Predictors of Recurrent Breast Tumors in Women with DCIS

Institution: University of California, San Francisco
Investigator(s): Karla Kerlikowske, M.S., M.D. -
Award Cycle: 1996 (Cycle II) Grant #: 2RB-0197 Award: $372,771
Award Type: Research Project Awards
Research Priorities
Etiology and Prevention>Etiology: the role of environment and lifestyle



Initial Award Abstract (1996)
The purpose of this research is to better define which breast ductal carcinoma in situ (DCIS) will recur after surgery. DCIS is a premalignant breast lesion confined within the mammary ducts that is detected primarily by mammography. Prior to the widespread practice of mammographic screening for breast cancer, DCIS occurred at a rate equal to 1-2% of breast cancers diagnosed in the United States. However, DCIS now accounts for a number equal to over 12% of all newly diagnosed breast cancers. Some DCIS lesions will go on to develop into invasive breast cancer. In addition, some DCIS lesions that are removed by surgery recur. At present, we cannot distinguish which DCIS lesions have the potential to develop into invasive cancer from those that never will. We also cannot distinguish which DCIS lesions will recur after surgery from those that never will. Thus, there is not consensus among doctors about the best way to treat DCIS. Some doctors recommend mastectomy (removal of the entire breast), while others recommend lumpectomy (removal of the DCIS lesion and the area of breast surrounding the lesion) followed by radiation therapy, and others recommend lumpectomy alone. Until recently, most women with DCIS have been treated with mastectomy. However, since it is recommended that most women with early stage invasive cancer undergo lumpectomy, the routine use of mastectomy for patients with DCIS has been called into question.

We are specifically targeting women with DCIS who were treated with lumpectomy alone since we are interested in identifying prognostic (predictive) factors that may lead to more individually appropriate recommendations for treatment. Since very few women receive no surgical treatment for DCIS, studying untreated women is not feasible. Specifically, we propose to conduct a population-based, nested case-control study (i.e., a study which compares women who had DCIS which recurred -- cases -- to a similar group of women without a recurrence -- controls -- drawn from a defined population of DCIS cases) to measure factors that may influence whether DCIS lesions recur as either DCIS or invasive cancer following lumpectomy. The population base is the nine-county San Francisco Bay area.

We will measure several different types of factors, including: 1) epidemiologic and clinical factors including information such as the age of the woman when DCIS was first detected, how the DCIS lesion was first discovered, whether the woman had breast symptoms when DCIS was first discovered, and whether there is a family history of breast cancer, 2) tumor factors such as size of tumor and morphology (i.e., the form and structure) of the premalignant cells; and 3) molecular markers of tumor function including estrogen and progesterone receptors, proteins uniquely made by tumor cells (p53 and erbB-2) and Ki67 which measures how fast tumors grow. A "cohort" (i.e., women diagnosed with DCIS during the period 1983 to 1992, who were over 40 years of age, and were treated by lumpectomy alone) of approximately 935 women will be identified from nine San Francisco Bay Area counties through the regional SEER cancer registry and interviewed by telephone to obtain information on epidemiologic and clinical factors. From these, all recurrent cases -- either of DCIS or invasive cancer -- (estimated to be 140 of the 935 women) and two controls (women without recurrent disease) per case (or 280 controls) will be selected from this cohort for inclusion in the study. For the 420 women selected for the nested case-control study, paraffin embedded tissue blocks (initial and recurrent blocks) will be collected by the regional tumor registry for standardized pathology review and determination of molecular markers.

This will be the largest study to date to investigate women with DCIS who have breast tumors recur and the only one to look at molecular markers in a population-based sample. The results of this study will be used to make recommendations on prognostic factors useful in predicting the risk of recurrent breast tumors after lumpectomy. Such knowledge will be useful to patients and clinicians since it will provide a better basis for determining appropriate treatment, such that those women who have a low risk of disease recurrence may avoid receiving unnecessary radiation therapy and those that have a high risk of recurrence may consider radiation therapy in addition to lumpectomy or possibly mastectomy.


Final Report (2000)
The purpose of this research study was to better define which breast ductal carcinoma in situ (DCIS) lesions, a preinvasive breast lesion confined within mammary ducts, will recur after lumpectomy. We conducted a study to measure factors that may influence whether DCIS lesions recur as either DCIS or invasive cancer. We measured several factors including: 1) Epidemiologic and clinical factors such as the age of the woman when DCIS was first detected, how the lesion was first discovered, whether the woman had breast symptoms when DCIS was first discovered and whether there is a family history of breast cancer; 2) Tumor characteristics such as size and microscopic structure; and 3) Measurements of tumor function (molecular markers) including estrogen and progesterone receptors, proteins uniquely made by tumor cells (p53 and erbB-2) and Ki67 which measures how fast tumors grow. A total of 1,568 women with DCIS treated by lumpectomy alone from 1983 to 1994 were identified from the nine Bay Area counties: 210 were ineligible, 199 refused participation, 67 could not be located, 26 do not speak English and 6 had a doctor's request for no contact. Of the 1,060 remaining women, 83% are white, 6% Black, 7% Asian and 8% Hispanic.

A total of 208 (19.6%) women developed recurrent disease, i.e., in the same breast; of these, 56% where DCIS and 44% invasive, and 49 developed contralateral disease (4.6%), all over a median of 73 months. Paraffin embedded tissue blocks for initial and recurrent events for cases and controls were collected for pathology review and determination of molecular markers. A total of 134 initial materials for cases and 203 for controls and 118 initial and recurrent materials for cases have undergone pathological review and slides have been prepared for molecular marker assays.

The final numbers will be 172 initial materials for cases and 296 for controls, and 180 initial and recurrent materials for cases. Although CBCRP funding for the study has ended, during the next year we will analyze the epidemiology, pathology and molecular marker data and publish the results.

The results of this study will be used to make recommendations concerning factors useful in predicting the risk of recurrent breast tumors after lumpectomy. Such knowledge will be valuable to patients and clinicians since it will provide a basis for determining selective treatment, such that, those who have a low risk of disease recurrence may avoid receiving unnecessary radiation therapy and those that have a high risk of recurrence may consider radiation therapy in addition to lumpectomy or possibly mastectomy.

Publications:
Risk prediction for local versus regional/metastatic tumors after initial ductal carcinoma in situ diagnosis treated by lumpectomy



Characteristics Associated With Recurrence Among Women With Ductal Carcinoma In Situ Treated by Lumpectomy
Periodical:Journal National Cancer Institute
Index Medicus: J Natl Can Inst
Authors: Kerlikowske K, Molinaro A, Cha I, Ljung BM, Ernster VL, Stewart K, Chew K, Moore DH, Waldm
Yr: 2003 Vol: 95 Nbr: 22 Abs: Pg:1692-1702