Background
More than 300,000 breast cancers are diagnosed in the United States annually.1 Ductal carcinoma in situ (DCIS) accounts for 25% of newly diagnosed breast cancers and is treated surgically with additional radiation and hormone therapy depending on individualized risk assessment or patient preference.2 Survival rates of women with DCIS are estimated at >97%.3,4 However, depending on treatment, ≥20% of women with DCIS develop a second ipsilateral breast tumor within 10 years of initial diagnosis, half of which are invasive.2,5 Additionally, women treated for DCIS have an increased risk of contralateral invasive breast cancer, estimated at 6.2% over 20 years of follow-up.4,6,7
To address risk of recurrence and second cancers among patients with DCIS, the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Breast Cancer8 recommend DCIS surveillance with interval history and physical examination every 6 to 12 months for 5 years and then annually, with mammogram every 12 months. This recommendation is based on expert opinion, and there is no recommendation for patient self-examination as current standard of practice. In addition, older studies of invasive breast cancer show that clinical breast examinations (CBEs) only detect 15% of recurrences, with no contemporary large series describing modality of detection of recurrent or second breast cancer in DCIS survivors.9–11 More evidence is needed to better understand how second breast cancers are diagnosed in clinical practice and to refine the NCCN Guidelines on DCIS survivorship given the growing population of cancer survivors, the strain on the oncology workforce, and the greater adoption/increased utilization of telehealth in providing breast cancer care.12,13
To address the gap between guidelines and evidence, we performed a retrospective cohort study of patients undergoing surveillance following diagnosis of DCIS within a large integrated health care system to determine the method of detection of second breast cancers.
Methods
Patient Population and Setting
Patients included in this study were members of Kaiser Permanente Northern California, an integrated, prepaid, nonprofit health care delivery system that provides comprehensive medical services to 30% of the insured population in the region. Integrated systems are ideal for survivorship studies, because long-term follow-up is possible, patients are highly representative of the local population, and their care is documented in a single electronic health record (EHR). Kaiser’s Institutional Review Board approved the study and granted permission for a waiver of consent for study participants as allowed under the Common Rule.
Study Design
We conducted a retrospective cohort study of self-reported female patients aged ≥18 years who were diagnosed and treated for unilateral DCIS between January 1, 2008, and January 1, 2011, and followed them for at least 10 years, through January 1, 2021. Patients were included if they had at least 5 years of follow-up, with 62% (962/1,550) of the cohort having at least 10 years of follow-up. Patients with a concurrent diagnosis of invasive breast cancer, with a personal history of breast or another cancer (other than nonmelanoma skin cancer), or who completed initial treatment outside of our institution were excluded. We also excluded patients diagnosed with a second breast cancer within 6 months of the index lesion. Patients who developed DCIS and/or invasive cancer in either breast, the chest wall, axillary nodes, or a distant site were identified during the follow-up period, and the term “second breast cancer” was used to define this subset.
Data Sources
Patients diagnosed with a first primary DCIS were identified in our institution’s cancer registry, which provides data to the NCI’s SEER program and the statewide cancer registry. The institution-specific data elements and quality control procedures are identical to these registries.
Demographic data and medical comorbidities were collected from the EHR databases. Our institutional cancer registry was used to identify the date of diagnosis, laterality, tumor grade, and estrogen receptor (ER) status of the primary DCIS tumor. The Breast Cancer Tracking System (BCTS; an integrated breast cancer reporting system) was used to confirm the method of detection of the initial DCIS and patients with genetic mutations. Information on treatment for the initial DCIS diagnosis (lumpectomy, unilateral or bilateral mastectomy, sentinel lymph node biopsy, and radiation and/or endocrine therapy) was collected from both the BCTS and the EHR.
Patients who developed a second breast cancer during follow-up were identified from the EHR using diagnosis codes, the Northern California Cancer Registry, and the BCTS. For these patients, individual chart reviews were performed to document the site of the second cancer (locoregional breast or node, distant metastases), the type, and the method of detection. The method of detection was classified as either patient-detected (via palpable mass, nipple changes, skin changes, or other; for each case, we verified that a patient reported a specific symptom at the time of their visit and/or that laterality of the symptom correlated with imaging findings whether or not it was confirmed on physical examination), physician-detected (by primary care, surgical oncologist, survivorship provider, or medical oncologist), or imaging-detected (mammogram, MRI, ultrasound, or other imaging modality). Patients’ vital status (living, deceased, or unknown) was collected from the cancer registry, and an individual chart review was performed to determine the cause of death if applicable.
Lastly, we quantified the number of mammograms and clinic visits during years 1 through 6 following DCIS treatment to characterize compliance with NCCN-recommended surveillance.8
Statistical Analysis
Our primary outcome was the diagnosis and detection method of a second breast cancer within 10 years of the initial DCIS diagnosis. Descriptive statistics (frequencies, medians, quartiles, ranges, and proportions) were used for demographic, clinical characteristics, treatment methods, and follow-up of mammograms and clinic visits. To provide more detail regarding patients who did and did not develop a second cancer, we also reported data specifically for these patients. All statistical analysis was performed using SAS 9.4 (SAS Institute Inc.).
Results
We identified 1,550 eligible patients, with a median age of 59 years at initial DCIS diagnosis (range, 32–92 years) (Table 1). All patients had at least 5 years of follow-up, and 62.1% (n=962) had ≥10 years; the percentage of patients with 10 years of follow-up was similar between those who did and did not develop a second cancer (62.6% vs 62.0%, respectively).
Distribution of Patient Characteristics
During follow-up, 11.5% (n=179) of women developed a second breast cancer. Those who developed a second cancer were more commonly overweight (34.1% vs 31.6%) or obese (39.1% vs 31.7%) and less commonly normal weight (25.7% vs 36.0%). Additionally, those who did versus did not develop a second cancer were more commonly Black (14% vs 7%) (Table 1).
Additionally during the follow-up period, 11.3% (n=175) of patients completed genetic testing, comprising 89.4% (n=160) of patients who developed a second breast cancer compared with 1.1% (n=15) of patients who did not. Pathologic variants were identified in 3.4% (n=6) of patients who developed a second cancer and in 0.6% (n=8) of those who did not (Table 1).
Initial DCIS
Initial DCIS was detected on a screening mammogram in 82.6% (n=1,280), on a diagnostic mammogram in 15.8% (n=245), and via “other” methods in 1.6% (n=25) (Table 2). The initial tumor grade was low/intermediate in 60.2% (n=934) and high in 39.7% (n=616). ER status of initial DCIS was known in 98.0% (n=1,519) of patients. ER positivity was similar between those who did (80.4%; n=144) and did not (81.6%; n=1,119) develop a second breast cancer (Table 2).
Initial DCIS Tumor Characteristics and Treatment Received
Surgical management for initial DCIS (with or without concomitant sentinel lymph node biopsy) included lumpectomy in 75.0% (n=1,162), unilateral mastectomy in 21.1% (n=327), and bilateral mastectomy in 3.9% (n=61) (Table 2). Women who developed a second breast cancer were more likely to have undergone lumpectomy (82.3% vs 74.0%) and less likely to have undergone unilateral mastectomy (15.6% vs 21.8%) or bilateral mastectomy (1.7% vs 4.2%). Those who developed a second cancer were less likely to have received endocrine therapy (17.3% vs 29.7%). They were also less likely to receive radiation therapy (40.2% vs 44.9%) (Table 2).
Surveillance After DCIS
Patients completed a median number of 5 mammograms during years 1 through 6 of surveillance, and 92.5% of patients completed ≥4 mammograms in this time. After initial DCIS diagnosis, patients had a median number of 7 clinic visits with providers likely to complete a clinical examination (ie, medical oncology and obstetrics/gynecology) (data not shown).
Second Breast Cancers
We identified 179 (11.5%) second breast cancers diagnosed during the study surveillance period, with a median time to diagnosis of 57 months (IQR, 34–84) (Table 3).
Second Breast Cancer Characteristics (N=179)
The second breast cancers were ipsilateral in 43.0% (n=77), contralateral in 54.8% (n=98), and presented with distant metastases in 2.2% (n=4) of patients. Patients presented with locoregional disease in 96.6% (n=173) of cases, with locoregional with distant metastasis in 1.1% (n=2), and with distant metastases only in 2.2% (n=4). Histologically, 61.5% (n=110) of second breast cancers were invasive, 36.3% (n=65) were DCIS, and 2.2% (n=4) were Paget’s disease (Table 3).
Second breast cancers were detected via imaging in 74.3% (n=133) of cases. Of those detected via imaging, 99.2% (n=132) were detected on mammogram and 0.6% (n=1) by breast MRI (this patient had a mammogram within 1 year of second cancer diagnosis) (Table 4).
Mode of Detection of Second Breast Cancers (n=179)
A total of 36 (20.1%) second breast cancers were patient-detected. The presenting symptom was palpable mass in 52.8% (n=19), nipple changes in 25% (n=9), skin changes in 8.3% (n=3), and “other” symptoms in 13.9% (n=5), which were determined to be pain related to metastatic disease (Table 4). Of these 36 patients, 23 had a surveillance mammogram within 13 months of second cancer diagnosis; 8 had a mammogram within 2 years of second cancer diagnosis; 2 went ≥3 years without mammography; 1 declined surveillance mammography; 1 had undergone bilateral mastectomy, and therefore mammogram surveillance was not indicated; and 1 had not had a mammogram before the second cancer diagnosis (data not shown).
A total of 2.2% (n=4) of second breast cancers were physician-detected, with 1 case detected by each of the following specialists: adult/family medicine, surgical oncology, oncology survivorship, and gynecology (Table 4). Three of these patients had a mammogram performed within 1 year of diagnosis, and the fourth had not yet had a surveillance mammogram since initial DCIS treatment.
Lastly, 3.4% (n=6) of second breast cancers were detected incidentally on imaging or pathology from procedures performed by plastic surgery unrelated to oncologic care (Table 4). Five of these patients had completed a mammogram within 1 year of diagnosis, and 1 patient had undergone bilateral mastectomy and so was not completing surveillance mammograms.
Among women who developed a second breast cancer, there was a similar proportion of ipsilateral and contralateral second cancers in those treated with lumpectomy (48.0% vs 50.4%, respectively), lumpectomy with radiation (48.0% vs 52.0%), or bilateral mastectomy (33.3% vs 33.3%) (supplemental eTable 1, available with this article at JNCCN.org). A higher proportion of those treated initially with unilateral mastectomy developed a contralateral versus ipsilateral second cancer (75.0% vs 14.3%). Additionally, a higher proportion of those with a contralateral versus ipsilateral second breast cancer received endocrine therapy as part of their initial DCIS treatment (61.3% vs 38.7%). Of the contralateral second breast cancers, a higher proportion of cases were detected via imaging (57.9% vs 42.1%) or clinical examination (75.0% vs 25.0%). A similar proportion of ipsilateral and contralateral second cancers were patient-detected (44.4% and 38.9%, respectively).
Vital Status
Of the 179 women who developed a second breast cancer, 155 (86.7%) were alive at the end of the study period, 4 (2.2%) died of causes related to breast cancer, and 20 (11.2%) died of other causes (Table 3).
Discussion
In our contemporary cohort of 1,550 patients undergoing surveillance following treatment of DCIS, we found that CBE detected only 2% of second breast cancers, which is fewer than the number of cases detected incidentally from plastic surgery procedures unrelated to oncologic surveillance. The remaining 94.4% of second breast cancers were detected by imaging or by patient-reported symptoms. Although some studies have evaluated specific modalities of surveillance for second breast cancers, ours is one of the first to examine multiple surveillance methods (imaging, patient-reported, and physical examination).14
Although our study is unique in its focus on patients with DCIS, our findings are in agreement with prior studies of invasive breast cancer, showing that the proportion of relapses detected on CBE has decreased over time, from approximately 46% of cases prior to 2000 to <15% in more modern cohorts.9–11,15 In fact, some studies limit the value of CBE to being a safeguard against false-negatives, because screening mammography can miss up to 10% to 15% of palpable lesions.16–18 A significant finding in our study was that 20.1% of second breast cancers were detected by patient self-report, which is consistent with previous studies primarily in invasive disease. A 2007 meta-analysis found that 30% to 40% of locoregional relapses and new contralateral breast cancers were detected by breast self-examination.9 Similarly, a single-center retrospective review of second breast events following stage I or II breast cancer found that 40% of second breast malignancies were detected via patient report.19 Although the US Preventive Services Task Force recommends against breast self-examination for primary breast cancer screening, this recommendation is largely based on findings from a single randomized controlled trial that excluded women with a history of breast cancer.20,21 Additionally, most women in our study with a self-detected second breast cancer were compliant with mammogram screenings. Our findings suggest that posttreatment surveillance should focus on compliance with mammography and patient education regarding symptoms of a second breast cancer, including palpable masses, skin changes, and nipple discharge. Patients should also be reminded about the risk of second breast cancer involving the ipsilateral and contralateral breast, axillary lymph nodes, and chest wall.
Our finding that CBE is low yield for diagnosing second breast cancers in this population has important implications for in-person health care utilization and offers an opportunity to harness telehealth as a cost-effective and time-effective option for DCIS survivorship care. Telehealth rapidly expanded during the COVID-19 pandemic and is a sustainable model providing equitable oncologic patient care.22,23 Furthermore, telehealth is a tool that can help alleviate the shortages and geographic discrepancies faced by the oncology workforce across the United States.24,25 Providers could still use the telehealth platform to address patients’ social and emotional well-being, sexual health and fertility concerns, practical concerns regarding survivorship care, health promotion, comorbidity management, and care coordination. Additionally, although our findings show that screening physical examination at survivorship visits has limited utility for detecting second cancers, patients will still have touch points with the health care system (eg, to complete surveillance mammography), and it is reasonable for a physical examination to be performed if patients report symptoms or there is discordance between symptom and imaging findings.
We found that 3.4% of second cancers were detected incidentally on imaging or surgical pathology obtained from plastic surgery procedures. Prior studies show that between 0.06% and 5.45% of specimens obtained from breast reduction procedures lead to the identification of precursor and/or malignant lesions.26–28 Although unrelated to DCIS surveillance, this emphasizes the importance of specimen orientation, thorough pathologic review of submitted specimens, and ensuring that patients’ breast cancer screening is up to date.
In our study population, second breast cancers were ipsilateral in 43.0%, were contralateral in 54.8%, and presented with distant metastases in 2.2% (n=4) of patients. The reported ratios of ipsilateral versus contralateral second breast cancers in women treated for DCIS vary by treatment and were likely also influenced by the distribution of other characteristics of the cohort, such as age and family history.29–31 Although it is possible that we missed some second cancers, we do not think this influenced our overall finding that few second cancers are detected by clinical examination. Additionally, estimates are also impacted by the evolution of treatment for DCIS, including standardization of treatment protocols, focus on tumor grade when developing a treatment plan, accessibility of genetic counseling, more accurate pathologic review, and an emphasis on margins at the time of surgical resection.
The strengths of our study include the ability to follow a large cohort of DCIS survivors over a long period of time within an integrated health system. We captured an insured patient population with access to health care resources. The generalizability of data from our institution has been evaluated previously and demonstrated that our patients are similar to the background population.32 We were able to capture a spectrum of follow-up methods, including examinations with multiple types of providers, various imaging modalities, and specific patient-reported symptoms. Unlike prior studies that included all patients with early-stage breast cancer, we were able to evaluate patients without invasive breast cancer at diagnosis and provide more robust evidence for this specific patient population. The generalizability of data from our institution has been evaluated previously and demonstrated that our patients are similar to the background population.32 Additionally, data from the CanBrTest (CDC) reported mammogram compliance in a nationwide sample of women aged 40 to 64 years, stratified by insurance status, and found that 72% of those with insurance accessed mammography in 2019, which reflects the rates of mammography access in our population.33 To further validate our findings, future research could compare outcomes from our patient cohort (an insured patient population with good adherence to posttreatment surveillance) to a cohort from the generalized and/or uninsured population. However, because various barriers exist regarding different aspects of surveillance, this study is important in helping to highlight the most valuable methods for detecting second cancers in order to allocate resources to increase access to these methods.
There are limitations to our study. The retrospective design reflects biases inherent to our institutional practice that rely on observational data, and comparisons can be impacted by selection bias or confounding. Future studies using a prospective design would add value, such as time-to-event analysis for proportions of patients with second breast cancer detected via the 3 different modes. In our study, we only included patients with ≥5 years of follow-up to eliminate patients lost to follow-up soon after initial DCIS diagnosis. As a result, we may have excluded some patients with early events. However, a previous study from the Cancer Research Network (in which our institution provided 50% of patients) demonstrated low rates of disenrollment in the 5 years following cancer diagnosis.34 Additionally, the overall rate of second cancers detected in our cohort is similar to that in other studies. Lastly, our study follow-up period was through 2021 and included the first 18 months of the COVID-19 pandemic, during which time surveillance patterns may have been impacted. However, it is unlikely that this had a significant impact on our study findings, because most second breast cancers were diagnosed before COVID restrictions, and the total number of surveillance visits reflects the NCCN Guideline-based recommendations.
Conclusions
In this retrospective study with a median follow-up of 10 years, we found that 94.4% of second breast cancers were detected via mammography or as a result of patient-reported symptoms, and only 2.2% of second breast cancers were detected via routine CBE. Decreasing the need for in-person breast examinations could allow for other effective methods of survivorship monitoring, such as telehealth. Our findings highlight the importance of mammogram screening and patient education regarding self-detection, and can inform future NCCN recommendations for DCIS survivorship care.
Acknowledgments
The authors would like to thank the Kaiser Permanente Northern California Graduate Medical Education program for supporting this project, particularly Aida Shirazi.
References
- 1.↑
American Cancer Society. Cancer statistics center. Accessed July 18, 2022. Available at: https://cancerstatisticscenter.cancer.org
- 2.↑
Virnig BA, Tuttle TM, Shamliyan T, et al. Ductal carcinoma in situ of the breast: a systematic review of incidence, treatment, and outcomes. J Natl Cancer Inst 2010;102:170–178.
- 3.↑
Bijker N, Meijnen P, Peterse JL, et al. Breast-conserving treatment with or without radiotherapy in ductal carcinoma-in-situ: ten-year results of European Organisation for Research and Treatment of Cancer randomized phase III trial 10853—a study by the EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group. J Clin Oncol 2006;24:3381–3387.
- 4.↑
Narod SA, Iqbal J, Giannakeas V, et al. Breast cancer mortality after a diagnosis of ductal carcinoma in situ. JAMA Oncol 2015;1:888–896.
- 5.↑
Wapnir IL, Dignam JJ, Fisher B, et al. Long-term outcomes of invasive ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24 randomized clinical trials for DCIS. J Natl Cancer Inst 2011;103:478–488.
- 6.↑
Habel LA, Moe RE, Daling JR, et al. Risk of contralateral breast cancer among women with carcinoma in situ of the breast. Ann Surg 1997;225:69–75.
- 7.↑
Li CI, Malone KE, Saltzman BS, et al. Risk of invasive breast carcinoma among women diagnosed with ductal carcinoma in situ and lobular carcinoma in situ, 1988–2001. Cancer 2006;106:2104–2112.
- 8.↑
Gradishar WJ, Moran MS, Abraham J, et al. NCCN Clinical Practice Guidelines in Oncology: Breast Cancer. Version 4.2023. Accessed April 18, 2023. To view the most recent version, visit https://www.nccn.org
- 9.↑
Montgomery DA, Krupa K, Cooke TG. Follow-up in breast cancer: does routine clinical examination improve outcome? A systematic review of the literature. Br J Cancer 2007;97:1632–1641.
- 10.↑
Montgomery DA, Krupa K, Jack WJ, et al. Changing pattern of the detection of locoregional relapse in breast cancer: the Edinburgh experience. Br J Cancer 2007;96:1802–1807.
- 11.↑
van der Sangen MJ, van de Poll-Franse LV, Roumen RM, et al. The prognosis of patients with local recurrence more than five years after breast conservation therapy for invasive breast carcinoma. Eur J Surg Oncol 2006;32:34–38.
- 12.↑
Tang A, Neeman E, Vuong B, et al. Care in the time of COVID-19: impact on the diagnosis and treatment of breast cancer in a large, integrated health care system. Breast Cancer Res Treat 2022;191:665–675.
- 13.↑
Teicher S, Whitney RL, Liu R. Breast cancer survivors’ satisfaction and information recall of telehealth survivorship care plan appointments during the COVID-19 pandemic. Oncol Nurs Forum 2022;49:223–231.
- 14.↑
Byng D, Thomas SM, Rushing CN, et al. Surveillance imaging after primary diagnosis of ductal carcinoma in situ. Radiology 2023;307:e221210.
- 15.↑
Montgomery DA, Krupa K, Cooke TG. Alternative methods of follow up in breast cancer: a systematic review of the literature. Br J Cancer 2007;96:1625–1632.
- 16.↑
McDonald S, Saslow D, Alciati MH. Performance and reporting of clinical breast examination: a review of the literature. CA Cancer J Clin 2004;54:345–361.
- 17.↑
Smith RA, Saslow D, Sawyer KA, et al. American Cancer Society guidelines for breast cancer screening: update 2003. CA Cancer J Clin 2003;53:141–169.
- 18.↑
Siu AL, Calonge N, Petitti DB, et al. Screening for breast cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2016;164:279–296.
- 19.↑
Tun SM, Alluri S, Rastegar V, et al. Mode of detection of second events in routine surveillance of early stage breast cancer patients. Clin Breast Cancer 2022;22:e818–824.
- 20.↑
Thomas DB, Gao DL, Ray RM, et al. Randomized trial of breast self-examination in Shanghai: final results. J Natl Cancer Inst 2002;94:1445–1457.
- 21.↑
U.S. Preventive Services Task Force. Breast cancer: screening. Accessed August 8, 2022. Available at: https://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/breast-cancer-screening
- 22.↑
Neeman E, Lyon L, Sun H, et al. Future of teleoncology: trends and disparities in telehealth and secure message utilization in the COVID-19 era. JCO Clin Cancer Inform 2022;6:e2100160.
- 23.↑
Liu R, Sundaresan T, Reed ME, et al. Telehealth in oncology during the COVID-19 outbreak: bringing the house call back virtually. JCO Oncol Pract 2020;16:289–293.
- 24.↑
Shih YT, Kim B, Halpern MT. State of physician and pharmacist oncology workforce in the United States in 2019. JCO Oncol Pract 2021;17:e1–10.
- 25.↑
State of Cancer Care in America. 2022 snapshot: state of the oncology workforce in America. JCO Oncol Pract 2022;18:396.
- 26.↑
Razavi SA, Hart AM, Betarbet U, et al. The incidence of occult malignant and high-risk pathologic findings in breast reduction specimens. Plast Reconstr Surg 2021;148:534e–539e.
- 27.↑
Nergiz D, Süren D, Alikanoğlu AS, et al. Rate of incidental pathological lesions in reduction mammoplasty specimens and incidence of invasive breast carcinoma following breast reduction operation. Aesthetic Plast Surg 2022;46:83–90.
- 28.↑
Carlson GW. The management of breast cancer detected by reduction mammaplasty. Clin Plast Surg 2016;43:341–347.
- 29.↑
Arvold ND, Punglia RS, Hughes ME, et al. Pathologic characteristics of second breast cancers after breast conservation for ductal carcinoma in situ. Cancer 2012;118:6022–6030.
- 30.↑
Falk RS, Hofvind S, Skaane P, et al. Second events following ductal carcinoma in situ of the breast: a register-based cohort study. Breast Cancer Res Treat 2011;129:929–938.
- 31.↑
Miller ME, Muhsen S, Olcese C, et al. Contralateral breast cancer risk in women with ductal carcinoma in situ: is it high enough to justify bilateral mastectomy? Ann Surg Oncol 2017;24:2889–2897.
- 32.↑
Gomez SL, Shariff-Marco S, Von Behren J, et al. Representativeness of breast cancer cases in an integrated health care delivery system. BMC Cancer 2015;15:688.
- 33.↑
National Center for Health Statistics. Use of mammography among women aged 40 and over, by selected characteristics: United States, selected years 1987–2019. Accessed May 1, 2023. Available at: https://www.ncbi.nlm.nih.gov/books/NBK569311/table/ch3.tab33/
- 34.↑
Chubak J, Ziebell R, Greenlee RT, et al. The Cancer Research Network: a platform for epidemiologic and health services research on cancer prevention, care, and outcomes in large, stable populations. Cancer Causes Control 2016;27:1315–1323.
