Adjuvant Therapy for Node-Negative Breast Cancer The Use of Prognostic Factors in Selecting Patients David P. Winchester, MD, FACS Patients with invasive breast cancer and pathologically negative lymph nodes (NO) have a favorable 10-year survival rate, particu larly with sma ll «1 cm) primary tumors. Overall, however, 20% to 35% will ex perience recurrence with local therapy onl y. Adjuvant chemotherapy or tamoxifen have prolonged disease-free survival (DFS), but not overall surviva l (OS). Unanswered questions of optimal end point (DFS or OS) and the risk of treating many to benefit few have prompted clinicians to use prognostic indicators to fa cili tate treatment recommendations. Currently, the most readily available and accurate information comes from TNM staging, pathologic features, and hormone receptors. Ploidy, S-phase fraction, HER2-neu amplification or over-expression, and cathepsin-D may be useful prognostic indices. Until a more precise system of weighing several prognostic variables is developed, the decision to recommend adjuvant systemic therapy in this generally good prognosis group will have to be thoughtfully considered by patient and physician. Whenever possible. patients should be encouraged to enter clinical trials. Cancer 67:1741-1743, 1991.
T
HE CLINICAL ALERT released by the National Cancer
Institute in May 1988 emphasized the importance of considering adjuvant therapy in women with nodenegative breast cancer. Ten-year survival rates of 65% to 80% indicated far too many treatment failures for this subset of patients receiving local therapy only. linical trial data demonstrated significant disease-free survival prolongation without overall survival benefit when either adjuvant chemotherapy or tamoxifen were admi nistered. ' ·2 A number of issues arose in response to this widespread comm unication . The Program Committee proposed that an appropriate forum for discussion of these issues would be at a National onference on Ca ncer and the Changing HeaIthcare System. One of the major questions related to an assessment of benefit versus risk in this generally good prognosis group. How many patients would be treated unnecessarily to benefit others? Is benefit best defined by disease-free survival? an high- or low-risk Presented at the American Cancer ociety ational onference on ancer and the Changing Healthcare ystem, an Fra ncisco, alifornia. May 3-5. 1990. From the Depart ment of urgery. orthwestern University Medical enter. va nston, Illinois. Address for reprints: David P. Winchester. MD, FA S, Department of Surgery, Evanston Hospital, 2650 Ridge Ave nu e. Eva nston. IL 6020 I. Accepted for publication eptembcr 14, 1990.
1741
patients be identified by any prospective methods? What is a reasonable threshold number for treatment failures to initiate a recommendation for adjuvant therapy? This report considers an ever-expanding number of prognostic factors being developed by laboratorians and considered by clinicians as they try to make their best recommendations for their patients. The Importance of Accurate Staging The topic of adjuvant therapy for node-negative breast cancer was chosen to correspond to clinical alert terminology. The term is misleading and fails to accurately depict stage. After years of negotiation between the UICC and AJ C, there is worldwide agreement on the anatomic staging of this and most other cancers. The term "nodenegative" fails to take tumor size into account. The controversy surrounding adjuvant treatment is primarily focused on T I NOMO, Stage I patients. Most published series dealing with the subject of nOde-negative breast cancer include some patients with T 2 and T3 primary tumors, placing them in Stages 11 or III. Few would argue that these patients should not recei e adju ant therapy. Meaningful comparison of data can only be made if uniform staging is utilized . Future clinical trials hould keep this simple, yet critically important, fact in mind.
1742
CANCER
March 15 Supplement 1991
Time-Tested, Available, and Cost-Effective Prognostic Factors The surgical pathologist can provide a wealth of prognostic information to the managing physician. Numerous · 3-6 have shown that tumor size, nuclear grade and stu d les histologic grade are important prognostic determin~nts. Some cell types, such as tubular, confer a good prognosis. Tumor necrosis and lymphatic and blood vessel invasion are unfavorable findings. 6 The mitotic rate of a tumor is an important component of grading. Rosen et af.? demonstrated a particularly strong association between tumor size in Stage I patients and long-term survival. When stratified by tumor size, T I NOMO patients with tumors 1.0 cm or less in diameter had a significantly better 20year recurrence-free survival (86%) than did T INOMO patients with tumors 1.1 to 2.0 cm (69%). Many medical oncologists do not administer adjuvant therapy to the latter subset of patients as reported by Carbone et at. at an American Cancer Society workshop on the Management of Limited Breast Cancer. s Thus, careful attention to pathologic TNM stage and histopathologic features can provide key clinical information. The aforementioned factors should be widely available at a reasonable cost. Years ago, hormone receptor determination was a promising laboratory examination. Today's standards for quality assurance being developed by the Joint Commission on Accreditation of Healthcare Organizations may well require documentation that estrogen receptor values are determined for patients with invasive breast carcinoma over .1.0 cm in diameter. 9 Many institutions can accurately determine receptor status with immunohistochemical techniques, using much smaller amounts of tumor. In patients with node-negative disease, the value of estrogen receptor determination is controversial. 10, 1I The largest series reported by Fisher et af. 10 and McGuire et at. 12 demonstrated that estrogen receptor was a highly significant prognostic determinant in node-negative patients. Most clinicians regard estrogen receptor negativity as a poor prognostic factor, which may lead them to recommend adjuvant therapy. However, the reports on estrogen receptor represent a classical example of the difference between statistical significance and clinical pertinence. McGuire et al. 13 indicated that despite a highJy significant statistical separation, the magnitude of the difference over a long period of observation was only 8% to 9%. Other Prognostic Factors in Evolution Ploidy and S-phase fraction have been the subject of numerous recent reports. 14- 18 Ploidy alone mayor may not be a significant prognostic indicator for Stage r or ~ode-negative subsets of patients. However, S-phase frachon alone or with ploidy is more consistently predictive
Vol. 67
of recurrence. Diploid, low S-phase patients generally do well, whereas high S-phase fraction, diploid or aneuploid connotes a significantly higher likelihood of recurrence. Considerable variability exists, however, from laboratory to laboratory in S-phase fraction determination. Laboratories seek a separation point between low and high Sphase that will provide the best prognostic differentiation. Fu~ure st~dies in these laboratories should test the separahon pomt prospectively. Thymidine labeling index (TLI) is a more accurate metho~ for determi~ng S-phase, but it is time-consuming, expenSive, and reqUires fresh tissue. It is not likely to reach widespread clinical applicability. Histochemical identification of~-phase specific antigens may offer a more rapid and practIcal method for assessing proliferative activity than either TLI or flow cytometry. Molecular biologic technology shows real promise for the future in the area of breast cancer research. HER-2 neu amplification or overexpression, as determined by So~thern or Western blot techniques, has been the subject of mtense recent investigation. Slamon el al. 19 reported a correlation of HER-2 oncogene amplification with poorer prognosis. As techniques become more standardized, it appears as though this variable is highly predictive of recurrence in node-positive patients but fails to act as a discriminant in the node-negative population where this information is critically needed. It is possible that the lack of significance in the node-negative population relates to the small number of patients relapsing, especially on a sho.rt-term basis. Thus, it is possible that this prognostic vanable may take on significance as larger numbers of patients are followed over a longer period of time. Cathepsin D, a proteolytic enzyme widely distributed in human tissue, is an estrogen-regulated autocrine mit~gen. ~t may become a clinically relevant prognostic predictor 10 node-negative disease, as recently reported by T.andon ~l al. 20 Among 199 women with node-negative disease, high levels of cathepsin D significantly predicted reduced disease-free survival and overall survival. An elevated level of cathepsin D was the most important independent factor in predicting shortened disease-free survival and overall survival in a multivariate analysis of patients with node-negative disease. Future Directions Clinicians caring for Stage 1 breast cancer patients are keenly aware of risks and benefits in treating this generally good prognosis group of patients. Although a great deal of information is currently available on prognostic factors a clinically relevant and accurate prognostic index or serie~ of individual prognostic indices remains to be developed. This goal should be accomplished but will require future research examining the most promising prognostic factors
No.6
ADJUVANT THERAPY FOR NODE-NEGATIVE BREAST CANCER
as a group applied to a large number of patients. The AJCC and the Commission on Cancer are working with private industry to develop a pocket-sized electronic stager. The AJCC plans to convene a group of experts in prognostic factors for breast cancer in the near future to examine the feasibility of developing and testing a prognostic index for breast cancer, particularly Stage 1. If this cou ld be programmed into a pocket stager and risk for relapse quantified for the clinician, perhaps a more objective decision regarding adjuvant therapy could be reached between clinician and patient. REFERENCES I. Fisher B, Costantino J, Redmond C el al. A randomized clinical trial evaluating tamox.ifen in the treatment of patients with node negative breast cancer who have estrogen-receptor positive tumors. N Engll Med 1989; 320:479-484. 2. Mansour EG, Gray R. Shatila AH el al. Efficacy of adjuvant chemotherapy in high-risk node negative breast cancer. N Engl 1 Med 1989; 320:485-490. 3. Fisher B, Slack NH , Bross (OJ el al. Cancer of the breast: Size of neoplasm and prognosis. Cancer 1969; 24: I071-1 080. 4. Nemoto T , Vana J, Bedwani RN el al. Management and survival of female breast cancer: Results of a national survey by the American College of Surgeons. Cancer 1980; 45:2917-2924. 5. Carter CL, Allen C, Henson DE. Relation of tumor size, lymph node status and survival in 24,740 breast cancer cases. Cancer 1989; 63: 181-187. 6. Fisher ER. Prognostic and therapeutic significance of pathological features of breast cancer. NCI Monogr 1986: 1:29-34. 7. Rosen PP, Groshen S, Saigo PE el al. A long-term follow-up study of survival in stage 1 (TI NoMo) breast carcinoma. 1 Clin Oncol 1987; 7: 3SS- 366.
Winchester
1743
8. Carbone PP, Nixon OW, Fennelly J el al. Adjuvant systemic therapy. Cancer 1990; (Suppl) 6S:21 08-21 09. 9. Winchester DP. Personal communication, 1990. 10. Fisher B, Redmond C, Fisher ER el al. Relative worth of estrogen or progesterone receptor and pathologic characteristics of differentiation as indicators of prognosis in node negative breast cancer patients: Findings from National Surgical Adjuvant Breast and Bowel Project Protocol B-
06.1 Ciin Onco11988 ; 6:1076-1087 . II. Kinne OW, Ashikari R, Butler A el al. Estrogen receptor protein in breast cancer as a predictor of recurrence. Cancer 1981' 47:23642367. ' 12. McGuire WL, Clark GM , Dressler LG el al. Role of steroid hormone receptors as prognostic factors in primary breast cancer. NCI MOllogr 1986; I: 19-23. 13. McGuire WL, Tandon AK, Allred DC el al. Commentaries: How to use prognostic factors in axillary node negative breast cancer patients. lNCI 1990; 82:1006-10IS. 14. Hedley OW. Flow cytometry using paraffin-embedded tissue: Five years on. Cylometry 1989; 10:229-241 . I S. Dressler LG, Seamer LC, Owens MA el al. D A flow cytometry and prognostic factors in 1331 frozen breast cancer specimens. Ca/lCer 1988; 6 1:420-427. 16. K~lIioniemi OP, Blanco G, Alavaikko M el al. Improving the prognostiC value of DNA flow cytometry in breast cancer by combining DNA index and S-phase fraction : A proposed classification of DNA histograms in breast cancer. Cancer 1988: 62:2183-2190. 17. Clark GM, Dressler LG, Owens MA el al. Prediction of relapse or survival in patients with node negative breast cancer by DNA flow cytometry. N Engll Med 1989; 320:627-633 . . 18. Mc~uire WL, Dressler LG. Emerging impact of flow cytometry 10 predlctlOg recurrence and survival in breast cancer patients. JNCI 1985; 7S:40S-4 1O. 19. Siamon DJ, Clark GM, Wong SG el al. Human breast cancer: Correlation of relapse and survival with amplification of the HER-2 neu oncogene. Science 1987; 23S: 177-182. 20. Tandon AK, Clark GM, Chamness GC el al. Cathepsin D and prognosis in breast cancer. N Engl 1 Med 1990; 322:297-302.
T
HE CLINICAL ALERT released by the National Cancer
Institute in May 1988 emphasized the importance of considering adjuvant therapy in women with nodenegative breast cancer. Ten-year survival rates of 65% to 80% indicated far too many treatment failures for this subset of patients receiving local therapy only. linical trial data demonstrated significant disease-free survival prolongation without overall survival benefit when either adjuvant chemotherapy or tamoxifen were admi nistered. ' ·2 A number of issues arose in response to this widespread comm unication . The Program Committee proposed that an appropriate forum for discussion of these issues would be at a National onference on Ca ncer and the Changing HeaIthcare System. One of the major questions related to an assessment of benefit versus risk in this generally good prognosis group. How many patients would be treated unnecessarily to benefit others? Is benefit best defined by disease-free survival? an high- or low-risk Presented at the American Cancer ociety ational onference on ancer and the Changing Healthcare ystem, an Fra ncisco, alifornia. May 3-5. 1990. From the Depart ment of urgery. orthwestern University Medical enter. va nston, Illinois. Address for reprints: David P. Winchester. MD, FA S, Department of Surgery, Evanston Hospital, 2650 Ridge Ave nu e. Eva nston. IL 6020 I. Accepted for publication eptembcr 14, 1990.
1741
patients be identified by any prospective methods? What is a reasonable threshold number for treatment failures to initiate a recommendation for adjuvant therapy? This report considers an ever-expanding number of prognostic factors being developed by laboratorians and considered by clinicians as they try to make their best recommendations for their patients. The Importance of Accurate Staging The topic of adjuvant therapy for node-negative breast cancer was chosen to correspond to clinical alert terminology. The term is misleading and fails to accurately depict stage. After years of negotiation between the UICC and AJ C, there is worldwide agreement on the anatomic staging of this and most other cancers. The term "nodenegative" fails to take tumor size into account. The controversy surrounding adjuvant treatment is primarily focused on T I NOMO, Stage I patients. Most published series dealing with the subject of nOde-negative breast cancer include some patients with T 2 and T3 primary tumors, placing them in Stages 11 or III. Few would argue that these patients should not recei e adju ant therapy. Meaningful comparison of data can only be made if uniform staging is utilized . Future clinical trials hould keep this simple, yet critically important, fact in mind.
1742
CANCER
March 15 Supplement 1991
Time-Tested, Available, and Cost-Effective Prognostic Factors The surgical pathologist can provide a wealth of prognostic information to the managing physician. Numerous · 3-6 have shown that tumor size, nuclear grade and stu d les histologic grade are important prognostic determin~nts. Some cell types, such as tubular, confer a good prognosis. Tumor necrosis and lymphatic and blood vessel invasion are unfavorable findings. 6 The mitotic rate of a tumor is an important component of grading. Rosen et af.? demonstrated a particularly strong association between tumor size in Stage I patients and long-term survival. When stratified by tumor size, T I NOMO patients with tumors 1.0 cm or less in diameter had a significantly better 20year recurrence-free survival (86%) than did T INOMO patients with tumors 1.1 to 2.0 cm (69%). Many medical oncologists do not administer adjuvant therapy to the latter subset of patients as reported by Carbone et at. at an American Cancer Society workshop on the Management of Limited Breast Cancer. s Thus, careful attention to pathologic TNM stage and histopathologic features can provide key clinical information. The aforementioned factors should be widely available at a reasonable cost. Years ago, hormone receptor determination was a promising laboratory examination. Today's standards for quality assurance being developed by the Joint Commission on Accreditation of Healthcare Organizations may well require documentation that estrogen receptor values are determined for patients with invasive breast carcinoma over .1.0 cm in diameter. 9 Many institutions can accurately determine receptor status with immunohistochemical techniques, using much smaller amounts of tumor. In patients with node-negative disease, the value of estrogen receptor determination is controversial. 10, 1I The largest series reported by Fisher et af. 10 and McGuire et at. 12 demonstrated that estrogen receptor was a highly significant prognostic determinant in node-negative patients. Most clinicians regard estrogen receptor negativity as a poor prognostic factor, which may lead them to recommend adjuvant therapy. However, the reports on estrogen receptor represent a classical example of the difference between statistical significance and clinical pertinence. McGuire et al. 13 indicated that despite a highJy significant statistical separation, the magnitude of the difference over a long period of observation was only 8% to 9%. Other Prognostic Factors in Evolution Ploidy and S-phase fraction have been the subject of numerous recent reports. 14- 18 Ploidy alone mayor may not be a significant prognostic indicator for Stage r or ~ode-negative subsets of patients. However, S-phase frachon alone or with ploidy is more consistently predictive
Vol. 67
of recurrence. Diploid, low S-phase patients generally do well, whereas high S-phase fraction, diploid or aneuploid connotes a significantly higher likelihood of recurrence. Considerable variability exists, however, from laboratory to laboratory in S-phase fraction determination. Laboratories seek a separation point between low and high Sphase that will provide the best prognostic differentiation. Fu~ure st~dies in these laboratories should test the separahon pomt prospectively. Thymidine labeling index (TLI) is a more accurate metho~ for determi~ng S-phase, but it is time-consuming, expenSive, and reqUires fresh tissue. It is not likely to reach widespread clinical applicability. Histochemical identification of~-phase specific antigens may offer a more rapid and practIcal method for assessing proliferative activity than either TLI or flow cytometry. Molecular biologic technology shows real promise for the future in the area of breast cancer research. HER-2 neu amplification or overexpression, as determined by So~thern or Western blot techniques, has been the subject of mtense recent investigation. Slamon el al. 19 reported a correlation of HER-2 oncogene amplification with poorer prognosis. As techniques become more standardized, it appears as though this variable is highly predictive of recurrence in node-positive patients but fails to act as a discriminant in the node-negative population where this information is critically needed. It is possible that the lack of significance in the node-negative population relates to the small number of patients relapsing, especially on a sho.rt-term basis. Thus, it is possible that this prognostic vanable may take on significance as larger numbers of patients are followed over a longer period of time. Cathepsin D, a proteolytic enzyme widely distributed in human tissue, is an estrogen-regulated autocrine mit~gen. ~t may become a clinically relevant prognostic predictor 10 node-negative disease, as recently reported by T.andon ~l al. 20 Among 199 women with node-negative disease, high levels of cathepsin D significantly predicted reduced disease-free survival and overall survival. An elevated level of cathepsin D was the most important independent factor in predicting shortened disease-free survival and overall survival in a multivariate analysis of patients with node-negative disease. Future Directions Clinicians caring for Stage 1 breast cancer patients are keenly aware of risks and benefits in treating this generally good prognosis group of patients. Although a great deal of information is currently available on prognostic factors a clinically relevant and accurate prognostic index or serie~ of individual prognostic indices remains to be developed. This goal should be accomplished but will require future research examining the most promising prognostic factors
No.6
ADJUVANT THERAPY FOR NODE-NEGATIVE BREAST CANCER
as a group applied to a large number of patients. The AJCC and the Commission on Cancer are working with private industry to develop a pocket-sized electronic stager. The AJCC plans to convene a group of experts in prognostic factors for breast cancer in the near future to examine the feasibility of developing and testing a prognostic index for breast cancer, particularly Stage 1. If this cou ld be programmed into a pocket stager and risk for relapse quantified for the clinician, perhaps a more objective decision regarding adjuvant therapy could be reached between clinician and patient. REFERENCES I. Fisher B, Costantino J, Redmond C el al. A randomized clinical trial evaluating tamox.ifen in the treatment of patients with node negative breast cancer who have estrogen-receptor positive tumors. N Engll Med 1989; 320:479-484. 2. Mansour EG, Gray R. Shatila AH el al. Efficacy of adjuvant chemotherapy in high-risk node negative breast cancer. N Engl 1 Med 1989; 320:485-490. 3. Fisher B, Slack NH , Bross (OJ el al. Cancer of the breast: Size of neoplasm and prognosis. Cancer 1969; 24: I071-1 080. 4. Nemoto T , Vana J, Bedwani RN el al. Management and survival of female breast cancer: Results of a national survey by the American College of Surgeons. Cancer 1980; 45:2917-2924. 5. Carter CL, Allen C, Henson DE. Relation of tumor size, lymph node status and survival in 24,740 breast cancer cases. Cancer 1989; 63: 181-187. 6. Fisher ER. Prognostic and therapeutic significance of pathological features of breast cancer. NCI Monogr 1986: 1:29-34. 7. Rosen PP, Groshen S, Saigo PE el al. A long-term follow-up study of survival in stage 1 (TI NoMo) breast carcinoma. 1 Clin Oncol 1987; 7: 3SS- 366.
Winchester
1743
8. Carbone PP, Nixon OW, Fennelly J el al. Adjuvant systemic therapy. Cancer 1990; (Suppl) 6S:21 08-21 09. 9. Winchester DP. Personal communication, 1990. 10. Fisher B, Redmond C, Fisher ER el al. Relative worth of estrogen or progesterone receptor and pathologic characteristics of differentiation as indicators of prognosis in node negative breast cancer patients: Findings from National Surgical Adjuvant Breast and Bowel Project Protocol B-
06.1 Ciin Onco11988 ; 6:1076-1087 . II. Kinne OW, Ashikari R, Butler A el al. Estrogen receptor protein in breast cancer as a predictor of recurrence. Cancer 1981' 47:23642367. ' 12. McGuire WL, Clark GM , Dressler LG el al. Role of steroid hormone receptors as prognostic factors in primary breast cancer. NCI MOllogr 1986; I: 19-23. 13. McGuire WL, Tandon AK, Allred DC el al. Commentaries: How to use prognostic factors in axillary node negative breast cancer patients. lNCI 1990; 82:1006-10IS. 14. Hedley OW. Flow cytometry using paraffin-embedded tissue: Five years on. Cylometry 1989; 10:229-241 . I S. Dressler LG, Seamer LC, Owens MA el al. D A flow cytometry and prognostic factors in 1331 frozen breast cancer specimens. Ca/lCer 1988; 6 1:420-427. 16. K~lIioniemi OP, Blanco G, Alavaikko M el al. Improving the prognostiC value of DNA flow cytometry in breast cancer by combining DNA index and S-phase fraction : A proposed classification of DNA histograms in breast cancer. Cancer 1988: 62:2183-2190. 17. Clark GM, Dressler LG, Owens MA el al. Prediction of relapse or survival in patients with node negative breast cancer by DNA flow cytometry. N Engll Med 1989; 320:627-633 . . 18. Mc~uire WL, Dressler LG. Emerging impact of flow cytometry 10 predlctlOg recurrence and survival in breast cancer patients. JNCI 1985; 7S:40S-4 1O. 19. Siamon DJ, Clark GM, Wong SG el al. Human breast cancer: Correlation of relapse and survival with amplification of the HER-2 neu oncogene. Science 1987; 23S: 177-182. 20. Tandon AK, Clark GM, Chamness GC el al. Cathepsin D and prognosis in breast cancer. N Engl 1 Med 1990; 322:297-302.
