30JDM CAN PHYSICIANS BETTER TARGET BREAST CANCER TREATMENTS TO WOMEN AGED 80 AND OLDER WITH EARLY STAGE DISEASE BY LIFE EXPECTANCY?

Monday, October 19, 2009
Grand Ballroom, Salons 1 & 2 (Renaissance Hollywood Hotel)
Mara Schonberg, MD, MPH1, Edward R. Marcantonio1, Donglin Li, MD, MPH2 and Ellen P. McCarthy, PhD, MPH1, (1)Division of General Medicine and Primary Care, Brookline, MA, (2)Beth Israel Deaconess Medical Center, Brookline, MA

Purpose: Decision-making around breast cancer treatment among women aged 80+ with early stage disease is challenging; especially since the benefits of radiation treatment (XRT) after breast conserving surgery (BCS) are debatable particularly among those in poor health.

Method: We used the linked Surveillance Epidemiology and End

Results -Medicare claims dataset from 1992-2002 to identify women aged 80+ with newly diagnosed early stage (I/II), estrogen receptor positive (ER+), lymph node negative (LN-), breast cancer.  We used the logrank test to determine conditions that were significantly associated with an increased risk of death from causes besides breast cancer within two years of diagnosis.  We considered 14 conditions from the Charlson Index and 14 conditions from Clinical Classification Software (e.g., pneumonia within 2 years).  To determine which conditions were most associated with death from non-breast cancer causes, we included in a Cox proportional hazards model all conditions (n=20) that were significantly associated with non-breast cancer death in bivariable analyses.  We used backwards elimination and set p <0.05 for retention.  We examined initial treatments (BCS+XRT, mastectomy, BCS alone, or no surgical treatment) among women with each significant condition and among women without these conditions.  Result: We identified 7,855 women aged 80+ with stage I/II, ER+, LN-, breast cancer.  Twelve conditions were associated with a significantly increased risk of non-breast cancer death in the multivariate model (the five with the greatest risk are listed below).  Treatments received for women with each condition are also reported.
Condition

n

Hazard ratio of non-breast cancer death in 2 years, 95% CI Treatments (%)

 

 

 

BCS+XRT

Mastectomy

BCS alone

None

Dementia

257

2.9 (2.4-3.7)

4

44

49

3

Parkinson’s

104

2.0 (1.4-2.9)

14

44

39

2

Chronic Renal Failure

92

 1.9 (1.3-2.8)  

23

30

37

2

Heart Failure

1,087

1.8 (1.5-2.1)

19

40

39

2

Pneumonia

335

1.5 (1.2-1.9)

13

42

42

3

 

 

 

 

 

 

 

Any significant condition

3,625

2.8 (2.5-3.3)

24

41

34

2

None

4,230

 

30

43

26

1

Conclusion: Many women aged 80+ with conditions that put them at increased risk of death from non-breast cancer causes received XRT after BCS for early stage disease. Although there is some targeting of treatments based on health for conditions such as dementia, XRT after BCS may be overused among older women with other conditions (e.g., heart failure).

Candidate for the Lee B. Lusted Student Prize Competition