Early And Severe Symptoms Predict Prophylactic Anastrozole Adherence

Symptom factors are associated with discontinuation of a 5-year preventative course of anastrozole in postmenopausal women

medwireNews: A third of women prescribed anastrozole do not adhere to treatment, suggests research published in the Annals of Oncology, with early onset of symptoms and symptom severity both linked to premature discontinuation.

The investigators collated data from 3763 postmenopausal women who participated in the International Breast Cancer Intervention Study (IBIS) II Prevention trial comparing anastrozole and placebo, and 2930 women who were randomly assigned to receive tamoxifen or anastrozole for the Ductal Carcinoma in Situ (DCIS) trial.

Ivana Sestak, from Queen Mary University London in the UK, and co-workers report that adherence to treatment in IBIS II Prevention was comparable in the anastrozole and placebo arms, at 65.7% versus 65.9%, with patients spending an average of 3.9 and 4.0 years on treatment, respectively.

These patterns also occurred in the DCIS trial, which demonstrated similar rates of both adherence to treatment in the anastrozole and tamoxifen arms (66.7 vs 65.8%) and mean time on study (3.99 vs 3.95 years), with the greatest rates of nonadherence occurring in the first 12–18 months of both the IBIS II Prevention and DCIS trials.

At the 6-month checkpoint of the IBIS II Prevention study, women using anastrozole were significantly more likely to experience arthralgia, hot flushes or night sweats, and gynaecological symptoms than controls. Both arthralgia and gynaecological symptoms were significant predictors of nonadherence after 4.5 years (hazard ratios [HR]=0.85 and 0.78, respectively), although the researchers note that the “absolute differences in adherence were small”.

Further analysis indicated that placebo-treated, but not anastrozole-treated, IBIS II Prevention patients reporting arthralgia at 6 months were significantly less adherent to treatment at 4.5 years than those without this side effect (HR=0.81), whereas anastrozole-treated patients were significantly less adherent if they reported gynaecological symptoms at the 6-month checkpoint than if they did not (HR=0.69).

Moreover, most symptoms reported at 6 months were described as mild or moderate in severity, and although nonadherence to treatment was significantly associated with increasing severity of all symptoms except eye diseases and osteoporosis, reports of mild symptoms were not associated with treatment discontinuation.

For the DCIS trial, only prior hormone replacement therapy use predicted nonadherence to treatment in multivariate analysis (odds ratio=0.81).

But women given anastrozole were significantly more likely to report arthralgia at 6 months than tamoxifen users (30.4 vs 20.3%), while the reverse was true for hot flushes or night sweats (40.6 vs 46.7%). And participants in either arm who experienced hot flushes were significantly more likely to adhere to treatment than those who did not (HR=1.18), as were those who were using anastrozole (HR=1.23). There was a trend towards this for tamoxifen users but this did not reach significance.

No other symptoms at 6 months were associated with nonadherence in the DCIS trial, the authors say. They emphasize that adherence was similar between symptom-free patients those with mild symptoms, whereas increasing severity of arthralgia, hot flushes and gynaecological symptoms was again significantly associated with greater nonadherence.

Acknowledging that patient-reported symptoms do not fully explain treatment nonadherence, Ivana Sestak et al conclude: “Identifying factors other than medication induced side-effects that explain non-adherence could help to target future intervention strategies to support medication taking behaviour.

 “Once interventions have been developed, they should be targeted at women within the first 18 months of therapy, as this is when medication cessation is most likely.”


Sestak I, Smith SG, Howell A, et al. Early participant-reported symptoms as predictors of adherence to anastrazole in the International Breast Cancer Intervention Studies II. Ann Oncol; Advance online publication 6 November 2017. https://doi.org/10.1093/annonc/mdx713

medwireNews is an independent medical news service provided by Springer Healthcare. © 2017 Springer Healthcare part of the Springer Nature group