubular toxicity by contrast media. Thus, a low level of albumin exacerbated CA-AKI. Likewise, lymphocytes can be used as a marker for nutritional status 23 , but various lineages of lymphocytes are also involved in the repair and protection of AKI 25 . We also revealed several risk factors for CA-AKI. Preexisting CKD, such as low eGFR and proteinuria, is the strongest risk factor for AKI. Serum uric acid is a risk factor for various heart and kidney diseases and was associated with an increased risk of AKI after coronary angiography 12 . The type and volume of the contrast material affected the incidence of AKI. Current guidelines recommend using iso-osmolar or low-osmolar contrast media with minimized volume (total contrast volume/GFR < 3.7) to prevent AKI 26 . All contrast media used in our patients were nonionic iso-osmolar or low-osmolarity media. The mean volume of contrast media injected was 243 mL, and 29.9% of patients received a dose over 3.7 in total contrast volume/GFR. For the high volume of contrast media associated with AKI, we should try to use contrast media cautiously in CKD patients. Diabetes mellitus was not related to the occurrence of AKI.
Various risk factors for mortality after PCI were identified, and AKI showed the highest HR on multiple Cox regression analysis. There is concern that CA-AKI may have been overestimated, affecting the negative effect of performing angiography in patients with CKD 27 . For instance, Ribitsch et al. showed that the occurrence of contrast-induced AKI did not negatively impact kidney function after three months or mortality after PCI in a prospective cohort 28 . However, Ribitsch's study included only 706 patients with a follow-up period of two years. With a large population and longer follow-up duration, our results clearly showed AKI after PCI as a risk factor for mortality; this is consistent with previous reports 29, 30 .
Old age, diabetes mellitus, smoking, and several laboratory values, such as RDW-CV and CRP, were independently associated with increased mortality. CRP, a well-known acute-phase reactant, reflects the degree of underlying inflammation 31 . RDW-CV is a hematological index used to explore the causes of anemia. However, it is now regarded as a prognostic marker that reflects the underlying inflammatory state and is related to patient outcomes 32 . Higher RDW-CV was also a predictor in patients undergoing PCI 33 in the previous report as well as CRP 34 . Our results are in line with those of previous studies, showing that inflammation played a crucial role in the initiation and progression of atherosclerosis, affecting clinical outcomes.
Meanwhile, a low PNI also increases the risk of mortality after PCI. A previous study also reported that PNI was associated with mortality in patients with stable coronary artery disease 19 . Nutritional influences on cardiovascular risk are well demonstrated. Poor quality of diet and undernutrition could aggravate insulin resistance and atherogenesis, promoting proinflammatory process, eventually leading to mortality 35,36 .
PNI did not show discrimination power for mortality in groups with AKI and eGFR < 30 ml/min/1.73 m 2 . This result could be due to the small population number in AKI (n = 271) groups and eGFR < 30 ml/min/1.73m 2 Figure 2. Cox hazard ratio of survival for PNI in various subgroups. A low level of PNI was associated with increased mortality in the group aged over 70 years, female sex, no AKI, and eGFR > 30 ml/min/1.73m 2 . PNI, prognostic nutritional index; AKI, acute kidney injury; eGFR, estimated glomerular filtration rate. www.nature.com/scientificreports/ (n = 154). Another study reported that malnutrition in AKI patients was associated with a higher incidence of complications, longer hospitalizations, and higher mortality 37 . We set the cutoff value of PNI as 47.8, defined by ROC curve using AKI prediction. ROC curve for mortality also set a similar cutoff value [data not shown]. However, we could not define a specific PNI cutoff value. After Onodera suggested the cutoff of PNI value as 45 13 , this is frequently used 17,38 . However, the researchers chose several cutoff levels according to study design, such as median, tertile, or ROC analysis 15,18,20,39 . In previous studies, the PNI cut-off value for defining undernutrition varied between 45 and 50. The optimal cut-off value and the classification of the groups by PNI remain unclear.
This study had several limitations. First, this was a single-center retrospective study. Many confounding factors could influence the prognosis of PCI. Second, the PNI was evaluated only once before the PCI. Laboratory changes after PCI were not counted for the analysis. Third, the detection rate of AKI may have been underdiagnosed because of the early discharge of patients. Moreover, we could not suggest AKI to CKD progression or kidney function decline after PCI due to limited follow-up serum creatinine data. We did not consider heart failure; heart failure is one of the important factors for AKI and mortality. In addition, AKI and PNI are highly correlated. To solve the multicollinearity problem, we categorized and integrated AKI/PNI variables. Finally, we did not perform another nutritional assessment, such as subjective global assessment or body composition. However, we analyzed a large number of patients and analyzed long-term patient outcomes using national statistical data related to death.
In conclusion, poor nutritional status, as assessed by PNI, was a risk factor for CA-AKI and all-cause mortality in patients undergoing PCI. This finding suggested that the PNI score, which can be easily determined using serum albumin and lymphocyte counts, was a useful prognostic marker in patients undergoing PCI. Clinicians should be careful when diagnosing and treating patients with coronary artery disease with low PNI.

Methods
Study population. We retrospectively screened 4,096 patients who underwent PCI at Pusan National University Hospital between January 2010 and December 2018. If the subjects underwent repetitive PCI during the study period, only the first case was included. Patients who already received hemodialysis due to kidney failure (n = 137), died within 48 h after PCI (n = 52) or did not have serum creatinine (n = 45) or PNI (n = 131) values, were excluded. A total of 3,731 subjects were included in the final analysis.
Data sources and measures. The data were retrieved from electronic medical records. The data collected included age, sex, body mass index, history of diabetes, hypertension, blood pressure, and laboratory findings, including serum creatinine, hemoglobin, albumin, and urine dipstick albumin results. Mortality data were obtained from the Microdata Integrated System (MDIS) 8 , generated by Statistics Korea.
CA-AKI was defined as a rise in serum creatinine ≥ 0.3-or 1.5 than the baseline value within 48 h after PCI according to KDIGO guidelines 40 . Estimated GFR (eGFR) was calculated through the CKD-EPI equation 41 . Proteinuria was defined as a urine dipstick test result of ≥ 1 + 42 . PNI was calculated as 10 × serum albumin (g/ dL) + 0.005 × total lymphocyte count (per mm 3 ) 13 .
Statistical analysis. All variables with a normal distribution were expressed as mean ± standard deviation.
Categorical variables were expressed as numbers and proportions, and comparisons were made using the chisquare test. Logistic regression analysis was conducted to analyze the factors associated with AKI. The ability to predict CA-AKI was evaluated using the area under the curve (AUC) in the receiver operating characteristic (ROC) curve. The cumulative survival rates were estimated using the Kaplan-Meier method, and differences between survival curves were compared using the log-rank test. Multivariate Cox regression models for mortality were constructed using the confounding factors. Significant variables in the univariate model were entered into a multivariable model. Among the variables collected, serum albumin levels and lymphocyte counts were not included in the multiple logistic or Cox regression models because they were included in the calculation of PNI. Variables included in all multivariable analyses were tested for multicollinearity. We also examined the association between PNI and mortality through a subgroup analysis using a fully adjusted multivariate Cox regression model. The patients were stratified according to age, sex, presence of diabetes mellitus, baseline eGFR, and CA-AKI. All analyses were performed using the R statistical software package (version 4.0.1; R Development Core Team; R Foundation for Statistical Computing, Vienna, Austria). Ethical statement. The study was conducted in accordance with the Declaration of Helsinki, and the study was approved by the Institutional Review Board of Pusan National University Hospital (H-1909-014-083). The need for informed consent was waived because of its retrospective design.