Aromatase inhibitors: what is the true cost?

Abstract

The major tenets of systemic anti-cancer therapy should be to improve either or both overall survival and quality of life with acceptable cost–benefit. In this issue of the Journal, Vitry et al. discuss the rising costs associated with the use of aromatase inhibitors (AIs) in the treatment of breast cancer.1 As they point out, this class of drugs is now the preferred choice in Australia despite the lack of evidence of improvement in overall survival and the cost being six times that of the alternative drug, tamoxifen. Tamoxifen, a selective oestrogen receptor modulator, has a proven survival benefit in the adjuvant setting, along with a well-understood side-effect profile that includes hot flushes (common) as well as deep vein thrombosis and uterine cancer at a combined rate of 0.2% per decade among women on treatment for 5 years.2 An overview of randomized trials of tamoxifen has shown an absolute improvement in survival of 9.2% at 15 years after diagnosis compared with no tamoxifen.2 For individual women the magnitude of benefit will vary with the risk of recurrence. The AIs prevent the synthesis of oestrogen from adrenal produced androgen by blocking the aromatase enzyme. They are suitable only for postmenopausal women. Their side effects include hot flushes and arthralgia; the latter can affect 15–25% of women.3 AIs can cause a fall in bone mineral density (BMD) and a subsequent increased risk of fracture during treatment. This increased risk resolves after stopping treatment.3 Patients on an AI should have their BMD measured at baseline and every 2 years while on treatment. A fall in BMD or a low baseline can be treated with an oral or i.v. bisphosphonate, with their inherent toxicities and costs. Adjuvant AI therapy has consistently been shown to be superior to tamoxifen in terms of disease-free survival (delaying recurrence) but has not been shown to improve overall survival.4 Ten-year survival data for 5 years of tamoxifen compared with 5 years of the AI anastrozole, the most mature AI trial to date, have recently been published. The absolute reduction in recurrence with anastrozole compared with tamoxifen was 4.3% at 10 years. There was no improvement in overall survival. The number of deaths after breast cancer recurrence in the anastrozole arm of the trial was reduced. However, the increase in the number of deaths from other cancers and other causes in those without breast cancer recurrence resulted in similar overall survival to tamoxifen.5 Letrozole, another AI compared with tamoxifen in the adjuvant setting, now has 6-year data.6 An earlier interim analysis had shown an improvement in disease-free survival and subsequent unblinding allowed cross-over to the AI arm. Intention-to-treat analysis has not shown a significant improvement in survival. Subgroup analyses excluding patients who were allowed to cross-over are underpowered and post-hoc but do show a small advantage in this group.6 It may be that a true survival advantage exists but is likely now never to be proven. This will be an ongoing problem in oncology as primary outcomes other than overall survival are now frequent and these early results often prompt cross-over in trials. Vitry et al. discuss the cost to the Pharmaceutical Benefits Scheme of the increase in prescribing of AIs. The true total cost of this treatment must include the cost of measuring, preventing and treating the effects on BMD, including the costs borne by patients of calcium and vitamin D supplementation and analgesia for arthralgia. This would seem likely to be more than the cost of managing the side effects of tamoxifen. We know from the chemotherapy published work that breast cancer patients will accept side effects for even very small advantages in survival, but it is less clear whether they accept this for delayed recurrence alone. Certainly, the compliance with any adjuvant hormone therapy is known to be poor. Aromatase inhibitors are not cheap. They offer a choice to women and their oncologists with regard to side-effect profile and have a role in delaying recurrence. They are just one example of new cancer treatments, often expensive, which provide improvements in disease-free survival in the adjuvant setting or progression-free survival in the metastatic setting but have no clear overall survival benefit. As a society, are we prepared to continue to bear the cost of these medications, or should we adopt a more hardline stance and only accept expensive new drugs if there is an improvement in quality of life and/or a clear and meaningful overall survival advantage?