Differential expression of prognostic biomarkers between interval and screen-detected breast cancers: does age or family history matter?

The aim of this study was to compare tumor expression of prognostic biomarkers between interval breast cancers and screen-detected breast cancers overall, and according to age at diagnosis and familial risk. Tissue micro-arrays were constructed from 98 breast cancers (47 interval and 51 screen-detected) diagnosed in women in the Cancer Genetics Network. Arrays were immuno-stained to compare protein expression of six biomarkers including estrogen and progesterone receptor (ER/PR), Her2/neu, EGFR, cytokeratin 5/6, and Ki67. Fisher's Exact test was used to compare expression between interval and screen-detected cancers. Interval cancers were larger (P = 0.04), higher stage (P < 0.001), and more likely to have lobular histology (P = 0.01) than screen-detected cancers. Overall, interval cancers more often overexpressed EGFR (P = 0.01) and were somewhat more likely to be ER- (55% vs. 43%, P = 0.3), and triple negative (ER-/PR-/Her2-) (21 vs. 12%, P = 0.26). A greater difference in the proportion of interval versus screen-detected tumors that were ER- (53 vs. 35%; P = 0.29), PR- (35 vs. 21%; P = 0.25) and EGFR+ (17 vs. 0%; P = 0.02) was evident among women over 50. There was a trend toward differential expression among women with familial risk for PR- (P = 0.005) and triple negative status (P = 0.02). This study provides new data indicating that EGFR may be important in the etiology of interval cancer and be a possible therapeutic target. Our data also suggest that biological differences between interval and screen-detected cancers are more defined in older women. Future studies to confirm this finding and to elucidate novel markers for characterizing interval cancers may be more beneficial to this subgroup.

Complementary approaches to assessing risk factors for interval breast cancer.

OBJECTIVE: To examine risk factors for interval breast cancer among women screened in a population-based mammography program. METHODS: Risk for interval cancer was assessed in terms of both the incidence per 10,000 negative screens and the proportion of all breast cancers diagnosed among screened women. Interval (N = 557) and screen-detected cancers (N = 1,545) were identified among 208,667 women receiving mammography in Colorado (1994-2001). Logistic regression was used to assess independent effects of multiple factors. RESULTS: Overall risk of interval cancer was 29.5/10,000 women screened. Incidence was higher in women >50 years (OR: 2.28, 1.86-2.80), with family history (OR: 2.23, 1.85-2.70), with dense breasts (OR: 3.84, 2.76-5.35), and using hormones (OR: 1.54, 1.20-1.97). Hispanics had lower incidence than Whites (OR: 0.52, 0.34-0.81). Interval cancers represented 26% of all cancers diagnosed. This proportion was higher in women <50 (OR: 1.41, 1.09-1.82) and in women with dense breasts (OR: 2.95, 1.94-4.48). CONCLUSIONS: Incidence of interval cancer increases with age, breast density, hormone use, and family history. Attempts to reduce occurrence of these cancers through more sensitive and/or intensive screening should focus on these subgroups. The disproportionate number of interval cancers associated with young age and dense breasts suggests these cancers result from both rapid growth and difficulties in detection.

Late-stage breast cancer among women with recent negative screening mammography: do clinical encounters offer opportunity for earlier detection?

BACKGROUND: Opportunities to prevent late-stage breast cancer within the course of usual care are needed. We evaluate whether clinical encounters offer such opportunities. METHODS: Within seven health care plans, we identified 1298 women aged more than 50 years with early (<3 cm), late-stage (> or = 3 cm), or metastatic invasive breast cancer diagnosed during 1995-1999, whose first screening mammogram 13-36 months prior to the diagnosis (index) was negative. We audited all care occurring in the health plans up to 36 months prior to the cancer diagnoses. Ordinal logistic regression compared the frequency of events by disease category. We hypothesized that during the 13-36 months prior to diagnosis, women with late-stage or metastatic breast cancer would have more symptoms and be more likely to have breast-related clinical visits but have less breast screening (clinical breast examination [CBE] or mammography) than women with early-stage disease, thereby indicating clinical opportunities for earlier detection. RESULTS: We found no differences in demographic characteristics across breast cancer stage among the 1298 women. Both before and after the negative index mammogram but during the 13-36 months prior to diagnosis, few women had breast symptoms (5% before index, 8% after), but many sought breast care (86% before index, 90% after) and screening CBE (62% before index, 43% after). Only the occurrence of screening CBE (odds ratio [OR] = 0.73, 95% confidence interval [CI] = 0.56 to 0.95) or screening mammograms (OR = 0.74, 95% CI = 0.57 to 0.97) after the negative index mammogram reduced odds of more severe disease at diagnosis. CONCLUSION: Although the mortality benefit of CBE, or one compared to two year mammography has not been established, we found that women with late-stage breast cancers undetected by screening mammography did not experience opportunities for earlier detection except through CBE or additional screening mammography.

Breast and cervical cancer screening: clinicians' views on health plan guidelines and implementation efforts.

BACKGROUND: Optimizing breast and cervical cancer screening rates within health plans requires clinician support for screening guidelines, an awareness of whether there are tools available and functioning to aid screening implementation, and a perception of collegial and leadership support for quality screening services. This study investigated clinicians' perceptions of guidelines, reminders for screening, and plan and practice commitment in order to assess where opportunities exist to improve the screening process. METHODS: A stratified sample of 761 primary care clinicians from three comprehensive health plans were surveyed to assess awareness of and agreement with guideline elements, perception of guidelines' usefulness, awareness of plan strategies to promote guideline adherence, perception of support for high-quality screening services, and ratings of plan efforts to maximize members' access. RESULTS: Clinician awareness of and agreement with guideline elements was high (98% breast, 94% cervical). Across guideline elements, agreement was lower for mammography than cervical screening, notably for upper age limit recommendations (58% breast, 79% cervical). Knowledge of systems that cue patients and clinicians that screening is due varied by cancer test, and clinician report and plan report data about the existence of systems were, at times, not congruent. Views about consistent operation of systems differed by test (mammograms, 74%-92%; Pap, 66%-84%). Clinicians rated local colleagues and local and plan medical leadership as very committed to high-quality screening, albeit with somewhat lower ratings for cervical testing. Although the majority rated overall plan efforts to maximize screening as very good or excellent, perceived consistency of systems to cue a woman that she is due for testing and perception of collegial support were independently and significantly related to ratings of plan efforts. CONCLUSIONS: Improvements in knowledge of systems that support guideline implementation varied, and action to ensure accurate perception of reminders, as well as consistent implementation of systems, may be important for improving screening rates and outcomes. Plan efforts and clinician efforts at the practice level are closely linked and need to be aligned to maximize screening rates. This requires plan and practice-level analyses of structures and processes that could be improved.

Characteristics of women refusing follow-up for tests or symptoms suggestive of breast cancer.

BACKGROUND: Delay in diagnosis of breast cancer can occur at several points on the diagnostic pathway. We examined characteristics of women with breast cancer who before diagnosis actively refused recommended follow-up of tests or symptoms suggestive of breast cancer. METHODS: We identified women aged 50 years or older diagnosed with late-stage (metastatic disease or tumors > or = 3 cm at diagnosis) and a matched sample of women with early-stage (tumors < 3 cm) breast cancer from 1995 to 1999. Using medical records, we investigated clinical characteristics, use of health care, and documentation of care refusal during the 3 years before diagnosis. We used logistic regression models to compare refusers to nonrefusers. RESULTS: Of the 2694 women studied, 7.2% refused provider follow-up advice during the 3 years. These women were more likely to have late-stage breast cancer at diagnosis than were nonrefusers (odds ratio [OR] = 1.9, 95% confidence interval [CI] = 1.4 to 2.6). They were more likely to be aged 75 years or older (OR = 1.9, 95% CI = 1.4 to 2.7 compared with age 50-64) or to have six or more children (OR = 2.3, 95% CI = 1.3 to 4.2 compared to women with one to two children). Clinical factors associated with refusal included low use of mammography, high use of clinical breast exam, and missed appointments. A minority of women who refused had a reason documented in the medical record; the most frequent reasons were avoidance-denial-fatalism, fear of diagnostic tests, and fear of surgery or disfigurement. CONCLUSIONS: Our results suggest that certain demographic and clinical characteristics are associated with women's refusal of diagnostic testing for breast cancer. Further study is needed on refusers' characteristics and on how such refusals affect outcomes. Efforts aimed at identifying and counseling women with abnormal results who refuse follow-up are warranted.

Building a research consortium of large health systems: the Cancer Research Network.

Critical questions about cancer prevention, care, and outcomes increasingly require research involving large patient populations and their care delivery organizations. The Cancer Research Network (CRN) includes 11 integrated health systems funded by the National Cancer Institute (NCI) to conduct collaborative cancer research. This article describes the challenges of constructing a productive consortium of large health systems, and explores the CRN's responses. The CRN was initially funded through an NCI cooperative agreement in 1999 and has since received a second 4-year grant. Leadership and policy development are provided through a steering committee, subcommittees, and an external advisory committee. The CRN includes integral and affiliated research projects supported by a Scientific and Data Resources Core. Three characteristics of the CRN intensified the general challenges of consortium research: 1) its members are large health systems with legitimate concerns about confidentiality of data about enrollees, providers, and the organization; 2) CRN research projects often generate highly sensitive data about quality of care; and therefore 3) each participating organization wants a strong voice in CRN direction. CRN experience to date confirms that a consortium of health systems with internal research capacity can address a range of important cancer research questions that would be difficult to study in other venues. The advantages and challenges of consortium research are explored, with suggestions for the development, execution, and management of multisystem population laboratories.

Assessing controversial direct-to-consumer advertising for hereditary breast cancer testing: reactions from women and their physicians in a managed care organization.

OBJECTIVE: To describe the impact on patients and physicians at a managed care organization (MCO) of a direct-to-consumer advertising (DTC-ad) campaign concerning testing for the BRCA1 and BRCA2 genes. STUDY DESIGN: Observational study. METHODS: In 2003, we mailed a 30-item questionnaire to 750 randomly chosen female members of Kaiser Permanente Colorado (KPCO) aged 25 to 54 years, and 100 female KPCO members with a history of breast cancer genetic referral. We mailed a 7-item questionnaire to 180 randomly chosen KPCO primary care providers. RESULTS: Of 394 patient respondents, 245 (62%) reported exposure to the DTC-ad of whom 63% reported that the DTC-ad caused no anxiety at all. A high level of perceived breast cancer risk and being of Hispanic ethnicity each were independently associated with reported anxiety due to the DTC-ad (adjusted odds ratio [OR] = 3.23, 95% confidence interval [CI] = 1.35, 7.73, and adjusted OR = 4.19, 95% CI = 1.48, 11.83, respectively). Greater knowledge was seen among respondents exposed to the DTC-ad than among those reporting no exposure (P = .015). Of the physician respondents, 84% reported that the DTC-ad caused no strain on the doctor-patient relationship, and nearly 80% reported no effect on daily clinical practice. Genetic referrals soared more than 200% compared with the prior year, when there was no advertising. CONCLUSION: The DTC-ad had a marked impact on genetic services, but little apparent negative impact on patients or primary care providers at an MCO.

Organizational barriers to physician participation in cancer clinical trials.

OBJECTIVE: To assess barriers to physician participation in cancer clinical trials among oncologists, oncology leaders, and health plan leaders. STUDY DESIGN: Mail survey of 221 oncologists combined with semistructured telephone interviews with oncology and plan leaders at 10 integrated healthcare systems. METHODS: The survey instrument examined physicians' involvement in clinical trials; their perception of the value of trials to them, their patients, and their organization; and the presence of infrastructure support for trials and associated resource constraints. The interviews investigated similar issues from the leaders' perspective. We used linear regression to model trial enrollment and standard qualitative techniques to analyze the interviews. RESULTS: Oncologists estimated they enrolled 7% of patients in trials. They expressed extremely favorable attitudes toward trials as a source of high-quality patient care and a benefit to themselves professionally. While positive attitudes toward trials were common, and were significant bivariate predictors of enrollment, organizational factors were the predominant predictors in multivariate analysis. The best combination of factors independently predicting enrollment related to organizational support for trials, subspecialty of the oncologist, and limitations of trial eligibility requirements. CONCLUSIONS: To increase trial participation, there is a critical need for infrastructure to support trials, especially additional support staff and research nurses. In addition, there is a need for better intra-organizational communication and consideration of the impact of trial design on internal health plan resources. This research supports the need to continue a national dialogue about the broadly defined benefits and costs of clinical trials to patients, physicians, and health plans.

Recommendations for care related to follow-up of abnormal cancer screening tests: accuracy of patient report.

When using patient self-report of processes of care as part of measuring quality performance, validity and reliability are important considerations. In this study, the congruence of patient report of recommendations of screening follow-up care was compared with record audit data. Survey data were collected from a random sample of patients with abnormal breast (n = 230) or cervical (n = 219) cancer screening tests from four health centers. With patient consent, record audits were conducted to validate self-report. Measures of congruence for recommendations for follow-up were calculated along with sensitivity and specificity for procedure-specific recommendations. Overall congruence was higher in the mammography sample (81%) as compared to the Pap sample (61%). Predictors of overall congruence for the abnormal Pap test sample included health plan, self-reported health status, and test result. There were no significant predictors of congruence for the abnormal mammogram test sample from the potential variables collected. Raw agreement rates support using patient self-report in assessing abnormal test follow-up recommendations.

The sensitivity of Medicare billing claims data for monitoring mammography use by elderly women.

Mammography use is monitored through Medicare billing claims; however, the sensitivity of this data source has not been previously described. This study included 10,852 Colorado women ages 65 and older with a mammogram in 1998 as registered by the Colorado Mammography Project who were Medicare fee-for-service (FFS) enrollees. These records were matched to Medicare billing data to assess the proportion of those mammograms submitted for payment to Medicare. The overall sensitivity of the FFS Medicare billing data for screening mammography was 85 percent. Medicare billing claims were less sensitive for younger women, African Americans, women with some college education, and women with supplementary private insurance. In Colorado, the Medicare FFS billing claims understates mammography usage by 15 percent. Care must be taken when comparing mammography use derived from Medicare billing claims, as the sensitivity of billing data can vary substantially by age, race, and socioeconomic status.

Implementation of systems strategies for breast and cervical cancer screening services in health maintenance organizations.

OBJECTIVE: To describe the systems strategies used to reduce failures in delivery of breast and cervical cancer screening services in HMOs with high performance rates for these services. STUDY DESIGN: Multiple case study. PARTICIPANTS AND METHODS: Seven HMOs participated in an assessment of their breast and cervical cancer screening policies and procedures. Current clinical practice guidelines were analyzed, and key informants were interviewed about organizational policies and procedures that ensure initial screening and follow-up of abnormal results. Data were analyzed across plans for several theoretically relevant domains, including leadership and policies, clinical decision support, delivery system design, clinical information systems, and patient self-management support. RESULTS: Practice guidelines were fundamentally similar across plans for both cancer screenings, although operationalization of risk and formatting of the written documents differed. These plans adopted a wide array of strategies, particularly in the clinical decision support, clinical information systems, and patient self-management support domains, but there is room for improvement. Differences among plans and between strategies for breast and cervical cancer screening provide new understanding of how to approach this problem. CONCLUSIONS: Organizations seeking to improve performance of breast and cervical cancer screening should consider multiple strategies aimed at multiple targets and should ensure that strategies used for one type of cancer are considered for others.

Caffeine, cajoling, and other strategies to maximize clinician survey response rates.

An ongoing objective in health services research is to increase response rates to clinician surveys to ensure generalizability of findings. Three HMOs in the Cancer Research Network participated in a primary care clinician survey to better understand organizational characteristics affecting adoption and implementation of breast and cervical cancer screening guidelines. A four-stage data collection strategy was implemented to maximize response. This included careful attention to survey design and layout, extensive piloting, choice of token incentive, use of "local champions," and denominator management. An overall response rate of 91% was attained, ranging from 83 to 100% among the plans (N = 621). Although the response rate after the second stage of data collection met commonly used standards, the authors argue for the four-stage method due to the possibility of differences when comparing early and late responders. This is important when multiple plans with differing structure and internal characteristics are surveyed.

Population-based estimate of the contribution of TP53 mutations to subgroups of early-onset breast cancer: Australian Breast Cancer Family Study.

Although germline TP53 mutations have been identified in women with breast cancer from families meeting Li-Fraumeni criteria, their contribution to breast cancer per se is not well known, but is thought to be minimal. We aimed to determine the prevalence of germline TP53 mutations in subgroups of early-onset breast cancer. Germline TP53 mutation status was assessed by DNA sequencing, screening for heterozygous single-nucleotide polymorphisms, and Multiplex Ligation-Dependent Probe Amplification analyses. From an Australian population-based series of invasive breast cancers, we studied (a) 52 women diagnosed before age 30 years unselected for family history [very early-onset (VEO)] and (b) 42 women diagnosed in their 30s with two or more first- or second-degree relatives with breast or ovarian cancer [early-onset family history (EO-FH)]. Of the VEO group, two (4%) had a mutation: G13203A (exon 6 missense) in a 24-year-old and a large 5,338-bp genomic deletion in a 26-year-old. Neither had a family cancer history that met Li-Fraumeni criteria. Of the EO-FH group, three (7%) had a mutation: T13240G (a known intron 5 splicing mutation) in a 36-year-old from a classic Li-Fraumeni family; G12299A (exon 4 missense) in a 33-year-old from a Li-Fraumeni-like family; and 14058delG (exon 7 frame-shift) in a 39-year-old with a family cancer history that did not meet Li-Fraumeni criteria. Germline TP53 mutations play a larger role in early-onset breast cancer than previously thought, and in this context, can be evident outside clinically defined Li-Fraumeni families.

Increasing utilization of cancer genetic counseling services using a patient navigator model.

Rarely has utilization of genetic counseling for Hereditary Breast and Ovarian Cancer (HBOC) been studied separately from utilization of testing. At Kaiser Permanente Colorado, consistently only 30% of all members referred for HBOC attend genetic counseling. To increase the volume of genetic counseling appointments, a patient navigator approach was pilot tested in a randomized-controlled trial over 3 months. A total of 125 members were referred for HBOC genetic counseling (55 randomized to PN, 70 randomized to usual care). Utilization of referrals for Navigator-assisted members was 44%, compared to 31% in the usual care arm (p=0.16). The patient navigator significantly decreased time to appointment, with over 80% of Navigator-assisted members seen for genetic counseling less than three months from referral date, compared to 32% in usual care (p=0.002). patient navigator assistance shortens time from referral to appointment for HBOC genetic counseling, and may increase utilization of such services.

Does having insurance affect differences in clinical presentation between Hispanic and non-Hispanic white women with breast cancer?

BACKGROUND: Hispanic women with breast cancer present differently than do non-Hispanic white (NHW) women. Lack of access to care has been offered as an explanation for these differences. In this study breast cancer presentation was examined in Hispanic women in a comprehensive, equal-access health care system. METHODS: Hispanic and NHW breast cancer cases registered between 1995 and 2004 in the Kaiser Permanente of Colorado Tumor Registry were compared by age at diagnosis, stage, tumor grade, size, and receptor status. Multivariate logistic regression was performed to generate age-adjusted odds ratios by ethnicity and each tumor characteristic. RESULTS: A total of 139 Hispanic women and 2118 NHW women with breast cancer were identified. Hispanic women had a mean average age at diagnosis of 56 years compared with 61 years for NHW women (P < .0001). Use of mammographic screening services in the prior 2 years was similar by ethnicity. Relative to NHW women, Hispanic women had age-adjusted odds ratios of 2.70 (95% confidence interval [CI]: 1.26-5.77) for having stage IV disease, 2.25 (95% CI: 1.39-3.67) for having poorly differentiated tumors, 2.16 (95% CI: 1.26-3.69) for having a tumor greater than 5 cm, and 1.88 (95% CI: 1.24-2.81) for having estrogen receptor-negative tumors. CONCLUSIONS: Despite equal access to health care services, differences persist in the size, stage, and grade of breast cancer for Hispanic women compared with NHW women. The results of the study suggest a biologic/genetic basis for these differences.

Genetic testing for susceptibility to breast and ovarian cancer: evaluating the impact of a direct-to-consumer marketing campaign on physicians' knowledge and practices.

PURPOSE: To assess the impact of direct-to-consumer marketing of genetic testing for risk of breast and ovarian cancer by a biotechnology company on: 1) physicians' knowledge; 2) reasons given when asking questions about the test; and 3) physicians' practice patterns in two pilot cities where the campaign took place and two control cities. METHODS: Survey of randomly selected family physicians, internists, obstetrician-gynecologists, and oncologists from May 1-May 21, 2003. RESULTS: Physicians' knowledge did not differ between pilot and control cities. Significant differences (pilot versus control cities) were seen in the reasons patients gave for asking questions about testing. More physicians in pilot cities (14%) than control cities (7%) reported an increase in the number of times they ordered genetic testing for breast and ovarian cancer risk in the previous 6 months (adjusted odds ratio 1.9, 95% confidence interval, 1.2-3.1). Awareness of professional guidelines and being in a practice with a policy on genetic testing for risk of breast and ovarian cancer were associated with physicians' behaviors and interest among patients in testing. CONCLUSIONS: Given the complexity and limitations of genetic testing for risk of breast and ovarian cancer, the development and broad dissemination of clinical guidelines and education of physicians are needed.

Impact of direct-to-consumer advertising for hereditary breast cancer testing on genetic services at a managed care organization: a naturally-occurring experiment.

PURPOSE: To describe the impact of Myriad Genetics, Inc.'s direct-to-consumer advertising (DTC-ad) campaign on cancer genetic services within two Managed Care Organizations, Kaiser Permanente Colorado (KPCO), Denver, Colorado, where the ad campaign occurred, and Henry Ford Health System (HFHS), Detroit, Michigan, where there were no advertisements. METHODS: The main outcome measures were the changes in number and pretest mutation probability of referrals approved for cancer genetic services at KPCO and HFHS during the campaign versus the year prior, and mutation probability of those undergoing testing. RESULTS: At KPCO, referrals increased 244% during the DTC-ad compared to the same time period a year earlier (P value<0.001). The proportion of referrals at high pretest probability of a mutation (10% or greater) dropped from 69% the previous year to 48% during the campaign (P value<0.001). There was no significant change in pretest mutation probability among women who underwent testing between the two time periods. HFHS reported no significant change between the two time periods for numbers or mutation probability of referrals, or for mutation probability of women tested. CONCLUSION: The DTC-ad caused significant increase in demand for cancer genetic services. In the face of potential future DTC-ad for inherited cancer risk, providers and payers need to consider the delivery of genetic services and genetic education for persons of all risk levels.

Cervical cancer in women with comprehensive health care access: attributable factors in the screening process.

BACKGROUND: Invasive cervical cancer is highly preventable, yet it continues to occur, even among women who have access to cancer screening and treatment services. To reduce cervical cancer among such women, reasons for its occurrence must be better understood. We examined factors associated with the diagnosis of cervical cancer among women enrolled in health plans. METHODS: We identified all cases of invasive cervical cancer (n = 833) diagnosed from January 1, 1995, through December 31, 2000, among women who were long-term members of seven prepaid comprehensive health plans and reviewed each woman's medical records for the 3 years prior to her cancer diagnosis. Women were classified into one of three categories based on Pap test histories 4-36 months before diagnosis: failure to screen with a Pap test, failure in detection by a Pap test, or failure in follow-up of an abnormal test result. RESULTS: The majority of cases (n = 464; 56%) were in women who had no Pap tests during the period 4-36 months prior to diagnosis. Of the remaining cases, 263 (32%) were attributed to Pap test detection failure and 106 (13%) to follow-up failure. Being older (odds ratio [OR] = 6.48, 95% confidence interval [CI] = 3.89 to 10.79) or living in an area of higher poverty (OR = 1.72, 95% CI = 1.11 to 2.67) or having a lower education level (OR= 1.52; 95% CI = 1.07 to 2.16) was associated with the likelihood of being assigned to the failure to screen category versus either of the other two categories. A total of 375 (81%) of the 464 patients who had not had Pap screening had had at least one outpatient visit 4-36 months prior to cancer diagnosis. The cancer diagnostic process was triggered by a routine screening examination in 44% of patients, whereas 53% of the patients presented with symptoms consistent with cervical cancer; the remaining 3% were identified fortuitously during the course of receiving noncervical care. CONCLUSIONS: To reduce the incidence of invasive cervical cancer among women with access to screening and treatment, Pap screening adherence should be increased. In addition, strategies to improve the accuracy of Pap screening could afford earlier detection of cervical cancer.

Mammography screening matters for young women with breast carcinoma: evidence of downstaging among 42-49-year-old women with a history of previous mammography screening.

BACKGROUND: Breast carcinoma is the leading cause of death for women between the ages of 40 and 49 years, yet questions linger regarding the effectiveness of screening mammography in reducing mortality rates among women of this age. In the current study, the authors report on the association between cancer stage at diagnosis and a history of mammography screening in a clinical setting that has emphasized informed choice regarding mammography screening for women in this age group. METHODS: Previous mammographic screening for 247 breast cancer patients 42-49 years of age who were diagnosed at Kaiser Permanente Colorado during 1994-2000 was evaluated relative to cancer stage. Cancer stage was dichotomized into early (American Joint Committee on Cancer [AJCC] Stages 0 and I) and late (AJCC Stages II-IV) and previous screening was defined as at least one normal screening mammogram within 24 months before the breast carcinoma diagnosis. RESULTS: Women who were screened were less likely to be diagnosed at a late stage than were women who were not screened (40% vs. 52% late stage, respectively). Adjusted for age, year of diagnosis, and family history, screened women were 0.56 (95% confidence interval = 0.32-0.97) times as likely as unscreened women to be diagnosed at a late stage. CONCLUSIONS: Women 42-49 of years with breast carcinoma who undergo regular screening mammography have a more favorable cancer stage than do women with breast carcinoma who do not undergo regular screening. This downstaging of breast carcinoma is likely to translate into improved breast carcinoma survival resulting from screening mammography for women 40-49 years of age.