|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Changes in the 2003 American Joint Committee on Cancer Staging for Breast Cancer Dramatically Affect Stage-Specific Survival
From the Departments of Radiation Oncology, Biomathematics, Surgical Oncology, and Breast Medical Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX. Address reprint requests to Thomas A. Buchholz, MD, Department of Radiation Oncology, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Unit 97, Houston, TX 77030; e-mail: tbuchhol{at}mdanderson.org.
Purpose: To evaluate how implementation of the 2003 American Joint Committee on Cancer (AJCC) staging system will affect stage-specific survival of breast cancer patients. Patients and Methods: Records of 1,350 patients treated on sequential institutional protocols with mastectomy and adjuvant doxorubicin-based chemotherapy were reviewed. Pathologic stage was assigned retrospectively according to the 1988 and the 2003 AJCC staging criteria. Overall stage-specific survival (OS) was calculated using the Kaplan-Meier method, and hypothetical differences were compared by the log-rank test. Results: Six hundred five of 1,087 patients with stage II disease according to the 1988 classification system had stage II disease according to the 2003 system. The 10-year OS for patients with stage II disease was significantly improved using the 2003 system (76% [2003] v 65% [1988]; P < .0001). Two hundred eighty-nine of 633 patients with stage IIb disease using the 1988 system were stage IIb with the 2003 system, and 10-year OS was 58% (1988) versus 70% (2003; P = .003). The number of patients with stage III disease increased from 207 (1988) to 443 (2003), and the 10-year OS changed from 45% (1988) to 50% (2003; P = .077). Most of this difference resulted from changes within stage IIIa: OS, 45% (1988) versus 59% (2003; P < .0001). Conclusion: Stage reclassification using the new AJCC staging system for breast cancer will result in significant changes in reported outcome by stage. It is imperative that careful attention is devoted to this effect so that accurate conclusions regarding the efficacy of new treatment strategies can be drawn.
TUMOR STAGING systems provide information about extent of disease that can be used to guide treatment recommendations and provide estimates of patient prognosis. In addition, the staging system provides a framework for reporting treatment outcomes and thereby permits the efficacy of new treatments to be assessed. Changes in the staging system are periodically required to incorporate new diagnostic and therapeutic advances that affect risks of disease recurrence and patient survival. The American Joint Committee for Cancer (AJCC) has recently published new staging criteria for breast cancer, which were implemented in January 2003.1 In this study, we examined the impact that changing the staging criteria will have on reporting stage-specific outcomes for breast cancer. We compared the stage-specific overall survival of 1,350 patients treated on prospective adjuvant chemotherapy breast cancer protocols at the University of Texas M.D. Anderson Cancer Center with mastectomy and doxorubicin-based chemotherapy using both the 2003 and 1988 AJCC staging criteria. We found that the change in the staging system had a dramatic effect on the reporting of stage-specific survival outcomes for breast cancer.
Patient, Tumor, and Treatment Characteristics We retrospectively reviewed the records of 1,501 patients with breast cancer who were treated with mastectomy and doxorubicin-based adjuvant systemic therapy with or without tamoxifen or radiation in five prospective clinical trials at the University of Texas M.D. Anderson Cancer Center between 1975 and 1994.2–8 Each protocol was reviewed and approved by the institutional review board, and participants gave written informed consent. One hundred fifty-one patients treated on these protocols were excluded from this analysis because insufficient information was available to assign a pathologic stage in both the 1988 and 2003 AJCC staging classifications, leaving a total of 1,350 patients who were assessable for this study. Each patient was retrospectively assigned a 1988 and a 2003 disease stage based on pathologic data from the chart.
End Points and Statistical Analysis
Staging Subgroups The AJCC pathologic staging criteria for breast cancer from 1988 and 2003 are listed in Table 1  . Table 2 demonstrates the distribution of patients in each stage subgroup using the 1988 and the 2003 staging systems. In general, the 2003 staging system shifted higher-risk patients from the stage II group into the stage III group. Using the 1988 staging system, 1,087 patients had stage II disease, whereas only 605 of these patients remained in stage II using 2003 system. The number of patients with stage III disease increased from 207 (1988) to 701 (2003). The number of patients in stage IIa decreased from 454 (1988) to 316 (2003), and the number of patients in stage IIb decreased from 633 (1988) to 289 (2003). Conversely, the number of patients in stage IIIa increased from 196 (1988) to 436 (2003). The number of patients with stage IIIb was relatively unchanged, and 258 patients were classified as stage IIIc (2003), a stage that did not exist in the 1988 system. Because most stage changes manifested in this cohort of patients were based on the number of pathologically involved axillary nodes, the 2003 staging criteria did not affect the number of patients in stage I (node-negative disease).
Overall Survival Figures 1A through 1C show overall survival curves for the patients with stage IIa, IIb, and IIIa disease defined according to the 1988 staging system. For each of these figures, the overall survival curves were then reclassified into stages IIa, IIb, and IIIa according to the 2003 system. Within each of the 1988 stage groupings, there was a highly significant difference in overall survival between the new stage groupings (P < .001, P = .0002, and P = .0013 for stage IIa, IIb, and IIIa, respectively).
Figures 2A through 2D superimpose the Kaplan-Meier stage-specific overall survival curves in the two staging systems. For each stage of disease, the patients whose stage was assigned according to the 2003 system had a better overall survival than those staged according to the 1988 standard. The changes in the 10-year and 15-year actuarial overall survival by stage are listed in Table 2 .
Regarding the topic of geographic migration during the Great Depression of the 1930s, humorist-philosopher Will Rogers once said, "When the Okies left Oklahoma and moved to California, they raised the intellectual level in both states."10 This effect has come to be known as the Will Rogers phenomenon and describes an important source of bias in clinical research. Feinsten et al10 first suggested that this phenomenon could cause improvement in survival rates between two different groups without there actually being a change in individual outcome. In this report, we have shown that the recent change in the breast cancer staging system has this effect. In our data, implementing the 2003 AJCC staging for breast cancer improved stage-specific overall survival by as much as 15%. Furthermore, the improvement in stage-specific overall survival for patients staged according to the 2003 staging system applied across nearly every disease stage. The data we report used the same group of patients for comparison, so our data obviously do not reflect a change in the inherent survival of the group. Rather, the result reflects a shifting of the population in each stage towards more advanced stages. As the poorer prognostic cohort of each stage was shifted toward a higher stage, the overall survival for any one stage was improved.
The data presented in Figure 1
The data we present do not imply that the new staging system is not an important improvement. Indeed, Bunnell and Winer11 conclude their editorial on the 2003 system by saying, "The stage is set for a time when we better understand the heterogeneity of breast cancer and use this information in making treatment decisions." Our data indicate, rather, that comparisons among patients staged with the different staging systems will be inaccurate and may be inappropriately interpreted as reflecting improvements in treatment efficacy when none exists. In addition, ongoing randomized trials that span the transition from the old to the new staging system need to consider these findings and consistently stage patients according to one rather than two systems. Finally, it is recommended that during this time of transition, investigational protocols and treatment reports should also include a clear statement about which staging system is used. Our review was limited by the diagnostic tests available at the time these patients were diagnosed, and so we do not have adequate information on whether the method of histologic or diagnostic detection of lymph nodes affects outcome. The patients whose outcome we analyzed were staged before the use of immunohistochemistry to detect micrometastatic disease in lymph nodes, and the patients in this report did not undergo routine computed tomography, ultrasound examination, or lymphoscintigraphy to detect abnormal internal mammary or infraclavicular nodes. As a result, the stage of very few patients were reclassified on the basis of clinically positive internal mammary nodes or infraclavicular nodes, and no patients were restaged on the basis of micrometastatic nodal disease or the method of detecting lymph node disease. In addition, because patients with stage IV disease were excluded from participation in these protocols, few patients with positive supraclavicular lymph nodes are included in our analysis. For these reasons, these data reflect restaging primarily as a result of incorporating the number of positive lymph nodes into the staging system and may therefore be an underestimate of the changes in overall survival using the 2003 staging system. Lastly, it should be noted that relatively few stage I patients were enrolled in these protocols. This impact is not likely to impact population-based outcomes because this reporting is not stage-specific. In conclusion, a significant percentage of patients whose 1988 stage of disease was IIa and IIb will be reclassified into higher disease stages using the 2003 staging system. Higher-risk patients are removed from the stage II groups, whereas the stage III groups seem to have an increased proportion of favorable tumors. This results in patients with an assigned 2003 stage of IIa/IIIa having significantly improved stage-specific overall survival compared with cohorts with similar stage disease defined by the 1988 staging system. It is imperative that careful attention is devoted to this effect so that accurate conclusions regarding the efficacy of new treatments can be drawn.
We thank Jessica Erwin and Ramani Krishnan for their contributions to the database.
Supported in part by grant Nos. CA16672 and T32CA77050 from the National Cancer Institute, the Nellie B. Connally Breast Cancer Research Fund, and a grant from the Stanford and Joan Alexander Foundation, Houston, TX. T.A.B. is supported by Department of Defense Breast Cancer Research Program Career Development Award, BC980154.
1. Singletary SE, Allred C, Ashley P, et al: Revision of the American Joint Committee on Cancer staging system for breast cancer. J Clin Oncol 20:3628–3636, 2002 2. Blumenschein GR, Buzdar AU, Hortobagyi GN: FAC + BCG as adjuvant therapy in breast cancer: An 8-year update. Recent Results Cancer Res 96:129–132, 1984[Medline] 3. Buzdar AU, Blumenschein GR, Smith TL, et al: Adjuvant chemotherapy with fluorouracil, doxorubicin, and cyclophosphamide, with or without Bacillus Calmette-Guerin and with or without irradiation in operable breast cancer: A prospective randomized trial. Cancer 53:384–389, 1984[CrossRef][Medline] 4. Buzdar AU, Hortobagyi GN, Marcus CE, et al: Results of adjuvant chemotherapy trials in breast cancer at M.D. Anderson Hospital and Tumor Institute. NCI Monogr 81–85, 1986 5. Buzdar AU, Hortobagyi GN, Smith TL, et al: Adjuvant therapy of breast cancer with or without additional treatment with alternate drugs. Cancer 62:2098–2104, 1988[CrossRef][Medline] 6. Buzdar AU, Hortobagyi GN: Ten-year results of FAC adjuvant chemotherapy trial in breast cancer. Breast Cancer Res Treat 13:117–121, 1989[Medline]
7. Buzdar AU, Sahin AA, Ross MA: Adjuvant therapy with escalating doses of doxorubicin and cyclophosphamide with or without leukocyte alpha-interferon for stage II or III breast cancer. Ann Intern Med 117:771–777, 1992 8. Theriault R, AU B, GN H: Irradiation (XRT) following mastectomy in patients treated with FAC adjuvant therapy: M.D. Anderson Cancer Center experience. Proc Am Soc Clin Oncol 17, 1998 (abstr) 9. Harris E, Albert A: Survivorship Analysis for Clinical Studies. New York, NY, Dekker, 1991 10. Feinstein AR, Sosin DM, Wells CK: The Will Rogers phenomenon: Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med 312:1604–1608, 1985[Abstract]
11. Bunnell CA, Winer EP: Lumping versus splitting: The splitters take this round. J Clin Oncol 20:3576–3577, 2002 Submitted March 10, 2003; accepted June 18, 2003.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
