**3.1 Immunotherapy and ductal carcinoma in situ**

With the increasing use of screening mammography, DCIS, the pre-invasive form of breast cancer, has become the most frequently diagnosed breast malignancy. DCIS is a heterogeneous disease both in terms of nuclear grade and expression of cell-surface receptors. DCIS is the non-obligate precursor of invasive ductal carcinoma. Low-grade DCIS may develop into invasive cancer slowly or not all. In contrast, high-grade DCIS almost always develops into invasive disease and often requires more aggressive surgical and adjuvant therapy [48].

Currently, in most clinical practices, DCIS lesions are examined for the over-expression of ER and PR. Interest in the correlation between the molecular biology of DCIS and its clinical aggressiveness has led to staining for other markers. The additional information about potential antigen targets on DCIS is useful to guide development of novel adjuvant vaccines against DCIS. Specifically, for the past 6 years we have routinely stained all of the tumors from our DCIS patients for the over-expression of HER-2. Other biologic markers that are may have prognostic significance include HER 1-4, Ki67, p21, bcl-2, p16 and COX-2, c-myc and survivin [10, 13, 49].

Multiple recent studies have concentrated on determining molecular phenotypes for DCIS similar to those described for invasive breast cancer. In 2010 *Tamimi et al* published a large case series of DCIS and invasive breast tumors that were analyzed using tissue microarray and immunostaining for ER, PR, HER-2, CK 5/6 and EGFR. The authors concluded that the same 5 molecular phenotypes used to describe invasive cancer were all identified among the cases of DCIS. The prevalence of the lesions was not consistent between the DCIS and invasive tumors; the Luminal A phenotype was significantly more frequent among the invasive cancers and the Luminal B and HER-2 molecular phenotypes were more frequent in DCIS. The triple negative (Basal-like) phenotype is very uncommon in DCIS [50]. This is consistent with other studies that show a higher prevalence of HER-2 over expression in DCIS compared to invasive breast cancer [51-53]. Additional work has expanded the traditional molecular profiling to incorporate many more biomarkers that have been found to be biologically relevant to invasive breast cancer, including p53, bcl-2 and P-cadherin. Bcl-2 was found to be one of the most common genes to be up regulated in the welldifferentiated sample of DCIS [53] and has also been reported to be a predictor of good prognosis in invasive breast cancers [54].

DCIS is a particularly attractive vaccine target because the elimination of this disease prevents the subsequent development of invasive breast cancer. In addition, a novel neoadjuvant vaccine would be ideal to reduce size of these lesions prior to surgery. This could theoretically decrease the amount of breast tissue required to obtain clear margins during surgical excision and could also prevent the risk of subsequent recurrence. In general, DCIS patients are often otherwise healthy and are therefore able to mount an immune response to vaccination. The long latency period between the diagnosis of *in situ* disease and the development of invasive breast cancer, as well as the minimal disease burden of DCIS, provides an ideal therapeutic window to administer preventative vaccines. This strategy of treating early disease is applicable to early invasive breast cancer as well as to other cancers with indolent courses or those that are diagnosed in early stages, such as prostate and colon cancers, chronic leukemia and lymphomas [48, 55].
