We read with interest the article by Bernardi et al. [1] published in the April 2012 issue of the World Journal of Surgery. In this retrospective analysis of 579 cases, incidence and risk factors for failure of intra operative localization of non palpable breast lesions by radio-guided occult lesion localization (ROLL) were identified. As this was a retrospective study the authors propose that prospective studies are required to affirm their findings.Recently we performed a prospective randomized controlled trial in which we compared 162 patients who underwent ROLL with 152 patients who underwent wireguided localization (WGL) for non-palpable breast cancers [2]. Of all patients who underwent ROLL, five (3 %) localization failures (i.e., defined as \25 % of the lesion enclosed in the surgical specimen) were encountered [3]. When comparing our prospective results with the results of Bernardi and colleagues, we show a similar proportion of localization failures.Bernardi et al. found that the main risk factors for failure were lesion size \5 mm, radiologist's inexperience, and tumor located in the central subareolar quadrant. Because of the small number of localization failures, it was not possible to determine risk factors for failure in our patients. Similar to the series of Bernardi et al., inexperience of the involved radiologists with the ROLL technique also seemed to play a role in our study, as four out of the five failures occurred within the first year after implementation of ROLL. For this reason we recommend performing ROLL under the supervision of an experienced team in at least the first 10 cases. In addition, we suggest use of a contrast medium mixed with the radiotracer for injection, a technique that has already been performed by others [4]. In this way, mammographic verification of correct placement of the radioactive tracer can be achieved, preventing localization failures resulting from incorrect preoperative localization of the tumor by the radiologist. In our series the majority of our failures potentially could have been avoided.Failure of intraoperative localization is a rare but serious complication of the ROLL procedure, one that should be prevented in every case. Taking the results of the retrospective study of Bernardi et al. and our prospective study together it can be concluded that the chance of failure of ROLL is \5 %, which could be minimized by adequate training of the team and use of contrast. References 1. Bernardi S, Bertozzi S, Londero AP et al (2012) Incidence and risk factors of the intraoperative localization failure of nonpalpable breast lesions by radio-guided occult lesion localization: a retrospective analysis of 579 cases.