95 Outcome of Patients with Metastatic Breast Carcinoma Treated at a Private Medical Oncology Clinic William F. Anderson, M.D., M.P.H.1 James E. Reeves, M.D.2 Abdalla Elias, M.D.3 Hans (J) Berkel, M.D., Ph.D.4 1 National Cancer Institute/Division of Cancer Prevention, Preventive Oncology Branch, Bethesda, Maryland. 2 Cancer Care Associates, Oklahoma City, Oklahoma. 3 Pathology Department, Glenwood Regional Medical Center, West Monroe, Louisiana. 4 Section of Cancer Control and Prevention, FeistWeiller Cancer Center, Louisiana State University Medical Center-Shreveport, Shreveport, Louisiana. BACKGROUND. Metastatic breast carcinoma usually is fatal. The median survival after recurrence is 2 years. Five-year survival after recurrence is approximately 10 –20%. Although encouraging data have been reported from clinical trials, two issues confound the translation of clinical research to the general medical community: patient selection bias and the absence of untreated controls. Therefore, the authors examined their community-based metastatic breast carcinoma patients, comparing their results with reports from clinical trials and with untreated historic controls. METHODS. This was a nonconcurrent cohort study of 407 community-based metastatic breast carcinoma patients treated at the authors’ private medical oncology clinic in northeast Louisiana. Prognostic variables were correlated with survival time. RESULTS. The median age of the patients at the time of diagnosis of breast carcinoma was 61 years. The median age at the time of first recurrence was 65 years. The median overall survival and median survival after recurrence were 4.5 years and 20.1 months, respectively. The 5-year and 10-year survival rates were 17% and 2.5%, respectively. CONCLUSIONS. Although clinical trial participants and the authors’ private practice cohort are incommensurable, outcomes in these disparate groups were similar but, unfortunately, not much different from those of untreated historic controls. Cancer 2000;88:95–107. © 2000 American Cancer Society. KEYWORDS: metastatic breast carcinoma, chemotherapy, endocrine therapy, patient selection bias, generalizability. B For their review, comments, and suggestions, the authors are indebted to Douglas L. Weed, M.D., Ph.D. (NCI/DCP/Preventive Oncology Branch, Bethesda, MD) and Kenneth C. Chu, Ph.D. (NCI/OSPR, Bethesda, MD). Address for reprints: William F. Anderson, M.D., M.P.H., NCI/DCP/Preventive Oncology Branch, Executive Plaza South, Suite T41, 6120 Executive Boulevard, MSC 7105, Bethesda, MD 208927105. Received February 18, 1999; revision received May 26, 1999; accepted July 12, 1999. © 2000 American Cancer Society reast carcinoma is a major public health problem. Although modern methods of early detection, intervention, and postoperative treatment have improved the prognosis for patients with primary breast carcinoma,1,2 not all patients are cured. Some women present de-novo with metastatic breast carcinoma (MBC), whereas others develop recurrent MBC. Historically, the median survival for MBC patients is a modest 2 years, with 1–3% of patients achieving a long term cure.3–7 Front-line, conventional dose anthracycline and alkylating agent-based chemotherapy protocols produce 50 –90% overall response rates,3 but responses usually are temporary. For patients with MBC who are resistant to standard cytotoxic chemotherapy, there is a 20 –25% partial response with salvage treatment8 and little hope of a cure.9 –13 Consequently, we are witnessing an enormous effort to increase the treatment options for these patients with novel drug combinations and treatment strategies.14 –20 The medical literature reports encour- 96 CANCER January 1, 2000 / Volume 88 / Number 1 aging data from both standard dose and high dose clinical trials;5,16,21 however, clinical research may not transfer to community-based breast carcinoma patients.22–24 Randomized trials have two serious flaws that threaten extrapolation to the general medical community: patient selection bias and the lack of untreated controls. To ensure patient safety and internal validity, clinical trial participation is dependent on rigorous eligibility criteria. These criteria may introduce systematic patient selection favoring the young, healthy, wealthy, and chemosensitive patient who is willing and able to travel for treatment.5,6,25–29 Although the experimental method of the randomized clinical trial is the preferred study design for internal validity and is the “gold standard” for evaluating therapeutic efficacy,30,31 patient selection bias threatens external validity and generalizability to the global breast carcinoma population. In addition and for many appropriate reasons, MBC clinical trials do not include untreated reference patients. Although the strongest possible evidence of therapeutic effect would come from a prospective trial that randomized a test regimen against symptomatic care,32 such a study would be unethical. Nonetheless, missing control data confounds extrapolation to the public, because untreated patients provide the backdrop against which we judge the true benefits of a therapeutic response.23,33 We are aware of the limitations of historic controls;34 however, without the notreatment control arm, we have lost a valuable means of evaluating clinical research.35 Therefore, we compared the outcomes of patients with MBC in our community-based medical oncology clinic with the outcomes of untreated historic controls and with the results from clinical trials. MATERIALS AND METHODS The Hematology/Oncology Clinic was a communitybased, private medical oncology clinic that operated in northeast Louisiana for 14 years, from 1981 to 1995. In 1990, we established the Hematology/Oncology Database (HODB), an observational data bank containing every breast carcinoma patient treated in the Hematology/Oncology Clinic. Data were collected retrospectively for patients who were diagnosed prior to 1990 and prospectively thereafter. For each patient, demographic as well as clinicopathologic prognostic information was obtained from multiple sources, including patient history and physical examinations, office and hospital records, roentgenographic and pathologic reports, and data from the Northeast Louisiana Regional Tumor Registry. For women who had not received long term fol- low-up in the clinic, self-administered mail surveys were obtained annually. Nonrespondents were contacted by telephone. Reported recurrences were confirmed with the patient and/or her physician. Roentgenographic and pathologic reports corroborated all recurrences, when available. We used the 1977 International Union Against Cancer criteria to determine responses to MBC treatment.36 A complete response (CR) was the disappearance of all known disease with no evidence of disease (NED). A partial response (PR) required a decrease . 50% in measurable lesions and no new lesions. No response showed no change or could not be determined. Patients Most of the patients in the Hematology/Oncology Clinic were assembled through physician referral: A few were accrued through self-referral. Consistent with a “real world” private medical oncology clinic, we nonconcurrently managed patients with primary breast carcinoma or MBC. We excluded no one from private practice or from the HODB, but this report includes only those women with MBC (n 5 407 women) at the time of last follow-up (December, 1998). Although all MBC patients visited the clinic sometime during the period 1981–1995, the dates of initial diagnoses and recurrence overlapped this period. The initial dates of diagnosis of primary breast carcinoma ranged from April 1958 to March 1995. The dates of systemic recurrence varied from July 1974 to November 1998. The 407 MBC patients presented to our clinic in three different ways. One group (n 5 61 patients; 15%) presented with de novo TNM Stage IV disease.37 A second group (n 5 188 patents; 46%) presented to the clinic with systemic recurrence. For this second group, primary breast carcinoma care was given elsewhere that, in some instances, preceded recurrence by many years. The median disease free interval (DFI; date of recurrence minus the date of diagnosis of primary breast carcinoma) was 3.2 years (range, , 1 year to 27 years). Twenty-five patients experienced disease recurrence .10 years after the date of diagnosis of primary breast carcinoma. Past medical history and past medical records documented initial pathologic diagnoses, prognostic factors, and adjuvant treatment. A third group (n 5 158 patients; 39%) came to our clinic without MBC but suffered a later systemic recurrence. Many patients in this last group received adjuvant hormonal and/or chemotherapy in our office, whereas others were followed without postoperative treatment. However, none of the patients in the third group had MBC on first presentation to our clinic; that is, they all Metastatic Breast Carcinoma/Anderson et al. 97 FIGURE 1. Kaplan–Meier productlimit analysis of 407 patients with metastatic breast carcinoma (MBC). experienced disease recurrence after their initial visit to the Hematology/Oncology Clinic. In this third group, MBC was not always treated in the Hematology/Oncology Clinic; that is, some patients received therapy in other clinics or in referral centers. Notwithstanding these varied presentations, the 407 MBC patients from the HODB were analyzed as a single cohort. The date of diagnosis of primary breast carcinoma was the starting point for this study. The date of study closure (or the date of last follow-up) was December 1998. Demographic and Clinicopathologic Characteristics Demographic and clinicopathologic data were examined as adjusted and unadjusted categoric variables. Demographic information was patient age at the time of initial diagnosis (# 50 years vs. . 50 years) and race (black vs. white). Patient age . 50 years was our surrogate measure for postmenopausal status. Traditional clinicopathologic prognostic factors included TNM tumor size (T4, T3, T2, and T1) and lymph node status (. 3, 1–3, and 0 positive axillary lymph nodes). Estrogen receptor status was coded as positive versus negative. Chemotherapy and hormonal treatments both before (adjuvant) and after disease recurrence were coded yes versus no. Chemotherapy was classified further as cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) or doxorubicin-based chemotherapy. Outcome For each patient, we recorded the date of birth, the date of diagnosis of primary breast carcinoma, the date of disease recurrence, and the date of death or of last contact if the patient was still living. Primary outcome measures included survival times: DFI, survival after recurrence (SAR), and overall survival (OS). DFI was the time between the date of diagnosis of primary breast carcinoma and the date of first systemic recurrence. Women with de novo Stage IV disease had no DFI.36 The site of recurrence was categorized as either visceral (pulmonary, pleura, hepatic, brain, bone marrow, and other), bone, locoregional, or other. Breast recurrences (ipsilateral or contralateral) were not coded as recurrences. Women with breast recurrences were not censored from the study: They were followed until systemic recurrence. We calculated SAR as the interval between the date of recurrence and the date of death or of the last contact for living patients. OS was the interval between date of diagnosis and the date of death from any cause or the date of last contact. OS corrected for cause of death determined cause specific mortality. Death without evidence of MBC was coded death-NED. Statistical Analysis Data were entered into a personal computer with Microsoft Windows 98 operating system (Microsoft Corp., Redmond, WA). The database was compiled in Microsoft Access 97 for Windows. Recorded information was checked repeatedly for consistency and accuracy. Statistical analyses and graphs were generated with Statistica for Windows, version 5.1 (Statsoft, Inc., Tulsa, OK).38 OS and SAR (Fig. 1) were estimated with the Kaplan–Meier product-limit method.39 The stratified log-rank test was used to compare survival distributions in two samples (Fig. 2).40 The Cox proportional hazards model was used to generate unadjusted and adjusted hazard ratios, which are expressed as the relative risk.41 The risk of a given prognostic variable is displayed relative to the reference variable, which has 98 CANCER January 1, 2000 / Volume 88 / Number 1 FIGURE 2. Survival after recurrence for patients who developed metastatic breast carcinoma prior to 1985 and after 1990. a value of 1.00 and is the last entry of each category (Table 2). With logistic regression, we calculated unadjusted and adjusted correlations, which are expressed as odds ratios.42 Contingency tables (Tables 3 and 4) were arranged to demonstrate the odds of dependent, dichotomized variables that had a given prognostic factor relative to the reference variable; i.e., 1.00, the last entry of each category. Relative risk and odds ratios were presented with 95% confidence intervals and P values. All P values were two-sided: P values # 0.05 were considered to be statistically significant. RESULTS Seven hundred ninety-seven (797) breast carcinoma patients were treated in the Hematology/Oncology Clinic between 1981 and 1995 and were entered into the Hematology/Oncology data base (HODB). This breast carcinoma cohort comprised 13% of our total private practice patient base and 25% of all breast carcinoma patients in Northeast Louisiana, based on data from the Northeast Louisiana Tumor Registry. Currently, 390 breast carcinoma patients (49.8%) remain recurrence free. Four hundred seven patients (51.1%) have MBC and are the subject of this report. Demographic and Clinicopathologic Characteristics At the time of diagnosis of primary breast carcinoma, the median age of the patients was 61 years (range, 21– 88 years). One hundred eleven patients (27%) were age # 50 years. Almost 75% (n 5 296 patients; 73%) were age . 50 years. Three hundred thirty-two patients (82%) were white, and 74 patients (18%) were black (see Table 1). The majority of the tumors (n 5 362 tumors; 89%) were infiltrating ductal carcinoma of no special type.43 The initial median tumor greatest dimension was 3.0 cm, with a range from 0.3 cm to .10 cm. Most of the primary tumors were T1 lesions (n 5 100 tumors; 25%) or T2 (n 5 167 tumors; 41%). In 18 tumors, the tumor size was unknown or was different than T1–T4. One hundred fifty-one patients (37%) had negative axillary lymph nodes at the time of diagnosis of primary breast carcinoma. Two hundred forty-nine patients (61%) were lymph node positive: 100 with 1–3 axillary lymph nodes and 149 with . 3 axillary lymph nodes. We were able to determine the initial TNM stage in all but 7 patients, yielding a successful staging rate of 98%. Early-stage tumors (Stage 0 and I) comprised a minority of the primary breast carcinomas (n 5 45 patients; 11%). Half of the MBC cohort (n 5 202 patients; 50%) presented initially with TNM Stage II disease. Stage III disease represented a large fraction of the cohort (n 5 92 patients; 23%). Sixty-one patients had de novo Stage IV disease. One hundred seventeen (29%) tumors were estrogen receptor negative, 172 (42%) were positive, and 118 (29%) were unknown. Many of the patients with unknown hormonal receptors developed primary breast carcinoma prior to 1980, before the routine measurement of estrogen and progesterone receptors. At the time of initial systemic recurrence, the median age of patients was 65 years (range, 23–90 years). The median DFI was 25 months (range, 0 days to 27 years). The DFI was #24 months for 198 patients (49%) and .24 months for the remaining 209 patients (51%). There were 193 visceral recurrences (47%), 130 bone recurrences (32%), and 82 (20%) locoregional recurrences. Metastatic Breast Carcinoma/Anderson et al. TABLE 1 Demographic and Clinicopathologic Characteristics for 407 Patients with Metastatic Breast Carcinoma Characteristic Total Percent Sample size Age at the time of diagnosis (yrs) #50 .50 Race Black White Other Initial TNM tumor size T4 T3 T2 T1 Other or unknown Initial axillary lymph nodes .3 1–3 0 Unknown Initial TNM stage 0 I II III IV Unknown Estrogen receptor status Negative Positive Unknown Disease free interval (months) #24 .24 Site of systemic recurrence Visceral Bone Locoregional Other Adjuvant radiotherapy Yes No Unknown Adjuvant systemic therapy Yes No Unknown Response to MBC treatment Yes No Unknown 407 100.0 111 296 27.3 72.7 74 332 1 18.2 81.6 0.2 75 47 167 100 18 18.4 11.5 41.0 24.6 4.4 149 100 151 7 36.6 24.6 37.1 1.7 4 41 202 92 61 7 1.0 10.1 49.6 22.6 15.0 1.7 117 172 118 28.7 42.3 29.0 198 209 48.6 51.4 193 130 82 2 47.4 31.9 20.1 0.5 138 262 7 33.9 64.4 1.7 220 185 2 54.1 45.5 0.5 142 130 135 34.9 31.9 33.2 MBC: metastatic breast carcinoma. Treatment At the time of diagnosis of primary breast carcinoma, adjuvant radiotherapy was given to 138 patients (34%). More than half (n 5 220 patients; 54%) had adjuvant systemic hormonal therapy and/or chemotherapy. 99 One hundred forty patients (34%) received postoperative hormonal treatment: 138 with tamoxifen and 2 with oophorectomy. One hundred thirty-eight patients (34%) received adjuvant systemic chemotherapy. Chemotherapy protocols included CMF for 60 patients, doxorubicin-based regimes for 61 patients, and less common programs, containing thiotepa and melphalan, for 17 patients. At the time of initial systemic recurrence, treatment after recurrence in our clinic consisted of systemic hormonal therapy and/or chemotherapy with or without symptomatic radiotherapy. Patients with nonlife-threatening postmenopausal MBC were treated with front-line antihormonal therapy; i.e., tamoxifen or megestrol acetate. Initial systemic chemotherapy, including CMF and doxorubicin-based protocols, was reserved for premenopausal patients, visceral recurrences, or antihormonal failures. Some women received MBC treatment elsewhere both before and after closure of the Hematology/Oncology Clinic. Outside office records have not always permitted an accurate assessment of response to therapy. Other patients are actively receiving chemohormonal therapy, and it is too early to judge their response to treatment. Consequently, large amounts of MBC treatment-related data were incomplete (n 5 135 patients; 33%). As of the date of last follow-up (December 1998), there was sufficient information on 272 patients (67%) to judge their response to systemic treatment after disease recurrence. One hundred forty-two patients (35%) had either a partial or complete response, and 130 patients (32%) did not respond. Outcome Outcome-related follow-up was 100% complete. Calculated from the dates of diagnosis to the date of study closure, the median follow-up for this analysis was 13 years (range, 4 – 41 years). There were 35 living MBC patients and 372 deceased MBC patients. The median OS estimated with Kaplan–Meier plots was 4.5 years (Fig. 1). The median SAR was 20.1 months (range, 0 days to 17 years) (Fig. 1). Seventy patients (17%) achieved long term survival after recurrence, defined as SAR $ 5 years. Fifty-six long term survivors have died: 51 with MBC and 5 with NED. Ten patients (2.5%) are living $10 years after systemic recurrence. Although the treatment philosophy was consistent during the period of clinic operation, 1981–1995, we noted a temporal shift for SAR (Fig. 2). SAR worsened in later years. The SAR for women who experienced disease recurrence after 1990 and for those who 100 CANCER January 1, 2000 / Volume 88 / Number 1 experienced disease recurrence before 1985 was 15.8 months and 27.4 months, respectively (P , 0.001). vant chemotherapy (yes vs. no) was the only statistically significant variable (P , 0.001). Unadjusted and Adjusted Cox Proportional Hazards Model for SAR DISCUSSION Table 2 displays unadjusted SAR by prognostic variables. The shortened median SAR was associated with increased hazard ratio (increased relative risk) for black women compared with white women, larger tumor sizes, .3 positive axillary lymph nodes compared with negative lymph nodes, and negative estrogen receptor levels. Increased relative risk also was observed for short DFI (#24 months vs. .24 months), site of systemic recurrence (visceral and bone disease vs. locoregional disease), adjuvant chemotherapy (yes vs. no), and response to MBC treatment (yes vs. no). In a multivariate (adjusted) model, statistical significance was maintained for initial lymph node status, DFI, site of systemic recurrence, and response to MBC treatment (Table 2). Unadjusted and Adjusted Correlations for Short Term Survivors versus Long Term Survivors after Disease Recurrence We investigated short term survivors (#2 years) compared with long term survivors ($5 years) (Table 3). Univariate (unadjusted) analysis revealed that short term survivors, compared with long term survivors, had a greater likelihood of being black, having TNM T4 lesions, having .3 axillary lymph nodes, and being estrogen receptor negative. Short term survivors also had a greater likelihood of having visceral and bone recurrences versus locoregional recurrences, having had adjuvant systemic chemotherapy, and having responded to MBC treatment. In the multivariate (adjusted) model (Table 3), correlation with site of systemic recurrence, adjuvant systemic therapy, and response to MBC treatment remained statistically significant. Unadjusted and Adjusted Correlations for Early versus Late Systemic Recurrence We examined later MBC diagnosis (after 1990) versus earlier diagnosis (before 1985) with prognostic variables (Table 4). Unadjusted odds ratios demonstrated no secular difference in patient age at the time of diagnosis of primary breast carcinoma, TNM tumor size, initial axillary lymph node status, or response to MBC treatment. However, women who were diagnosed after 1990 versus women who were diagnosed prior to 1985 were more likely to be estrogen receptor negative, to have visceral and bone metastases versus locoregional metastases, and to have had adjuvant chemotherapy. In the adjusted model (Table 4), adju- MBC is a heterogeneous disease with an acknowledged 2-year median SAR. Decades of clinical research have focused on the translation of systemic treatment into the general medical community with the expectation of improved outcome. However, in our community-based oncology clinic, we noted a median SAR of ,2 years and an adverse secular trend toward shorter survival times. SAR in the Hematology/Oncology clinic was correlated with race, TNM tumor size, axillary lymph node status, estrogen receptor status, DFI, site of systemic recurrence, prior adjuvant systemic therapy, and response to MBC treatment. In multivariate models, the strongest predictors of SAR were the site of recurrence and response to MBC treatment (Table 2). For patients with MBC after 1990, SAR was 15.8 months compared with 27.4 months for women with MBC before 1985 (Fig. 2). History of adjuvant systemic therapy was the only significant variable in the adjusted model (Table 4). We suspect a causal association between adjuvant chemotherapy and the negative temporal trend. Although the MBC populations studied in earlier reviews rarely received adjuvant therapy,44 there is little doubt that SAR is compromised by adjuvant treatment.4 Adjuvant treatment may induce acquired resistance or select resistant malignant clones;45,46 however, for whatever reason, recurrence after adjuvant treatment identifies a subgroup of patients with a lower response to further therapy and shorter survival after first disease recurrence.47 After 1990, 74% of our MBC patients had a history of prior adjuvant hormonal therapy or systemic chemotherapy. Prior to 1985, 67% of the women were hormonal and/or chemotherapy naı̈ve. Similar to our study, Debonis et al. noted a negative secular trend in median survival after disease recurrence.35 SAR decreased from 31.97 months to 18.72 months from 1955 to 1980, but those authors did not comment on the effect of adjuvant treatment.35 Clark et al. observed a negative association of adjuvant treatment with SAR but reported the disappearance of statistical significance in their multivariate model.4 They postulated that the decision to administer adjuvant therapy was based on colinear factors, such as lymph node status and estrogen receptor status. However, Clark et al.’s explanation would not explain our finding, because tumor size and lymph node status were similar before 1985 and after 1990 (Table 4). The choice to administer adjuvant chemotherapy probably was related more Metastatic Breast Carcinoma/Anderson et al. 101 TABLE 2 Cox Proportional Hazards Model for Survival after Recurrencea Variable Unadjusted relative risk Sample size Age at the time of diagnosis (yrs) #50 .50 Race Black White Other Initial TNM tumor size T4 T3 T2 T1 Other or unknown Initial axillary lymph nodes .3 1–3 0 Unknown Estrogen receptor status Negative Positive Unknown Disease free interval (months) #24 .24 Site of systemic recurrence Visceral Bone Locoregional Other Adjuvant systemic therapy Yes No Other Response to MBC treatment Yes No Unknown Adjusted relative risk Race (black vs. white) Initial TNM tumor size (T4 vs. T1) Initial TNM tumor size (T3 vs. T1) Initial lymph node status (.3 vs. 0) Estrogen receptor status (negative vs. positive) Disease free interval (#24 months vs. .24 months) Site of systemic recurrence (visceral vs. locoregional) Site of systemic recurrence (bone vs. locoregional) Adjuvant systemic therapy (yes vs. no) Response to MBC treatment (yes vs. no) Total Median SAR (months) Relative risk 95% CI P value 407 20.1 — — — 111 296 23.5 18.3 0.90 1.00 0.71–1.13 — 0.35 — 74 332 1 14.8 21.8 — 1.41 1.00 — 1.08–1.84 — — 0.01 — — 75 47 167 100 18 12.6 21.0 23.2 22.4 1.87 1.54 1.06 1.00 1.35–2.58 1.07–2.21 0.81–1.38 — ,0.001 0.02 0.68 — 149 100 151 7 16.2 24.2 23.2 1.49 0.94 1.00 1.18–1.90 0.72–1.23 — ,0.001 0.64 — 117 172 118 11.6 23.3 1.56 1.00 1.22–2.01 — ,0.001 — 198 209 16.6 24.8 1.46 1.00 1.19–1.79 — ,0.001 — 193 130 82 2 13.2 22.6 33.2 1.98 1.50 1.00 1.50–2.60 1.12–2.02 — ,0.001 ,0.01 — 220 185 2 23.8 28.9 1.47 1.00 1.19–1.81 — ,0.001 — 142 130 135 38.2 10.9 — 0.24 1.00 — 0.18–0.32 — — ,0.001 — — — — — — — — — — — — — — — — — — — — — — 1.25 1.28 0.87 1.49 1.09 1.49 3.62 1.70 1.26 0.24 0.83–1.88 0.85–1.93 0.53–1.43 1.04–2.11 0.78–1.53 1.07–2.07 2.32–5.67 1.10–2.62 0.87–1.82 0.16–0.35 0.29 0.24 0.57 0.03 0.61 0.02 ,0.001 0.02 0.23 ,0.001 SAR: survival after recurrence; CI: confidence interval; MBC: metastatic breast carcinoma. a Hazard ratios are expressed as relative risk. 102 CANCER January 1, 2000 / Volume 88 / Number 1 TABLE 3 Demographic and Clinicopathologic Characteristics for Survival after Recurrence (<2 years vs. >5 years)a Survival after relapse Variable Unadjusted odds ratio Sample size Actuarial median SAR (months) Age at the time of diagnosis (yrs) #50 .50 Race Black White Initial TNM tumor size T4 T3 T2 T1 Other or unknown Initial axillary lymph nodes .3 1–3 0 Unknown Estrogen receptor status Negative Positive Unknown Site of systemic recurrence Visceral Bone Locoregional Other Adjuvant systemic therapy Yes No Other Response to MBC treatment Yes No Unknown Adjusted odds ratio Race (black vs. white) Initial TNM tumor size (T4 vs. T1) Initial axillary lymph nodes (.3 vs. 0) Estrogen receptor status (negative vs. positive) Site of systemic recurrence (visceral vs. locoregional) Site of systemic recurrence (bone vs. locoregional) Adjuvant systemic therapy (yes vs. no) Response to MBC treatment (yes vs. no) <2 yrs >5 yrs Odds ratio 95% CI P value 268 11.3 70 78.1 — — — — — ,0.001 69 199 17 53 1.08 1.00 0.58–1.99 — 0.80 — 54 214 5 65 3.28 1.00 1.25–8.57 — 0.02 — 61 34 102 60 11 3 5 37 21 4 7.12 2.38 0.96 1.00 — 2.01–25.24 0.82–6.91 0.52–1.81 — — ,0.01 0.11 0.91 — — 113 60 91 4 13 21 34 2 3.25 1.07 1.00 — 1.61–6.53 0.57–2.01 — — ,0.001 0.84 — — 97 107 64 10 31 29 2.81 1.00 — 1.30–6.06 — — ,0.01 — — 140 88 39 1 22 20 27 1 4.41 3.05 1.00 — 2.26–8.60 1.52–6.09 — — ,0.001 ,0.01 — — 167 99 2 21 49 0 3.94 1.00 — 2.22–6.96 — — ,0.001 — — 62 113 93 46 5 19 0.06 1.00 — 0.02–0.16 — — ,0.001 — — — 14.00 0.46 1.87 25.25 5.62 5.75 0.04 — 0.82–240 1.00 0.07 0.56–6.30 4.81–132.6 1.36–23.22 1.59–20.76 0.01–0.19 0.31 ,0.001 0.02 ,0.01 ,0.001 1.64 — — 0.44–6.07 — — SAR: survival after recurrence; CI: confidence interval; MBC: metastatic breast carcinoma. a Correlations are expressed as odds ratios. to the date of disease recurrence. We established the Hematology/Oncology Clinic in 1981. Many patients who experienced recurrences prior to 1985 simply were never offered adjuvant treatment for primary breast carcinoma. Our study has several weaknesses. First, the HODB was a retrospective series of consecutive cases: a longitudinal study with uncontrollable vagaries and biases.48 We bring creditability to the project with meticulous record keeping (e.g., the 100% complete Metastatic Breast Carcinoma/Anderson et al. 103 TABLE 4 Demographic and Clinicopathologic Characteristics for Temporal Shift (>1990 vs. <1985)a Variable Unadjusted odds ratio Sample size Actuarial median SAR (months) Age at the time of diagnosis (yrs) #50 .50 Race Black White Other Initial TNM tumor size T4 T3 T2 T1 Other or unknown Initial axillary lymph nodes .3 1–3 0 Unknown Estrogen receptor status Negative Positive Unknown Site of systemic recurrence Visceral Bone Locoregional Other Adjuvant systemic therapy Yes No Other Response to MBC treatment Yes No Unknown Adjusted odds ratio Race (Black vs. White) Estrogen receptor status (negative vs. positive) Site of systemic recurrence (visceral vs. locoregional) Site of systemic recurrence (bone vs. locoregional) Adjuvant systemic therapy (yes vs. no) >1990 <1985 Odds ratio 95% CI P value 154 15.8 96 27.4 — — — — — ,0.001 49 105 24 72 1.40 1.00 0.79–2.49 — 0.25 — 35 118 1 8 88 0 3.26 1.00 — 1.44–7.41 — — ,0.01 — — 35 17 59 39 4 9 11 47 22 7 2.19 0.87 0.71 1.00 — 0.89–5.42 0.35–2.20 0.37–1.36 — — 0.09 0.77 0.30 — — 63 43 47 1 37 20 36 3 1.30 1.65 1.00 — 0.72–2.37 0.83–3.28 — — 0.38 0.15 — — 49 85 20 27 21 48 0.45 1.00 — 0.23–0.88 — — 0.02 — — 80 51 22 1 39 29 28 0 2.61 2.24 1.00 — 1.32–5.15 1.08–4.62 — — 0.01 0.03 — — 115 39 0 31 63 2 5.99 1.00 — 3.40–10.55 — — ,0.001 — — 49 47 58 39 32 25 0.86 1.00 — 0.46–1.59 — — 0.62 — — — — — — — — — — — — 2.46 0.54 2.66 1.85 7.78 0.88–6.89 0.25–1.18 0.98–7.20 0.64–5.37 3.59–17.04 0.08 0.12 0.053 0.26 ,0.001 CI: confidence interval; SAR: survival after recurrence; MBC: metastatic breast carcinoma. a Correlations are expressed as odds ratios. follow-up and 98% successful TNM staging). Nevertheless, accurate records cannot correct an obvious systematic error: The Hematology/Oncology Clinic was a referral-based private practice in the rural Southeast. Our clinic physicians did not recruit, randomize, or select our cohort of patients in any way; nonetheless, the patients were chosen for us. Consulting physician preferences as well as the socioeconomic fac- tors in the community defined a vague sampling method. Although this method of patient selection is not ideal and clearly is subject to bias, patient accrual through physician referral is representative of a “realworld” situation and not is unique to our northeast Louisiana referral base. Consistent with a medical oncology private practice, patients were sent to us for the administration of endocrine therapy and/or cytotoxic chemotherapy. 104 CANCER January 1, 2000 / Volume 88 / Number 1 TABLE 5 Comparison of Survival for Metastatic Breast Carcinoma Patients Treated with Hormonal Therapy, Standard Dose Chemotherapy, and High Dose Chemotherapy Study Sample size Median age (yrs) Rx SAR (months) 3 yr SAR (%) 5 yr SAR (%) Reference San Antonio New Zealand POA E2181 MDACC ABMTR HODB 1015 141 250 501 1581 3398 407 58 53–57 57 55 51.7 44 65 Not recorded SDCT SDCT SDCT SDCT HDCT SDCT and HT 23 10–12 14.8 20.9 19 19 20.1 — 9.0 20.0 30.0 28.0 30.0 30.0 — — — 18.0 12.2 10.0 17.0 4 60 59 58 16–17 25 — Rx: treatment after recurrence; SAR: median survival after recurrence; 3 yr SAR: 3-year survival after recurrence; 5 yr SAR: 5-year survival after recurrence; POA: Piedmont Oncology Association; E2181: Eastern Cooperative Oncology Group E2181; MDACC: M.D. Anderson Cancer Center; ABMTR: Autologous Blood and Marrow Transplant Registry; HODB: Hematology/Oncology Clinic; SDCT: standard dose chemotherapy; HDCT: high dose chemotherapy; HT: hormonal therapy. Compared with most clinical trials, treatment in our clinic was conservative; however, herein is a second flaw: Large numbers of disease recurrences were treated in other clinics, and it has been difficult to obtain accurate treatment-related information from outside offices. We are unable to classify the responses to MBC treatment for 135 patients (33%). However, given the fact that the median SAR for this group is 18.1 months, which is only slightly lower than the median SAR for the entire group, it is doubtful that the details of their treatment would have altered our final conclusions substantially. It also should be noted that our main outcome measure was survival time and not response to therapy.32 Survival-related information was 100% complete even for those patients with incomplete treatment records. As of the date of study closure (December 1998), no one is lost to follow-up with regard to vital status and/or cause of death if deceased. Thus, if therapy is considered to be an integral part of the disease process, then OS and SAR can be calculated without knowledge of treatment response.4,49 Although it is difficult to compare our series of consecutive cases with other study groups, it actually is difficult to compare most breast cancer data bases. For other data bases, because, without an accepted comprehensive prognostic index,50 –52 we risk comparing “apples with oranges.” Different study groups simply are incommensurable.53 Conventional-dose and high dose chemotherapy data bases are not only different from one another; but also, and perhaps more importantly, neither is representative of the general medical community.5,6,16,17,21,25–29 For example, patient age is a major risk factor of breast carcinoma. Projected over a life span of 851 years, the risk of developing breast carcinoma is 12.8% or 1 in 8 women.54 –56 Typical breast carcinoma patients have a median age of 60 – 64 years, similar to our data base. However, ,2% of breast carcinoma patients age . 60 years participate in clinical trials.57 Likewise, the Autologous Blood and Marrow Transplant Registry (ABMTR) reports a median age of 44 years for MBC patients25 compared with 65 years in our clinic. Notwithstanding the obvious barriers to comparisons, it appears nonetheless that our clinic patients comprised a very poor prognostic cohort. Six of our patients actually died on the first day of disease recurrence; these patients surely would not have been included in most clinical studies. Thus, perhaps it is not surprising that our patients faired poorly: A greater wonder is that their outcome was similar to studies with more aggressive treatment philosophies. Although clinical end points were not standardized and outcome measures differed from study to study, a few examples illustrate the point (Table 5). Despite a wide range of median ages and various therapies, HODB median SAR compares favorably with all standarddose chemotherapy trials as well as with ABMTR results with high dose chemotherapy and autotransplants.4,16,17,25,58 – 60 When discussing breast carcinoma patient outcome, there is an additional sobering fact: Untreated breast carcinoma patients and even untreated MBC patients can survive for 5–10 years.33,61– 63 Consider Bloom et al.’s account of the Middlesex Cancer Charity Ward.33 Founded in England in 1745, the Middlesex Hospital opened a Cancer Charity Ward in 1792 as an asylum for the terminally ill. From 1747 to 1903, there were 5526 women with carcinoma, including 1950 (35%) with “malignant disease of the breast.” Bloom et al. located the clinical records of 356 untreated breast carcinoma patients. He reported on 250 of these patients with necropsy confirmation of breast carcinoma. The median ages were 51– 60 years, 23% had locally advanced disease, and 74% had systemic metastases. Patients were treated with opiates, dressings, Metastatic Breast Carcinoma/Anderson et al. lead sera, Coley toxin, and thymic extract. None had surgery, radiotherapy, or hormonal therapy. The median survival was 2.7 years (range, 2 months to 18 years). The 5-year and 10-year overall survival rates were 18% and 4%, respectively, respectable results for most clinical trials. All patients were followed to death; 95% experienced a breast carcinoma death, and 5% died of an intervening illness. There were no spontaneous remissions. CONCLUSIONS In summary, although MBC usually is fatal, a number of patients can survive with and without treatment. A great mystery of contemporary medical oncology is that survival times are similar for dissimilar treatments.64 Most MBC patients will respond to standard cytotoxic systemic chemotherapy, but the majority of responses are temporary and partial. Proponents of chemotherapy argue that conventional-dose chemotherapy is not enough and that dose intensification will be the answer. However, uncontrolled studies of high dose chemotherapy with autotransplants65 have not been validated in randomized clinical trials.29,66,67 It may be wise to recognize that dose intensification could be the medical equivalent of past surgical excesses:24 More treatment may not be better treatment.68 For the past 100 years, there has been little gain in survival time for patients with MBC. Sadly, and despite our best efforts, many patients with advanced breast carcinoma have outcomes similar to untreated historic controls.69,70 In addition, the widespread use of adjuvant chemotherapy may complicate further future MBC clinical trials by selecting a drug-resistant subpopulation. However, It would be a disservice to the medical profession and to MBC patients alike if our comments were viewed as nihilistic and/or were interpreted to endorse no treatment. We recognize the benefit of a therapeutic response. In our data base as well as others, patient response to MBC treatment was associated with prolonged SAR (P , 0.001). Along with prolonging survival, treatment also can improve the quality of life. Treatment is a reality for most MBC patients; however, there is a real dilemma in deciding which protocol to use. We acknowledge the inherent slowness of translational research;71 however, patient selection bias and incomplete control data prevent unconditional extrapolations from contemporary clinical trials to the general medical community. Given the retrospective nature of our data and other potentially confounding variables, we also cannot draw any definite conclusions regarding the benefits of a specific drug regimen. However, for a disease 105 that most patients will not survive, it seems reasonable to be practical and truthful.72,73 MBC treatment should achieve maximum palliation and prolong survival time with limited toxicity.74 For example, front-line antihormonal therapy for older women with nonlife-threatening MBC will not jeopardize survival even for estrogen receptor negative disease.13 Similarly, in our clinic, a conservatively administered community standard of care did not compromise survival time and/or breast carcinoma specific mortality. Future progress awaits the development of new chemotherapeutic agents (possibly taxanes) and/or the emergence of treatment strategies resulting from a better knowledge of molecular carcinogenesis. REFERENCES 1. Chu KC, Tarone RE, Kessler LG, Ries LA, Hankey BF, Miller BA,et al. Recent trends In U. S. breast cancer incidence, survival, and mortality rates. J Natl Cancer Inst 1996;88(21): 1571–9. 2. Early Breast Cancer Trialist’s Collaborative Group. Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy. 133 randomized trials involving 31,000 recurrences and 24,000 death among 75,000 women. Lancet 1992;339:1–15,71– 85. 3. Bearman SI, Shpall EJ, Jones RB, Cagnoni PJ, Ross M. Highdose chemotherapy with autologous hematopoietic progenitor cell support for metastatic and high-risk primary breast cancer. Semin Oncol 1996;23(Suppl 2):60 –7. 4. Clark GM, Sledge GW, Osborne CK, McGuire WL. Survival from first recurrence: relative importance of prognostic factors in 1,015 breast cancer patients. J Clin Oncol 1987;5(1): 55– 61. 5. Ayash LJ, Wheeler C, Fairclough D, Schwartz G, Reich E, Warren D, et al. Prognostic factors for prolonged progression-free survival with high-dose chemotherapy with autologous stem-cell support for advance breast cancer. J Clin Oncol 1995;13(8):2043–9. 6. Rahman ZU, Frye DK, Buzdar AU, Smith TL, Asmar L, Champlin RE, et al. Impact of selection process on response rate and long-term survival of potential high-dose chemotherapy candidates treated with standard-dose doxorubicincontaining chemotherapy in patients with metastatic breast cancer. J Clin Oncol 1997;15(10):3171–7. 7. Rosen PP, Groshen S, Saigo PE, Kinne DW, Hellman S. Pathological prognostic factors in Stage I (T1N0M0) and Stage II (T1N1M0) breast carcinoma: a study of 644 patients with median follow-up of 18 years. J Clin Oncol 1989;7(9): 1239 –51. 8. Seidman AD. Chemotherapy for advanced breast cancer: a current perspective. Semin Oncol 1996;23(Suppl 2):55–9. 9. Falkson G, Gelman RS, Pandya KJ, Osborne CK, Tormey D, Cummings FJ, et al. Eastern Cooperative Oncology Group randomized trials of observation versus maintenance therapy for patients with metastatic breast cancer in complete remission following induction treatment. J Clin Oncol 1998; 16:1669 –76. 10. Schapira DV, Urban N. A minimalist policy for breast cancer surveillance. JAMA 1991;265(3):380 –2. 106 CANCER January 1, 2000 / Volume 88 / Number 1 11. Norton L. Metastatic breast cancer length and quality of life. N Engl J Med 1991;325:1370 –1. 12. Fisher B, Osborne CK, Margolese R, Bloomer W. Treatment of metastatic breast cancer. In: Holland JF, Frei EI, Bast RC, Kufe DW, Morton DL, Weichselbaum RR, editors. Cancer medicine, 3rd ed. Philadelphia: Lea & Febiger, 1993:1719. 13. Muss HB. The role of chemotherapy and adjuvant therapy in the management of breast cancer in older women. Cancer 1994;74(7 Suppl):2165–71. 14. Budman DR, Berry DA, Cirrincione CT, Henderson IC, Wood WC, Weiss RB, et al. Dose and dose intensity as determinants of outcome in the adjuvant treatment of breast cancer. J Natl Cancer Inst 1998;90:1205–11. 15. Wood WC, Budman DR, Korzun AH, Cooper MR, Younger J, Hart RD, et al. Dose and dose intensity of adjuvant chemotherapy for Stage II, node-positive breast carcinoma. N Engl J Med 1994;330:1253–59. 16. Greenberg PAC, Hortobagyi GN, Smith TL, Ziegler LD, Frye DK, Buzdar AU. Long-term follow-up of patients with complete remission following combination chemotherapy for metastatic breast cancer. J Clin Oncol 1996;14(8):2197–205. 17. Rahman ZU, Frye DK, Smith TL, Asmar L, Theriault RL, Buzdar AU, et al. Results and long term follow-up for 1581 patients with metastatic breast carcinoma treated with standard dose doxorubicin-containing chemotherapy. Cancer 1999;85:104 –11. 18. Norton L. Evolving concepts in the systemic drug therapy of breast cancer. Semin Oncol 1997;24(4)(Suppl 10):S10-3–10. 19. Hudis CA, Norton L. Adjuvant drug therapy for operable breast cancer. Semin Oncol 1996;23(4):475–93. 20. Hudis CA, Norton L. Adjuvant drug treatment for resectable breast cancer. Oncologist 1997;2:351– 8. 21. Sledge GW. Should we dream the impossible dream? The meaning of long-term survival in metastatic breast cancer. J Clin Oncol 1996;14(8):2191–3. 22. General Accounting Office. Breast cancer patients’ survival. Washington, DC: General Accounting Office of the United States, 1989. 23. Mueller CB. The case against the use of adjuvant chemotherapy in breast cancer. Bull Am Coll Surg 1993;78:25–31. 24. Editorial. Breast cancer: have we lost our way? Lancet 1993; 341:343– 4. 25. Antman KH, Rowlings PA, Vaughan WP, Pelz CJ, Fay JW, Fields KK, et al. High-dose chemotherapy with autologous hematopoietic stem-cell support for breast cancer in North America. J Clin Oncol 1997;15(5):1870 –9. 26. Crump M, Goss PE, Prince M, Girouard C. Outcome of extensive evaluation before adjuvant therapy in women with breast cancer and 10 or more positive axillary lymph nodes. J Clin Oncol 1996;14(1):66 –9. 27. Canellos GP. Selection bias in trials of transplantation for metastatic breast cancer: have we picked the apple before it was ripe? J Clin Oncol 1997;15(10):3169 –70. 28. Garcia-Carbonero R, Hidalgo M, Paz-Ares L, Calzas J, Gomez H, Guerra JA, et al. Patient selection in high-dose chemotherapy trials: relevance in high-risk breast cancer. J Clin Oncol 1997;15(10):3178 – 84. 29. Savarese DMF, Hsieh C-C, Stewart FM. Clinical impact of chemotherapy dose escalation in patients with hematologic malignancies and solid tumors. J Clin Oncol 1997;15(8): 2981–95. 30. LeLorier J, Gregoire G, Benhaddad A, Lapierre J, Derderian F. Discrepancies between meta-analyses and subsequent 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. large randomized, controlled trials. N Engl J Med 1997;337(8):536 – 42. Byar DP. Why data bases should not replace randomized clinical trials. Biometrics 1980;36:336 – 42. A’Hern RP, Ebbs SR, Baum MB. Does chemotherapy improve survival in advanced breast cancer? A statistical overview. Br J Cancer 1988;57(6):615– 8. Bloom HJG, Richardson WW, Harries EJ. Natural history of untreated breast cancer (1805–1933). Br Med J 1962;2:213– 21. Diehl LF, Perry DJ. A comparison of randomized concurrent control groups with matched historical control groups: are historical controls valid? J Clin Oncol 1986;4(7):1114 –20. Debonis D, Terz JJ, Eldar S, Hill LR. Survival of patients with metastatic breast cancer diagnosed between 1955 and 1980. J Surg Oncol 1991;48(3):158 – 63. Hayward JL, Carbone PP, Heuson J-C, Kumaoka S, Segaloff A, Rubens RD. Assessment of response to therapy in advanced breast cancer. Cancer 1977;39(3):1289 –94. Fleming ID, Cooper JS, Henson DE, Hutter RVP, Kennedy BJ, Murphy GP, et al. Breast AJCC cancer staging manual, 5th ed. Philadelphia: Lippincott-Raven, 1997:171– 80. StatSoft, Inc. Statistica for Windows, version 5.1. Tulsa, OK: Statsoft, Inc., 1998. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 81. Peto R, Peto J. Asymptomatically efficient rank invariant test procedures. J R Stat Soc 1972;135(A):18598. Cox DR. Regression models and life-tables. J R Stat Soc 1972;34(B):187–220. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research: principles and quantitative methods. New York: John Wiley & Sons, 1982. Simpson JF, Page DL. Status of breast cancer prognostication based on histopathologic data. Am J Clin Pathol 1994; 102(Suppl 1):S3– 8. Saphner T, Tormey DC, Gray R. Annual hazard rates of recurrence for breast cancer after primary therapy. J Clin Oncol 1996;14(10):2738 – 46. Rubens RD, Bajetta E, Bonneterre J, Klijn JG,Lonning PE,, Paridaens R. Treatment of relapse of breast cancer after adjuvant systemic therapy-review and guidelines for future research. Eur J Cancer 1994;30A(1):106 –11. Bonneterre J, Mercier M. Response to chemotherapy after relapse in patients with or without previous adjuvant chemotherapy for breast cancer. Cancer Treat Rev 1993; 19(Suppl B):21–30. Ahmann FR, Jones SE, Moon TE. The effect of prior adjuvant chemotherapy on survival in metastatic breast cancer. J Surg Oncol 1988;37(2):116 –22. Moses LE. The series of consecutive cases as a device for assessing outcomes of interventions. In: Bailar JC, Mosteller F, editors. Medical uses of statistics, 2nd ed. Boston: New England Journal of Medicine Books, 1992:125– 40. Mueller CB, Jefferies W. Cancer of the breast: its outcome as measured by the rate of dying and causes of death. Ann Surg 1975;182:334 – 41. Black WC, Welch HG. Advances in diagnostic imaging and overestimation of disease prevalence and the benefits of therapy. N Engl J Med 1993;328(17):1237– 43. Shipp MA. A predictive model for aggressive non-Hodgkin’s lymphoma. N Engl J Med 1993;329(14):987–94. Metastatic Breast Carcinoma/Anderson et al. 52. Yamamoto N, Watanabe T, Katsumata N, Omuro Y, Ando M, Fukuda H, et al. Construction and validation of a practical prognostic index for patients with metastatic breast cancer. J Clin Oncol 1998;16: 2401– 8. 53. Bailar JC. Diagnostic drift in the reporting of cancer incidence. J Natl Cancer Inst 1998;90(11):863– 4. 54. Editorial. Assessing the odds [editorial]. Lancet 1998;350: 1563. 55. Feuer EJ, Wun LM, Boring CC, Flanders WD, Timmel MJ, Tong T. The lifetime risk of developing breast cancer. J Natl Cancer Inst 1993;85(11):892–7. 56. Phillips K-A, Glendon G, Knight JA. Putting the risk of breast cancer in perspective. N Engl J Med 1999;340(2):141– 4. 57. American College of Physicians. Oncology medical knowledge self-assessment program. Philadelphia: American College of Physicians, 1999. 58. Falkson G, Gelman R, Falkson CI, Glick J, Harris J. Factors predicting for response, time to treatment failure, and survival in women with metastatic breast cancer treated with DAVTH: a prospective Eastern Cooperative Oncology Group study. J Clin Oncol 1991;9(12):2153– 61. 59. Muss HB, Case LD, Richards F II, White DR, Cooper MR, Cruz JM, et al. Interrupted versus continuous chemotherapy in patients with metastatic breast cancer. The Piedmont Oncology Association [see comments]. N Engl J Med 1991; 325(19):1342– 8. 60. Perez DJ, Harvey VJ, Robinson BA, Atkinson CH, Dady PJ, Kirk AR, et al. A randomized comparison of single-agent doxorubicin and epirubicin as first-line cytotoxic therapy in advanced breast cancer. J Clin Oncol 1991;9(12):2148 – 52. 61. Daland EM. Untreated cancer of the breast. Surg Gynecol Obstet 1927;44:264 – 8. 62. Greenwood M. A report On the natural duration of cancer. Rep Public Health Med Subj HMSO 1926;33:1–26. 63. Nathanson IT, Welch CE. Life expectancy and incidence of malignant disease. I. Carcinoma of the breast. Am J Cancer 1936;28:40 –53. 107 64. Norton L. Kinetic concepts in the treatment of breast cancer. Recent Results Cancer Res 1993;127:1– 6. 65. Peters WP, Ross M, Vredenburgh JJ, Meisenberg B, Marks LB, Winer E, et al. High-dose chemotherapy and autologous bone marrow support as consolidation after standard-dose adjuvant therapy for high-risk primary breast cancer. J Clin Oncol 1993;11(6):1132– 43. 66. Rodenhuis S, Richel DJ, van der Wall E, Schornagel JH, Baars JW, Koning CC,et al. Randomised trial of high-dose chemotherapy and haemopoietic progenitor-cell support in operable breast cancer with extensive axillary lymph-node involvement [see comments]. Lancet 1998;352(9127):515–21. 67. Stadtmauer EA, O’Neill A, Goldstein LJ, et al. Phase III randomized trial of high-dose chemotherapy (HDC) and stem cell support (SCT) shows no difference in overall survival or severe toxicity compared to maintenance chemotherapy with cyclophosphamide, methotrexate and 5-fluorouracil (CMF) for women with metastatic breast cancer who are responding to conventional induction chemotherapy: the “Philadelphia” Intergroup Study (PBT-1). Proc Am Soc Clin Oncol 1999;18:1a. 68. Hudis CA, Munster PN. High-dose therapy for breast cancer. Semin Oncol 1999;26:35– 47. 69. Gardner B, Feldman J. Are positive axillary nodes in breast cancer markers for incurable disease? Ann Surg 1993;218(3): 270 – 8. 70. Fox MS. On the diagnosis and treatment of breast cancer. JAMA 1979;241(5):489 –94. 71. Cannistra SA. “Cancer defeated”: not if, but when—introducing the Biology of Neoplasia Series. J Clin Oncol 1997; 15(11):3297– 8. 72. Smith TJ. Knowing when to bow out gracefully. In: Oncology News International, vol 5. New York: 1996:1,22. 73. Smith TJ, Swisher K. Telling the truth about terminal cancer. JAMA 1998;279(21):1746 – 8. 74. Hortobagyi GN. Treatment of breast cancer. N Engl J Med 1998;339(14):974 – 84.