- Email: [email protected]
Tumor Site and Breast Cancer Prognosis
Accepted Manuscript Tumor Site and Breast Cancer Prognosis Charalampos Siotos, MD, Michael McColl, BS, Kevin Psoter, PhD, Richard C. Gilmore, MD, Mohamad E. Sebai, MD, Kristen P. Broderick, MD, Lisa K. Jacobs, MD, Stephanie Irwin, PhD, Gedge D. Rosson, MD, Mehran Habibi, MD PII:
S1526-8209(18)30299-4
DOI:
10.1016/j.clbc.2018.05.007
Reference:
CLBC 812
To appear in:
Clinical Breast Cancer
Received Date: 4 May 2018 Accepted Date: 22 May 2018
Please cite this article as: Siotos C, McColl M, Psoter K, Gilmore RC, Sebai ME, Broderick KP, Jacobs LK, Irwin S, Rosson GD, Habibi M, Tumor Site and Breast Cancer Prognosis, Clinical Breast Cancer (2018), doi: 10.1016/j.clbc.2018.05.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Tumor Site and Breast Cancer Prognosis
RI PT
Authors: Charalampos Siotos, MD1, Michael McColl, BS1, Kevin Psoter, PhD2, Richard C. Gilmore, MD2, Mohamad E. Sebai, MD2, Kristen P. Broderick, MD1, Lisa K. Jacobs, MD2,
SC
Stephanie Irwin, PhD2, Gedge D. Rosson, MD1, Mehran Habibi, MD2
1. Department of Plastic & Reconstructive Surgery, Johns Hopkins Hospital, Baltimore, MD,
M AN U
USA.
2. Department of Surgery, Johns Hopkins Hospital, Baltimore, MD, USA.
Mehran Habibi, MD
TE D
Corresponding Author:
Johns Hopkins Medicine
EP
4940 Eastern Ave
AC C
Baltimore, MD 21224 Work: 410-550-1226 [email protected]
Declarations of interest: None Funding: None
ACCEPTED MANUSCRIPT 1
RI PT
Tumor Site and Breast Cancer Prognosis
MicroAbstract
The present study is a retrospective review of a prospectively collected institutional cancer
SC
registry. We evaluated more than 5,000 to explore the role of primary tumor location in breast cancer prognosis. Our results indicate that tumors of different primary tumor site in the breast
M AN U
behave differently and survival differences should be expected.
Abstract
TE D
Introduction/Background: Various factors affect breast cancer prognosis; however, little information is available regarding the role of primary tumor site. The purpose of this study was to compare the survival and clinicopathologic characteristics of breast cancer patients by primary
EP
tumor location.
Materials and Methods: We analyzed a prospectively collected single-institution breast cancer
AC C
registry. Univariate and multivariable analyses were used to evaluate the association of tumor site with positive lymph node status, presence of metastasis, time to recurrence, or death and events of recurrence or death. Patients with tumors originating from the upper-outer quadrant were the reference group. Results: From, 2003-2015, 5,295 patients with breast cancer were identified. Tumors originated from the upper-outer quadrant (36.2%), upper-inner (13.1%), lower-outer (9.8%), lower-inner
ACCEPTED MANUSCRIPT 2
(7.6%), nipple (1.2%), axillary tail (0.3%) or overlapping (24.7%). Tumors originated from overlapping lesions (OR=1.58; 95%CI:1.36-1.83) had higher odds of positive axillary lymph nodes, whereas tumors from the upper-inner (OR=0.68; 95%CI:0.56-0.84) and lower-inner
RI PT
quadrant (OR=0.72; 95%CI:0.56-0.93) had lower odds. Tumors from the lower-outer quadrant were associated with lower risk of death (HR=0.64; 95%CI:0.46-0.88), whereas tumors from overlapping lesions had higher risk (HR=1.28; 95%CI:1.05-1.55).
SC
Conclusion: The site of primary tumor may be an important characteristic affecting the
M AN U
prognosis of patients with breast cancer.
AC C
EP
TE D
Keywords: breast cancer; primary tumor site; recurrences; survival; prognosis.
ACCEPTED MANUSCRIPT 3
Introduction Establishing a prognosis is a critical step in the treatment of breast cancer. Starting with an
RI PT
informed estimate about the course of disease allows the physician and patient to make proper personalized decisions about treatment options. Many patient and tumor characteristics are currently used for prognosis in breast cancer patients, such as: patient age, hormone receptor status, tumor size, lymph node involvement, and histologic grade.1 Primary tumor site has been
SC
shown to affect survival in various other forms of cancer such as colorectal, urological, sarcoma,
M AN U
and Merkel cell carcinoma.2–6 The vast colorectal cancer literature has also established a relationship between tumor site and response to adjuvant therapy.7 The mechanism has been suggested to be related to a varying presence of mutations, such as TP53, between sites.8 A better understanding of primary tumor site location and its effect on the prognosis of breast cancer is an
tumor site.
TE D
important first step that may lead to future targeted therapies for breast cancer based on primary
Primary tumor location has yet to be adopted as a standard of prognosis although there have been
EP
multiple studies suggesting a substantial prognostic utility.9–15 The lack of agreement on existence of biological differences among tumors of different sites makes inclusion of tumor site
AC C
as a prognostic value challenging. However, its role has also been underlined by studies conducted in Asia, which agreed that tumors from inner breast quadrants tend to have less favorable prognosis.16,17 In this context, the aim of this study was to evaluate the association between tumor site and selected patient outcomes at a high volume academic hospital system in the U.S.
ACCEPTED MANUSCRIPT 4
RI PT
Materials and Methods The local institutional review board (under protocol IRB00046361) approved this study. The study population included all females who were diagnosed with and treated for breast cancer
SC
from 2003-2015 and were included in the Johns Hopkins Sidney Kimmel Comprehensive Cancer Center (SKCCC) tumor registry, an ongoing prospective registry initiated in 1999. We excluded
M AN U
women with bilateral breast cancer, male patients, and patients with insufficient data. Demographic and clinical characteristics of interest were extracted. The primary exposure of interest was tumor site and was classified according to the International Classification of Diseases for Oncology Third Edition (ICD-O-3)18 which includes: breast upper-outer quadrant
TE D
(UOQ), breast upper-inner quadrant (UIQ), breast lower-outer (LOQ), breast lower-inner (LIQ), breast central portion, breast axillary tail, breast overlapping lesion, and nipple. Our endpoints of interest included lymph node status and metastasis during the follow-up period.
EP
In addition, we evaluated breast cancer recurrence, time to recurrence, progression free survival,
AC C
and time to death among patients with breast cancer over the study period.
Statistical analysis
Demographic and tumor characteristics were summarized by tumor site. Individuals entered risk sets at the date of diagnosis and were followed until death for the survival analysis, or until date of first cancer recurrence for the progression analysis. Individuals who did not experience these
ACCEPTED MANUSCRIPT 5
outcomes were censored at their last clinical visit occurring prior to December 31, 2015. Kaplan Meier plots were produced to display the time to death and progression free survival for all tumor sites. Initially, univariate logistic regression was used to evaluate the association of tumor Similarly, Cox proportional hazards
RI PT
site with positive lymph node status and recurrence.
regression was used to evaluate the association between tumor site and time to recurrence, death and recurrence or death. Next, multivariable regression was used to evaluate the association of
SC
tumor site with each of the previously described outcomes after adjustment for patient and tumor characteristics, including age of patient, stage of disease, lymph nodes status, grade, estrogen and
M AN U
progesterone sensitivity of the tumor, receipt of surgical treatment, chemotherapy, radiation, and hormonal therapy with each outcome. For these analyses, tumors originating from the upperouter quadrant were taken as the reference group. Finally, we evaluated patient and tumor characteristics and their association with time to death and progression free survival for each
TE D
tumor site. Due to the minimal number of individuals with cancer originating from the nipple and breast axilla, we excluded these sites from regression analyses. A p-value <0.05 was considered statistically significant for all comparisons. All analyses were performed using STATA Version
AC C
EP
13.1 (StataCorp LP, College Station, TX).
Results
Demographic characteristics A total of 6,763 individuals were diagnosed with breast cancer at our institution between 20032015, of which 5,295 (78.3%) met inclusion criteria with available demographic and tumor data. Breast cancer most commonly originated from the UOQ (n=1,919), followed by overlapping
ACCEPTED MANUSCRIPT 6
lesion (n=1,310), UIQ (n=695), LOQ (n=518), LIQ (n=403), central portion (n=372), nipple (n=62), and least commonly from the axillary tail (n=16). Baseline characteristics of the patients are presented in Table 1. The median follow-up was 4.68 years (Interquartile Range-IQR=2.18-
RI PT
7.57 years). Outcomes
SC
Outcomes are displayed in Table 2. The highest mortality was seen in tumors originating from the central portion (16.9%), followed by LIQ (13.9%), overlapping lesions (13.6%), UOQ
M AN U
(12.5%), UIQ (11.2%), nipple (9.7%), LOQ (8.3%), and axillary tail (6.3%). Recurrences were identified in overlapping lesions (5.8%), LIQ (5.0%), central (4.3%), UOQ (4.1%), LOQ (3.5%), UIQ (2.7), axillary tail and nipple (0%). In contrast, positive lymph nodes were present in 56.3% of tumors from the axillary tail, 38.9% overlapping lesions, 33.9% nipple, 28.7% UOQ, 148%
TE D
LOQ, 26.9% central, 22.6% LIQ, and 21.6% from the UIQ.
Kaplan Meier survival and progression free survival are presented in Figures 1 and 2. In univariate analysis, breast malignant pathologies that originated from overlapping lesions
EP
(OR=1.58, 95% CI [1.36, 1.83]) had statistically significantly higher odds of positive lymph nodes (N1/N2/N3) in comparison to UOQ tumors (Table 2). In contrast, tumors from the UIQ
AC C
(OR=0.68, 95% CI [0.56, 0.84]) and the LIQ (OR=0.72, 95% CI [0.56, 0.93]) had statistically significantly lower odds of positive lymph nodes. After adjustment for individual and tumor characteristics, tumors that originated from the LOQ had statistically significantly lower risk of death (HR=0.70, 95% CI [0.50, 0.97]) and progression free survival (HR=0.75, 95% CI [0.56, 1.00]) in comparison to tumors from the
ACCEPTED MANUSCRIPT 7
UOQ (Figure 1). Breast cancer patients had similar odds for recurrence and risk of recurrence regardless of the tumor site of origin. (Figure 2, Table 3)
tumor site separately.
RI PT
We evaluated risk factors for death (Table 4) and progression free survival (Table 5) for each Older age and higher stage were associated with a statistically
significantly increased risk of death and progression free survival for all tumors except those
associated with lower risk for these outcomes.
SC
originating from the UOQ and central portion. Radiation and hormonal therapy were generally
M AN U
Metastasis to bones was the most common site for patients presenting with stage IV disease regardless of the origin site of the breast cancer, followed by liver, lung, and finally brain (Table 6).
TE D
Discussion
Our study suggests differences in survival among tumors distributed throughout different regions of the breast. Our results reinforce the findings of large-scale studies of national breast cancer
EP
data from various countries, such as our finding that breast cancer usually originates in the upper
AC C
outer quadrant as well as the poor prognosis associated with tumors in the medial half of the breast. Conversely, our finding that tumors originating in the lower outer region of the breast have the best prognosis contradicts the common finding of upper outer tumors having the best outcomes.
This study finds agreement with others that have shown the upper outer quadrant to be the most common location of breast tumors.12,17 This commonality is accompanied by superiority in survival among all primary tumor site locations according to multiple studies.19,20 The more
ACCEPTED MANUSCRIPT 8
favorable survival outcomes of upper outer tumors has been credited to having the most complete surgical management of the tumor burden of all regions thanks to improvements in management of the axilla.21 Interestingly, Hwang et al. (2017) reported more advanced
RI PT
clinicopathologic features for upper-outer tumors, including axillary lymph node metastasis, yet they were not associated with a worse prognosis.17 This was hypothesized to be a result of increased lymphatic flow through the axillary nodes rather than the internal mammary nodes as
SC
compared to more medial tumors. Our finding of significantly lower rates of axillary lymph node metastasis in both inner quadrants supports this claim. Metastasis to the internal mammary nodes
M AN U
has long been blamed for poor prognosis among breast cancer patients given their “hidden” nature. The difficulty in detection on physical exam and imaging leads to understaging and undertreatment.9,17 Consequently, internal mammary node metastases are associated with a 30% decline in 10 year survival.19 Upper outer tumors may have a better prognosis due to their
TE D
distance from these internal mammary nodes and likelihood of axillary node metastasis which is more easily detected and surgically managed.
Lymph node metastasis was the most common in overlapping tumors in our study. Positive
EP
lymph nodes have been found to be most common in both central (periareolar) tumors and upper
AC C
outer quadrant tumors in previous studies.17,21 Due to this high rate of lymph node metastasis, overlapping tumors would be expected to be associated with poorer survival as this metastasis has been shown many times over to be the most telling factor of a poor prognosis.22,23 Our data initially showed an increased hazard ratio for time to recurrence, progression free survival, and time to death for these overlapping tumors. However, when this data was adjusted for patient and tumor characteristics, including age of patient, stage of disease, lymph nodes status, grade, estrogen and progesterone sensitivity of the tumor, receipt of surgical treatment, chemotherapy,
ACCEPTED MANUSCRIPT 9
radiation, and hormonal therapy, these findings were no longer significant. Although it seems a logical association between increased axillary lymph node metastasis and worse outcomes, some studies have found paradoxical results.10,11,16,24 One of these studies attributed the increase in
RI PT
survival with an increased rate of lymph node metastasis to a difference in treatment strategies.24 The study explained that node positive patients are more likely to receive aggressive treatment such as adjuvant chemotherapy.24 So although their clinicopathological characteristics favored a
SC
poor prognosis, the aggressive treatment they received kept the survival rates in line with other tumor locations. This paradox has been investigated in the reverse as well. Hwang et al. (2017)
M AN U
found that lower-inner quadrant tumors had the worst survival, but relatively favorable clinicopathological characteristics.17 They suggested that hidden metastases to the internal mammary nodes and an absence of aggressive therapy were to blame. This hypothesis was strengthened by the fact that the overall survival in lower-inner quadrant node negative patients
TE D
was actually not significantly different than other tumor locations when looking at only those patients who received chemotherapy and radiation. Colleoni et al. (2005) also found that their result of poorer outcomes in medial tumors was even worse when looking only at node-negative
EP
patients.9 This was likely due to the absence of aggressive treatment due to the negative lymph nodes. Unsurprisingly, the survival differences among primary tumor site regions in the breast
AC C
are partly related to treatment decisions surrounding risk factors such as axillary lymph node metastasis.
The lateral or outer half of the breast has been found by many studies to have the most favorable outcomes.9–11,16,17 Our data shows that tumors from the lower outer quadrant are associated with better survival than tumors from other regions. Although there is agreement that the outer half of the breast carries a better prognosis, many studies show the upper outer quadrant to have the best
ACCEPTED MANUSCRIPT 10
survival.12,16 As discussed previously, most of these studies attribute the difference to hidden metastases to the internal mammary nodes in the medial half combined with undertreatment. However, some studies have discredited this idea. For example, one study showed that medial-
RI PT
central tumors actually have no greater probability of death or distant metastases than lateral tumors despite metastasizing to the internal mammary nodes more often.25 And one study even asserted that the lower-outer quadrant actually has the highest amount of lymphatic drainage to
SC
the internal mammary nodes at 42.9% which could nullify the theory entirely.26 The underlying mechanism for the favorability of lateral tumors requires further investigation. Until the
M AN U
mechanism is fully understood, it will be difficult to establish primary tumor site as a standard of prognosis
Age at diagnosis and tumor stage were associated with higher risk of death and progression free survival in the vast majority of sites. Recurrences were very rare in our population causing
TE D
progression free survival to approximate time to death. The associations between advanced age and worse outcomes are actually somewhat contested in the breast cancer literature with multiple studies finding an association and others not.27 However, our results suggest older patients face
EP
higher risks of mortality and recurrence events even after adjusting for other demographic and
AC C
clinical characteristics. Some studies with similar results have hypothesized that this may be due to differences in treatment strategies in older patients, such as foregoing adjuvant radiotherapy and axillary lymph node dissections.27 This less aggressive treatment may be due to an abundance of comorbidities, although studies have suggested that even when controlling for comorbidities, older patients receive substandard care.28 Other studies suggest that there are inherent biological differences in play that offer younger women a better prognosis.29 Tumor
ACCEPTED MANUSCRIPT 11
stage is expected to be associated with higher risk as a higher stage indicates more aggressive and advanced disease.30
RI PT
A history of radiation therapy or hormonal therapy was protective at many tumor sites and resulted in a lower risk of death and disease free progression. Chemotherapy, however, did not affect risk. Radiotherapy has been shown to significantly reduce the risk of local recurrence as well as reducing long-term breast cancer mortality. However, it has also been demonstrated to
SC
increase all-cause mortality due to a significant increase in vascular related deaths.31 Hormonal
M AN U
therapy also improved progression free survival and time to death in many tumor sites. A largescale review of randomized trials demonstrated the efficacy of hormonal therapy for ER-positive disease in 15-year survival.32 The rate of ER positivity in each of our tumor sites approximates 80%. Although not causally related, the results support the potential benefit of hormonal therapy
TE D
in this setting.
Of the few cases with distant metastases, the most common site was bone. Breast cancer has been shown in multiple studies to present most commonly as osteolytic metastases in bone.33
EP
This is due to the overexpression of genes in these cancer cell lines with functions such as bone marrow homing and extravasation, pericellular proteolysis and invasion, angiogenesis,
AC C
osteoclastogenesis, growth factor regulation, and extracellular matrix alteration.34 Our results reflect this established trend although the occurrence of metastasis was too low in our study for formal statistical analysis.
Our study is limited by our inability to separate cancer-related mortality from all-cause mortality and its retrospective nature. However, it is based on a prospectively collected database with a large number of patients treated at our institution.
ACCEPTED MANUSCRIPT 12
Conclusion Conflicting data still exists about primary tumor site origin and its association with recurrence
RI PT
and survival. Overall, the data that we present is in agreement with the evidence published in the current literature, however there are a few noticeable differences as well. This points to more than just geography affecting the hazard ratios of each tumor site location. It is likely that proximity to lymph nodes is accompanied by genetic and other variations in its effect on
SC
survival. Further study to truly differentiate what it the cause of the variable outcomes in each
M AN U
region is warranted. If a truly causal relationship exists, primary tumor site should be incorporated as a standard of prognosis in breast cancer which will enable physicians to better educate their patients and treat their disease. Clinical Practice Points
TE D
Various factors affect breast cancer prognosis; however, little information is available regarding the role of primary tumor site. This extensive work evaluates the survival and clinicopathologic characteristics of breast cancer patients by the location of the primary tumor. Our results indicate
EP
that tumors of different primary tumor site in the breast behave differently and survival differences should be expected. In particular, tumors originated from overlapping lesions had
AC C
higher odds of positive axillary lymph nodes, whereas tumors from the upper-inner and lowerinner quadrant had lower odds. Tumors from the lower-outer quadrant were associated with lower risk of death, whereas tumors from overlapping lesions had higher risk. This knowledge has serious implications on breast cancer prognosis and is essential in order to provide better education to our patients with breast cancer.
ACCEPTED MANUSCRIPT 13
References 1.
Colzani E, Liljegren A, Johansson AL V, et al. Prognosis of patients with breast cancer:
RI PT
Causes of death and effects of time since diagnosis, age, and tumor characteristics. J Clin Oncol. 2011;29(30):4014-4021. doi:10.1200/JCO.2010.32.6462. 2.
Aoyama T, Kashiwabara K, Oba K, et al. Clinical impact of tumor location on the colon
SC
cancer survival and recurrence: Analyses of pooled data from three large phase III randomized clinical trials. Cancer Med. 2017. doi:10.1002/cam4.1208. Jafarabadi MA, Mohammadi SM, Hajizadeh E, Kazemnejad A, Fatemi SR. Does the
M AN U
3.
prognosis of colorectal cancer vary with tumor site? Gastroenterol Hepatol from Bed to Bench. 2011;4(4):199-209. 4.
Dutta R, Abdelhalim A, Martin JW, et al. Effect of tumor location on survival in urinary
TE D
bladder adenocarcinoma: A population-based analysis. Urol Oncol Semin Orig Investig. 2016;34(12):531.e1-531.e6. doi:10.1016/j.urolonc.2016.06.009. 5.
Smith CA, Martinez SR, Tseng WH, et al. Predicting survival for well-differentiated
EP
liposarcoma: The importance of tumor location. J Surg Res. 2012;175(1):12-17.
6.
AC C
doi:10.1016/j.jss.2011.07.024. Smith VA, MaDan OP, Lentsch EJ. Tumor Location Is an Independent Prognostic Factor in Head and Neck Merkel Cell Carcinoma. Otolaryngol Neck Surg. 2012;146(3):403-408. doi:10.1177/0194599811431789.
7.
Aljehani MA, Morgan JW, Guthrie LA, et al. Association of Primary Tumor Site With Mortality in Patients Receiving Bevacizumab and Cetuximab for Metastatic Colorectal
ACCEPTED MANUSCRIPT 14
Cancer. JAMA Surg. 2017;92350. doi:10.1001/jamasurg.2017.3466. 8.
Russo A, Bazan V, Iacopetta B, et al. The TP53 colorectal cancer international
RI PT
collaborative study on the prognostic and predictive significance of p53 mutation: influence of tumor site, type of mutation, and adjuvant treatment. J Clin Oncol. 2005;23(30):7518-7528.
Colleoni M, Zahrieh D, Gelber RD, et al. Site of primary tumor has a prognostic role in
SC
9.
operable breast cancer: The International Breast Cancer Study Group experience. J Clin
10.
M AN U
Oncol. 2005;23(7):1390-1400. doi:10.1200/JCO.2005.06.052.
Zucali BR, Mariani L, Marubini E, et al. Early Breast Cancer : Evaluation of the Prognostic Role of the Site of the Primary Tumor. 1998;16(4):1363-1366.
11.
Gaffney DK, Tsodikov A, Wiggins CL. Diminished survival in patients with inner versus quadrant
breast
cancers.
J
Clin
Oncol.
2003;21(3):467-472.
TE D
outer
doi:10.1200/JCO.2003.12.047. 12.
Sohn VY, Arthurs ZM, Sebesta JA, Brown TA. Primary tumor location impacts breast
Lohrisch BC, Jackson J, Jones A, Mates D, Olivotto IA. Relationship Between Tumor
AC C
13.
EP
cancer survival. Am J Surg. 2008;195(5):641-644. doi:10.1016/j.amjsurg.2007.12.039.
Location and Relapse in 6,781 Women With Early Invasive Breast Cancer. J Clin Oncol. 2007;18(15):2828-2835.
14.
Nemoto T, Natarajan S, Bedwani R, Vana J, Murphy GP. Breast Cancer in the Medial
Half: Results of 1978 National Survey of the American College of Surgeons. 1983. 15.
Giuliano AE, Connolly JL, Edge SB, Mittendorf EA. Breast Cancer — Major Changes in
ACCEPTED MANUSCRIPT 15
the American Joint Committee on Cancer Eighth Edition Cancer Staging Manual. A Cancer Journals Clin. 2017;67(0):290-303. doi:10.3322/caac.21393. Bao J, Yu K, Jiang Y, Shao Z, Di G. The Effect of Laterality and Primary Tumor Site on
RI PT
16.
Cancer-Specific Mortality in Breast Cancer : A SEER Population-Based Study. 2014;9(4). doi:10.1371/journal.pone.0094815.
Hwang KT, Kim J, Kim EK, et al. Poor Prognosis of Lower Inner Quadrant in Lymph
SC
17.
Node-negative Breast Cancer Patients Who Received No Chemotherapy: A Study Based
M AN U
on Nationwide Korean Breast Cancer Registry Database. Clin Breast Cancer. 2017;(July):169-184. doi:10.1016/j.clbc.2016.12.011. 18.
Fritz A, Percy C, Jack A, et al. International Classification of Diseases for Oncology: Third Edition. World Heal Organ. 2000:240 pages.
Veronesi U, Cascinelli N, Greco M, et al. Prognosis of breast cancer patients after
TE D
19.
mastectomy and dissection of internal mammary nodes. Ann Surg. 1985;202(6):702-707. 20.
Sahoo PK, Jana D, Mandal PK, Basak S. Effect of lymphangiogenesis and
EP
lymphovascular invasion on the survival pattern of breast cancer patients. Asian Pacific J
21.
AC C
Cancer Prev. 2014;15(15):6287-6293. doi:10.7314/APJCP.2014.15.15.6287. Kroman N, Wohlfahrt J, Mouridsen HT, Melbye M. Influence of tumor location on breast
cancer prognosis. Int J Cancer. 2003;105(4):542-545. doi:10.1002/ijc.11116.
22.
Michaelson JS, Chen LL, Silverstein MJ, et al. How cancer at the primary site and in the
lymph nodes contributes to the risk of cancer death. Cancer. 2009;115(21):5095-5107. doi:10.1002/cncr.24592.
ACCEPTED MANUSCRIPT 16
23.
Andersson Y. Breast cancer survival in relation to the metastatic tumor burden in axillary lymph nodes. J Clin Oncol. 2010;28(17):2868-2873. Glick JH, Gelber RD, Goldhirsch A, Senn HJ. Adjuvant therapy of primary breast cancer.
RI PT
24.
4th International Conference on Adjuvant Therapy of Primary Breast Cancer St. Gallen, Switzerland. Ann Oncol. 1992;3(10):801-807.
Fisher B, Wolmark N, Redmond C, Deutsch M, Fisher ER. Findings from NSABP
SC
25.
protocol no. b04: Comparison of radical mastectomy with alternative treatments. II. The biologic
significance of
1981;48(8):1863-1872. 26.
medialcentral
breast
cancers.
Cancer.
M AN U
clinical and
Shahar KH, Buchholz TA, Delpassand E, et al. Lower and central tumor location correlates with lymphoscintigraphy drainage to the internal mammary lymph nodes in
27.
TE D
breast carcinoma. Cancer. 2005;103(7):1323-1329. doi:10.1002/cncr.20914. Brandt J, Garne JP, Tengrup I, Manjer J. Age at diagnosis in relation to survival following
28.
Schonberg MA, Marcantonio ER, Li D, Silliman RA, Ngo L, McCarthy EP. Breast cancer
29.
AC C
0429-x.
EP
breast cancer: a cohort study. World J Surg Oncol. 2015;13:33. doi:10.1186/s12957-014-
among the oldest old: Tumor characteristics, treatment choices, and survival. J Clin Oncol. 2010;28(12):2038-2045. doi:10.1200/JCO.2009.25.9796.
Sant M, Allemani C, Capocaccia R, et al. Stage at diagnosis is a key explanation of differences in breast cancer survival across Europe. Int J Cancer. 2003;106(3):416-422. doi:10.1002/ijc.11226.
ACCEPTED MANUSCRIPT 17
30.
Howlader N, Noone A, Krapcho M, et al. SEER Cancer Statistics Review. Bethesda, MD Available at: https://seer.cancer.gov/csr/1975_2014/. Assessed on December 12, 2017. Noël G, Mazeron JJ. Favourable and unfavourable effects on long-term survival of
RI PT
31.
radiotherapy for early breast cancer: an overview of the randomised trials. Early Breast Cancer
Trialists’
Collaborative
Group.
Lancet.
32.
SC
doi:S0140673600022637 [pii].
2000;355(9217):1757-1770.
Early Breast Cancer Trialists' Collaborative Group (EBCTCG). Effects of chemotherapy
overview
of
the
M AN U
and hormonal therapy for early breast cancer on recurrence and 15-year survival: an randomised
trials.
Lancet.
2005;365(9472):1687-1717.
doi:10.1016/S0140-6736(05)66544-0. 33.
Boyce BF, Yoneda T, Guise T a. Factors regulating the growth of metastatic cancer in
34.
TE D
bone. Endocr Relat Cancer. 1999;6(3):333-347.
Kang Y, Siegel PM, Shu W, et al. A multigenic program mediating breast cancer
AC C
6.
EP
metastasis to bone. Cancer Cell. 2003;3(6):537-549. doi:10.1016/S1535-6108(03)00132-
Legends and titles for figures and tables Figure 1: Kaplan-Meier survival estimates for overall survival of patients with breast cancer arising from different sites.
Figure 2: Kaplan-Meier estimates for time to recurrence of patients with breast cancer arising from different sites.
ACCEPTED MANUSCRIPT 18
Table 1: Baseline characteristics and survival outcomes of patients with tumors originating from multiple sites.
RI PT
Table 2: Unadjusted logistic regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from 2003 to 2015.
SC
Table 3: Cox proportional hazards regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from
M AN U
2003 to 2015.
Table 4: Factors that affect survival of patients with breast cancer, startified by tumor site origin. Table 5: Factors that affect progression free survival of patients with breast cancer, startified by tumor site origin. Statistically significant differences are bolded.
AC C
EP
by tumor site origin.
TE D
Table 6: Patterns of metastatis at the time of diagnosis for patients with breast cancer stratified
ACCEPTED MANUSCRIPT
Table 1: Baseline characteristics and survival outcomes of patients with tumors originating from multiple
Age (mean, SD)
Breast, upperinner quadrant (n=695) 57.1 (12.78)
57.4 (12.59)
60.3 (14.90)
58.6 (12.33)
III
336 (17.5) 799 (41.6) 549 (28.6) 171 (8.9)
21 (33.9) 16 (25.8) 20 (32.3) 3 (4.8)
IV
64 (3.3)
2 (3.2)
73 98 (14.1) (19.6) 147 345 (39.5) (49.6) 94 210 (25.3) (30.2) 37 29 (4.2) (10.0) 21 (5.7) 13 (1.9)
I
238 (12.4) 869 (45.3) 686 (35.8) 1 (0.1)
II
Grade
II III IV
EP
I
AC C
0
TE D
Stage
6 (9.7)
28 (45.2) 15 (24.2) 0 (0.0)
56 (15.1) 175 (47.0) 107 (28.8) 2 (0.5)
Breast, lowerinner quadrant (n=403)
Breast, lowerouter quadrant (n=518)
SC
Breast, central portion (n=372)
M AN U
Breast, Nipple upper(n=62) outer quadrant (n=1919)
RI PT
sites.
101 (14.5) 326 (46.9) 224 (32.2) 2 (0.3)
Breast, axillary tail (n=16)
Breast, overlapping lesion (n=1310)
57.4 (12.56)
55.7 (12.21)
56.7 (11.03)
55.0 (12.67)
88 (21.8)
92 (17.8)
1 (6.3)
214 (16.3)
173 (42.9) 109 (27.1) 22 (5.5)
225 (43.4) 139 (26.8) 5 (31.3)
393 (30.0)
11 (2.7)
390 (29.8)
5 (31.3) 5 (31.3) 4 (25.0) 1 (6.3)
50 (12.4)
57 (11.0)
122 (9.3)
182 (45.2) 142 (35.2) 0 (0.0)
249 (48.1) 175 (33.8) 1 (0.2)
2 (12.5) 6 (37.5) 5 (31.3) 0 (0.0)
390 (29.8) 245 (18.7) 68 (5.2)
633 (48.3) 467 (35.7) 3 (0.2)
13 (21.0)
32 (8.6)
42 (6.0)
29 (7.2)
36 (7.0)
3 (18.8)
85 (6.5)
41 (66.1) 19 (30.7) 2 (3.2)
272 (73.1) 79 (21.2) 15 (4.0)
545 (78.4) 129 (18.6) 14 (2.0)
312 (77.4) 68 (16.9)
370 (71.4) 106 (20.5) 31 (6.0)
3
43 (2.2)
0 (0.0)
6 (1.6)
7 (1.0)
11 (2.7)
7 (43.8) 6 (37.5) 3 (18.8) 0 (0.0)
801 (61.2)
2
1368 (71.3) 405 (21.1) 103 (5.4)
White
1400 (72.3) 390 (20.3) 129 (6.7)
46 (74.2) 13 (21.0) 3 (4.8)
277 (74.5) 76 (20.4) 19 (5.1)
474 (68.2) 168 (24.2) 53 (7.6)
272 (67.5) 103 (25.6) 28 (7.0)
369 (71.2) 118 (22.8) 31 (6.0)
11 (68.8) 4 (25.0) 1 (6.3)
923 (70.5)
1082 (56.4) 156 (8.1)
38 (61.3) 1 (1.6)
206 (55.4) 32 (8.6)
394 (56.7) 52 (7.5)
219 (54.3) 45 (11.2)
302 (58.3) 41 (7.9)
747 (57.0)
482 (25.1) 199 (10.4)
12 (19.4) 11 (17.7)
88 (23.7) 46 (12.4)
179 (25.8) 70 (10.1)
99 (24.6) 40 (9.9)
118 (22.8) 57 (11.0)
8 (50.0) 3 (18.8) 3 (18.8) 2 (12.5)
4 (6.5) 29 (46.8) 29 (46.8)
35 (9.4) 215 (57.8) 122 (32.8)
27 (3.9) 408 (58.7) 260 (37.4)
23 (5.7) 217 (53.9) 163 (40.5)
30 (5.8) 243 (46.9) 245 (47.3)
1 (6.3) 7 (43.8) 8 (50.0)
90 (6.9) 396 (30.2)
N stage 0 1
Former Missing Surgery No surgery Lumpectomy Mastectomy Adjuvant therapy
133 (6.9) 1113 (58.0) 673 (35.1)
TE D
Active
AC C
AfricanAmerican Other Smoking Never
11 (2.1)
M AN U
Race
12 (3.0)
SC
unknown
RI PT
125 (6.5)
EP
ACCEPTED MANUSCRIPT
309 (23.6) 128 (9.8) 72 (5.5)
313 (23.9) 74 (5.7)
116 (8.9) 318 (24.3) 129 (9.9)
824 (62.9)
ACCEPTED MANUSCRIPT
Outcomes Death Any recurrence
405 (21.1) 1220 (63.6) 287 (15.0) 7 (0.4)
14 (22.6) 31 (50.0) 17 (27.4) 0 (0.0)
86 (23.1) 215 (57.8) 69 (18.6) 2 (0.5)
121 (17.4) 490 (70.5) 83 (11.9)
80 (19.9)
1494 (77.9) 1297 (67.6)
49 (79.0) 40 (64.5)
306 (82.3) 269 (72.3)
239 (12.5) 79 (4.1)
6 (9.7) 0 (0.0)
265 (51.2) 212 (40.9) 307 (59.3)
9 (56.3) 9 (56.3) 11 (68.8)
277 (68.7) 44 (10.9)
104 (20.1) 342 (66.0) 72 (13.9)
1 (0.1)
2 (0.5)
0 (0.0)
2 (12.5) 8 (50.0) 6 (37.5) 0 (0.0)
537 (77.3) 468 (67.3)
329 (81.6) 293 (72.7)
413 (79.7) 362 (69.9)
14 (87.5) 12 (75.0)
1049 (80.1)
56 (13.9)
43 (8.3)
1 (6.3)
178 (13.6)
20 (5.0)
18 (3.5)
0 (0.0)
76 (5.8)
RI PT
219 (54.3) 140 (34.7) 227 (56.3)
SC
400 (57.6) 256 (36.8) 420 (60.4)
M AN U
PR+
222 (59.7) 126 (33.9) 226 (60.8)
TE D
Sentinel LN biopsy Other dissection Unknown Hormonal sensitivity ER+
27 (43.6) 16 (25.8) 30 (48.4)
63 78 (11.2) (16.9) 16 (4.3) 19 (2.7)
EP
Hormonal Therapy Lymph Node Surgery No
1157 (60.3) 732 (38.1) 1119 (58.3)
AC C
Radiation Therapy Chemotherapy
685 (52.3) 602 (46.0) 810 (61.8)
172 (13.1) 898 (68.6) 233 (17.8) 7 (0.5)
932 (71.2)
ACCEPTED MANUSCRIPT
Table 2: Unadjusted logistic regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from 2003 to
Time to death HR (95% CI) REF
Positive lymph nodes (N1/2/3) OR (95% CI) REF
1.05 (0.60, 1.81)
0.76 (0.45, 1.31)
1.09 (0.84, 1.41)
1.08 (0.82, 1.43)
0.91 (0.71, 1.17)
0.65 0.39, 1.09)
0.65 (0.39, 1.07)
0.86 (0.68, 1.10)
0.90 (0.70, 1.17)
0.68 (0.56, 0.84)
1.22 (0.74, 2.01)
1.09 (0.67, 1.78)
1.07 (0.81, 1.40)
1.08 (0.81, 1.44)
0.72 (0.56, 0.93)
SC
REF
Progression free HR (95% CI) REF
TE D
M AN U
Time to recurrence HR (95% CI)
0.84 (0.50, 1.41)
0.77 (0.46, 1.29)
0.70 (0.52, 0.94)
0.64 (0.46, 0.88)
0.99 (0.80, 1.23)
1.43 (1.04, 1.98)
1.74 (1.27, 2.38)
1.31 (1.10, 1.57)
1.28 (1.05, 1.55)
1.58 (1.36, 1.83)
EP
Breast, upperouter quadrant Breast, central portion Breast, upperinner quadrant Breast, lowerinner quadrant Breast, lowerouter quadrant Breast, overlapping lesion
Any recurrence OR (95% CI) REF
RI PT
2015.
AC C
Bold font represents statistically signficant outcomes. REF=Reference group for the comparisons, OR=Odds Ratio, HR=Hazard Ratio, CI=Confidence Interval.
ACCEPTED MANUSCRIPT
Table 3: Cox proportional hazards regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from
1.19 (0.67, 2.09)
0.90 (0.52, 1.54)
1.10 (0.84, 1.43)
1.11 (0.83, 1.47)
0.71 (0.42, 1.19)
0.69 (0.41, 1.14)
0.97 (0.76, 1.23)
0.99 (0.76, 1.29)
1.37 (0.82, 2.30)
1.34 (0.81, 2.20)
1.21 (0.92, 1.59)
1.26 (0.94, 1.70)
0.84 (0.49, 1.43)
0.77 (0.46, 1.30)
0.75 (0.56, 1.00)
0.70 (0.50, 0.97)
1.10 (0.78, 1.56)
1.32 (0.94, 1.84)
1.15 (0.96, 1.39)
1.16 (0.95, 1.42)
SC
Time to death HR (95% CI)
REF
Progression free HR (95% CI) REF
M AN U
Breast, upperouter quadrant Breast, central portion Breast, upperinner quadrant Breast, lowerinner quadrant Breast, lowerouter quadrant Breast, overlapping lesion
Time to recurrence HR (95% CI) REF
TE D
Any recurrence OR (95% CI)
RI PT
2003 to 2015.
REF
EP
Bold font represents statistically signficant outcomes. REF=Reference group for the
AC C
comparisons, OR=Odds Ratio, HR=Hazard Ratio, CI=Confidence Interval.
ACCEPTED MANUSCRIPT
Table 4: Factors that affect survival of patients with breast cancer, startified by tumor site origin.
Nodal stage Grade Estrogen positive Progesterone positive Chemotherapy Radiation therapy Hormonal therapy
Breast, lowerouter quadrant 1.06 (1.03, 1.10) 1.87 (1.09, 3.22) 0.86 (0.41, 1.81) 1.78 (1.00, 3.18) 0.93 (0.50, 1.72) 1.48 (0.38, 5.75) 0.73 (0.23, 2.28) 0.93 (0.34, 2.56) 0.47 (0.17, 1.34) 0.52 (0.23, 1.16)
Bold font represents statistically signficant outcomes. Outcomes are Hazard ratio (95%
AC C
EP
Confidence Interval).
Breast, overlapping lesion
RI PT
Breast, lowerinner quadrant 1.05 (1.03, 1.08) 1.62 (1.00, 2.64) 0.77 (0.45, 1.32) 1.31 (0.78, 2.18) 2.73 (1.50, 4.97) 0.76 (0.24, 2.43) 2.58 (0.89, 7.47) 1.63 (0.73, 3.66) 0.83 (0.39, 1.76) 0.35 (0.17, 0.71)
SC
Sentinel
1.07 (1.04, 1.09) 2.03 (1.40, 2.94) 1.40 (0.89, 2.21) 1.45 (0.87, 2.41) 1.38 (0.85, 2.22) 0.93 (0.26, 3.33) 1.85 (0.64, 5.40) 0.54 (0.25, 1.17) 0.38 (0.18, 0.83) 0.61 (0.28, 1.35)
Breast, upperinner quadrant 1.06 (1.03, 1.08) 2.07 (1.45, 2.96) 0.72 (0.44, 1.19) 0.80 (0.49, 1.29) 0.98 (0.63, 1.53) 0.98 (0.39, 2.49) 0.74 (0.35, 1.59) 1.18 (0.63, 2.24) 0.78 (0.41, 1.48) 0.54 (0.30, 0.98)
M AN U
Tumor stage
Breast, central portion
TE D
Age at diagnosis
Breast, upperouter quadrant 1.03 (1.02, 1.05) 1.91 (1.58, 2.30) 0.97 (0.76, 1.24) 1.09 (0.89, 1.34) 1.26 (0.98, 1.62) 1.53 (0.94, 2.48) 0.98 (0.64, 1.50) 0.82 (0.59, 1.16) 0.66 (0.47, 0.94) 0.35 (0.25, 0.49)
1.02 (1.01, 1.03) 2.16 (1.72, 2.71) 0.94 (0.72, 1.23) 1.23 (1.01, 1.49) 1.17 (0.87, 1.56) 0.80 (0.43, 1.49) 0.67 (0.40, 1.13) 0.46 (0.31, 0.68) 0.95 (0.65, 1.39) 0.53 (0.33, 0.83)
ACCEPTED MANUSCRIPT
Table 5: Factors that affect progression free survival of patients with breast cancer, startified by tumor site origin. Statistically significant differences are bolded.
Nodal stage Grade Estrogen positive Progesterone positive Chemotherapy Radiation therapy Hormonal therapy
Breast, lowerouter quadrant 1.04 (1.01, 1.07) 1.87 (1.17, 2.98) 0.78 (0.40, 1.50) 1.36 (0.80, 2.30) 1.04 (0.61, 1.78) 1.54 (0.46, 5.15) 0.77 (0.29, 2.07) 0.90 (0.38, 2.15) 0.50 (0.21, 1.22) 0.61 (0.30, 1.25)
EP
Bold font represents statistically signficant outcomes. Outcomes are Hazard ratio (95%
AC C
Confidence Interval).
Breast, overlapping lesion
RI PT
Breast, lowerinner quadrant 1.04 (1.02, 1.07) 1.60 (1.02, 2.51) 0.74 (0.45, 1.21) 1.35 (0.84, 2.18) 2.54 (1.48, 4.34) 1.03 (0.36, 2.94) 1.75 (0.70, 4.36) 1.52 (0.74, 3.14) 0.85 (0.42, 1.70) 0.39 (0.20, 0.76)
SC
Sentinel
1.05 (1.03, 1.07) 1.93 (1.36, 2.72) 1.53 (0.99, 2.36) 1.46 (0.90, 2.37) 1.20 (0.77, 1.86) 1.23 (0.39, 3.91) 1.24 (0.48, 3.22) 0.49 (0.24, 1.00) 0.38 (0.19, 0.78) 0.44 (0.21, 0.91)
Breast, upperinner quadrant 1.04 (1.02, 1.06) 2.01 (1.44, 2.79) 0.88 (0.55, 1.41) 0.78 (0.50, 1.21) 0.89 (0.59, 1.35) 0.81 (0.34, 1.89) 0.68 (0.33, 1.39) 1.20 (0.67, 2.16) 0.63 (0.34, 1.18) 0.59 (0.33, 1.05)
M AN U
Tumor stage
Breast, central portion
TE D
Age at diagnosis
Breast, upperouter quadrant 1.03 (1.02, 1.04) 1.79 (1.49, 2.13) 1.00 (0.80, 1.25) 1.22 (1.01, 1.49) 1.29 (1.02, 1.63) 1.74 (1.11, 2.72) 0.84 (0.57, 1.25) 0.79 (0.57, 1.09) 0.54 (0.39, 0.75) 0.37 (0.27, 0.51)
1.02 (1.00, 1.03) 1.89 (1.54, 2.32) 0.93 (0.73, 1.19) 1.24 (1.04, 1.48) 1.19 (0.92, 1.55) 0.97 (0.55, 1.70) 0.68 (0.43, 1.08) 0.59 (0.41, 0.84) 0.98 (0.69, 1.40) 0.63 (0.42, 0.95)
ACCEPTED MANUSCRIPT
Table 6: Patterns of metastatis at the time of diagnosis for patients with breast cancer stratified
RI PT
by tumor site origin.
Metastasis to Liver 11
Metastasis to Lung 14
Metastasis to Brain 4
1
0
0
0
Breast, central portion
8
1
Breast, upper-inner quadrant Breast, lower-inner quadrant Breast, lower-outer quadrant Breast, axillary tail
2
2
4 9 0 45
AC C
EP
TE D
Breast, overlapping lesion
2
0
1
0
1
0
0
4
2
0
0
1
0
21
20
2
M AN U
Breast, upper-outer quadrant Nipple
SC
Metastasis to Bone 27
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S1526-8209(18)30299-4
DOI:
10.1016/j.clbc.2018.05.007
Reference:
CLBC 812
To appear in:
Clinical Breast Cancer
Received Date: 4 May 2018 Accepted Date: 22 May 2018
Please cite this article as: Siotos C, McColl M, Psoter K, Gilmore RC, Sebai ME, Broderick KP, Jacobs LK, Irwin S, Rosson GD, Habibi M, Tumor Site and Breast Cancer Prognosis, Clinical Breast Cancer (2018), doi: 10.1016/j.clbc.2018.05.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Tumor Site and Breast Cancer Prognosis
RI PT
Authors: Charalampos Siotos, MD1, Michael McColl, BS1, Kevin Psoter, PhD2, Richard C. Gilmore, MD2, Mohamad E. Sebai, MD2, Kristen P. Broderick, MD1, Lisa K. Jacobs, MD2,
SC
Stephanie Irwin, PhD2, Gedge D. Rosson, MD1, Mehran Habibi, MD2
1. Department of Plastic & Reconstructive Surgery, Johns Hopkins Hospital, Baltimore, MD,
M AN U
USA.
2. Department of Surgery, Johns Hopkins Hospital, Baltimore, MD, USA.
Mehran Habibi, MD
TE D
Corresponding Author:
Johns Hopkins Medicine
EP
4940 Eastern Ave
AC C
Baltimore, MD 21224 Work: 410-550-1226 [email protected]
Declarations of interest: None Funding: None
ACCEPTED MANUSCRIPT 1
RI PT
Tumor Site and Breast Cancer Prognosis
MicroAbstract
The present study is a retrospective review of a prospectively collected institutional cancer
SC
registry. We evaluated more than 5,000 to explore the role of primary tumor location in breast cancer prognosis. Our results indicate that tumors of different primary tumor site in the breast
M AN U
behave differently and survival differences should be expected.
Abstract
TE D
Introduction/Background: Various factors affect breast cancer prognosis; however, little information is available regarding the role of primary tumor site. The purpose of this study was to compare the survival and clinicopathologic characteristics of breast cancer patients by primary
EP
tumor location.
Materials and Methods: We analyzed a prospectively collected single-institution breast cancer
AC C
registry. Univariate and multivariable analyses were used to evaluate the association of tumor site with positive lymph node status, presence of metastasis, time to recurrence, or death and events of recurrence or death. Patients with tumors originating from the upper-outer quadrant were the reference group. Results: From, 2003-2015, 5,295 patients with breast cancer were identified. Tumors originated from the upper-outer quadrant (36.2%), upper-inner (13.1%), lower-outer (9.8%), lower-inner
ACCEPTED MANUSCRIPT 2
(7.6%), nipple (1.2%), axillary tail (0.3%) or overlapping (24.7%). Tumors originated from overlapping lesions (OR=1.58; 95%CI:1.36-1.83) had higher odds of positive axillary lymph nodes, whereas tumors from the upper-inner (OR=0.68; 95%CI:0.56-0.84) and lower-inner
RI PT
quadrant (OR=0.72; 95%CI:0.56-0.93) had lower odds. Tumors from the lower-outer quadrant were associated with lower risk of death (HR=0.64; 95%CI:0.46-0.88), whereas tumors from overlapping lesions had higher risk (HR=1.28; 95%CI:1.05-1.55).
SC
Conclusion: The site of primary tumor may be an important characteristic affecting the
M AN U
prognosis of patients with breast cancer.
AC C
EP
TE D
Keywords: breast cancer; primary tumor site; recurrences; survival; prognosis.
ACCEPTED MANUSCRIPT 3
Introduction Establishing a prognosis is a critical step in the treatment of breast cancer. Starting with an
RI PT
informed estimate about the course of disease allows the physician and patient to make proper personalized decisions about treatment options. Many patient and tumor characteristics are currently used for prognosis in breast cancer patients, such as: patient age, hormone receptor status, tumor size, lymph node involvement, and histologic grade.1 Primary tumor site has been
SC
shown to affect survival in various other forms of cancer such as colorectal, urological, sarcoma,
M AN U
and Merkel cell carcinoma.2–6 The vast colorectal cancer literature has also established a relationship between tumor site and response to adjuvant therapy.7 The mechanism has been suggested to be related to a varying presence of mutations, such as TP53, between sites.8 A better understanding of primary tumor site location and its effect on the prognosis of breast cancer is an
tumor site.
TE D
important first step that may lead to future targeted therapies for breast cancer based on primary
Primary tumor location has yet to be adopted as a standard of prognosis although there have been
EP
multiple studies suggesting a substantial prognostic utility.9–15 The lack of agreement on existence of biological differences among tumors of different sites makes inclusion of tumor site
AC C
as a prognostic value challenging. However, its role has also been underlined by studies conducted in Asia, which agreed that tumors from inner breast quadrants tend to have less favorable prognosis.16,17 In this context, the aim of this study was to evaluate the association between tumor site and selected patient outcomes at a high volume academic hospital system in the U.S.
ACCEPTED MANUSCRIPT 4
RI PT
Materials and Methods The local institutional review board (under protocol IRB00046361) approved this study. The study population included all females who were diagnosed with and treated for breast cancer
SC
from 2003-2015 and were included in the Johns Hopkins Sidney Kimmel Comprehensive Cancer Center (SKCCC) tumor registry, an ongoing prospective registry initiated in 1999. We excluded
M AN U
women with bilateral breast cancer, male patients, and patients with insufficient data. Demographic and clinical characteristics of interest were extracted. The primary exposure of interest was tumor site and was classified according to the International Classification of Diseases for Oncology Third Edition (ICD-O-3)18 which includes: breast upper-outer quadrant
TE D
(UOQ), breast upper-inner quadrant (UIQ), breast lower-outer (LOQ), breast lower-inner (LIQ), breast central portion, breast axillary tail, breast overlapping lesion, and nipple. Our endpoints of interest included lymph node status and metastasis during the follow-up period.
EP
In addition, we evaluated breast cancer recurrence, time to recurrence, progression free survival,
AC C
and time to death among patients with breast cancer over the study period.
Statistical analysis
Demographic and tumor characteristics were summarized by tumor site. Individuals entered risk sets at the date of diagnosis and were followed until death for the survival analysis, or until date of first cancer recurrence for the progression analysis. Individuals who did not experience these
ACCEPTED MANUSCRIPT 5
outcomes were censored at their last clinical visit occurring prior to December 31, 2015. Kaplan Meier plots were produced to display the time to death and progression free survival for all tumor sites. Initially, univariate logistic regression was used to evaluate the association of tumor Similarly, Cox proportional hazards
RI PT
site with positive lymph node status and recurrence.
regression was used to evaluate the association between tumor site and time to recurrence, death and recurrence or death. Next, multivariable regression was used to evaluate the association of
SC
tumor site with each of the previously described outcomes after adjustment for patient and tumor characteristics, including age of patient, stage of disease, lymph nodes status, grade, estrogen and
M AN U
progesterone sensitivity of the tumor, receipt of surgical treatment, chemotherapy, radiation, and hormonal therapy with each outcome. For these analyses, tumors originating from the upperouter quadrant were taken as the reference group. Finally, we evaluated patient and tumor characteristics and their association with time to death and progression free survival for each
TE D
tumor site. Due to the minimal number of individuals with cancer originating from the nipple and breast axilla, we excluded these sites from regression analyses. A p-value <0.05 was considered statistically significant for all comparisons. All analyses were performed using STATA Version
AC C
EP
13.1 (StataCorp LP, College Station, TX).
Results
Demographic characteristics A total of 6,763 individuals were diagnosed with breast cancer at our institution between 20032015, of which 5,295 (78.3%) met inclusion criteria with available demographic and tumor data. Breast cancer most commonly originated from the UOQ (n=1,919), followed by overlapping
ACCEPTED MANUSCRIPT 6
lesion (n=1,310), UIQ (n=695), LOQ (n=518), LIQ (n=403), central portion (n=372), nipple (n=62), and least commonly from the axillary tail (n=16). Baseline characteristics of the patients are presented in Table 1. The median follow-up was 4.68 years (Interquartile Range-IQR=2.18-
RI PT
7.57 years). Outcomes
SC
Outcomes are displayed in Table 2. The highest mortality was seen in tumors originating from the central portion (16.9%), followed by LIQ (13.9%), overlapping lesions (13.6%), UOQ
M AN U
(12.5%), UIQ (11.2%), nipple (9.7%), LOQ (8.3%), and axillary tail (6.3%). Recurrences were identified in overlapping lesions (5.8%), LIQ (5.0%), central (4.3%), UOQ (4.1%), LOQ (3.5%), UIQ (2.7), axillary tail and nipple (0%). In contrast, positive lymph nodes were present in 56.3% of tumors from the axillary tail, 38.9% overlapping lesions, 33.9% nipple, 28.7% UOQ, 148%
TE D
LOQ, 26.9% central, 22.6% LIQ, and 21.6% from the UIQ.
Kaplan Meier survival and progression free survival are presented in Figures 1 and 2. In univariate analysis, breast malignant pathologies that originated from overlapping lesions
EP
(OR=1.58, 95% CI [1.36, 1.83]) had statistically significantly higher odds of positive lymph nodes (N1/N2/N3) in comparison to UOQ tumors (Table 2). In contrast, tumors from the UIQ
AC C
(OR=0.68, 95% CI [0.56, 0.84]) and the LIQ (OR=0.72, 95% CI [0.56, 0.93]) had statistically significantly lower odds of positive lymph nodes. After adjustment for individual and tumor characteristics, tumors that originated from the LOQ had statistically significantly lower risk of death (HR=0.70, 95% CI [0.50, 0.97]) and progression free survival (HR=0.75, 95% CI [0.56, 1.00]) in comparison to tumors from the
ACCEPTED MANUSCRIPT 7
UOQ (Figure 1). Breast cancer patients had similar odds for recurrence and risk of recurrence regardless of the tumor site of origin. (Figure 2, Table 3)
tumor site separately.
RI PT
We evaluated risk factors for death (Table 4) and progression free survival (Table 5) for each Older age and higher stage were associated with a statistically
significantly increased risk of death and progression free survival for all tumors except those
associated with lower risk for these outcomes.
SC
originating from the UOQ and central portion. Radiation and hormonal therapy were generally
M AN U
Metastasis to bones was the most common site for patients presenting with stage IV disease regardless of the origin site of the breast cancer, followed by liver, lung, and finally brain (Table 6).
TE D
Discussion
Our study suggests differences in survival among tumors distributed throughout different regions of the breast. Our results reinforce the findings of large-scale studies of national breast cancer
EP
data from various countries, such as our finding that breast cancer usually originates in the upper
AC C
outer quadrant as well as the poor prognosis associated with tumors in the medial half of the breast. Conversely, our finding that tumors originating in the lower outer region of the breast have the best prognosis contradicts the common finding of upper outer tumors having the best outcomes.
This study finds agreement with others that have shown the upper outer quadrant to be the most common location of breast tumors.12,17 This commonality is accompanied by superiority in survival among all primary tumor site locations according to multiple studies.19,20 The more
ACCEPTED MANUSCRIPT 8
favorable survival outcomes of upper outer tumors has been credited to having the most complete surgical management of the tumor burden of all regions thanks to improvements in management of the axilla.21 Interestingly, Hwang et al. (2017) reported more advanced
RI PT
clinicopathologic features for upper-outer tumors, including axillary lymph node metastasis, yet they were not associated with a worse prognosis.17 This was hypothesized to be a result of increased lymphatic flow through the axillary nodes rather than the internal mammary nodes as
SC
compared to more medial tumors. Our finding of significantly lower rates of axillary lymph node metastasis in both inner quadrants supports this claim. Metastasis to the internal mammary nodes
M AN U
has long been blamed for poor prognosis among breast cancer patients given their “hidden” nature. The difficulty in detection on physical exam and imaging leads to understaging and undertreatment.9,17 Consequently, internal mammary node metastases are associated with a 30% decline in 10 year survival.19 Upper outer tumors may have a better prognosis due to their
TE D
distance from these internal mammary nodes and likelihood of axillary node metastasis which is more easily detected and surgically managed.
Lymph node metastasis was the most common in overlapping tumors in our study. Positive
EP
lymph nodes have been found to be most common in both central (periareolar) tumors and upper
AC C
outer quadrant tumors in previous studies.17,21 Due to this high rate of lymph node metastasis, overlapping tumors would be expected to be associated with poorer survival as this metastasis has been shown many times over to be the most telling factor of a poor prognosis.22,23 Our data initially showed an increased hazard ratio for time to recurrence, progression free survival, and time to death for these overlapping tumors. However, when this data was adjusted for patient and tumor characteristics, including age of patient, stage of disease, lymph nodes status, grade, estrogen and progesterone sensitivity of the tumor, receipt of surgical treatment, chemotherapy,
ACCEPTED MANUSCRIPT 9
radiation, and hormonal therapy, these findings were no longer significant. Although it seems a logical association between increased axillary lymph node metastasis and worse outcomes, some studies have found paradoxical results.10,11,16,24 One of these studies attributed the increase in
RI PT
survival with an increased rate of lymph node metastasis to a difference in treatment strategies.24 The study explained that node positive patients are more likely to receive aggressive treatment such as adjuvant chemotherapy.24 So although their clinicopathological characteristics favored a
SC
poor prognosis, the aggressive treatment they received kept the survival rates in line with other tumor locations. This paradox has been investigated in the reverse as well. Hwang et al. (2017)
M AN U
found that lower-inner quadrant tumors had the worst survival, but relatively favorable clinicopathological characteristics.17 They suggested that hidden metastases to the internal mammary nodes and an absence of aggressive therapy were to blame. This hypothesis was strengthened by the fact that the overall survival in lower-inner quadrant node negative patients
TE D
was actually not significantly different than other tumor locations when looking at only those patients who received chemotherapy and radiation. Colleoni et al. (2005) also found that their result of poorer outcomes in medial tumors was even worse when looking only at node-negative
EP
patients.9 This was likely due to the absence of aggressive treatment due to the negative lymph nodes. Unsurprisingly, the survival differences among primary tumor site regions in the breast
AC C
are partly related to treatment decisions surrounding risk factors such as axillary lymph node metastasis.
The lateral or outer half of the breast has been found by many studies to have the most favorable outcomes.9–11,16,17 Our data shows that tumors from the lower outer quadrant are associated with better survival than tumors from other regions. Although there is agreement that the outer half of the breast carries a better prognosis, many studies show the upper outer quadrant to have the best
ACCEPTED MANUSCRIPT 10
survival.12,16 As discussed previously, most of these studies attribute the difference to hidden metastases to the internal mammary nodes in the medial half combined with undertreatment. However, some studies have discredited this idea. For example, one study showed that medial-
RI PT
central tumors actually have no greater probability of death or distant metastases than lateral tumors despite metastasizing to the internal mammary nodes more often.25 And one study even asserted that the lower-outer quadrant actually has the highest amount of lymphatic drainage to
SC
the internal mammary nodes at 42.9% which could nullify the theory entirely.26 The underlying mechanism for the favorability of lateral tumors requires further investigation. Until the
M AN U
mechanism is fully understood, it will be difficult to establish primary tumor site as a standard of prognosis
Age at diagnosis and tumor stage were associated with higher risk of death and progression free survival in the vast majority of sites. Recurrences were very rare in our population causing
TE D
progression free survival to approximate time to death. The associations between advanced age and worse outcomes are actually somewhat contested in the breast cancer literature with multiple studies finding an association and others not.27 However, our results suggest older patients face
EP
higher risks of mortality and recurrence events even after adjusting for other demographic and
AC C
clinical characteristics. Some studies with similar results have hypothesized that this may be due to differences in treatment strategies in older patients, such as foregoing adjuvant radiotherapy and axillary lymph node dissections.27 This less aggressive treatment may be due to an abundance of comorbidities, although studies have suggested that even when controlling for comorbidities, older patients receive substandard care.28 Other studies suggest that there are inherent biological differences in play that offer younger women a better prognosis.29 Tumor
ACCEPTED MANUSCRIPT 11
stage is expected to be associated with higher risk as a higher stage indicates more aggressive and advanced disease.30
RI PT
A history of radiation therapy or hormonal therapy was protective at many tumor sites and resulted in a lower risk of death and disease free progression. Chemotherapy, however, did not affect risk. Radiotherapy has been shown to significantly reduce the risk of local recurrence as well as reducing long-term breast cancer mortality. However, it has also been demonstrated to
SC
increase all-cause mortality due to a significant increase in vascular related deaths.31 Hormonal
M AN U
therapy also improved progression free survival and time to death in many tumor sites. A largescale review of randomized trials demonstrated the efficacy of hormonal therapy for ER-positive disease in 15-year survival.32 The rate of ER positivity in each of our tumor sites approximates 80%. Although not causally related, the results support the potential benefit of hormonal therapy
TE D
in this setting.
Of the few cases with distant metastases, the most common site was bone. Breast cancer has been shown in multiple studies to present most commonly as osteolytic metastases in bone.33
EP
This is due to the overexpression of genes in these cancer cell lines with functions such as bone marrow homing and extravasation, pericellular proteolysis and invasion, angiogenesis,
AC C
osteoclastogenesis, growth factor regulation, and extracellular matrix alteration.34 Our results reflect this established trend although the occurrence of metastasis was too low in our study for formal statistical analysis.
Our study is limited by our inability to separate cancer-related mortality from all-cause mortality and its retrospective nature. However, it is based on a prospectively collected database with a large number of patients treated at our institution.
ACCEPTED MANUSCRIPT 12
Conclusion Conflicting data still exists about primary tumor site origin and its association with recurrence
RI PT
and survival. Overall, the data that we present is in agreement with the evidence published in the current literature, however there are a few noticeable differences as well. This points to more than just geography affecting the hazard ratios of each tumor site location. It is likely that proximity to lymph nodes is accompanied by genetic and other variations in its effect on
SC
survival. Further study to truly differentiate what it the cause of the variable outcomes in each
M AN U
region is warranted. If a truly causal relationship exists, primary tumor site should be incorporated as a standard of prognosis in breast cancer which will enable physicians to better educate their patients and treat their disease. Clinical Practice Points
TE D
Various factors affect breast cancer prognosis; however, little information is available regarding the role of primary tumor site. This extensive work evaluates the survival and clinicopathologic characteristics of breast cancer patients by the location of the primary tumor. Our results indicate
EP
that tumors of different primary tumor site in the breast behave differently and survival differences should be expected. In particular, tumors originated from overlapping lesions had
AC C
higher odds of positive axillary lymph nodes, whereas tumors from the upper-inner and lowerinner quadrant had lower odds. Tumors from the lower-outer quadrant were associated with lower risk of death, whereas tumors from overlapping lesions had higher risk. This knowledge has serious implications on breast cancer prognosis and is essential in order to provide better education to our patients with breast cancer.
ACCEPTED MANUSCRIPT 13
References 1.
Colzani E, Liljegren A, Johansson AL V, et al. Prognosis of patients with breast cancer:
RI PT
Causes of death and effects of time since diagnosis, age, and tumor characteristics. J Clin Oncol. 2011;29(30):4014-4021. doi:10.1200/JCO.2010.32.6462. 2.
Aoyama T, Kashiwabara K, Oba K, et al. Clinical impact of tumor location on the colon
SC
cancer survival and recurrence: Analyses of pooled data from three large phase III randomized clinical trials. Cancer Med. 2017. doi:10.1002/cam4.1208. Jafarabadi MA, Mohammadi SM, Hajizadeh E, Kazemnejad A, Fatemi SR. Does the
M AN U
3.
prognosis of colorectal cancer vary with tumor site? Gastroenterol Hepatol from Bed to Bench. 2011;4(4):199-209. 4.
Dutta R, Abdelhalim A, Martin JW, et al. Effect of tumor location on survival in urinary
TE D
bladder adenocarcinoma: A population-based analysis. Urol Oncol Semin Orig Investig. 2016;34(12):531.e1-531.e6. doi:10.1016/j.urolonc.2016.06.009. 5.
Smith CA, Martinez SR, Tseng WH, et al. Predicting survival for well-differentiated
EP
liposarcoma: The importance of tumor location. J Surg Res. 2012;175(1):12-17.
6.
AC C
doi:10.1016/j.jss.2011.07.024. Smith VA, MaDan OP, Lentsch EJ. Tumor Location Is an Independent Prognostic Factor in Head and Neck Merkel Cell Carcinoma. Otolaryngol Neck Surg. 2012;146(3):403-408. doi:10.1177/0194599811431789.
7.
Aljehani MA, Morgan JW, Guthrie LA, et al. Association of Primary Tumor Site With Mortality in Patients Receiving Bevacizumab and Cetuximab for Metastatic Colorectal
ACCEPTED MANUSCRIPT 14
Cancer. JAMA Surg. 2017;92350. doi:10.1001/jamasurg.2017.3466. 8.
Russo A, Bazan V, Iacopetta B, et al. The TP53 colorectal cancer international
RI PT
collaborative study on the prognostic and predictive significance of p53 mutation: influence of tumor site, type of mutation, and adjuvant treatment. J Clin Oncol. 2005;23(30):7518-7528.
Colleoni M, Zahrieh D, Gelber RD, et al. Site of primary tumor has a prognostic role in
SC
9.
operable breast cancer: The International Breast Cancer Study Group experience. J Clin
10.
M AN U
Oncol. 2005;23(7):1390-1400. doi:10.1200/JCO.2005.06.052.
Zucali BR, Mariani L, Marubini E, et al. Early Breast Cancer : Evaluation of the Prognostic Role of the Site of the Primary Tumor. 1998;16(4):1363-1366.
11.
Gaffney DK, Tsodikov A, Wiggins CL. Diminished survival in patients with inner versus quadrant
breast
cancers.
J
Clin
Oncol.
2003;21(3):467-472.
TE D
outer
doi:10.1200/JCO.2003.12.047. 12.
Sohn VY, Arthurs ZM, Sebesta JA, Brown TA. Primary tumor location impacts breast
Lohrisch BC, Jackson J, Jones A, Mates D, Olivotto IA. Relationship Between Tumor
AC C
13.
EP
cancer survival. Am J Surg. 2008;195(5):641-644. doi:10.1016/j.amjsurg.2007.12.039.
Location and Relapse in 6,781 Women With Early Invasive Breast Cancer. J Clin Oncol. 2007;18(15):2828-2835.
14.
Nemoto T, Natarajan S, Bedwani R, Vana J, Murphy GP. Breast Cancer in the Medial
Half: Results of 1978 National Survey of the American College of Surgeons. 1983. 15.
Giuliano AE, Connolly JL, Edge SB, Mittendorf EA. Breast Cancer — Major Changes in
ACCEPTED MANUSCRIPT 15
the American Joint Committee on Cancer Eighth Edition Cancer Staging Manual. A Cancer Journals Clin. 2017;67(0):290-303. doi:10.3322/caac.21393. Bao J, Yu K, Jiang Y, Shao Z, Di G. The Effect of Laterality and Primary Tumor Site on
RI PT
16.
Cancer-Specific Mortality in Breast Cancer : A SEER Population-Based Study. 2014;9(4). doi:10.1371/journal.pone.0094815.
Hwang KT, Kim J, Kim EK, et al. Poor Prognosis of Lower Inner Quadrant in Lymph
SC
17.
Node-negative Breast Cancer Patients Who Received No Chemotherapy: A Study Based
M AN U
on Nationwide Korean Breast Cancer Registry Database. Clin Breast Cancer. 2017;(July):169-184. doi:10.1016/j.clbc.2016.12.011. 18.
Fritz A, Percy C, Jack A, et al. International Classification of Diseases for Oncology: Third Edition. World Heal Organ. 2000:240 pages.
Veronesi U, Cascinelli N, Greco M, et al. Prognosis of breast cancer patients after
TE D
19.
mastectomy and dissection of internal mammary nodes. Ann Surg. 1985;202(6):702-707. 20.
Sahoo PK, Jana D, Mandal PK, Basak S. Effect of lymphangiogenesis and
EP
lymphovascular invasion on the survival pattern of breast cancer patients. Asian Pacific J
21.
AC C
Cancer Prev. 2014;15(15):6287-6293. doi:10.7314/APJCP.2014.15.15.6287. Kroman N, Wohlfahrt J, Mouridsen HT, Melbye M. Influence of tumor location on breast
cancer prognosis. Int J Cancer. 2003;105(4):542-545. doi:10.1002/ijc.11116.
22.
Michaelson JS, Chen LL, Silverstein MJ, et al. How cancer at the primary site and in the
lymph nodes contributes to the risk of cancer death. Cancer. 2009;115(21):5095-5107. doi:10.1002/cncr.24592.
ACCEPTED MANUSCRIPT 16
23.
Andersson Y. Breast cancer survival in relation to the metastatic tumor burden in axillary lymph nodes. J Clin Oncol. 2010;28(17):2868-2873. Glick JH, Gelber RD, Goldhirsch A, Senn HJ. Adjuvant therapy of primary breast cancer.
RI PT
24.
4th International Conference on Adjuvant Therapy of Primary Breast Cancer St. Gallen, Switzerland. Ann Oncol. 1992;3(10):801-807.
Fisher B, Wolmark N, Redmond C, Deutsch M, Fisher ER. Findings from NSABP
SC
25.
protocol no. b04: Comparison of radical mastectomy with alternative treatments. II. The biologic
significance of
1981;48(8):1863-1872. 26.
medialcentral
breast
cancers.
Cancer.
M AN U
clinical and
Shahar KH, Buchholz TA, Delpassand E, et al. Lower and central tumor location correlates with lymphoscintigraphy drainage to the internal mammary lymph nodes in
27.
TE D
breast carcinoma. Cancer. 2005;103(7):1323-1329. doi:10.1002/cncr.20914. Brandt J, Garne JP, Tengrup I, Manjer J. Age at diagnosis in relation to survival following
28.
Schonberg MA, Marcantonio ER, Li D, Silliman RA, Ngo L, McCarthy EP. Breast cancer
29.
AC C
0429-x.
EP
breast cancer: a cohort study. World J Surg Oncol. 2015;13:33. doi:10.1186/s12957-014-
among the oldest old: Tumor characteristics, treatment choices, and survival. J Clin Oncol. 2010;28(12):2038-2045. doi:10.1200/JCO.2009.25.9796.
Sant M, Allemani C, Capocaccia R, et al. Stage at diagnosis is a key explanation of differences in breast cancer survival across Europe. Int J Cancer. 2003;106(3):416-422. doi:10.1002/ijc.11226.
ACCEPTED MANUSCRIPT 17
30.
Howlader N, Noone A, Krapcho M, et al. SEER Cancer Statistics Review. Bethesda, MD Available at: https://seer.cancer.gov/csr/1975_2014/. Assessed on December 12, 2017. Noël G, Mazeron JJ. Favourable and unfavourable effects on long-term survival of
RI PT
31.
radiotherapy for early breast cancer: an overview of the randomised trials. Early Breast Cancer
Trialists’
Collaborative
Group.
Lancet.
32.
SC
doi:S0140673600022637 [pii].
2000;355(9217):1757-1770.
Early Breast Cancer Trialists' Collaborative Group (EBCTCG). Effects of chemotherapy
overview
of
the
M AN U
and hormonal therapy for early breast cancer on recurrence and 15-year survival: an randomised
trials.
Lancet.
2005;365(9472):1687-1717.
doi:10.1016/S0140-6736(05)66544-0. 33.
Boyce BF, Yoneda T, Guise T a. Factors regulating the growth of metastatic cancer in
34.
TE D
bone. Endocr Relat Cancer. 1999;6(3):333-347.
Kang Y, Siegel PM, Shu W, et al. A multigenic program mediating breast cancer
AC C
6.
EP
metastasis to bone. Cancer Cell. 2003;3(6):537-549. doi:10.1016/S1535-6108(03)00132-
Legends and titles for figures and tables Figure 1: Kaplan-Meier survival estimates for overall survival of patients with breast cancer arising from different sites.
Figure 2: Kaplan-Meier estimates for time to recurrence of patients with breast cancer arising from different sites.
ACCEPTED MANUSCRIPT 18
Table 1: Baseline characteristics and survival outcomes of patients with tumors originating from multiple sites.
RI PT
Table 2: Unadjusted logistic regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from 2003 to 2015.
SC
Table 3: Cox proportional hazards regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from
M AN U
2003 to 2015.
Table 4: Factors that affect survival of patients with breast cancer, startified by tumor site origin. Table 5: Factors that affect progression free survival of patients with breast cancer, startified by tumor site origin. Statistically significant differences are bolded.
AC C
EP
by tumor site origin.
TE D
Table 6: Patterns of metastatis at the time of diagnosis for patients with breast cancer stratified
ACCEPTED MANUSCRIPT
Table 1: Baseline characteristics and survival outcomes of patients with tumors originating from multiple
Age (mean, SD)
Breast, upperinner quadrant (n=695) 57.1 (12.78)
57.4 (12.59)
60.3 (14.90)
58.6 (12.33)
III
336 (17.5) 799 (41.6) 549 (28.6) 171 (8.9)
21 (33.9) 16 (25.8) 20 (32.3) 3 (4.8)
IV
64 (3.3)
2 (3.2)
73 98 (14.1) (19.6) 147 345 (39.5) (49.6) 94 210 (25.3) (30.2) 37 29 (4.2) (10.0) 21 (5.7) 13 (1.9)
I
238 (12.4) 869 (45.3) 686 (35.8) 1 (0.1)
II
Grade
II III IV
EP
I
AC C
0
TE D
Stage
6 (9.7)
28 (45.2) 15 (24.2) 0 (0.0)
56 (15.1) 175 (47.0) 107 (28.8) 2 (0.5)
Breast, lowerinner quadrant (n=403)
Breast, lowerouter quadrant (n=518)
SC
Breast, central portion (n=372)
M AN U
Breast, Nipple upper(n=62) outer quadrant (n=1919)
RI PT
sites.
101 (14.5) 326 (46.9) 224 (32.2) 2 (0.3)
Breast, axillary tail (n=16)
Breast, overlapping lesion (n=1310)
57.4 (12.56)
55.7 (12.21)
56.7 (11.03)
55.0 (12.67)
88 (21.8)
92 (17.8)
1 (6.3)
214 (16.3)
173 (42.9) 109 (27.1) 22 (5.5)
225 (43.4) 139 (26.8) 5 (31.3)
393 (30.0)
11 (2.7)
390 (29.8)
5 (31.3) 5 (31.3) 4 (25.0) 1 (6.3)
50 (12.4)
57 (11.0)
122 (9.3)
182 (45.2) 142 (35.2) 0 (0.0)
249 (48.1) 175 (33.8) 1 (0.2)
2 (12.5) 6 (37.5) 5 (31.3) 0 (0.0)
390 (29.8) 245 (18.7) 68 (5.2)
633 (48.3) 467 (35.7) 3 (0.2)
13 (21.0)
32 (8.6)
42 (6.0)
29 (7.2)
36 (7.0)
3 (18.8)
85 (6.5)
41 (66.1) 19 (30.7) 2 (3.2)
272 (73.1) 79 (21.2) 15 (4.0)
545 (78.4) 129 (18.6) 14 (2.0)
312 (77.4) 68 (16.9)
370 (71.4) 106 (20.5) 31 (6.0)
3
43 (2.2)
0 (0.0)
6 (1.6)
7 (1.0)
11 (2.7)
7 (43.8) 6 (37.5) 3 (18.8) 0 (0.0)
801 (61.2)
2
1368 (71.3) 405 (21.1) 103 (5.4)
White
1400 (72.3) 390 (20.3) 129 (6.7)
46 (74.2) 13 (21.0) 3 (4.8)
277 (74.5) 76 (20.4) 19 (5.1)
474 (68.2) 168 (24.2) 53 (7.6)
272 (67.5) 103 (25.6) 28 (7.0)
369 (71.2) 118 (22.8) 31 (6.0)
11 (68.8) 4 (25.0) 1 (6.3)
923 (70.5)
1082 (56.4) 156 (8.1)
38 (61.3) 1 (1.6)
206 (55.4) 32 (8.6)
394 (56.7) 52 (7.5)
219 (54.3) 45 (11.2)
302 (58.3) 41 (7.9)
747 (57.0)
482 (25.1) 199 (10.4)
12 (19.4) 11 (17.7)
88 (23.7) 46 (12.4)
179 (25.8) 70 (10.1)
99 (24.6) 40 (9.9)
118 (22.8) 57 (11.0)
8 (50.0) 3 (18.8) 3 (18.8) 2 (12.5)
4 (6.5) 29 (46.8) 29 (46.8)
35 (9.4) 215 (57.8) 122 (32.8)
27 (3.9) 408 (58.7) 260 (37.4)
23 (5.7) 217 (53.9) 163 (40.5)
30 (5.8) 243 (46.9) 245 (47.3)
1 (6.3) 7 (43.8) 8 (50.0)
90 (6.9) 396 (30.2)
N stage 0 1
Former Missing Surgery No surgery Lumpectomy Mastectomy Adjuvant therapy
133 (6.9) 1113 (58.0) 673 (35.1)
TE D
Active
AC C
AfricanAmerican Other Smoking Never
11 (2.1)
M AN U
Race
12 (3.0)
SC
unknown
RI PT
125 (6.5)
EP
ACCEPTED MANUSCRIPT
309 (23.6) 128 (9.8) 72 (5.5)
313 (23.9) 74 (5.7)
116 (8.9) 318 (24.3) 129 (9.9)
824 (62.9)
ACCEPTED MANUSCRIPT
Outcomes Death Any recurrence
405 (21.1) 1220 (63.6) 287 (15.0) 7 (0.4)
14 (22.6) 31 (50.0) 17 (27.4) 0 (0.0)
86 (23.1) 215 (57.8) 69 (18.6) 2 (0.5)
121 (17.4) 490 (70.5) 83 (11.9)
80 (19.9)
1494 (77.9) 1297 (67.6)
49 (79.0) 40 (64.5)
306 (82.3) 269 (72.3)
239 (12.5) 79 (4.1)
6 (9.7) 0 (0.0)
265 (51.2) 212 (40.9) 307 (59.3)
9 (56.3) 9 (56.3) 11 (68.8)
277 (68.7) 44 (10.9)
104 (20.1) 342 (66.0) 72 (13.9)
1 (0.1)
2 (0.5)
0 (0.0)
2 (12.5) 8 (50.0) 6 (37.5) 0 (0.0)
537 (77.3) 468 (67.3)
329 (81.6) 293 (72.7)
413 (79.7) 362 (69.9)
14 (87.5) 12 (75.0)
1049 (80.1)
56 (13.9)
43 (8.3)
1 (6.3)
178 (13.6)
20 (5.0)
18 (3.5)
0 (0.0)
76 (5.8)
RI PT
219 (54.3) 140 (34.7) 227 (56.3)
SC
400 (57.6) 256 (36.8) 420 (60.4)
M AN U
PR+
222 (59.7) 126 (33.9) 226 (60.8)
TE D
Sentinel LN biopsy Other dissection Unknown Hormonal sensitivity ER+
27 (43.6) 16 (25.8) 30 (48.4)
63 78 (11.2) (16.9) 16 (4.3) 19 (2.7)
EP
Hormonal Therapy Lymph Node Surgery No
1157 (60.3) 732 (38.1) 1119 (58.3)
AC C
Radiation Therapy Chemotherapy
685 (52.3) 602 (46.0) 810 (61.8)
172 (13.1) 898 (68.6) 233 (17.8) 7 (0.5)
932 (71.2)
ACCEPTED MANUSCRIPT
Table 2: Unadjusted logistic regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from 2003 to
Time to death HR (95% CI) REF
Positive lymph nodes (N1/2/3) OR (95% CI) REF
1.05 (0.60, 1.81)
0.76 (0.45, 1.31)
1.09 (0.84, 1.41)
1.08 (0.82, 1.43)
0.91 (0.71, 1.17)
0.65 0.39, 1.09)
0.65 (0.39, 1.07)
0.86 (0.68, 1.10)
0.90 (0.70, 1.17)
0.68 (0.56, 0.84)
1.22 (0.74, 2.01)
1.09 (0.67, 1.78)
1.07 (0.81, 1.40)
1.08 (0.81, 1.44)
0.72 (0.56, 0.93)
SC
REF
Progression free HR (95% CI) REF
TE D
M AN U
Time to recurrence HR (95% CI)
0.84 (0.50, 1.41)
0.77 (0.46, 1.29)
0.70 (0.52, 0.94)
0.64 (0.46, 0.88)
0.99 (0.80, 1.23)
1.43 (1.04, 1.98)
1.74 (1.27, 2.38)
1.31 (1.10, 1.57)
1.28 (1.05, 1.55)
1.58 (1.36, 1.83)
EP
Breast, upperouter quadrant Breast, central portion Breast, upperinner quadrant Breast, lowerinner quadrant Breast, lowerouter quadrant Breast, overlapping lesion
Any recurrence OR (95% CI) REF
RI PT
2015.
AC C
Bold font represents statistically signficant outcomes. REF=Reference group for the comparisons, OR=Odds Ratio, HR=Hazard Ratio, CI=Confidence Interval.
ACCEPTED MANUSCRIPT
Table 3: Cox proportional hazards regression evaluating the association between site of breast cancer tumor and selected outcomes amongst women treated at Johns Hopkins Hospital, from
1.19 (0.67, 2.09)
0.90 (0.52, 1.54)
1.10 (0.84, 1.43)
1.11 (0.83, 1.47)
0.71 (0.42, 1.19)
0.69 (0.41, 1.14)
0.97 (0.76, 1.23)
0.99 (0.76, 1.29)
1.37 (0.82, 2.30)
1.34 (0.81, 2.20)
1.21 (0.92, 1.59)
1.26 (0.94, 1.70)
0.84 (0.49, 1.43)
0.77 (0.46, 1.30)
0.75 (0.56, 1.00)
0.70 (0.50, 0.97)
1.10 (0.78, 1.56)
1.32 (0.94, 1.84)
1.15 (0.96, 1.39)
1.16 (0.95, 1.42)
SC
Time to death HR (95% CI)
REF
Progression free HR (95% CI) REF
M AN U
Breast, upperouter quadrant Breast, central portion Breast, upperinner quadrant Breast, lowerinner quadrant Breast, lowerouter quadrant Breast, overlapping lesion
Time to recurrence HR (95% CI) REF
TE D
Any recurrence OR (95% CI)
RI PT
2003 to 2015.
REF
EP
Bold font represents statistically signficant outcomes. REF=Reference group for the
AC C
comparisons, OR=Odds Ratio, HR=Hazard Ratio, CI=Confidence Interval.
ACCEPTED MANUSCRIPT
Table 4: Factors that affect survival of patients with breast cancer, startified by tumor site origin.
Nodal stage Grade Estrogen positive Progesterone positive Chemotherapy Radiation therapy Hormonal therapy
Breast, lowerouter quadrant 1.06 (1.03, 1.10) 1.87 (1.09, 3.22) 0.86 (0.41, 1.81) 1.78 (1.00, 3.18) 0.93 (0.50, 1.72) 1.48 (0.38, 5.75) 0.73 (0.23, 2.28) 0.93 (0.34, 2.56) 0.47 (0.17, 1.34) 0.52 (0.23, 1.16)
Bold font represents statistically signficant outcomes. Outcomes are Hazard ratio (95%
AC C
EP
Confidence Interval).
Breast, overlapping lesion
RI PT
Breast, lowerinner quadrant 1.05 (1.03, 1.08) 1.62 (1.00, 2.64) 0.77 (0.45, 1.32) 1.31 (0.78, 2.18) 2.73 (1.50, 4.97) 0.76 (0.24, 2.43) 2.58 (0.89, 7.47) 1.63 (0.73, 3.66) 0.83 (0.39, 1.76) 0.35 (0.17, 0.71)
SC
Sentinel
1.07 (1.04, 1.09) 2.03 (1.40, 2.94) 1.40 (0.89, 2.21) 1.45 (0.87, 2.41) 1.38 (0.85, 2.22) 0.93 (0.26, 3.33) 1.85 (0.64, 5.40) 0.54 (0.25, 1.17) 0.38 (0.18, 0.83) 0.61 (0.28, 1.35)
Breast, upperinner quadrant 1.06 (1.03, 1.08) 2.07 (1.45, 2.96) 0.72 (0.44, 1.19) 0.80 (0.49, 1.29) 0.98 (0.63, 1.53) 0.98 (0.39, 2.49) 0.74 (0.35, 1.59) 1.18 (0.63, 2.24) 0.78 (0.41, 1.48) 0.54 (0.30, 0.98)
M AN U
Tumor stage
Breast, central portion
TE D
Age at diagnosis
Breast, upperouter quadrant 1.03 (1.02, 1.05) 1.91 (1.58, 2.30) 0.97 (0.76, 1.24) 1.09 (0.89, 1.34) 1.26 (0.98, 1.62) 1.53 (0.94, 2.48) 0.98 (0.64, 1.50) 0.82 (0.59, 1.16) 0.66 (0.47, 0.94) 0.35 (0.25, 0.49)
1.02 (1.01, 1.03) 2.16 (1.72, 2.71) 0.94 (0.72, 1.23) 1.23 (1.01, 1.49) 1.17 (0.87, 1.56) 0.80 (0.43, 1.49) 0.67 (0.40, 1.13) 0.46 (0.31, 0.68) 0.95 (0.65, 1.39) 0.53 (0.33, 0.83)
ACCEPTED MANUSCRIPT
Table 5: Factors that affect progression free survival of patients with breast cancer, startified by tumor site origin. Statistically significant differences are bolded.
Nodal stage Grade Estrogen positive Progesterone positive Chemotherapy Radiation therapy Hormonal therapy
Breast, lowerouter quadrant 1.04 (1.01, 1.07) 1.87 (1.17, 2.98) 0.78 (0.40, 1.50) 1.36 (0.80, 2.30) 1.04 (0.61, 1.78) 1.54 (0.46, 5.15) 0.77 (0.29, 2.07) 0.90 (0.38, 2.15) 0.50 (0.21, 1.22) 0.61 (0.30, 1.25)
EP
Bold font represents statistically signficant outcomes. Outcomes are Hazard ratio (95%
AC C
Confidence Interval).
Breast, overlapping lesion
RI PT
Breast, lowerinner quadrant 1.04 (1.02, 1.07) 1.60 (1.02, 2.51) 0.74 (0.45, 1.21) 1.35 (0.84, 2.18) 2.54 (1.48, 4.34) 1.03 (0.36, 2.94) 1.75 (0.70, 4.36) 1.52 (0.74, 3.14) 0.85 (0.42, 1.70) 0.39 (0.20, 0.76)
SC
Sentinel
1.05 (1.03, 1.07) 1.93 (1.36, 2.72) 1.53 (0.99, 2.36) 1.46 (0.90, 2.37) 1.20 (0.77, 1.86) 1.23 (0.39, 3.91) 1.24 (0.48, 3.22) 0.49 (0.24, 1.00) 0.38 (0.19, 0.78) 0.44 (0.21, 0.91)
Breast, upperinner quadrant 1.04 (1.02, 1.06) 2.01 (1.44, 2.79) 0.88 (0.55, 1.41) 0.78 (0.50, 1.21) 0.89 (0.59, 1.35) 0.81 (0.34, 1.89) 0.68 (0.33, 1.39) 1.20 (0.67, 2.16) 0.63 (0.34, 1.18) 0.59 (0.33, 1.05)
M AN U
Tumor stage
Breast, central portion
TE D
Age at diagnosis
Breast, upperouter quadrant 1.03 (1.02, 1.04) 1.79 (1.49, 2.13) 1.00 (0.80, 1.25) 1.22 (1.01, 1.49) 1.29 (1.02, 1.63) 1.74 (1.11, 2.72) 0.84 (0.57, 1.25) 0.79 (0.57, 1.09) 0.54 (0.39, 0.75) 0.37 (0.27, 0.51)
1.02 (1.00, 1.03) 1.89 (1.54, 2.32) 0.93 (0.73, 1.19) 1.24 (1.04, 1.48) 1.19 (0.92, 1.55) 0.97 (0.55, 1.70) 0.68 (0.43, 1.08) 0.59 (0.41, 0.84) 0.98 (0.69, 1.40) 0.63 (0.42, 0.95)
ACCEPTED MANUSCRIPT
Table 6: Patterns of metastatis at the time of diagnosis for patients with breast cancer stratified
RI PT
by tumor site origin.
Metastasis to Liver 11
Metastasis to Lung 14
Metastasis to Brain 4
1
0
0
0
Breast, central portion
8
1
Breast, upper-inner quadrant Breast, lower-inner quadrant Breast, lower-outer quadrant Breast, axillary tail
2
2
4 9 0 45
AC C
EP
TE D
Breast, overlapping lesion
2
0
1
0
1
0
0
4
2
0
0
1
0
21
20
2
M AN U
Breast, upper-outer quadrant Nipple
SC
Metastasis to Bone 27
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT