- Email: [email protected]
Risk of Recurrence or Contralateral Breast Cancer More than 5 Years After Diagnosis of Hormone Receptor-Positive Early-Stage Breast Cancer
Original Study
Risk of Recurrence or Contralateral Breast Cancer More than 5 Years After Diagnosis of Hormone Receptor-Positive Early-Stage Breast Cancer Sheridan Wilson,1 Caroline Speers,1 Scott Tyldesley,1 Stephen Chia,1 Hagen Kennecke,1 Susan Ellard,2 Caroline Lohrisch1 Abstract Adjuvant hormone therapy in years 5 to 10 has shown a modest survival advantage that must be balanced against the side effects associated with continued use. We identified cancer cases with < 10% recurrence risk > 5 years after breast cancer diagnosis to inform discussions between clinicians and patients about the individual benefit of continuing hormone therapy after 5 years. Background: Three large studies have shown a survival benefit from 10 years of adjuvant hormone therapy (AHT). We evaluated the risk of an event 5 years after the initial breast cancer (BC) diagnosis and identified the prognostic factors to assist clinicians considering extended AHT. Patients and Methods: Patients newly referred to the BC Cancer Agency with stage I to III estrogen receptor-positive BC diagnosed from 1989 to 2004 who had undergone AHT were identified by the BC Cancer Agency’s Breast Cancer Outcomes Unit. Cases with recurrence, death, or contralateral BC occurring within the first 5 years were excluded. The 10-year event-free survival (EFS) and 95% confidence intervals (CIs) were calculated using the Kaplan-Meier method. This provided estimates of recurrence risk after the fifth year following the diagnosis. The histopathologic and age variables were examined for prognostic value by univariate analysis. Results: Within our cohort, 6615 women were postmenopausal and 1886 were premenopausal at the BC diagnosis. The median follow-up period was 11 years. The 10-year EFS for women aged < 50 years with stage I, II, and III disease was 94.8% (95% CI, 92.8%-96.3%), 88.3% (95% CI, 86.0%-90.2%), and 80.4% (95% CI, 73.6%-85.6%), respectively. Among women aged 50 years, the corresponding EFS rates were 94.8% (95% CI, 93.8%-95.6%), 86.3% (95% CI, 85.0%-87.5%), and 73.8% (95% CI, 69.1%-77.8%). EFS varied significantly by grade. The 10-year recurrence risk was < 10% with stage I cancer (any grade) and for stage II (node-negative and node-positive), grade I cancer. Conclusion: Our data have identified BCs associated with a very low recurrence risk 5 to 10 years after diagnosis, providing women with such cancers confidence about a decision to discontinue AHT after 5 years. Clinical Breast Cancer, Vol. -, No. -, --- ª 2015 Elsevier Inc. All rights reserved. Keywords: Adjuvant hormone therapy, Five versus 10 years of therapy, Good prognosis groups, Hormone receptor-positive breast cancer, Late recurrence risk
Introduction Breast cancer (BC) accounts for 5000 deaths among Canadian women annually and is the leading cause of cancer-related death in women < 50 years old.1 Most BCs arising in premenopausal and 1
BC Cancer Agency, Vancouver Cancer Centre, Vancouver, BC, Canada BC Cancer Agency, Centre for the Southern Interior, Kelowna, BC, Canada
2
Submitted: Apr 16, 2015; Accepted: Nov 10, 2015 Address for correspondence: Caroline Lohrisch, MD, FRCPC, Medical Oncology Department, BC Cancer Agency, Vancouver Cancer Centre, 600 West 10th Avenue, Vancouver, BC V5Z 4E6 Canada E-mail contact: [email protected]
1526-8209/$ - see frontmatter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clbc.2015.11.002
postmenopausal women are estrogen receptor-positive (ERþ). In this population, adjuvant hormone therapy (AHT) significantly reduces the risk of BC recurrence and BC-related death.1,2 The efficacy of adjuvant tamoxifen was established within large randomized trials reported in the 1980s.3-6 The updated results from the Early Breast Cancer Trialists’ Collaborative Group meta-analysis of > 20,000 women receiving approximately 5 years of tamoxifen within 20 randomized trials demonstrated that the benefit of tamoxifen continues well beyond the treatment period, reducing the 15-year risks of BC recurrence and death.2,6-8 Since 2004, the American Society of Clinical Oncology guidelines have recommended an aromatase inhibitor (AI) as a part of AHT for
Clinical Breast Cancer Month 2015
-1
Risk More Than 5 Years After HR+ BC Diagnosis postmenopausal hormone receptor-positive (HRþ) BC based on the results of many randomized controlled trials.9 Late relapses (> 5 years from diagnosis) of ERþ BC remain an important problem. A significant proportion of ERþ BC recurrences and two thirds of BC deaths occur > 5 years from diagnosis.1 Until recently, the optimal duration of AHT was considered 5 years, because the initial trials of > 5 years failed to show a consistent benefit.3,10 However, the findings from 3 large randomized controlled trials now support a further survival advantage of 2.5% to 3% by extending the duration of AHT from 5 years to 10 for both pre- and postmenopausal women with HRþ early-stage BC. MA.17, a randomized placebo-controlled trial, evaluated the addition of 5 years of letrozole after 5 years of tamoxifen, and the Adjuvant Tamoxifen Longer Against Shorter (ATLAS) and Adjuvant Tamoxifen Treatment offers More trials investigated the extension of tamoxifen to 10 years.11-13 Although the results of the subgroup analyses of these trials suggested that all stage and age subgroups might benefit from extended AHT, the survival benefits were lower in the groups with a lower baseline risk. Women with very low or no residual risk of recurrence after 5 years are still at risk of harm from continued AHT but can benefit only by virtue of secondary prevention (0.5%-1.0% absolute reduction, determined by the excess incidence of contralateral breast cancer [CLBC] observed in the control arms of the MA.17 and ATLAS trials). The ATLAS study reported a 1.5% excess risk of endometrial cancer (0.2% excess mortality) and a 0.3% excess risk of pulmonary embolus but no overall increase in non-BC mortality associated with 10 versus 5 years of tamoxifen. In the MA.17 study, at 30 months of follow-up, numerically, but not statistically significantly, more cardiovascular events, osteoporosis diagnoses, and fractures were observed in the letrozole group. In addition, a 5% excess rate of grade 1 to 3 arthralgia was found. A challenge for oncologists is how to apply these important findings to a modern-day population. The evidence for extended tamoxifen use was generated before the availability of adjuvant AIs. Modern chemotherapy, hormone therapy, and radiotherapy techniques are known to influence BC recurrence and survival rates. Thus, the benefit of extended AHT might be less for women diagnosed and treated today. Considering the possible serious harm and long-term effects on quality of life associated with extended AHT, an understanding of the patient and tumor characteristics associated with recurrence > 5 years after the diagnosis of HRþ BC will help clinicians counsel women with low residual recurrence risk who might reasonably and safely choose to stop AHT after 5 years. We examined the prognostic factors for BC events in women alive and free of BC relapse or CLBC after 5 years of AHT. Our primary objective was to identify patients with an AHT event-free survival (EFS) rate at 10 years of > 90%, including the lower boundary of the 95% confidence intervals (CIs), who, on this basis, might reasonably be spared another 5 years of AHT.
Patients and Methods
2
-
Population-based data are prospectively collected and maintained by the BC Cancer Agency (BCCA) Breast Cancer Outcomes Unit for cases diagnosed since 1989. This database captures all patients with BC referred to the BCCA and represents an estimated 85% of
Clinical Breast Cancer Month 2015
all BCs diagnosed in British Columbia annually. Referred women diagnosed with stage I to III invasive ERþ BC from 1989 to 2004 and who had begun AHT were identified for the present study. The exclusion criteria were previous invasive or in situ disease and synchronous bilateral BC. Women aged < 50 years whose initial AHT was an AI were excluded. Disease stage was assigned using the American Joint Committee on Cancer Staging criteria. The pathology characteristics abstracted included the degree of ER positivity, tumor size and grade, and nodal status. ER positivity was divided into low, moderate, and high expression using the following definitions: for the biochemical method, low, 10 to 50 fmol/mg; moderate, 51 to 100 fmol/mg; and high, > 100 fmol/mg; for the immunohistochemical method, low, 1þ or Allred score of 3 or 4; moderate, 2þ or Allred score of 5 or 6; and high, 3þ or Allred score of 7 or 8. When ER expression was only described as positive, it was assigned a value of moderate. The study population was dichotomized based on age (< 50 or 50 years) at diagnosis to serve as a surrogate for menopausal status. Patients underwent surgery followed by adjuvant radiotherapy and systemic therapy, directed by evidence-based provincial guidelines maintained and disseminated by the BCCA.14 The primary outcome was EFS, defined as the interval to the first of local (breast or chest wall), regional (axillary, infra- and/or supraclavicular or internal mammary lymph nodes), or distant recurrence; new CLBC diagnosis; or death from BC without previous documented recurrence. The analysis of the 10-year EFS was confined to women without a BC event or death within the first 5 years after the diagnosis. BC-specific survival (BCSS), defined as the interval from diagnosis to BC death, and overall survival (OS), defined as the interval from diagnosis to death from any cause, were also calculated for the 2 age cohorts. Survival was estimated using the Kaplan-Meier method and compared using log-rank tests. All statistical analyses were performed with SPSS, version 14.0, software (SPSS Inc., Chicago IL). The Research Ethics Board of the BCCA, University of British Columbia, approved the study.
Results Before the exclusions, 10,815 ERþ stage I to III BC cases diagnosed from 1989 to 2004 and referred to the BCCA were identified. After excluding the cases with previous invasive or in situ BC and those with synchronous bilateral BC, 10,414 unique cases remained. During the first 5 years after diagnosis, EFS events occurred in 1443 cases, representing 13% of incident cases and 54% of all EFS events. This left 8526 patients (82%) alive and free of a BC event at 5 years after the initial diagnosis, forming the study population for our analysis. The median follow-up period was 11 years. A total of 1207 EFS events occurred > 5 years from diagnosis among these 8526 cases, representing 12% of the 10,414 incident cases and 45% of all events. The patient and tumor characteristics are listed in Table 1. A total of 25 patients (1.3%) aged < 50 years at diagnosis who had received an AI as their initial AHT were excluded. Of the remaining 1886 women aged < 50 years at diagnosis, 1692 (90%) were premenopausal and 147 (8%) were postmenopausal (the menopausal status was unknown in 47 patients [2%]). The patients for whom the tumor grade was unknown were included in the overall EFS calculations but excluded from the analysis by tumor grade.
Sheridan Wilson et al Table 1 Patient and Tumor Characteristics for Patients Remaining Event-Free 5 Years After Initial Diagnosis
Table 1 Continued Age (Years)
Age (Years) Characteristic
<50 (n [ 1886)
‡50 (n [ 6615)
Age at diagnosis (years)
P Value NA
Median
45.0
64
Range
24-49
50-92
2.0
1.8
Range
0.1-9.9
0.1-9.9 <.001
Tumor size group, cm 0.1-0.5
49 (2.6)
196 (3.0)
0.6-1.0
222 (11.8)
940 (14.2)
1.1-2.0
798 (42.3)
2900 (43.8)
2.1-3.0
459 (24.3)
1680 (25.4)
3.1-5.0
222 (11.8)
583 (8.8)
5.1-9.9
130 (6.9)
279 (4.2)
6 (0.3)
37 (0.6)
Negative
955 (50.6)
4046 (61.2)
Positive
931 (49.4)
2560 (38.7)
Unknown
0 (0.0)
9 (0.1)
Median nodes sampled
11
10
Range nodes sampled
1-42
1-48
<.001
Positive nodes (n)
Positive nodal group
704 (37.3)
1423 (21.5)
2 (0.1)
13 (0.2)
654 (71.5)
1908 (75.1)
4-9
212 (23.2)
507 (20.0)
10
48 (5.2)
119 (4.7)
1 (0.1)
6 (0.2) <.001
LVI status Negative
1212 (64.3)
4639 (70.1)
Positive
617 (32.7)
1700 (25.7)
Unknown
57 (3.0)
276 (4.2)
1
392 (20.8)
1695 (25.6)
2
926 (49.1)
3363 (50.8)
3
537 (28.5)
1337 (20.2)
31 (1.6)
220 (3.3)
7 (0.4)
32 (0.5)
<.001
Grade
Initial surgery
.02
Breast-conserving surgery
1070 (56.7)
3952 (59.7)
Total mastectomy
809 (42.9)
2631 (39.8)
No breast/chest wall or nodal RT
366 (19.4)
1937 (29.3)
Breast/chest wall alone
814 (43.2)
3242 (49.0)
<.001
Initial radiotherapy
P Valuea
<.001
Yes
1381 (73.2)
1545 (23.4)
No
505 (26.8)
5070 (76.6)
Data presented as n (%), unless otherwise noted. Abbreviations: LVI ¼ lymphovascular invasion; NA ¼ not applicable; RT ¼ radiation therapy. a P values calculated from known values only; test statistics used were c2 test for categorical variables and Mann-Whitney U test for continuous variables.
The 10-year outcomes for women without a BC cancer event during the first 5 years after diagnosis are listed in Table 2. Considering all stages and grades, the 10-year EFS approached 90% for women aged 50 and those aged < 50 years. The event rates in the second 5 years after diagnosis for stage I and II cancer were comparable between the younger and older cohort, except for stage III disease; women aged 50 years with stage III disease had greater event rates. For the entire study population, the older and younger cohorts had similar BCSS. No difference was found between BCSS and OS in women aged < 50 years at diagnosis; however, OS was 10% lower than BCSS in women aged > 50 years.
Effect of Grade and Stage .08
1-3
No breast surgery
‡50 (n [ 6615)
Ten-Year Mortality and BC Event Rates
<.001
Nodal status
Unknown
Breast/chest wall plus regional nodes
<50 (n [ 1886)
Initial chemotherapy
Median
Unknown No. of positive nodes
a
Regional nodal RT alone
Tumor size, cm
Unknown
Characteristic
The effect of grade on the 10-year EFS was considered for each disease stage (I, II, and III; Table 3). EFS decreased as the stage and grade increased in both age groups. Stage III disease was associated with at least a 20% recurrence risk after 5 years, regardless of patient age or tumor grade. Among those with stage III cancer, some grade and age subgroups had large 95% CIs owing to the small case and event numbers. The effect of grade was also minimal in stage I disease, which was characterized by an excellent prognosis beyond 5 years, with point estimates for EFS of > 93% for all age and grade subgroups. Using a more conservative approach that considered the 95% CI, women aged < 50 years at diagnosis of stage I, grade 3 disease had a lower boundary of the 95% CI of < 90% for EFS, but all others had a lower 95% CI boundary of 90%. To further clarify the risk for those with stage II cancer, we explored the effect of nodal status and grade on EFS. A low grade was more predictive of a low risk of recurrence than was nodal status (Figure 1). The point estimates of recurrence were < 10% for stage II, grade 1 cancer in both younger and older women; however, the upper boundary of the 95% CI was > 10% for women aged < 50 years. Grade 2 and 3 cancers had a > 10% recurrence risk at 10 years, whether node negative or positive. No additional discriminatory effect was seen for the strength of ER positivity (P ¼ .66) or receipt of chemotherapy (P ¼ .09) on univariate analysis; however, the presence of lymphovascular invasion (LVI) was associated with significantly worse EFS (82.6%; 95% CI, 80.5-84.5) compared with no LVI (89.6%; 95% CI, 88.3-90.8; P < .001).
Clinical Breast Cancer Month 2015
-3
Risk More Than 5 Years After HR+ BC Diagnosis Table 2 Ten-Year Outcomes for Women Without an Event in Years 0 to 5 10-Year Outcome
Age ‡50 Years (n [ 6615)
Age <50 Years (n [ 1886)
EFS BCSS OS
89 (88.1-89.7) 94.9 (94.3-95.4) 84.6 (83.7-85.5)
89.8 (88.3-91.2) 96.3 (95.3-97.1) 95.5 (94.4-96.4)
Data presented as % (95% confidence interval). Abbreviations: BCSS ¼ breast cancer-specific survival; EFS ¼ event-free survival; OS ¼ overall survival.
Discussion With recurrence and death rates in the control arm of the ATLAS study of 25% and 15% at 15 years, respectively, an absolute survival benefit of 2.8% for 10 years of tamoxifen was observed, representing a number needed to treat of 36 to avoid 1 death. Populations with a lower event rate can expect a smaller absolute benefit, against which the risks of serious and bothersome toxicity must be balanced. This provided a motivation to identify those with an excellent prognosis and patients who might safely stop AHT after 5 years. Within our data set of 10,414 unique cases, the overall EFS rate over 10 years was 27%, with an almost equal proportion of events occurring within and after 5 years from the initial diagnosis. We postulated that women with a risk in years 5 through 10 of < 10% could choose to avoid extended hormone therapy without meaningfully altering their recurrence risk. Because some events are curable local recurrences and/ or CLBCs, the omission of extended hormone therapy in BC populations with low event rates > 5 years after the original diagnosis would be exceptionally unlikely to affect BCSS or OS. In our study, the event rates at 10 years were similar for women aged < 50 and those aged 50 years in similar stage and grade categories. These observations and the subgroup analyses of the ATLAS study support the finding that extended therapy benefits are generalizable across age groups. We found a correlation between risk and stage and grade in both age groups. Among younger and older women with stage I cancer of any grade and no relapse in the first 5
years, the combined rates of recurrence, CLBC, and BC death without recurrence at 10 years were < 7%. Relatively few women aged < 50 years had stage I, grade 3 cancer, and the event rate in this subgroup was small. The lower boundary of the 95% CI for EFS was < 90%, suggesting caution in estimating the risk for this subgroup. Although the recurrence risks were greater for node-positive than for node-negative stage II cancer, consistent with the findings from other reports, our analysis suggests that grade, which is reflective of tumor biology, has a greater influence on recurrence risk than nodal status in those with stage II disease.15,16 Using point estimates, stage II, grade 1 cancer cases were associated with excellent outcomes in both older (EFS, 93%) and younger (EFS, 94%) women and in the node-negative (EFS, 95%) and node-positive (EFS, 92%) subsets, suggesting that, similar to women with stage I cancer, these groups can safely avoid 10 years of hormone therapy with minimal or no effect on their recurrence rates or BCSS. If taking a more conservative approach by considering the lower boundary of the 95% CI to estimate risk, women whose stage II, grade 1 cancer was associated with LVI or who had been diagnosed when they were < 50 years did not meet the 10% threshold risk. Grade 2 and 3, stage II tumors resulted in a > 10% event risk at 10 years, regardless of nodal status. Among those with stage III cancer, no grade or age subset had a < 10% recurrence risk after 5 years from diagnosis. For women aged < 50 years at diagnosis, the 10-year BCSS and OS were similar, suggesting few competing causes of mortality in younger women. In contrast, women aged 50 years at diagnosis had an approximately 10% lower OS than BCSS at 10 years, highlighting the need to consider comorbidities in estimating the survival benefits of extended AHT in older women. Using our results, we have constructed an algorithm to aid patients and clinicians considering AHT (Figure 2).
Study Limitations The present study had several important limitations to consider before applying these data. Given the era in which these cancer
Table 3 EFS Stratified by Age, Stage, and Grade for Women Without an Event in Years 0 to 5 Age ‡50 Years at Diagnosis Stage and Grade
n
Events
10-Year EFS (%)
n
Events
2798
204
94.8 (93.8-91.0)
677
47
94.8 (92.8-96.3)
Grade 1
1037
57
95.4 (93.7-96.6)
225
11
97.3 (94.0-98.8)
Grade 2
1338
111
94.6 (93.1-95.7)
334
28
93.4 (90.0-95.6)
Grade 3
364
29
94.8 (91.8-96.8)
118
8
94.4 (88.0-97.5)
Stage I
Stage II
3330
589
86.3 (85.0-87.5)
991
156
88.3 (86.0-90.2)
605
61
93.0 (90.5-94.9)
151
15
93.8 (87.7-96.9)
Grade 2
1773
307
86.6 (84.8-88.2)
496
85
87.7 (84.4-90.4)
Grade 3
823
187
81.9 (78.9-84.5)
344
56
87.9 (83.8-91.1)
487
130
73.8 (69.1-77.8)
187
38
80.4 (73.6-85.6)
Grade 1
53
11
90.5 (78.6-95.9)
16
1
90.9 (50.8-98.7)
Grade 2
252
67
74.3 (67.8-79.8)
96
24
77.3 (66.7-84.8)
Grade 3
150
40
72.0 (62.8-79.3)
75
13
82.6 (70.0-90.3)
Data in parentheses are 95% confidence intervals. Abbreviation: EFS ¼ event-free survival.
-
10-Year EFS (%)
Grade 1
Stage III
4
Age <50 Years at Diagnosis
Clinical Breast Cancer Month 2015
Sheridan Wilson et al Figure 1 Event-Free Survival for Stage II Estrogen Receptor-Positive Breast Cancer Stratified by Nodal Status for Age (A) < 50 Years and (B) ‡ 50 Years, by Grade for Age (C) < 50 Years and (D) ‡ 50 Years, and by Grade for (E) Node-Negative and (F) NodePositive Disease
cases were diagnosed, human epidermal growth factor-2 (HER2) status was unavailable for our study population, and no patient received adjuvant trastuzumab. The routine use of trastuzumab
would have resulted in lower event rates than we observed, with the effect likely to be more pronounced in higher grade subgroups, because HER2þ status occurs most frequently with grade 3
Clinical Breast Cancer Month 2015
-5
Risk More Than 5 Years After HR+ BC Diagnosis Figure 2 Algorithm to Guide 5 Versus 10 Years of Hormone Therapy After Breast Cancer, According to 10% Residual Threshold Risk of Event-Free Survival Event in Years 5 to 10 After Diagnosis
Age at diagnosis
<10% risk of event
<50 years
Stage I Grades 1, 2, 3 Stage II Grade 1 Stage I Grade 1 Stage II Grade 1
50+ years
<10% risk of event, including 95% CI Stage I Grades 1, 2 Stage II Grade 1 Stage I Grade 1 Stage II Grade 1
>10% risk of event in years 510 Stage II Grades 2, 3 Stage III All grades Stage II Grades 2, 3 Stage III All grades
Hormone receptor positive breast cancer with no recurrence, death, or CLBC in first 5 years after diagnosis
Stage I
Stage II, node negative and node positive 5 years AHT : grade 1, 2 OR Figure 2. Algorithm to guide 5 versus 10 years of hormone therapy after breast cancer, based on 10% residual
5 or 10 years AHT: Grade 1 Or Low risk by RS or similar assay
5 or 10 years AHT: grade 3 and <50 years old at dx OR intermediate risk by RS or similar assay (any age)
10 years AHT: Grade 2 or 3 OR Intermediate or high RS OR high risk by similar assay
Stage III
10 years AHT: all grades
10 years AHT: high risk by RS or similar assay
Abbreviations: AHT ¼ adjuvant hormone therapy; CI ¼ confidence interval; CLBC ¼ contralateral breast cancer; RS ¼ 21-gene recurrence score assay.
6
-
histologic features.17 Other contemporary treatments such as modern chemotherapy and regional radiotherapy reduce relapse rates; therefore, the event rates observed in our cohorts might have been a percentage point or 2 greater than expected today. This effect has been seen in a number of recent clinical trials, in which the control arm event rates were lower than had been estimated when designing the trial, presumably owing to improvements in elements of care not being studied but systematically applied within the trial populations. We were unable to confirm AHT adherence in our population. Previous reports, including from our own institution, have suggested high nonadherence rates in the general population.18 Because ours was a population-based study, we assumed an average rate of nonadherence and would anticipate better outcomes in a perfectly compliant population. When discussing extended AHT with a woman known to have been compliant for the first 5 years, it is likely that her residual recurrence risk will not be more than observed for her age, stage, and grade cohort in the present data set. Hormone therapy recommendations changed during the study period; increasing for postmenopausal women from 2 to 3 years in 1992 and to 5 years in 1995. In 1993, 3 years of tamoxifen use was
Clinical Breast Cancer Month 2015
introduced for premenopausal women and was extended to 5 years in 1996. Thus, many of the younger women in the present study had received less than optimal exposure to tamoxifen by current standards. We observed an EFS rate of 87.7% and 87.9% among women aged < 50 years with stage II, grade 2 or 3 cancer, respectively. Given a 30% relative risk reduction with 5 years of tamoxifen, these cohorts might, with modern treatment standards, experience an event risk of close to 10% at 10 years if event free by year 5; however, this remains speculative.1 The use of AI therapy for 2 to 5 years was introduced for postmenopausal ERþ BC at BCCA in July 2005. Only women diagnosed in the most recent 2 to 3 years of the present study would thus have received an adjuvant AI. The relatively modest additional benefit of AIs compared with tamoxifen would be unlikely to improve the 10-year BC event rate to < 10% in the stage and grade cohorts aged > 50 years with a demonstrated risk > 10% in our study population.9 Also, 12% of younger women and 7% of older women with stage I cancer in our cohorts filled 1 prescription for an AI after 5 years of tamoxifen. This might result in 1% to 2% lower recurrence rates and would not meaningfully alter the excellent prognosis we observed for those with stage I disease.
Sheridan Wilson et al The changes in hormone therapy policies described would improve the EFS rates for all groups. Similarly, improvements in chemotherapy and local therapy over time might result in lower rates than we observed. Thus, we have postulated that the groups of women for whom we observed a < 10% event rate would still experience a low event rate in the context of contemporary therapy. We believe these data can accurately identify populations with a low residual risk after 5 years of AHT.
inform their decision to continue or discontinue AHT after 5 years.
Acknowledgments This work was partly funded by an unrestricted educational grant from Astra Zeneca.
Disclosure Conclusion Our study was designed to provide estimates of a BC event, including CLBC, > 5 years from diagnosis, to help contextualize the potential benefits of extended AHT. Tissue-based assays such as the 21-gene recurrence score assay have been shown to provide prognostic information about recurrence in years 5 to 10 and might be an additional mechanism to identify women who can safely stop after 5 years of AHT.19 This test is not available in all jurisdictions or for many women diagnosed several years earlier; however, pathologic variables, which are readily available for all cases, appear to provide good risk discrimination. The side effects experienced in the first 5 years and competing potential causes of mortality will also have relevance for an individual contemplating this choice. We believe that women of all ages who remain free of recurrence 5 years from the initial diagnosis of ERþ stage I cancer of any grade (with the possible exception of women aged < 50 years at diagnosis of a stage I, grade 3 cancer) or stage II, grade I cancer (with the possible exception of those aged < 50 years at diagnosis or with associated LVI) might reasonably avoid 10 years of hormone therapy because they have a 10-year combined risk of recurrence, CLBC, or BC death of 10%. Women with such cancer who wish to discontinue AHT after 5 years can be reassured that the additional protection of longer therapy is exceedingly small and possibly negligible for them. For other stage and grade cancer combinations, in which the BC event risk after 5 years is > 10%, a careful discussion of the potential benefits and harm of continued AHT is warranted.
Clinical Practice Points Recurrences are distributed across > 10 years after a diagnosis of
HRþ BC. Most of the recurrence risk reduction conferred by AHT results from the first 5 years of use, with more modest benefits in years 5 to 10 of continued use. Our population-based research suggests that women with stage II grade 1 cancer and stage I cancer, who have not experienced recurrence in the first 5 years can expect a < 10% future recurrence risk, making discontinuation of hormone therapy after 5 years a safe option for them. Many women struggle with hormone therapy-related side effects but also with fears of BC recurrence. The recurrence estimates for the second 5 years after diagnosis, stratified by the initial disease presentation, will better
The authors have stated that they have no conflicts of interest.
References 1. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: An overview of the randomised trials. Lancet 2005; 365: 1687-717. 2. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), Davies C, Godwin J, Gray R, et al. Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: patient-level meta-analysis of randomised trials. Lancet 2011; 378:771-84. 3. Fisher B, Dignam J, Bryant J, et al. Five versus more than five years of tamoxifen for lymph node-negative breast cancer: updated findings from the National Surgical Adjuvant Breast and Bowel Project B-14 randomized trial. J Natl Cancer Inst 2001; 93:684-90. 4. Singh L, Wilson AJ, Baum M, et al. The relationship between histological grade, oestrogen receptor status, events and survival at 8 years in the NATO (“nolvadex”) trial. Br J Cancer 1988; 57:612-4. 5. “Nolvadex” Adjuvant Trial Organisation. Controlled trial of tamoxifen as a single adjuvant agent in the management of early breast cancer. Br J Cancer 1988; 57:608-11. 6. Adjuvant tamoxifen in the management of operable breast cancer: the Scottish trial. Report from the Breast Cancer Trials Committee, Scottish Cancer Trials Office (MRC), Edinburgh. Lancet 1987; 2:171-5. 7. Stewart HJ, Forrest AP, Everington D, et al. Randomised comparison of 5 years of adjuvant tamoxifen with continuous therapy for operable breast cancer: the Scottish Cancer Trials Breast Group. Br J Cancer 1996; 74:297-9. 8. Fisher B, Costantino J, Redmond C, et al. A randomized clinical trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have estrogen-receptor-positive tumors. N Engl J Med 1989; 320:479-84. 9. Dowsett M, Cuzick J, Ingle J, et al. Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol 2010; 28:509-18. 10. Delozier T, Spielmann M, Mace-Lesec’h J, et al. Tamoxifen adjuvant treatment duration in early breast cancer: initial results of a randomized study comparing short-term treatment with long-term treatment. Federation Nationale des Centres de Lutte Contre le Cancer Breast Group. J Clin Oncol 2000; 18:3507-12. 11. Goss PE, Ingle JN, Martino S, et al. Efficacy of letrozole extended adjuvant therapy according to estrogen receptor and progesterone receptor status of the primary tumor: National Cancer Institute of Canada clinical trials group MA.17. J Clin Oncol 2007; 25:2006-11. 12. Davies C, Pan H, Godwin J, et al. Long-term effects of continuing adjuvant tamoxifen to 10 years versus stopping at 5 years after diagnosis of oestrogen receptorpositive breast cancer: ATLAS, a randomised trial. Lancet 2013; 381:805-16. 13. Gray R, Rea D, Handley K, et al. aTTom: long-term effects of continuing adjuvant tamoxifen to 10 years versus stopping at 5 years in 6,953 women with early breast cancer. J Clin Oncol 2013; 31(suppl), abstract 5. 14. BC Cancer Agency. BC Cancer Agency, Health Professionals Information. Available at: http://www.bccancer.bc.ca/HPI/default.htm. Accessed 2014. 15. Kennecke HF, Olivotto IA, Speers C, et al. Late risk of relapse and mortality among postmenopausal women with estrogen responsive early breast cancer after 5 years of tamoxifen. Ann Oncol 2007; 18:45-51. 16. Brewster AM, Hortobagyi GN, Broglio KR, et al. Residual risk of breast cancer recurrence 5 years after adjuvant therapy. J Natl Cancer Inst 2008; 100:1179-83. 17. Hu Z, Fan C, Oh DS, et al. The molecular portraits of breast tumors are conserved across microarray platforms. BMC Genomics 2006; 7:96. 18. Hershman DL, Kushi LH, Shao T, et al. Early discontinuation and nonadherence to adjuvant hormonal therapy in a cohort of 8,769 early-stage breast cancer patients. J Clin Oncol 2010; 28:4120-8. 19. Dowsett M, Cuzick J, Wale C, et al. Prediction of risk of distant recurrence using the 21-gene Recurrence Score in node-negative and node-positive postmenopausal patients with breast cancer treated with anastrozole or tamoxifen: a TransATAC study. J Clin Oncol 2010; 28:1829-34.
Clinical Breast Cancer Month 2015
-7
Risk of Recurrence or Contralateral Breast Cancer More than 5 Years After Diagnosis of Hormone Receptor-Positive Early-Stage Breast Cancer Sheridan Wilson,1 Caroline Speers,1 Scott Tyldesley,1 Stephen Chia,1 Hagen Kennecke,1 Susan Ellard,2 Caroline Lohrisch1 Abstract Adjuvant hormone therapy in years 5 to 10 has shown a modest survival advantage that must be balanced against the side effects associated with continued use. We identified cancer cases with < 10% recurrence risk > 5 years after breast cancer diagnosis to inform discussions between clinicians and patients about the individual benefit of continuing hormone therapy after 5 years. Background: Three large studies have shown a survival benefit from 10 years of adjuvant hormone therapy (AHT). We evaluated the risk of an event 5 years after the initial breast cancer (BC) diagnosis and identified the prognostic factors to assist clinicians considering extended AHT. Patients and Methods: Patients newly referred to the BC Cancer Agency with stage I to III estrogen receptor-positive BC diagnosed from 1989 to 2004 who had undergone AHT were identified by the BC Cancer Agency’s Breast Cancer Outcomes Unit. Cases with recurrence, death, or contralateral BC occurring within the first 5 years were excluded. The 10-year event-free survival (EFS) and 95% confidence intervals (CIs) were calculated using the Kaplan-Meier method. This provided estimates of recurrence risk after the fifth year following the diagnosis. The histopathologic and age variables were examined for prognostic value by univariate analysis. Results: Within our cohort, 6615 women were postmenopausal and 1886 were premenopausal at the BC diagnosis. The median follow-up period was 11 years. The 10-year EFS for women aged < 50 years with stage I, II, and III disease was 94.8% (95% CI, 92.8%-96.3%), 88.3% (95% CI, 86.0%-90.2%), and 80.4% (95% CI, 73.6%-85.6%), respectively. Among women aged 50 years, the corresponding EFS rates were 94.8% (95% CI, 93.8%-95.6%), 86.3% (95% CI, 85.0%-87.5%), and 73.8% (95% CI, 69.1%-77.8%). EFS varied significantly by grade. The 10-year recurrence risk was < 10% with stage I cancer (any grade) and for stage II (node-negative and node-positive), grade I cancer. Conclusion: Our data have identified BCs associated with a very low recurrence risk 5 to 10 years after diagnosis, providing women with such cancers confidence about a decision to discontinue AHT after 5 years. Clinical Breast Cancer, Vol. -, No. -, --- ª 2015 Elsevier Inc. All rights reserved. Keywords: Adjuvant hormone therapy, Five versus 10 years of therapy, Good prognosis groups, Hormone receptor-positive breast cancer, Late recurrence risk
Introduction Breast cancer (BC) accounts for 5000 deaths among Canadian women annually and is the leading cause of cancer-related death in women < 50 years old.1 Most BCs arising in premenopausal and 1
BC Cancer Agency, Vancouver Cancer Centre, Vancouver, BC, Canada BC Cancer Agency, Centre for the Southern Interior, Kelowna, BC, Canada
2
Submitted: Apr 16, 2015; Accepted: Nov 10, 2015 Address for correspondence: Caroline Lohrisch, MD, FRCPC, Medical Oncology Department, BC Cancer Agency, Vancouver Cancer Centre, 600 West 10th Avenue, Vancouver, BC V5Z 4E6 Canada E-mail contact: [email protected]
1526-8209/$ - see frontmatter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clbc.2015.11.002
postmenopausal women are estrogen receptor-positive (ERþ). In this population, adjuvant hormone therapy (AHT) significantly reduces the risk of BC recurrence and BC-related death.1,2 The efficacy of adjuvant tamoxifen was established within large randomized trials reported in the 1980s.3-6 The updated results from the Early Breast Cancer Trialists’ Collaborative Group meta-analysis of > 20,000 women receiving approximately 5 years of tamoxifen within 20 randomized trials demonstrated that the benefit of tamoxifen continues well beyond the treatment period, reducing the 15-year risks of BC recurrence and death.2,6-8 Since 2004, the American Society of Clinical Oncology guidelines have recommended an aromatase inhibitor (AI) as a part of AHT for
Clinical Breast Cancer Month 2015
-1
Risk More Than 5 Years After HR+ BC Diagnosis postmenopausal hormone receptor-positive (HRþ) BC based on the results of many randomized controlled trials.9 Late relapses (> 5 years from diagnosis) of ERþ BC remain an important problem. A significant proportion of ERþ BC recurrences and two thirds of BC deaths occur > 5 years from diagnosis.1 Until recently, the optimal duration of AHT was considered 5 years, because the initial trials of > 5 years failed to show a consistent benefit.3,10 However, the findings from 3 large randomized controlled trials now support a further survival advantage of 2.5% to 3% by extending the duration of AHT from 5 years to 10 for both pre- and postmenopausal women with HRþ early-stage BC. MA.17, a randomized placebo-controlled trial, evaluated the addition of 5 years of letrozole after 5 years of tamoxifen, and the Adjuvant Tamoxifen Longer Against Shorter (ATLAS) and Adjuvant Tamoxifen Treatment offers More trials investigated the extension of tamoxifen to 10 years.11-13 Although the results of the subgroup analyses of these trials suggested that all stage and age subgroups might benefit from extended AHT, the survival benefits were lower in the groups with a lower baseline risk. Women with very low or no residual risk of recurrence after 5 years are still at risk of harm from continued AHT but can benefit only by virtue of secondary prevention (0.5%-1.0% absolute reduction, determined by the excess incidence of contralateral breast cancer [CLBC] observed in the control arms of the MA.17 and ATLAS trials). The ATLAS study reported a 1.5% excess risk of endometrial cancer (0.2% excess mortality) and a 0.3% excess risk of pulmonary embolus but no overall increase in non-BC mortality associated with 10 versus 5 years of tamoxifen. In the MA.17 study, at 30 months of follow-up, numerically, but not statistically significantly, more cardiovascular events, osteoporosis diagnoses, and fractures were observed in the letrozole group. In addition, a 5% excess rate of grade 1 to 3 arthralgia was found. A challenge for oncologists is how to apply these important findings to a modern-day population. The evidence for extended tamoxifen use was generated before the availability of adjuvant AIs. Modern chemotherapy, hormone therapy, and radiotherapy techniques are known to influence BC recurrence and survival rates. Thus, the benefit of extended AHT might be less for women diagnosed and treated today. Considering the possible serious harm and long-term effects on quality of life associated with extended AHT, an understanding of the patient and tumor characteristics associated with recurrence > 5 years after the diagnosis of HRþ BC will help clinicians counsel women with low residual recurrence risk who might reasonably and safely choose to stop AHT after 5 years. We examined the prognostic factors for BC events in women alive and free of BC relapse or CLBC after 5 years of AHT. Our primary objective was to identify patients with an AHT event-free survival (EFS) rate at 10 years of > 90%, including the lower boundary of the 95% confidence intervals (CIs), who, on this basis, might reasonably be spared another 5 years of AHT.
Patients and Methods
2
-
Population-based data are prospectively collected and maintained by the BC Cancer Agency (BCCA) Breast Cancer Outcomes Unit for cases diagnosed since 1989. This database captures all patients with BC referred to the BCCA and represents an estimated 85% of
Clinical Breast Cancer Month 2015
all BCs diagnosed in British Columbia annually. Referred women diagnosed with stage I to III invasive ERþ BC from 1989 to 2004 and who had begun AHT were identified for the present study. The exclusion criteria were previous invasive or in situ disease and synchronous bilateral BC. Women aged < 50 years whose initial AHT was an AI were excluded. Disease stage was assigned using the American Joint Committee on Cancer Staging criteria. The pathology characteristics abstracted included the degree of ER positivity, tumor size and grade, and nodal status. ER positivity was divided into low, moderate, and high expression using the following definitions: for the biochemical method, low, 10 to 50 fmol/mg; moderate, 51 to 100 fmol/mg; and high, > 100 fmol/mg; for the immunohistochemical method, low, 1þ or Allred score of 3 or 4; moderate, 2þ or Allred score of 5 or 6; and high, 3þ or Allred score of 7 or 8. When ER expression was only described as positive, it was assigned a value of moderate. The study population was dichotomized based on age (< 50 or 50 years) at diagnosis to serve as a surrogate for menopausal status. Patients underwent surgery followed by adjuvant radiotherapy and systemic therapy, directed by evidence-based provincial guidelines maintained and disseminated by the BCCA.14 The primary outcome was EFS, defined as the interval to the first of local (breast or chest wall), regional (axillary, infra- and/or supraclavicular or internal mammary lymph nodes), or distant recurrence; new CLBC diagnosis; or death from BC without previous documented recurrence. The analysis of the 10-year EFS was confined to women without a BC event or death within the first 5 years after the diagnosis. BC-specific survival (BCSS), defined as the interval from diagnosis to BC death, and overall survival (OS), defined as the interval from diagnosis to death from any cause, were also calculated for the 2 age cohorts. Survival was estimated using the Kaplan-Meier method and compared using log-rank tests. All statistical analyses were performed with SPSS, version 14.0, software (SPSS Inc., Chicago IL). The Research Ethics Board of the BCCA, University of British Columbia, approved the study.
Results Before the exclusions, 10,815 ERþ stage I to III BC cases diagnosed from 1989 to 2004 and referred to the BCCA were identified. After excluding the cases with previous invasive or in situ BC and those with synchronous bilateral BC, 10,414 unique cases remained. During the first 5 years after diagnosis, EFS events occurred in 1443 cases, representing 13% of incident cases and 54% of all EFS events. This left 8526 patients (82%) alive and free of a BC event at 5 years after the initial diagnosis, forming the study population for our analysis. The median follow-up period was 11 years. A total of 1207 EFS events occurred > 5 years from diagnosis among these 8526 cases, representing 12% of the 10,414 incident cases and 45% of all events. The patient and tumor characteristics are listed in Table 1. A total of 25 patients (1.3%) aged < 50 years at diagnosis who had received an AI as their initial AHT were excluded. Of the remaining 1886 women aged < 50 years at diagnosis, 1692 (90%) were premenopausal and 147 (8%) were postmenopausal (the menopausal status was unknown in 47 patients [2%]). The patients for whom the tumor grade was unknown were included in the overall EFS calculations but excluded from the analysis by tumor grade.
Sheridan Wilson et al Table 1 Patient and Tumor Characteristics for Patients Remaining Event-Free 5 Years After Initial Diagnosis
Table 1 Continued Age (Years)
Age (Years) Characteristic
<50 (n [ 1886)
‡50 (n [ 6615)
Age at diagnosis (years)
P Value NA
Median
45.0
64
Range
24-49
50-92
2.0
1.8
Range
0.1-9.9
0.1-9.9 <.001
Tumor size group, cm 0.1-0.5
49 (2.6)
196 (3.0)
0.6-1.0
222 (11.8)
940 (14.2)
1.1-2.0
798 (42.3)
2900 (43.8)
2.1-3.0
459 (24.3)
1680 (25.4)
3.1-5.0
222 (11.8)
583 (8.8)
5.1-9.9
130 (6.9)
279 (4.2)
6 (0.3)
37 (0.6)
Negative
955 (50.6)
4046 (61.2)
Positive
931 (49.4)
2560 (38.7)
Unknown
0 (0.0)
9 (0.1)
Median nodes sampled
11
10
Range nodes sampled
1-42
1-48
<.001
Positive nodes (n)
Positive nodal group
704 (37.3)
1423 (21.5)
2 (0.1)
13 (0.2)
654 (71.5)
1908 (75.1)
4-9
212 (23.2)
507 (20.0)
10
48 (5.2)
119 (4.7)
1 (0.1)
6 (0.2) <.001
LVI status Negative
1212 (64.3)
4639 (70.1)
Positive
617 (32.7)
1700 (25.7)
Unknown
57 (3.0)
276 (4.2)
1
392 (20.8)
1695 (25.6)
2
926 (49.1)
3363 (50.8)
3
537 (28.5)
1337 (20.2)
31 (1.6)
220 (3.3)
7 (0.4)
32 (0.5)
<.001
Grade
Initial surgery
.02
Breast-conserving surgery
1070 (56.7)
3952 (59.7)
Total mastectomy
809 (42.9)
2631 (39.8)
No breast/chest wall or nodal RT
366 (19.4)
1937 (29.3)
Breast/chest wall alone
814 (43.2)
3242 (49.0)
<.001
Initial radiotherapy
P Valuea
<.001
Yes
1381 (73.2)
1545 (23.4)
No
505 (26.8)
5070 (76.6)
Data presented as n (%), unless otherwise noted. Abbreviations: LVI ¼ lymphovascular invasion; NA ¼ not applicable; RT ¼ radiation therapy. a P values calculated from known values only; test statistics used were c2 test for categorical variables and Mann-Whitney U test for continuous variables.
The 10-year outcomes for women without a BC cancer event during the first 5 years after diagnosis are listed in Table 2. Considering all stages and grades, the 10-year EFS approached 90% for women aged 50 and those aged < 50 years. The event rates in the second 5 years after diagnosis for stage I and II cancer were comparable between the younger and older cohort, except for stage III disease; women aged 50 years with stage III disease had greater event rates. For the entire study population, the older and younger cohorts had similar BCSS. No difference was found between BCSS and OS in women aged < 50 years at diagnosis; however, OS was 10% lower than BCSS in women aged > 50 years.
Effect of Grade and Stage .08
1-3
No breast surgery
‡50 (n [ 6615)
Ten-Year Mortality and BC Event Rates
<.001
Nodal status
Unknown
Breast/chest wall plus regional nodes
<50 (n [ 1886)
Initial chemotherapy
Median
Unknown No. of positive nodes
a
Regional nodal RT alone
Tumor size, cm
Unknown
Characteristic
The effect of grade on the 10-year EFS was considered for each disease stage (I, II, and III; Table 3). EFS decreased as the stage and grade increased in both age groups. Stage III disease was associated with at least a 20% recurrence risk after 5 years, regardless of patient age or tumor grade. Among those with stage III cancer, some grade and age subgroups had large 95% CIs owing to the small case and event numbers. The effect of grade was also minimal in stage I disease, which was characterized by an excellent prognosis beyond 5 years, with point estimates for EFS of > 93% for all age and grade subgroups. Using a more conservative approach that considered the 95% CI, women aged < 50 years at diagnosis of stage I, grade 3 disease had a lower boundary of the 95% CI of < 90% for EFS, but all others had a lower 95% CI boundary of 90%. To further clarify the risk for those with stage II cancer, we explored the effect of nodal status and grade on EFS. A low grade was more predictive of a low risk of recurrence than was nodal status (Figure 1). The point estimates of recurrence were < 10% for stage II, grade 1 cancer in both younger and older women; however, the upper boundary of the 95% CI was > 10% for women aged < 50 years. Grade 2 and 3 cancers had a > 10% recurrence risk at 10 years, whether node negative or positive. No additional discriminatory effect was seen for the strength of ER positivity (P ¼ .66) or receipt of chemotherapy (P ¼ .09) on univariate analysis; however, the presence of lymphovascular invasion (LVI) was associated with significantly worse EFS (82.6%; 95% CI, 80.5-84.5) compared with no LVI (89.6%; 95% CI, 88.3-90.8; P < .001).
Clinical Breast Cancer Month 2015
-3
Risk More Than 5 Years After HR+ BC Diagnosis Table 2 Ten-Year Outcomes for Women Without an Event in Years 0 to 5 10-Year Outcome
Age ‡50 Years (n [ 6615)
Age <50 Years (n [ 1886)
EFS BCSS OS
89 (88.1-89.7) 94.9 (94.3-95.4) 84.6 (83.7-85.5)
89.8 (88.3-91.2) 96.3 (95.3-97.1) 95.5 (94.4-96.4)
Data presented as % (95% confidence interval). Abbreviations: BCSS ¼ breast cancer-specific survival; EFS ¼ event-free survival; OS ¼ overall survival.
Discussion With recurrence and death rates in the control arm of the ATLAS study of 25% and 15% at 15 years, respectively, an absolute survival benefit of 2.8% for 10 years of tamoxifen was observed, representing a number needed to treat of 36 to avoid 1 death. Populations with a lower event rate can expect a smaller absolute benefit, against which the risks of serious and bothersome toxicity must be balanced. This provided a motivation to identify those with an excellent prognosis and patients who might safely stop AHT after 5 years. Within our data set of 10,414 unique cases, the overall EFS rate over 10 years was 27%, with an almost equal proportion of events occurring within and after 5 years from the initial diagnosis. We postulated that women with a risk in years 5 through 10 of < 10% could choose to avoid extended hormone therapy without meaningfully altering their recurrence risk. Because some events are curable local recurrences and/ or CLBCs, the omission of extended hormone therapy in BC populations with low event rates > 5 years after the original diagnosis would be exceptionally unlikely to affect BCSS or OS. In our study, the event rates at 10 years were similar for women aged < 50 and those aged 50 years in similar stage and grade categories. These observations and the subgroup analyses of the ATLAS study support the finding that extended therapy benefits are generalizable across age groups. We found a correlation between risk and stage and grade in both age groups. Among younger and older women with stage I cancer of any grade and no relapse in the first 5
years, the combined rates of recurrence, CLBC, and BC death without recurrence at 10 years were < 7%. Relatively few women aged < 50 years had stage I, grade 3 cancer, and the event rate in this subgroup was small. The lower boundary of the 95% CI for EFS was < 90%, suggesting caution in estimating the risk for this subgroup. Although the recurrence risks were greater for node-positive than for node-negative stage II cancer, consistent with the findings from other reports, our analysis suggests that grade, which is reflective of tumor biology, has a greater influence on recurrence risk than nodal status in those with stage II disease.15,16 Using point estimates, stage II, grade 1 cancer cases were associated with excellent outcomes in both older (EFS, 93%) and younger (EFS, 94%) women and in the node-negative (EFS, 95%) and node-positive (EFS, 92%) subsets, suggesting that, similar to women with stage I cancer, these groups can safely avoid 10 years of hormone therapy with minimal or no effect on their recurrence rates or BCSS. If taking a more conservative approach by considering the lower boundary of the 95% CI to estimate risk, women whose stage II, grade 1 cancer was associated with LVI or who had been diagnosed when they were < 50 years did not meet the 10% threshold risk. Grade 2 and 3, stage II tumors resulted in a > 10% event risk at 10 years, regardless of nodal status. Among those with stage III cancer, no grade or age subset had a < 10% recurrence risk after 5 years from diagnosis. For women aged < 50 years at diagnosis, the 10-year BCSS and OS were similar, suggesting few competing causes of mortality in younger women. In contrast, women aged 50 years at diagnosis had an approximately 10% lower OS than BCSS at 10 years, highlighting the need to consider comorbidities in estimating the survival benefits of extended AHT in older women. Using our results, we have constructed an algorithm to aid patients and clinicians considering AHT (Figure 2).
Study Limitations The present study had several important limitations to consider before applying these data. Given the era in which these cancer
Table 3 EFS Stratified by Age, Stage, and Grade for Women Without an Event in Years 0 to 5 Age ‡50 Years at Diagnosis Stage and Grade
n
Events
10-Year EFS (%)
n
Events
2798
204
94.8 (93.8-91.0)
677
47
94.8 (92.8-96.3)
Grade 1
1037
57
95.4 (93.7-96.6)
225
11
97.3 (94.0-98.8)
Grade 2
1338
111
94.6 (93.1-95.7)
334
28
93.4 (90.0-95.6)
Grade 3
364
29
94.8 (91.8-96.8)
118
8
94.4 (88.0-97.5)
Stage I
Stage II
3330
589
86.3 (85.0-87.5)
991
156
88.3 (86.0-90.2)
605
61
93.0 (90.5-94.9)
151
15
93.8 (87.7-96.9)
Grade 2
1773
307
86.6 (84.8-88.2)
496
85
87.7 (84.4-90.4)
Grade 3
823
187
81.9 (78.9-84.5)
344
56
87.9 (83.8-91.1)
487
130
73.8 (69.1-77.8)
187
38
80.4 (73.6-85.6)
Grade 1
53
11
90.5 (78.6-95.9)
16
1
90.9 (50.8-98.7)
Grade 2
252
67
74.3 (67.8-79.8)
96
24
77.3 (66.7-84.8)
Grade 3
150
40
72.0 (62.8-79.3)
75
13
82.6 (70.0-90.3)
Data in parentheses are 95% confidence intervals. Abbreviation: EFS ¼ event-free survival.
-
10-Year EFS (%)
Grade 1
Stage III
4
Age <50 Years at Diagnosis
Clinical Breast Cancer Month 2015
Sheridan Wilson et al Figure 1 Event-Free Survival for Stage II Estrogen Receptor-Positive Breast Cancer Stratified by Nodal Status for Age (A) < 50 Years and (B) ‡ 50 Years, by Grade for Age (C) < 50 Years and (D) ‡ 50 Years, and by Grade for (E) Node-Negative and (F) NodePositive Disease
cases were diagnosed, human epidermal growth factor-2 (HER2) status was unavailable for our study population, and no patient received adjuvant trastuzumab. The routine use of trastuzumab
would have resulted in lower event rates than we observed, with the effect likely to be more pronounced in higher grade subgroups, because HER2þ status occurs most frequently with grade 3
Clinical Breast Cancer Month 2015
-5
Risk More Than 5 Years After HR+ BC Diagnosis Figure 2 Algorithm to Guide 5 Versus 10 Years of Hormone Therapy After Breast Cancer, According to 10% Residual Threshold Risk of Event-Free Survival Event in Years 5 to 10 After Diagnosis
Age at diagnosis
<10% risk of event
<50 years
Stage I Grades 1, 2, 3 Stage II Grade 1 Stage I Grade 1 Stage II Grade 1
50+ years
<10% risk of event, including 95% CI Stage I Grades 1, 2 Stage II Grade 1 Stage I Grade 1 Stage II Grade 1
>10% risk of event in years 510 Stage II Grades 2, 3 Stage III All grades Stage II Grades 2, 3 Stage III All grades
Hormone receptor positive breast cancer with no recurrence, death, or CLBC in first 5 years after diagnosis
Stage I
Stage II, node negative and node positive 5 years AHT : grade 1, 2 OR Figure 2. Algorithm to guide 5 versus 10 years of hormone therapy after breast cancer, based on 10% residual
5 or 10 years AHT: Grade 1 Or Low risk by RS or similar assay
5 or 10 years AHT: grade 3 and <50 years old at dx OR intermediate risk by RS or similar assay (any age)
10 years AHT: Grade 2 or 3 OR Intermediate or high RS OR high risk by similar assay
Stage III
10 years AHT: all grades
10 years AHT: high risk by RS or similar assay
Abbreviations: AHT ¼ adjuvant hormone therapy; CI ¼ confidence interval; CLBC ¼ contralateral breast cancer; RS ¼ 21-gene recurrence score assay.
6
-
histologic features.17 Other contemporary treatments such as modern chemotherapy and regional radiotherapy reduce relapse rates; therefore, the event rates observed in our cohorts might have been a percentage point or 2 greater than expected today. This effect has been seen in a number of recent clinical trials, in which the control arm event rates were lower than had been estimated when designing the trial, presumably owing to improvements in elements of care not being studied but systematically applied within the trial populations. We were unable to confirm AHT adherence in our population. Previous reports, including from our own institution, have suggested high nonadherence rates in the general population.18 Because ours was a population-based study, we assumed an average rate of nonadherence and would anticipate better outcomes in a perfectly compliant population. When discussing extended AHT with a woman known to have been compliant for the first 5 years, it is likely that her residual recurrence risk will not be more than observed for her age, stage, and grade cohort in the present data set. Hormone therapy recommendations changed during the study period; increasing for postmenopausal women from 2 to 3 years in 1992 and to 5 years in 1995. In 1993, 3 years of tamoxifen use was
Clinical Breast Cancer Month 2015
introduced for premenopausal women and was extended to 5 years in 1996. Thus, many of the younger women in the present study had received less than optimal exposure to tamoxifen by current standards. We observed an EFS rate of 87.7% and 87.9% among women aged < 50 years with stage II, grade 2 or 3 cancer, respectively. Given a 30% relative risk reduction with 5 years of tamoxifen, these cohorts might, with modern treatment standards, experience an event risk of close to 10% at 10 years if event free by year 5; however, this remains speculative.1 The use of AI therapy for 2 to 5 years was introduced for postmenopausal ERþ BC at BCCA in July 2005. Only women diagnosed in the most recent 2 to 3 years of the present study would thus have received an adjuvant AI. The relatively modest additional benefit of AIs compared with tamoxifen would be unlikely to improve the 10-year BC event rate to < 10% in the stage and grade cohorts aged > 50 years with a demonstrated risk > 10% in our study population.9 Also, 12% of younger women and 7% of older women with stage I cancer in our cohorts filled 1 prescription for an AI after 5 years of tamoxifen. This might result in 1% to 2% lower recurrence rates and would not meaningfully alter the excellent prognosis we observed for those with stage I disease.
Sheridan Wilson et al The changes in hormone therapy policies described would improve the EFS rates for all groups. Similarly, improvements in chemotherapy and local therapy over time might result in lower rates than we observed. Thus, we have postulated that the groups of women for whom we observed a < 10% event rate would still experience a low event rate in the context of contemporary therapy. We believe these data can accurately identify populations with a low residual risk after 5 years of AHT.
inform their decision to continue or discontinue AHT after 5 years.
Acknowledgments This work was partly funded by an unrestricted educational grant from Astra Zeneca.
Disclosure Conclusion Our study was designed to provide estimates of a BC event, including CLBC, > 5 years from diagnosis, to help contextualize the potential benefits of extended AHT. Tissue-based assays such as the 21-gene recurrence score assay have been shown to provide prognostic information about recurrence in years 5 to 10 and might be an additional mechanism to identify women who can safely stop after 5 years of AHT.19 This test is not available in all jurisdictions or for many women diagnosed several years earlier; however, pathologic variables, which are readily available for all cases, appear to provide good risk discrimination. The side effects experienced in the first 5 years and competing potential causes of mortality will also have relevance for an individual contemplating this choice. We believe that women of all ages who remain free of recurrence 5 years from the initial diagnosis of ERþ stage I cancer of any grade (with the possible exception of women aged < 50 years at diagnosis of a stage I, grade 3 cancer) or stage II, grade I cancer (with the possible exception of those aged < 50 years at diagnosis or with associated LVI) might reasonably avoid 10 years of hormone therapy because they have a 10-year combined risk of recurrence, CLBC, or BC death of 10%. Women with such cancer who wish to discontinue AHT after 5 years can be reassured that the additional protection of longer therapy is exceedingly small and possibly negligible for them. For other stage and grade cancer combinations, in which the BC event risk after 5 years is > 10%, a careful discussion of the potential benefits and harm of continued AHT is warranted.
Clinical Practice Points Recurrences are distributed across > 10 years after a diagnosis of
HRþ BC. Most of the recurrence risk reduction conferred by AHT results from the first 5 years of use, with more modest benefits in years 5 to 10 of continued use. Our population-based research suggests that women with stage II grade 1 cancer and stage I cancer, who have not experienced recurrence in the first 5 years can expect a < 10% future recurrence risk, making discontinuation of hormone therapy after 5 years a safe option for them. Many women struggle with hormone therapy-related side effects but also with fears of BC recurrence. The recurrence estimates for the second 5 years after diagnosis, stratified by the initial disease presentation, will better
The authors have stated that they have no conflicts of interest.
References 1. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: An overview of the randomised trials. Lancet 2005; 365: 1687-717. 2. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), Davies C, Godwin J, Gray R, et al. Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: patient-level meta-analysis of randomised trials. Lancet 2011; 378:771-84. 3. Fisher B, Dignam J, Bryant J, et al. Five versus more than five years of tamoxifen for lymph node-negative breast cancer: updated findings from the National Surgical Adjuvant Breast and Bowel Project B-14 randomized trial. J Natl Cancer Inst 2001; 93:684-90. 4. Singh L, Wilson AJ, Baum M, et al. The relationship between histological grade, oestrogen receptor status, events and survival at 8 years in the NATO (“nolvadex”) trial. Br J Cancer 1988; 57:612-4. 5. “Nolvadex” Adjuvant Trial Organisation. Controlled trial of tamoxifen as a single adjuvant agent in the management of early breast cancer. Br J Cancer 1988; 57:608-11. 6. Adjuvant tamoxifen in the management of operable breast cancer: the Scottish trial. Report from the Breast Cancer Trials Committee, Scottish Cancer Trials Office (MRC), Edinburgh. Lancet 1987; 2:171-5. 7. Stewart HJ, Forrest AP, Everington D, et al. Randomised comparison of 5 years of adjuvant tamoxifen with continuous therapy for operable breast cancer: the Scottish Cancer Trials Breast Group. Br J Cancer 1996; 74:297-9. 8. Fisher B, Costantino J, Redmond C, et al. A randomized clinical trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have estrogen-receptor-positive tumors. N Engl J Med 1989; 320:479-84. 9. Dowsett M, Cuzick J, Ingle J, et al. Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol 2010; 28:509-18. 10. Delozier T, Spielmann M, Mace-Lesec’h J, et al. Tamoxifen adjuvant treatment duration in early breast cancer: initial results of a randomized study comparing short-term treatment with long-term treatment. Federation Nationale des Centres de Lutte Contre le Cancer Breast Group. J Clin Oncol 2000; 18:3507-12. 11. Goss PE, Ingle JN, Martino S, et al. Efficacy of letrozole extended adjuvant therapy according to estrogen receptor and progesterone receptor status of the primary tumor: National Cancer Institute of Canada clinical trials group MA.17. J Clin Oncol 2007; 25:2006-11. 12. Davies C, Pan H, Godwin J, et al. Long-term effects of continuing adjuvant tamoxifen to 10 years versus stopping at 5 years after diagnosis of oestrogen receptorpositive breast cancer: ATLAS, a randomised trial. Lancet 2013; 381:805-16. 13. Gray R, Rea D, Handley K, et al. aTTom: long-term effects of continuing adjuvant tamoxifen to 10 years versus stopping at 5 years in 6,953 women with early breast cancer. J Clin Oncol 2013; 31(suppl), abstract 5. 14. BC Cancer Agency. BC Cancer Agency, Health Professionals Information. Available at: http://www.bccancer.bc.ca/HPI/default.htm. Accessed 2014. 15. Kennecke HF, Olivotto IA, Speers C, et al. Late risk of relapse and mortality among postmenopausal women with estrogen responsive early breast cancer after 5 years of tamoxifen. Ann Oncol 2007; 18:45-51. 16. Brewster AM, Hortobagyi GN, Broglio KR, et al. Residual risk of breast cancer recurrence 5 years after adjuvant therapy. J Natl Cancer Inst 2008; 100:1179-83. 17. Hu Z, Fan C, Oh DS, et al. The molecular portraits of breast tumors are conserved across microarray platforms. BMC Genomics 2006; 7:96. 18. Hershman DL, Kushi LH, Shao T, et al. Early discontinuation and nonadherence to adjuvant hormonal therapy in a cohort of 8,769 early-stage breast cancer patients. J Clin Oncol 2010; 28:4120-8. 19. Dowsett M, Cuzick J, Wale C, et al. Prediction of risk of distant recurrence using the 21-gene Recurrence Score in node-negative and node-positive postmenopausal patients with breast cancer treated with anastrozole or tamoxifen: a TransATAC study. J Clin Oncol 2010; 28:1829-34.
Clinical Breast Cancer Month 2015
-7