Home - St. Luke's Hospital
Letter from Dr. Van Amburg

2011 Cancer Registry Report

New Infusion Center

Linear Accelerator Upgrade

Breast Advantage Program

High Risk Breast Clinic

NAPBC Accreditation

SonoCine Update

Gynecological Cancer Treatment with the da Vinci Surgical System

Social Workers Provide Comfort to Patients

Community Support and Screenings

Progress in Prediction of Breast Cancer Outcome

From Our Patient

2011 Oncology Peer Review Committee Roster



Back to Main Site
Progress in Prediction of Breast Cancer Outcome: Infiltrating Lobular Carcinoma.
John S. Meyer, M.D. Department of Pathology

Breast cancer consists chiefly of a group of tumors derived from epithelium, cells that line the ducts of the lobules, where milk is secreted, and the ducts that convey milk to the nipple. These tumors are called carcinomas in distinction from sarcomas, malignancies of non-epithelial tissues like muscle, fat, bone, and connective tissues. In situ carcinomas are confined to the ducts and have not invaded the surrounding fat and connective tissues. In situ carcinomas do not spread, but they are forerunners of life-threatening invasive carcinomas. About 70 percent of invasive carcinomas are designated invasive ductal, although they can be separated into subtypes by modern techniques. Ten percent are classified as lobular carcinoma which originate in the coils of small ducts of the lobules. Other types of invasive carcinomas distinguishable under the microscope are termed tubular carcinoma, mucinous carcinoma, and medullary carcinoma, which have relatively good prognosis, and other , rare types. In the last 4 decades much progress has been made in characterizing common infiltrating ductal carcinoma by means other than simple microscopy. Notably, breast carcinomas are now analyzed for estrogen and progesterone receptors, protein molecules that are responsible for responses of breast cells to those hormones. More recently proteins on the cell surface that respond to growth factors have been identified, notably HER2. Therapies have been developed that block the effect of hormones and other growth factors and interfere with growth of the tumor.

Growth rate of breast carcinomas is an important determinant of tumor aggressiveness. It is determined largely by the rate at which cells of the tumor divide. Cell division can be observed microscopically as mitotic figures. Mitotic figures are relatively few because division occurs in about 20 minutes, whereas the cell needs a day or more between divisions to grow and replicate DNA of the chromosomes before it can divide again. Thus counting of mitotic figures is not an accurate means of measuring the rate of proliferation. The markers most often used are radioactive thymidine or bromodeoxyuridine, which are DNA precursors detectible by special microscopic techniques. S-phase measurement by these methods became available in the 1970s. A more accurate assessment of growth rate can be made by labeling cancer cells with a radioisotope that is incorporated into DNA.

More recently, a protein called Ki-67 (one of a series of monoclonal antibodies that attach to specific proteins in cells developed at the University of Kiel, Germany) was shown to be present in and restricted to cells that were committed to growth and DNA synthesis or in mitosis, covering essentially any phase of the cell cycle. The Ki-67 protein can be detected microscopically by modern immune-staining methods, and is said to measure the "growth fraction" of the cell population. Growth fractions of human cells range from essentially zero in tissues such as nerves, muscle, fat and brain, a few percent in slow-growing tumors, to virtually 100 percent in fast-growing, highly malignant tumors.

This study includes patients with invasive lobular carcinoma who were diagnosed at St. Luke's Hospital in the years 1985 through 2002. One hundred sixty-eight patients had one or more axillary lymph nodes examined, of whom 94 did not have axillary metastasis (node-negative), 74 had axillary metastasis (node-positive), and 36 had no lymph nodes examined. Outcome data were available for 75 of the node-negative patients, 63 of the node-positive patients, and 29 of the patients with no axillary lymph nodes examined. Adjuvant systemic therapy was tabulated, and survival analysis was done by the method of Kaplan-Meier with log rank testing for significance of differences .

The tertile cutoffs for Ki-67 index for our entire group of infiltrating lobular carcinomas were 4.9% and 8.2%. Using these cutoffs that divided the group of lobular carcinomas into high, mid and low thirds, we found notably better survival for patients with lower third Ki-67 indices in comparison with patients with mid or high values. Statistical testing showed this result to be highly significant, P = 0.0001 (null hypothesis rejected). Probability of freedom from recurrent tumor within ten years for the 77 patients available for analysis was projected as approximately 94% for the low group, 88% for the mid group, and only 33% for the high group. Four patients in the node-negative group had the atypical variant of lobular carcinoma. When they were excluded from the analysis, the result was essentially unchanged. Only one of the atypical lobular carcinoma patients had relapsed. These results suggest that only minimal benefit of adjuvant systemic treatment of node-negative patients could be expected if Ki-67 indices are in the low or mid thirds (below 8.2%).

Results were different for 63 patients with metastasis to axillary lymph nodes. Projected relapse-free survivals at ten years were 61% for low Ki-67 index, 55% for mid, and 47% for high. Statistical analysis showed these differences to be too small to accept with confidence (P = 0.48, null hypothesis not rejected). Five atypical lobular carcinoma patients were included in this group, one of whom relapsed, and their exclusion from the analysis left the results essentially unchanged.

Thirty-six patients had no lymph nodes removed during initial therapy. Projected relapse-free survivals at ten yers were 90% for the low Ki-67 group, 89% for the mid group, and 23% for the high group. Two atypical lobular carcinomas were in the cohort with no lymph nodes, and neither patient relapsed.

Other variables with potential relationship to outcome are age of patient, size of the primary tumor, maximum dimension of the primary tumor, nuclear size, nucleolar size, angiolymphatic invasion, and estrogen and progesterone receptor detection. Size was not a significant predictor for the node-negative patients , although it was predictive when all nodal classes were analyzed in aggregate (univariate P =0.035). In proportional hazards analysis of the node-negative cohort, the only predictors were Ki-67 index (P <0.0001) and size of tumor (P = 0.015).

The patients in this study received various types of cytotoxic therapy or chemotherapy, and in four patients both systemic and adjuvant treatments. Interaction between systemic therapy and the proliferative index of the tumor can influence relapse-free survival and overall survival. Analysis of relapse-free survival by adjuvant therapy subset showed that the Ki-67 index predicted outcome in node-negative patients who received either chemotherapy or hormonal therapy and in the group who received no systemic adjuvant therapy. This result is evidence that increased risk of progression of more highly proliferative carcinomas is maintained despite adjuvant systemic therapy.

Conclusions: Invasive lobular carcinomas vary in proliferation rate as measured by Ki-67 index. Atypical lobular carcinomas tend to have higher Ki-67 indices than the typical variant, but their ranges overlap. Atypical lobular carcinoma can be considered to be within the lobular carcinoma spectrum rather than a separate and distinct entity. A Ki-67 index of 8.9% or below by a microscopic count of 1,600 cells is a strong, independent predictor of relapse-free survival in node-negative invasive lobular carcinoma in patients receiving no adjuvant systemic therapy or in patients treated with hormonal or cytotoxic means.

Center for Cancer Care | Home - St. Luke's Hospital

232 South Woods Mill Road - Chesterfield, MO 63017 - 314-434-1500
© St. Luke's Hospital
Website Privacy Policy | Website Terms and Conditions
Patient Notice of Privacy Policies (PDF)