Journal of Clinical Lipidology
Volume 6, Issue 1 , Pages 19-26, January 2012

Evaluation of the incidence and risk factors for development of fenofibrate-associated nephrotoxicity

  • Rebecca L. Attridge, PharmD, MSc

      Affiliations

    • University of the Incarnate Word, 4301 Broadway CPO#99, San Antonio, TX 78209, USA
    • University of Texas at Austin, Austin, TX, USA
    • University of Texas Health Science Center at San Antonio, San Antonio, TX, USA
    • Corresponding Author InformationCorresponding author.
    • ,
  • William D. Linn, PharmD

      Affiliations

    • University of the Incarnate Word, 4301 Broadway CPO#99, San Antonio, TX 78209, USA
    • University of Texas at Austin, Austin, TX, USA
    • ,
  • Laurajo Ryan, PharmD, MSc

      Affiliations

    • University of Texas at Austin, Austin, TX, USA
    • University of Texas Health Science Center at San Antonio, San Antonio, TX, USA
    • ,
  • Jim Koeller, MSc

      Affiliations

    • University of Texas at Austin, Austin, TX, USA
    • University of Texas Health Science Center at San Antonio, San Antonio, TX, USA
    • ,
  • Christopher R. Frei, PharmD, MSc

      Affiliations

    • University of Texas at Austin, Austin, TX, USA
    • University of Texas Health Science Center at San Antonio, San Antonio, TX, USA

Received 20 May 2011; accepted 30 August 2011. published online 14 September 2011.

Article Outline

Background

Fenofibrate-associated nephrotoxicity has been described in two randomized controlled trials and several observational studies. However, little is known regarding its incidence and the population(s) at risk.

Objective

This study aims to quantify the incidence and identify potential risk factors for development of nephrotoxicity in patients receiving fenofibrate.

Methods

A retrospective, observational study was conducted in the South Texas Veterans Health Care System. Data were collected regarding baseline demographics, concurrent medical conditions, medications, laboratory results, and fenofibrate use.

Results

Within 6 months after initiation of fenofibrate in 428 patients, 115 (27%) experienced an increase in serum creatinine of ≥0.3 mg/dL. Any renal disease (P = .001), chronic kidney disease (P = .01), and diabetes (P = .02) were significantly more prevalent in patients with fenofibrate-associated nephrotoxicity. Patients with nephrotoxicity had significantly greater serum creatinine (1.2 [SD 0.3] vs. 1.1 mg/dL [SD 0.3], P = .0002) and lower estimated glomerular filtration rate (72 [SD 20] vs 81 mL/min/1.73m2 [SD 20], P < .0001) at baseline. These patients also had greater use of calcium channel blockers (P = .0003), furosemide (P = .02), and angiotensin-converting enzyme inhibitors (P = .02). The incidence of nephrotoxicity was significantly greater in patients initiated on high-dose versus those on low-dose fenofibrate (P = .002). In a multivariable regression model, renal disease (P = .02), high-dose fenofibrate (P = .001), and dihydropyridine calcium channel blocker use (P = .02) were determined to be independent predictors of development of increased serum creatinine on fenofibrate.

Conclusion

This observational study suggests fenofibrate-associated nephrotoxicity occurs more frequently than previously reported, particularly in patients with renal disease and in those receiving high-dose fenofibrate or concomitant calcium channel blockers.

Keywords: Fenofibrate, Fenofibrate-associated nephrotoxicity, Fibric aid derivatives, Nephrotoxicity

 

Back to Article Outline

Introduction 

Fenofibrate, which was approved by the Food and Drug Administration in 1998, is a fibric acid derivative indicated for treatment of hypertriglyceridemia and dyslipidemia. Despite being well-tolerated in the majority of patients, several studies have described the development of nephrotoxicity in patients on fenofibrate.1, 2, 3, 4, 5, 6 Two prospective studies have been published evaluating this effect in small samples and both describe an increase in serum creatinine (SCr) in patients who are initiated on fenofibrate.7, 8 However, the incidence, risk factors for development, and mechanism of fenofibrate-associated nephrotoxicity are unknown.

This adverse effect was not fully elucidated at the time of drug approval and is still underrecognized in the clinical setting.2, 9, 10 The hypothesis of this study is that an elevation in SCr after initiation of fenofibrate occurs more commonly than currently appreciated, particularly in patients with a pre-existing renal dysfunction. We performed a single-center, retrospective, observational study to evaluate our hypothesis.

Back to Article Outline

Methods 

The primary aim of this retrospective study was to quantify the incidence of fenofibrate-associated nephrotoxicity, defined as an increase in SCr ≥ 0.3 mg/dL within the first 6 months after the initiation of fenofibrate. Secondary end points included evaluation of the following: 1) medications and comorbidities that predispose a patient to risk of nephrotoxicity; 2) the incidence of fenofibrate-associated nephrotoxicity in patients with and without renal disease or diabetes; 3) the effect of low or high dose on incidence of nephrotoxicity; and 4) the time course for nephrotoxicity. The time course for nephrotoxicity was analyzed by the use of changes in SCr and glomerular filtration rate (GFR) while patients were receiving therapy and the temporal relationship between initiation of therapy and increase in SCr.

Patients were selected if they were initiated on fenofibrate between January 1, 2004, and July 10, 2008. Data were collected on both living and deceased patients by use of the electronic medical record system of the South Texas Veterans Health Care System. Patients were included if they were prescribed fenofibrate therapy for at least 1 month. Patients were excluded for the following: younger than 18 years of age, no baseline SCr value, no follow-up SCr value within six months, and no initial clinic notes (Fig. 1). The study received approval from the University of Texas Health Science Center San Antonio Institutional Review Board and the Veterans Administration Research and Development Committee.

Data collected included baseline patient demographics as well as all concurrent medical conditions and medications. Over-the-counter medication use was not captured unless use was stated in the clinic note at the time fenofibrate initiation. Fenofibrate dose, duration, and adverse effects were collected. Baseline laboratory data and two sets of follow-up laboratory data were obtained from outpatient visits. All charts were reviewed for alternate etiologies of nephrotoxicity, such as initiation of a nonsteroidal anti-inflammatory drug or contrast-induced nephropathy.

Nephrotoxicity was defined as an increase in SCr of ≥0.3 mg/dL in the first 6 months after the initiation of fenofibrate. Because there are no evidence-based definitions of drug-induced nephrotoxicity, this study applied the Acute Kidney Injury Network definition.11, 12, 13 Low-dose fenofibrate was defined as 45, 48, 50, or 96 mg, whereas high-dose fenofibrate was defined as 145, 160, or 290 mg daily. Any renal disease was defined as the presence of chronic kidney disease (CKD), glomerulonephritis, nephropathy, or renal carcinoma. Baseline SCr was calculated as the mean of up to three SCr values from outpatient visits before the initiation of fenofibrate. Follow-up laboratory data were the mean of up to two sets of laboratory data from outpatient visits after fenofibrate initiation. Stage of CKD was determined by the estimated GFR via the four-variable Modification of Diet in Renal Disease equation.14

Data were analyzed by use of the statistical program JMP 7.0® (SAS Corporation, Cary, NC). Comparisons were considered statistically significant if the P value was less than an a priori alpha level of .05. First, all baseline characteristics listed in Table 1 were tested via two-way statistical tests to see whether any of these were associated with increased SCr. Chi-square and the Fisher exact tests were used for analysis of dichotomous and categorical data. Continuous variables were tested for normality by use of the Shapiro-Wilk W Test. Normally distributed variables were reported as mean and standard deviation and compared by the Student’s t test. Non-normally distributed variables were reported as median and interquartile range and were analyzed by the Wilcoxon Rank Sum test.

Table 1. Baseline patient information
Characteristic<0.3 mg/dL increase (n = 313)≥0.3 mg/dL increase (n = 115)Odds ratio and 95% confidence intervalP value
Patient demographics
Age, yr59 (SD 10)60 (SD 9) .4
Male300 (96%)111 (97%) .8
Ethnicity .8
White143 (47%)53 (46%)
Hispanic144 (47%)51 (44%)
Black6 (2%)3 (3%)
Unknown14 (5%)8 (7%)
Weight, kg98 (SD 19)100 (SD 20) .3
Height, cm175 (SD 8)175 (SD 8) .6
BMI, kg/m232 (SD 6)32 (SD 5) .7
Smoker112 (36%)36 (31%) .4
Alcohol use131 (42%)47 (41%) .9
Deceased15 (5%)4 (4%) .8
Medical history
Diabetes168 (54%)76 (66%)1.68 (1.08–2.63).02∗
Duration of diabetes, yr5 (3–10)7 (3–14)−.2
Recent HbA1c, %7.3 (SD 1.7)7.3 (SD 1.9)−.9
Microalbuminuria32 (10%)15 (13%)1.32 (0.68–2.53).4
Macroalbuminuria13 (4%)7 (6%)1.5 (0.58–3.85).4
Albumin excretion, mg/dL13 (8–84)23 (12–118)−.1
Chronic kidney disease15 (5%)14 (12%)2.75 (1.28–5.90).01∗
Diabetic nephropathy9 (3%)4 (4%)1.22 (0.37–4.03).8
Diabetic retinopathy4 (1%)1 (1%)0.68 (0.07–6.13)1.0
Any renal disease20 (6%)20 (17%)3.08 (1.59–5.98).001∗
Any hepatic disease11 (4%)4 (4%)1.0 (0.31–3.17)1.0
Coronary artery disease88 (28%)29 (25%)0.86 (0.53–1.40).5
Heart failure13 (4%)7 (6%)1.5 (0.58–3.85).4
Stroke/transient ischemic attack5 (2%)4 (4%)2.22 (0.59–8.41).3
Dyslipidemia247 (79%)83 (72%)0.69 (0.42–1.13).1
Hypertriglyceridemia47 (15%)22 (19%)1.34 (0.77–2.34).3
Hypertension201 (64%)81 (70%)1.33 (0.84–2.11).2
Concomitant medications
Nonsteroidal anti-inflammatory drugs83 (27%)29 (25%)0.93 (0.57–1.53).9
ACE inhibitor154 (49%)71 (62%)1.67 (1.08–2.58).02∗
Angiotensin receptor blocker18 (6%)9 (8%)1.39 (0.61–3.19).4
Furosemide17 (5%)14 (12%)2.41 (1.15–5.07).02∗
Hydrochlorothiazide57 (18%)28 (24%)1.45 (0.86–2.42).2
Other diuretics9 (3%)9 (8%)2.87 (1.11–7.42).03∗
Calcium channel blocker66 (21%)41 (36%)2.07 (1.30–3.31).003∗

ACE, angiotensin-converting enzyme; BMI, body mass index.

∗Denotes statistically significant result with P < .05.

We collected variables that have been shown to be of importance in other studies or that we know to be important from our clinical experience. We tested all of the variables collected (shown in Table 1) for their association with increased SCr. Then, correlation matrices were constructed to identify and exclude variables that were related. In the case where two were related, we kept the one with the strongest predictive value. All variables found to be significant (P < .05) with two-way tests were included in the final multivariable regression model. Increased SCr was the dependent variable. All other variables were entered simultaneously. Variables that remained significant in the multivariable regression model were considered to be independent predictors of increased SCr.

Back to Article Outline

Results 

Our retrospective study included 428 patients from the veteran population. Within 6 months after initiation of fenofibrate, 115 (27%) patients experienced an increase in SCr of ≥0.3 mg/dL. Both groups were similar with respect to age, gender, ethnicity, body mass index, and substance abuse (Table 1). Diabetes, CKD, and any renal disease were significantly more prevalent in patients who experienced an increase in SCr after initiation of fenofibrate compared with those who did not (Table 1). The presence of hepatic disease, coronary artery disease, heart failure, stroke, dyslipidemia, hypertriglyceridemia, and hypertension was similar in both groups. Duration of diabetes and hemoglobin A1c did not differ between groups.

Patients with fenofibrate-associated nephrotoxicity had statistically significantly greater usage of ACE inhibitors (P = .02), furosemide (P = .02), other diuretics (spironolactone, triameterene, and acetazolamide; P = .03), and calcium channel blockers (P = .003; Table 1). There was no difference between groups in use of nonsteroidal anti-inflammatory drugs, cyclooxygenase-2 inhibitors, angiotensin receptor blockers, or hydrochlorothiazide. We evaluated use of concomitant nephrotoxic medications, including aminoglycosides and amphotericin B, and found no use in patients in either group. In the group that developed an increase in SCr with fenofibrate, one patient was on tacrolimus and one patient on cyclosporine; both patients had therapeutic levels throughout the duration of fenofibrate therapy. Patients in both groups were similar with respect to drug therapy for dyslipidemia and diabetes.

Patients who experienced fenofibrate-associated nephrotoxicity had a significantly higher baseline SCr (P = .0002) and blood urea nitrogen (P = .0001) than those that did not (Table 2). The estimated GFR was lower at baseline in patients who developed nephrotoxicity (P < .0001).

Table 2. Baseline renal function
Characteristic<0.3 mg/dL increase (n = 313)≥0.3 mg/dL increase (n = 115)P value
Baseline SCr, mg/dL1.1 (SD 0.3)1.2 (SD 0.3).0002∗
Baseline eGFR by MDRD, mL/min/1.73 m281 (SD 20)72 (SD 20)<.0001∗
Blood urea nitrogen, BUN, mg/dL16 (SD 6)19 (SD 8).0001∗
BUN:SCr ratio15 (SD 5)16 (SD 4).1

BUN, blood urea nitrogen; SCr, serum creatinine.

∗Denotes statistically significant result with P < .05.

Because of the statistically significant higher prevalence of CKD, any renal disease, and diabetes in the group with nephrotoxicity, we sought to determine the incidence of nephrotoxicity in patients with and without these conditions. We found that patients with CKD, any renal disease, or diabetes had significantly greater rates of nephrotoxicity than patients without these diseases (P < .05 for all comparisons; Fig. 2).

  • View full-size image.
  • Figure 2 

    Incidence of fenofibrate-associated nephrotoxicity. Any renal disease includes patients with CKD (n = 29), obstructive nephropathy (n = 1), IgA nephropathy (n = 1), renal cancer (n = 1), nephrolithiasis (n = 1), membranous glomerulonephritis (n = 2), previous nephrectomy (n = 2), and focal segmental glomerulosclerosis (n = 2). ∗P < .05 for all comparisons.

The incidence of nephrotoxicity was also significantly greater in patients initiated on high-dose versus low-dose fenofibrate (77% vs 24%, P = .002). Patients with renal disease were more likely to be on high-dose fenofibrate therapy, though this finding was not statistically significant (P = .3).

Predictors of increased SCr 

Any renal disease, diabetes, and use of furosemide, dihydropyridine calcium channel blockers, ACE inhibitors, and high dose at initiation were found to be significant predictors of increased SCr on bivariable analysis. Patients with diabetes had a much greater use of ACE inhibitors (64% vs 38%) and furosemide (11% vs 2%) compared with patients without diabetes. Those with any renal disease had greater use of ACE inhibitors (78% vs 50%), furosemide (20% vs 6%), and calcium channel blockers (43% vs 23%) compared with patients without renal disease. In contrast, there was no use of “other diuretics” in patients with renal disease, whereas 5% of patients without renal disease were on other diuretics.

Spironolactone comprised most of the “other diuretics” group, and its use is contraindicated in patients with a creatinine clearance <30 mL/min; therefore, the group “other diuretics” was excluded from the model. All CKD patients are included in the “any renal disease” group, so CKD was also excluded. Calcium channel blocker use was composed mostly of patients on amlodipine or felodipine, with very few patients on verapamil and diltiazem; only dihydropyridine calcium channel blocker use was included in the model.

By using a multivariable regression model, we introduced the following covariates: any renal disease, diabetes, high dose at initiation, dihydropyridine calcium channel blocker, ACE inhibitor, and furosemide. Renal disease (P = .02), high-dose fenofibrate (P = .001), and dihydropyridine calcium channel blockers (P = .02) were significantly associated with increased SCr in our final multivariable regression model (Table 3). This finding suggests these three factors are independent predictors for the development of SCr elevations after fenofibrate initiation.

Table 3. Independent risk factors for fenofibrate-associated nephrotoxicity
CovariateOdds ratio and 95% Confidence intervalP value
Renal disease2.38 (95% CI 1.17–4.83).02∗
Diabetes1.36 (95% CI 0.84–2.21).2
High dose fenofibrate2.28 (95% CI 1.38–3.86).001∗
Furosemide1.76 (95% CI 0.77–3.92).2
Dihydropyridine calcium channel blockers1.84 (95% CI 1.12–3.02).02∗
ACE inhibitor1.30 (95% CI 0.81–2.10).3

ACE, angiotensin-converting enzyme.

∗Denotes statistically significant result with P < .05.

Time course of fenofibrate-associated nephrotoxicity 

The mean peak SCr in patients who experienced nephrotoxicity was 1.7 mg/dL (SD 0.5; Table 4). This reflects a mean change in SCr from baseline to peak of 0.5 mg/dL (SD0.3; 1.2 vs. 1.7 mg/dL, P < .0001). The mean GFR at the time of peak SCr decreased from 72 mL/min/1.73 m2 (SD 20) to 48 mL/min/1.73 m2 (SD 14; P < .0001), representing a 33% decrease in less than a 6-month time period. The follow-up blood urea nitrogen was significantly different between groups. Higher serum levels were observed in the group who developed nephrotoxicity (22 [SD 8] vs. 16 mg/dL [SD 5], P < .0001). The blood urea nitrogen to SCr ratio was not statistically different between groups on follow-up (15 [SD 4] vs 15 mg/dL [SD 4], P = .4).

Table 4. Renal function trend on fenofibrate therapy
Characteristic<0.3 mg/dL increase (n = 313)≥0.3 mg/dL increase (n = 115)P value
Peak SCr on fenofibrate, mg/dL1.1 (SD 0.3)1.7 (SD 0.5)<.0001∗
Minimum eGFR by MDRD, mL/min/1.73 m275 (SD 19)48 (SD 14)<.0001∗
Change in SCr, mg/dL0.1 (SD 0.1)0.5 (SD 0.3)<.0001∗

eGFR, estimated glomerular filtration rate; MDRD, Modification of Diet in Renal Disease; SCr, serum creatinine.

Calculated using ideal body weight or adjusted body weight if body mass index >30 kg/m2.

∗Denotes statistically significant result with P < .05.

The median time to a SCr lab drawn was 63 days (34−115) and the median time to peak SCr was 106 days (70−169). We also evaluated the temporal relationship between initiation of fenofibrate and the increase in SCr (Fig. 3). By approximately 90 days, more than 60% of patients in the nephrotoxicity group had at least one SCr lab drawn. Of those patients, 56% already had an increase in SCr of ≥0.3 mg/dL.

The peak SCr returned to baseline in 70% (n = 80) patients, leaving 30% of patients with a persistent increase in creatinine. Of those patients who did return to baseline, the mean time to return to baseline was 4 months. Fenofibrate was discontinued in 17 (15%) patients at the time of the peak in SCr. However, 10 of these patients did not return to baseline SCr in 12 months after discontinuation. Only 4 of 10 patients returned to baseline at 24 months after discontinuation.

Back to Article Outline

Discussion 

Our retrospective study suggests that approximately one in four patients initiated on fenofibrate will develop an increase in SCr of ≥0.3 mg/dL. This alarmingly high incidence is increased even further to one in two patients with concomitant renal disease or diabetes. Unfortunately, fenofibrate-associated nephrotoxicity is not confined to patients with risk factors because one in five patients who had neither renal disease nor diabetes still experienced this adverse effect. This incidence is even more concerning in light of recent prospective trials in which investigators have failed to show a reduction in cardiovascular events or death in patients receiving fenofibrate.15, 16

Our analysis of the time period for development of nephrotoxicity was limited to the first 6 months after initiation of fenofibrate in part to distinguish fenofibrate-associated nephrotoxicity from progression of CKD or diabetic nephropathy. More than 50% of the patients in our study had diabetes, which is the leading cause of kidney failure in the United States.14 Hypertension will develop in 25% of diabetic patients, with an average rate of decline in GFR of 10 to 12 mL/min/year in untreated patients.17, 18 CKD affects 11% of the United States population and hypertension is the most important modifiable risk in progression of CKD.14, 19 Normal GFR in young adults is between 120 and 130 mL/min/1.73m2 with a predicted decline of 1 mL/min/1.73m2 per year.14, 19 The expected rate of progression of CKD depends on control of modifiable factors, but patients with a GFR less than 60 mL/min/1.73 m2 may progress to kidney failure (GFR <15 mL/min/1.73 m2) within 10 years.14, 20 Patients with CKD and a GFR <60 mL/min/1.73 m2 are considered “fast progressors” if the decline is greater than 4 mL/min/1.73 m2 per year.20 The patients in our study who experienced fenofibrate-associated nephrotoxicity had a decrease in GFR of 24 mL/min/1.73 m2 in a 6-month period, compared with predicted progression of diabetic nephropathy or chronic kidney disease of less than 12 mL/min/1.73 m2 in a year. This finding clearly delineates nephrotoxicity from progression of pre-existing disease.

Dihydropyridine calcium channel blocker use was significantly related to development of fenofibrate-associated nephrotoxicity on bivariable analysis and in the multivariable regression model. Review of animal and human data show conflicting results as to the effects of dihydropyridine calcium channel blockers on renal hemodynamics.21, 22, 23, 24, 25 In general, although dihydropyridine calcium channel blockers decrease afferent arteriole resistance, they also fully impair renal autoregulation, resulting in glomerular hypertension.22, 23, 26, 27, 28 The class has variable effects on progression of proteinuria or renal disease, with no clear benefit.18, 22, 29 Diabetic patients treated with dihydropyridine calcium channel blockers still have histologic progression of renal disease despite adequate blood pressure control.21, 27 The clinical relevance of significantly greater use in the group with nephrotoxicity is unknown.

The use of high-dose fenofibrate also seems to predispose the patient to risk of nephrotoxicity. This effect could be the result of accumulation of fenofibrate or inappropriate use of high-dose therapy in patients with renal impairment. Dosage adjustment of fenofibrate is recommended in patients with a creatinine clearance below 50 mL/min because of accumulation of the active metabolite fenofibric acid.9, 10, 17 In our population, 18 patients with a baseline creatinine clearance less than 50 mL/min received an inappropriately high dose of fenofibrate. Seven of these patients went on to develop nephrotoxicity on therapy, which highlights the need to carefully follow these dosing recommendations.

The overall incidence of fenofibrate-associated nephrotoxicity was found to be much greater than observed in the large, randomized, controlled trials in which fenofibrate use was evaluated, such as the FIELD and ACCORD trials. However, the probable reason for this difference is that patients with a baseline SCr >1.5 mg/dL were excluded from randomization in both the FIELD and the ACCORD trials, thus excluding the population that our study deemed most susceptible to development of fenofibrate-associated nephrotoxicity. Both the FIELD and ACCORD studies did report a statistically significantly greater SCr in patients treated with fenofibrate compared to the control group.15, 16

In four previous studies authors found that elevations in SCr correlated with increased blood urea nitrogen in patients on fenofibrate while two studies found no change.1, 2, 3, 4, 6, 8 In our study we found a parallel increase in SCr and blood urea nitrogen, with no significant change in the blood urea nitrogen to SCr ratio.

Several potential mechanisms of fenofibrate-associated nephrotoxicity have been hypothesized. The most two most commonly published mechanisms revolve around alteration of renal hemodynamics. Fenofibrate down-regulates the expression of the inducible cyclo-oxygenase-2 enzyme, which is essential for maintenance of prostaglandins.30, 31 This decreases dilation of the afferent arteriole and compromises glomerular capillary pressure and perfusion of the kidneys.1, 4 In diabetic rat models, reduction in prostaglandin production was coupled with a small decrease in GFR.30 PPARα agonism may also lead to alterations in renal hemodynamics. Fenofibrate, through binding PPARα, increases expression of cytochrome P450 enzymes which mediate formation of arachidonic acid metabolites that modulate renal function and vascular tone.32, 33, 34, 35 Knowledge of the true mechanism could have important implications for clinical practice. In our study, the SCr returned to baseline in 70% of patients maintained on fenofibrate; 30% of patients had a persistent increase in SCr while on fenofibrate. After discontinuation of fenofibrate, 25 of 35 (71%) returned to baseline SCr at 12 months, and 4 of 10 (40%) returned to baseline SCr at 24 months. This suggests a reversible mechanism of fenofibrate toxicity in most patients; however, the effected persisted in a small minority, thereby suggesting an irreversible mechanism in these patients.

The limitations of our study center on the retrospective, observational design. While this epidemiologic design is useful to evaluate the incidence of an adverse effect, we were limited by the data we were able to capture. Over-the-counter medication use was not accounted for unless explicitly stated in the medical chart. Data on patients who were re-challenged on fenofibrate therapy were not obtained. The prescribers did not routinely provide indications for fenofibrate or document their rationale for using greater doses of fenofibrate. Our population consisted predominantly of patients of white or Hispanic ethnicity, potentially limiting the generalizability of our results to other populations.

Given the retrospective nature of our study, we were only able to evaluate laboratory data from the existing medical records. This could create a selection bias if laboratory tests were ordered more frequently in patients at risk for renal compromise. We sought to minimize the potential bias by excluding patients who did not receive SCr monitoring within the first 6 months after fenofibrate initiation. In addition, the temporal relationship between initiation of fenofibrate and the increase in SCr was strongly affected by the time to follow-up. Most patients returned to clinic for follow-up visits within 3 to 6 months after initiation of fenofibrate, limiting the information we could collect on renal function changes in the first 2 months after the initiation of fenofibrate.

The definition of nephrotoxicity in our study was based on an increase in SCr of ≥0.3 mg/dL within 6 months after initiation of fenofibrate. This definition was based on the Acute Kidney Injury Network guidelines and has been implemented in other studies.13, 36 Data were not collected on other clinical outcomes, such as time to initiation of hemodialysis or changes in GFR requiring renal dose adjustment of other medications. However, with a mean decline in GFR of 24 mL/min/1.73 m2 in a 6-month period, we predict that many patients could require renal dose adjustment.

Back to Article Outline

Conclusion 

Fenofibrate-associated nephrotoxicity is a common adverse effect that merits increased recognition and monitoring. Growing evidence substantiates the prevalence of this effect and the biologic plausibility of these relationships seems high because the independent predictors of nephrotoxicity appear biologically consistent. Our study supports that fenofibrate-associated nephrotoxicity may occur in any patient, with or without predisposing risk factors for renal impairment. However, this risk is greatest in those with pre-existing renal disease, as seen by a doubling of the incidence in this group of patients. Additional studies are needed to validate our findings, further describe the incidence, and identify the populations predisposed to fenofibrate-associated nephrotoxicity. Until further data can be used to examine this effect, caution should be exercised when using fenofibrate in patients with renal disease or diabetes, especially with high-dose therapy. We recommend that all patients receive routine SCr monitoring after initiation of fenofibrate, but particularly in patients with pre-existing renal disease or diabetes.

Back to Article Outline

Financial disclosures 

CRF is supported by the U.S. National Institutes of Health in the form of a NIH/KL2 career development award (3UL1RR025767).

Back to Article Outline

References 

  1. Broeders N, Knoop C, Antoine M, Tielemans C, Abramowicz D. Fibrate-induced increase in blood urea and creatinine: is gemfibrozil the only innocuous agent?. Nephrol Dial Transplant. 2000;15:1993–1999
  2. Angeles C, Lane BP, Miller F, Nord EP. Fenofibrate-associated reversible acute allograft dysfunction in 3 renal transplant recipients: biopsy evidence of tubular toxicity. Am J Kidney Dis. 2004;44:543–550
  3. Lipscombe J, Lewis GF, Cattran D, Bargman JM. Deterioration in renal function associated with fibrate therapy. Clin Nephrol. 2001;55:39–44
  4. McQuade CR, Griego J, Anderson J, Pai AB. Elevated serum creatinine levels associated with fenofibrate therapy. Am J Health Syst Pharm. 2008;65:138–141
  5. Nissen SE, Nicholls SJ, Wolski K, et al. Effects of a potent and selective PPAR-alpha agonist in patients with atherogenic dyslipidemia or hypercholesterolemia: two randomized controlled trials. JAMA. 2007;297:1362–1373
  6. Ritter JL, Nabulsi S. Fenofibrate-Induced elevation in serum creatinine. Pharmacotherapy. 2001;21:1145–1149
  7. Ansquer JC, Dalton RN, Causse E, Crimet D, Malicot KL, Foucher C. Effect of fenofibrate on kidney function: a 6-week randomized crossover trial in healthy people. Am J Kidney Dis. 2008;51:904–913
  8. Hottelart C, El Esper N, Rose F, Achard J-M, Fournier A. Fenofibrate increases creatininemia by increasing metabolic production of creatinine. Nephron. 2002;92:536–541
  9. Antara package insert. Emeryville: Oscient Pharmaceutical; 2007;
  10. Tricor package insert. Chicago: Abbott Pharmaceutical; 2007;
  11. Pannu N. An overview of drug-induced acute kidney injury. Crit Care Med. 2008;36:S216–S23
  12. Schetz M, Dasta J, Goldstein S, Golper T. Drug-induced acute kidney injury. Curr Opin Crit Care. 2005;11:555–565
  13. Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care. 2007;11:R31
  14. National Kidney Foundation: K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis. 2002;39:S1–S266
  15. Keech A, Simes RJ, Barter P, et al. FIELD study investigators Effects of long-term fenofibrate therapy on cardiovascular events in 9795 people with type 2 diabetes mellitus (the FIELD study): randomised controlled trial. Lancet. 2005;366:1849–1861
  16. The ACCORD Study Group. Effects of combination lipid therapy in type 2 diabetes mellitus. N Engl J Med. 2010;362:1563–1574
  17. Keating GM, Croom KF. Fenofibrate: a review of its use in primary dyslipidaemia, the metabolic syndrome and type 2 diabetes mellitus. Drugs. 2007;67:121–153
  18. Tarif N, Bakris GL. Preservation of renal function: the spectrum of effects by calcium-channel blockers. Nephrol Dial Transplant. 1997;12:2244–2250
  19. Bailie GR, Uhlig K, Levey AS. Clinical practice guidelines in nephrology: evaluation, classification, and stratification of chronic kidney disease. Pharmacotherapy. 2005;25:491–502
  20. Stevens LA, Greene T, Levey AS. Surrogate end points for clinical trials of kidney disease progression. Clin J Am Soc Nephrol. 2006;1:874–884
  21. Abbott KC, Bakris G. Renal effects of antihypertensive medications: an overview. J Clin Pharmacol. 1993;33:392–399
  22. Bakris GL. The effects of calcium antagonists on renal hemodynamics, urinary protein excretion, and glomerular morphology in diabetic states. J Am Soc Nephrol. 1991;2:S21–S29
  23. Griffin KA, Picken M, Bakris GL, Bidani AK. Comparative effects of selective T- and L-type calcium channel blockers in the remnant kidney model. Hypertension. 2001;37:1268–1272
  24. Hayashi K, Wakino S, Sugano N, Ozawa Y, Homma K. Ca2+ channel subtypes and pharmacology in the kidney. Circ Res. 2007;100:342–353
  25. Robles NR. Calcium antagonists and renal failure progression. Ren Fail. 2008;30:247–255
  26. Bakris GL, Griffin KA, Picken MM, Bidani AK. Combined effects of an angiotensin converting enzyme inhibitor and a calcium antagonist on renal injury. J Hypertens. 1997;15:1181–1185
  27. Gashti CN, Bakris GL. The role of calcium antagonists in chronic kidney disease. Curr Opin Nephrol Hypertens. 2004;13:155–161
  28. Sarafidis PA, Khosla N, Bakris GL. Antihypertensive therapy in the presence of proteinuria. Am J Kidney Dis. 2007;49:12–26
  29. Nathan S, Pepine CJ, Bakris GL. Calcium antagonists: effects on cardio-renal risk in hypertensive patients. Hypertension. 2005;46:637–642
  30. Chen YJ, Quilley J. Fenofibrate treatment of diabetic rats reduces nitrosative stress, renal cyclooxygenase-2 expression, and enhanced renal prostaglandin release. J Pharmacol Exp Ther. 2008;324:658–663
  31. Wilson MW, Lay LT, Chow CK, Tai HH, Robertson LW, Glauert HP. Altered hepatic eicosanoid concentrations in rats treated with the peroxisome proliferators ciprofibrate and perfluorodecanoic acid. Arch Toxicol. 1995;69:491–497
  32. Varghese Z, Moorhead JF, Ruan XZ. The PPARalpha ligand fenofibrate: meeting multiple targets in diabetic nephropathy. Kidney Int. 2006;69:1490–1391
  33. Zhao X, Quigley JE, Yuan J, Wang MH, Zhou Y, Imig JD. PPAR-alpha activator fenofibrate increases renal CYP-derived eicosanoid synthesis and improves endothelial dilator function in obese Zucker rats. Am J Physiol Heart Circ Physiol. 2006;290:H2187–H2195
  34. Muller DN, Theuer J, Shagdarsuren E, et al. A peroxisome proliferator-activated receptor-alpha activator induces renal CYP2C23 activity and protects from angiotensin II-induced renal injury. Am J Pathol. 2004;164:521–532
  35. Alonso-Galicia M, Frohlich B, Roman RJ. Induction of P450 activity improves pressure-natriuresis in Dahl S rats. Hypertension. 1998;31:232–236
  36. Felker GM, Lee KL, Bull DA, Redfield MM, Stevenson LW. Diuretic strategies in patients with acute decompensated heart failure. N Engl J Med. 2011;364:797–805

PII: S1933-2874(11)00708-2

doi:10.1016/j.jacl.2011.08.008

Journal of Clinical Lipidology
Volume 6, Issue 1 , Pages 19-26, January 2012

Article Tools

* Image Usage & Resolution