Alendronate Treatment in Women With Normal to Severely Impaired Renal Function: An Analysis of the Fracture Intervention Trial

Authors


  • Presented, in part, as an abstract at the 27th American Society for Bone and Mineral Research Annual Meeting, Nashville, TN, September 23–27, 2005.

    Dr Hochberg has served as a consultant to Merck & Co., Roche, Proctor & Gamble, Novartis, GlaxoSmithKline, and Amgen. Dr Cauley receives research funding from Eli Lilly & Co., Novartis, Merck & Co., and Pfizer & Co. She also serves as a consultant to Eli Lilly & Co. and Novartis in addition to serving on the Speakers Bureau for Merck & Co. All other authors state that they have no conflicts of interest.

Abstract

To determine if alendronate had differential effects on BMD and fracture by renal function, we performed a secondary data analysis of women participating in the FIT. Alendronate increased BMD and decreased fractures to a similar degree among women with and without reduced renal function. There was no increase in adverse events among women with impaired renal function treated with alendronate. Alendronate is safe and effective among this group of women with reduced renal function.

Introduction: Alendronate is cleared by the kidney and may have sustained effects on bone in subjects with impaired renal function. We hypothesized that, with decreasing renal function, alendronate treatment would result in greater increases in BMD and greater decreases in fractures and that the frequency of adverse events would be increased.

Materials and Methods: We studied women participating in the Fracture Intervention Trial (FIT), a randomized controlled trial of alendronate or placebo (n = 6458). We estimated baseline creatinine clearance (eGFR) using the Cockcroft Gault Formula.

Results: Five hundred eighty-one (9.9%) participants had a severely reduced eGFR (<45 ml/minute). Alendronate increased BMD regardless of eGFR, but women with reduced eGFR had a 5.6% (95% CI: 4.8–6.5) increase in total hip BMD compared with 4.8% (95% CI: 4.6–5.0) among women with normal to moderate renal dysfunction (interaction: p = 0.04). Compared with placebo, alendronate increased spine BMD by 6.6 ± 5.8%, but there was no significant interaction for the increase in spine BMD (interaction: p = 0.75). Treatment with alendronate reduced the risk of clinical fractures to a similar degree in those with (OR: 0.78; 95% CI: 0.51–1.21) and without reduced renal function (OR: 0.80; 95% CI; 0.70–0.93; p for interaction = 0.89). Treatment with alendronate reduced the risk of spine fractures to a similar degree in those with (OR: 0.72; 95% CI: 0.31–1.7) and without reduced renal function (OR: 0.50; 95% CI: 0.32–0.76; p for interaction = 0.44). There were no differences in adverse events by renal function.

Conclusions: Alendronate is safe and effective at increasing BMD and decreasing fractures among this group of women with reduced renal function.

INTRODUCTION

Alendronate, which is almost exclusively cleared by the kidney,(1) may have sustained effects on bone in subjects with impaired renal function. We hypothesized that with decreasing renal function treatment with alendronate would result in greater increases in BMD and greater decreases in fractures. We further hypothesized that the frequency of adverse events would be increased with worsening renal function. To test these hypotheses, we conducted a secondary data analysis among women enrolled in the Fracture Intervention Trial (FIT), a multicenter randomized, placebo-controlled trial of alendronate designed to reduce osteoporotic fractures among women with low BMD at the femoral neck.(2)

MATERIALS AND METHODS

Participants

Women were enrolled in FIT if they were 55–80 years old, were at least 2 years postmenopausal, and had a femoral neck BMD of ≤0.68 g/cm2 (QDR 2000; Hologic, Waltham, MA, USA). Women with serum creatinine >1.27 mg/dl, serum PTH >85 pg/ml in isolation, or serum PTH >65 pg/ml in combination with abnormal serum calcium, alkaline phosphatase, or phosphate were excluded from FIT.

A total of 6459 women were randomly assigned to treatment in FIT: 2027 in the Vertebral Fracture arm and 4432 in the Clinical Fracture Arm. In this analysis, we pooled the data from the clinical fracture and vertebral fracture arms. All women gave informed consent, and the study was approved by the relevant institutional review boards. Recruitment strategies, screening visit protocols, and stratification techniques are described in detail in prior publications.(2)

BMD

BMD was measured at the whole body, femoral neck, total hip, and lumbar spine at study entry and completion-48 months for the clinical fracture arm and 36 months for the vertebral fracture arm-with DXA. We classified women as having osteoporosis using the standard WHO definition: BMD at either the femoral neck, total hip, or lumbar spine of ≤2.5 SD below the mean BMD for young adult women (a T score of ≤ −2.5).(3) Similarly, women who had T scores between −1 and −2.5 were classified as having osteopenia, and women with T scores > −1 were classified as having “normal BMD.”

A detailed description of the BMD quality control procedures used in FIT may be found in earlier publications.(2) Briefly, all densitometry operators participating in FIT received central training to ensure use of standardized procedures for participant positioning and data analysis, and expert densitometry technicians at the coordinating center reviewed a sample of all scans to assure adherence to standardize techniques. BMD quality assurance included on-site and central monitoring of machine performance based on daily phantom scanning, circulation of gold standard phantoms, and software change control.

Fractures

Clinical fractures

Participants were contacted every 3 months and queried about fractures. We considered all fractures diagnosed by a community physician confirmed by written reports of radiographs or other tests over 4.5 years of follow-up. Pathologic fractures (i.e., fractures caused by a malignant neoplasm in the bone, either primary or metastatic), fractures caused by trauma, and facial and skull fractures were excluded. Clinical vertebral fractures were defined as those that came to medical attention and were reported to the clinical centers by the participants. A copy of the spine film obtained from the participant's personal physician was sent to the coordinating center. A blinded study radiologist compared this film with baseline radiography and defined an incident clinical vertebral fracture using a semiquantitative method.(4)

Radiographic vertebral fractures

We also included radiographic vertebral fractures, identified by morphometric and semiquantitative techniques, over a mean follow-up of 3 years for the vertebral fracture arm and 4 years for the clinical fracture arm. The assessment and diagnosis of fractures in FIT has been described previously.(2,5)

Laboratory evaluations

All women who participated in FIT had extensive laboratory testing, the details of which have been described elsewhere.(2,6) Included in these tests were measurements of serum calcium, phosphate, creatinine, alkaline phosphatase, and PTH, which were measured at study entry. Serum creatinine was also measured yearly until study completion. All blood tests were performed by a single, central laboratory using standardized methods, and reference ranges were determined from this laboratory's data (Nichols Laboratory). We estimated glomerular filtration rates (GFRs) from data obtained at study entry using the Cockcroft Gault Formula: {(140 − age)[lean body mass (kg)]/[serum creatinine (mM)]} >1.2.(7) Lean body mass was obtained from whole body DXA. We chose not to use the modification of diet in renal disease index (MDRD) to estimate the GFR, because the MDRD may underestimate the true GFR, particularly in older women with normal GFR.(8)

We defined categories of CrCl based the modified National Kidney Foundation classification of chronic kidney disease (K/DOQI Guidelines 2002).(9) Based on this classification, an estimated GFR (eGFR) of <45 ml/minute is considered severely reduced renal function; an eGFR of 45–59 ml/minute is considered moderately reduced renal function; and an eGFR of ≤60 ml/minute is considered normal renal function.

Adverse event assessment

Adverse events were assessed in FIT every 3 months, over the phone or at clinic visits, using standardized questionnaires.(10) In our analyses, we evaluated serious adverse events defined as those that were life threatening, disabling, or required hospitalization. The specific serious adverse events that we assessed included gastrointestinal diseases (e.g., esophageal, duodenal or gastric ulcers, diverticular disease), cardiac arrhythmias, coronary heart disease, cerebrovascular disease, cancer, and death. The decision to hospitalize FIT participants was independent of the study and was at the discretion of each participant's attending physician. We also evaluated renal adverse events, which included worsening renal function, the development of new onset renal failure, renal cysts, renal colic, urolithiasis, and pyelonephritis. Serious adverse events, but not renal adverse events, were validated.

Statistical analysis

Our analysis focused on whether CrCl based on eGFR modified the effect of alendronate on BMD or fracture risk. To do this, we tested for an interaction effect between eGFR and a treatment variable. We examined the associations between three categories of eGFR (<45, 45–59, and ≤60 ml/minute), alendronate, BMD, and fracture and dichotomized eGFR, comparing women with severely reduced values (<45 ml/minute) with women with moderately reduced to normal renal function. We analyzed BMD and fracture risk using linear and logistic regression, respectively. We repeated these analyses limited to women with osteoporosis.

Figure Figure 1.

Proportion of women with severely reduced (<45 ml/minute), moderately reduced, or normal estimated GFR.

We also examined the frequency of serious adverse events by eGFR and by treatment arm. In all analyses, we considered a two-tailed p < 0.05 to be statistically significant and did not adjust for multiple comparisons. Analyses were performed using STATA version 7 (College Station, TX, USA).

RESULTS

Of the 6458 women that we studied, 581 women (9.9%) had an eGFR of <45 ml/minute, and 2409 (37.3%) had an eGFR between 45 and 59 ml/minute (Fig. 1). Note that we did not find any difference in outcomes among subjects with an eGFR of 45–59 ml/minute and those with an eGFR of ≤60 ml/minute. As a result, these groups were pooled, and the pooled data are presented. Compared with women with an eGFR ≤45 ml/minute, women with an eGFR <45 ml/minute were significantly older, weighed less, had lower BMD at the hip and spine, had a greater prevalence of vertebral fractures, reported more fractures after 45 years of age, had poorer self-rated health, and had a higher serum creatinine at study entry (Table 1).

Compared with placebo-treated women, alendronate increased total hip BMD by 4.9 ± 8.7% over a mean follow-up of 3 years. Women with a reduced eGFR had a 5.6% (95% CI: 4.8–6.5) increase in total hip BMD compared with 4.8% (95% CI: 4.6–5.0) among women with normal to moderate renal dysfunction (interaction: p = 0.04). Compared with placebo, alendronate increased spine BMD by 6.6 ± 5.8%, but in contrast to effects on hip BMD, there was no significant interaction between renal function and the increase in BMD (interaction: p = 0.75). Similarly, there was no significant interaction between renal function and the increase in femoral neck (p = 0.32). Among women with osteoporosis, the annualized percent changes in total hip, femoral neck, and spine BMD were similar among those with a CrCl <45 ml/minute and those with CrCl ≤45 ml/minute, with no evidence of an interaction between renal function and treatment assignment (Table 2).

Table Table 1.. Baseline Characteristics by eGFR*
original image

Among all subjects followed for an average of 4 years, there were 907 clinical fractures, including 76 hip fractures. Women with reduced eGFR were also at an increased risk of any clinical fracture (OR = 1.3; 95% CI: 1.0–1.6) and at increased risk of spine fractures (OR = 2.5; 95% CI: 1.6–3.9) compared with women with an eGFR of ≤45 ml/minute.

Compared with placebo-treated women, alendronate reduced the risk of clinical fractures (OR = 0.8; 95% CI: 0.70–0.9) regardless of renal function. Furthermore, treatment with alendronate reduced the risk of clinical fractures to a similar degree in those with (OR: 0.78; 95% CI: 0.51–1.21) and without reduced renal function (OR: 0.80; 95% CI: 0.70–0.93; p for interaction = 0.89). There were 119 morphometric spine fractures over a maximum period of 4 years of follow-up (3 years in the vertebral fracture arm and 4 years in the clinical fracture arm). Overall, compared with placebo-treated women, alendronate reduced the risk of spine fractures (OR: 0.54; 95% CI: 0.37–0.78). Treatment with alendronate reduced the risk of morphometric spine fractures to a similar degree in those with (OR: 0.72; 95% CI: 0.31–1.7) and without reduced renal function (OR: 0.50; 95% CI: 0.32–0.76; p for interaction = 0.44). Limiting our fracture analyses to women with osteoporosis did not substantially change our results (Table 3).

There was a small significant increase in serum creatinine from baseline to the 3-year follow-up (mean at baseline: 1.05 ± 0.16 versus mean at follow-up: 1.06 ± 0.16; p < 0.00001), but the increase in serum creatinine over the study period was the same in those with and without reduced renal function (mean increase in both groups: 0.01 ± 0.10; p = 0.88) and was the same in both the placebo- and alendronate-treated groups (mean increase: 0.01 ± 0.10; p = 0.99). Of the 6458 women in the study, 6426 reported an adverse event, but there was no difference in the frequency of reported events among women with reduced renal function (576 of 581 women or 99.1%) compared with women without reduced renal function (5850 of 5877 or 99.5%; p = 0.189). More specifically, there was no difference in the frequency of gastrointestinal events (4.5% in women with reduced renal function compared with 5.2% among women without reduced renal function; p = 0.5), cerebrovascular events (2.2% in both groups; p = 0.9), cardiovascular events (2.6% in women with reduced renal function compared with 3.2; p = 0.4), arrhythmias (2.4% in women with reduced renal function compared with 2.1; p = 0.7), malignancies (4.3% in women with reduced renal function compared with 5.0 among women without reduced renal function; p = 0.4), and death (1.6% in women with reduced renal function compared with 1.9; p = 0.5). Furthermore, there was no difference in the frequency of renal adverse events in women with reduced renal function (12 of 581 or 2.1%) compared with those without (137 of 5877 or 2.3%; p = 0.68).

Table Table 2.. Change in BMD With Alendronate by eGFR
original image
Table Table 3.. Fracture Risk With Alendronate by eGFR
original image

DISCUSSION

We found alendronate was equally effective at increasing femoral neck and spine BMD and decreasing spine and nonspine fractures in women with and without reduced eGFR. Conversely, increases in total hip BMD with alendronate did differ by baseline renal function, with slightly greater increases among women with reduced eGFR. Possible explanations for the discordance between eGFR and changes in femoral neck, total hip, and spine BMD include the fact that this was a chance finding. Alternatively, it may be caused by the poorer precision at the femoral neck in combination with unknown changes in spine BMD over time such as degenerative arthritis and vascular calcification.(11,12) Among the 581 women with severely reduced eGFR followed for up to 4 years, there was no differential increase in serum creatinine, nor was there an increase in the incidence of severe or renal related adverse events compared with women with a normal eGFR. Our findings suggest that among osteoporotic women with normal serum creatinine, but reduced eGFR, alendronate is safe and effective.

Alendronate is cleared exclusively by the kidney through glomerular filtration and active tubular secretion.(1,13) We hypothesized that decreasing renal function would be associated with increased skeletal retention of the drug, with greater effects on BMD and fracture reduction. Our analyses did not support this hypothesis. However, we did find that, compared with women with a normal eGFR, women with a reduced eGFR had a 25% increase risk in spine fracture and were at double the risk of clinical fractures, yet the reduction in fracture risk with alendronate was similar in both groups. Previous studies have shown an increased risk of hip fractures among patients with end-stage renal disease,(14,15) and more recently, among women with moderate renal insufficiency (estimated GFR < 60 ml/minute).(16)

It is important to note that women were excluded from participating in FIT if serum creatinine was >1.27 mg/dl, yet 581 or 9.9% of the participants had an eGFR that was severely reduced. This prevalence is similar to data reported by Miller et al.(17); in their pooled analysis of nine risedronate fracture trials, they found that 7% of participants had severe renal impairment (eGFR < 30 ml/minute) despite a normal serum creatinine at study entry. The discordance between serum creatinine and creatinine clearance is not surprising: creatinine production is correlated with muscle mass, and with increasing age there is a decrease in muscle mass that may result in a “normal serum creatinine” in the setting of reduced clearance.(18)

Fracture risk increases with increasing age, and bisphosphonates are effective, even in women ≤80 years of age, at reducing this risk.(19) However, physicians have been limited in the prescription of bisphosphonates for osteoporotic older patients because of concerns of concomitant renal impairment, which is common is this population. For example, data from the Third National Health and Nutrition Examination Survey (NHANES III) report that among women >80 years of age with osteoporosis by BMD criteria, just over one half of these women (54%) have severe renal compromise (defined as an eGFR < 35 ml/minute by Cockcroft-Gault).(20) Furthermore, most of the large randomized trials of oral bisphosphonate therapies excluded women with an elevated serum creatinine. The paucity of data and the safety concerns of using bisphosphonates in subjects with renal impairment mean that a large proportion of patients at risk for osteoporotic fractures are not being treated. Our findings suggest that patients with normal serum creatinine and reduced creatinine clearance have an increase in fractures and that alendronate can be used safely and effectively in this population.

An important question that is yet to be addressed is at what creatinine clearance, if any, would bisphosphonates such as alendronate, be contraindicated. A concern with administering bisphosphonates in patients with end-stage renal failure is that antiresorptive agents may promote the development of adynamic bone disease. It has been hypothesized that adynamic bone disease may be associated with an increase in fracture, although there is no direct evidence to support this hypothesis.(15,21) In our analyses, we did not find evidence that alendronate increased the risk of fractures among women with a reduced eGFR. However, to definitively answer this, clinically relevant question trials of bisphosphonates in patients with severe chronic kidney disease and end-stage renal failure are needed.

Our analysis adds to the small body of literature concerning the use of oral bisphosphonates in patients with impaired renal function.(17,22) Our findings are consistent with a recently published analysis on the safety and efficacy of risedronate by eGFR. The risedronate analysis, which pooled data from nine clinical trials, included >9000 postmenopausal women taking risedronate or placebo for an average of 2 years.(17) The authors reported that risedronate preserved hip and spine BMD and prevented vertebral fractures regardless of creatinine clearance as estimated by Cockcroft-Gault. Furthermore, there was no difference in the incidence of overall adverse events or renal-related events by eGFR, nor was there an increase in serum creatinine over the follow-up. Our safety data are also in keeping with a retrospective review of oral alendronate and risedronate use among men and women with impaired renal function.(23) Data from 181 patients, 21 of whom were men, were included in the analyses, and 31 patients had a reduced clearance (eGFR < 35 ml/minute). The authors reported that there was no difference in the incidence of adverse drug events or changes in eGFR over the 2-year follow-up among patients with and without reduced clearance.

Our study has some limitations. This was a secondary post hoc data analysis, and, although we had data on >6000 women, we had limited power to assess interactions between renal function, alendronate therapy, and fracture outcomes. Our findings are limited to women with a reduced eGFR. As such, we cannot comment on the safety on efficacy of bisphosphonate use in women with end-stage renal disease or those on dialysis. Finally, we did not have a direct measure of GFR but rather estimated it by applying the Cockcroft-Gault formula. However, the Cockcroft-Gault is highly correlated to GFR as measured by creatinine clearance and is a well-accepted indirect measure of GFR.(24)

Considered together, our study and the few other studies of bisphosphonate use among patients with reduced renal function suggest that oral bisphosphonates are effective at increasing BMD and decreasing fracture risk and are not associated with an increase in serum creatinine, reduction in creatinine clearance, or an increase in adverse events. Based on our data, we suggest that alendronate can be prescribed to older women if serum creatinine is in the normal range, regardless of clearance, and that even among subjects with reduced eGFR, but not end-stage renal failure, alendronate is safe and effective.

Acknowledgements

SAJ is supported by a New Investigator Award from the Canadian Institutes of Health Research.

Ancillary