Abstract
The incidence of ductal carcinoma in situ (DCIS) has increased dramatically with the widespread adoption of mammographic screening. It is estimated that as of 2005, half a million women were living with DCIS; this figure is expected to increase to one million by 2020 [1,2]. The management of DCIS remains a challenge for clinicians owing to the lack of validated biomarkers to predict outcomes. Despite the marked heterogeneity in tumor grade, histologic subtypes, hormone receptor status and clinical presentations, most patients diagnosed with DCIS are managed with a ‘one-size-fits-all’ approach: breast conservation with resection of the involved breast segment followed by adjuvant whole breast radiation or total mastectomy. While these strategies have proven success in controlling local disease and preventing invasive carcinoma, it is likely that only a subset of patients benefit from aggressive local therapy. Identification of this subset has been elusive and standard histopathologic factors have failed. The Van Nuys Prognostic Index (VNPI) based on histopathologic indictors (e.g., high nuclear grade, necrosis, margin width, patient age and size) has been used clinically for predicting local recurrence [3]. However, the VNPI requires intensive pathologic substaging of the extent of DCIS in the resected specimen, which is not uniformly practical across all centers. Recently, invasive carcinomas have been classified into distinct molecular subtypes (luminal A/B, HER2-like, basal-like and normal-like), which confer prognostic clinical significance [4]. To date, few studies have attempted to classify DCIS into similar molecular-based subtypes [5–7]. There is emerging evidence on limited data to suggest cDNA microarray gene-expression profiles can segregate DCIS into similar distinct molecular subtypes as invasive carcinoma [8]; however, a significant proportion of patients with DCIS show tumor heterogeneity [5,8,9], making it difficult to stratify DCIS cases into a single subtype and thus diminishing the prognostic significance of these molecular subtypes, compared with invasive carcinoma. DCIS, as a preinvasive malignancy, is associated with long-term survival rates of 98–99%. Overall survival as a primary end point of treatment is less relevant than invasive carcinoma recurrence-free survival. Identification of a subset of patients at high risk for subsequent invasive carcinoma would permit optimal individualized therapy.
Method
In a recent population-based case–controlled retrospective study, Kerilikowske et al. [10] report a comprehensive analysis of the biomarkers estrogen receptor (ER), progesterone receptor, p53, Her2/neu, p16, Ki67 and COX-2 in a cohort of 329 women with DCIS managed with lumpectomy alone. The specific aim of the study was to identify factors of DCIS that were associated with subsequent recurrent invasive carcinoma or recurrent DCIS. Competing risk models were used to determine factors associated with risk of subsequent invasive cancer versus DCIS, and cumulative incidence survival functions were used to estimate the 8-year risk.
Results
Patients presenting with palpable DCIS and patients who were triple positive for p16, COX-2, and Ki67 were found, on multivariate analysis, to have a significantly increased risk of subsequent invasive carcinoma with a 5- and 8-year risk of 19.6 and 27.3%, respectively. Neither nuclear grade nor histopathologic characteristics were associated with subsequent invasive carcinoma. Recurrent DCIS was associated with disease extent of 10 mm or more, positive resection margins, high nuclear grade and extensive necrosis, factors included in the VNPI. Recurrent DCIS was also associated with the biomarker combinations ER-Her2+ and ER-Ki67+. Overall, utilizing the biomarker combinations, 27% of the study cohort were classified as high risk for subsequent invasive carcinoma and 44% were classified in the very-low- or low-risk subgroup.
Significance of results
The study by Kerlikowske et al. has several strengths. The patients were derived from a large population-based cohort of women with DCIS treated with lumpectomy alone with a risk assessment over a 5–8 year period. Multiple pathologic factors and relevant biomarkers were analyzed that would predict the risk of these women developing invasive breast carcinoma. It is shown in the study that DCIS presenting as a palpable lesion and a combination of the biomarkers p16+, COX-2+ and Ki67+, emerged as strong predicting factors consistent with other studies that have been published in the literature. The study also analyzed for the first time p16 expression by immunohistochemistry, a well-established biomarker for cervical dysplasia as well as invasive carcinoma following DCIS. Isolated studies have demonstrated the expression of p16 in basal-like breast carcinoma [10]. It is not clear from the study if there is an association of p16 with ER status. Further studies are needed to address the significance of p16 expression in DCIS as well as invasive carcinomas. There are also limitations to the study. Only a subset of the tissue blocks was available for biomarker analysis. The patients were managed with surgery alone so it is not clear if adjuvant radiation and/or hormonal therapy would mitigate subsequent invasive carcinoma risk in the high-risk cohort.
In a similar pursuit to identify novel biomarkers that can identify high-risk DCIS, we have recently reported the role of novel transcription factors FOXA1 and GATA-3 in relation to hormonal status as well as predictors of invasive carcinoma [11]. ER status has been the leading candidate biomarker in DCIS since it plays a key role in the treatment outcome in breast cancer patients. Absence of ER was shown to be a significant predictor of invasive carcinoma recurrence in DCIS [12]; however, the two broad groups of invasive carcinoma, namely ER+ and ER−, are yet to be well understood in DCIS.
Recently, several groups have investigated the functional interaction between the FOXA1 winged helix transcription factor and GATA-3, a zinc finger transcription factor, their role in suppressing ER-dependent breast cancer cell growth and tumorigenesis, and maintenance of breast luminal cell differentiation in vivo. Their use as prognostic clinical biomarkers has recently been studied in invasive carcinoma but not in DCIS [12–15]. FOXA1, originally called the HNF3α, is an ubiquitous transcription factor expressed in the liver, breast, prostate, lung, colon and pancreas that has activator and repressor activity. As an activator, FOXA1 has the unique ability to bind to its target sites on silent or compact chromatin to initiate regulatory events. FOXA1 may also act as a growth inhibitor via binding and activation of the p27 promoter, located within the BRCA1-responsive element [12,13]; thus, it plays a critical role in suppressing ER-dependent breast cancer cell growth and tumorigenesis in vivo [13]. The clinical studies have further suggested that the presence of FOXA1 preserves the invasive carcinoma in a more differentiated form with a decreased metastatic potential, thus acting as an essential prognostic biomarker [12,13].
GATA-3, a member of the zinc finger DNA binding proteins, has been demonstrated in vivo to be highly expressed in the mammary luminal epithelial cells, which are responsible for both the development and maintenance of luminal cells fate [4,5]; thus, it is associated with ER expression in breast carcinomas [14]. In vivo breast cancer models have shown that GATA-3 maintains tumor differentiation and suppresses metastatic potential [14]. Similar to FOXA1, GATA-3 is a promising biomarker that plays a vital role in tumor differentiation hormonal responsive status, and can help to predict clinical outcome [12–15]. Understanding the complex relationship between ERα, FOXA1 and GATA-3 would help to refine the hormone-responsive phenotype in invasive breast carcinoma. However, this association has not yet been defined in DCIS.
The specific aims of our study were to analyze the expression of novel biological transcription markers FOXA1 and GATA-3 for the first time in DCIS, along with established markers MIB-1 (Ki67) and HER2/neu in ER+ and ER− DCIS patients. We wanted to investigate if there was a similar association between FOXA1/GATA-3 with ERα in DCIS. Our secondary goal was to determine if the expression profile of FOXA1 and GATA-3 along with other biomarkers (ER, progesterone receptor, Ki67/MIB-1 and HER2/neu) would stratify the DCIS patients as high risk or low risk in terms of recurrence.
In our initial pilot study we have shown that a strong expression of FOXA1 and GATA-3, low Ki67 and low/absent HER2/neu expression were characteristically seen in our ER+ DCIS groups, which is a similar finding to that previously described in invasive carcinoma [15]. We also observed a low GATA-3 expression, and strong FOXA1 expression in ER− DCIS case. Interestingly, in our study, a higher percentage of ER−, HER2/neu+ cases in the recurrent group showed low GATA-3 expression and strong FOXA1 expression compared with the nonrecurrent group. Further work needs to be carried out on a larger cohort of DCIS patients with recurrence to better understand which variables are best able to predict recurrence and guide therapy decision strategies. The maintenance of FOXA1 and GATA-3 expression in ER− DCIS needs to be evaluated further, since these transcription factors may offer new promising targets for therapy.
Future perspective
In conclusion, DCIS is heterogeneous disease confined to the breast ducts. Our understanding of the biology of DCIS is evolving, but it remains unknown what proportion of DCIS lesions will progress to invasive carcinoma. We are in need of specific biomarkers to predict the clinical outcome of ductal carcinoma in situ to permit individualized therapy and shed the one-size-fits-all approach.
Executive summary
Ductal carcinoma in situ (DCIS) is a heterogeneous disease confined to the breast ducts.
Breast conservation with resection of the involved breast segment followed by adjuvant whole breast radiation or total mastectomy is modes of treatment for DCIS.
It is shown in the study that DCIS presenting as a palpable lesion, and a combination of the biomarkers p16+, COX-2+, Ki67+, emerged as strong predicting factors consistent with other studies that were published in the literature.
Although our understanding of the biology of DCIS is evolving, it remains unknown what proportion of DCIS lesions will progress to invasive carcinoma. We are in need of specific biomarkers to predict the clinical outcome of DCIS to permit individualized therapy and shed the ‘one-size-fits-all’ approach.
Footnotes
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.