REVIEW
OF
THERAPEUTICS
Adverse Reactions Associated with Systemic Polymyxin Therapy Julie Ann Justo,1,* and John A. Bosso1,2 1
Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, Columbia and Charleston, South Carolina; 2Division of Infectious Diseases, Department of Medicine, Medical University of South Carolina College of Medicine, Charleston, South Carolina
The systemic polymyxins, colistin and polymyxin B, are increasingly used for multidrug-resistant bacterial infections and have a long history of dose-limiting toxicity. This review summarizes the most recent available information about the mechanisms, incidence, risk factors, and minimization strategies for polymyxin toxicity. Nephrotoxicity is related to polymyxin exposure with both size of dose and length of therapy associated with frequency. Newer studies have questioned conventional thinking that the relative risk of nephrotoxicity is lower for colistin than polymyxin B, especially in light of evolving dosing practices. Neurotoxicities and hypersensitivity reactions are less common than nephrotoxicity. New techniques to minimize or avoid polymyxin toxicities are now emerging including a growing interest in clinical assays for therapeutic drug monitoring and the development of novel, less toxic agents (e.g., polymyxin derivatives) for the treatment of multidrug-resistant bacterial infections. KEY WORDS polymyxin, polymyxin B, colistin, toxicity. (Pharmacotherapy 2014;**(**):**–**) doi: 10.1002/phar.1493
Polymyxins were introduced into clinical use more than 50 years ago, and it was quickly recognized that these agents had both nephrotoxic and neurotoxic potential. Patients with preexisting renal compromise appeared to be especially susceptible to both types of toxicity.1 In 1970, a study reported on a series of 317 courses of intravenous colistin (administered as the inactive prodrug colistimethate sodium [CMS]) with an adverse event rate of 25.1% including 20.2% nephrotoxicity, 7.3% neurotoxicity, and 2.2% allergic reactions.2 Approximately 10% of the nephrotoxicity took the form of acute tubular necrosis (ATN), and the most common neurotoxicities were paresthesias. Importantly, these
authors found relationships between the occurrence of nephrotoxicity and neurotoxicity and the magnitude of dose and length of therapy. Most adverse effects occurred within the first few days of beginning therapy and were generally reversible.2 As the use of polymyxins, namely colistin and polymyxin B, has reemerged for treatment of infections caused by multidrugresistant bacteria, a historical perspective is important when reassessing the nature of polymyxin-related adverse effects. We review more current information related to polymyxin toxicity including mechanisms, incidence, risk factors, and minimization strategies. Nephrotoxicity
Sources of Support: The authors have nothing to disclose. *Address for correspondence: Julie Ann Justo, Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, University of South Carolina, Coker Life Sciences Building, 715 Sumter Street, Columbia, SC 29208; e-mail: [email protected]. Ó 2014 Pharmacotherapy Publications, Inc.
Nephrotoxicity is the most clinically concerning adverse effect associated with polymyxins given its relative frequency and potential severity. Accumulation of polymyxins in renal cells, particularly renal proximal tubule cells, appears to result in ATN and renal dysfunction. Renal cell accumulation is likely driven by selective
2
PHARMACOTHERAPY Volume **, Number **, 2014
polymyxin reabsorption by receptors such as megalin on the renal brush border membrane.3 Data from murine models of renal toxicity suggest that cellular accumulation may be a saturable nonpassive process because frequent dosing of polymyxin B (i.e., every 6 hrs) was associated with a significantly greater degree of renal accumulation and cytotoxicity than the equivalent daily dose administered once every 24 hours.4 Oxidative stress appears to play a significant role in the development of renal toxicity as well.5, 6 The exact mechanism of nephrotoxicity is thought to be similar for colistin and polymyxin B because the chemical structures of their components differ by only a single amino acid. However, current in vitro studies comparing cytotoxicities between colistin and polymyxin B are conflicting.7, 8 Incidence It is difficult to determine the true incidence of nephrotoxicity associated with systemic polymyxin use. Not only have dosing practices changed, but clinical definitions of nephrotoxicity vary from one report to the next. More recently, authors use the risk, injury, and failure and loss, and end-stage renal disease (RIFLE) criteria, a consensus-based definition of acute renal failure that adds needed consistency to studies of drugrelated nephrotoxicity.9 However, despite the wide adoption of RIFLE criteria, individual reported rates for colistin and polymyxin B continue to vary substantially. Most reports have focused on colistin use, with rates of acute kidney injury (AKI) ranging from 33–61%.10–17 In other recent trials using other definitions of nephrotoxicity, rates ranged from 18–26% for colistin and as high as 51% for polymyxin B.18–20 Another aspect that has been addressed in some, but not all studies, is the proportion of patients with AKI who subsequently required renal replacement therapy (range 0–28%).14, 15, 17, 20, 21 There are many possible explanations for the wide range of rates of AKI. The patient populations varied, although many studies focused on critically ill patients who would be expected to have large variations in both polymyxin pharmacokinetics and risk factors for AKI. Also, most studies were relatively small and retrospective in design. Of obvious importance was the variation in overall patient exposure to the polymyxin, an issue discussed later in this article. Reports on the relative rates of nephrotoxicity between colistin and polymyxin B are conflict-
ing.7, 10, 11, 18 In one of the more recent comparative studies, risk of AKI in 173 critically ill adults was assessed retrospectively.10 All patients received either polymyxin for at least 72 hours (colistin, n=106; polymyxin B, n=67). There was a significantly higher rate of AKI with colistin (60.4%) versus polymyxin B (41.8%, p=0.02), which persisted after adjustment for potential confounders such as age, APACHE II score, and baseline renal function (adjusted hazard ratio 2.27, p=0.002). However, it should be noted that while polymyxin B dosing was consistent across patients (continuous infusion of 15,000– 25,000 units/kg/day), three different dosing schemes were used in the colistin patients (fixed dose of 150 mg colistin base activity [CBA] every 12 hrs or 5 mg CBA/kg/day of ideal or actual body weight). Although these results are noteworthy, another recent retrospective study using multivariable analysis found no difference in nephrotoxicity between polymyxin agents, although this study evaluated a cohort with a relatively lower proportion of critically ill patients.11 All comparative evaluations to date are limited by the potential for variable in vivo exposures of colistin versus polymyxin B. Variable in vivo exposure is a key factor because the risk of nephrotoxicity is dose dependent.11, 12, 19 Historically, colistin (i.e., CMS) was thought to be less nephrotoxic than polymyxin B.22 The findings of some more recent in vitro and in vivo studies suggest the cytotoxicity of polymyxin B may be comparable or even lower than that of colistin.7, 10 Even as early as the 1960s and 1970s, investigators noted that decreased rates of nephrotoxicity with colistin were likely due to the lower doses and potency of colistin (and CMS) compared with polymyxin B.2, 22 To overcome the inefficient conversion of CMS to colistin (20–25%) and achieve adequate colistin exposure, current pharmacokinetic data suggest more aggressive CMS dosing may be necessary (more than 300 mg CBA daily, the current product-recommended maximum dose).13 Pharmacokinetic-pharmacodynamic (PK-PD) models predict that higher colistin doses are needed when average steady-state colistin plasma concentrations of 2.5 mg/L or greater are desired, specifically for patients with good renal function (creatinine clearance greater than 70 ml/min/ 1.73 m2). The recent findings of higher rates of AKI with colistin versus polymyxin B may be explained by the higher CMS doses that result in greater intrarenal conversion to colistin.23
POLYMYXIN ADVERSE EFFECTS Justo et al However, higher doses of polymyxin B have recently been suggested based on the availability of new PK-PD data in critically ill patients.24 Thus the issue is far from settled. As new dosing strategies for both agents continue to evolve, this issue needs to be studied prospectively. Risk Factors There are many risk factors for polymyxin nephrotoxicity2,11,12,14,16,19,25–29 (Table 1). Risk factor assessment usually centers around dosing issues. The magnitude of the daily dose, total cumulative dose, and length of therapy have all been associated with risk of nephrotoxicity. At the same time, other authors either did not evaluate these factors or failed to find associations with one or more. More important, however, is the observation that high rates of nephrotoxicity may be seen with a range of doses including what are today considered appropriate doses.10, 12 Weight-based dosing must also be evaluated, with potentially heightened awareness in obese patients. Two retrospective studies of colistin and polymyxin B use showed that increased body mass index was associated with AKI,14, 29 raising the question of whether obesity itself is a risk factor for polymyxin-induced nephrotoxicity. Other data have suggested that colistin dosing based on actual body weight (ABW), instead of the product-recommended ideal body weight (IBW), may increase the risk of AKI with colistin in particular.12, 16 Contemporary pharmacokinetic data regarding polymyxin B in obesity are limited to a single 250kg patient on continuous renal replacement therapy and suggest that ABW-dosing produced similar drug exposure as nonobese patients.24, 30 Table 1. Risk Factors for Polymyxin Nephrotoxicity Risk factor Polymyxin exposure Length of therapy2, 25–28 Daily dose11, 12, 16, 19 Cumulative dose2, 14, 25 Other drug exposure Concomitant nephrotoxinsa;10–12, 14–17 Patient factors Obesity14, 29 Preexisting renal dysfunction/disease1, 10, Age10, 16, 17, 29 Diabetes29 Hypertension10 Hypoalbuminemia15 Hyperbilirubinemia15 a
27
Including agents such as vancomycin, aminoglycosides, vasopressors, and calcineurin inhibitors.
3
The concomitant use of other nephrotoxins has, not surprisingly, been noted as a risk factor for polymyxin nephrotoxicity.10–12, 14–17 Preexisting renal disease or dysfunction is also a recognized risk factor. First reported in 1962,1 several other authors have corroborated this observation, including one study with renal transplant patients.10, 27 Table 1 lists other reported risk factors. As with the interpretation of results from any small study, the validity of reported associations should be considered with caution to determine if they are true causal relationships or simply a function of common practices in critically ill patients. Neurotoxicity Polymyxins are also known for causing neurotoxicities including paresthesias, apnea, nausea and vomiting, dizziness, myopathy, neuropathy, confusion, psychosis, and seizures.2, 26 Paresthesias (typically perioral), nausea and vomiting, and dizziness are most common and often benign. Polymyxin-associated neurotoxicity including more serious adverse reactions (e.g., apnea and respiratory distress) appear to stem from neuromuscular blockade which may be due to the inhibition of synaptic acetylcholine release, competitive inhibition of acetylcholine, or prolonged depolarization and calcium depletion.26 Incidence The propensity for neurotoxicity with polymyxins appears lower than that for nephrotoxicity. Most reported incidence rates are 7% or less, with paresthesias being the most common reaction.2 However, higher rates of neurotoxicity (27–29%) have been noted in older studies with intramuscular colistin and in more modern studies of patients with cystic fibrosis receiving higher colistin doses (6–8 mg/kg/day).26, 31 Reports of serious neurologic reactions, such as apnea, myopathy, ataxia, and severe dizziness, generally stem from older studies as well1, 2 and were exceedingly rare in the last few decades. Multiple recent cohort studies with polymyxins actually reported no cases of neurotoxicity.18, 25 Yet with the resurgence of polymyxin use, five additional cases of respiratory failure (one with convulsions) were reported in recent years.32–35 These incidence estimates should be interpreted cautiously because the true incidence of neurotoxicity may be underreported, especially in the
4
PHARMACOTHERAPY Volume **, Number **, 2014
critically ill population where subjective symptoms and neurologic status are difficult to assess routinely. Risk Factors Similar to nephrotoxicity, the risk of neurotoxicity is linked to increased drug exposure, either as greater length of therapy or higher polymyxin dose.1, 26 Another identified risk factor, renal dysfunction, is likely linked to increased drug exposure as well. Older reports of more serious reactions, such as apnea, ataxia, and severe dizziness, frequently occurred in patients with renal impairment.1, 2 Product-recommended renal dosage adjustment of polymyxins was not consistently performed in these cases, which may have increased the risk of dose-dependent toxicity. Patients with myasthenia gravis may also be at increased risk for adverse reactions such as respiratory arrest, secondary to neuromuscular blockade.36 Additional risk factors for neurotoxicity include hypoxia, female sex, and concomitant medications (i.e., sedatives, anesthetics, muscle relaxants, narcotics, or corticosteroids).2, 26 Other Adverse Reactions Most of the literature regarding polymyxin toxicity focuses on nephrotoxicity followed by neurotoxicity. Whether this reflects the low incidence of other potential side effects or their lack of import is unclear. Early reports of allergic or hypersensitivity reactions to the polymyxins were published,2 and, over the years, others have reported cases of pruritus with and without rashes.26 It is also well known that inhaled colistin is irritating to the airways causing complaints ranging from minor irritation (cough, sore throat) to bronchoconstriction. It has been speculated that both types of reactions are mediated by histamine.26 Timing, Treatment, and Reversibility The onset of nephrotoxicity and neurotoxicity most commonly occurs during the first few days of polymyxin therapy. One study reported that most cases of nephrotoxicity and neurotoxicity associated with colistin developed during the first 4 days of therapy (64% and 83%, respectively).2 Immediate reactions may also occur during administration including transient paresthesias or apnea with intravenous infusion. Treatment generally involves dosage reduction
or drug discontinuation and supportive care. Renal replacement therapy has been used in cases of nephrotoxicity and/or neurotoxicity as support for renal dysfunction26 and for CMS/ colistin removal (particularly continuous renal replacement therapy).35 Of note, renal replacement therapy is unlikely to decrease polymyxin B concentrations significantly (only 6–12% of polymyxin B is cleared by continuous renal replacement therapy).30 For neurotoxicity, mechanical ventilation provides support during respiratory failure,32–35 and treatment with calcium or cholinesterase inhibitors has shown mixed results.26 Nephrotoxicity, neurotoxicity, and other adverse reactions are generally reversible following drug discontinuation,20, 26 particularly upon prompt recognition of the reaction. However, resolution rates and outcomes may be poorer for nephrotoxic patients requiring renal replacement therapy.15 Minimization Strategies Despite the significant toxicity of polymyxins, their use is often necessary as a last-line option for treatment of life-threatening multidrug-resistant bacterial infections. Clinicians should therefore consider further research and potential use of minimization strategies, examples of which are summarized here. Promising data from animal models support the use of concomitant antioxidants, for example melatonin and ascorbic acid, to prevent polymyxin-induced nephrotoxicity.5, 6 One study found rats treated with colistin and concomitant high-dose ascorbic acid (200 mg/kg/day, equivalent to ~2 g/day in a 60-kg adult) over 7 days exhibited significantly decreased urinary excretion of N-acetyl-b-D-glucosaminidase, a sensitive marker for renal tubular damage, compared with those given colistin alone (p
OF
THERAPEUTICS
Adverse Reactions Associated with Systemic Polymyxin Therapy Julie Ann Justo,1,* and John A. Bosso1,2 1
Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, Columbia and Charleston, South Carolina; 2Division of Infectious Diseases, Department of Medicine, Medical University of South Carolina College of Medicine, Charleston, South Carolina
The systemic polymyxins, colistin and polymyxin B, are increasingly used for multidrug-resistant bacterial infections and have a long history of dose-limiting toxicity. This review summarizes the most recent available information about the mechanisms, incidence, risk factors, and minimization strategies for polymyxin toxicity. Nephrotoxicity is related to polymyxin exposure with both size of dose and length of therapy associated with frequency. Newer studies have questioned conventional thinking that the relative risk of nephrotoxicity is lower for colistin than polymyxin B, especially in light of evolving dosing practices. Neurotoxicities and hypersensitivity reactions are less common than nephrotoxicity. New techniques to minimize or avoid polymyxin toxicities are now emerging including a growing interest in clinical assays for therapeutic drug monitoring and the development of novel, less toxic agents (e.g., polymyxin derivatives) for the treatment of multidrug-resistant bacterial infections. KEY WORDS polymyxin, polymyxin B, colistin, toxicity. (Pharmacotherapy 2014;**(**):**–**) doi: 10.1002/phar.1493
Polymyxins were introduced into clinical use more than 50 years ago, and it was quickly recognized that these agents had both nephrotoxic and neurotoxic potential. Patients with preexisting renal compromise appeared to be especially susceptible to both types of toxicity.1 In 1970, a study reported on a series of 317 courses of intravenous colistin (administered as the inactive prodrug colistimethate sodium [CMS]) with an adverse event rate of 25.1% including 20.2% nephrotoxicity, 7.3% neurotoxicity, and 2.2% allergic reactions.2 Approximately 10% of the nephrotoxicity took the form of acute tubular necrosis (ATN), and the most common neurotoxicities were paresthesias. Importantly, these
authors found relationships between the occurrence of nephrotoxicity and neurotoxicity and the magnitude of dose and length of therapy. Most adverse effects occurred within the first few days of beginning therapy and were generally reversible.2 As the use of polymyxins, namely colistin and polymyxin B, has reemerged for treatment of infections caused by multidrugresistant bacteria, a historical perspective is important when reassessing the nature of polymyxin-related adverse effects. We review more current information related to polymyxin toxicity including mechanisms, incidence, risk factors, and minimization strategies. Nephrotoxicity
Sources of Support: The authors have nothing to disclose. *Address for correspondence: Julie Ann Justo, Department of Clinical Pharmacy and Outcomes Sciences, South Carolina College of Pharmacy, University of South Carolina, Coker Life Sciences Building, 715 Sumter Street, Columbia, SC 29208; e-mail: [email protected]. Ó 2014 Pharmacotherapy Publications, Inc.
Nephrotoxicity is the most clinically concerning adverse effect associated with polymyxins given its relative frequency and potential severity. Accumulation of polymyxins in renal cells, particularly renal proximal tubule cells, appears to result in ATN and renal dysfunction. Renal cell accumulation is likely driven by selective
2
PHARMACOTHERAPY Volume **, Number **, 2014
polymyxin reabsorption by receptors such as megalin on the renal brush border membrane.3 Data from murine models of renal toxicity suggest that cellular accumulation may be a saturable nonpassive process because frequent dosing of polymyxin B (i.e., every 6 hrs) was associated with a significantly greater degree of renal accumulation and cytotoxicity than the equivalent daily dose administered once every 24 hours.4 Oxidative stress appears to play a significant role in the development of renal toxicity as well.5, 6 The exact mechanism of nephrotoxicity is thought to be similar for colistin and polymyxin B because the chemical structures of their components differ by only a single amino acid. However, current in vitro studies comparing cytotoxicities between colistin and polymyxin B are conflicting.7, 8 Incidence It is difficult to determine the true incidence of nephrotoxicity associated with systemic polymyxin use. Not only have dosing practices changed, but clinical definitions of nephrotoxicity vary from one report to the next. More recently, authors use the risk, injury, and failure and loss, and end-stage renal disease (RIFLE) criteria, a consensus-based definition of acute renal failure that adds needed consistency to studies of drugrelated nephrotoxicity.9 However, despite the wide adoption of RIFLE criteria, individual reported rates for colistin and polymyxin B continue to vary substantially. Most reports have focused on colistin use, with rates of acute kidney injury (AKI) ranging from 33–61%.10–17 In other recent trials using other definitions of nephrotoxicity, rates ranged from 18–26% for colistin and as high as 51% for polymyxin B.18–20 Another aspect that has been addressed in some, but not all studies, is the proportion of patients with AKI who subsequently required renal replacement therapy (range 0–28%).14, 15, 17, 20, 21 There are many possible explanations for the wide range of rates of AKI. The patient populations varied, although many studies focused on critically ill patients who would be expected to have large variations in both polymyxin pharmacokinetics and risk factors for AKI. Also, most studies were relatively small and retrospective in design. Of obvious importance was the variation in overall patient exposure to the polymyxin, an issue discussed later in this article. Reports on the relative rates of nephrotoxicity between colistin and polymyxin B are conflict-
ing.7, 10, 11, 18 In one of the more recent comparative studies, risk of AKI in 173 critically ill adults was assessed retrospectively.10 All patients received either polymyxin for at least 72 hours (colistin, n=106; polymyxin B, n=67). There was a significantly higher rate of AKI with colistin (60.4%) versus polymyxin B (41.8%, p=0.02), which persisted after adjustment for potential confounders such as age, APACHE II score, and baseline renal function (adjusted hazard ratio 2.27, p=0.002). However, it should be noted that while polymyxin B dosing was consistent across patients (continuous infusion of 15,000– 25,000 units/kg/day), three different dosing schemes were used in the colistin patients (fixed dose of 150 mg colistin base activity [CBA] every 12 hrs or 5 mg CBA/kg/day of ideal or actual body weight). Although these results are noteworthy, another recent retrospective study using multivariable analysis found no difference in nephrotoxicity between polymyxin agents, although this study evaluated a cohort with a relatively lower proportion of critically ill patients.11 All comparative evaluations to date are limited by the potential for variable in vivo exposures of colistin versus polymyxin B. Variable in vivo exposure is a key factor because the risk of nephrotoxicity is dose dependent.11, 12, 19 Historically, colistin (i.e., CMS) was thought to be less nephrotoxic than polymyxin B.22 The findings of some more recent in vitro and in vivo studies suggest the cytotoxicity of polymyxin B may be comparable or even lower than that of colistin.7, 10 Even as early as the 1960s and 1970s, investigators noted that decreased rates of nephrotoxicity with colistin were likely due to the lower doses and potency of colistin (and CMS) compared with polymyxin B.2, 22 To overcome the inefficient conversion of CMS to colistin (20–25%) and achieve adequate colistin exposure, current pharmacokinetic data suggest more aggressive CMS dosing may be necessary (more than 300 mg CBA daily, the current product-recommended maximum dose).13 Pharmacokinetic-pharmacodynamic (PK-PD) models predict that higher colistin doses are needed when average steady-state colistin plasma concentrations of 2.5 mg/L or greater are desired, specifically for patients with good renal function (creatinine clearance greater than 70 ml/min/ 1.73 m2). The recent findings of higher rates of AKI with colistin versus polymyxin B may be explained by the higher CMS doses that result in greater intrarenal conversion to colistin.23
POLYMYXIN ADVERSE EFFECTS Justo et al However, higher doses of polymyxin B have recently been suggested based on the availability of new PK-PD data in critically ill patients.24 Thus the issue is far from settled. As new dosing strategies for both agents continue to evolve, this issue needs to be studied prospectively. Risk Factors There are many risk factors for polymyxin nephrotoxicity2,11,12,14,16,19,25–29 (Table 1). Risk factor assessment usually centers around dosing issues. The magnitude of the daily dose, total cumulative dose, and length of therapy have all been associated with risk of nephrotoxicity. At the same time, other authors either did not evaluate these factors or failed to find associations with one or more. More important, however, is the observation that high rates of nephrotoxicity may be seen with a range of doses including what are today considered appropriate doses.10, 12 Weight-based dosing must also be evaluated, with potentially heightened awareness in obese patients. Two retrospective studies of colistin and polymyxin B use showed that increased body mass index was associated with AKI,14, 29 raising the question of whether obesity itself is a risk factor for polymyxin-induced nephrotoxicity. Other data have suggested that colistin dosing based on actual body weight (ABW), instead of the product-recommended ideal body weight (IBW), may increase the risk of AKI with colistin in particular.12, 16 Contemporary pharmacokinetic data regarding polymyxin B in obesity are limited to a single 250kg patient on continuous renal replacement therapy and suggest that ABW-dosing produced similar drug exposure as nonobese patients.24, 30 Table 1. Risk Factors for Polymyxin Nephrotoxicity Risk factor Polymyxin exposure Length of therapy2, 25–28 Daily dose11, 12, 16, 19 Cumulative dose2, 14, 25 Other drug exposure Concomitant nephrotoxinsa;10–12, 14–17 Patient factors Obesity14, 29 Preexisting renal dysfunction/disease1, 10, Age10, 16, 17, 29 Diabetes29 Hypertension10 Hypoalbuminemia15 Hyperbilirubinemia15 a
27
Including agents such as vancomycin, aminoglycosides, vasopressors, and calcineurin inhibitors.
3
The concomitant use of other nephrotoxins has, not surprisingly, been noted as a risk factor for polymyxin nephrotoxicity.10–12, 14–17 Preexisting renal disease or dysfunction is also a recognized risk factor. First reported in 1962,1 several other authors have corroborated this observation, including one study with renal transplant patients.10, 27 Table 1 lists other reported risk factors. As with the interpretation of results from any small study, the validity of reported associations should be considered with caution to determine if they are true causal relationships or simply a function of common practices in critically ill patients. Neurotoxicity Polymyxins are also known for causing neurotoxicities including paresthesias, apnea, nausea and vomiting, dizziness, myopathy, neuropathy, confusion, psychosis, and seizures.2, 26 Paresthesias (typically perioral), nausea and vomiting, and dizziness are most common and often benign. Polymyxin-associated neurotoxicity including more serious adverse reactions (e.g., apnea and respiratory distress) appear to stem from neuromuscular blockade which may be due to the inhibition of synaptic acetylcholine release, competitive inhibition of acetylcholine, or prolonged depolarization and calcium depletion.26 Incidence The propensity for neurotoxicity with polymyxins appears lower than that for nephrotoxicity. Most reported incidence rates are 7% or less, with paresthesias being the most common reaction.2 However, higher rates of neurotoxicity (27–29%) have been noted in older studies with intramuscular colistin and in more modern studies of patients with cystic fibrosis receiving higher colistin doses (6–8 mg/kg/day).26, 31 Reports of serious neurologic reactions, such as apnea, myopathy, ataxia, and severe dizziness, generally stem from older studies as well1, 2 and were exceedingly rare in the last few decades. Multiple recent cohort studies with polymyxins actually reported no cases of neurotoxicity.18, 25 Yet with the resurgence of polymyxin use, five additional cases of respiratory failure (one with convulsions) were reported in recent years.32–35 These incidence estimates should be interpreted cautiously because the true incidence of neurotoxicity may be underreported, especially in the
4
PHARMACOTHERAPY Volume **, Number **, 2014
critically ill population where subjective symptoms and neurologic status are difficult to assess routinely. Risk Factors Similar to nephrotoxicity, the risk of neurotoxicity is linked to increased drug exposure, either as greater length of therapy or higher polymyxin dose.1, 26 Another identified risk factor, renal dysfunction, is likely linked to increased drug exposure as well. Older reports of more serious reactions, such as apnea, ataxia, and severe dizziness, frequently occurred in patients with renal impairment.1, 2 Product-recommended renal dosage adjustment of polymyxins was not consistently performed in these cases, which may have increased the risk of dose-dependent toxicity. Patients with myasthenia gravis may also be at increased risk for adverse reactions such as respiratory arrest, secondary to neuromuscular blockade.36 Additional risk factors for neurotoxicity include hypoxia, female sex, and concomitant medications (i.e., sedatives, anesthetics, muscle relaxants, narcotics, or corticosteroids).2, 26 Other Adverse Reactions Most of the literature regarding polymyxin toxicity focuses on nephrotoxicity followed by neurotoxicity. Whether this reflects the low incidence of other potential side effects or their lack of import is unclear. Early reports of allergic or hypersensitivity reactions to the polymyxins were published,2 and, over the years, others have reported cases of pruritus with and without rashes.26 It is also well known that inhaled colistin is irritating to the airways causing complaints ranging from minor irritation (cough, sore throat) to bronchoconstriction. It has been speculated that both types of reactions are mediated by histamine.26 Timing, Treatment, and Reversibility The onset of nephrotoxicity and neurotoxicity most commonly occurs during the first few days of polymyxin therapy. One study reported that most cases of nephrotoxicity and neurotoxicity associated with colistin developed during the first 4 days of therapy (64% and 83%, respectively).2 Immediate reactions may also occur during administration including transient paresthesias or apnea with intravenous infusion. Treatment generally involves dosage reduction
or drug discontinuation and supportive care. Renal replacement therapy has been used in cases of nephrotoxicity and/or neurotoxicity as support for renal dysfunction26 and for CMS/ colistin removal (particularly continuous renal replacement therapy).35 Of note, renal replacement therapy is unlikely to decrease polymyxin B concentrations significantly (only 6–12% of polymyxin B is cleared by continuous renal replacement therapy).30 For neurotoxicity, mechanical ventilation provides support during respiratory failure,32–35 and treatment with calcium or cholinesterase inhibitors has shown mixed results.26 Nephrotoxicity, neurotoxicity, and other adverse reactions are generally reversible following drug discontinuation,20, 26 particularly upon prompt recognition of the reaction. However, resolution rates and outcomes may be poorer for nephrotoxic patients requiring renal replacement therapy.15 Minimization Strategies Despite the significant toxicity of polymyxins, their use is often necessary as a last-line option for treatment of life-threatening multidrug-resistant bacterial infections. Clinicians should therefore consider further research and potential use of minimization strategies, examples of which are summarized here. Promising data from animal models support the use of concomitant antioxidants, for example melatonin and ascorbic acid, to prevent polymyxin-induced nephrotoxicity.5, 6 One study found rats treated with colistin and concomitant high-dose ascorbic acid (200 mg/kg/day, equivalent to ~2 g/day in a 60-kg adult) over 7 days exhibited significantly decreased urinary excretion of N-acetyl-b-D-glucosaminidase, a sensitive marker for renal tubular damage, compared with those given colistin alone (p
