The decrease in breast-cancer incidence in 2003 in the united states

T h e n e w e ng l a n d j o u r na l o f m e dic i n e The Decrease in Breast-Cancer Incidence
in 2003 in the United States
Peter M. Ravdin, Ph.D., M.D., Kathleen A. Cronin, Ph.D., Nadia Howlader, M.S., Christine D. Berg, M.D., Rowan T. Chlebowski, M.D., Ph.D., Eric J. Feuer, Ph.D., Brenda K. Edwards, Ph.D., and Donald A. Berry, Ph.D.
age-adjusted incidence of breast cancer by an av- erage of about 0.5% per year, a rise that was par- An initial analysis of data from the National Cancer ticularly evident among women who were 50 years Institute’s Surveillance, Epidemiology, and End Re- of age or older2 (Fig. 1). Changes in reproductive sults (SEER) registries shows that the age-adjust- factors, in the use of menopausal hormone- ed incidence rate of breast cancer in women in the replacement therapy, in mammographic screening, United States fell sharply (by 6.7%) in 2003, as in environmental exposures, and in diet have all compared with the rate in 2002. Data from 2004 been proposed to explain the trend. Of these fac- showed a leveling off relative to the 2003 rate, tors, only the use of hormone-replacement therapy with little additional decrease. Regression analy- changed substantially between 2002 and 2003.
sis showed that the decrease began in mid-2002 In this report, we provide additional data from and had begun to level off by mid-2003. A com- 2004 that show little change in breast-cancer in- parison of incidence rates in 2001 with those in cidence between 2003 and 2004. A comparison of 2004 (omitting the years in which the incidence incidence rates in 2001 with those in 2004 (omit- was changing) showed that the decrease in annual ting the years in which the incidence was in the age-adjusted incidence was 8.6% (95% confidence process of changing) showed that the decrease in interval [CI], 6.8 to 10.4). The decrease was evident annual age-adjusted incidence was 8.6% (95% CI, only in women who were 50 years of age or older 6.8 to 10.4).
and was more evident in cancers that were estro- The decrease in breast-cancer incidence began gen-receptor–positive than in those that were in mid-2002 and occurred shortly after the high- estrogen-receptor–negative. The decrease in breast- ly publicized series of reports from the random- cancer incidence seems to be temporally related ized trial of the Women’s Health Initiative, which to the first report of the Women’s Health Initia- reported a significant increase in the risks of cor- tive and the ensuing drop in the use of hor- onary heart disease and breast cancer associated mone-replacement therapy among postmenopaus- with the use of estrogen–progestin combination al women in the United States. The contributions therapy.3 By the end of 2002, the use of hormone- of other causes to the change in incidence seem replacement therapy had decreased by 38% in the less likely to have played a major role but have not United States, with approximately 20 million few- er prescriptions written in 2003 than in 2002.4,5 The analyses we report here used information Major changes in cancer incidence and death from the SEER Program of the National Cancer rates, as detected in cancer-registry data, provide Institute (NCI) collected from nine cancer regis- unique opportunities to examine questions related tries reporting on 9% of the U.S. population. to the cause, prevention, detection, and treatment Trends in the incidence of female breast cancer of cancer. In a preliminary report, we suggested were age-adjusted to the standard population in that such a major change in breast-cancer inci- the year 2000 and were adjusted for reporting de- dence occurred in 2003 in the United States.1 In lays. Joinpoint (version 3.0) statistical software contrast, the 1990s saw an increase in the annual (http:/ srab.cancer.gov/joinpoint/) was used for fit- n engl j med 356;16 www.nejm.org april 19, 2007 Downloaded from nejm.org on March 19, 2012. For personal use only. No other uses without permission. Copyright 2007 Massachusetts Medical Society. All rights reserved. ting trends over time and to evaluate when changes in trends occurred. The number of patients with unknown estrogen-receptor status changed from 15% in 2001 to 8% in 2004; to adjust for this change, multiple imputation was used to generate estrogen-receptor values for missing data.
Comparison of incidence rates in 2001 with rates in 2004 (omitting the years in which the in- cidence was rapidly changing) showed that the de- crease in annual age-adjusted incidence was evi- dent only in women who were 50 years of age or Rate per 100,000 Women
more. During that period, there was an increase of 1.3% (95% CI, −3.1 to 5.8) in incidence for women below the age of 50 years, a decrease of 11.8% (95% CI, 9.2 to 14.5) for women between the ages of 50 and 69 years, and a decrease of 11.1% (95% CI, 7.9 to 14.2) for women 70 years Year of Diagnosis
Figure 1. Annual Incidence of Female Breast Cancer (1975–2004).
Data are from nine of the NCI’s SEER registries. SEER sites include San years, the decrease was more evident in those with Francisco, Connecticut, Detroit (metropolitan area), Hawaii, Iowa, New estrogen-receptor–positive tumors (14.7%; 95% CI, Mexico, Seattle–Puget Sound, Utah, and Atlanta (metropolitan area). 11.6 to 17.4) than in those with estrogen-recep- tor–negative tumors (1.7%; 95% CI, −4.6 to 8.0). The decreases were similar for localized disease One possibility (11.3%; 95% CI, 8.0 to 14.6) and more advanced seems unlikely. T disease (13.6%; 95% CI, 9.2 to 17.9) and were dence in 2003 wa evident in primary breast cancers (13.7%; 95% CI, tries, there was n 4-C ange SIZE
11.0 to 16.4) but not in contralateral second pri- in the incidence mary or later breast cancers, for which there cer in women during this period, and the l AUTHOR, PLEASE NOTE: ower
was a nonsignificant increase (9.4%; 95% CI, rates continued in 2004. Could the change have Figure has been redrawn and type has been reset.
Please check carefully.
been related to a major decrease in the rate of Figure 2A shows the quarterly, age-adjusted screening mam incidence rates of breast cancer in women between 3.2% in this rate was reported for women between the ages of 50 and 69 years, categorized according the ages of 50 and 65 years for 2003, as compared to estrogen-receptor status. The data for change with that for 2000,6 such a change would seem in trend were examined with the use of Joinpoint insufficient to explain the observation. A change statistical software. Changes in trend in mid-2002 in screening patterns specific to women who for- and mid-2003 were evident for all patients and for merly received hormone-replacement therapy is patients with estrogen-receptor–positive tumors also a possibility. For example, if women who but not for those with estrogen-receptor–negative discontinued hormone-replacement therapy also tumors. However, the low incidence of estrogen- stopped receiving mammograms, an apparent de- receptor–negative tumors limited the statistical crease in incidence could result. Although visits ability to detect a change in trend. For all patients, to physicians would probably decrease among the quarterly changes in rate were an increase of women who discontinued hormone-replacement 0.08% (95% CI, −0.60 to 0.77) in the first time therapy, no published data are available showing interval, a decrease of 4.43% (95% CI, −12.66 to a substantial decrease in mammographic screen- 4.75) in the next time interval defined by Join- ing in such women. Another possible explanation point analysis, and a decrease of 0.04% (95% CI, is that a decrease in incidence is expected in a −1.56 to 1.50) in the last time interval.
heavily screened population, similar to that re- What might have been responsible for the ported for prostate cancer. No sudden decrease sharp decline in breast-cancer incidence, followed has yet been reported for breast-cancer incidence by a relative stabilization at a lower incidence rate? in heavily screened populations. n engl j med 356;16 www.nejm.org april 19, 2007 Downloaded from nejm.org on March 19, 2012. For personal use only. No other uses without permission. Copyright 2007 Massachusetts Medical Society. All rights reserved. T h e n e w e ng l a n d j o u r na l o f m e dic i n e estrogen-receptor status is minor. Thus, a drop in screening would result in an approximately equal decrease in estrogen-receptor–positive and estro- gen-receptor–negative tumors, an expectation that Discontinuation of hormone-replacement ther- apy could have caused a decreased incidence of breast cancer by direct hormonal effects on the growth of occult breast cancers, a change that would have been expected to affect predomi- nantly estrogen-receptor–positive tumors. If the decrease in breast-cancer incidence had been as- Quarterly Rate per 100,000 Women
sociated with discontinuation of hormone-replace- ment therapy, the rapidity of change suggested Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 that clinically occult breast cancers stopped pro- gressing or even regressed soon after discontin- Year of Diagnosis
uation of the therapy. The hypothesis that hor- mone withdrawal can rapidly influence the growth of breast cancer is supported by anecdotal reports of regression of breast cancer after discontinua- tion of hormone-replacement therapy.8 A cessation of such therapy was associated with a reduction in the proliferative index of breast-cancer cells within 1 month in women with estrogen-recep- tor–positive tumors but not in those with estro- gen-receptor–negative tumors in the same set- ting,9 and responses within weeks after estrogen deprivation have been seen in clinical trials of No. of Prescriptions (millions)
neoadjuvant hormones. An early effect of tamoxi- fen was seen in the Breast Cancer Prevention Trial, in which the cumulative rates of invasive breast Reporting Year
cancer in the tamoxifen group and the placebo group appeared to diverge within the first few Figure 2. Quarterly Incidence of Breast Cancer in Women between the
months and differed statistically at the end of the Ages of 50 and 69 Years, According to Estrogen-Receptor (ER) Status,
first year.10 An analysis of 51 epidemiologic stud- and the Number of Prescriptions for Hormone-Replacement Therapy
(2000–2004).
ies showed that an elevated risk of breast cancer of the NCI’s SEER registries, with trends mod- after the use of hormone-replacement therapy had istical software (Joinpoint). Trends were largely if not wholly disappeared within 5 years o the standard population in the year 20 Revised after discontinuation of therapy, although a more for reporting delays. Panel B shows the number4-C of prescri SIZE
detailed analysis of the time course of changes in the United States for the combined estH/T risk within this period was not presented.11 and the conjugated equine estrogen Premarin, according to year. AUTHOR, PLEASE NOTE:
Notably, the change in the use of hormone- Figure has been redrawn and type has been reset.
Please check carefully.
replacement therapy also followed a time course One of the arguments against changes in mam- that was similar to the decline in breast-cancer in- 04-19-07 on for the cidence, with a sharp decline followed by a rela- decline is that the effect was mainly on estrogen- tive stabilization at a new, lower level. The total receptor–positive tumors. Breast cancers that are number of prescriptions for the two most com- detected on mammography are more likely to be monly prescribed forms of hormone-replacement estrogen-receptor–positive than are tumors not therapy in the United States — Premarin and detected on mammography (80% vs. 70%),7 but Prempro — had their steepest declines starting the difference in the percentages according to in 2002 and particularly in 2003 (62 million pre- n engl j med 356;16 www.nejm.org april 19, 2007 Downloaded from nejm.org on March 19, 2012. For personal use only. No other uses without permission. Copyright 2007 Massachusetts Medical Society. All rights reserved. scriptions in 2000, 61 million in 2001, 47 million apy has already occurred, but important questions in 2002, 27 million in 2003, 21 million in 2004, remain. Can we expect only a delay in the appear- and 18 million in 2005)12 (Fig. 2B).
ance of clinically detectable tumors, with no re- Other medications can influence the incidence duction in long-term incidence, or will there be a of breast cancer. These drugs include tamoxifen long-term reduction? A change in the hormonal and raloxifene, and there is some evidence for milieu may have slowed the growth of tumors beneficial effects of nonsteroidal antiinflamma- slightly or temporarily. If this is the case, as the tory drugs, statins, and calcium and vitamin D use of hormone-replacement therapy stabilizes, supplements. However, none of these agents were breast-cancer incidence should rise again. Alter- used by a substantial portion of postmenopausal natively, the change in hormonal milieu may have women or showed substantial change in use dur- a more profound effect, similar to that of hor- ing the period from 2000 to 2004.12,13 Therefore, monal adjuvant therapy.16 the drugs are unlikely candidates for causing the We believe that the data are most consistent with a direct effect of hormone-replacement ther- When the results of the Women’s Health Initia- apy on preclinical disease, but this conclusion does tive hormone trial were announced, women were not rule out some contribution from changes in asked to discontinue their study medications (pla- screening mammography. In any case, attempts cebo or hormone) but were encouraged to con- to understand the rapid reduction in incidence tinue undergoing annual mammography. These using theoretical models of breast-cancer evolu- women continue to be followed for clinical out- tion and the effects of screening and treatment come, and a report of follow-up of the combined — such as those of the NCI’s Cancer Interven- estrogen-plus-progestin trial is anticipated later tion and Surveillance Modeling Network17 — may this year. This report will provide the highest level lead to new insights into the development and of evidence concerning the influence of cessa- prevention of breast cancer.
tion of hormone-replacement therapy on the in- No potential conflict of interest relevant to this article was re- cidence of breast cancer. Other observers have ported.
noted a decline in breast-cancer incidence after From the Department of Biostatistics, M.D. Anderson Cancer 2002. A report from a subgroup of California reg- Center, Houston (P.M.R., D.A.B.); the Division of Cancer Control istries also showed a sharp decrease in breast- and Population Sciences (K.A.C., N.H., E.J.F., B.K.E.) and the Di- vision of Cancer Prevention (C.D.B.), National Cancer Institute, cancer incidence in 2003 and suggested that it ex- Bethesda, MD; and the Los Angeles Biomedical Research Insti- tended into 2004.14 A recent analysis of national tute at Harbor–UCLA Medical Center, Torrance, CA (R.T.C.).
cancer data by Jemal et al.15 showed a decline in 1. Ravdin PM, Cronin KA, Howlander N, Chlebowski RT, Berry
the incidence of breast cancer in 2003 but did not DA. A sharp decrease in breast cancer incidence in the United States comment on its clinical relevance. The joinpoint in 2003. Breast Cancer Res Treat 2006;100:Suppl:S2. abstract.
in that study was done with annual (rather than 2. Howe HL, Wu X, Ries LAG, et al. Annual report to the nation
on the status of cancer, 1975-2003, featuring cancer among U.S. quarterly) data. Annual rates obscure within-year Hispanic/Latino populations. Cancer 2006;107:1711-42.
trends, in this case within years 2002 and 2003. 3. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and ben-
In addition, the statistical method used by Jemal efits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative et al. cannot select the final year in a range (in randomized controlled trial. JAMA 2002;288:321-33.
this case, 2003) as demonstrating a discontinuity. 4. Buist DSM, Newton KM, Miglioretti DL, et al. Hormone
It is possible that the ultimate understanding therapy prescribing patterns in the United States. Obstet Gyne- of the effect of cessation of hormone-replacement 5. Hersh AL, Stefanick ML, Stafford RS. National use of post-
therapy will be complex; it will probably depend menopausal hormone therapy: annual trends and response to on more than one mechanism and will be affect- recent evidence. JAMA 2004;291:47-53.
6. Health, United States 2006, with chartbook on trends in the
ed in different ways by various forms of post- health of Americans. Hyattsville, MD: National Center for Health menopausal hormone-replacement therapy. The Statistics, November 2006:313-4. (DHHS publication no. 2006- time course of the decrease in breast-cancer in- 1232.) 7. Porter PL, El-Bastawissi AY, Mandelson MT, et al. Breast tu-
cidence is of both practical and theoretical in- mor characteristics as predictors of mammographic detection: terest. Our data suggest that much of the decrease comparison of interval- and screen-detected cancers. J Natl Can- in breast-cancer incidence that is attributable to cer Inst 1999;91:2020-8.
8. Powles TJ, Hickish T. Breast cancer response to hormone
changes in the use of hormone-replacement ther- replacement therapy withdrawal. Lancet 1995;345:1442.
n engl j med 356;16 www.nejm.org april 19, 2007 Downloaded from nejm.org on March 19, 2012. For personal use only. No other uses without permission. Copyright 2007 Massachusetts Medical Society. All rights reserved. 9. Prasad R, Boland GP, Cramer A, Anderson E, Knox WF, Bun-
MH, Ballard-Barbash R. Tamoxifen use for breast cancer che- dred NJ. Short-term biologic response to withdrawal of hormone moprevention among U.S. women. Eur J Cancer Suppl 2004;2: replacement therapy in patients with invasive breast carcinoma. 17-8.
14. Clarke CA, Glaser SL, Uratsu SL, Selby JV, Kushi LH, Her-
10. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for rington LJ. Recent declines in hormone therapy utilization and
prevention of breast cancer: report of the National Surgical Ad- breast cancer incidence: clinical and population-based evidence. juvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst J Clin Oncol 2006;24:49e-50e.
15. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer
11. Collaborative Group on Hormonal Factors in Breast Cancer. statistics, 2007. CA Cancer J Clin 2007;57:43-66.
Breast cancer and hormone replacement therapy: collaborative 16. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG).
reanalysis of data from 51 epidemiological studies of 52,705 women Effects of chemotherapy and hormonal therapy for early breast with breast cancer and 108,411 women without breast cancer. cancer on recurrence and 15-year survival: an overview of the Lancet 1997;350:1047-59. [Erratum, Lancet 1997;350:1484.] randomised trials. Lancet 2005;365:1687-717.
12. Drug Topics. Drugs by units in the United States in specific 17. Berry DA, Cronin KA, Plevritis SK, et al. Effect of screening
years. (Accessed March 29, 2007, at http://www.drugtopics.com/ and adjuvant therapy on mortality from breast cancer. N Engl J 13. Freedman AN, Graubard BI, McCaskill-Stevens W, Gail Copyright 2007 Massachusetts Medical Society.
n engl j med 356;16 www.nejm.org april 19, 2007 Downloaded from nejm.org on March 19, 2012. For personal use only. No other uses without permission. Copyright 2007 Massachusetts Medical Society. All rights reserved.

Source: http://math.oit.edu/~reidj/classes/S414materials/Week%202/The%20Decrease%20in%20Breast%20Cancer%20Incidence%20in%202003%20in%20the%20United%20States.pdf