GUEST EDITORIAL INVITED EDITORIAL

perspective A review of the randomized clinical trials of adjuvant trastuzumab in human epidermal growth factor receptor (HER2)–positive breast cancer demonstrates the great benefit of this monoclonal antibody whose potential cardiotoxicity is a fascinating problem for the cardiologist 1. The direct cardiotoxicity of trastuzumab, a relatively new cardiomyopathy, is actually quite limited in the absence of prior anthracycline exposure. Concurrent blockage of the HER2 cardiac receptor is believed to make repair of myocardial damage more difficult 2,3, but this effect varies substantially from patient to patient. It is unpredictable, it does not appear to be dosedependent, and it is reversible. About 80 % of patients receiving trastuzumab as adjuvant treatment will maintain a normal left ventricular ejection fraction (LVEF), and therefore questions naturally arise about how otherwise to monitor for potential cardiotoxicity. Cardiac injury can occur in the absence of systolic dysfunction. The LVEF measurement can vary from centre to centre (different methods could be used) and even throughout the day by up to 10 % in individual patients—a level of fall that could result in treatment being inappropriately suspended. To complicate matters further, the definition of cardiotoxicity as measured by LVEF can vary substantially among clinical studies, as can stoppage rules 4. Fortunately, the risk of severe cardiotoxicity remains rare (0.6%–4%), and if such a complication occurs, cessation of trastuzumab therapy can often be followed by contractile improvement given conventional treatment for heart failure. The demonstrable recovery of LVEF can even allow for reintroduction of trastuzumab if the cancer prognosis demands it. The management of an asymptomatic drop of LVEF is difficult to define. In 14.2 % of patients treated with trastuzumab (as in the B-31 N 9831 trial), an asymptomatic LVEF drop triggered stoppage of treatment. In the HERA (Herceptin a Adjuvant) trial, in which trastu-zumab was given sequentially, and in the subgroup without anthracycline exposure of the Breas