Update: Breast Cancer Prevention

1
Update Breast Cancer Prevention

• Erin Dunn Snyder, MD
• GIM Noon Conference
• October 16, 2007

2
(No Transcript)
3
TIME October 15, 2007
4
Road Map

• Breast cancer Screening
• Digital Mammography
• Computer aided detection
• MRI
• Primary Prevention of Breast cancer
• Lifestyle changes
• Diet
• Exercise
• Medication
• Selective Estrogen Receptor Modulators

5
Breast Cancer Screening

• Guidelines recommend yearly mammogram for average
  risk women, beginning at age 40-50
• Other modalities clinical and self exams,
  ultrasound, digital mammogram, breast MRI

6
Digital Mammography

• Allows radiologist to manipulate the image AFTER
  it has been acquired.
• to optimize contrast in certain areas of the
  breast.
• Potentially allows less radiation dose per study
• Electronic storage
• Easy data transfer
• Substantially more expensive than conventional
  film mammography systems

7
Diagnostic Performance of Digital versus Film
Mammography for Breast Cancer Screening

• 49,528 asymptomatic women at 33 sites in the US
  and Canada
• All women underwent digital and film mammography
• Five different digital systems used
• Examinations independently interpreted by 2
  radiologists
• Scored both on 7 point malignancy scale, and
  standard Bi-Rads scale
• Women returned for follow-up mammogram at one
  year

Pisano, et al. NEJM 20053531773-83
8
Digital vs. Film Mammography

• Positive cancer verified within 455 days of
  initial mammogram
• Negative negative follow-up mammogram, or
  negative biopsy
• Subgroup Analysis
• Age, race, breast density, menopausal status
• Baseline risk of breast cancer
• Type of digital mammogram system

9
Digital vs. Film Mammography
Difference in AUC 0.03 95 confidence interval
-0.02 to 0.08 P value 0.18
10
Digital vs. Film Mammography
Difference in AUC 0.15 95 CI 0.05 to 0.25 P
0.002
Difference in AUC 0.11 95 CI 0.04 to 0.18 P
0.003
11
Digital Mammography

• No difference seen in overall population
• No difference seen in subgroups of
• Race, Risk of breast cancer, Type of system
• Significant improvement in breast cancer
  detection for
• Women under age 50
• Women with dense breast tissue
• Pre- or Peri-menopausal women
• Mortality was not an end point

12
Digital vs. Film Mammography

• Four other studies
• None showed significant differences in the whole
  group screened
• 2007 Norwegian study
• Significant improvement in detection of DCIS
  (p

Vigeland, E. European J Radiology. 2007
13
Digital vs. Film Mammography

• UAB and TKC do use digital mammography for
  routine screening
• Likely it is as good as conventional film
  mammography, perhaps better in certain
  populations and with certain cancers
• Limited by availability, expense, and radiology
  expertise

14
Computer aided detection

• Mammograms digitized
• Software marks suspicious areas for further
  review
• FDA approval in 1998

15
Influence of Computer-Aided Detection on
Performance of Screening Mammography

• 43 Facilities, surveyed 1998-2002
• Breast Cancer Surveillance Consortium
• 222,135 women, 429,345 screening mammograms
• 2351 cases of breast cancer diagnosed within one
  year of screening

Fenton, et al. NEJM 2007356-1399-409
16
Performance of Computer-Aided Detection

• 36 facilities did not use CAD
• Patients were older, had denser breasts, less
  likely to have undergone recent mammography
• Radiologists had more experience
• 7 facilities implemented CAD during study
• No difference in patient or radiologist
  characteristics before and after implementation

17
Performance of Computer-Aided Detection
Unadjusted Performance of Screening MMG according
to use of CAD
18
Performance of Computer-Aided Detection
Number of cancers detected per 1000 screening
mammograms
19
Performance of Computer-Aided Detection
Difference in AUC0.048 P 0.005
20
Performance of Computer-Aided Detection

• Significantly decreases specificity and positive
  predictive value
• Significantly decreased accuracy
• Increased recall and biopsy rates
• No change in overall detection of cancer
• Perhaps some increased sensitivity in detection
  of DCIS

21
Breast MRI

• Compared to mammogram, Breast MRI is
• More sensitive, less specific
• Detects clinically occult cancers, at earlier
  stages
• Associated with higher recall and biopsy rates
• Expensive

22
Breast MRI

• Multiple studies in young, high risk women
• BRCA Positive patients or family members
• Lifetime risk 25 based on models
• Familial cancer syndromes

23
Screening with MRI and MMG of UK population at
high familial risk of breast cancer

• 649 women, 35-55 years old with
• Known BRCA1, BRCA2, TP53 mutation carriers
• First degree relative of carrier
• Strong family history of breast/ovarian cancer
• Family history of Li-Fraumeni syndrome
• Excluded previous breast cancer, expected life
  expectancy
• Women got annual Mammography and MRI for 2-7 years

Leach, et al. Lancet. 2005. 3651769-78
24
MRI vs. MMG for high risk women
25
Breast MRI

• ACS recommends MRI in addition to MMG yearly in
  High Risk women
• Still not widely available, although it is
  available at UAB
• Need skilled, experienced radiologist
• Need ability to do MRI guided biopsy

Saslow, et al. CA Cancer J Clin 2007 5775-89
26

• What about prevention?

27
Low Fat Diet

• Womens Health Initiative
• Postmenopausal women, age 50-79
• 48,835 randomized into Dietary Modification Trial
• Exclusions
• Breast or other cancers
• Expected survival
• Baseline diet with
• Adherence concerns

WHI investigators. JAMA 2006. 295629-42
28
WHI Diet and Breast Cancer

• Intervention group
• 18 behavioral modification group sessions in
  first year, quarterly thereafter
• Dietary modification goals
• Total fat intake
• At least 5 servings fruit and vegetables/day
• At least 6 servings grains/day
• Control group
• Dietary guidelines booklet, not asked to make
  changes

29
WHI Diet and Breast Cancer

• Intake monitored with questionnaire
• Baseline, year one, year three
• Subgroup also provided serum yearly for
• Estradiol, Estrone, Testosterone, SHBG
• Various antioxidant biomarkers, lipids, glucose,
  insulin
• MMG at baseline and every 2 years

30
WHI Diet and Breast Cancer

• Intervention group had
• Fewer calories from fat
• More servings fruit and vegetables/ day
• Early weight loss, able to maintain change longer
  than control group
• Minimal changes in blood biomarkers
• Modest decrease in LDL
• No change in HDL, triglycerides, insulin, glucose
• Greater reduction in estradiol, increase in SHBG

31
WHI Diet and Breast Cancer
32
WHI Diet and Breast Cancer

• 9 lower incidence of breast cancer at end of
  study, not significant
• Cancers more likely to be PR negative (p0.04)
• 22 reduction in cancer (CI .64-.95), in women
  with higher fat intake at baseline

33
Physical Activity

• Several case-control studies show a link between
  breast cancer and physical activity
• Known to
• Modify menstrual characteristics
• Increase SHBG levels
• Decrease weight, abdominal adiposity
• Improve insulin sensitivity

34
Long Term Recreational Physical Activity and Risk
of Breast Cancer

• Prospective cohort study
• California Teachers Cohort 133,479 current and
  retired female participants in CA State Teachers
  Retirement System as of 1995
• Cancer diagnosis determined via CA Cancer
  Registry linkages
• Exclusions
• Previous history of breast cancer
• 80 yrs at baseline
• Incomplete data

Dallal et al. Arch Internal Med 2007 167408-15
35
Recreational Activity and Breast Cancer

• Physical activity self reported via baseline
  questionnaire
• Moderate vs. strenuous activity
• Hours per week and months per year at specific
  age ranges
• Baseline Breast Cancer risk
• Age, race, family history, estrogen exposure,
  menopausal status
• BMI, smoking history, alcohol history
• MMG and breast biopsy history

36
Recreational Activity and Breast Cancer

• Difference only observed with 5 hrs/week of
  strenuous activity
• Invasive Breast Cancer RR 0.80 (CI 0.69-0.94)
• ER cancers RR 0.45 (CI 0.27-0.76)
• In Situ Breast Cancer RR 0.69 (CI 0.48-0.98)
• No difference seen in ER invasive cancers

37
Selective Estrogen Receptor Modulators
38
Tamoxifen and Breast cancer

• 13,388 women randomized to Tamoxifen 20mg/day or
  placebo for 5 years
• Inclusion criteria at least one
• 60 years
• 35-59 with 5 year predicted cancer risk 1.66
• History of LCIS
• History of atypical hyperplasia
• MMG within 180 days of randomization without
  evidence of cancer
• No HRT, OCP, androgens within 3 months of
  randomization
• No history of DVT or PTE

Fisher et al. J National Ca Instit 2005971652-62
39
Tamoxifen and Breast cancer
40
Tamoxifen and Breast cancer

• Unblinded after 5 years due to positive results
• Tamoxifen reduced risk of invasive and non
  invasive breast cancer
• Rate of ER cancer significantly less
• Cancers less likely to be node
• No difference in breast cancer deaths

41
Tamoxifen and Breast cancer
Secondary end points rate of event per 1000 women
42
Tamoxifen vs. Raloxifene

• Study of Tamoxifen and Raloxifene
• 19,747 women randomized to Tamoxifen 20mg/day or
  Raloxifene 60mg/day x 5years
• Same risk as tamoxifen trial
• No stroke, PTE, DVT, uncontrolled A Fib, DM, or
  HTN, current use of warfarin
• No malignancy within last 5 years

Vogel et al. JAMA 2006 295 2727-2751
43
Tamoxifen vs. Raloxifene
44
Tamoxifen vs. Raloxifene
Secondary end points rate of event per 1000 women
45
Tamoxifen vs. Raloxifene
Secondary end points rate of event per 1000 women
46
SERM and Breast cancer
All trials of Tamoxifen and Raloxifene vs placebo
Cuszick, J et al. Lancet 2003 361296-300
47
SERM and Endometrial cancer
All trials of Tamoxifen and Raloxifene vs placebo
Cuszick, J et al. Lancet 2003 361296-300
48
SERM and Thromboembolic Disease
All trials of Tamoxifen and Raloxifene vs placebo
Cuszick, J et al. Lancet 2003 361296-300
49
SERM and Breast Cancer

• Tamoxifen shown to prevent the development of
  breast cancer
• Increased risk PE, endometrial cancer, cataracts
• Raloxifene as good as Tamoxifen in invasive
  breast cancer prevention
• Less PE, endometrial hyperplasia, cataracts
• Both approved for primary prevention of breast
  cancer in postmenopausal women
• Tamoxifen 20mg qd x 5 years
• Raloxifene 60mg qd x 5 years

50
Summary

• Breast cancer screening
• Digital Mammogram
• Increased accuracy for younger women, women with
  dense breast tissue
• Computer Aided detection
• Decreased accuracy due to decreased specificity
• Significant increase in recall and biopsy rate
• MRI
• Only studied in high risk women
• Combination of MRI and MMG more sensitive than
  either modality alone

51
Summary

• Primary Prevention via lifestyle changes
• Low fat diet might play a role, difficult to
  study
• More benefit in women with higher baseline fat
  intake
• Recreational activity associated with decreased
  incidence of ER tumors
• Had to be strenuous activity, 5 hrs per week
  over lifetime

52
Summary

• Primary Prevention via SERMs
• Tamoxifen and Raloxifene can prevent invasive
  cancers
• No difference between the two in cancer incidence
• Raloxifene with less endometrial hyperplasia but
  not uterine cancers
• Raloxifene with less PE and cataract
• Raloxifene known to decrease risk of vertebral
  fracture