Originally published as JCO Early Release 10.1200/JCO.2005.04.0014 on May 15 2006

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jagsi, R. Articles by Pierce, L. J. Search for Related Content PubMed PubMed Citation Articles by Jagsi, R. Articles by Pierce, L. J.

Stroke Rates and Risk Factors in Patients Treated With Radiation Therapy for Early-Stage Breast Cancer

Reshma Jagsi, Kent A. Griffith, Todd Koelling, Rachel Roberts, Lori J. Pierce

From the Radiation Oncology Department; University of Michigan Cancer Center Biostatistics Core; and the University of Michigan School of Medicine, Internal Medicine Department, Ann Arbor, MI

Address reprint requests to Lori J. Pierce, MD, Department of Radiation Oncology, University of Michigan School of Medicine, 4308 Cancer and Geriatrics Center, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0010; e-mail: ljpierce{at}umich.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
PURPOSE: To examine whether stroke risk is elevated in American breast cancer patients treated with modern techniques, as well as whether supraclavicular radiation therapy (RT) is associated with increased risk.

METHODS: Observed rates of stroke in 820 eligible early-stage breast cancer patients treated at the University of Michigan Hospital (Ann Arbor, MI) were compared with expected rates. Relationships between potential risk factors and actuarial rate of first stroke were analyzed.

RESULTS: Median follow-up was 6.8 years. Twenty patients had at least one cerebrovascular accident (CVA) in follow-up; 35 patients had at least one CVA or transient ischemic attack (CVA/TIA). The standardized incidence ratios were 1.74 (0.94 to 2.37) for CVA and 1.68 (1.003 to 2.06) for CVA/TIA. The absolute excess risk per 1,000 patients per year was 1.67 for CVA and 2.76 for CVA/TIA. On bivariate analysis, factors significantly associated with actuarial rate of first CVA included hypertension (P = .002), age (P < .0001), coronary artery disease (P = .001), atrial fibrillation (P = .009), and supraclavicular RT (P = .021). Factors associated with CVA/TIA were hypertension (P < .001), coronary artery disease (P = .002), and age (P < .0001). Tamoxifen use alone was not significant (P = .19), but tamoxifen combined with baseline hypertension led to increased risk of CVA/TIA (log-rank P < .0001). On multivariate analysis, only age (P < .001) and hypertension (P = .003) remained significant predictors of CVA/TIA. Age was the only significant predictor of CVA alone (P < .001).

CONCLUSION: American breast cancer survivors may have an elevated risk of stroke compared with the general population, but the absolute excess risk is low. This study found no significant association between supraclavicular RT and stroke after controlling for other factors.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
Meta-analyses of the numerous trials assessing the role of radiation therapy (RT) in the treatment of breast cancer have indicated that patients receiving RT have higher mortality from vascular causes than patients who are not treated with RT.¹ Much research to date has focused on radiation-related cardiac toxicity,^2,3 but few studies have addressed other potential causes of vascular mortality and morbidity in breast cancer patients, including stroke. However, one recent European study did reveal an increased rate of stroke in Swedish breast cancer patients, compared with the general population, and speculated that the use of RT might play a causal role in this finding.⁴

Indeed, studies have established an increased incidence of ischemic stroke^5,6 and carotid stenosis⁷ in head and neck cancer patients treated with RT. The supraclavicular fossa, including the proximal carotid artery, is often included as a target in breast cancer patients for indications such as node-positive disease. Therefore, it is critical to examine whether the common technique of supraclavicular RT is associated with stroke in breast cancer patients.

In this study, we examine whether the risk of stroke is increased in a series of early-stage breast cancer patients treated with modern radiotherapeutic techniques in the United States compared with the general population. We further seek to examine the association of a number of potential risk factors, including the administration of RT to the supraclavicular fossa, with the risk of stroke.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
From 1984 to 2000, 867 consecutive patients with American Joint Committee on Cancer 1999 stage I or II breast cancer were treated with breast-conserving surgery and RT at the University of Michigan Hospital (Ann Arbor, MI).⁸ After obtaining consent from the University of Michigan institutional review board to review patient records for the outlined study, we reviewed these patients’ inpatient and outpatient charts and found that 841 of these patients had records sufficient to determine the incidence of cerebrovascular accident (CVA) or transient ischemic attack (TIA) in follow-up. Patients found on chart review to have received other courses of RT to the neck, either before or after their primary breast cancer treatment (for other causes or for breast cancer recurrence), were excluded from analysis. Details of the patients’ breast cancer treatment were documented, including whether the radiation fields included the supraclavicular fossa, as well as any systemic therapies received. Adjuvant use of a systemic agent was defined as use of such an agent within 1 year of breast cancer diagnosis.

Charts were reviewed to determine whether patients had any established vascular risk factors at baseline (before their RT), including hypertension, diabetes, hypercholesterolemia, smoking, atrial fibrillation, history of coronary artery disease, or history of stroke. Hypertension was established if it was documented in the patient’s history, if medications included an antihypertensive agent for which there was no other obvious indication, or if two blood pressure readings in excess of 140 systolic or 90 diastolic were recorded within 1 year of one another without other explanation. Diabetes was established if listed in the history, if the patient was taking insulin or an oral hypoglycemic agent, or if the laboratory value of hemoglobin A1c exceeded 6.5. Hypercholesterolemia was established if listed in the history, if the patient was taking an antihyperlipidemic agent, if total fasting serum cholesterol exceeded 220, or if fasting LDL exceeded 160. Details of smoking history were recorded, including whether the patient had ever smoked, start and quit dates, pack-years, and whether the patient was actively smoking at the time of RT. Atrial fibrillation was established by history or ECG findings. History of coronary artery disease was also noted, including prior myocardial infarction, prior percutaneous coronary intervention or bypass surgery, catheterization-documented coronary stenosis, or documented history of angina. History of stroke was also assessed, including prior CVA causing a clinical neurological deficit of longer than 24 hours, infarction noted on brain imaging, or history of TIA without persistent neurologic deficit.

Charts were also reviewed to establish the occurrence of CVAs and TIAs in follow-up. CVAs were defined as events causing a clinical, neurologic deficit that persisted longer than 24 hours, generally confirmed by imaging studies. Only primary ischemic or thromboembolic events were included in this definition; primary cerebral hemorrhage was not coded as a CVA in either the observed population or in the expected rate calculations described below. TIAs were documented based on a treating physician’s documented clinical assessment of a transient neurologic deficit of less than 24 hours in duration. Two stroke end points were used in the analyses: one for CVA alone, and one for either CVA or TIA (CVA/TIA).

Observed rates of CVA and CVA/TIA were compared with expected age, sex, and race-specific rates to obtain standardized incidence ratios (SIRs). Expected rates were calculated from the National Hospital Discharge Surveys (NHDS), annual cross-sectional, population-based surveys of hospital discharge events for the United States civilian population, obtained from the Centers for Disease Control and Prevention.⁹ For purposes of this analysis, NHDS data from 1994 through 2003 were used to calculate the expected rates of stroke events. The NHDS uses the International Classification of Disease, ninth revision, clinical modification (ICD-9-CM) with annual addenda to codify the type of hospital discharge. Before the 1994 survey year, the ICD-9-CM did not differentiate between ischemic cerebrovascular events associated with and without infarction. In order to calculate expected rates comparable with our CVA event definition, which required infarction, only surveys from 1994 and thereafter were used. For each survey year, the NHDS database was queried by relevant ICD-9-CM diagnostic codes for the outcomes of interest. The total number of discharges for the particular outcome of interest was calculated by summing the survey weights per record (according to the NHDS instructions) by age, race, Midwest region, and ICD-9 codes. The total discharges were totaled within 5-year age groups, which overlapped with the age distribution for the observed breast cancer patients. Five-year age groups started at ages 20 to 24 years, and continued, up to a final age group of older than 85. The total discharges were also totaled by the race category assigned to the patient (white, black, or other). The total discharges by type (CVA or CVA/TIA), age group, and race were added together for all 10 surveys considered. The population estimates for the United States, published by the US Census Bureau,¹⁰ for the 10 survey years were used as the denominator for expected rate calculations.

The person-years of follow-up in the University of Michigan Radiation Oncology Breast database were calculated from the date of initiation of RT, until the patient’s date of death or last known contact date (defined strictly as the presence of records adequate to establish whether a stroke event had occurred). For each patient, the person-years of follow-up contributed to each 5-year age group were calculated, and the total person-years for the study population were added up by race category. The expected number of events for each age group was calculated by multiplying the NHDS rate by the total person-years for each age group. The expected number of events was then added up for all age groups to give an aggregate estimate. The observed-to-expected ratios—or SIRs—were then calculated. CIs were calculated by iteratively resampling our population with replacement (1,000 samples), the bootstrap technique, and calculating the SIR for each sample. Ninety-five percent CIs for the SIR estimates were constructed empirically from the boot-strap distributions, using the 2.5th percentile and 97.5th percentile as the confidence limits.¹¹

Next, actuarial curves were analyzed to examine the relationship between potential risk factors and the actuarial rate of first stroke in the patients treated at the University of Michigan; we analyzed the rate of stroke for each of the two stroke end points (CVA and CVA/TIA). The time to the first stroke was calculated from the initiation of radiotherapy. Patients without a stroke event were censored on the date of their death or last contact.

The bivariate relationships between potential risk factors and the actuarial rate of first CVA and first CVA/TIA were analyzed, using the product-limit method of Kaplan-Meier and the log-rank test statistic. To develop the best multivariate model for each end point, all characteristics were offered simultaneously to a single Cox proportional hazards regression model, using a backward elimination algorithm for variable inclusion. Characteristics found to be at least marginally significant (P < .1) were retained and compared with bivariate analyses for evidence of confounding. As none was found, final models only retained significant characteristics, with significance defined by the usual P value is less than 5% criteria. Possible two-way interactions between significant characteristics were also explored.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
Of the 867 cases reviewed, 841 patients had medical records sufficient for cerebrovascular event follow-up. Of these patients, 15 were excluded from analysis because each of these patients received an additional course of RT to the neck, either before or after treatment for their primary breast cancers, other cancers, nonmalignant conditions, or breast cancer recurrence. This left 826 patients who were analyzable for time to first cerebrovascular event, with a median follow-up of 6.8 years (minimum, 0.1 years; maximum, 20.3 years). Excluded cases (N = 41) were similar to the analyzable sample for age, race, radiation dose administered to the breast, and whether radiation was administered to the supraclavicular nodes (Table 1). For six of 826 patients (0.7%), medical records could not identify the patient’s racial group. Therefore, 820 patients were used for comparison with the general population’s expected cerebrovascular events rates and for calculation of SIRs. This sample yielded a total of 6,144 person-years of observation.

Figure 1 shows the cumulative risk of CVA during follow-up for the 826 patients in the University of Michigan series. At 10 years of follow-up, the cumulative risk was 2.7%. Table 4 presents the hazard ratios determined for a variety of potential risk factors for each of the two stroke end points (CVA alone and CVA/TIA). On bivariate analysis, factors significantly associated with actuarial rate of first CVA included hypertension (P = .002), age (P < .0001), history of coronary artery disease (P = .001), atrial fibrillation (P = .009), and supraclavicular RT (P = .021). Factors associated with CVA/TIA were hypertension (P < .001), history of coronary artery disease (P = .002), and age (P < .0001), with supraclavicular RT (P = .072) and atrial fibrillation (P = .051) trending towards significance. Tamoxifen use alone was not significant (P = .19), but tamoxifen combined with baseline hypertension did lead to increased risk of CVA/TIA (log-rank P < .0001). Figure 2 shows the unadjusted product-limit estimates for stroke failure by baseline hypertension status and adjuvant tamoxifen therapy.

View larger version (5K): [in this window] [in a new window]   Fig 1. Cumulative incidence of first ischemic cerebrovascular accident (excluding transient ischemic attacks) in 826 patients treated for early-stage breast cancer with breast-conserving surgery and radiation therapy at the University of Michigan Hospital from 1984 to 2000, from the time of breast cancer treatment.

View larger version (10K): [in this window] [in a new window]   Fig 2. Product limit estimate of first cerebral vascular event (either persistent cerebrovascular accident or transient ischemic attack) by baseline hypertension and adjuvant tamoxifen use.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
This study suggests that survivors of early-stage breast cancer in the United States may have an elevated risk of stroke compared with the general American population. This is consistent with findings of the only other study of which we are aware that explored stroke rates in breast cancer patients—a recent study of Swedish patients treated for breast cancer, which also found an increased relative risk of stroke.⁴ The mechanisms for this finding has yet to be established.

Nilsson et al,⁴ lacking specific information regarding the treatment characteristics of their patient population, were left to speculate whether RT to the supraclavicular region, as is commonly used in breast cancer patients with certain indications such as node-positive disease,¹² could have played a causal role in increasing stroke risk. Indeed, RT for head and neck cancer has been associated with increased rates of stroke and carotid stenosis.^5,6,7 However, the carotid vessels generally receive a substantially lower dose from the standard supraclavicular field for the treatment of breast cancer than from fields employed in the treatment of the neck for many head and neck malignancies. Therefore, it was particularly important in this study to explore this concern directly in a series of breast cancer patients rather than by extrapolation from the head and neck cancer experience.

Our study, which we believe is the first study in breast cancer patients to examine stroke rates in a data set also including details of breast cancer treatment and baseline cardiovascular risk factors, did not reveal any significant association between supraclavicular radiation therapy and stroke incidence, after controlling for other risk factors. These results are in agreement with the findings in the most recent Overview update.¹³ In that publication, the Early Breast Cancer Trialists’ Collaborative Group reported greater detail regarding the causes of nonbreast cancer death in 32,800 patients treated in trials of surgery with and without RT.¹³ While the incidence of heart disease was found to be significantly greater in those women who received RT, no significant excess mortality from RT was observed due to stroke (ratio of rates, 1.12; 2P = .3). These results are consistent with the findings of our study, which focused on stroke incidence rather than mortality alone, in order to account also for the morbidity that may result from stroke even when it is not a cause of death. However, it should be noted that given the size of our series, it is impossible to rule out a small increase in the risk of stroke due to the use of a supraclavicular field. Therefore, it is particularly important for physicians to continue supporting the ongoing National Cancer Institute of Canada MA-20 trial,¹⁴ as well as to anticipate the results of the European Organisation for Research and Treatment of Cancer trial of regional RT,¹⁵ to prospectively validate our results.

A number of other mechanisms are also important to consider. Tamoxifen use has been associated with an increased risk of stroke in other studies,^16,17 and while our study did not find tamoxifen use alone to be a significant independent risk factor for stroke, its use may contribute to the final risk. In particular, these data suggest that there may be an increased risk of stroke in tamoxifen users with baseline hypertension, a finding that may not have achieved independent significance on multivariate analysis due to the sample size of the study.

Particularly important is the possibility that the relationship between breast cancer and stroke is mediated by confounding risk factors. For example, obesity is a risk factor for both breast cancer and stroke,¹⁸ and may be the true underlying cause of the association between breast cancer and stroke. While the complex interplay between cytokines, hormones, and lipoproteins in the genesis of both breast cancer and stroke is not yet fully understood, it is likely that several of the same circulating factors may predispose to both conditions.^19,20

In any case, the findings of this study are reassuring insofar as they establish that while stroke rates may be increased in patients treated for early-stage breast cancer, they are still low in absolute terms, with an absolute excess risk of 1.67 per 1,000 patients per year for true CVAs. While it is possible that breast cancer treatments, such as RT to the supraclavicular region, tamoxifen, and chemotherapy, each may contribute a small amount to the increased risk suggested by this study, the present results suggest that the absolute increase in risk is small compared with the risk of recurrent breast cancer in the absence of adjuvant therapy. Moreover, the observation of increased risk of stroke in breast cancer survivors may well be confounded by other factors, such as obesity, rather than a treatment-related toxicity. Therefore, oncologists and breast cancer patients alike should not alter treatment recommendations due to concerns that breast cancer treatment causes substantial cerebrovascular morbidity. Rather, additional research should be directed towards gaining a better understanding of the underlying causes of the small increase in stroke risk found in this population.

It is important to reflect on the potential sources of bias in our analysis. The first stems from the temporal differences between the University of Michigan series, which spanned the years 1984 to 2000, and the comparison data set from the NHDS, which spanned 1994 to 2003. Methods of diagnosing stroke, and particularly imaging modalities, have advanced since the early years of our institutional series. Consequently, it is possible that advanced imaging modalities allowed physicians in the NHDS survey years to rule out stroke as the etiology of a persistent neurologic deficit of acute onset in some cases in which stroke would otherwise have been considered the most likely diagnosis, thereby contributing to some of the difference we observed. In contrast, it is also possible that advances in imaging modalities may have led to increased rates of diagnosis of stroke in the NHDS data set, which, if anything, would bias our study toward the null (ie, reduce the overall standardized incidence ratio toward 1). We believe that the net impact of temporal changes in diagnostics is unlikely to be substantial, particularly because neurologists continue to rely heavily on clinical skills of the history and physical examination in establishing the diagnosis of stroke. Furthermore, epidemiologic studies of stroke rates in the United States suggest that incidence rates have been stable^21,22 or have even slightly increased^23,24during the late 1980s and 1990s, such that any bias introduced due to temporal changes in stroke incidence would be toward the null.

Another potential source of bias is the fact that the NHDS documents only those strokes leading to hospitalization. While the vast majority of strokes do lead to hospitalization, a portion of the difference in rates of strokes we found might be due to the occurrence of strokes that did not lead to hospitalization and thus were unrecognized in the population-based comparison sample. However, in general the intensity of surveillance for strokes was not substantially greater in our series than in the NHDS surveys, insofar as we were only able to recognize stroke events that were clearly documented in patients’ medical records. Thus, strokes too minor to require hospitalization would be unlikely to have been recorded in the Michigan series either. Furthermore, strokes so severe that they caused death before hospitalization would not have been recorded in either the institutional series or the comparison data set.

This study relies heavily on the accuracy of the NHDS sampling strategy and US Census Bureau’s yearly population estimates. For our calculation of the standardized incidence ratio, the mean expected rate of CVA and CVA/TIA has been calculated from the 1994 to 2003 surveys, which have compatible ICD-9 CM coding to establish the occurrence of cerebral infarction. The mean rates of these surveys, by 5-year age group and race categories, have been accepted as population parameters for our calculations. Given the complexity and detail of NHDS sampling strategy, we believe this to be a reasonable assumption. This assumption assumes no sampling variability for the expected event rates. The variability of our SIR estimates is therefore entirely derived from the variability of our local cohort, as captured by bootstrapping our sample. A similar bootstrap technique of the NHDS datasets was not possible. Therefore it is possible that our estimates of variability slightly underestimate the true variability in our calculations, and as a result the width of the CIs reported may be smaller than necessary to capture 95% confidence. However, we believe that any variability contributed by the NHDS sampling strategy would be slight when compared with the variability in follow-up in our local sample, and would not appreciably alter the width of the SIR CIs. Nevertheless, it is possible that if a bootstrap technique could be applied to the NHDS datasets, the CIs might widen enough that the finding of increased stroke incidence might lose statistical significance at the .05 level, although a highly suggestive trend would likely remain in any case.

In summary, this study suggests that American early-stage breast cancer patients treated with modern techniques of breast conserving surgery and RT at a single institution may have a higher risk of stroke than age- and race-matched controls. Treatment-related risk factors were not found to be significantly associated with the risk of stroke, after controlling for other risk factors, although a minor role cannot be excluded. Common underlying risk factors for breast cancer and stroke may also be important in explaining the finding of excess risk. Ultimately, even if in part treatment-related, the absolute magnitude of the increased risk of stroke suggested by this study is reassuringly small, particularly when compared to the risk of breast cancer recurrence in the absence of adjuvant treatment. Therefore, concerns about potential cerebrovascular toxicity do not appear to justify withholding appropriate treatments for breast cancer, including supraclavicular radiation therapy and tamoxifen, when indicated.

	Authors’ Disclosures of Potential Conflicts of Interest

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
The authors indicated no potential conflicts of interest.

Author Contributions TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES   Conception and design: Reshma Jagsi, Kent A. Griffith, Todd Koelling, Lori J. Pierce Financial support: Lori J. Pierce Administrative support: Rachel Roberts, Lori J. Pierce Provision of study materials or patients: Lori J. Pierce Collection and assembly of data: Reshma Jagsi, Rachel Roberts Data analysis and interpretation: Reshma Jagsi, Kent A. Griffith, Todd Koelling, Lori J. Pierce Manuscript writing: Reshma Jagsi Final approval of manuscript: Reshma Jagsi, Kent A. Griffith, Todd Koelling, Rachel Roberts, Lori J. Pierce

 

	NOTES

 
Presented in abstract form at the San Antonio Breast Cancer Symposium, San Antonio, TX, December 8-12, 2005.

The authors certify that this is an original unpublished work.

Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Authors' Disclosures of... Author Contributions REFERENCES

 
1. Early Breast Cancer Trialists’ Collaborative Group: Favourable and unfavourable effects on long-term survival of radiotherapy for early breast cancer: An overview of the randomised trials. Lancet 355:1757-1770, 2000[CrossRef][Medline]

2. Marks LB, Yu X, Prosnitz RG, et al: The incidence and functional consequences of RT-associated cardiac perfusion defects. Int J Radiat Oncol Biol Phys 63:214-223, 2005[CrossRef][Medline]

3. Giordano SH, Kuo YF, Freeman JL, et al: Risk of cardiac death after adjuvant radiotherapy for breast cancer. J Natl Cancer Inst 97:419-424, 2005[Abstract/Free Full Text]

4. Nilsson G, Holmberg L, Garmo H, et al: Increased incidence of stroke in women with breast cancer. Eur J Cancer 41:423-429, 2005[CrossRef][Medline]

5. Dorresteijn LD, Kappelle AC, Boogerd W, et al: Increased risk of ischemic stroke after radiotherapy on the neck in patients younger than 60 years. J Clin Oncol 20:282-288, 2002[Abstract/Free Full Text]

6. Haynes JC, Machtay M, Weber RS, et al: Relative risk of stroke in head and neck carcinoma patients treated with external cervical irradiation. Laryngoscope 112:1883-1887, 2002[CrossRef][Medline]

7. Cheng SW, Ting AC, Lam LK, et al: Carotid stenosis after radiotherapy for nasopharyngeal carcinoma. Arch Otolaryngol Head Neck Surg 126:517-521, 2000[Abstract/Free Full Text]

8. Pierce LJ, Griffith KA, Hayman JA, et al: Conservative surgery and radiotherapy for stage I/II breast cancer using lung density correction: 10-year and 15-year results. Int J Radiat Oncol Biol Phys 61:1317-1327, 2005[CrossRef][Medline]

9. National Center for Health Statistics: National Hospital Discharge Survey. http://www.cdc.gov/nchs/about/major/hdasd/nhdsdes.htm

10. United States Census: American Community Survey. http://www.census.gov

11. Efron B, Tibshirani RJ: Monographs on Statistics and Applied Probability 57: An Introduction To The Bootstrap. Boca Raton, FL, Chapman & Hall/CRC, 1998

12. Recht A, Edge SB, Solin LJ, et al: Postmastectomy radiotherapy: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 19:1539-1569, 2001[Abstract/Free Full Text]

13. Early Breast Cancer Trialists’ Collaborative Group: Effect of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: An overview of the randomised trials. Lancet 366:2087-2106, 2005[Medline]

14. Olivotto I, Chua B, Elliott E, et al: A clinical trial of breast radiation therapy versus breast plus regional radiation therapy in early-stage breast cancer: The MA20 trial. Clin Breast Cancer 5:361-363, 2003

15. European Organisation for Research and Treatment of Cancer (EORTC): Phase III randomized trial investigating the role of internal mammary and medial supraclavicular (IM-MS) lymph node chain irradiation in stage I-III breast cancer (Joint study of the EORTC Radiotherapy Cooperative Group and the EORTC Breast Cancer Cooperative Group EORTC 22922/10925). http://www.eortc.be/protoc/details.asp?protocol=22922

16. Braithwaite RS, Chlebowski RT, Lau J, et al: Meta-analysis of vascular and neoplastic events associated with tamoxifen. J Gen Intern Med 18:937-947, 2003[CrossRef][Medline]

17. Fisher B, Costantino JP, Wickerham DL, et al: Tamoxifen for prevention of breast cancer: Report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 90:1371-1388, 1998[Abstract/Free Full Text]

18. Rose DP, Komninou D, Stephenson GD: Obesity, adipocytokines, and insulin resistance in breast cancer. Obes Rev 5:153-165, 2004[CrossRef][Medline]

19. Gonullu G, Ersoy C, Ersoy A, et al: Relation between insulin resistance and serum concentrations of IL-6 and TNF-alpha in overweight or obese women with early stage breast cancer. Cytokine 31:264-269, 2005[CrossRef][Medline]

20. Stoll BA: Western nutrition and the insulin resistance syndrome: A link to breast cancer. Eur J Clin Nutr 53:83-87, 1999[CrossRef][Medline]

21. Ahmed A, Ness J, Howard G, et al: Cerebrovascular diseases as primary hospital discharge diagnoses: National trend (1970-2000) among older adults. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 60:1328-1332, 2005[Abstract/Free Full Text]

22. Shahar E, McGovern PG, Pankow JS, et al: Stroke rates during the 1980s. Stroke 28:275-279, 1997[Abstract/Free Full Text]

23. Fang J, Alderman MH: Trend of stroke hospitalization, United States, 1988-1997. Stroke 32:2221-2226, 2001[Abstract/Free Full Text]

24. Broderick JP, Phillips SJ, Whisnant JP, et al: Incidence rate of stroke in the eighties: The end of the decline of stroke? Stroke 20:577-582, 1989[Abstract/Free Full Text]

Submitted August 24, 2005; accepted March 28, 2006.

This article has been cited by other articles:

				M. Morrogh, P. I. Borgen, and T. A. King The Importance of Local Control in Early-Stage Breast Cancer: A Historical Review and a Discussion of Ongoing Issues Ann. Surg. Oncol., December 1, 2007; 14(12): 3310 - 3320. [Full Text] [PDF]

				M. J. Hooning, L. D.A. Dorresteijn, B. M.P. Aleman, A. C. Kappelle, J. G.M. Klijn, W. Boogerd, and F. E. van Leeuwen Decreased Risk of Stroke Among 10-Year Survivors of Breast Cancer J. Clin. Oncol., December 1, 2006; 24(34): 5388 - 5394. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jagsi, R. Articles by Pierce, L. J. Search for Related Content PubMed PubMed Citation Articles by Jagsi, R. Articles by Pierce, L. J.