Patient Safety

Intraoperative Hydroxyethyl Starch 70/0.5 Is Not Related to Acute Kidney Injury in Surgical Patients

Retrospective Cohort Study

Endo, Arata MD; Uchino, Shigehiko MD; Iwai, Kennichi MD; Saito, Keita MD; Sanui, Masamitsu MD; Takinami, Masanori MD; Uezono, Shoichi MD

Author Information

From the Intensive Care Unit, Department of Anesthesiology, Jikei University School of Medicine, Tokyo, Japan.

Accepted for publication July 13, 2012.

Published ahead of print November 9, 2012

Supported by departmental funding. No other funding was received.

The authors declare no conflicts of interest.

This report was previously presented, in part, at the 2010 annual meeting of the Japanese Society of Anesthesiologists.

Reprints will not be available from the authors.

Address correspondence to Shigehiko Uchino, MD, Intensive Care Unit, Department of Anesthesiology, Jikei University School of Medicine, 3-19-18, NishiShinbashi, Minato-ku, Tokyo, Japan, 105-8471. Address e-mail to [email protected].

Anesthesia & Analgesia 115(6):p 1309-1314, December 2012. | DOI: 10.1213/ANE.0b013e31826ba8d7

Free
SDC

BACKGROUND:

Although high-molecular-weight hydroxyethyl starch (HES) has been reported to cause acute kidney injury (AKI), it is not clear whether low-molecular-weight HES (6% HES 70/0.5) can be a risk for AKI or not. We hypothesized that intraoperative 6% HES 70/0.5 administration is not related to postoperative AKI.

METHODS:

A retrospective chart review was conducted to identify adult surgical patients with intraoperative blood loss of ≥1000 mL at a university hospital. AKI was defined as >50% increase in serum creatinine from the preoperative value within 7 days after the operation according to the RIFLE (Risk, Injury, Failure, Loss, or End-stage kidney disease) criteria. We compared the incidence of AKI between patients with and without intraoperative HES administration. Multivariate logistic regression analysis and propensity score matching were also conducted to elucidate the impact of HES on postoperative AKI.

RESULTS:

Among 14,332 surgical cases, 846 patients met the inclusion criteria. In patients given HES (a median dose of 1000 mL, n = 635), 12.9% developed AKI, compared with 16.6% (−3.7%, −1.7% to 9.1%) in patients without HES (n = 211). Multivariate logistic regression analysis showed that HES was not an independent risk factor for postoperative AKI (odds ratio: 0.76, 0.48–1.21). Using the propensity score, 179 pairs were matched. In patients with HES, 12.3% developed AKI, compared with 14.5% in patients without HES (−2.2%, −4.9% to 9.3%).

CONCLUSION:

In this uncontrolled retrospective chart review, intraoperative 6% HES 70/0.5 in a low dose was not related to postoperative AKI in patients with major intraoperative blood loss. Randomized controlled trials are warranted to further evaluate the safety and efficacy of low-molecular-weight HES.

Hydroxyethyl starch (HES) is often used for perioperative fluid resuscitation. However, HES has been reported to cause significant adverse effects, including acute kidney injury (AKI) and coagulopathy.^1–8 Recently, a new generation of HES with a molecular weight of 130 kDa and molar substitution rate of 0.4 (6% HES 130/0.4), has gained popularity because of its potentially improved safety profile.⁹^,¹⁰ However, its renal safety has not been established.^11–14

In Japan, low-molecular-weight HES with a molecular weight of 70 kDa and a molar substitution rate of 0.5 (6% HES 70/0.5), which is not frequently used in other countries, has been used for >30 years. Although no apparent renal-associated adverse effects have been reported in this country, its safety has hardly been studied in a scientific manner.¹⁵ Therefore, it seems important to evaluate the safety of HES 70/0.5 and we conducted a retrospective study to evaluate the association between intraoperative 6% HES 70/0.5 administration and postoperative AKI.

METHODS

IRB approval was obtained with informed patient consent waived because of the retrospective nature of the study. All patients who underwent operations and stayed at least 1 day at Jikei University Hospital between January 2006 and April 2008 were retrospectively screened. We included those who had intraoperative blood loss of ≥1000 mL in this study (intraoperative blood loss was routinely counted in all operations by nurses). Exclusion criteria were patients younger than 18 years old, patients with end-stage kidney disease and/or preoperative renal replacement therapy (RRT), cesarean delivery cases (incorrect volume of blood loss due to contamination of amniotic fluid), and patients with no preoperative serum creatinine measured within 6 weeks. All patients had creatinine measured at least once within 7 days postoperatively. We divided the study patients into 2 groups according to HES administration.

Patient characteristics were retrieved from the operation database (Japanese Association of Anesthesia database system): age, sex, body mass index, comorbidities (hypertension, diabetes, chronic obstructive pulmonary disease), surgical state (elective or emergent), ASA physical status, anesthetic method (general anesthesia or not), and hospital unit. In addition, serum creatinine values measured within 6 weeks preoperatively (most recent one was chosen if >2 creatinine values were available) and within 7 days postoperatively were collected from the hospital laboratory database. The estimated glomerular filtration rate (eGFR) was calculated with the modified Modification of Diet in Renal Disease study equation for Japan.¹⁶ Intraoperative data were also retrieved from the operation database: operation time, amount of blood loss, urinary output, crystalloids and HES (Hespander^®; Fresenius Kabi Japan Inc., Tokyo, Japan) administered, and blood transfused. All anesthetic management decisions including type and amount of fluid given were left to the discretion of the attending physicians. Only HES was available in the operating rooms as synthetic colloids.

The primary outcome of this study was the incidence of postoperative AKI. We determined AKI by the Risk, Injury, Failure, Loss, or End-stage kidney disease (RIFLE) criteria.¹⁷ Only creatinine criteria were used because of the lack of information for postoperative urine volume. The peak serum creatinine within 7 days after the operation was used for the determination of AKI. The secondary outcomes of the study were RRT requirement, intensive care unit (ICU) admission, length of hospital stay (the duration between the day of surgery and of hospital discharge), and hospital mortality, which were obtained from ICU and hospital databases.

Data are expressed as medians with 25th to 75th percentiles or percentages. The Mann-Whitney U test was used to compare continuous variables and the Fisher exact test or χ² test was used to compare nominal variables. Multivariate logistic regression analysis was performed to elucidate the impact of HES on the incidence of postoperative AKI, using all demographical variables presented in Table 1: age, male sex, body mass index, hypertension, diabetes, chronic obstructive pulmonary disease, emergent surgery, ASA physical status >II, hospital unit, preoperative eGFR <60, general anesthesia, operation time, blood loss, urinary output, crystalloid administration, HES administration, and type and amount of transfusions. Patients with missing data were excluded from the analysis. Backward stepwise elimination process was used to remove variables whose multivariate P value was >0.05. HES administration was forced to remain in the analysis.

T1-10 — Table 1:
Demographics of Study Patients With and Without HES Administration

To reduce imbalances between study groups, the propensity score for HES administration was calculated using multivariate logistic regression analysis. Variables included in the final model to calculate the propensity score were: actual operation time (the issue of scheduled versus actual duration is addressed in the Discussion section), hypertension, urine output, crystalloid administration, transfusion of concentrated red cells and platelets, autotransfusion, orthopedics, cardiac surgery, and vascular surgery. Using the method of Beattie et al.,¹⁸ individual patients with HES administration “were matched 1:1 to patients without HES administration based on similar propensity scores. A 5 to 1 computerized greedy matching technique was used for the matching process; cases were first matched to controls that had a propensity score (logit transform) that was identical in all 5 digits. Those that did not match were then matched to controls on 4 digits of the propensity score. This continued down to a 1-digit match on the propensity score for those remaining unmatched.¹⁹ Measured covariates in these matched pairs were also compared using the Mann-Whitney U test for continuous variables and the Fisher exact test or χ² test for nominal variables, respectively. All statistical analyses were performed using the StatView version 5.0 statistical software package (SAS Institute Inc., Cary, NC).

RESULTS

The clinical records of 14,332 surgical cases were reviewed. We excluded 2104 pediatric cases, 215 cases with end-stage kidney disease, 8 cases with preoperative renal replacement therapy, 10,996 cases with intraoperative blood loss of <1000 mL, 159 cesarean delivery cases, and 4 cases with no preoperative serum creatinine measured. Finally, 846 patientsmet the inclusion criteria (Fig. 1). HES was administered intraoperatively in 635 patients (75.1%), with a median amount of 1000 mL (500–1500 mL) and a mean of 1300 mL. Serum creatinine was measured immediately after their operation in 59.2%, and the next day in 98.3% of all study patients. On average, creatinine was measured 2.7 times (median 3, range 1–7) within 72 hours postoperatively.

F1-10 — Figure 1:
Study participant flow diagram. ESKD = end-stage kidney disease; RRT = renal replacement therapy.

Preoperative patient characteristics and intraoperative data are shown in Table 1. Patients in the HES group more frequently had a history of hypertension (HES: 14.5%; control: 4.3%; P < 0.0001), and had slightly but significantly more blood loss compared with the control group (1.7 L vs 1.5 L, P = 0.011). Albumin (39.1% vs 17.1%, P < 0.0001) and red blood cell transfusion (45.7% vs 30.3%, P = 0.0001) were also more frequently administered in the HES group. In contrast, slightly less crystalloid (3.0 L vs 3.6 L, P = 0.0014) and platelet transfusion (8.8% vs 13.7%, P = 0.047) was administered in the HES group.

Outcomes of study patients are shown in Table 2. The incidence of AKI and postoperative peak serum creatinine within the 7 days after the operation were not different between the 2 groups (12.9% in HES group vs 16.6% in the control group, absolute difference of −3.7%, −1.7% to 9.1%). RRT was required less frequently in the HES group compared with the control group (1.4% vs 4.3%, P = 0.023). ICU admission, length of hospital stay, and hospital mortality were similar between the 2 groups.

T2-10 — Table 2:
Outcomes of Patients With and Without HES Administration

In the multivariate logistic regression analysis (Table 3), 4 variables were found to be independent risk factors for postoperative AKI: preoperative eGFR of <60 mL/min, ASA physical status >II, age, and amount of intraoperative blood loss. Incidence of postoperative AKI was not significantly different between groups. By multivariate analysis, HES administration was also not an independent risk factor for AKI (odds ratio 0.76, 95% confidence interval 0.48–1.21, P = 0.25). Multivariate logistic regression analysis was repeated using the volume of HES administered as an independent variable and we found no relation to postoperative AKI (P = 0.35).

T3-10 — Table 3:
Multivariate Logistic Regression Analysis for Postoperative Acute Kidney Injury

In the course of propensity score matching, we successfully matched 179 patients with and without HES administration. Preoperative patient characteristics and intraoperative data are shown in Table 4. There was no significant difference between groups in preoperative patient characteristics. However, patients in the HES group had slightly, but significantly, more blood loss compared with the control group (1.66 L vs 1.47 L, P = 0.0052), and albumin was more frequently administered in the HES group (26.8% vs 17.3%, P = 0.041).

T4-10 — Table 4:
Propensity-Matched Comparison of Study Patients With and Without HES Administration

Outcomes of matched patients are shown in Table 5 and Figure 2. The incidence of AKI and postoperative peak serum creatinine within the 7 days after the operation were not different between the 2 groups. Twenty-two patients (12.3%) in the HES group and 26 patients (14.5%) in the control group developed AKI (absolute difference of −2.2%, −4.9% to 9.3%). The incidence of RIFLE failure (2.8% vs 1.1%), RRT requirement (3.4% vs 2.8%), ICU admission (46.4% vs 44.7%), and hospital mortality (3.4% vs 5.0%) were also similar between the 2 groups. However, length of hospital stay in the HES group was significantly longer than in the control group (31 days vs 23 days, P = 0.0046).

T5-10 — Table 5:
Propensity-Matched Comparison of Study Outcomes With and Without HES Administration

F2-10 — Figure 2:
The incidence of postoperative acute kidney injury in matched patients with major intraoperative blood loss with or without hydroxyethyl starch (HES) administration.

DISCUSSION

We have retrospectively assessed the impact of 6% HES 70/0.5 on renal function after major intraoperative blood loss. Our results showed that intraoperative 6% HES 70/0.5 administration was not associated with postoperative AKI. To minimize the selection biases, we conducted a propensity score matching, which supported these results.

HES solution is among the most popular nonprotein intravascular volume expanders.²⁰^,²¹ These solutions differ in their percentage concentration, molecular weight, proportion of hydroxyethyl units (molar substitution rate), and the ratio of substitution at the C2 and C6 positions on the glucose ring (C2/C6 ratio).²² Several prospective studies have suggested that the newer HES might be less nephrotoxic than other colloids.^23–25 However, although a significant number of clinical studies have been conducted, the renal safety of the newer HES has not been established. Some suggest that the newer HES is safe,¹²^,^23–25 whereas others do not.⁸^,¹¹^,¹³^,²⁶

The only published study evaluating the relationship between 6% HES 70/0.5 administration and AKI was conducted by Suzuki et al.¹⁵ They retrospectively studied 31 patients who had >5000 mL of blood loss during surgery and survived for >1 month. Thirteen patients (41.9%) developed AKI. The amount of HES infused in the AKI patients (53 mL/kg) did not differ from that in the non-AKI patients (55 mL/kg), and there was no relationship between the amount of HES infused and serum creatinine changes.

Our study has several limitations. First, we do not have information for postoperative administration of HES or any other nephrotoxic drugs (iodinated contrast media, antibiotics, vasopressors, diuretics, or angiotensin-converting enzyme inhibitors). We also could not collect information for other potentially confounding factors (e.g., perioperative hemodynamic status, postoperative infectious complications). We could not retrieve the postoperative urine output data and we used only creatinine criteria to determine AKI by the RIFLE criteria. The missing information could have changed our findings. Second, this is a single-center, retrospective, observational study, which includes significant limitations in the external validity and the generalizability. For example, anesthesiologists in our hospital often prescribed HES for intraoperative fluid resuscitation (approximately 75% of the study patients received HES). To eliminate selection bias, we conducted propensity score matching, and confirmed that the incidence of AKI was similar between the 2 groups. Third, coagulopathy, another concern associated with HES administration,⁵ was not evaluated in this study. In fact, the HES group had more blood loss and longer hospital stay in the propensity-matched cohort. Our study design cannot differentiate whether the higher blood loss was related to HES administration or indicates baseline imbalances. Coagulopathy induced by HES administration could have potentially contributed to intraoperative hemorrhage. However, the difference in the amount of blood loss was clinically small (1.66 L vs 1.47 L) and the percentage of patients transfused was not different between the 2 groups. The reason for longer hospital stay in the HES group is uncertain, considering the similar patient characteristics and hospital mortality. Future studies for 6% HES 70/0.5 will need to evaluate its safety including blood loss and length of hospital stay. Fourth, our study cohort consisted of a mixed surgical population with mostly ASA physical status I or II, thus was at low risk to develop AKI. Our sample size (especially in the matched comparisons) is therefore too small to detect a statistical difference in AKI incidence. Fifth, the median amount of HES administered intraoperatively was relatively small (1000 mL), which might not have been enough to cause postoperative AKI. Although we attempted to conduct a subgroup analysis for patients given a larger amount of HES, such patients also had a very large amount of intraoperative blood loss and we could not properly match these patients with those without HES administration. In the additional multivariate analysis using the volume of HES administered as an independent variable, HES was not found to be related to AKI. The study by Suzuki et al.¹⁵ also did not find an increased incidence of AKI in patients who received >50 mL/kg HES intraoperatively, but only 31 patients were studied. Finally, we used actual operation time instead of scheduled operation time to conduct propensity analysis, which causes bias in logistic regression results.²⁷ We could not avoid this potential bias because of lack of information for scheduled operation time.

In conclusion, the findings of our uncontrolled retrospective chart review suggest that intraoperative administration of a low dose of 6% HES 70/0.5 is not related to postoperative AKI in patients with major intraoperative blood loss. Randomized controlled trials are warranted to further evaluate the safety and efficacy of 6% HES 70/0.5.

DISCLOSURES

Name: Arata Endo, MD.

Contribution: This author helped conduct the study and write the manuscript.

Attestation: Arata Endo has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.

Name: Shigehiko Uchino, MD.

Contribution: This author helped design the study, analyze the data, and write the manuscript.

Attestation: Shigehiko Uchino has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.

Name: Kennichi Iwai, MD.