The most common noninvasive lesion of the breast is ductal carcinoma in situ (DCIS). Only a fraction of DCIS lesions will progress to invasive breast tumors, but it is not currently possible to predict which ones. Therefore, although surveillance may be recommended for early stage lesions, treatment of DCIS is typically aggressive, including options like mastectomy, lumpectomy and radiation, all of which have serious systemic side effects. Some patients also undergo endocrine therapy, which can have life-threatening side effects including stroke, blood clots, bone loss and elevated risks of certain cancers. There is an urgent clinical need for minimally invasive therapies that can be selectively targeted to prevent progression of premalignant breast lesions without producing systemic toxicity.

RNA interference (RNAi) has been used to treat various tumor types in rodent models, including xenografts of human mammary tumors in mice, with few adverse side effects. But the development of RNAi-based cancer therapeutics relies on accurate identification of the genes to be targeted for silencing and on optimization of techniques for delivering RNAi effectively.

Results from a recent study now suggest that researchers led by Donald Ingber (Harvard University, Boston, MA) have addressed both challenges in developing a promising new treatment for DCIS. Brock's team applied computational gene network modeling to identify the gene HoxA1 as a driver of early mammary cancer progression in transgenic mice. Next, they delivered HoxA1 RNAi molecules formulated as lipidoid nanoparticles to premalignant lesions in the epithelium of intact mammary glands of mice by direct injection through the nipple. This local route of delivery is relatively noninvasive and circumvents liver uptake and accumulation of nanoparticles, which can complicate systemic delivery.

This RNAi silencing of HoxA1 prevented loss of hormone receptor expression, suppressed cell proliferation and reduced mammary tumor incidence by 75% in mice during a 21-week study (Sci. Transl. Med. 6, 217ra2; 2014). The results show that computational methods can be used to identify oncogene candidates for RNAi-based treatment and that localized gene silencing can reduce tumor incidence in mice. The strategy may be translatable to the prevention of mammary tumor progression in humans, although additional testing will be needed first. Amy Brock, a co-author of the paper who is now at The University of Texas at Austin, said, “Localized delivery of a therapeutic opens up new options for patients and doctors. We see this as a platform technology that can be personalized to individual patients and individual tumor types.”