Health & Medical Cancer & Oncology

Molecular Phenotypes of DCIS Predict Recurrence

Molecular Phenotypes of DCIS Predict Recurrence


DCIS is a heterogeneous disease, with molecular phenotypes seen in invasive cancer presenting in primary DCIS using IHC surrogate markers. These molecular phenotypes can be used to independently predict for both overall and, importantly, invasive recurrence. DCIS subtypes have the potential to identify women more likely to develop disease recurrence following surgery and will benefit from maximal adjuvant therapy, compared with those with ER+, HER2− (Luminal A) lesions, who could potentially avoid adjuvant treatment.

Patients who undergo BCS are more likely to suffer a disease recurrence compared with mastectomy patients. Those with involved surgical margins are at highest risk. The optimal radial surgical margin width in DCIS remains controversial in the literature. Our unit guidelines (changed in 1997) suggest that a radial margin ≥1 mm is acceptable, based on previously published data.

Before 2000, we did not use radiotherapy if margins were clear by >1 mm or in small tumours <2 cm in size, and thus radiotherapy use was under 18% overall. The low recurrence rate, particularly in the Luminal A ER-positive/HER2-negative DCIS, reflects a low risk of recurrence in the absence of radiotherapy and indicates a difference in the molecular phenotypes with regard to recurrence and radiotherapy sensitivity. Although this is non-randomised data, the long follow-up and the fact that the other non-Luminal A subtypes recurred within 5 years indicate this is a genuine finding.

We found that patients diagnosed with high-grade tumours were more likely to recur than intermediate or low-grade tumours (HR 2.28, 0.024); however, the presence of comedo necrosis within the tumour failed to predict for recurrence. Patient age was not an independent predictor of recurrence, but the dataset used for analysis lacks patients of a younger age, as the majority of DCIS lesions are diagnosed only on screening mammmogram and UK screening did not commence until patients were 50 years old.

High Ki67 expression (>14%) was an independent predictor for invasive recurrence in DCIS, whereas previous studies have failed to show the predictive value of Ki67 in DCIS. Previous studies have only investigated overall recurrence and have not looked specifically at invasive recurrence alone. Kerlikowske et al. combined Ki67 with ER and HER2 expression data to show that patients with ER-negative DCIS express high levels of Ki67 (>10%) and HER2-positive had a greater risk of developing a local DCIS recurrence than other patient groups, which accords with our findings.

The molecular phenotype of DCIS is an independent predictor for both overall and, more importantly, invasive recurrence, when compared with the Luminal A group. Molecular phenotypes do exist in DCIS, but previous studies did not examine the relationship between receptor expression and overall or invasive recurrence. Gene signatures have yet to provide prognostic information for DCIS tumours, although an OncotypeDX for DCIS is currently being validated. Determination of molecular phenotypes of DCIS using surrogate immunohistochemical markers is an economically viable approach to aid identification of women at high risk of recurrence. ER-negative/HER2-positive patients need maximal adjuvant treatment to avoid invasive recurrence. In contrast, lower risk ER-positive/HER2-negative patients whose invasive relapse is significantly less could potentially avoid overtreatment with radiotherapy. The HER2 status of pure DCIS is not currently routinely measured in clinical practice in the UK, but clinicians should consider routine assessment of HER2 status for DCIS as well as invasive breast cancer.

Ethics Committee Approval

All trials received ethics committee approval and patients screened or entered agreed to their data being used for follow-up studies.

Leave a reply