Skip to main content
. 2024 Apr 6;12(7):795. doi: 10.3390/healthcare12070795

Table 2.

Summary of the main key points of the study.

Key Points
Prognostic value of clinical and histopathological factors
  • Some weaknesses in the robustness of the traditional CPFs for risk stratification:
    • -
      Low to intermediate grade: Show a moderate agreement due to various grading systems, interobserver variability, and heterogeneity within DCIS.
    • -
      Tumour size: Difficult to be measured with precision, often considered an estimate.
    • -
      Surgical margins: Minimum margin clearance varies across the studies and guidelines.
  •     -

    The age factor should always be considered in the cost/benefit treatment evaluation.
RT Treatment
  • PBI
    • ○
      Suitable for low-risk DCIS patients (non-palpable presentation, tumour size ≤ 2–2.5 cm, low or intermediate grade, free-surgical margins of at least 3 mm).
  • Tumour bed boost
    • ○
      Beneficial in non-low-risk DCIS patients (≤50 years, symptomatic palpable tumours, size ≥ 1.5 cm, multifocality, intermediate or high nuclear grade, central necrosis, comedo histology, surgical margins less than 10 mm).
  • Postmastectomy RT
    • ○
      Not routinely considered for DCIS, requires an individualized cost/benefit assessment, weighing up some risk factors, such as positive/close margins, high grade, multifocality, comedonecrosis, and age < 50 years.
Biomolecular factors

  • Overexpression of HER2: Frequent in DCIS, uncertain clinical significance, not performed routinely. It is statistically significantly associated with in situ LR [120]; predicts a more effective response to RT for in situ LRs [120].

  • Hormonal receptor status (ER+): One of the most important prognostic factors for LRs [11,107,108]; endocrine therapy (tamoxifen) alone or in combination with RT reduces all BC events [113].

  • Ki-67: Usually associated with high-grade DCIS [118]; incidence of LRs increases with the increase in Ki-67 [119,120]. It is shown to be also a predictor of radiotherapy response [104].
Decision support tools: imaging biomarkers

  • Identification of the DCIS extent and the recognition of multicentricity or multifocality are of paramount importance for the management of DCIS.

    Imaging biomarkers and radiomic features may be useful in the characterization of DCIS and in the prediction of recurrence risk:
  •     -

    Surface–volume ratio was significantly different between high nuclear grade and non-high nuclear grade DCIS [135];

  •     -

    Higher parenchymal signal enhancement ratio around the tumour at preoperative MRI and larger histologic tumour size were independent factors associated with worse recurrence-free survival after conservative surgery [136];

  •     -

    Higher functional tumour volume of lesion and signal enhancement ratio were significantly associated with the risk of developing a recurrence [137].


  • Computer-aided detection (CAD) systems proven to be especially beneficial in mammography tests involving dense breast tissue [139].
Decision support tools: traditional prediction models

  • The Van Nuys Prognostic Index (VPNI) [148]—current application is limited.

    Main key features:
    • -
      Built on retrospective monoinstitutional series;
    • -
      External validation failed;
    • -
      Does not consider endocrine therapy and hormonal receptor status;
    • -
      Threshold for RT recommendation of 20% at 12 years deemed too high;
    • -
      Refined with the addition of margin status width.

  • The Memorial Sloan Kettering Cancer Centre Nomogram (MSKCC) [52]

    Main key features:
    • -
      Includes 10 CPFs and treatment variables (e.g., the use of endocrine therapy/RT);
    • -
      Available online;
    • -
      Biomolecular markers and tumour size not included;
    • -
      Externally validated in five studies with moderate discrimination power and variable calibration.
Decision support tools: biomolecular prediction models

  • The Oncotype DX DCIS Score [166,169]: Multigene expression assay (12 genes) for estimates of 10-year risk of any LR after BCS, regardless of tamoxifen and RT.

    Main key features:
    • -
      No measure of predictive accuracy (discrimination and calibration);
    • -
      Validated in highly selected population;
    • -
      Failed to discriminate intermediate risk from high risk in a larger cohort [164];
    • -
      Margin status not considered;
    • -
      Designed to be unaltered by endocrine therapy;
    • -
      Costly;
    • -
      Improved accuracy by incorporating age, size and year of diagnosis (Refined DS Score);
    • -
      It was shown to be the most impactful factor in the decision making, reducing anxiety and decisional conflict.

  • The DCISion score: Comprises seven cancer-related genes (among which HER2 and Ki-67), related with recurrence and progression, and four CPFs; discriminates patients at low and elevated LR risk and assesses the benefit of RT [177]. The higher the score, the higher the absolute risks of total LRs and invasive LRs [176,177,181,182].

    Main key features:
    • -
      Need to be validated in prospective randomized trial;
    • -
      It is the most important driving factor in the decision making;
    • -
      The novel version considers the EGFR/HER2/KRAS biomarker expression, identifying patients who remained at higher risk after RT.