Fax: (608) 265-6905
Survival in patients with metastatic recurrent breast cancer after adjuvant chemotherapy
Little evidence of improvement over the past 30 years
Article first published online: 12 OCT 2012
Copyright © 2012 American Cancer Society
Volume 119, Issue 6, pages 1140–1148, 15 March 2013
How to Cite
Tevaarwerk, A. J., Gray, R. J., Schneider, B. P., Smith, M. L., Wagner, L. I., Fetting, J. H., Davidson, N., Goldstein, L. J., Miller, K. D. and Sparano, J. A. (2013), Survival in patients with metastatic recurrent breast cancer after adjuvant chemotherapy. Cancer, 119: 1140–1148. doi: 10.1002/cncr.27819
- Issue published online: 4 MAR 2013
- Article first published online: 12 OCT 2012
- Manuscript Accepted: 6 AUG 2012
- Manuscript Revised: 14 JUL 2012
- Manuscript Received: 19 APR 2012
- breast cancer;
- distant disease recurrence;
- distant recurrence-free interval;
- metastatic disease
Population-based studies have shown improved survival for patients diagnosed with metastatic breast cancer over time, presumably because of the availability of new and more effective therapies. The objective of the current study was to determine whether survival improved for patients who developed distant recurrence of breast cancer after receiving adjuvant therapy.
Adjuvant chemotherapy trials coordinated by the Eastern Cooperative Oncology Group that accrued patients between 1978 and 2002 were reviewed. Survival after distant disease recurrence was estimated for progressive time periods, and adjusted for baseline covariates in a Cox proportional hazards model.
Of the 13,785 patients who received adjuvant chemotherapy in 11 trials, 3447 (25%) developed distant disease recurrence; the median survival after recurrence was 20 months (95% confidence interval, 19 months-21 months). Factors associated with inferior survival included a shorter distant recurrence-free interval (DRFI), estrogen receptor-negative and progesterone receptor-negative disease, the number of positive axillary lymph nodes present at the time of diagnosis, and black race (P < .0001 for all). When the time period of recurrence was added to the model, it was not found to be significantly associated with survival for the general population with disease recurrence. Survival improved over time only in those patients with hormone receptor-negative disease with a DRFI ≤ 3 years, both among the 5 most recent and the entire trial data sets (P = .01 and P = .05, respectively).
In contrast to reports from population-based studies, no general improvement in survival was observed over the last 30 years for patients who developed distant disease recurrence after adjuvant chemotherapy after adjusting for DRFI. Improved survival for patients with hormone receptor-negative disease with a short DRFI suggests a benefit from trastuzumab. Cancer 2013. © 2012 American Cancer Society.
Metastatic breast cancer (MBC) remains the second leading cause of cancer death in women, with > 40,000 dying each year in the United States and > 400,000 dying globally each year.1 Although breast cancer mortality rates have declined in the United States because of screening and improved systemic adjuvant therapy,2 the disease remains incurable for those patients with distant metastases.3 Evidence suggests that survival has modestly improved in the era of modern systemic therapy,4-6 indicating that further improvement may be possible with new therapeutic approaches. A potential secondary benefit is that the identification of effective new agents for the treatment of MBC may produce survival gains when used as adjuvant therapy for patients with localized disease: notable examples include anthracyclines, taxanes, and trastuzumab.7-9
To determine whether survival has improved for patients who developed distant disease recurrence after receiving adjuvant therapy, we undertook a review of adjuvant phase 3 trials conducted by the Eastern Cooperative Oncology Group (ECOG) over a period of approximately 30 years.
MATERIALS AND METHODS
Eleven phase 3 adjuvant breast cancer trials conducted by ECOG that treated patients using adjuvant chemotherapy and had at least 5 years of follow-up were identified and included in the current analysis (Table 1).10-21 Participants who did not receive adjuvant chemotherapy were excluded from this analysis.
|Trial||Treatment Arms||Key Characteristics||Enrollment Dates, Month/Year||No. Enrolled||No. Receiving Chemotherapy||No. With Distant Recurrence|
|E517710||CMF/P/PT||Premenopausal||2/1978 to 2/1982||662||662||284|
|E617710, 11||CMFP vs CMFPT vs observation||Postmenopausal||2/1978 to 7/1981||265||170||92|
|E118012, 13||CMFP vs observation||Lymph node negative||5/1981 to 5/1988||541||276||47|
|E418114||CMFPT 4 vs 12 vs 12+ continued T||Postmenopausal||2/1982 to12/1986||961||961||392|
|E518115||CMFPT vs alternating chemotherapy with continued T vs observation||Premenopausal||2/1982 to 7/87||658||658||224|
|E318116||CAFTH with RT vs observation||Locally advanced, resectable||1/1982 to 2/1987||332||332||171|
|“Recent” Adjuvant Trialsa|
|E318917||CAF vs multidrug||ER-negative||8/1989 to 4/1993||646||646||212|
|E518818||CAF vs CAF+Z vs CAF+ZT||Premenopausal, ER-positive, lymph node-positive||7/1989 to 2/1994||1536||1536||533|
|E219019||CAF vs CAF plus high-dose chemotherapy||≥10 positive axillary lymph nodes||8/1991 to 8/1998||540||540||246|
|E219720||AC vs AD||0-3 positive axillary lymph nodes||7/1998 to 1/2000||2952||2952||293|
|E119921||AC + taxanes||Positive lymph nodes or high-risk, lymph node-negative||10/1999 to 1/2002||5052||5052||953|
The primary study endpoint was survival after distant disease recurrence, defined as the time from the date of distant disease recurrence to the date of death (or date patient was last known alive). Survival after distant disease recurrence was determined with adjustment for baseline covariates in a Cox proportional hazards model. The models included calendar year of (distant) recurrence, age at diagnosis of recurrence, estrogen receptor (ER) and progesterone receptor (PgR) status, number of positive axillary lymph nodes at the time of diagnosis, size of the primary tumor at diagnosis, race, and distant recurrence-free interval (DRFI). Human epidermal growth factor receptor 2 (HER2) was not available for inclusion in the model. DRFI was defined as the time from study entry to the date of distant disease recurrence. Because DRFI is strongly associated with survival after recurrence and the potential for “gap time” bias,22 log-rank tests for other covariates were computed stratified on DRFI (0-3 years, > 3 to 6 years, and > 6 years). Estimates of survival after distant disease recurrence were weighted averages of the Kaplan-Meier estimates computed within DRFI groups. Therapeutic intervals of interest were identified between the first on-study date (1978) and the last point at which data were censored (2010). Recurrences were assigned to 1 of 6 time periods: 1978 to 1983, 1984 to 1988, 1989 to 1993, 1994 to 1998, 1999 to 2003, and 2004 to 2010. The most recent time period (2004-2010) was used as the comparator. Analyses were based on all 3447 patients with distant disease recurrence. In addition, we examined the outcomes for 2237 patients who received “recent” chemotherapy regimens (E3189, E5188, E2190, E2197, and E1199) between 1989 and 2002, given that the type of adjuvant chemotherapy used might influence resistance in the metastatic setting.
Characteristics of Included Trials
The study population included 13,785 patients enrolled on 11 adjuvant ECOG trials between 1978 and 2002 (Table 1). Among the 13,785 patients enrolled, 3447 (25%) had distant recurrences, 814 (5.9%) had local recurrences only, and 20 (0.1%) had unknown sites of disease recurrence. Table 2 shows the characteristics of both all adjuvant participants and the subset of patients with distant recurrence. The median survival after distant recurrence in the 3447 patients was 20 months (95% confidence interval, 19 months-21 months). Estimated 5-year and 10-year survival rates were 16.3% and 6.4%, respectively.
|Characteristic||Entire Adjuvant Study Population Treated With Chemotherapy||Cases With Distant Recurrence From Adjuvant Study Population||Cases With Distant Recurrence From Recent Trials Only|
|No. of adjuvant trials||11||11||5 most recent: E3189, E5188, E2190, E2197, E1199|
|No. of patients||13,785||3447||2237|
|Median age at time of initial diagnosis (range), y||49 (19-85)||48 (19-80)||47 (19-80)|
|Median age at time of distant recurrence (range), y||—||52 (21-85)||50 (22-83)|
|Hormone receptor status|
|ER positive||8822 (64%)||2093 (61%)||1388 (62%)|
|PgR positive||7977 (58%)||1794 (52%)||1359 (61%)|
|Tumor size at diagnosis, cm|
|0-2.0||5261 (38%)||933 (27%)||613 (28%)|
|2.1-5.0||7103 (52%)||1994 (58%)||1285 (57%)|
|≥5.1||1300 (10%)||501 (15%)||323 (14%)|
|No. of positive axillary LNs at diagnosis|
|0||2771 (20%)||275 (8%)||225 (10%)|
|1-3||6400 (46%)||1281 (37%)||843 (38%)|
|4-9||2746 (20%)||981 (29%)||560 (25%)|
|≥10||1792 (13%)||898 (26%)||603 (27%)|
|Unknown||301 (2%)||45 (1%)||43 (2%)|
|White||11,817 (85%)||2972 (86%)||1873 (84%)|
|Black||1137 (8%)||298 (9%)||221 (10%)|
|Other||530 (4%)||132 (4%)||100 (4%)|
|≤3||—||1781 (52%)||1144 (51%)|
|3-6||—||987 (29%)||700 (31%)|
|>6||—||679 (20%)||393 (18%)|
Analysis of Distant Recurrence
Covariates found to be significantly associated with inferior survival after distant disease recurrence included ER-negative and PR-negative disease (Fig. 1A). However, DRFI was found to be the most strongly associated with survival after disease recurrence (Figs. 1B and 1C). Black race and increasing number of positive axillary lymph nodes at the time of diagnosis were also found to be significant.
Table 3 shows the estimated hazards ratio (HR) from the Cox proportional hazards model. Patients with a shorter DRFI fared significantly worse than those with a longer DRFI (HR of 2.44 for DRFI > 6 years vs ≤ 3 years [P < .0001] and HR of 1.43 for DRFI > 6 years vs > 3 to 6 years [P < .0001]). Patients with ER-negative and PR-negative disease also were found to have significantly shorter survival after recurrence (HR, 1.35 [P < .0001] and HR, 1.33 [P < .0001], respectively), as did patients of black race (HR, 1.34; P < .0001) and those with more positive lymph nodes at the time of diagnosis (HR of 1.17 for 1-3 lymph nodes vs 0 lymph nodes; HR of 1.35 for 4-9 lymph nodes vs 0 lymph nodes; and HR of 1.33 for > 9 lymph nodes vs 0 lymph nodes [P < .0001]). Age, primary tumor size, and year of disease recurrence were not found to be statistically significant.
|All 11 Studies (n = 3431)||5 Recent Studies (n = 2231)|
|Factor||Comparison||HR||95% CI||P||HR||95% CI||P|
|DRFI, y||≥3 vs >6||2.44||2.16-2.76||<.0001||2.22||1.87-2.63||<.0001|
|>3, ≤6 vs >6||1.43||1.26-1.62||1.38||1.16-1.63|
|ER status||Negative vs positive||1.35||1.23-1.49||<.0001||1.34||1.16-1.56||<.0001|
|Unknown vs positive||0.88||0.52-1.48||0.54||0.24-1.21|
|PgR status||Negative vs positive||1.33||1.20-1.47||<.0001||1.27||1.09-1.47||.007|
|Unknown vs positive||1.04||0.90-1.21||1.43||0.78-2.62|
|No. of positive axillary LNsa||1-3 vs 0||1.17||1.01-1.36||<.0001||1.05||0.89-1.25||.15|
|4-9 vs 0||1.35||1.16-1.58||1.16||0.97-1.40|
|>9 vs 0||1.33||1.14-1.56||1.17||0.97-1.40|
|Race||Black vs white||1.34||1.18-1.53||<.0001||1.38||1.19-1.60||<.0001|
|Other vs white||0.88||0.72-1.07||0.82||0.65-1.03|
|Unknown vs white||0.95||0.68-1.33||1.06||0.75-1.50|
|Age at distant recurrence, y||40-49 vs <40||0.95||0.85-1.07||.30||0.96||0.83-1.11||<.0001|
|50-59 vs <40||0.99||0.88-1.11||1.06||0.91-1.23|
|60-69 vs <40||1.01||0.89-1.16||1.20||1.00-1.43|
|>69 vs <40||1.13||0.95-1.33||1.68||1.31-2.15|
|Tumor size, cm||2.1-5.0 vs ≤2.0||1.11||1.01-1.21||.12||1.13||1.02-1.27||.07|
|>5.0 vs ≤2.0||1.07||0.95-1.21||1.19||1.03-1.39|
|Unknown vs ≤2.0||0.87||0.53-1.44||0.96||0.56-1.64|
|Time period of distant recurrenceb||1978-1983 vs 2004-2010||0.93||0.76-1.14||.24||—||—|
|1984-1988 vs. 2004-2010||0.99||0.85-1.15||—||—|
|1989-1993 vs 2004-2010||1.02||0.87-1.18||1.29||1.06-1.58||.005|
|1994-1998 vs 2004-2010||0.97||0.84-1.12||1.13||0.97-1.31|
|1999-2003 vs. 2004-2010||0.89||0.77-1.02||0.98||0.84-1.14|
If survival after distant recurrence was evaluated without stratification for DRFI for all 11 studies, survival was found to significantly improve over time (Fig. 2A). However, once stratified by DRFI, there was no significant improvement in survival over time noted (Fig. 2B). The improvement observed in the unadjusted analysis for the entire population most likely reflects patients with more favorable characteristics who were recruited by later adjuvant trials (recent adjuvant trials selected for more patients with ER-positive disease). Table 4 illustrates this point. Survival after distant disease recurrence by time period of recurrence stratified by DRFI and hormone receptor status is shown in Figures 3A to 3D. Only among patients with hormone receptor-negative disease who developed disease recurrence within 3 years did there appear to be any improvement in survival over time.
|Calendar Year of Recurrence|
|No. of patients||246||602||471||622||897||609|
|Median age at time of initial diagnosis (range), y||47 (23-73) [Unknown in 4]||52 (20-79)||47 (25-78)||44 (24-77)||48 (24-80)||50 (19-77)|
|Median age at time of distant recurrence (range), y||49 (25-74) [Unknown in 4]||55 (21-81)||50 (25-83)||48 (28-85)||52 (25-82)||57 (22-83)|
|ER status||Negative||129 (52%)||278 (46%)||191 (41%)||217 (35%)||394 (44%)||123 (20%)|
|Positive||117 (48%)||323 (54%)||278 (59%)||403 (65%)||493 (55%)||479 (80%)|
|PR status||Negative||61 (25%)||266 (44%)||194 (41%)||198 (32%)||411 (46%)||137 (23%)|
|Positive||48 (20%)||192 (32%)||238 (51%)||394 (63%)||461 (51%)||461 (76%)|
|Unknown||137 (56%)||144 (24%)||39 (8%)||30 (5%)||25 (3%)||11 (2%)|
|Tumor size, cm||≤2||59 (24%)||156 (26%)||110 (23%)||194 (31%)||235 (26%)||179 (29%)|
|>2, ≤5||140 (57%)||354 (59%)||279 (59%)||337 (54%)||523 (59%)||361 (60%)|
|>5 cm||45 (18%)||91 (15%)||80 (17%)||85 (14%)||134 (15%)||66 (11%)|
|No. of positive axillary LNs at diagnosis||0||1 (0%)||31 (5%)||11 (2%)||5 (1%)||147 (16%)||80 (13%)|
|1-3||83 (34%)||199 (33%)||159 (34%)||214 (34%)||354 (40%)||272 (45%)|
|4-9||94 (38%)||205 (34%)||185 (39%)||174 (28%)||172 (19%)||151 (25%)|
|>9||65 (26%)||165 (28%)||116 (25%)||229 (37%)||220 (25%)||103 (17%)|
|Race||White||219 (89%)||543 (90%)||398 (85%)||520 (84%)||767 (86%)||525 (86%)|
|Black||18 (7%)||40 (7%)||48 (10%)||64 (10%)||76 (8%)||52 (9%)|
|Other||8 (3%)||19 (3%)||22 (5%)||36 (6%)||32 (4%)||15 (3%)|
|Unknown||1||0||3||2||22 (2%)||17 (3%)|
|DRFI, y||<3||214 (87%)||394 (65%)||262 (56%)||243 (39%)||620 (69%)||48 (8%)|
|3-6||32 (13%)||167 (28%)||97 (21%)||246 (40%)||125 (14%)||320 (53%)|
|>6||0 (0%)||41 (7%)||112 (24%)||133 (21%)||152 (17%)||241 (40%)|
Analysis of Recent Trials Only
In the 5 most recent trials studied, factors found to be associated with survival after distant disease recurrence included DRFI, ER/PR expression, and race (Fig. 2C) (Table 3). However, in contrast to the entire study population, older age at the time of disease recurrence and time period of recurrence beginning in 1999 (compared with before 1994) were found to be significantly associated with improved survival. However, this improvement over time was again confined to those patients with hormone receptor-negative disease who developed disease recurrence within 3 years of diagnosis (Figs. 4A-4E).
Among the many phase 3 MBC trials performed over the past 30 years, survival was found to significantly improve in only a small number of cases.23-26 Despite the difficulty in demonstrating improved survival in individual trials, population-based studies suggest that MBC patients now survive modestly longer than in the past.4-6 This improvement could be because of the increased availability of drugs that when used individually have minimal effect in prolonging survival but when used sequentially may produce modest survival gains. Other suggested explanations include the impact of improved imaging.4 Better imaging may lead to an apparent prolongation in the interval between the diagnosis of disease recurrence and death by identifying MBC at earlier time points. Better imaging also increases the percentage of women with de novo MBC (women with recurrent breast cancer have inferior survival compared with women with de novo disease).3
To determine whether the perceived improvement in survival demonstrated by population-based studies also was evident in clinical trial populations, we evaluated survival after recurrence among participants in ECOG clinical trials who received adjuvant chemotherapy. The results of the current analysis suggest that for women who develop distant recurrence after adjuvant chemotherapy, the availability of new cytotoxic and biologic agents has not broadly translated into improved survival. The exception appears to be among patients with hormone receptor-negative disease, who develop disease recurrence within 3 years of diagnosis, where survival improved for the period of recurrence beginning in 1999 (compared with prior to 1994). Therefore, we hypothesize that this may reflect an effect of trastuzumab rather than improved cytotoxic therapy: trastuzumab became commercially available in 1998, and HER2-positive disease is associated with early recurrence.
Table 5 summarizes the characteristics of this and other reports evaluating trends in metastatic survival. There are several key differences between the current analysis and the other reports studied.4-6, 27-29 First, we included only patients whose disease recurred after they received adjuvant chemotherapy for early-stage breast cancer and thus were more likely to have drug-resistant disease (de novo disease was included in other analyses.) Second, we adjusted for multiple prognostic covariates in multivariate models, including DRFI. Giordano et al4 noted large differences in outcome by the year of recurrence, which was not evident when the analysis was adjusted for DRFI. However, not all of the prior analyses shown in Table 5 were adjusted for DRFI and other covariates. This may have contributed to the improved survival over time observed by others. Indeed, survival was found to improve over time in our data set if the survival analysis was not stratified by DRFI, thereby highlighting the importance of controlling for this variable when evaluating survival over time.
|Study||No. With MBCa (and % With De Novo Stage IV)||Time Period Examined||Source||Median Survival, Months||DFI or DRFI in Model||Survival Over Time|
|Chia 20075||2150 (21.4%)||1991-2001||Population-based registry||15-22b||No||Improved|
|Dabakuyo 200829||1459 (12%)||1982-2005||Population-based registry||Relative survival reported||No||Improved|
|Dawood 201027||2091 (22.4%)||1991-2007||Single-institution database||28.6||No||Improvedc|
|Dawood 20086||15,438 (100%)||1988-2003||Population-based registry||23||No||Improved|
|Giordano 20044||834 (0%)||1974-1994||Single-institution database||21||Yes||Not improvedd|
|Largillier 200828||1038 (0%)||1975-2005||Single-institution database||23.1||Yes||Not improved|
|Current study||3477 (0%)||1978-2010||Multiinstitution database||20||Yes||Not improved overall|
The current study has several strengths and limitations. Strengths include the large sample size, long follow-up, the fact that it was conducted by the same group of investigators, and its use of standardized treatment regimens and prospective data collection. The heterogeneity of the population over time is a limitation imposed by evolving eligibility criteria for the 11 adjuvant trials. To account for this, the analysis was not based on the time period of enrollment but rather on the time period of recurrence while controlling for factors such as age, ER/PR expression, and race. Other limitations include a lack of information regarding sites of disease recurrence, treatments used after recurrence, and the potential for lead-time bias because of improved imaging. In addition, information regarding HER2 status and anti-HER2 therapy was absent. Although we hypothesized that the survival benefit for those patients diagnosed with an early recurrence after 1999 was because of the availability of trastuzumab, it is not possible for us to determine with certainty that this finding reflects the improved survival for patients with HER2-positive disease demonstrated in prospective clinical trials.23, 30, 31
The results of the current study provided evidence over a 30-year time period of an improvement in survival among patients with hormone receptor-negative disease who developed disease recurrence within 3 years after adjuvant chemotherapy for localized disease, but not for the population as a whole. Survival appeared to improve over the past 30 years for the entire population, but this effect was not found to be persistent when the survival analysis was adjusted for DRFI and other covariates. This suggests that the survival improvements observed among population-based studies may not reflect outcomes for all subsets of women whose disease recurs after adjuvant chemotherapy. There remains a critical need for the development of more effective therapies for patients with MBC, especially those who have developed disease recurrence after receiving adjuvant chemotherapy.
This study was coordinated by the Eastern Cooperative Oncology Group (Robert L. Comis, MD, Chair) and supported in part by Public Health Service grants CA23318, CA66636, CA21115, CA21076, CA49883, CA17145, CA16116, CA39229, CA27525, and CA14958 and from the National Cancer Institute, National Institutes of Health, and the Department of Health and Human Services. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 20Concurrent doxorubicin plus docetaxel is not more effective than concurrent doxorubicin plus cyclophosphamide in operable breast cancer with 0 to 3 positive axillary nodes: North American Breast Cancer Intergroup Trial E 2197. J Clin Oncol. 2008; 26: 4092-4099., , , et al.
- 25EMBRACE (Eisai Metastatic Breast Cancer Study Assessing Physician's Choice Versus E7389) investigators. Eribulin monotherapy versus treatment of physician's choice in patients with metastatic breast cancer (EMBRACE): a phase 3 open-label randomised study. Lancet. 2011; 377: 914-923., , , et al;
- 31Randomized phase II trial of the efficacy and safety of trastuzumab combined with docetaxel in patients with human epidermal growth factor receptor 2-positive metastatic breast cancer administered as first-line treatment: the M77001 study group. J Clin Oncol. 2005; 23: 4265-4274., , , et al.