Acute Renal Failure and Mortality After Open-Heart Surgery in Infants

Abstract

Acute renal failure (ARF) is a major complication in infants who undergo cardiac surgery. The aim of this investigation was to identify possible risk factors for ARF and mortality in this patients group. Out of 64 patients, 21 (32.8%) cases developed acute renal failure and overall mortality rate was 25%. The mortality rate was higher in the infants who developed ARF than those who did not (66.7% and 4.7%, respectively, p < 0.05). Also, ARF was positively correlated with mortality (r:0.70, p < 0.0001). The nonsurvivors had lower mean serum albumin than did the survivors (p < 0.05), and serum albumin level was negatively correlated with mortality (r = − 0.34, p < 0.05). For the patients with serum albumin level < 3.5 g/dL, the unadjusted odds ratio for mortality was 4.3 (CI 95%:1.05 − 17.86). Total bypass time and aorta clamping time were significantly longer in the nonsurvivor group than in the survivor group (p < 0.05 for both). In conclusion, the significant risk factors for mortality in these patients were development of ARF, low serum albumin level, and long total bypass and aorta clamping times, which may be predictive of poor prognosis.

Keywords:

• acute renal failure
• cardiac surgery
• childhood
• mortality

Introduction

Cardiac surgery in infants is associated with high mortality, and the reported rates range from 30%–79%.[1-3] Acute renal failure (ARF) remains a major problem after these procedures, and is a major cause of death in this patient group. Although significant advances have been made in cardiac surgery and intensive care techniques for children, the incidence of ARF in this group has changed little during the past decade. Current reports document ARF rates of 1% to 39% for children who have undergone heart surgery.[1&2] The etiology of renal insufficiency after cardiac surgery in children is poorly understood, but ischemic injury of the kidney due to inadequate perfusion is believed to be a major factor.[3]

Several studies have investigated predictors of outcome in children who develop ARF after open-heart surgery.[1], [3] Some of the factors that have been identified as indicators of poor prognosis in this patient group include age, complexity of the cardiac lesion, duration of cardiopulmonary bypass, central venous hypertension, systolic arterial hypotension, and need for dialysis.[3] Hypoalbuminemia has been identified as an independent predictor of high mortality in adult patients with end-stage renal disease or ARF.[4] However, the literature contains little information about the relationship between hypoalbuminemia and mortality in pediatric patients with ARF.[4]

Our aim in this study was to identify possible risk factors for ARF and mortality in pediatric patients who undergo cardiac surgery. We also examined the link between serum albumin level and mortality in children with postoperative ARF.

Patients and Methods

The study involved 64 pediatric patients, (22 females and 42 males; median age 3 months; age range 0–22 months) who underwent open-heart surgery at our center in a 6-month period. Data related to the surgical procedures and clinical course of each case were obtained from hospital records. Every child received furosemide (1–3 mg/kg/day) in the postoperative period. For each case, we recorded pre- and postoperative clinical and biochemical data as follows: serum levels of albumin, blood urea nitrogen, creatinine, and C-reactive protein (C-RP); hematocrit; platelet count; central venous pressure; systemic arterial pressure; cardiopulmonary bypass time; aorta clamping time; fluid intake and urine output; inotropic drugs received and their dosages; and arterial blood gas parameters.

Analyses for hematocrit and platelets were performed by cell analyzer (Celldyne 3700, Abott, USA). Serum albumin, creatinine, and BUN levels were measured by quantitative calorimetric and C-RP by the turbimetric method (P Moduler Autoanalyzer Hitachi, Rooche, USA). Systolic and diastolic arterial hypotension were defined according to reference levels. Central venous hypertension was defined as central venous pressure > 9 cm of H₂O. Low cardiac output was defined as low mixed venous oxygen saturation (< 60% for one hour) or echocardiographic evidence of a low flow state.

Acute renal failure was defined as serum creatinine > 1.2 mg/dL and/or urine output < 0.5 mL/kg/h for more than 4 hours with no response to adjustments of fluid therapy, diuretics, or inotropic support. All patients who developed ARF were treated with peritoneal dialysis. In each case, this was performed using a Tenckhoff pediatric catheter placed via a midline subumbilical incision. Standard dialysis solution containing lactate as a buffer was used. The exchange volume was 10–20 mL/kg body weight. The peritoneal dialysis procedure included 15 minutes of dwelling time and 30–40 minutes of drainage time. The frequency of peritoneal exchange and the dextrose concentration was established according to patient needs. In the survivors, peritoneal dialysis was generally stopped when the oliguria or anuria resolved. Recovery from ARF was defined as reduction of the serum creatinine level to the baseline value after stoppage of peritoneal dialysis.

Statistical Analysis

The chi-square test and Fisher’s exact test were used to analyze differences between proportions. The association between baseline and the intra- and postoperative variables and mortality was assessed by logistic regression. The variables used to compute the logistic model included the postoperative albumin levels, aorta clamping, total by-pass time, presence of sepsis, and postoperative hypotension. Adjusted odds ratios and 95% confidence intervals (CI) were calculated by conditional multiple logistic regression. P values < 0.05 were considered to indicate statistical significance.

Results

Patient demographic data and the significant results in the survivor and nonsurvivor groups are summarized in Table 1. Twenty-one (32.8%) patients developed ARF after surgery. Twenty patients were oliguric and one was nonoliguric. The cause of ARF was considered to be cardiac surgery–related renal ischemia and possible acute tubular necrosis in all patients. In addition, the presence of sepsis in five patients and the use of nephrotoxic antibiotics (Vancomycin and amikacin) in two patients were the other conditions that contributed to the development of ARF. A total of 16 patients died (overall mortality 25%). The death rate in the ARF patients (14/21, 66.7%) was significantly higher than that in the non-ARF patients (2/43, 4.7%), and there was a significant positive correlation between ARF and mortality (r: 0.70, p < 0.0001). The rates of postoperative hypotension in the survivor and nonsurvivor groups were 14/48 (29.2%) and 12/16 (75%), respectively (p < 0.05). The nonsurvivors had significantly lower mean postoperative serum albumin than did the survivors (3.7 ± 0.49 gr/dL vs. 3.16 ± .0.44 gr/dL, respectively; p < 0.05) (Figure 1). There was also negative correlation between postoperative serum albumin level and mortality (r = − 0.34, p < 0.05). When results in the nonsurvivor and survivor groups were compared, the nonsurvivors had significantly longer total bypass time (179.26 ± 63.39 minute vs. 119.06 ± 48.33 minute, p < 0.05) and significantly longer aorta clamping time (106.42 ± 42.44 minute vs. 71.59 ± 31.3 minute, p < 0.05). In the patients with postoperative serum albumin level < 3.5 gr/dL, the unadjusted odds ratio (OR) for mortality was 4.3 (p = 0.01, CI 95%: 1.05–17.86). The unadjusted OR for postoperative hypotension was 5.71 (p = 0.03, CI 95%:1.49–21.88), OR for total bypass time was 1.01 (p = 0.006, CI 95%:1.005–1.032), and OR for aorta clamping time was 1.02 (p = 0.026, CI 95%: 1.002–1.045). Using multivariate analysis, postoperative serum albumin level (OR: 3.95, CI 95%: 1.19–16.16) and postoperative hypotension (OR: 1.45, CI 95%: 1.01–4.87) were independently associated with mortality. Age, sex, use of inotropic drugs, serum levels of creatinine, C-reactive protein, and the presence of sepsis were not associated with high mortality.

Table 1. Patient demographic data and the significant results in the survivor and nonsurvivor groups

	Survivors (n:48)	Nonsurvivors (n:16)
Median age (months) (range)	4 months (0–22 months)	3 months (0–22 months)
Mean body weight (kg)	5.49 ± 3.81	5.72 ± 3.34
Sex (F/M)	16/32	6/10
Total bypass time (min)	119.06 ± 48.33	179.26 ± 63.39
Aorta clamping time (min)	71.59 ± 31.3	106.42 ± 42.44
Acute renal failure (n)	4 (8.3%)	12 (75%)
Postoperative hypotension (n)	14 (29%)	12 (75%)
Low cardiac output(n)	10 (16.6%)	21 (81%)
Mean postoperative serum albumin level (gr/dL)	3.71 ± 0.49	3.16 ± 0.44
Postoperative C-reactive protein (mg/dL)	46.64 ± 19.98	49.87 ± 15.99
Presence of sepsis	4 (8.5 %)	2 (12.5%)

^*p < 0.05.

Figure 1 Postoperative serum albumin level of < 3.5 g/dL was significantly associated with high risk of mortality (OR: 4.3, CI 95%:1.05–17.86).

Discussion

Acute renal failure remains a frequent and serious complication of cardiac surgery in infants and is associated with high mortality in this patient group. Identifying the risk factors for ARF after cardiac surgery allows physicians to recognize and treat high-risk patients early, which decreases mortality. Recent studies have documented hypoalbuminemia as a predictor of mortality in adult patients with ARF.[4&5] In experimental models, renal failure itself neither suppresses albumin synthesis nor results in a reduced albumin pool;[6] however, accelerated protein breakdown is one of the main metabolic changes that occurs in ARF. Increased protein catabolism in ARF is associated with acute disease processes, uremic toxins, inflammatory mediators, metabolic acidosis, circulating proteases, and hormonal imbalance. Moshage et al. have shown that albumin is a negative acute-phase reactive protein, and that albumin synthesis is suppressed under inflammatory conditions.[7] Recent work has revealed that severity of hypoalbuminemia is significantly linked with tumor necrosis-factor-alpha level and with interleukin-6 level.[8] Investigation has also shown that patients who have undergone open-heart surgery secrete increased amounts of inflammatory mediators.[9]

In this study, we evaluated the link between postoperative hypoalbuminemia and mortality in infants who develop ARF after open-heart surgery. The mean serum albumin level in the nonsurvivor group of infants was significantly lower than that in the survivors, and hypoalbuminemia (< 3.5 g/dL) was associated with a fourfold increase in mortality.

The duration of severe renal ischemia is of paramount importance in determining the development and severity of postoperative ARF in infants who undergo cardiac surgery. Reports in the literature have identified cardiopulmonary bypass time and circulatory arrest time as risk factors for post-cardiopulmonary bypass surgery–ARF in children.[3], [10] Our results from this study confirm that these features are important determinants of renal morbidity in infants after cardiac surgery. It has also been reported that duration of arterial hypotension (more than 12 hours in particular) and/or central venous hypertension are independent risk factors for ARF in these patients.[3] We found that postoperative hypotension was more frequent in the nonsurvivor patients, but that the frequencies of central venous hypertension in the two groups were similar. Our analysis showed that age, sex, sepsis, serum creatinine level, and serum C-reactive protein level were not associated with high mortality in infants who have undergone open heart surgery.

In line with previous reports we conclude that development of ARF in pediatric patients after cardiac surgery is associated with high mortality. Low serum albumin level, long total cardiopulmonary bypass time, aorta clamping time, and severe hypotension are significant risk factors for mortality in infants after open-heart surgery. The presence of these factors may be predictive for poor prognosis.