Evaluation of Renal Function with Administration of 6% Hydroxyethyl Starch and 4% Gelatin in Major Abdominal Surgeries: A Pilot Study.

BACKGROUND: Synthetic colloids, both starches and gelatins, are commonly used as intravascular fluid replacements on account of increased vascular persistence. The safety on renal outcomes during perioperative use is poorly understood. AIMS: We evaluated renal outcomes of hydroxyethyl starch 6% (HES) and gelatins 4% (G) in patients undergoing elective abdominal surgery. The primary outcome was serum creatinine measurements at baseline, 12 h, 36 h, and 1 week postoperatively (T0, T12, T36, and D7). The secondary outcomes were measurements of prothrombin time (PT), international normalized ratio (INR), fibrinogen, and activated partial thromboplastin time (aPTT) at baseline, 12 h, and 36 h postoperatively. SETTING AND
DESIGN: A prospective randomized study was conducted at a tertiary care institute.
MATERIALS AND METHODS: Seven-five adult patients received either HES (Group H) or gelatin (Group G) at 20-ml/kg body weight or only crystalloids (Group C) during surgery. Statistical tests used were one-way ANOVA, Student's t-test, Pearson correlation method, and Chi-square test.
RESULTS: Serum creatinine assessed at T0, T12, T36, and D7 was comparable between the three groups. PT/INR and aPTT showed no significant increase in values of T12 and T36 in comparison to T0. Fibrinogen level was significantly higher in Group C at T12 and T36. Intraoperative vasopressor use, need for product transfusion, length of intensive care unit stay, and return of bowel function were similar between the three groups.
CONCLUSIONS: Intraoperative use of HES (130/0.4) or gelatin (4%) at 20-ml/kg body weight was not associated with renal dysfunction or altered PT and aPTT in adult patients undergoing elective abdominal major surgeries.

INTRODUCTION

Synthetic colloids allow fluids to remain in the intravascular compartment and provide better hemodynamic stability.[1] Safety of colloids in clinical practice has been a matter of constant debate even though there are several reports establishing their safe use even in the context of sepsis and otherwise.[23456] Although the beneficial effect on the maintenance of intravascular oncotic pressure remains undisputed, the side effects have limited its use. Newer generations of synthetic colloids have been introduced to negate the undesirable effects. The European Society of Critical Care Medicine recommends avoidance of starches in patients with renal dysunction or sepsis resulting in a decline in its use even for perioperative fluid resuscitation. Hydroxyethyl starches (HES) are associated with increased incidence of renal injury among septic patients,[7] while gelatins are known to cause coagulation abnormalities as well as anaphylaxis due to its bovine component.[89] Fluid restriction has gained increasing acceptance in recent years for improved postoperative outcomes.[10] We believed that the administration of colloids would help in a restrictive fluid regimen and proposed to evaluate the safety of these colloids among nonseptic stable patients undergoing surgery.

Aim of the study

The primary outcome was serum creatinine measured at baseline and at 12 h, 36 h, and 1 week postoperatively (T0, T12, T36, and D7). The secondary outcomes were measurements of prothrombin time (PT), international normalized ratio (INR), fibrinogen, and activated partial thromboplastin time (aPTT) at baseline, 12 h, and 36 h postoperatively (T0, T12, and T36).

MATERIALS AND METHODS

Following approval from the Institutional Ethics Committee, 75 patients scheduled to undergo elective abdominal surgery were included in this study [Figure 1]. Patients were randomly allocated into one of the three groups by computer-generated random sequence of numbers and allocated as Group H, Group G, and Group C. The numbers were allotted from opaque sealed envelopes.

Figure 1

Consort flowchart

Patients belonging to the American Society of Anesthesiologists (ASA) physical status Classes I and II, aged between 18 and 75 years undergoing abdominal surgery, were included. Patients with renal dysfunction defined as serum creatinine ≥1.3 mg/dL disorders of coagulation and history of allergic disorders necessitating treatment were excluded.

Group H received 20-ml/kg 6% HES 130/0.4 (maximum of up to 1 L). Group G received 20-ml/kg 4% gelatin in balanced salt solution (maximum of up to 1 L). Group C received crystalloids only.

Induction and maintenance of anesthesia were according to standard protocols. An epidural catheter was placed at the discretion of the anesthesiologist, prior to induction. The fasting requirements were replaced according to each case to about 8 ml/kg of balanced salt solution after induction prior to the commencement of surgery. The maintenance fluid requirement was given at a rate of 4 ml/kg/h of Ringer's lactate (RL; Fresenius Kabi)/acetated balanced salt solution (Kabilyte®; Fresenius Kabi). After replacement of 1-L crystalloid solution, patients in Group H were administered 20 ml/kg of 6% HES 130/0.4 (Volulyte®; Fresenius Kabi) or 4% gelatin (Gelaspan®; B Braun) in Group G to a maximum of 1 L in each patient, and in Group C, RL or acetated balanced salt solution was administered.

A fall in blood pressure, defined as fall in mean arterial pressure <20% from baseline, was managed by fluids and noradrenaline infusion according to standard protocols. Blood loss of >10% of estimated blood volume or hemoglobin (Hb) <8.0 mg/dL intraoperatively was considered an indication for transfusion of packed red cells.

Renal dysfunction was defined as per the Risk, Injury, Failure, Loss of function, and End-stage renal disease (RIFLE) criteria for acute kidney injury (AKI). A rise in creatinine levels of more than 1.5 times the baseline value was considered evidence for renal risk. Creatinine values were measured at baseline, 12 h, 36 h, and 1 week postoperatively, and percentage differences between the time points were computed for assessing inclusivity in the RIFLE criteria for AKI.

The coagulation abnormalities were measured as a change in PT/INR and aPTT from the baseline at 12 h and 36 h after surgery. The total volume of intravenous fluids used was calculated. The need for any product transfusion in these study groups was also assessed.

Statistical analysis was done using IBM SPSS Statistics 20 Windows (SPSS Inc., Chicago, IL, USA). For all the continuous variables, the results were given in mean ± standard deviation and for categorical variables as percentage. To compare the mean difference of numerical variables between the three groups, one-way ANOVA was used. Bonferroni multiple comparison test was used to compare intergroup differences in variables showing significant P value with ANOVA. To compare the mean difference of numerical variables between the two groups, Student's t-test was applied. To study the relationship between the two variables, Pearson correlation coefficient was computed. To test the statistical significance of the association of categorical variables, Chi-square test was used. Probability value (P value) < 0.05 was considered statistically significant.

RESULTS

The demographic variables among the three groups were comparable except for age which was higher in Group C. The types of surgeries between the three groups were comparable [Table 1].

Table 1

Demographic variables

Variables	Category	Groups			P
		C, n (%)	G, n (%)	H, n (%)
Age years	NA	60.24±11.43	50.96±14.45	52.4±12.64	0.028
Gender	Female	8 (32)	9 (36)	10 (40)	NA
	Male	17 (68)	16 (64)	15 (60)
ASA physical status	I	2 (8)	7 (28)	7 (28)	0.137
	II	23 (92)	18 (72)	18 (72)
Type of anesthesia	GA	7 (28)	5 (20)	4 (16)	0.573
	GA+EA	18 (72)	20 (80)	21 (84)
Type of surgery	LAP	10 (40)	8 (32)	5 (20)	0.658
	OPEN	14 (56)	16 (64)	19 (76)
	LAP + OPEN	1 (4)	1 (4)	1 (4)

GA+EA=General anesthesia + epidural anesthesia, LAP=Laparoscopic surgery, OPEN=Open surgical procedure, LAP+OPEN=Laparoscopic procedure converted to open, NA=Not available, ASA=American Society of Anesthesiologists

The mean creatinine measured at the defined time points (T0, T12, T36, and D7) was compared between the groups and was comparable [Figure 2]. The percentage changes of creatinine between the three groups were also comparable [Table 2].

Figure 2

Mean creatinine at defined time points

Table 2

Percentage differences in creatinine values

Groups	% Change (Mean±SD)
	T0 versus T12	T0 versus T36	T0 versus D7
C	−2±22.52	−10.9±19.49	−14.75±19.7
G	1.88±22.85	−8.66±15.94	−14.27±29.3
H	−4.03±17.19	−9.93±19.21	−16.36±17.56
P	0.603	0.910	0.951

SD=Standard deviation

Coagulation parameters such as PT, INR, aPTT, and fibrinogen were compared at the T0, T12, and T36 time points. The mean values as well as the percentage variations of PT/INR and aPTT between the three groups were comparable [Tables 3 and 4].

Table 3

Coagulation parameters

Groups	Mean±SD
	T0	T12	T36
PT (s)
C	14.47±1.37	16.84±1.49	17±2.05
G	14.42±1.25	17.38±2.57	17.29±2.69
H	14.64±1.33	17.2±2.29	17.08±2.33
P	0.828	0.667	0.907
INR
C	0.99±0.11	1.19±0.12	1.19±0.17
G	0.96±0.15	1.23±0.22	1.23±0.24
H	1±0.11	1.22±0.21	1.21±0.20
P	0.475	0.642	0.866
aPTT (s)
C	29.68±4.02	32.29±4.91	33.1±5.1
G	32.36±4.74	33.29±4.80	35.02±5.30
H	31.48±4.44	33.89±6.51	34.2±6.23
P	0.098	0.579	0.472
Fibrinogen (mg/dL)
C	313.92±74.55	410.8±86.16	532.28±126.58
G	279.72±71.55	353.64±87.77	463.56±122.34
H	335.24±91.84	371.76±75.41	444.60±109.06
P	0.052	0.052	0.029
Post hoc analysis	C versus H, P=0.035
	C versus G, P=0.137
	G versus H, P=1.000

T0=Baseline, T12=12 h after surgery, T36=36 h after surgery, D7=7th POD, POD=Postoperative days, SD=Standard deviation, PT=Prothrombin time, aPTT=Activated partial thromboplastin time

Table 4

Percentage changes in prothrombin time, international normalized ratio, activated partial thromboplastin time, and fibrinogen

Groups	% Change (Mean±SD)
	T0 versus T12	T0 versus T36
PT
C	16.84±10.31	17.88±13.91
G	20.46±12.33	19.92±15.70
H	17.44±10.27	16.92±13.88
P	0.464	0.759
INR
C	20.45±12.89	21.16±17.66
G	31.70±30.76	32.04±41.99
H	21.85±13.07	20.79±16.88
P	0.118	0.280
aPTT
C	8.995±9.79	12.02±13.12
G	4.128±16.25	9.38±16.79
H	7.654±12.9	8.89±13.40
P	0.411	0.717
Fibrinogen
C	34.59±27.9	77.74±59.87
G	31.14±36.17	76.85±65.14
H	14.83±20.91	39.08±42.46
P	0.042	0.027
Post hoc analysis	C versus G, P=0.947	C versus G, P=1.000
	C versus H, P=0.029	C versus H, P=0.016
	G versus H, P=0.313	G versus H, P=0.159

T0=Baseline, T12=12 h postoperatively, T36=36 h postoperatively, D7=1 week postoperatively, INR=International normalized ratio, aPTT=Activated partial thromboplastin time, SD=Standard deviation, PT=Prothrombin time

Compared to baseline, the fibrinogen levels at 12 h and 36 h were significantly higher in all the groups. The intergroup analysis in percentage changes showed that values for Group C were significant at 12 and 36 h postoperatively [Table 4]. There was a significantly higher difference between Group C and Group H (P = 0.029, 0.016) and not so when compared to Group G at the defined time points (P = 0.947, 1.000). Group G and H showed no difference in the values of percentage difference at 12 and 36 h.

Six patients in Group C, five in Group G, and seven in Group H needed perioperative transfusion of packed cells, and this was not significant between the three groups. Only one patient in Group C needed transfusion of plasma while none in Groups G and H needed transfusion of plasma.

The mean volume of intravenous fluid used showed no significant difference in between the groups [Figure 3].

Figure 3

Mean fluid administered during surgery

The length of the intensive care unit stay in Group C was 63.52 ± 53.05 h versus 40.04 ± 20.79 h and 58.08 ± 28.07 h in Groups G and H, respectively, and this was comparable between the three groups.

DISCUSSION

The debates on the safety of colloids have been overshadowing their clinical benefits in perioperative use for several years.[11] We proposed to evaluate the safety of administration of limited volumes of synthetic colloids in ASA physical status Classes I and II patients in a perioperative setting in elective abdominal surgery.

Two of the reported adverse effects with the use of colloids are renal dysfunction and coagulation abnormalities.[1213] Coagulation abnormalities reported involve reductions in Factor VIII and von Willebrand factor and platelet dysfunction due to the binding of the HES molecule to the platelets. Most of the adverse effects reported involve the larger molecular weight starches and were not seen when the molecular weight was reduced to 130 and the molar substitution to 0.4.[12] However, among patients admitted to the intensive care, the use of even 130/0.4 HES was shown to be associated with a poorer renal outcome in comparison to normal saline.[1314] The change to the usage of gelatins instead of HES did not decrease the incidence of acute renal failure and morbidity in a retrospective cohort of surgical intensive care patients.[15]

The occurrence of AKI is assessed by the RIFLE criteria put forth by the Kidney Disease Improving Global Outcomes (KDIGO) clinical practice guideline for AKI.[16] We evaluated the earliest evidence by this definition, risk, as the clinical criteria for identifying patients with renal dysfunction following colloid administration at a volume of 20 ml/kg. We excluded confounding factors contributing to postoperative renal dysfunction by excluding patients with baseline creatinine more than 1.3 mg/dL. As the allocation of patients to the different fluid strategies was randomized, we believed that confounding causes for renal dysfunction would be equally distributed between the two groups. Intraoperatively, we had ensured adequate blood pressure by a standardized protocol that incorporated a vasopressor at blood pressure values <20% from the baseline and transfusions when the Hb fell below 8 g/dL.

We did not find a significant change in creatinine as compared with the baseline values by either absolute values or percentage difference from the baseline in our study group. A follow-up period of 1 week did not show deterioration in creatinine subsequently. The RIFLE criteria warrant a substantial change in creatinine to document a risk, but the creatinine values were within the normal range in all the three groups in our study. We believe that although our numbers were small, this may offer some insights in the safety of the use of limited volumes of low-molecular-weight hetastarch and gelatins in a perioperative setting in low-risk surgical patients. This appears to match the findings in most recent literature.[171819]

Another deterrent to the routine perioperative use of colloids is the occurrence of coagulopathy. We measured coagulation by the standard tests, PT, aPTT, and fibrinogen at baseline, 12 h, and 36 h postoperatively. We felt that this could be appropriate time points for documentation of drug- or fluid-induced coagulation dysfunction. There were no changes in PT or INR in any of the patients in our study. Studies evaluating coagulation profiles with 6% HES 130/0.4 and 4% gelatins in the perioperative setting have incorporated findings from thromboelastography.[20] Sawhney et al.[20] compared standard and dynamic tests of coagulation (thromboelastogram) following the administration of HES, gelatin, and crystalloids in trauma patients and showed derangements in the thromboelastogram with administration of both HES and gelatins. However, Peng et al.[21] did not find alterations in the thromboelastogram following colloid preloading with HES versus albumin in pediatric intracranial surgery. The results from Sawhney's study could be related to alterations in coagulation due to the underlying trauma, and we could not see any alteration in standard tests at the doses of colloids used in our study. Due to financial implications in performing additional tests, we did not perform dynamic tests for our patients, but we could not find clinically abnormal coagulation among our patients.

Fibrinogen showed a significant increase postoperatively in all the three groups but was significantly higher in the crystalloid group (P = 0.042, 0.027) at 12 h and 36 h postoperatively. Fibrinogen is an acute-phase reactant and would rise in the postoperative period; however, a significant elevation only in the crystalloid group cannot be accounted for. A possible explanation could be that although the surgical mix was similar across the groups, more complex and intense cases were distributed in Group C resulting in higher postoperative inflammatory responses in this group.

The use of red blood cells and fresh frozen plasma showed no significant differences between the three study groups in this study. An infusion of 6% HES 130/0.4 has been associated with lesser need for allogeneic transfusion comparison to HES 200/0.5;[22] this could be related the platelet phagocytosis with the larger molecule. Expansion of the intravascular volume with the colloid may have reduced the need for transfusions, but transfusions was limited to one-fourth of the study patients and may have been too small to document any benefit.

A study looking at the overall fluid balance, perioperative blood loss, volume of red blood cells, and fresh frozen plasma administered, as well as the number of patients who received transfusions, showed significantly lower number in children who received HES and underwent cardiac surgery.[23] This retrospective analysis over 8 years included both 130/0.4 and 200/0.5 HES and appeared to suggest that HES was as efficacious as albumin in children undergoing cardiac surgery. Surprisingly, we did not find any conservation in the use of fluids in the gelatin or HES group in comparison to crystalloids in our study. Administration of either 1L of succinylated gelatin or HES produced similar volume expansions as measured by hematocrit in healthy surgical patients.[24] There is limited literature available on the perioperative use of gelatins in surgery and in some countries has been used as a volume expander for fluid conservation and also in major surgeries including renal and liver transplant. Anaphylaxis though rare during administration can be catastrophic, and there are reports during administration as a plasma expander during surgery.[25] We did not encounter allergy or anaphylaxis during administration in our patients.

A recent review on goal-directed therapy in abdominal surgery compared the use of two molecular weights of starches in pancreatic surgery.[17] While the 200/0.5 starch had poorer renal outcomes, both starches reduced the use of crystalloids. We could not document such a difference perhaps because there was a trend toward early use of vasopressors at our unit which could have offset the need for additional volumes.

Our study did have a few limitations. As there was limited literature in the past that demonstrated renal outcomes in elective surgeries with the newer starches or gelatins, an adequate sample size could not be obtained and we conducted this as a pilot study. This study targets outcomes regarding renal dysfunction with the use of different fluids. Hence, we could only conclude that it was safe to use both the starches and gelatin at the volume used in the surgical patients. We did not include more sensitive tests of renal dysfunction postoperatively as markers of renal dysfunction. The use of neutrophil gelatinase-associated lipocalin may have provided more definitive results of the safety of these colloids. We had employed only the standard or static tests of coagulation as an assessment of coagulation disturbances. We may have had a superior validation with the use of dynamic tests of coagulation.

CONCLUSIONS

The intraoperative use of 20 ml/kg of HES 130/0.4 or 4% gelatin was not associated with postoperative renal dysfunction or altered coagulation in low-risk adult patients undergoing elective major abdominal major surgeries.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.