Augmented Renal Clearance in Patients with Subarachnoid Hemorrhage

Abstract

Background

Patients with subarachnoid hemorrhage (SAH) typically exhibit hyperdynamic cardiovascular hemodynamics, which may lead to increased medication clearance. The aims of this study were to evaluate the actual creatinine clearance (CrCl_A) in an aneurysmal SAH population and the effect of the development of cerebral vasospasm (CV) along with its treatment to better understand if this population exhibits augmented renal clearance (ARC).

Methods

This was a prospective, single-center study in a neurosciences ICU at a university hospital. A total of 20 patients were consented and provided a 24-h urine sample to measure the CrCl_A. If patients experienced CV, a 24-h urine collection was repeated during vasospasm treatment. CrCl_A was measured using a modified Jaffe assay.

Results

Among the 20 patients enrolled, the mean SAH CrCl_A was 325.93 ± 135.20 ml/min 1.73 m² and this differed significantly from the SAH estimated creatinine clearance (CrCl_E) 144.93 ± 42.82 ml/min 1.73 m² (p < 0.001). Four patients developed CV; the mean CV CrCl_A was 558.43 ± 356.12 ml/min 1.73 m² and there was no significant difference when compared to those patients’ mean SAH CrCl_A (246.91 ± 84.14 ml/min 1.73 m², p = 0.16).

Conclusions

ARC was present in 100 % of the patients with recent SAH enrolled. Although ARC remained present in the patients who experienced CV, their creatinine clearance was not significantly further augmented. Further work is needed to clarify the impact of such clearances on renally excreted medications and how the development and treatment of CV further augment these findings.

Introduction

Non-traumatic subarachnoid hemorrhage (SAH) is a neurologic emergency characterized by extravasation of blood into spaces covering the central nervous system that are filled with cerebrospinal fluid. The leading cause of non-traumatic SAH is rupture of an aneurysm, which accounts for 80 % of all cases. Aneurysmal SAH is a common and devastating condition occurring in approximately 30,000 patients in the United States annually, with a mortality rate approaching 50 % [1]. Cerebral vasospasm (CV) is a common complication of SAH and is often discovered by angiographic evidence of arterial narrowing. CV may lead to delayed cerebral ischemia (DCI), defined as symptomatic vasospasm, or infarction on CT attributable to vasospasm, which occurs in approximately 21 % of SAH patients. The presence of DCI is statistically associated with more hospital complications [2, 3].

Much like other types of brain injury, the normal physiologic response to SAH can lead to fever, leukocytosis, alterations in the levels of acute phase proteins, and an acute surge of sympathetic activity [4]. This response, along with common clinical interventions for SAH (aggressive fluid resuscitation and the use of vasopressors), can result in changes in vascular tone and cardiovascular function. This may lead to alterations in glomerular filtration rate (GFR) and enhanced renal elimination of circulating solutes, often referred to as augmented renal clearance (ARC).

The most widely accepted estimate for renal function is the GFR or creatinine clearance (CrCl). ARC is defined by a creatinine clearance greater than ‘normal’ values: 120 ml/min 1.73 m² in women and 130 ml/min 1.73 m² in men [5]. Currently, data that describe the extent of ARC in the critically ill are relatively scarce and no consensus has yet been reached on an upper limit of normal filtration (though it has been preliminarily defined as a CrCl greater than 150 ml/min 1.73 m² in women and greater than 160 ml/min 1.73 m² in men) [6, 7]. ARC has been described in various critically ill populations including burn, traumatic brain injury, trauma, and sepsis [7-10]. To this point, ARC in the SAH population has only been theorized, but we believe that SAH patients are similar to other critically ill patients who are known to exhibit ARC.

The purpose of this study was to evaluate the actual creatinine clearance (CrCl_A) in an aneurysmal SAH population and the effect of the development of CV along with its treatment to better understand if this population exhibits ARC. The hypothesis was that patients with SAH have ARC and that implementation of hyperperfusion therapy further increases GFR and ARC.

Materials and Methods

Study Design, Setting, and Patient Selection

This was a single center, prospective study conducted at a tertiary-level, university-affiliated facility. The study was approved by the institutional review board and written informed consent was obtained from either the patient or their legal representative before initiation of any study procedures. The study population consisted of patients admitted to the neuroscience ICU between January 1, 2013 and April 30, 2014. Eligible participants were aged between 18 and 65, presented with or developed a non-traumatic SAH, and had a hospital length of stay (LOS) greater than 48 hours. Exclusion criteria included traumatic SAH, pregnancy, patients with pre-existing renal failure (chronic kidney disease stages 3, 4, and 5), nephrectomy, or those on nephrotoxic intravenous antibiotics at time of consent.

Management of SAH patients in our institution is consistent with published guidelines [11]. Specifically, blood pressure is rigorously controlled upon admission (typical goal is systolic blood pressure less than 140–150 mmHg). Nimodipine 60 mg by mouth or per tube every 4 h and atorvastatin 80 mg by mouth or per tube daily are initiated on hospital day 1. Digital subtraction angiography is performed as soon as is feasible to describe the location and morphology of the aneurysm. Aneurysms are secured via endovascular coiling or surgical clipping as soon as possible. After the aneurysm is secured, patients are kept euvolemic and the blood pressure is typically permitted to rise to no greater than a systolic blood pressure of 200 mmHg. Transcranial Dopplers (TCD) are performed and evaluated daily to assess for presence of CV. Sedating medications are avoided whenever possible. SAH patients typically reside in the intensive care unit for frequent, and often invasive, monitoring. When CV is suspected, euvolemia is confirmed and vasopressors are often initiated to target elevated mean arterial pressures (MAP) to typically greater than 100–110 mmHg [11]. Milrinone may also be used to increase cardiac output and/or increase cerebral perfusion [12]. Additionally, neuroendovascular procedures (infusion of spasmolytics and balloon angioplasty) may be performed.

Data Collection and Analysis

Those who met inclusion criteria and consented were enrolled in the study typically in the first 2–3 days of hospitalization. A 24-h urine sample was collected to measure the patient’s actual creatinine clearance (termed SAH CrCl_A). During the 24-h collection period, the urine was stored on ice. The total urine volume was measured and 5 ml aliquots were removed and stored at −80 °C until all patients were enrolled. If patients experienced CV, the urine collection was repeated during vasospasm treatment (hyperperfusion therapy) as described above (termed CV CrCl_A).

After enrollment each patient’s medical record was reviewed to gather demographic information. Additionally, we collected information regarding surgical interventions performed, SAH severity scores (Hunt and Hess grade and Fisher grade), pertinent laboratory values, TCD values, and information specific to SAH treatment including the type and amount of fluids, vasopressors, and inotropes administered.

Determination of Urine Creatinine Concentrations and Creatinine Clearances

The creatinine concentration of the urine samples was measured using a modified Jaffe assay (Cayman’s Creatinine (urine) Assay Kit, 2012, Item No. 500701; Caymen Chemical Company). Urine samples were thawed on ice and diluted with nanopure water to a 1:100 dilution and creatinine standards were prepared as described in the kit. Then, 15 μl of each creatinine standard dilution was added to the plate. After that, each urine sample was added to the wells in triplicate according to the plate map. Next, 150 μl of Alkaline Picrate Solution (2 ml sodium borate, 6 ml surfactant, 12 ml color reagent, 3.6 ml sodium hydroxide) was added to all wells. The samples were incubated at room temperature on a shaker for 10 minutes. The plate was placed on the plate reader (BioTek^® Synergy H1) and the absorbance was read at 500 nm. Then, 5 μl of the acid solution provided in the kit was added to all wells. The plate was incubated at room temperate on a shaker for 20 minutes. The plate was placed back on the plate reader and the absorbance was read at 500 nm. Once the assay was run, the urine creatinine concentration was calculated using the linear equation determined by the creatinine standard curve. The CrCL_A was calculated and subsequently normalized to body surface area (BSA) of 1.73 m² utilizing the equation below:

$${\text{CrCl}}_{\mathrm{A}}=[({\mathrm{U}}_{\text{CR}}\times \text{UV})∕({\mathrm{S}}_{\text{CR}}\times 1440)]\times (1.73∕\text{BSA})$$

The CrCl_A was then compared to the estimated creatinine clearance (CrCl_E) for both SAH and CV time points. The CrCl_E was calculated by utilizing the serum creatinine measured at various time points including admission, during SAH, and if the patient developed CV [7, 13]. The CrCl_E was calculated using the Cockcroft-Gault formula standardized to BSA. This formula was chosen as it was demonstrated by Hoste and colleagues to most closely correlate to CrCl_A in a critically ill population [14]. This formula is defined below: [15]

$${\text{CrCl}}_{\mathrm{E}}=\frac{(140-\text{Age})\times \text{Wt}\times 1.73}{{\mathrm{S}}_{\text{cr}}\times 72\times \text{BSA}}\times 0.85\text{if female}$$

Statistical Analysis

An a priori sample size calculation was not performed; however, we sat out to enroll 20 SAH patients based on previous studies that define ARC in a specific critically ill population [7, 10, 14]. Basic descriptive analysis including mean, standard deviation, and percentages were used to summarize demographics, baseline characteristics, and results. Additionally, paired student t-tests were performed to determine the difference between the CrCl_E and CrCl_A at various time points for the SAH patients.

Results

Demographic Data

Twenty patients were enrolled into the study and all the patients completed the urine collection study. The demographics of the study population are largely reflective of the typical SAH population with 60 % females and a mean age of 52.1 ± 10.4 years (Table 1). The admission CrCl_E was 117.5 ± 26.6 ml/min 1.73 m². All of the patients had a non-traumatic SAH and 5 of the patients had no aneurysm identified via angiogram. The other 15 patients had an aneurysm or vascular malformation in various locations. Fourteen of those were secured by a coiling procedure and one patient, who had a ruptured arterio-venous malformation (AVM), underwent an embolization procedure. The median Hunt and Hess score was 3 (IQR 2–4) and the median Fisher score was 3 (IQR 3). Four of the 20 patients enrolled (20 %) developed CV and all but one patient who developed CV received norepinephrine to maintain MAPs 100–110 mmHg. Thirty-day mortality was low (5 %).

Table 1.

Demographic data of prospective aneurysmal subarachnoid hemorrhage patients^*

Characteristic	Value (n = 20)	Variation
Age (years)	52.14	10.36
Female gender (n %)	12 (60 %)
Actual body weight (kg)	84.10	13.42
Height (cm)	168.24	8.62
Body mass index (kg/m2)	29.82	4.96
Body surface area (m2)	1.93	0.17
Admission serum creatinine (mg/dL)	0.75	0.25
Admission Cockcroft-Gault CrCl (ml/min 1.73 m2)	117.54	26.61
Day of SAH urine collection (hospital day)	2.95	1.54
Fluid balance during SAH urine collection (ml)	+442.21	+1344.99
SAH characteristics
Hunt and hess scale (median, IQR)	3.00	2–4
Fisher grade (median, IQR)	3.00	3
Vasospasm (n %)	4 (20 %)

^*All values are represented as mean and standard deviation unless otherwise stated above

Comparison of Estimated versus Actual Creatinine Clearance

The urine collection study was typically performed within the first 2–3 days after ictus. None of the patients exhibited clinical or radiographic signs of CV during SAH urine collection. The fluid balance during SAH urine collection was essentially euvolemic (Table 1), with a daily positive fluid balance of 441 ± 1344 ml. The estimated creatinine clearance appeared to increase from admission to the time of SAH urine collection, which may signify the increase in renal perfusion and clearance attendant with SAH (Fig. 1a). When evaluating actual creatinine clearance, 100 % of the patients met ARC criteria. As shown in Fig. 1b when comparing the SAH CrCl_E to SAH CrCl_A there was a statistically significant difference (144.9 ± 42.8 and 325.9 ± 135.2 ml/min 1.73 m², respectively, p < 0.001). In addition, the estimated creatinine clearance during CV under estimated the CV CrCl_A (Fig. 1c, 241.9 ± 131.2 and 558.4 ± 356.1 ml/min 1.73 m², respectively, p = 0.07). For those patients who developed CV, their SAH CrCl_A did not differ significantly from CV CrCl_A (Fig. 1d; p = 0.16).

Fig. 1.

Estimated vs actual creatinine clearance. a Bar graph comparing admission CrCl_E (mean ± standard deviation) to SAH CrCl_E (p = 0.001). b Bar graph comparing SAH CrCl_E to SAH CrCl_A (p < 0.001). c Bar graph comparing CrCl_E to CrCl_A in patients who developed CV (p = 0.075). d Bar graph comparing SAH CrCl_A to CV CrCl_A (p = 0.16)

Discussion

We believe this study is the first to describe the impact of SAH on creatinine clearance and the presence of ARC in the SAH population. In fact, 100 % of the 20 SAH patients studied met the criteria previously published for ARC by having a CrCl_A of greater than 120 for women and 130 ml/min 1.73 m² for men [5].

Due to the theoretical mechanisms of ARC, patients who develop CV may have further augmented clearance because of the initiation of hyperdynamic therapy used to attenuate the ischemic results of CV. Four (20 %) of the patients in this study developed CV, and three of the four patients received norepinephrine to maintain MAPs of 100–110 mmHg. The fourth patient actually received intravenous nicardipine infusion to reduce MAPs into the goal range. There did not appear to be a significant difference between the SAH and CV CrCl_A in this patient population. Additionally, there does not appear to be a difference in the CrCl if the patient receives treatment to achieve MAP goals or if the patient is auto-hypertensive noting no significant difference between the three patients on norepinephrine compared to the patient on nicardipine. The reason for the lack of a difference between SAH CrCl_A and the CV CrCl_A measured while receiving hyperdynamic therapy is not evident. It may be that glomerular filtration and renal autoregulation of blood flow is already maximized by the stress response initiated during SAH and further augmentation via hyperdynamic therapy has little effect. The present analysis is valuable as a pilot study, but further study on the impact of hyperdynamic therapy on renal function is necessary.

The implication of augmented creatinine clearance is potentially important from the standpoint of adequate dosing of nearly all renally cleared medications. Dose reductions to avoid drug accumulation and potential toxicity in the ICU are common and generally successfully managed by clinicians. Conversely, increasing the dose in response to ARC is seldom a topic of conversation. Alterations in the pharmacokinetic profile of medications that are not typically monitored by serum concentrations (such as meropenem, piperacillin/tazobactam, and levetiracetam) have been demonstrated in these populations as a result of ARC [16-19].

Udy and colleagues conducted a study in which they defined a strong association between elevated CrCl_A and subtherapeutic trough concentrations for various β-lactams, including piperacillin and meropenem in a cohort of critically ill patients. They found that 72 % of the patients did not meet their therapeutic pharmacokinetic-pharmacodynamic (PK-PD) target with ARC identified as a predictor of low trough concentrations confirmed by multivariate analysis [17]. This finding highlights the strong possibility that many critically ill patients receive inadequate drug exposure with standard dosing, confirming the need for aggressive dosing in this population. This suggests the need for routine application of CrCl_A measurement in patients where ARC is suspected or implementation of therapeutic drug monitoring in pharmaceuticals not typically followed this way.

Targeted pharmacokinetic studies of renally cleared medications are needed in the acute phase of illness after aneurysmal SAH and other ARC states. In addition, the time course of ARC is not well defined in states such as TBI or SAH. Further research on when ARC is most impactful and the declination of ARC back to baseline renal function is necessary.

This study has several limitations with the first being the small sample size of the SAH population. Although the sample size is small, it is similar to other studies that define ARC in various populations [7, 10, 14]. In the future, concomitant study of serial concentrations of renally eliminated medications and CrCl_A may help to better define the nature of ARC in this population and the impact on medication pharmacokinetics. It is possible that the urine collection was not executed appropriately for all patients (e.g., not stored on ice for a portion of the collection period, urine volumes inadvertently discarded instead of collected). To our knowledge, this did not occur. In fact, if urine was lost or the creatinine broke down in the urine sample, the impact would have been to reduce the measured creatinine clearance.

Conclusion

In conclusion, ARC appears to be present in patients with recent non-traumatic SAH. While serum creatinine and CrCl_E are readily available methods to estimate glomerular filtration, it appears that these estimates may not be accurate for patients who are hospitalized for SAH. The degree of ARC observed in these patients likely has an impact on the pharmacokinetics of renally cleared pharmaceuticals and this information should encourage clinicians to utilize more frequent therapeutic drug monitoring. Additionally, more appropriate dosing regimens for renally eliminated pharmaceuticals must be studied in populations who exhibit ARC.