How much higher is my recurrence risk if I stop after 2 years?

The short version: no trial has ever randomly assigned women to stop anastrozole at 2 years versus continue to 5 years, so the exact number you are asking for does not exist in the evidence.

The closest data comes from the ATAC trial, which compared 5 years of anastrozole to 5 years of tamoxifen in over 9,000 women [1, 2]. In the hormone-receptor-positive group, anastrozole reduced recurrence by about 17% compared to tamoxifen over the full 5 years (hazard ratio 0.83) [1, 2]. But that is a comparison of two different drugs, not a test of stopping early. The DATA trial tested 6 years versus 3 years of anastrozole after initial tamoxifen and found a hazard ratio of 0.79 for disease-free survival, but the result did not reach statistical significance (p=0.066) [13]. That means even the benefit of going from 3 to 6 years was uncertain in that study.

What the evidence does show is that recurrence risk in hormone-receptor-positive breast cancer does not front-load in the first two years. The ATAC trial's 100-month analysis showed the benefit of anastrozole over tamoxifen grew over time, not that it was concentrated early [2]. The annual recurrence rate in years 5-10 was actually higher than in years 0-5 in one ATAC sub-study: 2.80% per year versus 1.82% [9]. That means stopping at 2 years removes protection during a period when the risk of recurrence is still substantial.

The 2025 retrospective Mexican study of 301 patients found that longer therapy duration consistently reduced recurrence, with an odds ratio of 0.18 for therapy beyond 36 months versus less than 36 months [11]. That is a large effect, but it is observational, not randomized, and the confidence intervals are wide.

My call: stopping at 2 years instead of 5 means giving up a proven reduction in recurrence that continues beyond year 2, but the exact percentage increase in your personal risk is unknown because no trial tested this question. The evidence suggests the increase is real and may be substantial, especially if your tumor was high-risk. Confidence: moderate, because the direct trial is missing but the indirect evidence is consistent.