ABSTRACT
Colistin therapy is used as the last line of defense against life-threatening Gram-negative infections. Nephrotoxicity is the major dose-limiting side effect that impedes optimal dosing of patients. This study aims to examine the nephroprotective effect of the plasma volume expander gelofusine against colistin-induced nephrotoxicity. Renal protection was assessed in mice that were subcutaneously injected with colistin sulfate (14 mg/kg of body weight × 6 doses every 2 h; accumulated dose, 84 mg/kg) and simultaneously injected in the intraperitoneal region with gelofusine (75, 150, 300, or 600 mg/kg × 6). At 2 and 20 h after the last colistin dose, mice were euthanized, and the severity of renal alteration was examined histologically. Histological findings in mice revealed that colistin-induced nephrotoxicity was ameliorated by gelofusine in a dose-dependent manner, whereas significant histological abnormalities were detected in the kidneys of mice in the colistin-only group. The impact of coadministered gelofusine on colistin pharmacokinetics was investigated in rats. Rats were administered a single intravenous dose of gelofusine at 400 mg/kg 15 min prior to the intravenous administration of colistin (1 mg/kg). Gelofusine codosing did not alter the pharmacokinetics of colistin in rats; however, gelofusine did significantly lower the accumulation of colistin in the kidney tissue of mice. This is the first study demonstrating the protective effect of gelofusine against colistin-induced nephrotoxicity. These findings highlight the clinical potential of gelofusine as a safe adjunct for ameliorating the nephrotoxicity and increasing the therapeutic index of polymyxins.
INTRODUCTION
During the last 2 decades, there has been a significant increase in the incidence of infections caused by multidrug-resistant Gram-negative bacteria. This dire situation is exacerbated by the scarcity of novel antibiotics in the developmental pipeline. These compounding factors have led to the rebirth of polymyxin B and polymyxin E (also known as colistin) as a last line of defense against infections caused by Gram-negative “superbugs” (1). Polymyxins were launched in the 1950s and were not subjected to modern drug development procedures (1). The clinical usefulness of polymyxins is hampered by a very narrow therapeutic window that is imposed by dose-limiting nephrotoxicity, which occurs in up to 60% of patients (2, 3). This highlights the urgency to develop novel therapeutic approaches that incorporate nephroprotective agents to ameliorate this unwanted side effect. Such a kidney-sparing effect would serve to widen the therapeutic window, thereby allowing therapy with higher daily doses of polymyxins to achieve greater antibacterial efficacy and minimize resistance.
The precise mechanisms of the renal uptake and toxicity of polymyxins remain unknown. However, recent studies from our laboratory have demonstrated that polymyxin-induced nephrotoxicity is coincident with apoptosis in kidney tubular cells, which appears to be at least partially mediated by caspase activation, the generation of reactive oxygen species, and mitochondrial dysfunction (4–6). The renal handling mechanisms of polymyxins almost certainly contribute to their propensity for causing tubular cell damage (7, 8). For both colistin and polymyxin B, only a very small fraction of the drug is excreted after filtration through the glomeruli into the tubular urine (9). The majority of the filtered load undergoes extensive tubular reabsorption; indeed, the extent of polymyxin reabsorption approximates or exceeds the reabsorption of water, implicating the involvement of a carrier-mediated process (7, 8). The result is an extensive accumulation of polymyxin in tubular cells, which undoubtedly contributes significantly to polymyxin-induced renal tubular damage (6, 9). Available evidence points to the involvement of megalin and PEPT2 carrier systems in the renal reabsorption of polymyxins; these systems are known to scavenge other peptides and proteins from tubular urine (10–14).
Gelofusine is a negatively charged plasma expander composed of 4% (wt/vol) succinylated bovine gelatin polypeptides, with a mean molecular mass of 30,000 Da. Clinically, the dosage and infusion rate (usually 500 ml/30 min) for gelofusine are adjusted depending on the amount of blood loss and restoration of a stable hemodynamic situation. Notably, gelofusine has been shown to decrease the renal tubular reabsorption of proteins and peptides (15, 16). We hypothesized that gelofusine can decrease the renal reabsorption and accumulation of colistin, a lipopeptide antibiotic, in kidney tubular cells and thereby attenuate nephrotoxicity. In the present study, we examined the nephroprotective effect of gelofusine in a mouse model of colistin-induced nephrotoxicity. The effects of coadministration of gelofusine on the pharmacokinetics of colistin in rats were also examined. The findings highlight the clinical potential of gelofusine as a nephroprotective agent to ameliorate polymyxin-induced nephrotoxicity.
RESULTS
Effect of gelofusine on colistin-induced nephrotoxicity in a mouse model.The kidneys of mice in the (i) saline control (0.9% saline) and (ii) gelofusine3600-only control (cumulative succinylated bovine gelatin [SBG] dose, 3,600 mg/kg of body weight) were deemed healthy (Table 1). In contrast, the kidneys of mice treated with colistin alone and colistin plus saline (cumulative dose of colistin sulfate, 84 mg/kg) displayed extensive histological damage seen as marked tubular degeneration, necrosis, tubular dilation, cast formation, and infiltration of inflammatory cells (Fig. 1 and Table 1). Kidney damage was absent in the mice of the colistin plus gelofusine3600 group (cumulative dose of SBG, 3,600 mg/kg) (Fig. 1). The nephroprotective effect of gelofusine in the lower gelofusine codosage groups (i.e., gelofusine1800, gelofusine900, and gelofusine450, corresponding to cumulative SBG doses of 1,800, 900, and 450 mg/kg, respectively) was less pronounced (Table 1). All doses of gelofusine were well tolerated, and the animals did not display any adverse reactions.
Histological results for individual kidneys collected from the colistin-induced nephrotoxicity mouse model
Representative histological images of hematoxylin & eosin-stained kidney sections from mice collected 20 h following treatment with (A) colistin (cumulative dose, 84 mg/kg) coadministered with gelofusine (cumulative SBG dose, 3,600 mg/kg) showing normal renal cortex and glomeruli, no significant kidney damage (overall score 0, SQS 0); (B) colistin plus saline showing kidney damage, such as mild acute tubular damage dilation, prominent nuclei, and a few pale tubular casts (SQS 1, overall score 6); (C) colistin plus saline showing moderate to severe acute tubular damage with necrosis of tubular epithelial cells and numerous tubular casts (acute tubular necrosis) (SQS 2, overall score 24). The arrows in panels B and C show severely damaged and dilated tubules, degeneration, necrosis of tubular epithelial cells, and tubular casts.
Furthermore, we quantified the concentration of colistin in kidney tissue from mice treated with (i) colistin plus saline (cumulative colistin dose, 84 mg/kg) and (ii) colistin plus gelofusine3600 (cumulative colistin dose, 84 mg/kg plus cumulative SBG dose 3,600 mg/kg). The concentration of colistin in kidney tissue was significantly lower (P < 0.01) in the colistin plus gelofusine3600 treatment group (colistin, 48.6 ± 4.2 μg/g of tissue) than in the kidney tissue of mice in the colistin plus saline (colistin, 132 ± 38.7 μg/g of tissue) treatment group.
Effect of gelofusine on colistin pharmacokinetics in rats.The mean plasma concentration-time profiles and pharmacokinetic parameters in rats following intravenous administration of colistin (1 mg/kg) and colistin (1 mg/kg) plus gelofusine (SBG dose, 400 mg/kg) are presented in Fig. 2 and Table 2. The mean plasma concentration-time profiles and urinary recovery rates were similar between the two groups, and there were no significant differences in pharmacokinetic parameters between groups. The urinary recovery was very low, with <3% of the administered dose excreted in the urine as unchanged colistin in both groups. Consequently, similar renal clearances of colistin were obtained among these groups (Table 2). Likewise, the plasma protein binding of colistin in the presence (51.7% ± 4.0%) and absence (55.9% ± 2.4%) of gelofusine was comparable.
The mean plasma colistin concentration-versus-time profile in rats after a single intravenous dose of 1 mg/kg colistin coadministered with saline (●) or gelofusine (SBG dose, 400 mg/kg) (■) 15 min prior to colistin dosing. Data points are the mean ± standard deviation (SD) (n = 4).
Pharmacokinetic parameters for colistin following a single intravenous dose of 1 mg/kg colistin administered 15 min after either saline or gelofusinea
DISCUSSION
Gram-negative “superbugs” are a significant global medical challenge. Polymyxins are increasingly used as a last-line therapy; however, possibly due to suboptimal use, resistance to polymyxins is being increasingly reported (17). Unfortunately, simply administering higher doses of polymyxins is not possible due to a high incidence of nephrotoxicity. Fundamentally, the attenuation of polymyxin-induced nephrotoxicity may widen the therapeutic window to allow more effective killing of bacteria and reduce the emergence of resistance. The present study has identified a potential new therapeutic option, the plasma expander gelofusine, as an adjunct to attenuate dose-limiting nephrotoxicity.
The nephroprotective effect of gelofusine was examined in a mouse nephrotoxicity model using a fixed colistin dose (cumulative dose, 84 mg/kg) and four different gelofusine dosage regimens (cumulative SBG doses, 450, 900, 1,800, 3,600 mg/kg). Notably, based on animal scaling considerations, the doses we have employed in the animal studies are within the range clinically used in humans (18, 19). The kidneys of all mice in the colistin-only group and in the colistin and saline group had marked histological transformation, whereas in the colistin plus gelofusine3600 group, no mice exhibited any histological signs of kidney damage. It should be noted that the scoring system provides an overall score and semiquantitative score (SQS) that derive from the product of the severity of lesions and the percentages of kidney slices affected (20). Notably, the nephroprotective action of gelofusine was dose dependent, as reflected by the decreased level of protection afforded by the coadministration of lower doses of gelofusine. These histological observations and the significant reduction in total colistin concentration in kidneys of the colistin plus gelofusine3600 group suggest that colistin and gelofusine compete for uptake by the proximal tubular cells.
The pharmacokinetics of colistin in rats following intravenous administration of a single dose (1.0 mg/kg) were comparable to those in rats coadministered colistin (1.0 mg/kg) plus gelofusine (SBG dose, 400 mg/kg). Importantly, the coadministration of gelofusine produced no significant alteration in the plasma protein binding, urinary recovery, renal clearance, or plasma profile of colistin. Pharmacokinetic studies were conducted in rats because of the ability to collect multiple blood samples of adequate volume in a longitudinal manner from each animal. Colistin was administered intravenously, as this route provides the most unequivocal description of the pharmacokinetics, because the extent of bioavailability is not a potentially confounding issue. The intravenous dose used in rats was lower than the subcutaneous dose administered to mice, even after considering “dose equivalence” based upon allometric scaling across the species. The dose in mice was designed to lead to nephrotoxicity, and it was otherwise well tolerated by the animals. However, rats are substantially more sensitive to colistin than are mice. In rats, pharmacokinetic studies with intravenous colistin have been conducted with doses of 1 mg/kg or less (7, 21, 22). There is no evidence for the pharmacokinetics of colistin changing with repeat dosing, and linear pharmacokinetics of colistin in rodents has been demonstrated (21, 23).
We have recently garnered data that show that the renal toxicity of polymyxins is associated with their extensive accumulation in cultured renal proximal tubular cells, which results in apoptotic cell death (5, 6). Following filtration through glomeruli, colistin undergoes extensive renal tubular reabsorption by carrier-mediated pathways, and less than 3% of the administered dose was recovered in the urine, consistent with earlier reports (9). This highly efficient reabsorption mechanism plays a critical role in the extensive accumulation of polymyxins within the kidneys (6). As a result, even though the kidneys are exposed to a very large amount of this nephrotoxic compound via glomerular filtrate, the renal clearance makes a very small contribution to the total body clearance of colistin. As per our pharmacokinetic findings, it is then not surprising that the total body clearance of colistin would not be significantly impacted if gelofusine was inhibiting the reabsorption of colistin by the kidney.
Gelofusine has been reported to cause proteinuria of albumin and β2-microglobulin by competitively inhibiting tubular reabsorption (16, 24, 25). Moreover, gelofusine has been demonstrated to efficiently lower the renal reabsorption of nephrotoxic radiolabeled integrin and somatostatin peptides used for imaging of tumors in animals and humans (15, 26–28). Although the precise mechanism for this inhibition of renal reabsorption is not understood, it is purported to involve the megalin receptor, for a few reasons. First, megalin has been shown to be essential for the renal uptake of many filtered peptides and proteins, including albumin, β2-microglobulin, and the aforementioned somatostatin peptides (29). Second, gelofusine has been shown to competitively inhibit the tubular reabsorption of peptide and protein substrates of megalin (27, 30). Third, because gelofusine consists of a heterogeneous mixture of both cationic and anionic peptides of various sizes and structures, it has been suggested to interfere with all four of the charged amino acid repeats on the megalin receptor, thereby effectively blocking the renal reabsorption of a variety of molecules (31). Taken together, this all sits well with the available evidence which points to a role for megalin in the renal reabsorption of polymyxin B and colistin (10, 11, 13). Therefore, it is tenable to imagine that gelofusine competitively inhibits the carrier-mediated renal reabsorption of colistin and may also modify the endosomal/lysosomal trafficking of colistin (32). This dual effect has the potential to influence colistin nephrotoxicity. Extended time-course studies in animals are required prior to conducting trials to ascertain the role of gelofusine in preventing or ameliorating colistin-induced nephrotoxicity in patients.
Polymyxins are important last-line antibiotics, although their clinical usefulness is severely hampered by dose-limiting nephrotoxicity. For the first time, we provide evidence that the clinically available plasma expander gelofusine ameliorates colistin-induced nephrotoxicity. Gelofusine adjunct therapy may represent a promising approach to minimize nephrotoxicity and warrants further evaluation in patients.
MATERIALS AND METHODS
Animals.Swiss mice (female, 6 weeks; body weight, 20 to 25 g) and Sprague-Dawley rats (male; body weight, 300 to 350 g) were obtained from Monash Animal Services. Animals were housed in microisolated metabolic cages in a temperature- (21 ± 3°C) and humidity-controlled facility with a 12-h light-dark cycle (06:00 to 18:00) and acclimatized for 2 days.
Effect of gelofusine on colistin-induced nephrotoxicity in mice.Five groups of mice (n = 6 each) were employed to examine the effect of gelofusine on colistin-induced nephrotoxicity. Mice were dosed with six doses of colistin sulfate injected subcutaneously (s.c.) every 2 h and simultaneously injected intraperitoneally (i.p.) with either saline (control) or gelofusine. The doses of gelofusine are provided as the amount of succinylated bovine gelatin (SBG). The five groups were as follows: (i) a saline control group (0.9% saline s.c.); (ii) the gelofusine3600-only group (600 mg/kg SBG every 2 h; cumulative SBG dose, 3,600 mg/kg); (iii) a colistin-only control (14 mg/kg colistin every 2 h; cumulative dose, 84 mg/kg); (iv) colistin plus saline (14 mg/kg colistin every 2 h; cumulative dose, 84 mg/kg plus 0.9% saline every 2 h); and (v) colistin plus gelofusine3600 (14 mg/kg colistin every 2 h; cumulative colistin dose, 84 mg/kg plus 600 mg/kg SBG every 2 h; cumulative SBG dose, 3,600 mg/kg). At 20 h after the last dose, mice were euthanized by isoflurane inhalation overdose. The right kidney from each mouse was collected and placed in a tube containing 10% neutral buffered formalin pending histological examination, and the left kidney was placed in a preweighed 14-ml plastic tube which was weighed again and stored at −20°C, pending homogenization and analysis of the colistin concentration, as described below.
The influence of lower SBG doses on colistin-induced nephrotoxicity was examined subsequently in three additional groups (n = 3 per group) (i) colistin plus gelofusine1800 (14 mg/kg colistin every 2 h; cumulative dose, 84 mg/kg plus 300 mg/kg SBG every 2 h; cumulative SBG dose, 1,800 mg/kg); (ii) colistin plus gelofusine900 (14 mg/kg colistin every 2 h; cumulative dose, 84 mg/kg plus 150 mg/kg SBG every 2 h; cumulative SBG dose, 900 mg/kg); and (iii) colistin plus gelofusine450 (14 mg/kg colistin every 2 h; cumulative dose, 84 mg/kg plus 75 mg/kg SBG every 2 h; cumulative SBG dose, 450 mg/kg). The mice were euthanized by isoflurane inhalation 20 h after the last dose, and the right kidney from each mouse was removed and placed in a tube containing 10% neutral buffered formalin for blinded histological examination.
Histological examination of kidneys.All histological assessments were performed as previously described (20). Briefly, the nature and severity of the histological changes were initially graded as grade 1, mild acute tubular damage with tubular dilation, prominent nuclei, and a few pale tubular casts; grade 2, severe acute tubular damage with necrosis of tubular epithelial cells and numerous tubular casts; or grade 3, acute cortical necrosis/infarction of tubules and glomeruli with or without papillary necrosis. The grades were given the following scores: grade 1 = 1, grade 2 = 4, and grade 3 = 10. The percentages of the kidney slices affected were scored as follows: <1% = 0, 1 to <5% = 1, 5 to <10% = 2, 10 to <20% = 3, 20 to <30% = 4, 30 to <40% = 5, and ≥40% = 6. The overall kidney histology score was calculated as the product of the percentage score and the grade score. Finally, a simplified semiquantitative score (SQS) (a scale of 0 to +5) for renal histological changes was assigned as follows: SQS 0 = no significant change (overall score, <1); SQS +1 = mild damage (overall score, 1 to <15); SQS +2 = mild to moderate damage (overall score, 15 to <30); SQS +3 = moderate damage (overall score, 30 to <45); SQS +4 = moderate to severe damage (overall score, 45 to <60); and SQS 5 = severe damage (overall score, 60).
Effect of gelofusine on colistin pharmacokinetics in rats.Rats were divided into two groups (n = 4 each) and received either of the following bolus dose combinations via the jugular vein: (i) 0.9% saline (3 ml) 15 min prior to colistin (1 mg/kg) (control) or (ii) gelofusine, SBG dose of 400 mg/kg, 15 min prior to colistin (1 mg/kg). Blood (200 μl) was collected prior to the administration of saline or gelofusine and at 10, 20, 30, 60, 90, 120, 180, 240, and 360 min after colistin administration. Samples were centrifuged immediately (10,000 × g, 10 min) at 4°C for the separation of plasma. Urine was collected in a chilling chamber from the metabolic cage prior to administration of the dose combinations and over 0 to 24 h. All samples were stored at −80°C pending determination of colistin concentrations. High-performance liquid chromatography (HPLC) was utilized to determine the colistin concentrations in plasma, urine, and kidney tissue (mice) (7). Kidney samples (0.08 g of tissue/ml of Milli-Q water) were homogenized (Kinematica Polytron PT-DA 3007/2EC homogenizer; Kinematica, Luzernerstrasse, Switzerland), followed by ultrasonication for 10 s (Sonics Vibra-Cell VCX500). For all matrices (plasma, urine, and kidney homogenate), calibration curves were constructed with colistin concentrations ranging from 0.10 to 8.0 mg/liter. The limit of quantification (LOQ) for colistin in plasma, urine, and kidney homogenate samples was 0.10 mg/liter; for kidney homogenate, the LOQ corresponds to 1.25 μg/g of kidney tissue. The intraday accuracy and reproducibility were <15% at 0.30 mg/liter and <10% at 6.0 mg/liter. The data were analyzed using WinNonlin (noncompartmental model, version 5.2; Pharsight Corp., Cary, NC).
Effect of gelofusine on colistin plasma protein binding.The plasma protein binding of colistin was determined by ultracentrifugation with SBG-dosed rat plasma and healthy gelofusine-free rat plasma, respectively (33). The colistin concentration was determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS), as we have previously described in detail (34).
ACKNOWLEDGMENTS
This study and J.L., T.V., and R.L.N. are supported by a research grant from the National Institute of Allergy and Infectious Diseases of the National Institutes of Health (grant R01 AI111965). J.L., T.V., and R.L.N. are also supported by the Australian National Health and Medical Research Council (NHMRC; grants APP1026109 and APP1085637). E.K.S. is an appointed Young Ambassador for the American Society for Microbiology (ASM) and acknowledges support from the Australian Postgraduate Award.
The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases or the National Institutes of Health. T.V. is an Australian NHMRC Industry Career Development Research Fellow. J.L. is an Australian NHMRC Senior Research Fellow.
FOOTNOTES
- Received 14 May 2017.
- Returned for modification 12 June 2017.
- Accepted 9 September 2017.
- Accepted manuscript posted online 18 September 2017.
- Copyright © 2017 American Society for Microbiology.
