- Email: [email protected]
Aromatase Inhibitor Versus Tamoxifen in Postmenopausal Woman With Advanced Breast Cancer: A Literature-Based Meta-Analysis
Original Study
Aromatase Inhibitor Versus Tamoxifen in Postmenopausal Woman With Advanced Breast Cancer: A Literature-Based Meta-Analysis Hong-Bin Xu,1 Yu-Jin Liu,2 Ling Li1 Abstract Clinical trials have reported conflicting results as to whether Aromatase inhibitors (AIs) as first-line hormonal therapy improve outcome over tamoxifen in postmenopausal women with advanced breast cancer. We performed a metaanalysis comparing primary and secondary endpoints of AIs to tamoxifen as first-line hormonal therapy in postmenopausal women with advanced breast cancer. The event-based odds ratio (OR) with 95% confidence interval (95% CI) were derived, and a test of heterogeneity was applied. Six eligible trials (2560 patients) were selected from 488 studies that initially were identified. A significant difference in favoring AIs over tamoxifen was observed in overall response rate (ORR; OR, 1.56; 95% CI, 1.17-2.07; P ⫽ .002) and clinical benefit (CB; OR, 1.70; 95% CI, 1.24-2.33; P ⫽ .0009).Whereas the trend toward an improved overall survival (OS) rate was not significant (OR, 1.95; 95% CI, 0.88-4.30; P ⫽ .10).Toxicities did not differ significantly except vaginal bleeding (OR, 0.30; 95% CI, 0.16-0.56; P ⫽ .0002) and thromboembolic event (OR, 0.47; 95% CI, 0.28-0.77; P ⫽ .003). AIs appeared to be effective and feasible compared with tamoxifen as first-line hormonal therapy in postmenopausal women with advanced breast cancer. Further prospective, randomized, controlled trials will be necessary. Clinical Breast Cancer, Vol. 11, No. 4, 246-51 © 2011 Published by Elsevier Inc. Keywords: Advanced breast cancer, Aromatase inhibitor, Hormonal therapy, Meta-analysis, Postmenopausal women, Tamoxifen
Introduction Breast cancer is the most common cause of cancer death in elderly women. With the aging of the general population, the association between cancer and aging has become more evident and paramount as a pandemic public health concern. As such, formidable increases in the incidence and prevalence of breast cancer can be expected in the coming decades if the older population continues to expand at the present rate.1 Eighty percent of breast tumors occurring in elderly women are rich in hormone receptors, whereas the remaining 20% of women have aggressive tumors that have few hormone receptors.2,3 Knowledge of the steroid receptor content of human breast cancer is important for deciding the proper treatment for advanced 1 Department of Clinical Pharmacy, Shanghai Tenth People’s Hospital, Tongji University School of Medicine, Shanghai, China 2 Department of Interventional Therapy, Shanghai Tenth People’s Hospital, Tongji University School of Medicine, Shanghai, China
Submitted: Oct 17, 2010; Accepted Dec 21, 2010 Address for correspondence: Ling Li, Shanghai Tenth People’s Hospital, Department of Clinical Pharmacy, No. 301, YanChang Middle Road, Shanghai, China Fax: ⫹86 2166307668; e-mail contact: [email protected]
246
Clinical Breast Cancer August 2011
breast cancer. Endocrine therapy is the established treatment in women with hormone-sensitive tumors, as manifested by positive receptor status, a long disease-free interval, and primarily soft tissue disease. Tamoxifen, a competitive estrogen-receptor (ER) antagonist, was until recently the standard first-line hormonal therapy for hormone receptor-positive breast cancer in postmenopausal women.4 The development of resistance to tamoxifen, and long-term toxicities including thromboembolic events and endometrial cancer, has led to increasing use of alternative hormonal therapies including third-generation aromatase inhibitors (AIs; letrozole, exemestane, and anastrozole). These drugs are very selective and potent in the suppression of aromatase activity.5 AIs decrease the risk of thromboembolic and cerebrovascular events compared with tamoxifen, and are also associated with a lower incidence of endometrial cancer and fewer vaginal bleeding events than tamoxifen.6,7 Generally, adverse events with AIs are predictable and manageable, whereas tamoxifen may be associated with life-threatening events in a minority of patients.8 Several studies have compared third generation AIs to tamoxifen as first-line hormonal therapy in postmenopausal woman with advanced breast cancer.9-14 However, individually, these trials found
1526-8209/$ - see frontmatter © 2011 Published by Elsevier Inc. doi: 10.1016/j.clbc.2011.06.003
that overall response rate (ORR), clinical benefit, overall survival (OS), and toxicities were statistically inconsistent. The purpose of this study was to perform a meta-analysis to examine whether third generation AIs are effective compared with tamoxifen as first-line hormonal therapy for postmenopausal women with advanced breast cancer.
Methods The meta-analysis was performed according to a prospectively written protocol and analysis plan.
Objective of the Current Study The current literature-based meta-analysis was performed to evaluate the efficacy (ORR, clinical benefit [CB], and OS) and the toxicity profile of AIs compared with tamoxifen as first-line hormonal therapy in postmenopausal women with advanced breast cancer.
Definition of Outcome Efficacy was assessed using ORR and CB as the primary outcome and OS as the secondary outcome. The ORR is the percentage of patients who have a complete or partial tumor response, and the CB defined as the proportion of patients in each arm with a complete response, a partial response, or a stable disease. The OS is defined as the time between the date of random assignment and the date of death, or the last date of follow-up for censored patients according to World Health Organization criteria. Regarding toxicity, we considered the incidence of adverse events (including hot flashes, nausea, bone pain, vaginal bleeding, and thromboembolic event) using the National Cancer Institute common toxicity criteria.
Selection of Trials Only published, randomized trials were selected for evaluation when they investigated the first-line chemotherapy of histologically confirmed advanced breast cancer. As a consequence, all studies performed in the adjuvant or neoadjuvant setting were excluded.
Search Strategy A literature search was carried out in December 2009 to identify all published randomized trials, comparing AIs to tamoxifen as firstline hormonal therapy in postmenopausal women with advanced breast cancer. The search was performed using MEDLINE (from 1966 to December 2009), ClinicalTrials.gov, and the Cochrane Central Register of Controlled Trials, with the following search terms in all the possible combinations, “endocrine therapy,” “hormonal therapy,” “aromatase inhibitor,” “tamoxifen,” “postmenopausal,” and “breast cancer.” The search was limited to trials that were randomized, controlled, and published in the English language. When two or more articles reported the same data, the most recently updated data were included. References of the identified articles were also checked and principal investigators were asked if they were aware of other trials.
Data Collection Data abstraction was performed by two independent observers who extracted the data from the respective trials and verified the results by comparison. The following data were collected from the identified trials: first author’s name, journal, year of publication, randomization number, treatment schedule, number of patients,
median age of patients, performance status, menopausal and hormone receptor status, and median time to progression, median OS, objective response rate, clinical benefit rate, OS rate, and the incidence of toxicities (including hot flashes, nausea, pain bone, vaginal bleeding, and thromboembolic event) were selected in any treatment group.
Statistical Methods The AIs group was considered an investigational treatment, and the tamoxifen group was used as control treatment. The outcomes were represented by dichotomous variables: the ORR, CB, and toxicity analysis was performed by considering the number of patients evaluable for response and toxicity; OS was calculated by applying an intent-to-treat analysis, respectively. Heterogeneity was assessed by the Q-statistic (using a 10% significant level), and interpretation of the I2 statistic (percentage of variation as a result of heterogeneity, with higher values indicating a greater degree of heterogeneity). A fixed effects model was adopted unless there was evidence of significant unexplained heterogeneity, in which case a random effects model was used. The odds ratio (OR) was estimated for each considered endpoint, and 95% confidence interval (CI) were derived.15,16 Publication bias is a common concern in meta-analysis that is related to the tendency of journals to favor the publication of large and positive studies. We chose a commonly used method for detecting publication bias, which is a graphical plot of estimates of the log OR from the individual studies versus the SE of log OR.
Results The two assessors agreed on the selection of six RCTs from 488 studies that initially were identified.9-14 The main characteristics of the six randomized trials are listed in Table 1. Six randomized trials including 2657 patients were found. The main characteristics of the six randomized trials are listed in Table 1. One was randomized phase II trials,11 and five were randomized phase III trials.9,10,12-14 Two randomized trials compared tamoxifen to exemestane,9,11 one randomized trial compared tamoxifen to letrozole,10 and three randomized trials compared tamoxifen to anastrozole.12-14
Overall Response Rate All the trials reported this outcome representing a total of 2650 patients.9-14 The ORR was 33.0% (436/1323) and 24.9% (330/ 1327) for AIs arm and tamoxifen arm. The Q-statistics for heterogeneity was statistically significant (P ⫽ .04) and the I2-statistic indicated heterogeneity (58.1%). The overall OR was 1.56 (95% CI, 1.17-2.07; P ⫽ .002) by random effect model which suggested that there was significant difference between the AIs arm and tamoxifen arm (Figure 1). Three trials reported ER and/or progesterone receptor (PgR)-positive ORR representing a total of 1031 patients.9-11 The outcome was 58.4% (295/505) and 57.2% (301/526) for AIs arm and tamoxifen arm. The Q-statistics for heterogeneity was statistically significant (P ⬍ .0001) and the I2-statistic indicated heterogeneity (89.4%). The ORR was 1.41 (95% CI, 0.57-3.51; P ⫽ .46) by random effect model which suggested that there was no difference between the AIs arm and the tamoxifen arm (Figure 2).
Clinical Breast Cancer August 2011
247
AIs Versus Tamoxifen for Advanced Breast Cancer Table 1 Characteristics of Trials Characteristic
Paridaens 2008 Mouridsen 2004 Paridaens 2003 Alfredo 2003 Nabholtz 2000 Bonneterre 2000
Chemotherapy Schedule Arm A
Exemestane
Letrozole
Exemestane
Anastrozole
Anastrozole
Anastrozole
Arm B
Tamoxifen
Tamoxifen
Tamoxifen
Tamoxifen
Tamoxifen
Tamoxifen
Sample Size
371
907
120
238
353
668
37-87
31-96
37-87
55-77
30-92
34-92
ECOG ⱕ 2
KPS ⱖ 50%
ECOG ⱕ 2
ECOG ⱕ 2
NR
NR
ER and/or PgR-positive
93.3
66.1
92.5
NR
85.3
43.1
Both negative
0.3
0.3
0
NR
3.4
1.6
Age Performance Status ER/PgR Status (%)
Both unknown
6.4
33.6
7.5
NR
0.3
55.1
Allowed
Allowed
Allowed
Not allowed
Allowed
Allowed
Arm A
9.9 months
10.3 months
NR
18.0 months
11.1 months
8.2 months
Arm B
5.8 months
6.5 months
NR
7.0 months
5.6 months
8.3 months
Arm A
45.6
30.2
44.6
35.5
21.1
32.9
Arm B
31.2
20.3
17.5
26.5
17.0
32.6
Arm A
75.3
50.1
62.5
82.6
59.1
56.2
Arm B
66.6
38.1
43.9
55.6
45.6
55.5
Arm A
47.4
67.7
40.4
NR
NR
NR
Arm B
33.5
77.0
18.5
NR
NR
NR
Arm A
54.9
81.2
NR
39.7
NR
NR
Arm B
57.1
70.9
NR
11.1
NR
NR
Previous hormone treatment Efficacy Median time to progression
Overall response rates (%)
Clinical benefit (%)
ER and/or PgR-positive overall response rates (%)
Overall survival (%)
Abbreviations: ER ⫽ estrogen receptor; PgR ⫽ progesterone receptor; NR ⫽ not reported; EGCO ⫽ Eastern Cooperative Oncology Group; KPS ⫽ Karnofsky Performance Status.
Clinical Benefit All the trials reported CB representing a total of 2650 patients.9-14 The CB was 59.8% (791/1323) and 49.3% (654/ 1327) for the AIs arm and the tamoxifen arm. The Q-statistics showed the presence of heterogeneity (P ⫽ .005) and the I2statistic detected heterogeneity (70.2%). The pooled results by random effect model demonstrated a statistical difference in the clinical benefit rate between patients who had AIs treatment and those who had tamoxifen treatment (OR, 1.70; 95% CI, 1.242.33; P ⫽ .0009; Figure 3).
Overall Survival Three studies reported OS as one of the outcomes representing a total of 1516 patients.9,10,12 The OS was increased for the AIs arm (516/756 ⫽ 68.3%) compared with the tamoxifen arm (443/760 ⫽ 58.3%). The pooled OR by random effect model was 1.95 with a 95% CI ranged from 0.88 to 4.30. The Q-statistics for heterogeneity was statistically significant with P ⬍ .0001 and the I2-statistic detected heterogeneity (89.6%). The
248
Clinical Breast Cancer August 2011
ORR suggests that there is no difference between the AIs arm and the tamoxifen arm in terms of OS with P value .10 (Figure 4).
Toxicities Among the adverse events, we compared hot flashes, nausea, bone pain, vaginal bleeding, and thromboembolic event. All the trials reported the incidence rate of hot flashes,9-14 five trials reported nausea and bone pain,9-11,13,14 four trials reported vaginal bleeding,9,12-14 and three trials reported thromboembolic event.12-14 The Q-statistics of hot flashes (P ⫽ .006) and nausea (P ⫽ .009) showed the presence of heterogeneity; bone pain (P ⫽ .92), vaginal bleeding (P ⫽ .82), and thromboembolic events (P ⫽ .19) demonstrated the homogeneity. We obtained pooled results by random effect model or fixed effect model. The pooled results were as follows: hot flashes (OR, 0.99; 95% CI, 0.701.40; P ⫽ .95), nausea (OR, 1.06; 95% CI, 0.72-1.55; P ⫽ .77), bone pain (OR, 0.99; 95% CI, 0.81-1.22; P ⫽ .94), vaginal bleeding (OR, 0.30; 95% CI, 0.16-0.56; P ⫽ .0002), and thromboembolic event (OR, 0.47; 95% CI, 0.28-0.77; P ⫽ .003; Table 2). Hot flashes, nausea, and
Hong-Bin Xu et al Figure 1 Overall Response Rate of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category Bonneterre 2000 Nabholtz 2000 Alfredo 2003 Paridaens 2003 Mouridsen 2004 Paridaens 2008
aromarase inhibitor n/N
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
112/340
107/328
22.00
1.01 [0.73, 1.40]
36/171 43/121 25/56 137/453
31/182 31/117 10/57 92/454
14.78 14.19 8.04 22.73
1.30 [0.76, 2.21] 1.53 [0.88, 2.66] 3.79 [1.60, 8.98] 1.71 [1.26, 2.31]
83/182
59/189
18.26
1.85 [1.21, 2.82]
100.00
1.56 [1.17, 2.07]
1323 1327 Total (95% CI) Total events: 436 (aromarase inhibitor), 330 (tamoxifen) Test for heterogeneity: χ? = 11.94, df = 5 (P = 0.04), I? = 58.1% Test for overall effect: Z = 3.05 (P = 0.002) 0.1 0.2
0.5 1
2
5 10
Figure 2 Overall Response Rate by ER and/or PgR Positive of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category Paridaens 2003 Mouridsen 2004 Paridaens 2008 Total (95% CI)
aromarase inhibitor n/N
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
23/57 199/294
10/54 235/305
28.54 36.29
2.98 [1.25, 7.08] 0.62 [0.43, 0.90]
73/154
56/167
35.18
1.79 [1.14, 2.80]
505
526
100.00
1.41 [0.57, 3.51]
Total events: 295 (aromarase inhibitor), 301 (tamoxifen) Test for heterogeneity: χ? = 18.83, df = 2 (P = 0.0001), I? = 89.4% Test for overall effect: Z = 0.74 (P = 0.46) 0.01
bone pain were not different between the AIs and tamoxifen groups. However, vaginal bleeding and thromboembolic events were significantly more frequent in the tamoxifen group (P ⫽ .0002).
Publication Bias The evaluation of publication bias showed that the Egger test was not significant (P ⫽ .292). The funnel plots of the log OR versus the SE of log OR for publication bias formed a distinct funnel shape regardless of the study variance, and also did not show some asymmetry (figure not shown). These results indicate there was not a potential for publication bias.
Discussion Breast cancer is the most common form of cancer affecting women.1,17 Advanced breast cancer in postmenopausal women remains an incurable disease, and so treatment is aimed at palliation and improved quality of life, inhibition of disease progression, and
0.1
1
10
100
improvement in survival time where possible. Endocrine therapy is now a widely accepted treatment modality and is primarily directed at reducing the synthesis of estrogen or alternatively blocking estrogen receptors in breast cancer that are hormone-sensitive. Tamoxifen has been the accepted gold-standard first-line treatment for advanced breast cancer in postmenopausal women since the late 1970s. Tamoxifen acts by blocking the binding of estrogen to its receptor and has an ORR of 30%-35% when used as first-line therapy for advanced breast cancer.18 Adverse effects that have been associated with tamoxifen can be classified as either owing to its antiestrogenic actions (eg, hot flashes, vaginal bleeding, discharge, or dryness) or as more general effects (eg, nausea, vomiting, tumor flare, and skin rash). The serious and long-term adverse effects of tamoxifen in the treatment of advanced breast cancer have led to increasing use of alternative hormonal therapies such as the new generation AIs.
Clinical Breast Cancer August 2011
249
AIs Versus Tamoxifen for Advanced Breast Cancer Figure 3 Clinical Benefit of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category
aromarase inhibitor n/N
Bonneterre 2000 Nabholtz 2000 Alfredo 2003 Paridaens 2003 Mouridsen 2004 Paridaens 2008
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
191/340
182/328
20.56
1.03 [0.76, 1.40]
101/171 100/121 35/56 227/453
83/182 65/117 25/57 173/454
17.47 13.31 10.35 21.62
1.72 [1.13, 2.62] 3.81 [2.10, 6.91] 2.13 [1.01, 4.53] 1.63 [1.25, 2.12]
137/182
126/189
16.68
1.52 [0.97, 2.39]
100.00
1.70 [1.24, 2.33]
1323 1327 Total (95% CI) Total events: 791 (aromarase inhibitor), 654 (tamoxifen) Test for heterogeneity: χ? = 16.78, df = 5 (P = 0.005), I? = 70.2% Test for overall effect: Z = 3.33 (P = 0.0009) 0.1 0.2
0.5 1
2
5 10
Figure 4 Overall Survival of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category Alferdo 2003 Mouridsen 2004 Paridaens 2008
aromarase inhibitor n/N
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
29.64 35.89
5.26 [2.66, 10.40] 1.77 [1.30, 2.42]
48/121 368/453
13/117 322/454
100/182
108/189
34.47
0.91 [0.61, 1.38]
756
760
100.00
1.95 [0.88, 4.30]
Total (95% CI)
Total events: 516 (aromarase inhibitor), 443 (tamoxifen) Test for heterogeneity: χ? = 19.28, df = 2 (P < 0.0001), I? = 89.6% Test for overall effect: Z = 1.65 (P = 0.10) 0.01
0.1
1
10
100
Table 2 Overall Analysis - Toxicities Outcome
RCTs
Patients
OR
95% CI
P Value
Heterogeneity (P)
Hot Flashes
6
2653
0.99
0.70-1.40
.95
0.006
Nausea
5
2415
1.06
0.72-1.55
.77
0.009
Bone Pain
5
2415
0.99
0.81-1.22
.94
0.92
Vaginal Bleeding
4
1626
0.30
0.16-0.56
.0002
0.82
Thromboembolic Event
3
1255
0.47
0.28-0.77
.003
0.19
Abbreviations: RCTs ⫽ randomized controlled trials; OR ⫽ odds ratio; CI ⫽ confidence interval.
AIs inhibit the conversion of androgens into estrogens, thereby significantly reducing circulating estrogens in postmenopausal women. Their efficacy in breast cancer is well established.19 Our meta-analysis indicated that AIs were more effective than tamoxifen in terms of ORR and CB as first-line hormonal therapy in post-
250
Clinical Breast Cancer August 2011
menopausal women with advanced breast cancer. Furthermore, there was no difference in clinically relevant toxicities between the two treatments except vaginal bleeding and thromboembolic events. The six trials we included used AIs including the steroidal compound exemestane as well as nonsteroidal compounds such as anas-
Hong-Bin Xu et al trozole and letrozole.20-22 Two trials used exemestane,9,11 one trial used letrozole,10 and three trials used anastrozole.12-14 The outcomes (ORR, CB, and OS rate) and the incidence rate of toxicities (hot flashes, nausea, bone pain, vaginal bleeding, and thromboembolic events) were inconsistent in these trials, leading to controversies about treatment preference. We attempted to identify which agent is more effective and less toxic as first-line hormonal therapy in postmenopausal women with advanced breast cancer. Although there was no difference between therapies in the incidence of hot flashes, nausea, or bone pain, the incidence of vaginal bleeding or thromboembolic events was statistically significant. Therefore, we could carefully suggest that AIs are superior to tamoxifen in terms of toxicities. One limitation in our meta-analysis is the data source extracted from published studies not based on individual patient data. Metaanalysis based on published data tends to overestimate treatment effects compared with individual data analysis. However, analysis using individual data may include fewer studies if authors do not agree to submit their full database to the analyzing group. In general, an individual database meta-analysis would give a more robust estimation for the association; therefore, we should interpret the results with care, especially for a positive result. Although the risk of publication bias exists in any meta-analysis, whether based on individual patient data or not, we feel this was not an important aspect of our study. An important unexplored issue in our analysis to consider is the treatment cost, to be balanced with the benefit of AIs. Unfortunately, no planned analysis across all randomized trials evaluating this modern topic is available.
Conclusion Although considering all drawbacks and controversies related to literature-based meta-analysis,23 AIs have better ORRs and clinical benefit rates than tamoxifen and in terms of toxicities, is not inferior to tamoxifen; therefore, we could suggest AIs instead of tamoxifen for those postmenopausal patients with hormone receptor-positive advanced breast cancer.
Acknowledgments The authors thank Cochrane Collaboration for providing the software Review Manager used in this review. We also thank all the patients who had been treated in these trials.
Disclosure There was no funding for this article. Conflict of interest: We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service, and/or
company that could be construed as influencing the position presented in, or the review of, the manuscript entitled, “Aromatase inhibitor versus tamoxifen in postmenopausal woman with advanced breast cancer: a literature-based meta-analysis”.
References 1. Yancik R, Ries LA. Aging and cancer in America: demographic and epidemiologic perspectives. Hematol Oncol Clin North Am 2000; 14:17-23. 2. Alghanem AA, Hussain S. The effect of age on estrogen and progesterone receptor in primary breast cancer. J Surg Oncol 1985; 30:29-32. 3. Elwood JM, Godolphin W. Oestrogen receptors in breast tumours: associations with age, menopausal status and epidemiological and clinical features in 735 patients. Br J Cancer 1980; 42:635-44. 4. Jaiyesimi IA, Buzdar AU, Decker DA, et al. Use of tamoxifen for breast cancer: twenty-eight years later. J Clin Oncol 1995; 13:513-29. 5. Ragaz J. Status of aromatase inhibitors in relation to other breast cancer treatment modalities. Endocr Relat Cancer 1999; 6:277-91. 6. Goss PE, Ingle JN, Martino S, et al. A randomized trial of letrozole in postmenopausal women after five years of tamoxifen therapy for early-stage breast cancer. N Engl J Med 2003; 349:1793-802. 7. Gerber B, Krause A, Reimer T, et al. Anastrozole versus tamoxifen treatment in postmenopausal women with endocrine- responsive breast cancer and tamoxifeninduced endometrial pathology. Clin Cancer Res 2006; 12:1245-50. 8. Perez EA. Safety of aromatase inhibitors in the adjuvant setting. Breast Cancer Res Treat 2007; 105 suppl 1:75-89. 9. Paridaens RJ, Dirix LY, Louk V, et al. Phase III study comparing exemestane with tamoxifen as first-line hormonal treatment of metastatic breast cancer in postmenopausal women: The European Organisation for Research and Treatment of Cancer Breast Cancer Cooperative Group. J Clin Oncol 2008; 26:4883-90. 10. Mouridsen H, Chaudri-Ross HA. Efficacy of first-line letrozole versus tamoxifen as a function of age in postmenopausal women with advanced breast cancer. Oncologist 2004; 9:497-506. 11. Paridaens RJ, Dirix LY, Lohrisch C, et al. Mature results of a randomized phase II multicenter study of exemestane versus tamoxifen as first-line hormone therapy for postmenopausal women with metastatic breast cancer. Ann Oncol 2003; 14:1391-8. 12. Milla-Santos A, Milla L, Portella J, et al. Anastrozole versus tamoxifen as first-line therapy in postmenopausal patients with hormone-dependent advanced breast cancer: a prospective, randomized, phase III study. Am J Clin Oncol 2003; 26:317-22. 13. Bonneterre J, Thurlimann B, Robertson JF. Anastrozole versus tamoxifen as firstline therapy for advanced breast cancer in 668 postmenopausal women: results of the Tamoxifen or Arimidex Randomized Group Efficacy and Tolerability study. J Clin Oncol 2000; 18:3748-57. 14. Nabholtz JM, Buzdar A, Pollak M, et al. Anastrozole is superior to tamoxifen as first-line therapy for advanced breast cancer in postmenopausal women: results of a North American multicenter randomized trial. Arimidex Study Group. J Clin Oncol 2000; 18:3758-67. 15. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177-88. 16. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in metaanalyses. BMJ 2003; 327:557-60. 17. Kimmick GG, Balducci L. Breast cancer and aging: clinical interactions. Hematol Oncol Clin North Am 2000; 14:213-34. 18. Rose C, Mouridsen HT. Endocrine therapy of advanced breast cancer. Acta Oncol 1988; 27:721-8. 19. Dixon JM. Exemestane and aromatase inhibitors in the management of advanced breast cancer. Expert Opin Pharmacother 2004; 5:307-16. 20. Paridaens R, Thomas J, Wildiers J, et al. Safety, activity and estrogen inhibition by exemestane in postmenopausal women with advanced breast cancer: A phase I study. Anticancer Drugs 1998; 9:675-83. 21. Jones S, Vogel C, Arkhipov A, et al. Multicenter, phase II trial of exemestane as third-line hormonal therapy of postmenopausal women with metastatic breast cancer: Aromasin Study Group. J Clin Oncol 1999; 17:3418-25. 22. Hamilton A, Piccart M. The third-generation non-steroidal aromatase inhibitors: a review of their clinical benefits in the second-line hormonal treatment of advanced breast cancer. Ann Oncol 1999; 10:377-84. 23. Predbois P, Buyse M. Meta-analysis based on abstracted data: a step in the right direction, but only a first step. J Clin Oncol 2004; 22:3839-41.
Clinical Breast Cancer August 2011
251
Aromatase Inhibitor Versus Tamoxifen in Postmenopausal Woman With Advanced Breast Cancer: A Literature-Based Meta-Analysis Hong-Bin Xu,1 Yu-Jin Liu,2 Ling Li1 Abstract Clinical trials have reported conflicting results as to whether Aromatase inhibitors (AIs) as first-line hormonal therapy improve outcome over tamoxifen in postmenopausal women with advanced breast cancer. We performed a metaanalysis comparing primary and secondary endpoints of AIs to tamoxifen as first-line hormonal therapy in postmenopausal women with advanced breast cancer. The event-based odds ratio (OR) with 95% confidence interval (95% CI) were derived, and a test of heterogeneity was applied. Six eligible trials (2560 patients) were selected from 488 studies that initially were identified. A significant difference in favoring AIs over tamoxifen was observed in overall response rate (ORR; OR, 1.56; 95% CI, 1.17-2.07; P ⫽ .002) and clinical benefit (CB; OR, 1.70; 95% CI, 1.24-2.33; P ⫽ .0009).Whereas the trend toward an improved overall survival (OS) rate was not significant (OR, 1.95; 95% CI, 0.88-4.30; P ⫽ .10).Toxicities did not differ significantly except vaginal bleeding (OR, 0.30; 95% CI, 0.16-0.56; P ⫽ .0002) and thromboembolic event (OR, 0.47; 95% CI, 0.28-0.77; P ⫽ .003). AIs appeared to be effective and feasible compared with tamoxifen as first-line hormonal therapy in postmenopausal women with advanced breast cancer. Further prospective, randomized, controlled trials will be necessary. Clinical Breast Cancer, Vol. 11, No. 4, 246-51 © 2011 Published by Elsevier Inc. Keywords: Advanced breast cancer, Aromatase inhibitor, Hormonal therapy, Meta-analysis, Postmenopausal women, Tamoxifen
Introduction Breast cancer is the most common cause of cancer death in elderly women. With the aging of the general population, the association between cancer and aging has become more evident and paramount as a pandemic public health concern. As such, formidable increases in the incidence and prevalence of breast cancer can be expected in the coming decades if the older population continues to expand at the present rate.1 Eighty percent of breast tumors occurring in elderly women are rich in hormone receptors, whereas the remaining 20% of women have aggressive tumors that have few hormone receptors.2,3 Knowledge of the steroid receptor content of human breast cancer is important for deciding the proper treatment for advanced 1 Department of Clinical Pharmacy, Shanghai Tenth People’s Hospital, Tongji University School of Medicine, Shanghai, China 2 Department of Interventional Therapy, Shanghai Tenth People’s Hospital, Tongji University School of Medicine, Shanghai, China
Submitted: Oct 17, 2010; Accepted Dec 21, 2010 Address for correspondence: Ling Li, Shanghai Tenth People’s Hospital, Department of Clinical Pharmacy, No. 301, YanChang Middle Road, Shanghai, China Fax: ⫹86 2166307668; e-mail contact: [email protected]
246
Clinical Breast Cancer August 2011
breast cancer. Endocrine therapy is the established treatment in women with hormone-sensitive tumors, as manifested by positive receptor status, a long disease-free interval, and primarily soft tissue disease. Tamoxifen, a competitive estrogen-receptor (ER) antagonist, was until recently the standard first-line hormonal therapy for hormone receptor-positive breast cancer in postmenopausal women.4 The development of resistance to tamoxifen, and long-term toxicities including thromboembolic events and endometrial cancer, has led to increasing use of alternative hormonal therapies including third-generation aromatase inhibitors (AIs; letrozole, exemestane, and anastrozole). These drugs are very selective and potent in the suppression of aromatase activity.5 AIs decrease the risk of thromboembolic and cerebrovascular events compared with tamoxifen, and are also associated with a lower incidence of endometrial cancer and fewer vaginal bleeding events than tamoxifen.6,7 Generally, adverse events with AIs are predictable and manageable, whereas tamoxifen may be associated with life-threatening events in a minority of patients.8 Several studies have compared third generation AIs to tamoxifen as first-line hormonal therapy in postmenopausal woman with advanced breast cancer.9-14 However, individually, these trials found
1526-8209/$ - see frontmatter © 2011 Published by Elsevier Inc. doi: 10.1016/j.clbc.2011.06.003
that overall response rate (ORR), clinical benefit, overall survival (OS), and toxicities were statistically inconsistent. The purpose of this study was to perform a meta-analysis to examine whether third generation AIs are effective compared with tamoxifen as first-line hormonal therapy for postmenopausal women with advanced breast cancer.
Methods The meta-analysis was performed according to a prospectively written protocol and analysis plan.
Objective of the Current Study The current literature-based meta-analysis was performed to evaluate the efficacy (ORR, clinical benefit [CB], and OS) and the toxicity profile of AIs compared with tamoxifen as first-line hormonal therapy in postmenopausal women with advanced breast cancer.
Definition of Outcome Efficacy was assessed using ORR and CB as the primary outcome and OS as the secondary outcome. The ORR is the percentage of patients who have a complete or partial tumor response, and the CB defined as the proportion of patients in each arm with a complete response, a partial response, or a stable disease. The OS is defined as the time between the date of random assignment and the date of death, or the last date of follow-up for censored patients according to World Health Organization criteria. Regarding toxicity, we considered the incidence of adverse events (including hot flashes, nausea, bone pain, vaginal bleeding, and thromboembolic event) using the National Cancer Institute common toxicity criteria.
Selection of Trials Only published, randomized trials were selected for evaluation when they investigated the first-line chemotherapy of histologically confirmed advanced breast cancer. As a consequence, all studies performed in the adjuvant or neoadjuvant setting were excluded.
Search Strategy A literature search was carried out in December 2009 to identify all published randomized trials, comparing AIs to tamoxifen as firstline hormonal therapy in postmenopausal women with advanced breast cancer. The search was performed using MEDLINE (from 1966 to December 2009), ClinicalTrials.gov, and the Cochrane Central Register of Controlled Trials, with the following search terms in all the possible combinations, “endocrine therapy,” “hormonal therapy,” “aromatase inhibitor,” “tamoxifen,” “postmenopausal,” and “breast cancer.” The search was limited to trials that were randomized, controlled, and published in the English language. When two or more articles reported the same data, the most recently updated data were included. References of the identified articles were also checked and principal investigators were asked if they were aware of other trials.
Data Collection Data abstraction was performed by two independent observers who extracted the data from the respective trials and verified the results by comparison. The following data were collected from the identified trials: first author’s name, journal, year of publication, randomization number, treatment schedule, number of patients,
median age of patients, performance status, menopausal and hormone receptor status, and median time to progression, median OS, objective response rate, clinical benefit rate, OS rate, and the incidence of toxicities (including hot flashes, nausea, pain bone, vaginal bleeding, and thromboembolic event) were selected in any treatment group.
Statistical Methods The AIs group was considered an investigational treatment, and the tamoxifen group was used as control treatment. The outcomes were represented by dichotomous variables: the ORR, CB, and toxicity analysis was performed by considering the number of patients evaluable for response and toxicity; OS was calculated by applying an intent-to-treat analysis, respectively. Heterogeneity was assessed by the Q-statistic (using a 10% significant level), and interpretation of the I2 statistic (percentage of variation as a result of heterogeneity, with higher values indicating a greater degree of heterogeneity). A fixed effects model was adopted unless there was evidence of significant unexplained heterogeneity, in which case a random effects model was used. The odds ratio (OR) was estimated for each considered endpoint, and 95% confidence interval (CI) were derived.15,16 Publication bias is a common concern in meta-analysis that is related to the tendency of journals to favor the publication of large and positive studies. We chose a commonly used method for detecting publication bias, which is a graphical plot of estimates of the log OR from the individual studies versus the SE of log OR.
Results The two assessors agreed on the selection of six RCTs from 488 studies that initially were identified.9-14 The main characteristics of the six randomized trials are listed in Table 1. Six randomized trials including 2657 patients were found. The main characteristics of the six randomized trials are listed in Table 1. One was randomized phase II trials,11 and five were randomized phase III trials.9,10,12-14 Two randomized trials compared tamoxifen to exemestane,9,11 one randomized trial compared tamoxifen to letrozole,10 and three randomized trials compared tamoxifen to anastrozole.12-14
Overall Response Rate All the trials reported this outcome representing a total of 2650 patients.9-14 The ORR was 33.0% (436/1323) and 24.9% (330/ 1327) for AIs arm and tamoxifen arm. The Q-statistics for heterogeneity was statistically significant (P ⫽ .04) and the I2-statistic indicated heterogeneity (58.1%). The overall OR was 1.56 (95% CI, 1.17-2.07; P ⫽ .002) by random effect model which suggested that there was significant difference between the AIs arm and tamoxifen arm (Figure 1). Three trials reported ER and/or progesterone receptor (PgR)-positive ORR representing a total of 1031 patients.9-11 The outcome was 58.4% (295/505) and 57.2% (301/526) for AIs arm and tamoxifen arm. The Q-statistics for heterogeneity was statistically significant (P ⬍ .0001) and the I2-statistic indicated heterogeneity (89.4%). The ORR was 1.41 (95% CI, 0.57-3.51; P ⫽ .46) by random effect model which suggested that there was no difference between the AIs arm and the tamoxifen arm (Figure 2).
Clinical Breast Cancer August 2011
247
AIs Versus Tamoxifen for Advanced Breast Cancer Table 1 Characteristics of Trials Characteristic
Paridaens 2008 Mouridsen 2004 Paridaens 2003 Alfredo 2003 Nabholtz 2000 Bonneterre 2000
Chemotherapy Schedule Arm A
Exemestane
Letrozole
Exemestane
Anastrozole
Anastrozole
Anastrozole
Arm B
Tamoxifen
Tamoxifen
Tamoxifen
Tamoxifen
Tamoxifen
Tamoxifen
Sample Size
371
907
120
238
353
668
37-87
31-96
37-87
55-77
30-92
34-92
ECOG ⱕ 2
KPS ⱖ 50%
ECOG ⱕ 2
ECOG ⱕ 2
NR
NR
ER and/or PgR-positive
93.3
66.1
92.5
NR
85.3
43.1
Both negative
0.3
0.3
0
NR
3.4
1.6
Age Performance Status ER/PgR Status (%)
Both unknown
6.4
33.6
7.5
NR
0.3
55.1
Allowed
Allowed
Allowed
Not allowed
Allowed
Allowed
Arm A
9.9 months
10.3 months
NR
18.0 months
11.1 months
8.2 months
Arm B
5.8 months
6.5 months
NR
7.0 months
5.6 months
8.3 months
Arm A
45.6
30.2
44.6
35.5
21.1
32.9
Arm B
31.2
20.3
17.5
26.5
17.0
32.6
Arm A
75.3
50.1
62.5
82.6
59.1
56.2
Arm B
66.6
38.1
43.9
55.6
45.6
55.5
Arm A
47.4
67.7
40.4
NR
NR
NR
Arm B
33.5
77.0
18.5
NR
NR
NR
Arm A
54.9
81.2
NR
39.7
NR
NR
Arm B
57.1
70.9
NR
11.1
NR
NR
Previous hormone treatment Efficacy Median time to progression
Overall response rates (%)
Clinical benefit (%)
ER and/or PgR-positive overall response rates (%)
Overall survival (%)
Abbreviations: ER ⫽ estrogen receptor; PgR ⫽ progesterone receptor; NR ⫽ not reported; EGCO ⫽ Eastern Cooperative Oncology Group; KPS ⫽ Karnofsky Performance Status.
Clinical Benefit All the trials reported CB representing a total of 2650 patients.9-14 The CB was 59.8% (791/1323) and 49.3% (654/ 1327) for the AIs arm and the tamoxifen arm. The Q-statistics showed the presence of heterogeneity (P ⫽ .005) and the I2statistic detected heterogeneity (70.2%). The pooled results by random effect model demonstrated a statistical difference in the clinical benefit rate between patients who had AIs treatment and those who had tamoxifen treatment (OR, 1.70; 95% CI, 1.242.33; P ⫽ .0009; Figure 3).
Overall Survival Three studies reported OS as one of the outcomes representing a total of 1516 patients.9,10,12 The OS was increased for the AIs arm (516/756 ⫽ 68.3%) compared with the tamoxifen arm (443/760 ⫽ 58.3%). The pooled OR by random effect model was 1.95 with a 95% CI ranged from 0.88 to 4.30. The Q-statistics for heterogeneity was statistically significant with P ⬍ .0001 and the I2-statistic detected heterogeneity (89.6%). The
248
Clinical Breast Cancer August 2011
ORR suggests that there is no difference between the AIs arm and the tamoxifen arm in terms of OS with P value .10 (Figure 4).
Toxicities Among the adverse events, we compared hot flashes, nausea, bone pain, vaginal bleeding, and thromboembolic event. All the trials reported the incidence rate of hot flashes,9-14 five trials reported nausea and bone pain,9-11,13,14 four trials reported vaginal bleeding,9,12-14 and three trials reported thromboembolic event.12-14 The Q-statistics of hot flashes (P ⫽ .006) and nausea (P ⫽ .009) showed the presence of heterogeneity; bone pain (P ⫽ .92), vaginal bleeding (P ⫽ .82), and thromboembolic events (P ⫽ .19) demonstrated the homogeneity. We obtained pooled results by random effect model or fixed effect model. The pooled results were as follows: hot flashes (OR, 0.99; 95% CI, 0.701.40; P ⫽ .95), nausea (OR, 1.06; 95% CI, 0.72-1.55; P ⫽ .77), bone pain (OR, 0.99; 95% CI, 0.81-1.22; P ⫽ .94), vaginal bleeding (OR, 0.30; 95% CI, 0.16-0.56; P ⫽ .0002), and thromboembolic event (OR, 0.47; 95% CI, 0.28-0.77; P ⫽ .003; Table 2). Hot flashes, nausea, and
Hong-Bin Xu et al Figure 1 Overall Response Rate of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category Bonneterre 2000 Nabholtz 2000 Alfredo 2003 Paridaens 2003 Mouridsen 2004 Paridaens 2008
aromarase inhibitor n/N
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
112/340
107/328
22.00
1.01 [0.73, 1.40]
36/171 43/121 25/56 137/453
31/182 31/117 10/57 92/454
14.78 14.19 8.04 22.73
1.30 [0.76, 2.21] 1.53 [0.88, 2.66] 3.79 [1.60, 8.98] 1.71 [1.26, 2.31]
83/182
59/189
18.26
1.85 [1.21, 2.82]
100.00
1.56 [1.17, 2.07]
1323 1327 Total (95% CI) Total events: 436 (aromarase inhibitor), 330 (tamoxifen) Test for heterogeneity: χ? = 11.94, df = 5 (P = 0.04), I? = 58.1% Test for overall effect: Z = 3.05 (P = 0.002) 0.1 0.2
0.5 1
2
5 10
Figure 2 Overall Response Rate by ER and/or PgR Positive of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category Paridaens 2003 Mouridsen 2004 Paridaens 2008 Total (95% CI)
aromarase inhibitor n/N
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
23/57 199/294
10/54 235/305
28.54 36.29
2.98 [1.25, 7.08] 0.62 [0.43, 0.90]
73/154
56/167
35.18
1.79 [1.14, 2.80]
505
526
100.00
1.41 [0.57, 3.51]
Total events: 295 (aromarase inhibitor), 301 (tamoxifen) Test for heterogeneity: χ? = 18.83, df = 2 (P = 0.0001), I? = 89.4% Test for overall effect: Z = 0.74 (P = 0.46) 0.01
bone pain were not different between the AIs and tamoxifen groups. However, vaginal bleeding and thromboembolic events were significantly more frequent in the tamoxifen group (P ⫽ .0002).
Publication Bias The evaluation of publication bias showed that the Egger test was not significant (P ⫽ .292). The funnel plots of the log OR versus the SE of log OR for publication bias formed a distinct funnel shape regardless of the study variance, and also did not show some asymmetry (figure not shown). These results indicate there was not a potential for publication bias.
Discussion Breast cancer is the most common form of cancer affecting women.1,17 Advanced breast cancer in postmenopausal women remains an incurable disease, and so treatment is aimed at palliation and improved quality of life, inhibition of disease progression, and
0.1
1
10
100
improvement in survival time where possible. Endocrine therapy is now a widely accepted treatment modality and is primarily directed at reducing the synthesis of estrogen or alternatively blocking estrogen receptors in breast cancer that are hormone-sensitive. Tamoxifen has been the accepted gold-standard first-line treatment for advanced breast cancer in postmenopausal women since the late 1970s. Tamoxifen acts by blocking the binding of estrogen to its receptor and has an ORR of 30%-35% when used as first-line therapy for advanced breast cancer.18 Adverse effects that have been associated with tamoxifen can be classified as either owing to its antiestrogenic actions (eg, hot flashes, vaginal bleeding, discharge, or dryness) or as more general effects (eg, nausea, vomiting, tumor flare, and skin rash). The serious and long-term adverse effects of tamoxifen in the treatment of advanced breast cancer have led to increasing use of alternative hormonal therapies such as the new generation AIs.
Clinical Breast Cancer August 2011
249
AIs Versus Tamoxifen for Advanced Breast Cancer Figure 3 Clinical Benefit of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category
aromarase inhibitor n/N
Bonneterre 2000 Nabholtz 2000 Alfredo 2003 Paridaens 2003 Mouridsen 2004 Paridaens 2008
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
191/340
182/328
20.56
1.03 [0.76, 1.40]
101/171 100/121 35/56 227/453
83/182 65/117 25/57 173/454
17.47 13.31 10.35 21.62
1.72 [1.13, 2.62] 3.81 [2.10, 6.91] 2.13 [1.01, 4.53] 1.63 [1.25, 2.12]
137/182
126/189
16.68
1.52 [0.97, 2.39]
100.00
1.70 [1.24, 2.33]
1323 1327 Total (95% CI) Total events: 791 (aromarase inhibitor), 654 (tamoxifen) Test for heterogeneity: χ? = 16.78, df = 5 (P = 0.005), I? = 70.2% Test for overall effect: Z = 3.33 (P = 0.0009) 0.1 0.2
0.5 1
2
5 10
Figure 4 Overall Survival of Aromarase Inhibitor Comparing to Tamoxifen in Advanced Breast Cancer
Study or sub-category Alferdo 2003 Mouridsen 2004 Paridaens 2008
aromarase inhibitor n/N
tamoxifen n/N
OR (random) 95% CI
Weight %
OR (random) 95% CI
29.64 35.89
5.26 [2.66, 10.40] 1.77 [1.30, 2.42]
48/121 368/453
13/117 322/454
100/182
108/189
34.47
0.91 [0.61, 1.38]
756
760
100.00
1.95 [0.88, 4.30]
Total (95% CI)
Total events: 516 (aromarase inhibitor), 443 (tamoxifen) Test for heterogeneity: χ? = 19.28, df = 2 (P < 0.0001), I? = 89.6% Test for overall effect: Z = 1.65 (P = 0.10) 0.01
0.1
1
10
100
Table 2 Overall Analysis - Toxicities Outcome
RCTs
Patients
OR
95% CI
P Value
Heterogeneity (P)
Hot Flashes
6
2653
0.99
0.70-1.40
.95
0.006
Nausea
5
2415
1.06
0.72-1.55
.77
0.009
Bone Pain
5
2415
0.99
0.81-1.22
.94
0.92
Vaginal Bleeding
4
1626
0.30
0.16-0.56
.0002
0.82
Thromboembolic Event
3
1255
0.47
0.28-0.77
.003
0.19
Abbreviations: RCTs ⫽ randomized controlled trials; OR ⫽ odds ratio; CI ⫽ confidence interval.
AIs inhibit the conversion of androgens into estrogens, thereby significantly reducing circulating estrogens in postmenopausal women. Their efficacy in breast cancer is well established.19 Our meta-analysis indicated that AIs were more effective than tamoxifen in terms of ORR and CB as first-line hormonal therapy in post-
250
Clinical Breast Cancer August 2011
menopausal women with advanced breast cancer. Furthermore, there was no difference in clinically relevant toxicities between the two treatments except vaginal bleeding and thromboembolic events. The six trials we included used AIs including the steroidal compound exemestane as well as nonsteroidal compounds such as anas-
Hong-Bin Xu et al trozole and letrozole.20-22 Two trials used exemestane,9,11 one trial used letrozole,10 and three trials used anastrozole.12-14 The outcomes (ORR, CB, and OS rate) and the incidence rate of toxicities (hot flashes, nausea, bone pain, vaginal bleeding, and thromboembolic events) were inconsistent in these trials, leading to controversies about treatment preference. We attempted to identify which agent is more effective and less toxic as first-line hormonal therapy in postmenopausal women with advanced breast cancer. Although there was no difference between therapies in the incidence of hot flashes, nausea, or bone pain, the incidence of vaginal bleeding or thromboembolic events was statistically significant. Therefore, we could carefully suggest that AIs are superior to tamoxifen in terms of toxicities. One limitation in our meta-analysis is the data source extracted from published studies not based on individual patient data. Metaanalysis based on published data tends to overestimate treatment effects compared with individual data analysis. However, analysis using individual data may include fewer studies if authors do not agree to submit their full database to the analyzing group. In general, an individual database meta-analysis would give a more robust estimation for the association; therefore, we should interpret the results with care, especially for a positive result. Although the risk of publication bias exists in any meta-analysis, whether based on individual patient data or not, we feel this was not an important aspect of our study. An important unexplored issue in our analysis to consider is the treatment cost, to be balanced with the benefit of AIs. Unfortunately, no planned analysis across all randomized trials evaluating this modern topic is available.
Conclusion Although considering all drawbacks and controversies related to literature-based meta-analysis,23 AIs have better ORRs and clinical benefit rates than tamoxifen and in terms of toxicities, is not inferior to tamoxifen; therefore, we could suggest AIs instead of tamoxifen for those postmenopausal patients with hormone receptor-positive advanced breast cancer.
Acknowledgments The authors thank Cochrane Collaboration for providing the software Review Manager used in this review. We also thank all the patients who had been treated in these trials.
Disclosure There was no funding for this article. Conflict of interest: We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service, and/or
company that could be construed as influencing the position presented in, or the review of, the manuscript entitled, “Aromatase inhibitor versus tamoxifen in postmenopausal woman with advanced breast cancer: a literature-based meta-analysis”.
References 1. Yancik R, Ries LA. Aging and cancer in America: demographic and epidemiologic perspectives. Hematol Oncol Clin North Am 2000; 14:17-23. 2. Alghanem AA, Hussain S. The effect of age on estrogen and progesterone receptor in primary breast cancer. J Surg Oncol 1985; 30:29-32. 3. Elwood JM, Godolphin W. Oestrogen receptors in breast tumours: associations with age, menopausal status and epidemiological and clinical features in 735 patients. Br J Cancer 1980; 42:635-44. 4. Jaiyesimi IA, Buzdar AU, Decker DA, et al. Use of tamoxifen for breast cancer: twenty-eight years later. J Clin Oncol 1995; 13:513-29. 5. Ragaz J. Status of aromatase inhibitors in relation to other breast cancer treatment modalities. Endocr Relat Cancer 1999; 6:277-91. 6. Goss PE, Ingle JN, Martino S, et al. A randomized trial of letrozole in postmenopausal women after five years of tamoxifen therapy for early-stage breast cancer. N Engl J Med 2003; 349:1793-802. 7. Gerber B, Krause A, Reimer T, et al. Anastrozole versus tamoxifen treatment in postmenopausal women with endocrine- responsive breast cancer and tamoxifeninduced endometrial pathology. Clin Cancer Res 2006; 12:1245-50. 8. Perez EA. Safety of aromatase inhibitors in the adjuvant setting. Breast Cancer Res Treat 2007; 105 suppl 1:75-89. 9. Paridaens RJ, Dirix LY, Louk V, et al. Phase III study comparing exemestane with tamoxifen as first-line hormonal treatment of metastatic breast cancer in postmenopausal women: The European Organisation for Research and Treatment of Cancer Breast Cancer Cooperative Group. J Clin Oncol 2008; 26:4883-90. 10. Mouridsen H, Chaudri-Ross HA. Efficacy of first-line letrozole versus tamoxifen as a function of age in postmenopausal women with advanced breast cancer. Oncologist 2004; 9:497-506. 11. Paridaens RJ, Dirix LY, Lohrisch C, et al. Mature results of a randomized phase II multicenter study of exemestane versus tamoxifen as first-line hormone therapy for postmenopausal women with metastatic breast cancer. Ann Oncol 2003; 14:1391-8. 12. Milla-Santos A, Milla L, Portella J, et al. Anastrozole versus tamoxifen as first-line therapy in postmenopausal patients with hormone-dependent advanced breast cancer: a prospective, randomized, phase III study. Am J Clin Oncol 2003; 26:317-22. 13. Bonneterre J, Thurlimann B, Robertson JF. Anastrozole versus tamoxifen as firstline therapy for advanced breast cancer in 668 postmenopausal women: results of the Tamoxifen or Arimidex Randomized Group Efficacy and Tolerability study. J Clin Oncol 2000; 18:3748-57. 14. Nabholtz JM, Buzdar A, Pollak M, et al. Anastrozole is superior to tamoxifen as first-line therapy for advanced breast cancer in postmenopausal women: results of a North American multicenter randomized trial. Arimidex Study Group. J Clin Oncol 2000; 18:3758-67. 15. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177-88. 16. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in metaanalyses. BMJ 2003; 327:557-60. 17. Kimmick GG, Balducci L. Breast cancer and aging: clinical interactions. Hematol Oncol Clin North Am 2000; 14:213-34. 18. Rose C, Mouridsen HT. Endocrine therapy of advanced breast cancer. Acta Oncol 1988; 27:721-8. 19. Dixon JM. Exemestane and aromatase inhibitors in the management of advanced breast cancer. Expert Opin Pharmacother 2004; 5:307-16. 20. Paridaens R, Thomas J, Wildiers J, et al. Safety, activity and estrogen inhibition by exemestane in postmenopausal women with advanced breast cancer: A phase I study. Anticancer Drugs 1998; 9:675-83. 21. Jones S, Vogel C, Arkhipov A, et al. Multicenter, phase II trial of exemestane as third-line hormonal therapy of postmenopausal women with metastatic breast cancer: Aromasin Study Group. J Clin Oncol 1999; 17:3418-25. 22. Hamilton A, Piccart M. The third-generation non-steroidal aromatase inhibitors: a review of their clinical benefits in the second-line hormonal treatment of advanced breast cancer. Ann Oncol 1999; 10:377-84. 23. Predbois P, Buyse M. Meta-analysis based on abstracted data: a step in the right direction, but only a first step. J Clin Oncol 2004; 22:3839-41.
Clinical Breast Cancer August 2011
251