4/23/2021

1

DUCTAL CARCINOMA IN SITU: RISK ASSESSMENT

Catherine Park, M.D.

April 24 2021

British Journal of Cancer (2019) 121:285–292; https://doi.org/10.1038/s41416‐019‐0478‐6 

British Journal of Cancer (2019) 121:285–292; https://doi.org/10.1038/s41416‐019‐0478‐6 

Evolutionary models have beenproposed to describe the clonal progression of DCIS into IBC

Nor

mal

duc

tIB

CD

CIS

DCIS lesion

vessels

stro

ma

angiogenesis

In situ hybridization

EDA+ FN

Remodeling of ECM

4/23/2021

2

Outcomes are worse in patients with DENSE TILS

AGENDA

1.Incidence and natural history of pure DCIS2.Trials that Selected GOOD RISK DCIS

• RCT and Single Arm3.Risk Stratification

• Molecular Diagnostics• Oncotype DCIS• PRELUDEDX

• MSKCC nomogram5.Prospective Observation Trials

Increase in screening--> 30% rise in cancer incidence

Cancer.gov SEER data base of 18 registries, 1988-2011, N=108,196 individuals

eFigure 1. Twenty-year breast cancer-specific survival after DCIS, by age of diagnosis

4/23/2021

3

Effect of radiotherapy on BCS mortalityNSABP B-17

PROSPECTIVE RANDOMIZED TRIALS OF RT FOR DCIS

Trials N

Median follow up (years)

IBTRRisk

reduction PNo RT RT

NSABP B171 813 17.25 35% 20% 47-52% <0.001

EORTC 108532 1010 15.8 31% 18% 48% <0.001

UK, Aust, NZ3 475 12.7 26% 9% 69% <0.0001

SweDCIS4 1046 17 32% 20% 37.5% <0.001

1Wapnir, et al. J Natl Cancer Inst 2011; 103:4782Donker, et al. J Clin Oncol 2013; 31:40543Cuzick, et al. Lancet Oncol 2011;12:21 (Only those not receiving tamoxifen, and randomized to No RT

or RT are included)4Warnberg, et al. J Clin Oncol 2014; 32:3613

~50% Risk Reduction

DCIS: MANY TREATMENT OPTIONS

Treatment options for DCISWide excision +/- RT +/- endocrine therapyMastectomyBilateral mastectomy

Radiation: WBI, hypofractionated, APBI

Survival is excellent with all; IBTR varies dramatically

4/23/2021

4

Can we OMIT RT in highly selected patients with favorable DCIS?

RTOG 9804—Randomized trial for Good-Risk DCIS

ECOG-ACRIN E5194—Prospective WEA for Good-Risk DCIS

Tamoxifen optional (~62%)

Median f/u= 7 yrs6.7 vs 0.9% LR

Median f/u 12 yrs

30% of patients received tamoxifen

4/23/2021

5

Wide Excision Alone: ECOG 5194: 5 yr update

Hughes et al, JCO VOL 27 � NUMBER 32 � NOVEMBER 10 2009

Low-intermediate gradeSize <2.5 cmMargins >3mm

High gradeSize <1.0 cmMargins >3mm

10.5%

18.0 %

Ipsilateral breast events (IBEs) for cohort 1 and cohort 2. Cohort 1 was defined as low- or intermediate-grade ductal carcinoma in situ (DCIS), tumor size 2.5 cm or smaller. Cohort 2 was defined as high-grade DCIS, tumor size 1.0 cm or smaller.

24.5%

14.4%

ECOG 5194 12 year update

Selection of Patients with DCIS for Optimal Management is Complex

• Heterogeneity in biology/extent

• Difficulties assessing size and margins

• Protracted natural history (especially for low grade lesions) requires long follow up

• Inability to predict clinical outcome can lead to over- or under-treatment

Risk Assessment: Clinical Tools

Molecular Assays:

Oncotype DCIS

PreludeDx

Nomograms--MSKCC

4/23/2021

6

Molecular Diagnostic to Predict Risk

San Antonio Breast Cancer Symposium – December 6-10, 2011

Oncotype DCIS Score as a Predictor of Local Recurrence: ECOG E5194 Subgroup Analysis

Ipsilateral local recurrence:

DCIS Score HR (per 50 units) = 2.31

(95% CI: 1.15, 4.49,p=.02)

Adjusted for tamoxifen

Ipsilateral invasive recurrence:

DCIS Score HR (per 50 units) = 3.68

(95% CI: 1.34, 9.62,p=.01)

Unadjusted

22

Validation Study: Ontario Cohort

Patient CharacteristicsOntario Cohort (N=571)

Age ≥ 50 years 459 (81%)

Nuclear GradeLow 55 (10%)Intermediate 332 (58%)High 184 (32%)

Comedo Necrosis 350 (61%)Solid Subtype 358 (63%)Tumor Size

< 10 mm 150 (26%)> 10 mm 140 (25%)Missing 281 (49%)

Multifocality* 114 (20%)

ER+ by RT-PCR 541 (95%)HER2+ by RT-PCR 100 (17.5%)

*Presence of at least 2 foci of DCIS in the same quadrant at least 5 mm apartSikand et al. J Clin Path, 2005

4/23/2021

7

Study Design

Study population Population based cohort of cases diagnosed with pure DCIS in Ontario 1994-2003 Breast-conserving surgery alone

Negative resection margins

Statistical Analytical Plan Pre-specified study objectives, Laboratory assays, Endpoints

Oncotype DCIS Score Continuous variable (0 – 100) 3 pre-specified risk groups:

Low < 39Intermediate 39 – 54High > 55

Statistics

Cox proportional hazards models

Kaplan-Meier estimates to evaluate 10-year risk of recurrence by DCIS risk group (log rank tests used to compare risk groups)

7

Ontario Cohort Outcomes

Kaplan-Meier 10 yr Risk of LR by DCIS Score Risk Group

Factors Associated with LR: Multivariable Analysis

4/23/2021

8

Comparison of Ontario Cohort to E5194 Cohort: K-M 10 year LR

Multivariable Cox models for any local recurrence

EFFECT OF TUMOR SIZE and AGE

Study Design *A surgeon and a radiation oncologist independently completed pre‐ and post‐assay questionnaires about their recommendations for each patient. *Secondary objectives were to summarize patient clinicopathologic characteristics, to determine the distribution of DCIS Score results across the cohort and across clinicopathologic factors, and to evaluate the effect of DCIS Score results on physician estimates of local recurrence risk. *An exploratory objective was to assess changes in patient decisional conflict and anxiety using validated tools, pre‐ and post‐assay. 

4/23/2021

9

The number of patients with DCIS Score results in the low (blue), intermediate (yellow), and high (green) ranges are reported,along with mean (SD) and median (IQR) DCIS Score results for each risk group.DCIS ductal carcinoma in situ, SD standard deviation, IQRinterquartile range 

Distribution of DCIS score by Risk Group Distribution of DCIS Score by clinical-pathologic factors

Pre- and Post- assay radiotherapy recommendations

Low Intermediate High

• Swe-DCIS trial (n=1046, 1987-1999)

• BCS; randomized to RT/no RT

• PRELUDE DX• 7 biomarkers: HER2, PR, Ki67,

COX2, p16/INK4A, FOXO A1, and SIAH2

• Performed on 584 cases

• Requires at least 4 of 7 biomarkers to complete assay

4/23/2021

10

PRELUDEDX

BCS=BLUE

BCS+RT=ORANGE

The 10‐year absolute risks of IBC and IBE for patients with aDS of 3 or less (low risk) had an average 10‐year IBC risk of 4%(95% CI, 0%–9%) and IBE risk of 8% (95% CI, 0%–14%),adjusted for year of diagnosis 

PREDICT REGISTRY STUDY

Prospective Registry to Evaluate the effect of DCISionRT on treatment decisions in (n=2500) patients undergoing BCS for primary DCIS

Primary outcome: % of cases with changes in treatment recommendations at 5 years

Secondary outcome:

Function of demographic factors (age, ethnicity, FHx)

Function of Tumor factors: (size, grade, architecture, necrosis, palp, margins, ER)

Acta Oncologica, 2006; 45: 536543

Protocol Synopsis, continued

Study Groups/ Cohorts

Female patients with DCIS, 25 years and older. Patients must have histologically confirmed ductal carcinoma in situ (DCIS) in a single breast without evidence of invasive cancer (presence of lobular carcinoma in situ (LCIS) or other benign breast disease in addition to DCIS is acceptable)

Primary Outcome

Measures

Percent of Cases w/ Changes in Treatment Recommendation [Time: 5 years]The study will collect details on physician treatment recommendations before and after availability of the genomic test (DCISionRT) results. The data elements include type of surgery (lumpectomy, therapeutic mastectomy, contralateral prophylactic mastectomy), type of radiation therapy (none, IORT, APBI, whole breast RT) and endocrine therapy (yes, no). The main measure will be percent of cases in which treatment recommendations are changed after the test results become available.

Secondary Outcome

Measures

Function of Demographic Factors [Time Frame: 5 years]Percent of patients for which the recommended treatments change after DCISionRT results are known as a function of demographic factors (age groups <40, 40-50 and >50; ethnicity; family history)

Function of Tumor Factors [Time Frame: 5 years]Percent of patients for which the recommended treatments change after DCISionRT results are known as a function of tumor factors (tumor size, grade, architecture, necrosis, palpability, surgical margins, hormone receptor status).

Protocol Synopsis

NCT Number NCT03448926

Brief Title The PREDICT Registry

Official Title A Prospective Registry Study to Evaluate the Effect of the DCISionRT Test on Treatment Decisions in Patients with DCIS Following Breast Conserving Therapy

Brief Summary This is a prospective cohort study for patients diagnosed with ductal carcinoma in situ (DCIS) of the breast. The primary objective of the study is to create a de-identified database of patients, test results, treatment decisions and outcomes that can be queried to determine the utility of the DCISionRT test in the diagnosis and treatment of ductal carcinoma in situ of the breast.

Study Design Prospective Observational Cohort [Patient Registry]

The PREDICT Registry Study

Miami Breast Cancer Conference® - March 5-8, 2020

4/23/2021

11

MSKCC Nomogram

Based on 1681 patients

10 clinical/pathologic/treatment variables

Estimates probability of IBTR at 5 and 10 years after BCS

External validation

Rudloff U, et al., J Clin Oncol 2010

MULTIVARIABLE ANALYSIS: FACTORS ASSOCIATED WITH IBTR

VariableAdjusted

HRP

value

Age (per year) Continuous 0.98 0.03

Family history NoYes

1.001.34

0.07

Initial presentation

RadiologicClinical

1.001.39

0.09

Nuclear grade LowIntermediate/high

1.001.30

0.25

Necrosis AbsentPresent

1.001.13

0.50

Rudloff et al., 2010, J Clin Oncol

DCIS Nomogram

MULTIVARIABLE ANALYSIS: FACTORS ASSOCIATED WITH IBTR

Variable AdjustedHR

P value

Margin status Negative ≥2mmClose/positive

1.001.73

0.002

Number of excisions ≤2≥3

1.001.68

0.03

Time period 1991-19981999-2006

1.000.57

0.0006

Radiation therapy YesNo

1.002.67

<0.0001

Endocrine therapy YesNo

1.002.11

0.003

ResultsDCIS Nomogram

NOMOGRAM

Results

Rudloff et al., 2010, J Clin Oncol

DCIS Nomogram

4/23/2021

12

DCIS NOMOGRAM EXTERNAL VALIDATIONS

Author Year Institution AUC/C-index

Yi 2012 MDACC 0.63-0.65

Sweldens 2014Univ Hosp Leuven

Belgium0.67

Wang 2014Natl Cancer Centre

Singapore0.67

Collins 2015 Harvard 0.68

TRIAL COMETUS

LORISUK

LORDNetherland

s

LORETTAJCOGJapan

Design RCT RCT Patient preference

Single arm

Age >40 >48 >45 >40, <75

Endocrine Rx

Possible Possible Not allowed Tamoxifen

Primary outcome

2,5,7 yrs 10 yrs 10 yrs 5,10 yrs

Opened 2017 2014 2017 2017

Target (accrual 2020)

1200 (600)

Closed166

1240 (40)

340(60)

ACTIVE SURVEILLANCE in DCIS

PCORI: COMET STUDY

4/23/2021

13

SUMMARY:

For patients with low risk disease, aggressive therapies have limited absolute benefit

Individualized risk estimation based on clinical/pathologic factors and molecular analysis:

Selection criteria for ‘low risk’ in clinical trials (ECOG 5194, RTOG 9804)

MSKCC Nomogram

PreludeDx

Oncotype Dx

The level of acceptable risk to omit hormone or radiation therapy is not defined