• Coping with uncertainty: T1a,bN0M0 HER2-positive breast cancer, do we have a treatment threshold?

      Kelly, C M; Pritchard, K I; Trudeau, M; Andreopoulou, E; Hess, K; Pusztai, L; Department of Medical Oncology, Waterford Regional Hospital, Dunmore Road,, Waterford, Ireland. catherine.kelly@ucd.ie (2012-02-01)
      BACKGROUND: Recent retrospective studies have suggested that patients with T1a,bN0M0 human epidermal growth factor receptor 2 (HER2)-positive breast cancer are at a higher risk for recurrence and might benefit from adjuvant trastuzumab. The absolute benefits associated with treating this subgroup are uncertain. Design: We reviewed recent studies examining the prognostic value of HER2 in patients with node-negative T1a,b HER2-positive breast cancer. We calculated the number needed to treat (NNT) using baseline risk estimates for untreated T1a,bN0M0 breast cancer and the number needed to harm (NNH) using the incidence of cardiac events in each of the adjuvant trastuzumab clinical trials. RESULTS: Several studies were identified, each with limitations inherent to retrospective database analyses: small cohort sizes, lack of systematic HER2 testing in older specimens, variations in the use of adjuvant therapy and definitions of study end points, and lack of information relating to comorbidities. The 5-year disease-free survival in the pre-trastuzumab era ranged from 77% to 95%. Comparisons between small HER2 -positive and small HER2 -negative cancers showed numerically worse outcome for the HER2-positive cohort in some but not all studies. In many instances, the NNH was larger (26-250) than the NNT (13-35); however, in a subset of patients, the NNH was lower (6) than the NNT (13-35). CONCLUSIONS: Better prediction tools to estimate more precisely the risk for death due to comorbid illness versus breast cancer are needed. In some patients, the risks of therapy could outweigh the benefits. Treatment selection for T1a,bN0 HER2-positive cancers remains in the transition area between evidence- and subjective judgment-based medicine.