Renal function changes and NHS resource use in breast cancer patients with metastatic bone disease treated with IV zoledronic acid or oral ibandronic acid: a four-centre non-interventional study

Abstract

Aims: To describe renal function monitoring practice in patients with metastatic bone disease (MBD) treated with IV zoledronic acid (ZA) and oral ibandronic acid (IA), the management pathways and NHS hospital resources used.

Methods: Medical records of 189 patients; IA (91), ZA (98) with primary breast cancer and MBD were reviewed, and data collected on renal monitoring and hospital visits during bisphosphonate therapy. Time and motion review of resources to administer the bisphosphonates was also conducted.

Results: Only 30% of patients given ZA and no patient given IA had baseline creatinine clearance (CrCl) recorded. Calculated baseline CrCl suggested impaired renal function in 33% ZA and 29% IA patients. Dose reductions were not made correctly in 29 ZA and 2 IA patients whose monitoring suggested it. ZA patients made more clinic and day care attendances than IA-treated patients, at twice the cost. Staff activity and patient time per visit was higher with ZA than IA.

Conclusion: Although limited by retrospective design, these results demonstrate that in many patients, CrCl is not calculated before or during treatment with bisphosphonates. Renal function deteriorated in many patients during therapy. In view of these effects, practice should be reviewed to ensure appropriate dosing.

Key words::

• biphosphonates
• breast neoplasms
• health resources
• ibandronic acid
• renal insufficiency
• zoledronic acid

Introduction

Breast cancer, which is the leading cause of cancer death in Western women¹, metastasises most commonly to bone, affecting 65–75% of women with advanced disease². In addition to bone pain, such patients are at a high risk of skeletal events, with associated morbidity. The average survival time after diagnosis of metastatic disease is around 2.5 years and hence effective palliative treatment is essential, to allow patients to live as normal a life as possible³. The main goals of bisphosphonate therapy in patients with bone metastases are to relieve pain, to prevent skeletal events and to minimise disability⁴. The two most widely used agents in UK hospital practice, the amino-bisphosphonates zoledronic acid (Zometa, ZA, Novartis Pharmaceuticals) and ibandronate (Bondronat, IA, Roche Products Ltd) have benefits over earlier agents in terms of potency and efficacy^5,6.

Although there are some class effects, the safety and tolerability profiles of different bisphosphonates vary. The long-term renal safety of bisphosphonates is particularly important because of the renal dysfunction experienced by many patients due to their underlying disease or other treatment^7,8, IA and ZA differ significantly with respect to renal safety and renal monitoring requirements during therapy^9,10. The literature suggests renal function deterioration, associated with ZA use has been seen in 8.8–15.2% of ZA treated patients¹¹. There are also occasional reports of acute renal failure with ZA that resulted in the need for renal dialysis or death in some patients¹². In order to reduce the risk of renal damage, serum creatinine and creatinine clearance should be determined on initiation of IV ZA therapy and prior to each subsequent dose. The dose of ZA should be reduced or treatment withheld if renal function deteriorates and use of ZA is not recommended in patients with a CrCl < 30 ml/min¹³.

In contrast, in studies of the safety of both IV and oral IA over 4 years, there was no reported deterioration in renal function^14–16. For IA no dose adjustment is required in mild-to-moderate renal impairment, but IV and oral dose-reduction is recommended in patients with severe renal impairment, i.e. CrCl below 30 ml/min. Renal monitoring during treatment is recommended only according to clinical assessment of the individual patient. There is no evidence of a reduction in tolerability associated with an increase in exposure in patients with renal impairment¹⁷.

This study set out to describe the use of these bisphosphonates in patients with breast cancer and metastatic bone disease (MBD) in routine UK clinical practice, to show whether the required monitoring and dose reductions are routinely adhered to, and what are the consequences of ‘normal clinical practice’ for patients, particularly in terms of renal function. In addition, there is value in quantifying the NHS secondary care resources associated with use of bisphosphonates in the management of metastatic bone disease, in order to inform planning of NHS service delivery and this was a secondary aim of the study.

Methods

This observational study was conducted in four NHS secondary care trusts, with NHS research ethics committee (REC) and local management approval. Data to describe patient management pathways, renal function and adverse effects associated with the use of the two bisphosphonates were collected from patient medical notes and hospital computer systems as appropriate. All patients who received their first dose of IA or ZA after March 2005 were included, if they were female and aged ≥18 years, had a documented diagnosis of breast cancer with metastatic bone disease, and if details were available for a minimum of 3 months of treatment. Patients were excluded if breast cancer was not the primary tumour site or if they were enrolled in the BISMARK trial. Data were collected for the total duration of therapy, from the first dose of ZA or IA to death, change of bisphosphonate or first treatment for hypercalcaemia of malignancy. No data were collected on the resources required in managing adverse events associated with either drug.

Data on staff and resources used to dispense, prepare and administer ZA and IA in the hospital were collected, with staff and patient consent, via direct observation by a research nurse. The grade of healthcare professional and time taken for each step of drug preparation and administration in the day care unit for IV ZA and in the pharmacy for dispensing IA were recorded anonymously. Patient time associated with receiving each drug and the NHS equipment and consumables used were also recorded.

The direct hospital costs associated with provision of each bisphosphonate were estimated by application of NHS reference costs¹⁸ to the mean quantities of bisphosphonate supplied and the mean staff time required to dispense and administer it. The laboratory costs associated with monitoring renal function and the primary care cost of any bisphosphonate which may have been supplied by GPs (family doctors) were not included in the estimation.

Results

Data were collected from 189 patients (98 ZA, 91 IA). The median age at diagnosis was 60 years in both groups, but the median time from diagnosis to start of IA therapy was longer (13 months) than for ZA therapy (8 months) (Table 1). More of the patients given IA had received prior bisphosphonate therapy (45 vs. 23%), which was most commonly ZA (in 40 patients). More of the ZA patients were receiving concurrent chemotherapy (34 vs. 12%).

Table 1. Demographics and treatment pathways.

	Zoledronic acid N=98	Ibandronic acid N=91
Mean time, diagnosis of metastatic bone disease to start of ZA or IA therapy (months)	8 (range 0–89)	13 (range 0–79)
Patients having previous bisphosphonate therapy	23 (23%)	41 (45%)
Patients on concurrent chemotherapy	33 (34%)	11 (12%)

Differences between groups were not statistically significant except for number of patients on concurrent chemotherapy (p< 0.001).

Table 2 shows the results for renal function measurements found in the patients’ notes. Only 30% of patients given ZA and no patient given IA had a baseline CrCl (within 1 month before start of therapy) recorded in their medical notes. Serum creatinine (SCr) was recorded for 90% of ZA patients and 58% of IA patients and the researcher was able to calculate baseline CrCl, using the Cockroft–Gault formula, for 80 (82%) of ZA patients but for only 14 (15%) of IA patients (the remaining patients’ notes lacked a recently recorded body weight to allow the calculation).

Table 2. Renal monitoring and dosing of bisphosphonates.

	Zoledronic acid (N=98) No. of patients (%)		Ibandronic acid (N=91) No. of patients (%)
	SCr	CrCl	SCr	CrCl
Baseline value recorded	88 (89.8%)	29(29.6%)	53 (58.2%)	0
Baseline value could be calculated	—	80 (81.6%)	—	14 (15.4%)
Renal function at baseline	n=88	n=80	n=53	n=14
Normal*	81 (92.0%)	54 (67.5%)	39 (73.6%)	10 (71.4%)
Impaired	7 (8%)		14 (26.4%)
Mild impairment	—	12 (15.0%)	—	1 (7.1%)
Moderate impairment	—	12 (15.0%)	—	2 (14.3%)
Severe impairment	—	2 (2.5%)	—	1 (7.1%)
Renal function during treatment (measured/calculable at least once)	n=94	n=86	n=63	n=13
Normal* throughout	70 (74.5%)	52 (60.5%)	42 (66.7%)	8 (61.5%)
Impaired	24 (25.5%)		21 (33.3%)
Mild impairment	—	10 (11.6%)	—	3 (23.1%)
Moderate impairment	—	22 (25.6%)	—	0
Severe impairment	—	2 (2.3%)	—	2 (15.4%)
Patients requiring dose reductions		34 (39.5%)		2 (15.4%)
Patients who received any excessive doses		29/86 (33.7%)		NK

*Normal for serum creatinine (SCr), females = 53–98 μmol/L (0.6–1.1 mg/dL)¹⁹, creatinine clearance (CrCl) > 60 ml/min.

Impaired (any degree) = SCr > 98 μmol/L, CrCl ≤60 ml/min.

Mild impairment = CrCl 50–60 ml/min.

Moderate impairment = CrCl 30–49 ml/min.

Severe impairment = CrCl < 30 ml/min.

The data show baseline CrCl ≤60 ml/min (i.e., impaired renal function) in 26/80 (33%) ZA patients and in 4/14 (29%) IA patients. However, if SCr was used as the measure of renal function (normal reference range for females = 0.6–1.1 mg/dL (53–98 µmol/L)¹⁹, baseline renal function was impaired in 7/88 (8%) ZA patients and 14/53 (26%) IA patients.

During treatment, at least one SCr was recorded for 94 (96%) ZA patients and 63 (69%) IA patients and CrCl could be calculated in 86 and 13 of ZA and IA patients, respectively. In these patients, the results showed that 40% of ZA and 38% of IA-treated patients had CrCl ≤ 60 ml/min at some time during treatment, and SCr was raised in 26% and 33% of ZA and IA patients, respectively during treatment. The Summary of Product Characteristics (SPC) would recommend that ZA doses were reduced in all these patients, but this was not done correctly in 29 patients. For IA, the SPC would recommend dose reductions in the two patients with severe renal impairment, but this was not recorded.

The adverse events found in the medical records included identical numbers of reported GI symptoms, i.e. nausea, vomiting, diarrhoea and abdominal pain, with ZA and IA (n=12 in each group). Flu-like symptoms were reported only for ZA patients (n=4), as were problems with the IV line in 28 patients on 55 occasions, predominantly due to problems inserting the cannula.

Resource usage and the associated costs are shown in Tables 3 and 4. Mean numbers of inpatient, A&E and radiotherapy visits were low and did not differ between groups. Patients given ZA had more clinic visits and more day care attendances than those given IA. The median quantity of drug provided by the hospital per patient was 7.5 ZA infusions, each requiring a day care unit attendance, or 29 IA tablets, a single dispensing. This in turn cost the hospital more than twice as much to treat patients with ZA as with IA (Table 3).

Table 3. Number and cost of hospital visits and bisphosphonate supply from hospital.

Mean hospital resources	Zoledronic acid mean	Ibandronic acid mean	Unit costs (Department of Health Reference Costs 2006/07)	Zoledronic acid cost (UK£)	Ibandronic acid cost (UK£)
Clinic visits	9.62	6.06	107.00	1,026.25	646.57
Day care attendances	11.96	4.96	146.00	1,748.05	724.27
Total	21.58	11.02		2,774.30	1,370.84

Table 4. Mean hospital staff time and hospital cost per administration/supply of drug.

Staff type	Mean staff activity time (minutes)		Zoledronic acid cost (UK£)	Ibandronic acid cost (UK£)
	Zoledronic acid (n=19)	Ibandronic acid (n=10)
Pharmacist	0.13	4.30	0.11	3.69
Clerk/tech	0.00	1.72	0.00	0.81
Nurse	27.82	0.00	19.06	0.00
Total time	27.95	6.02	19.17	4.50

Table 4 shows that dispensing and administration of ZA took an average of 28 minutes of staff time, while IA took 6 minutes. Applying standard NHS reference costs to the detailed data, broken down by salary band of staff undertaking drug preparation and administration, this translated to a cost of £19.17 per ZA administration and £4.50 per IA dispensing in the secondary care environment.

Nineteen patients spent a median of 63 minutes in the unit to receive ZA, of which 37 minutes were in the infusion chair. When IA was dispensed for ten patients at the pharmacy, this took a median of 25 minutes.

Discussion

This study describes the patient monitoring and secondary care resources used during treatment of 189 patients with metastatic bone disease due to breast cancer with ZA or IA in four English NHS secondary care Trusts. Data from all eligible patients treated within the study period were collected, ensuring that the study results are representative of routine clinical practice at these sites. The scope of the study was limited by access to only the hospital and not the primary care records of the patients included and the estimates of cost of treatment are limited to the hospital costs, in order to inform secondary care budgeting. However it is likely that continuing supply of bisphosphonate was made in primary care, contributing to the total cost of therapy. The retrospective design did not allow for collection of data on the management of adverse events. This limits the estimation of costs to those directly attributable to the hospital administration or supply of the bisphosphonates studied.

The differences in demographics between the groups, with a longer time between diagnosis of bone disease and start of IA therapy, more patients in the IA group receiving previous bisphosphonate treatment and fewer IA patients receiving concurrent chemotherapy (Table 1), suggest that IA was used later in the course of the disease than ZA. The IA patients might therefore be expected to have a higher incidence of renal impairment at baseline than those given ZA and this is reflected in the study findings (Table 2). The course of bisphosphonate treatment in secondary care was also considerably shorter in IA (29 daily tablets) than in ZA-treated patients (7.5 monthly infusions) because in general, only the first prescription of IA was supplied by the hospital, with subsequent prescribing in primary care.

Although the prescribing information for both drugs recommends dosing according to CrCl level, Table 2 shows that CrCl was recorded for only 30% of ZA patients and none of the IA patients at the start of treatment. In addition, as only a few IA patients were receiving concurrent chemotherapy, not all had SCr and body weight data recorded, to allow the calculation of CrCl by the researcher. The data therefore give a very limited record of renal function status and its change during therapy in the IA patients. In 2/13 IA patients the researcher-calculated CrCl showed that renal function was severely impaired and the dose should have been reduced to one tablet weekly, but there was no record of this occurring.

The more complete data for ZA show that 33% of patients starting ZA had impaired renal function (CrCl ≤ 60 ml/min) at baseline and 40% had impaired renal function, as measured by CrCl, at some time during treatment. Thus more than a third of the ZA patients should, according to the recommendations in the prescribing information, have had doses reduced or withheld due to impaired renal function. In only five of these patients were all doses appropriately reduced or withheld and 29 patients (a third of those for whom records were found) received at least one excessive dose.

Elevations of SCr are another, less complete indication of renal impairment. In this audit 8% of ZA and 26% of IA patients showed elevated SCr at baseline and a quarter to a third of patients had elevated SCr during their ZA or IA therapy. The difference in elevated SCr between the two groups at baseline may reflect the difference in their disease status and possibly also the larger number of IA patients given prior bisphosphonate therapy. A German audit, covering 1,897 patients with breast cancer and metastatic bone disease showed that 8% of bisphosphonate naive patients had SCr > 1.2 mg/dL, while 11% of those receiving prior IA, 16% of those receiving prior pamidronate and 26% of those given prior ZA had SCr above this level²⁰.

Under phase III clinical trial conditions renal toxicity is an infrequent, but potentially very serious, side-effect associated with the administration of intravenous bisphosphonates. The acute renal failure associated with IV bisphosphonate administration may be clinically reversible, but varying degrees of irreversible impairment may persist and eventually lead to chronic renal failure²¹. High doses accompanied by high molar concentrations of some bisphosphonates have been shown to overload the renal elimination mechanism and the retained drug can damage renal cells²². This is more likely to occur in renally impaired patients, where drugs are cleared more slowly. The recommended dose reductions for bisphosphonates seek to prevent retention of drug within the kidneys with its resultant potential for renal damage. Widespread non-adherence to recommended doses, while not damaging all patients, puts all at risk.

The resource data (Tables 3–4) show that administration of ZA used more hospital resources, in terms of clinic visits, nursing time and IV chair occupancy than administration of IA. This results in less than half the mean hospital cost per patient for IA (£1,371) than ZA (£2,774) treatment in hospital. If the laboratory costs of renal monitoring had also been taken into consideration, this differential would have been even greater, as the study showed that renal monitoring was undertaken in many more patients receiving treatment with ZA (96%) than those receiving IA (69%). Each dispensing of IA was also associated with a lower cost than the administration of ZA (£4.50 vs. £19.17). Although it is intuitively obvious that dispensing of an oral product will require fewer resources than the preparation and administration of an IV infusion, empirical data to quantify such resource use is increasingly required by decision makers to inform guidance on preferred options for NHS treatment.

This study is limited to the hospital costs of treatment with each bisphosphonate, in order to inform secondary care budgeting. It does not include the costs associated with the continuing supply of bisphosphonates by GPs to patients discharged from hospital. There are no previously published studies of sufficient size to quantify the resource implications of administering these two drugs; an evidence gap which this study has attempted to address within the constraints already described.

Conclusion

This study demonstrates that many breast cancer patients with metastatic bone disease have impaired renal function, either before or during bisphosphonate therapy. It is well accepted that bisphosphonates are associated with a degree of renal damage, which may manifest as renal dysfunction. However, in spite of this and recommendations for renal monitoring in the SPC, it was not routine practice in four large NHS secondary care centres for renal function to be recorded before each bisphosphonate dose. The data show that 33% of the 98 ZA patients had some degree of renal impairment at the start of therapy and 40% showed renal impairment during therapy. These patients should have received a dose reduction, but although doses were reduced appropriately in five patients, 29 patients (a third of those with records) received at least one excessive dose and 22% of all the doses given to ZA patients were excessive according to the SPC. This four-centre study reveals significant short-comings in the monitoring of renal function in patients receiving bisphosphonate therapy. Current clinical practice should be reviewed in the light of these findings.

For all patients receiving IA, the data suggest renal function monitoring was not undertaken. Of the 13 patients for whom a CrCl could be calculated, five had some degree of renal impairment during treatment, which was severe enough in two patients to warrant dose reduction. Although the renal data in patients receiving IA was incomplete, the study does suggest that renal function should be monitored in patients receiving IA, as dose reduction may be necessary. Reliable systems for primary and secondary care to share renal function results should be in place to ensure that if a dose reduction of IA is needed, the prescriber is made aware of this.

The recording of staff activity and patient waiting times confirm that there are considerable cost savings for the hospital sector associated with each administration of oral IA compared with IV ZA. The oral formulation also removes the requirement for infusion chairs, thereby increasing chemotherapy unit capacity, and freeing patients from a number of hospital visits and long waiting times.

Transparency

Declaration of funding: This study was funded by an educational grant from Roche Products Ltd covering the cost of protocol drafting, obtaining approvals, data collection and analysis and manuscript preparation.

Declaration of financial/other relationships:S.H. and R.J.G. have disclosed that they have no relevant financial relationships; T.H. has disclosed that he has received honoraria from Roche and Novartis in the past (unrelated to this study); F.P. has disclosed that she is an employee of pH Associates, which received funding from Roche to complete this study. E.H. has disclosed that she is an employee of Roche Products Ltd.

The JME peer reviewers 1 and 2 have not received an honorarium for their review work on this manuscript. Both have disclosed that they have no relevant financial relationships.

AcknowledgmentsThe authors are grateful to Dr C Lowdell, Charing Cross Hospital and Dr O Al Salihi, Hammersmith Hospital for contributing data to this study.