NVVP/CME accreditatie:

E-learning (A); Interobserver analysis of histopathological classification of DCIS; stepping stone towards DCIS risk stratication and the Dutch Breast Pathology Course (B).

Secondary prevention of breast cancer: risk stratification for personalized management of screen-detected ductal carcinoma in situ (DCIS)

Jelle Wesseling1,2, Esther Lips2, Emiel Rutgers3

1Department of Pathology, Division of Diagnostic Oncology, 2Division of Molecular Pathology and 3Division of Surgical Oncology at the Netherlands Cancer Institute- Antoni van Leeuwenhoek Hospital

Since the introduction of population-based breast cancer screening and digital mammography, the incidence of precursor lesions has increased more than five-fold, without a decline in invasive breast cancer incidence. This suggests that overdiagnosis of precursor lesions exists. Most precursor lesions are Ductal Carcinoma In Situ (DCIS) cases. Indolent DCIS lesions might be managed by active surveillance sparing such women intensive treatment. A major problem is that we do not know yet which DCIS lesions will develop into invasive breast cancer and which will not.

Currently, we are performing an epidemiological follow-up analysis after a primary diagnosis of DCIS. A nested case-control study within this cohort is set up to find features predicting invasive ipsilateral breast cancer (iiBC). In another study, we evaluate the follow-up retrospectively of women referred to the hospital because of screen-detected microcalcifications. However, robust patient-tailored risk stratification between screen-detected indolent and aggressive DCIS is still lacking. In addition, risk prediction is severely hampered by the substantial interobserver variation among pathologists in classifying and grading such lesions.

To enable individualized DCIS risk stratification, it is essential to integrate clinical, morphological and molecular features. Strikingly, such an integrative approach is not available yet. Therefore, we propose here to develop, and validate a tool reliably distinguishing harmless from aggressive screen-detected DCIS that may help clinicians and women with DCIS to decide between management by active surveillance or more intensive treatment.

We will make use of the unique opportunity present in the Netherlands with well-organized registries connected by a state-of-the-art infrastructure allowing integration of data from population-based screening, hospital records, the cancer registry, and the national pathology network PALGA.

E-learning

We have compiled a well-documented nationwide representative DCIS patient cohort based on

on data from the Dutch Cancer Registry and the nationwide pathology database network system.

This cohort comprises data from over 10,000 women treated for primary DCIS between 1989 and 2004 with a median follow-up of 11.6 years. Within the cohort described above, we will focus on the subgroup of women referred to the hospital via population-based screening and managed by breast conserving treatment, with or without subsequent radiotherapy between 1993 and 2004. Three hundred fifty random cases of this screen-detected subgroup will be studied into detail in terms of clinical, imaging, histopathological and molecular data.

One of the necessary steps is to analyze inter- and intra-observer agreement of the histopathological classification of DCIS. Are we able to reliably classify DCIS to study its potential value in a risk prediction tool distinguishing harmless from aggressive screen-detected DCIS. Also, we will study which histopathological biomarkers are capable of predicting the risk of development of ipsilateral invasive breast cancer.

The e-learning entails just this, the classification of several histopathological variables concerning around 350 DCIS lesions using digitalized slides with web-enabled software providing rapid access to easy-to-use questionnaires.

We expect participants will need around 8 hours for this task (we would like each participating pathologist to review 150 DCIS cases).

The participant will acquire experience in the detailed classification of ductal carcinoma in situ.

The participant will receive feedback on how they scored compared to the other participants.

The participant will receive the results of the inter- and intra-observer agreement analysis and the risk prediction value of their classification.

Dutch Breast Pathology Course

This international ’Dutch Breast Pathology Course’ aims to explore our possibilities, challenges, and limitations in diagnosing and classifying breast cancer precursor lesions in the clinical context. Therefore, input will and has to be provided by different disciplines as well, i.e. epidemiology, radiology, surgery, radiotherapy and the patient perspective. Analysis of the e-learning will be presented, including a discussion on how to improve the histopathological classification of DCIS.

To achieve all this, we intend to encourage the participants to practice, interact, discuss, and come up with ideas how to solve this breast cancer precursor dilemma.

Inspiring and clinically relevant lectures

Informative case studies

Challenging debates

Lots of interaction and networking

Program at a glance

Detection, diagnostics and treatment of breast precursor lesions will be discussed in depth in relation to the prevention of overtreatment, including the relevance of precursor lesions in the low-grade spectrum, patient perspectives on treatment decisions and current trials on DCIS. The issue of interobserver variability will be highlighted and interactively discussed for improvement of our current classification system.