LYON, France, and SHEFFIELD, UK—The price of alendronate dropped by more than two thirds in the UK once the drug became available as a generic, but the change had no effect on judgments of its cost-effectiveness by the National Institute for Health and Clinical Excellence (NICE). The agency's subsequent Final Appraisal Determinations (FADs) on alendronate still recommended only generic alendronate and only for women with osteoporosis as defined by a bone mineral density (BMD) T-score <e;-2.5 SD. The recommendation is expected to dramatically affect prescribing in the National Health Service (NHS) and has aroused the ire of an international team of osteoporosis experts led by John A. Kanis, of the WHO collaborating centre for metabolic bone diseases, University of Sheffield Medical School in the UK.

Kanis et al reported in Bone that the NICE economic appraisal of alendronate for primary and secondary prevention of osteoporotic fracture was based on flawed assumptions and that more reasonable assumptions   showed that alendronate is cost-effective in a much wider setting.¹

Commenting on the Kanis re-analysis, Pierre D. Delmas, MD, of the University of Lyon, France, and Ethel S. Siris, MD, of the Toni Stabile osteoporosis center at Columbia University Medical Center in New York, said, "The conservatism of the recommendations contained in the FADs despite the recent large fall in the price of alendronate is a result of changing some assumptions in the absence of new evidence, and of introducing assumptions that are conceptually flawed. While NICE has a responsibility to ensure the effective use of resources, this should not be achieved through manipulation of economic models in order to reduce the apparent cost-effectiveness of treatment. Moreover, in order to avoid inequities across disease states, the consistency of approach is essential. In this respect, comparison with the recent NICE guidance on statins is particularly relevant, in terms of its use of absolute risk to determine intervention thresholds."²

NICE methodology criticized

Dr. Kanis collaborated on the cost-effectiveness analysis with

• Judith Adams, MD, clinical radiology department at the University of Manchester, UK
• Peter Selby, MD, department of medicine at Manchester Royal Infirmary, UK
• Cyrus Cooper, MD, MRC epidemiology resource centre, University of Southampton, UK
• Danielle Preedy, MD, National Osteoporosis Society in Bath, UK
• Juliet Compston, MD, University of Cambridge School of Clinical Medicine, UK
• Fred Borgström, MD, Karolinska Institute, Stockholm, Sweden, and
• Bengt Jönsson, PhD, Stockholm School of Economics, Sweden

The authors wrote, "The marked price reduction of alendronate, but without consequences on cost-effectiveness, might suggest that the NICE agenda is to seek to use cost-effectiveness modeling with ultra-conservative assumptions as a way of limiting access to medicines."

To re-examine this issue, they used a meta-analysis of the effects of alendronate on fracture risk that was conducted for the NICE appraisal, plus sensitivity analysis examination of other interventions such as raloxifene, strontium ranelate, and ibandronate. Clinical risk factors included low BMI, prior fragility fracture, parental history of hip fracture, long-term oral glucocorticoid use, rheumatoid arthritis (RA), current cigarette smoking, or >e;3 alcoholic drinks per day.

In contrast to the NICE conclusions, Kanis et al found that alendronate was cost-effective for

• primary prevention of fracture in women with osteoporosis, irrespective of age
• secondary prevention in postmenopausal women with prior fragility fracture, without need for BMD test
• primary prevention for postmenopausal patients with parental history of hip fracture, without need for BMD test, and
• primary or secondary fracture prevention in women with strong risk factors for fracture and BMD above threshold for osteoporosis

They commented, "A prior fragility fracture was a sufficiently strong risk factor that treatment was cost-effective even in women without other risk factors in whom BMD was not known."

Treatment recommendations for women with RA or who use glucocorticoids

Kanis et al added, "On the basis of our findings, treatment should be considered where the T-score for BMD at the femoral neck is <e;-1 SD for postmenopausal women with RA or [who are] committed to long-term oral glucocorticoids, and where the T-score for BMD at the femoral neck is <e;-2 SD for women with other secondary causes of osteoporosis, cigarette smokers, or women that drink >e;3 units of alcohol daily."

NICE recommended that alendronate was cost-effective only for secondary prevention of fracture in patients with osteoporosis and a prior fragility fracture. Furthermore, NICE "does not recommend treatment of women <75 years with a prior fragility fracture unless T-score for BMD is <e;-2.5 SD."

Kanis et al concluded that treatment of osteoporosis with alendronate in postmenopausal women "is highly cost-effective" even beginning at 50 years of age. They add, "Our findings that the presence of clinical risk factors and age modulate risk, and therefore cost-effectiveness, reinforce the view that treatment should be directed on the basis of fracture probability rather than on a BMD threshold." Kanis et al found that raloxifene can be cost-effective for both primary prevention and secondary prevention of fractures.

The difference between 10 years and life

In examining possible reasons for the different conclusions reached by the NICE analysis, Kanis et al singled out "changes in the time horizon and assumptions concerning side effects." NICE used a 10-year rather than a lifetime horizon, which had the effect of reducing the benefits and concentrating the "costs" attributable to drug side effects, which typical occur during the first few years of treatment. "The 10-year horizon captures all the costs of treatment (identification of patients and cost of treatment) but loses a component of the benefit. For example, an individual who dies after 9 years is dead for life, and not for 1 year, as would be assumed with a 10-year horizon."

They also criticized NICE for pooling risedronate and alendronate data for secondary prevention but attributing the result to alendronate, reducing the harm attributed to vertebral fractures, reducing the QALY benefit for primary prevention from ₤30,000 to ₤20,000 per year of fracture prevention, and assuming that the beneficial effect of drugs is much less for patients with risk factors (RA or long-term use of oral glucocorticoids) other than age, previous fracture, or low BMD.

In their accompanying editorial Drs. Delmas and Siris highlight the clinical impact of the FADs recommendation.

"The [FADs] have a mandatory status that obliges the commissioners of health care, the Primary Care Trusts (PCTs) to follow the recommendations. In contrast, the clinical guideline [to be issued later to include women who cannot take alendronate, who have osteopenia, or who are on long-term glucocorticoid therapy] does not have this status; [it serves only as an advisory]. In the current financial climate of the NHS, PCTs are likely to implement only mandatory recommendations, with the result that drugs other than generic alendronate would be removed from their formularies and management of glucocorticoid-induced osteoporosis and of women with osteopenia at high fracture risk would be neglected," Drs. Delmas and Siris ended.

References