national geographic
药物肾毒性的“6R”原则
篇名：The 6R’s of drug induced nephrotoxicity
turnover作者：Linda Awdishu，Ravindra L. Mehta
出处：BMC Nephrol. 2017; 18: 124.
PMCID: PMC5379580
摘要：
药物诱导的肾损伤是导致发病率增加和医疗保健利⽤率增加的常见不良事件。由于肾损伤的定义不同，并发危险因素评估不全和缺乏长期预后报告，导致现阶段我们对药物诱发肾脏疾病的认识有限。电⼦监控是识别易感⼈群的有⼒⼯具，改善对不良事件的认识，并对预防性策略或对早期⼲预损伤的提供决策⽀持。肾损伤和遗传易感性⽣物标志物可强化对药物不良反应的监测，以识别出易受伤害，并可能降低风险的患者。在本综述中，我们将提供⼀个6R框架来识别和管理药物诱发的肾损伤，即— 风险，识别，反应，肾脏⽀持，康复和研究！
Abstract
Drug induced kidney injury is a frequent adver event which contributes to morbidity and incread healthcare utilization. Our current knowledge of drug induced kidney dia is limited due to varying definitions of kidney injury, incomplete asssment of concurrent risk factors and lack of long term outcome reporting. Electronic surveillance prents a powerful tool to identify susceptible populations, improve recognition of events and provide decision support on preventative strategies or early intervention in the ca of injury. Rearch in the area of biomarkers for detecting kidney injury and genetic predisposition for this adver event will enhance detection of injury, identify tho susceptible to injury and likely mitigate risk. In this review we will prent a 6R framework to identify and mange drug induced kidney injury –risk, recognition, respon, renal support, rehabilitation and rearch.
Background背景
Drug-induced nephrotoxicity is increasingly recognized as a significant contributor to kidney dia including acute kidney injury (AKI) and chronic kidney dia (CKD). Nephrotoxicity has a wide spectrum, reflecting damage to different nephron gments bad upon individual drug mechanisms. Both glomerular and tubular injuries are recognized targets for drug toxicity and may result in acute or chronic functional changes. However, standard definitions of drug induced kidney d
ia (DIKD) are lacking, leading to challenges in recognition and reporting. The clinical manifestations of DIKD often go unrecognized, particularly in the tting of short drug exposures. This pos challenges in asssing the incidence, verity and long-term conquences of DIKD.
Our knowledge of the epidemiology of nephrotoxicity focus predominantly on drug induced AKI. Prospective cohort studies of AKI have documented the frequency of drug-induced nephrotoxicity to be approximately 14-26% in adult populations [1–3]. Nephrotoxicity is a significant concern in pediatrics with 16% of hospitalized AKI events being attributable primarily to a drug [4]. The epidemiology of tubular disorders is unclear as a standard definition is lacking and many published reports document tubular dysfunction leading to AKI. This may under-estimate the true incidence of tubular disorders since only cas associated with a change in rum creatinine (Scr) are recognized. However, frequent u of specific drugs, such as tenofovir, has led to greater attention to tubular injuries with documented frequencies of 12–22% of treated subjects in cohort studies [5, 6]. Glomerular injury is uncommon and most of the literature is limited to ca reports or ca ries. However, novel chemotherapeutic agents are increasingly being associated with this form of toxicity [7]. Given the challenges in the
reported epidemiology and outcomes of DIKD, we propo a novel framework to approach drug indu
ced nephrotoxicity focud on Risk asssment, early Recognition, targeted Respon, timely Renal support and Rehabilitation coupled with Rearch (the 6R approach). Risk风险
To evaluate the risk of nephrotoxicity, general questions can be applied to each causal drug. What is the predictable risk bad on the known pharmacology of the drug? What is the known risk, contributing risk factors and the typical timeline for injury? If the risks are known, how is this information ud clinically to predict the risk for an individual patient (i.e. clinical risk scores for contrast nephropathy)? How is this information ud to mitigate the risk?
gpl
Drug induced adver events can be classified into two categories: do dependent and idiosyncratic reactions. This categorization is important to consider in the context of drug induced kidney dia (DIKD) since the mechanisms for drug toxicity are different posing challenges in risk asssment. Do dependent reactions are predictable from the known pharmacology of the drug. For example, the risk of aminoglycoside induced nephrotoxicity increas with higher trough drug concentrations and longer duration of therapy [8]. Whereas, interstitial nephritis from proton pump inhibitors is an unpredictable idiosyncratic reaction, which is unlikely to be preventable or minimized.
When asssing the known risk of DIKD, often this information may be in the form of ca reports, a
dver event reporting from clinical trials or post marketing surveillance [9]. Prospective studies focud on determining the incidence of DIKD are few. Most studies are retrospective and are focud largely on drugs with predictable toxicities and therapeutic drug monitoring (TDM). Determining the incidence of idiosyncratic reactions is difficult since data is often limited to ca reports. Some studies attempt to demonstrate an association using claims data and diagnostic codes; however, the incidence of the AKI is variable between cohorts and likely overstated [10–12] Most importantly, the definition of DIKD has not been standardized, making interpretation of the epidemiology challenging. The information on drug specific DIKD risk is summarized in Table 1.
签署英文Risk factors contributing to the development of DIKD include patient specific factors, dia specific factors and process of care factors (Table 1). Common risk factors include age, causal drug single and/or cumulative do, underlying CKD and concurrent nephrotoxin exposures. In the ca of hospitalized patients, our experience is that a retrospective evaluation of DIKD almost always reveals the prescription of additional nephrotoxins concurrent to the causal drug (i.e. ketorolac prescribed to a patient receiving gentamicin and vancomycin). Minimizing the exposures may mitigate the development of DIKD.
Asssing kidney function is critical to the dosing of drugs and mitigation of DIKD. An important patie
nt specific risk factor is low rum Scr values due to reduced muscle mass, which may be age related or dia related (muscular dystrophy, spina bifida, etc.). This pos a challenge to asssment of kidney function using estimating equations. Pharmacists often “round” Scr values to an arbitrary threshold value in older patients or tho with low Scr values to account for low muscle mass. This practice is inaccurate and may lead to drug dosing errors in certain populations [13–16]. Currently, KDIGO guidelines on drug dosing advocate using either Cockcroft Gault or MDRD equation for drug dosing [17]. Since the drug information from manufacturers submitted to the U.S. Food and Drug Administration still utilizes the Cockcroft Gault equation for estimates of kidney function and no prospective studies have been conducted on clinical outcomes of the various equations, we feel that either equation could be ud in the abnce of kidney dia.
Published reports of DIKD have not consistently evaluated cas for the prence of common AKI risk factors. Subquently, risk factors specific to a causal agent have emerged but have not been validated in larger studies and across multiple drugs. As an example, drug interactions have emerged as an important risk factor for the development of AKI. Interactions leading to incread concentrations of anti-hypertensive medications, subquent hypotension and AKI have been reported [18]. In a study by Gandhi and colleagues, the risk for hospitalization with AKI was compare
d in patients receiving a prescription for amlodipine and one of two macrolide antimicrobials, clarithromycin or azithromycin. Clarithromycin is known to inhibit cytochrome P450 3A4 isoenzyme, which is involved in the metabolism of amlodipine, whereas azithromycin does not interact to the same extent. The authors found co-prescription with clarithromycin was associated with an odds ratio [OR], 1.98 [95% CI, 1.68–2.34] compared to co-prescription with azithromycin [18].
Identification of general DIKD risk factors is central to the development of clinical risk scores for the prediction and minimization of risk. For example, the identification of risk factors for contrast-induced nephropathy has led to the development of risk scores and evaluation of preventative treatments [19–23]. This has great applicability to the clinical tting, where an electronic medical record (EMR) can calculate the risk score and cardiologists or radiologists can prescribe preventative measures. Additionally the risk scores may predict long-term outcomes [24, 25].
thin
Recognition 识别
Currently, there is no standard definition of DIKD and incidence of nephrotoxicity varies depending on the definition employed and the causal drug. The most common drugs that cau DIKD include antibiotics, anti-rejection medications, antiviral agents, non-steroidal anti-inflammatory agents, anti-ulcer agents and chemotherapy.
Most studies have defined nephrotoxicity as 0.5 mg/dL or 50% ri in Scr over 24–72 h time frame and a minimum 24–48 h of drug exposure. However, the definitions po challenges since a 50% increa in Scr may not have high specificity for DIKD since the underlying dia being treated as well as other AKI risk factors could be significant to the attribution of risk. In the tting of fluctuating renal function or tho patients receiving renal replacement therapies, it is difficult to recognize DIKD. For example, if a critically ill patient develops AKI from psis, it may be difficult to recognize whether an antibiotic is causing additional injury to the susceptible kidney. Recognition is also complicated by the fact that the mechanism of kidney injury and time period for ont of injury varies by drug and some drugs cau injury by more than one mechanism. For instance, NSAIDS can result in AKI due to hemodynamic changes or acute interstitial nephritis (AIN), or nephrotic range proteinuria from glomerular injury.
In order to improve the recognition of DIKD in the literature, we convened an expert panel to develop connsus-bad definitions [26]. We propo that DIKD prents in one of four phenotypes: AKI, glomerular disorder, tubular disorder, or nephrolithiasis/crystalluria [26]. The clinical prentation of each phenotype is bad on primary and condary criteria. We suggest that at least one primary criterion must be met for all drugs suspected of causing DIKD [26]. For each phenotype definition, the following critical elements from the Bradford-Hill causal criteria must be met:
1. The drug exposure must be at least 24 h preceding the event.
2. There should be biological plausibility for the causal drug, bad on known mechanism
ingredient
of drug effect; metabolism and immunogenicity.
3. Complete data (including but not limited to co-morbidities, additional nephrotoxic
exposures, exposure to contrast agents, surgical procedures, blood pressure, urine
output) surrounding the period of drug exposure is required to account for concomitant risks and exposures to other nephrotoxic agents.
4. The strength of the relationship between the attributable drug and phenotype should belastbutnotleast
bad on drug exposure duration, extent of primary and condary criteria met and the
time cour of the injury.
In defining the time cour for DIKD, it is important to consider connsus definitions for AKI, acute kidney dia and CKD. Acute kidney injury develops in 7 days or less, injury beyond 7 days but le
ss than 90 days reflects acute kidney dia and beyond 90 days CKD [27]. Using the KDIGO definitions, the development of DIKD can similarly be divided into acute (1–7 days), sub-acute (8–90 days) and chronic (>90 days) post drug exposure [26]. This approach permits classification and tracking of injuries for duration and outcomes. Bad on this conceptual model, for each phenotype, thresholds could be established to detect DIKD, define its verity and ascertain recovery.
在线读英语单词The reference Scr ud for defining DIKD should as clo as possible to the event to meet the definition of AKI but may not always be available as in the ca of ambulatory care exposures. In this scenario, we recommend using the lowest Scr within 90 days of the event as the reference Scr. It is recognized that CKD is an important risk factor for the development of DIKD. Underlying kidney dia impacts the recognition of DIKD. We recommend using a Scr value greater than 90 days from the DIKD event to define the prence of CKD.
The standard definitions will become increasingly important when designing tools within the EMR to screen for DIKD. Such screening tools have been successful at identifying AKI and guiding the physician on the need for nephrology consultation [28]. Pharmacovigilence programs can identify patients who have been expod to nephrotoxic medications and develop AKI with high rum drug concentrations [29]. Additionally, the electronic screening tools can be customized. At risk patients,
such as tho receiving multiple nephrotoxins or prolonged nephrotoxin exposures, can be targeted. Identification of such patients can prompt interventions such as intensified Scr monitoring and improve the recovery of DIKD [30]. However, electronic screening and identification cannot establish causality. The tools are limited due to the complex interplay of risk factor asssment, concurrent multi-drug exposures, lack of TDM, comorbid conditions and lack of kidney damage biomarkers. It is important that DIKD cas are adjudicated for causality and an attribution of risk is estimated for each contributing drug or risk factor. In the ca of vancomycin, a pharmaco-vigilence program identified 32% of patients expod to vancomycin with high trough concentrations and AKI [29]. However, when the cas were adjudicated, only 8.4% of AKI cas were attributed to vancomycin toxicity [29]. Attribution of risk from each potential risk factor or from each causal drug in the ca of multi-drug injury is difficult since the asssments are bad on the individual patient prentation and might reflect a substantial degree of subjectivity depending on the adjudicator’s knowledge of DIKD and AKI epidemiology. We recommend when evaluating cas of DIKD, the consulting nephrologist document their causality asssment in the medical record including a percent attribution assigned to each causal drug with an overall likelihood to cau the DIKD, as well as a percent attribution for each of the identified concurrent AKI risk factors. Adver event causality scoring tools exist for general adver events as well as drug induced liver and skin injury (Naranjo,
Rucam, Liverpool), however, the tools have not been evaluated for the causality scoring of DIKD. Previous genomic studies of drug induced liver and skin injury have employed adjudication of cas by unbiad hepatologists or immunologists/dermatologists, respectively [31, 32]. Often, published ca reports lack the evaluation of causality using the scoring systems or adjudication. As the body of knowledge surrounding DIKD increas, we recommend employing the scoring tools in addition to adjudication of cas by a condary nephrologist when publishing ca reports or ries.
Respon 响应
Treatment of nephrotoxicity is dependent on the phenotype, verity of the injury and the underlying condition for which the medication was prescribed. The decision to stop or reduce the do of the offending drug requires a careful consideration of the risk versus benefit. In Type A reactions, do reduction may be sufficient to mitigate the injury (e.g. vancomycin or gentamicin). However, stage 2 AKI often warrants drug discontinuation. In the tting of vancomycin DIKD, a critical appraisal of other therapeutic options and do minimization is warranted. National guidelines on the u of vancomycin have recommended higher target
trough concentrations to obtain a high area under the curve (AUC) to minimum inhibitory concentration (MIC) ratio [33, 34]. However, the level of evidence for this recommendation was grade IIIb (limited evidence) [34]. With the widespread adoption of this recommendation [35], the rate of nephrotoxicity has incread. Meta-analysis conducted found the incidence of nephrotoxicity to be between 5-43% and target trough concentrations > 15 ng/mL to have a 2.67 odds ratio for the development of nephrotoxicity [36]. A more recent study of 1430 patients receiving vancomycin provides support for the association between concentrations and duration of therapy with risk of nephrotoxicity [37]. Post hoc analysis of prospective studies have examined the need for higher targets and demonstrated equivocal or lower cure rates with trough concentrations above 15 ng/mL for the treatment of staphylococcus aureus nosocomial acquired pneumonia [38, 39]. Additionally, the studies have demonstrated that alternative treatments such as linezolid or telavancin could be considered [39, 40]. Bad on the studies, we believe that DIKD from higher vancomycin trough concentrations is a real concern. However, prospective studies designed to evaluate the benefits and risks of high therapeutic concentrations need to be done. Type B DIKD, which is idiosyncratic, will require discontinuation of the offending drug and careful obrvation. Severe injuries or type B reactions often require longer periods of time to improve and may not completely resolve. When DIKD has been identified, the patient should be monitored carefully including daily asssment of Sc
r and urine output as changes in kidney function may lead to further injury or lack of clinical cure for infections. Concurrent risk factors for kidney injury should be addresd such as but not limited to hypotension, hyperglycemia, anemia, minimization of nephrotoxins or drug interactions, which may contribute to the injury. Do adjustments for kidney function should be made for other medications the patient is receiving. In some cas, timed urine collections for CLcr determination may be warranted to assist in the determination of renal function for the purpo of dosage adjustment. Where available, TDM should be employed and continued even after drug discontinuation in cas where supra-therapeutic concentrations are documented during the injury. Pharmacist consultation improves the achievement of target concentrations and improves clinical cure rates [41]. Additionally, documentation of the event is imperative to prevent future injuries from subquent exposures. Patients should be informed of the event to empower them to inform other healthcare providers of their susceptibility to the drug.
Often, the sub-phenotype is difficult to distinguish from laboratory parameters (i.e. ATN vs. AIN) and kidney biopsy information can guide treatment decisions. Several studies have demonstrated the importance of kidney biopsies for classifying the type of injury and establishing the causal drug in the tting of nephrotoxicity. Zaidan and colleagues published a ries of 222 kidney biopsies from HIV i
nfected patients, 59 cas demonstrated tubulopathy or interstitial nephritis with 52.5% attributable to a drug [42]. Tenofovir was identified as the most common culprit of tubular damage in this ries whereas infections and dysimmune syndromes accounted for the majority of interstitial nephritis cas [42]. Xie and colleagues published a ca ries of kidney injury from clindamycin, a previously unrecognized adver event [43]. Biopsy results documented the majority of cas with AIN (75%) and remainder with ATN (25%) [43]. Chu and colleagues demonstrated that only 79.2% of patients with biopsy proven acute tubular necrosis met the clinical criteria for AKI [44]. Most patients had a slower increa in Scr than current KDIGO definitions [44]. Kidney biopsy information in addition to careful consideration of the temporal and causal relationship to drugs can provide a more accurate diagnosis of DIKD. Additionally, DIKD is often caud by multiple drugs and determining causality can be difficult. Even with kidney biopsy data, it may be difficult to determine exact causality for multi-drug injury. Sequential discontinuation of suspected causal drugs and subquent re-challenge may assist in causality asssment.komatsu
Renal Support 肾⽀持
The need for renal support to treat DIKD is low (Table 1). The u of renal replacement therapy for DIKD is two-fold, firstly, dialysis can be utilized to remove the offending drug and minimize ongoing d北京培训机构
amage; additionally, dialysis can be utilized to support renal function to allow recovery. The decision to start renal replacement therapy is a complex one and generally rerved for vere injuries or cas in which the drug toxicity may be mitigated through removal by dialysis (example: vancomycin [45], aminoglycosides [46]).
