close

Вход

Забыли?

вход по аккаунту

?

ijc.31127

код для вставкиСкачать
Endometrial carcinoma recurrence according to race and ethnicity: An NRG
Oncology/Gynecologic Oncology Group 210 Study
AS Felix
1
2
3
4
5
4
6
7
, TM Brasky , DE Cohn , DG Mutch , WT Creasman , PH Thaker , JL Walker , RG Moore , SB
8
9
10
11
12
13
14
8
Lele , SR Guntupalli , LS Downs , CI Nagel , JF Boggess , ML Pearl , OB Ioffe , W Deng , DS
15
16
Miller , LA Brinton
Division of Epidemiology, Ohio State University College of Public Health, Columbus, OH;
Felix.20@osu.edu
Division of Cancer Prevention and Control, Ohio State University College of Medicine, Columbus, OH;
Theodore.Brasky@osumc.edu
Division of Gynecologic Oncology, Ohio State University College of Medicine, Columbus, OH;
David.Cohn@osumc.edu
Washington University School of Medicine, St. Louis, MO; mutchd@wudosis.wustl.edu [DGM];
thakerp@wudosis.wustl.edu [PHT]
Department of Obstetrics and Gynecology, Medical University of South Carolina, Charleston, SC;
creasman@musc.edu
Department of Obstetrics and Gynecology, University of Oklahoma, Oklahoma City, OK, USA; joanwalker@ouhsc.edu
Program in Women’s Oncology, Department of Obstetrics and Gynecology, Women and Infants
Hospital/Brown University, Providence, RI; richard_moore@urmc.rochester.edu
Gynecologic Oncology [SBL] & NRG Statistics and Data Management Center [WD], Roswell Park Cancer
Institute, Buffalo, NY; shashi.lele@roswellpark.org [SBL]; wdeng@gogstats.org [WD]
Gynecologic Oncology, University of Colorado Cancer Center, Aurora, CO;
saketh.guntupalli@ucdenver.edu
Gynecologic Oncology, University of Minnesota, Minneapolis, MN; downs008@umn.edu
Gynecologic Oncology, Case Western Reserve University, Cleveland, OH; christa.nagel@uhhospitals.org
Gynecologic Oncology Program, University of North Carolina, Chapel Hill, NC;
john_boggess@med.unc.edu
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process which may lead to
differences between this version and the Version of Record. Please cite this article as an
‘Accepted Article’, doi: 10.1002/ijc.31127
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Gynecologic Oncology, State University of New York at Stony Brook, Stony Brook, NY;
Michael.Pearl@stonybrookmedicine.edu
Anatomical Pathology, University of Maryland, College Park, MD; oioffe@fpi.umaryland.edu
Department of Obstetrics & Gynecology, University of Texas Southwestern Medical Center, Dallas, TX;
David.Miller@UTSouthwestern.edu
Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health,
Bethesda, MD; brintonl@exchange.nih.gov
Running title: Race and endometrial carcinoma recurrence
Keywords: Uterus Neoplasm; Racial disparities; Outcomes, Socioeconomic status; Tumor heterogeneity
Corresponding author: Ashley Felix, PhD, 1841 Neil Avenue, 346 Cunz Hall, Columbus, OH 43210,
Email: Felix.20@osu.edu;Telephone: 614-688-1477
Financial Support: This study was supported by National Cancer Institute grants to the Gynecologic
Oncology Group Administrative Office (CA 27469), the Gynecologic Oncology Group Statistical and Data
Center (CA 37517) and the NRG Oncology Grant number: 1 U10 CA180822 and U10 CA180868. In
addition, this research was supported in part by funds provided by the intramural research program of the
National Cancer Institute, National Institutes of Health.
Potential Conflicts of Interest: Dr. Cohn reported serving in an advisory/consulting role for Oncology
Analytics; Dr. Mutch reported serving in an advisory/consulting role for Lilly and paid to participate in a
speakers’ bureau for AstraZeneca; Dr. Thaker reported serving in an advisory/consulting role for Celsion
and received funding from Merck; Dr. Moore has received research funding from Angle Inc, and Fujirebio
Diagnostics; Dr. Lele reported serving in an advisory/consulting role and receiving honoraria from
Genetech; Dr. Guntupalli reported serving in an advisory/consulting role and receiving honoraria from
Genentech and Janssen and receiving funding from Bristol-Myers Squibb; and Dr. Pearl receives
royalties from our research foundation on behalf of VitaTex. This work has nothing to do with this
research paper or GOG 210 in general.
Number of figures/tables: 2/3
2
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 2 of 25
Page 3 of 25
International Journal of Cancer
NOVELTY AND IMPACT STATEMENT
Although racial disparities in outcomes among women with endometrial cancer (EC) are welldocumented, racial differences in recurrence risk according to tumor characteristics are not well defined.
Our study shows that within certain subgroups defined by tumor characteristics, non-Hispanic black and
Hispanic women have a higher risk of EC recurrence than non-Hispanic white women. Clinicians treating
women with EC should be aware of racial differences in recurrence risk, particularly among women
diagnosed with certain tumor characteristics.
3
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
ABSTRACT
Non-Hispanic black (NHB) women are more likely to experience an endometrial carcinoma (EC)
recurrence compared with Non-Hispanic white (NHW) women. The extent to which tumor characteristics,
socioeconomic status (SES), and treatment contribute to this observation is not well defined. In the NRG
Oncology/Gynecology Oncology Group (GOG) 210 Study we evaluated associations between race and
ethnicity with EC recurrence according to tumor characteristics with adjustment for potential confounders.
Our analysis included 3,199 non-Hispanic white, 532 non-Hispanic black and 232 Hispanic women with
EC. Recurrence was documented during follow-up. We used Cox regression to estimate hazard ratios
(HRs) and 95% confidence intervals (CIs) for associations between race and ethnicity with EC recurrence
in models stratified by histologic subtype (low-grade endometrioid, high-grade endometrioid, serous,
mixed cell, carcinosarcoma, clear cell) or stage (I, II, III) and adjusted for age, SES, body mass index,
smoking status, and treatment. In histologic subtype-stratified models, higher EC recurrence was noted in
NHB women with low-grade endometrioid (HR=1.94, 95% CI=1.21-3.10) or carcinosarcomas (HR=1.66,
95% CI=0.99-2.79) compared with NHWs. In stage-stratified models, higher EC recurrence was noted
among NHB women with stage I (HR=1.48, 95% CI=1.06-2.05) and Hispanic women with stage III
disease (HR=1.81, 95% CI=1.11-2.95). Our observations of higher EC recurrence risk among NHB and
Hispanic women, as compared with NHW women, were not explained by tumor characteristics, SES,
treatment or other confounders. Other factors, such as racial differences in tumor biology or other patient
factors, should be explored as contributors to racial disparities in EC recurrence.
Word count: 248/250
4
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 4 of 25
Page 5 of 25
International Journal of Cancer
INTRODUCTION
Endometrial carcinoma (EC) recurs in 2-15% of women with early stage disease and this
proportion can reach as high as 50% in women with advanced stage or non-endometrioid histologic
1
1
subtypes . Recurrence contributes to worse survival , and compared with white women, black women
2
are more likely to experience an EC recurrence . This observation has been attributed to a higher
3
prevalence of adverse tumor characteristics, receipt of less aggressive treatment and lower
4
socioeconomic status (SES) among black versus white women. Adjustment for tumor characteristics
attenuates associations between race and EC recurrence in some studies
5, 6
but not others
7-9
. Adding to
this unclear picture are heterogeneous study populations and lack of adjustment for factors associated
with EC recurrence. Further, the literature is hindered by relatively low numbers of black women with EC,
ranging from 14 to 257 black EC cases, limiting our understanding of the race-EC recurrence relationship.
Finally, studies have not examined recurrence risk among Hispanic women, a growing demographic in
the United States.
To clarify the effects of race and ethnicity on EC recurrence, we examined this relationship
according to clinically relevant tumor characteristics in the NRG Oncology/Gynecologic Oncology Group
(GOG) 210 Study, an observational study that included a large number of women with rare, aggressive
endometrial tumors. We also assessed the hypothesis that SES mediates the association between race,
ethnicity, and EC recurrence. In addition to inclusion of diverse histologic subtypes, this study has the
advantage of carefully collected epidemiological information and clinical outcomes along with a central
pathology review of tumor characteristics.
METHODS
Study population
Full details regarding the NRG Oncology/GOG 210 Study design have been published elsewhere
10
. Briefly, the NRG Oncology/GOG 210 Study was conducted from September 22, 2003 to December 1,
2011 at 62 U.S. institutions. Prior to surgery (hysterectomy, bilateral salpingo-oophorectomy, and lymph
node sampling at the discretion of the treating provider), consenting patients completed a selfadministered questionnaire that collected demographic and epidemiologic information
5
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
11
. Of 6,124 women
International Journal of Cancer
enrolled, 632 (10%) did not complete questionnaires and were excluded from further analyses. We
excluded women for the following reasons: incomplete surgical staging (n=20), final diagnosis not EC
(n=53), benign diagnoses (n=6), diagnosis of a second primary (n=2), misclassified pathologic diagnosis
based on central pathology review (n=49), inadequate material for pathology review (n=22), protocol
deviations (n=17), improper pre-protocol treatment (n=1), missing grade (n=23), mucinous tumors
(n=18), unusual histologic types (including squamous cell, undifferentiated, and de-differentiated
histologies) (n=111), and missing stage (n=5). We further excluded women with self-reported race and
ethnicity other than non-Hispanic black (NHB), non-Hispanic white (NHW), or Hispanic [Asian (n=95),
Native Hawaiian/Pacific Islander (n=12), American Indian (n=37), Other (n=46); N=231] and women with
stage IV disease, as they are not free of disease following primary surgery and thus, not at risk of
recurrence (n=236). Our analytic sample included 4,698 EC patients. This study was approved by
Institutional Review Boards at the National Cancer Institute and participating study centers. All
participants provided informed consent prior to participation.
Risk factor assessment
Questionnaires assessed demographic characteristics (age, race, annual income, highest level
education attained) and established EC risk factors including height, weight, reproductive factors,
diabetes, smoking status, oral contraceptive use, menopausal hormone use, tamoxifen use, and history
of breast cancer.
Tumor characteristics and outcome assessment
Pathology information was available from participating NRG Oncology/GOG institutions and
through specialized reviews
11
performed by the NRG Oncology/GOG Pathology Committee. Women with
the following International Classification of Diseases for Oncology (ICD-O-3) morphology codes were
included: endometrioid adenocarcinomas (8380-8383), adenocarcinoma with squamous metaplasia
(8570), adenosquamous carcinomas (8560), mucinous adenocarcinoma (8480, 8481), adenocarcinoma,
not otherwise specified (NOS, 8140), serous (8440-8441, 8460-8461), clear cell (8310), mixed cell
(8323), and carcinosarcoma (8950-8951, 8980). Women with endometrioid, adenocarcinoma with
6
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 6 of 25
Page 7 of 25
International Journal of Cancer
squamous metaplasia, adenosquamous carcinomas, mucinous adenocarcinoma, and adenocarcinoma
NOS were categorized as one broad category (n=3,392), which we further classified on the basis of
grade: grades 1 and 2 were considered low-grade endometrioid adenocarcinoma (n=2,778) and women
with grade 3 were considered high-grade endometrioid adenocarcinoma (n=614). Women with serous
(n=712), mixed cell (n=556), carcinosarcoma (n=331), or clear cell (n=174) tumors were considered
separately. Depth of myometrial invasion, stage according to International Federation for Gynecology and
12
Obstetrics 1988 criteria , pelvic and/or aortic lymph node involvement, peritoneal cytology results and
extra-uterine sites of metastasis were recorded. Information on recurrences, defined as evidence of
disease following complete response to primary therapy, vital status, cause of death, and date of events
was obtained through active follow-up of all patients, until death, loss to follow-up, or a total of 10 years of
follow-up. All patients were followed every three months for the first two years, every six months for the
next three years, and yearly for the next five years.
Statistical analysis
Epidemiological, tumor, and treatment characteristics were compared between NHW, NHB, and
Hispanic EC patients using chi-square tests. Kaplan-Meier estimates and log-rank tests were used to
compare survival distributions according to race and ethnicity stratified by histologic subtype and stage.
We used Cox proportional hazards regression models to estimate hazard ratios (HRs) and 95%
confidence intervals (CIs) for associations between race and EC recurrence. Models were stratified by
histologic subtype or stage and initially adjusted for age at diagnosis (<54, 55-59, 60-64, 65-69, ≥70),
adjuvant therapy (none, chemotherapy, radiation, chemotherapy and radiation), and either stage (I, II, III)
or histologic subtype (low-grade endometrioid, high-grade endometrioid, serous, mixed cell,
carcinosarcoma, clear cell), depending on the stratification variable. Epidemiological characteristics were
explored as confounders and were retained in models when inclusion of the factor changed estimates in
the minimally adjusted model by more than 10%. Final models were adjusted for age at diagnosis,
adjuvant treatment, histological subtype or stage, BMI, and smoking status. Missing values for each
covariate were modeled as a separate category. Deaths were treated as censoring events and
7
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Page 8 of 25
proportional hazards were assessed by including an interaction term between follow-up time (log scale)
and race and evaluating the Wald p-value
13
.
We used the Baron and Kenny approach
14
to determine the role of SES indicators (e.g. annual
income and education) in mediating the relationship between race and ethnicity with EC recurrence in
stratified models. In order for SES to mediate the relationship between race and EC recurrence, the
following criteria are needed: 1.) significant association between race and ethnicity with EC recurrence;
2.) significant association between annual income/education and race; and 3.) significant association
between annual income/education and EC recurrence in a model including race and ethnicity.
We performed several sensitivity analyses. First, to account for the possibility that deaths due to
EC were due to undetected recurrences, we examined progression, or time to recurrence or EC death as
15
the outcome. Second, we treated deaths from EC as competing events using the Fine-Gray model .
Third, as most EC recurrences occur within three years of diagnosis, we excluded women with follow-up
time less than one year, two years, and three years to account for the possibility that women did not have
enough follow-up time to experience the outcome of interest. Finally, we compared associations between
race and ethnicity with EC recurrence stratified by enrollment period (i.e. 2003-2007 vs. 2007-2011). We
tested the interaction by including a multiplicative interaction term between race and ethnicity and a
binary variable indicating the enrollment period. Statistical analyses were performed using SAS (version
9.3, SAS Institute, Cary, NC, USA) and Stata software (version 11, STATA Corp., Texas, USA). All P
values were two-sided.
RESULTS
Of 4,698 study participants, the majority were NHW (n=3,911), followed by NHB (n=532), and
Hispanic (any race, n=255). Distributions of epidemiological, tumor, and treatment characteristics
according to race and ethnicity are shown in Table 1. Compared to NHB or NHW women, Hispanic
women were significantly younger at EC diagnosis and more likely to: have less than high school
education, have lower annual incomes, have older ages at menarche, be never smokers, be never users
of oral contraceptives, and be premenopausal. Compared to NHW or Hispanic women, NHB women were
more likely to: be obese, be multiparous, have a history of diabetes, and be non-users of menopausal
8
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 9 of 25
International Journal of Cancer
hormones. As expected, NHB women more commonly had aggressive tumor characteristics, including
positive pelvic and/or aortic lymph node involvement, positive peritoneal biopsy, advanced stage, and
histologic subtypes other than low-grade endometrioid compared to NHW and Hispanic women. Adjuvant
chemotherapy, either with or without radiation, was more common among NHB (22%, 15%) than among
Hispanic (16.5%, 10.2%) or NHW women (12.6%, 12.2%).
After a median follow-up of 5 years (range: 1 day – 9.8 years), 753 recurrences occurred and 310
women died without a documented EC recurrence and were censored in the primary analysis. NHB
women had higher EC recurrence risk overall (log-rank P < 0.001, data not shown), and among women
with low-grade endometrioid (log-rank P = 0.02) or carcinosarcoma (log-rank P = 0.06) (Figure 1). In
Kaplan-Meier plots stratified by stage, we observed worse EC recurrence among NHB women for stage I
and II disease, while Hispanic women had worse EC recurrence among women with stage III disease
(Figure 2).
In the overall model NHB (HR=1.36, 95% CI=1.11-1.66, data not shown), but not Hispanic
women (HR=1.16, 95% CI=0.82-1.65) had higher EC recurrence risk compared with NHW women. We
observed higher EC recurrence among NHB women with low-grade endometrioid (HR=2.06, 95%
CI=1.30-3.27) or carcinosarcoma (HR=1.70, 95% CI=1.03-2.82) compared with NHW women (Table 2).
In addition, higher stage was associated with EC recurrence risk across all histologic subtypes, while
adjuvant radiation or adjuvant chemotherapy was significantly associated with higher EC recurrence
among women with low-grade endometrioid EC. The addition of annual income and education to the
multivariable models slightly attenuated estimates for race and EC recurrence: low-grade endometrioid
(HR=1.94, 95% CI=1.21-3.10), and carcinosarcoma (HR=1.66, 95% CI=0.99-2.79). Neither annual
income nor education were significantly associated with EC recurrence in these models, violating criterion
three of the Baron-Kenny method.
Table 3 shows associations between race and EC recurrence stratified by stage. The higher EC
recurrence among NHB compared to NHW women was primarily driven by strong relations for stage I
cancers (HR=1.57, 95% CI=1.14-2.16). The addition of annual income and education slightly attenuated
the associations for stage I (HR=1.46, 95% CI=1.05-2.03). Among women with stage III disease, Hispanic
9
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Page 10 of 25
women had an 81% higher risk of EC recurrence compared to NHW women (95% CI=1.11-2.95). Neither
income nor education were significantly associated with EC recurrence in these models.
We did not examine the association between race and ethnicity with EC recurrence in strata
defined by both histologic subtype and stage (e.g. low-grade endometrioid and stage I) due to low
numbers of events. We conducted several sensitivity analyses. We observed similar associations
between race and ethnicity with EC progression (EC recurrence and EC-specific death combined, data
not shown). In addition, modeling EC-specific death as a competing event did not materially change our
findings. Finally, because recruitment criteria changed in 2007, we evaluated the association between
race and ethnicity with EC recurrence separately among women enrolled before and after 2007.
Associations were similar in magnitude to the overall model. Analyses using the Fine-Gray model of
competing risks produced similar estimates as the Cox proportional hazards regression models (data not
shown).
DISCUSSION
Black women have a 93% higher mortality rate after EC diagnosis compared with white women,
16
demonstrating one of the most prominent racial disparities among common cancers . Our analysis
extends the literature on racial disparities in EC by demonstrating higher recurrence risk among NHB
women diagnosed with low-grade endometrioid, carcinosarcoma, and stage I tumors compared with their
NHW counterparts, while Hispanic women with stage III disease experienced higher recurrence risk. We
did not observe racial differences in recurrence risk for women in subgroups characterized by other tumor
characteristics, suggesting some specificity in the association.
Recurrence is an important contributor to death among EC patients. Typically occurring within 2
1
years of the initial diagnosis, recurrent ECs are frequently refractory to treatment . As such, surveillance
1
following completion of primary treatment focuses on symptoms of recurrence . Tumor characteristics
are the main tools guiding follow-up care based on the perceived risk of recurrence in a specific patient.
Broadly speaking, women with early stage or low-grade endometrioid cancers are considered to be at
lower risk of recurrence than those with higher-grade and advanced stage disease and thus have less
10
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 11 of 25
International Journal of Cancer
1
frequent surveillance over time . Our results suggest this surveillance strategy may be inadequate for
NHB women perceived to be at a low risk for recurrence.
Several studies have examined racial differences in EC recurrence with inconsistent findings
5-9
.
Two studies reported higher EC recurrence among black women; however, adjustment for tumor
characteristics attenuated the association
5, 6
while others have observed significant associations between
race and EC recurrence when adjusting for tumor characteristics or restricting to defined subgroups of
cases. Among 14 non-Caucasian and 84 Caucasian women with stage IIIA EC, Havrilesky et al.
8
observed a five times higher risk of recurrence among non-Caucasian women in models adjusted for age,
BMI, treatment, histology and number of lymph nodes removed. In a study of 44 black and 164 white
9
women with stages I-II EC, Ozen and colleagues observed a two-fold higher recurrence risk among
black women in models adjusted for histology. Interestingly, histology was not a predictor of recurrence in
this model. In GOG 137, a randomized trial of the effect of estrogen replacement therapy (ERT) vs.
placebo on recurrence and overall survival
17
, black women with early stage disease had an 11 times
7
higher EC recurrence compared with white women in the subgroup randomized to receive ERT . The
confidence interval reported in this study is imprecise, reflective of five recurrences among 59 black EC
cases in the ERT arm.
Our study agrees with the literature suggesting that NHB women are at higher risk of an EC
recurrence. However, our stratification by histology and stage within a large study population allowed us
to discern independent associations between race and ethnicity with EC recurrence. Notably, among
women with stage I disease or low-grade endometrioid tumors, characteristics indicative of favorable
outcomes, NHB experienced higher recurrence risk than NHW women. These findings are likely not
attributable to differences in SES or receipt of adjuvant treatment, as we included these potential
confounders in our models. Importantly, we only characterized SES along two dimensions, education and
income; however these are commonly accepted SES indicators
18
. In terms of treatment, all women in this
study underwent hysterectomy with bilateral oophorectomy; however, adjuvant treatment was prescribed
at the discretion of the treating clinician. Adjuvant treatment, an adjustment factor in our analyses, was
more frequent among NHB women in the overall study population and among women with stage I tumors,
reflective of a greater prevalence of adverse tumor characteristics among NHB women.
11
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Page 12 of 25
We speculate that despite stratifying by tumor characteristics and adjusting for treatment, SES,
and a number of other potential confounders, there is additional heterogeneity in these subgroups that we
cannot account for with our data. In particular, we cannot account for racial differences in molecular
profiles of stage 1 or low-grade endometrioid tumors that could contribute to the higher recurrence risk
among NHB women. To date, the hypothesis that NHB women have molecularly more aggressive tumors
has not been borne out by studies investigating differences in global gene expression
19, 20
; however,
some have identified racial differences in expression of candidate biomarkers. For example, mutations in
TP53 and ERBB2 are more frequent in ECs from black compared to white women, while methylation of
ribosomal RNA genes and mutations in PTEN are less prevalent among black as compared to white
women
21-25
, reflecting biologically more aggressive tumors. Whether these alterations differ within stage
or histology categories and are independently related to increased EC recurrence risk would need to be
examined in a setting with appropriate clinical data, tumor specimens, and a large number of EC cases.
In addition, there could be racial differences in behaviors following the cancer diagnosis, such as patient
adherence to follow-up schedules or changes in lifestyle factors after the cancer diagnosis that could
affect recurrence risk.
Several other findings warrant mention. We noted a higher recurrence risk among Hispanic
women diagnosed with stage III disease compared with similarly-staged NHW women. Again, this
observation is unlikely due to differences in SES, treatment, histology, or other confounders for which we
controlled. More likely, there are unaccounted differences in patient factors that underlie this association.
Additional studies examining EC recurrence among Hispanic patients are clearly needed. Also, among
women with low-grade endometrioid disease, receipt of adjuvant radiation or chemotherapy was
associated with increased recurrence risk. In this subgroup, 16% of women received adjuvant
radiotherapy and 4% received adjuvant chemotherapy. According to the National Comprehensive Cancer
Network (NCCN), use of adjuvant therapy among women with grades 1 or 2 endometrioid disease is
indicated in the presence of adverse risk factors, including age, positive lymphovascular space invasion,
tumor size, depth of invasion, and lower uterine segment involvement
26
. Although we adjusted for age
and stage, our inability to adjust for the other adverse characteristics, due to lack of data, may underlie
12
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 13 of 25
International Journal of Cancer
the increased risk of recurrence associated with receipt of adjuvant therapy among women with low-grade
endometrioid EC.
Our study has several strengths including the prospective design, long follow-up, central
pathology review, and large sample size. Despite the large overall sample size, we were unable to
explore race-EC recurrence associations among subgroups cross-classified by stage and histologic
subtype. The most clinically relevant ECs are the early-stage, low-grade endometrioid tumors given their
numerical predominance. Although our study included approximately 2,200 women with these
characteristics, a small number of recurrences limited our ability to examine this subgroup. Like previous
studies, we were unable to examine associations among women of other racial minority groups.
Nonetheless, this case-series is one of the largest and includes a diverse array of well-characterized
endometrial tumors among NHB, NHW, and Hispanic women.
In summary, we observed racial differences in recurrence following an EC diagnosis. Education
and income disparities are unlikely to account for these differences, nor are differences in the receipt of
treatment. Our results suggest that follow-up care plans should take into account higher EC recurrence
risk among NHB women diagnosed with less aggressive tumors (i.e. low-grade endometrioid or stage I
tumors). Future studies should also explore the molecular characteristics of recurrent ECs to increase our
understanding of the biology and develop biomarkers for risk stratification.
13
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Acknowledgements
This work is funded by National Cancer Institute grants U10 CA180868 (NRG Oncology
Operations), U10 CA180822 (NRG SDMC), CA27469 (GOG Administrative Office and Tissue Bank), and
CA37517 (GOG Statistical and Data Center). The authors would like to acknowledge the significant
contributions of the late Dr. D. Scott McMeekin who worked extensively on the GOG-210 study.
The following institutions participated in this study: Roswell Park Cancer Institute, University of
Alabama at Birmingham, Duke University Medical Center, Abington Memorial Hospital, Walter Reed Army
Medical Center, Wayne State University, University of Minnesota Medical School, Northwestern
University, University of Mississippi, University of Colorado-Anschutz Cancer Pavilion, University of
California at Los Angeles, Fred Hutchinson Cancer Research Center, Penn State Milton S. Hershey
Medical Center, University of Cincinnati, University of North Carolina, University of Iowa Hospitals and
Clinics, University of Texas Southwestern Medical Center, Indiana University Medical Center, Wake
Forest University Health Sciences, University of California Medical Center at Irvine – Orange Campus,
Magee Women’s Hospital – University of Pittsburgh Medical Center, University of New Mexico, Cleveland
Clinic Foundation, State University of New York at Stony Brook, Washington University School of
Medicine, Cooper Hospital/University Medical Center, Columbus Cancer Council/Ohio State University,
University of Massachusetts Memorial Health Care, Fox Chase Cancer Center, Women's Cancer Center
of Nevada, University of Oklahoma Health Sciences Center, University of Virginia, University of Chicago,
Mayo Clinic, Case Western Reserve University, Moffitt Cancer Center and Research Institute, Yale
University, University of Wisconsin Hospital, Women and Infants’ Hospital of Rhode Island, The Hospital
of Central Connecticut at New Britain General, GYN Oncology of West Michigan, PLLC and Community
Clinical Oncology Program.
14
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 14 of 25
Page 15 of 25
International Journal of Cancer
References
1. Salani R, Khanna N, Frimer M, Bristow RE, Chen LM. An update on post-treatment
surveillance and diagnosis of recurrence in women with gynecologic malignancies: Society of
Gynecologic Oncology (SGO) recommendations. Gynecol Oncol 2017;146: 3-10.
2. Collins Y, Holcomb K, Chapman-Davis E, Khabele D, Farley JH. Gynecologic cancer
disparities: a report from the Health Disparities Taskforce of the Society of Gynecologic Oncology.
Gynecol Oncol 2014;133: 353-61.
3. Randall TC, Armstrong K. Differences in treatment and outcome between African-American
and white women with endometrial cancer. J Clin Oncol 2003;21: 4200-6.
4. Setiawan VW, Yang HP, Pike MC, McCann SE, Yu H, Xiang YB, Wolk A, Wentzensen N,
Weiss NS, Webb PM, van den Brandt PA, van de Vijver K, et al. Type I and II endometrial cancers: have
they different risk factors? J Clin Oncol 2013;31: 2607-18.
5. Elshaikh MA, Munkarah AR, Robbins JR, Laser BS, Bhatt N, Cogan C, Siddiqui F. The impact
of race on outcomes of patients with early stage uterine endometrioid carcinoma. Gynecol Oncol
2013;128: 171-4.
6. Ko EM, Walter P, Clark L, Jackson A, Franasiak J, Bolac C, Havrilesky L, Secord AA, Moore
DT, Gehrig PA, Bae-Jump VL. The complex triad of obesity, diabetes and race in Type I and II
endometrial cancers: prevalence and prognostic significance. Gynecol Oncol 2014;133: 28-32.
7. Maxwell GL, Tian C, Risinger JI, Hamilton CA, Barakat RR. Racial disparities in recurrence
among patients with early-stage endometrial cancer: is recurrence increased in black patients who
receive estrogen replacement therapy? Cancer 2008;113: 1431-7.
8. Havrilesky LJ, Secord AA, O'Malley DM, Broadwater G, Bae-Jump V, Cohn DE, Gehrig PA.
Multicenter analysis of recurrence and survival in stage IIIA endometrial cancer. Gynecol Oncol 2009;114:
279-83.
9. Ozen A, Falchook AD, Varia MA, Gehrig P, Jones EL. Effect of race and histology on patterns
of failure in women with early stage endometrial cancer treated with high dose rate brachytherapy.
Gynecol Oncol 2015;138: 429-33.
10. Felix AS, Brinton LA, McMeekin DS, Creasman WT, Mutch D, Cohn DE, Walker JL, Moore
RG, Downs LS, Soslow RA, Zaino R, Sherman ME. Relationships of Tubal Ligation to Endometrial
Carcinoma Stage and Mortality in the NRG Oncology/ Gynecologic Oncology Group 210 Trial. J Natl
Cancer Inst 2015;107.
11. Brinton LA, Felix AS, McMeekin DS, Creasman WT, Sherman ME, Mutch D, Cohn DE,
Walker JL, Moore RG, Downs LS, Soslow RA, Zaino R. Etiologic heterogeneity in endometrial cancer:
Evidence from a Gynecologic Oncology Group trial. Gynecol Oncol 2013;129: 277-84.
12. Zaino RJ. FIGO staging of endometrial adenocarcinoma: a critical review and proposal.
International journal of gynecological pathology : official journal of the International Society of
Gynecological Pathologists 2009;28: 1-9.
15
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
13. Zhou B, Fine J, Laird G. Goodness-of-fit test for proportional subdistribution hazards model.
Statistics in medicine 2013;32: 3804-11.
14. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological
research: conceptual, strategic, and statistical considerations. Journal of personality and social
psychology 1986;51: 1173-82.
15. Fine JP, Gray RJ. A Proportional Hazards Model for the Subdistribution of a Competing Risk.
Journal of the American Statistical Association 1999;94: 13.
16. Jemal A, Ward EM, Johnson CJ, Cronin KA, Ma J, Ryerson B, Mariotto A, Lake AJ, Wilson R,
Sherman RL, Anderson RN, Henley SJ, et al. Annual Report to the Nation on the Status of Cancer, 19752014, Featuring Survival. J Natl Cancer Inst 2017;109.
17. Barakat RR, Bundy BN, Spirtos NM, Bell J, Mannel RS. Randomized double-blind trial of
estrogen replacement therapy versus placebo in stage I or II endometrial cancer: a Gynecologic
Oncology Group Study. J Clin Oncol 2006;24: 587-92.
18. Berkman L, Kawachi I, Glymour M. Social Epidemiology, Second ed.: Oxford University
Press, 2014.
19. Ferguson SE, Olshen AB, Levine DA, Viale A, Barakat RR, Boyd J. Molecular profiling of
endometrial cancers from African-American and Caucasian women. Gynecol Oncol 2006;101: 209-13.
20. Maxwell GL, Allard J, Gadisetti CV, Litzi T, Casablanca Y, Chandran U, Darcy KM, Levine
DA, Berchuck A, Hamilton CA, Conrads TP, Risinger JI. Transcript expression in endometrial cancers
from Black and White patients. Gynecol Oncol 2013;130: 169-73.
21. Kohler MF, Carney P, Dodge R, Soper JT, Clarke-Pearson DL, Marks JR, Berchuck A. p53
overexpression in advanced-stage endometrial adenocarcinoma. Am J Obstet Gynecol 1996;175: 124652.
22. Santin AD, Bellone S, Siegel ER, Palmieri M, Thomas M, Cannon MJ, Kay HH, Roman JJ,
Burnett A, Pecorelli S. Racial differences in the overexpression of epidermal growth factor type II receptor
(HER2/neu): a major prognostic indicator in uterine serous papillary cancer. Am J Obstet Gynecol
2005;192: 813-8.
23. Powell MA, Mutch DG, Rader JS, Herzog TJ, Huang TH, Goodfellow PJ. Ribosomal DNA
methylation in patients with endometrial carcinoma: an independent prognostic marker. Cancer 2002;94:
2941-52.
24. Maxwell GL, Risinger JI, Hayes KA, Alvarez AA, Dodge RK, Barrett JC, Berchuck A. Racial
disparity in the frequency of PTEN mutations, but not microsatellite instability, in advanced endometrial
cancers. Clin Cancer Res 2000;6: 2999-3005.
25. Clifford SL, Kaminetsky CP, Cirisano FD, Dodge R, Soper JT, Clarke-Pearson DL, Berchuck
A. Racial disparity in overexpression of the p53 tumor suppressor gene in stage I endometrial cancer. Am
J Obstet Gynecol 1997;176: S229-32.
16
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 16 of 25
Page 17 of 25
International Journal of Cancer
26. National Comprehensive Cancer Network. NCCN Guidelines for Treatment of Cancer by Site
(Uterine Neoplasms Version 2.2016).
17
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Figure legends
Figure 1. EC recurrence according to race and ethnicity in A) Low-grade endometrioid, B) High-grade
endometrioid, C) Serous, D) Mixed cell E) Carcinosarcoma, and F) Clear cell
Figure 2. EC recurrence according to race and ethnicity in A) Stage I, B) Stage II, C) Stage III, and D)
Stage IV
18
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 18 of 25
Page 19 of 25
International Journal of Cancer
Table 1. Epidemiological, tumor, and treatment characteristics of 3,911 Non-Hispanic White, 532 Non-Hispanic Black,
and 255 Hispanic women with endometrial carcinoma in the GOG 210 study
Non-Hispanic
White
(n=3,911)
Age at diagnosis
<54
55-59
60-64
65-69
≥70
Median (interquartile range)
Education
Less than high school
High school/GED
Some college/technical school
College graduate/beyond
Missing
Income
< $20,000
$20,000-$39,999
$40,000-$69,999
≥$70,000
Missing
Age at menarche
≤11
12
13
≥14
Missing
Parity
Nulliparous
1-2
≥3
Missing
2
BMI 1 year ago, kg/m
2
Normal (< 25 kg/m )
2
Overweight (25-30 kg/m )
2
Obese (≥ 30 kg/m )
Missing
Smoking status
Never
Current
Former
Missing
Diabetes
Race/Ethnicity
Non-Hispanic
Black
(n=532)
n (%)
Hispanic
1
p
(n=255)
<.0001
751 (19.2)
749 (19.2)
712 (18.2)
614 (15.7)
1085 (27.7)
63 (56, 70)
104 (19.6)
87 (16.4)
95 (17.9)
109 (20.5)
137 (25.8)
63.5 (57, 70)
90 (35.3)
44 (17.3)
47 (18.4)
31 (12.2)
43 (16.9)
59 (50, 66)
<.0001
320 (8.2)
1292 (33.0)
1102 (28.2)
1188 (30.4)
9 (0.2)
111 (20.9)
136 (25.6)
167 (31.4)
111 (20.9)
7 (1.3)
95 (37.3)
58 (22.8)
64 (25.1)
36 (14.1)
2 (0.8)
<.0001
685 (17.5)
861 (22.0)
962 (24.6)
954 (24.4)
449 (11.5)
209 (39.3)
121 (22.7)
91 (17.1)
47 (8.8)
64 (12.0)
118 (46.3)
51 (20.0)
33 (12.9)
25 (9.8)
28 (11.0)
0.03
815 (20.8)
1157 (29.6)
1030 (26.3)
808 (20.7)
101 (2.6)
123 (23.1)
126 (23.7)
124 (23.3)
126 (23.7)
33 (6.2)
52 (20.4)
69 (27.1)
56 (22.0)
65 (25.5)
13 (5.1)
740 (18.9)
1643 (42.0)
1404 (35.9)
124 (3.2)
56 (10.5)
182 (34.2)
267 (50.2)
27 (5.1)
40 (15.7)
93 (36.5)
112 (43.9)
10 (3.9)
<.0001
<.0001
879 (22.5)
874 (22.4)
1783 (45.6)
375 (9.6)
30 (5.6)
68 (12.8)
316 (59.4)
118 (22.2)
30 (11.8)
49 (19.2)
122 (47.8)
54 (21.2)
2472 (63.2)
253 (6.5)
1100 (28.1)
86 (2.2)
353 (66.4)
29 (5.5)
132 (24.8)
18 (3.4)
184 (72.2)
17 (6.7)
50 (19.6)
4 (1.6)
0.02
<.0001
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Never
Ever
Missing
Oral contraceptive use
Never
Ever
Missing
Menopausal status
Premenopausal
Postmenopausal
Missing
Menopausal hormone use
Never
Estrogen only
Progestin only
Estrogen + progestin
Missing
Tamoxifen use
Never
Ever
Missing
History of breast cancer
Never
Ever
Missing
Tumor characteristics
Histologic subtype
Low-grade endometrioid
High-grade endometrioid
Serous
Mixed cell
Carcinosarcoma
Clear cell
Stage
I
II
III
Myometrial invasion
Negative
Inner half
Outer half
Serosa
Pelvic lymph node involvement
No
Yes
2
Number of positive pelvic nodes, median (range)
Aortic lymph node involvement
No
Page 20 of 25
3002 (76.8)
730 (18.7)
179 (4.6)
337 (63.4)
159 (29.9)
36 (6.8)
171 (67.1)
70 (27.5)
14 (5.5)
1640 (41.9)
2171 (55.5)
100 (2.6)
190 (35.7)
311 (58.5)
31 (5.8)
121 (47.5)
121 (47.5)
13 (5.1)
330 (8.4)
3499 (89.5)
82 (2.1)
44 (8.3)
463 (87.0)
25 (4.7)
44 (17.3)
198 (77.7)
13 (5.1)
0.007
<.0001
<.0001
2601 (66.5)
291 (7.4)
172 (4.4)
808 (20.7)
39 (1.0)
452 (85.0)
36 (6.8)
9 (1.7)
31 (5.8)
4 (0.8)
199 (78.0)
21 (8.2)
11 (4.3)
22 (8.6)
2 (0.8)
0.23
3606 (92.2)
176 (4.5)
129 (3.3)
471 (88.5)
19 (3.6)
42 (7.9)
236 (92.6)
6 (2.4)
13 (5.1)
0.14
3529 (90.2)
296 (7.6)
86 (2.2)
475 (89.3)
35 (6.6)
22 (4.1)
236 (92.6)
11 (4.3)
8 (3.1)
2264 (57.9)
468 (12.0)
437 (11.2)
415 (10.6)
205 (5.2)
122 (3.1)
201 (37.8)
60 (11.3)
130 (24.4)
61 (11.5)
54 (10.2)
26 (4.9)
145 (56.9)
28 (11.0)
36 (14.1)
17 (6.7)
19 (7.5)
10 (3.9)
<.0001
<.0001
2949 (75.4)
275 (7.0)
687 (17.6)
345 (64.9)
58 (10.9)
129 (24.3)
190 (74.5)
28 (11.0)
37 (14.5)
0.06
972 (25.6)
1862 (49.0)
903 (23.7)
67 (1.8)
125 (24.0)
281 (53.8)
106 (20.3)
10 (1.9)
79 (32.0)
104 (42.1)
61 (24.7)
3 (1.2)
3294 (89.2)
400 (10.8)
2 (1, 3)
413 (83.6)
81 (16.4)
2 (1, 4)
210 (89.4)
25 (10.6)
3 (1, 5)
3316 (94.0)
410 (88.6)
202 (93.5)
0.43
0.17
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 21 of 25
International Journal of Cancer
Yes
213 (6.0)
53 (11.5)
2
Number of positive aortic nodes, median (range)
2 (1, 3)
2 (1, 4)
Peritoneal cytology
Negative
3356 (91.5)
446 (89.9)
Positive
312 (8.5)
50 (10.1)
Peritoneal biopsy
Negative
1587 (97.1)
281 (96.2)
Positive
47 (2.9)
11 (3.8)
Adjuvant therapy
None
2230 (57.0)
248 (46.6)
Chemotherapy
477 (12.2)
80 (15.0)
Radiation
682 (17.4)
85 (16.0)
Chemotherapy and radiation
493 (12.6)
117 (22.0)
Missing
29 (0.7)
2 (0.4)
2
1 P value was from two-sided χ test among women with non-missing data for the variable
2 Among patients for whom the procedure was performed
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
14 (6.5)
3 (1.5, 5)
0.24
210 (89.0)
26 (11.0)
0.23
130 (99.2)
1 (0.8)
<.0001
145 (56.9)
26 (10.2)
39 (15.3)
42 (16.5)
3 (1.2)
International Journal of Cancer
Page 22 of 25
Table 2. Multivariable hazard ratios (HRs) and 95% confidence intervals (CIs) for associations between race and risk of recurrence stratified by histologic subtype in the GOG 210 trial, 2003-2011
High-grade endometrioid
Serous
Mixed cell
Carcinosarcoma
Clear cell
Characteristic
Low-grade endometrioid
(recurrences=193, N=2,610)
(recurrences=119, N=556)
(recurrences=195, N=603)
(recurrences=107, N=493)
(recurrences=103, N=278)
(recurrences=36, N=158)
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
Race
164 (7.2)
102 (21.8)
143 (32.7)
68 (33.2)
29 (23.8)
NHW
1.00
1.00
1.00
1.00
1.00
1.00
85 (20.5)
1.00
1.00
1.00
1.00
1.00
1.00
22 (11.0)
15 (25.0)
41 (31.5)
26 (48.2)
5 (19.2)
NHB
2.06 (1.30, 3.27) 1.94 (1.21, 3.10)
1.30 (0.74, 2.31) 1.54 (0.85, 2.78)
0.99 (0.68, 1.46) 0.98 (0.65, 1.46)
19 (31.2)
1.50 (0.87, 2.56) 1.34 (0.76, 2.37)
1.70 (1.03, 2.82) 1.66 (0.99, 2.79)
0.66 (0.23, 1.88) 0.78 (0.25, 2.44)
7 (4.8)
2 (7.1)
11 (30.6)
9 (47.4)
2 (20.0)
Hispanic
1.08 (0.50, 2.33) 0.99 (0.45, 2.17)
0.44 (0.11, 1.82) 0.47 (0.11, 1.98)
1.40 (0.75, 2.62) 1.28 (0.67, 2.44)
3 (17.7)
0.94 (0.30, 3.00) 1.06 (0.33, 3.43)
1.71 (0.82, 3.58) 1.63 (0.73, 3.63)
1.40 (0.27, 7.36) 1.78 (0.29, 10.89)
Age
Stage
60 (15.5)
135 (6.1)
1.00
1.00
1.00
1.00
55 (17.2)
1.00
I
10 (23.8)
1.96 (1.19, 3.20) 1.96 (1.20, 3.22)
2.06 (1.03, 4.14) 2.04 (1.01, 4.14)
22 (35.5)
2.43 (1.45, 4.05)
II
21 (13.5)
49 (38.3)
III
37 (16.5)
1.69 (1.02, 2.80) 1.67 (1.01, 2.77)
3.01 (1.83, 4.96) 3.27 (1.95, 5.49)
118 (53.2)
5.93 (4.15, 8.47)
2
BMI 1 year ago, kg/m
2
1.00
1.00
23 (19.7)
1.00
1.00
1.00
42 (8.2)
40 (36.0)
Normal (< 25 kg/m )
2
31 (24.0)
1.19 (0.69, 2.05) 1.21 (0.69, 2.11)
0.86 (0.56, 1.32)
0.81 (0.52, 1.26) 0.79 (0.51, 1.24)
47 (33.6)
38 (7.8)
Overweight (25-30 kg/m )
2
55 (21.9)
1.19 (0.72, 1.96) 1.29 (0.76, 2.18)
0.79 (0.53, 1.16)
0.89 (0.61, 1.28) 0.86 (0.59, 1.25)
78 (28.5)
100 (7.8)
Obese (≥ 30 kg/m )
Smoking status
125 (7.4)
136 (36.0)
Never
1.00
1.00
77 (21.8)
1.00
1.00
1.00
13 (7.8)
8 (23.5)
Current
1.38 (0.77, 2.46) 1.35 (0.75, 2.43)
6 (17.1)
0.99 (0.43, 2.32) 1.03 (0.44, 2.44)
0.60 (0.29, 1.24)
48 (6.9)
47 (26.3)
Former
0.94 (0.67, 1.31) 0.95 (0.68, 1.32)
32 (20.9)
0.99 (0.65, 1.51) 0.96 (0.63, 1.48)
0.68 (0.48, 0.95)
Adjuvant treatment
None
106 (5.5)
1.00
1.00
34 (16.2)
1.00
1.00
50 (27.9)
1.00
22 (22.5)
69 (37.3)
Chemotherapy
3.11 (1.75, 5.52) 3.17 (1.78, 5.66)
31 (44.3)
1.78 (0.98, 3.23) 1.72 (0.93, 3.17)
0.55 (0.37, 0.82)
44 (10.6)
25 (41.7)
Radiation
1.48 (1.02, 2.16) 1.47 (1.01, 2.15)
36 (19.2)
1.06 (0.66, 1.71) 1.08 (0.66, 1.76)
1.14 (0.69, 1.86)
51 (29.1)
18 (13.0)
Chemotherapy and radiation
18 (20.7)
0.69 (0.36, 1.32) 0.62 (0.32, 1.20)
0.57 (0.38, 0.86)
1.62 (0.89, 2.96) 1.62 (0.88, 2.98)
Education
19 (7.6)
37 (36.3)
Less than high schoo
--1.00
12 (20.3)
--1.00
--62 (7.8)
64 (34.2)
High school/GED
--1.08 (0.63, 1.84)
40 (22.6)
--0.89 (0.44, 1.79)
--54 (7.2)
55 (33.1)
Some college/technical schoo
--1.14 (0.66, 2.00)
33 (20.1)
--0.92 (0.44, 1.90)
--57 (7.1)
37 (26.1)
College graduate/beyond
--1.27 (0.71, 2.26)
33 (21.6)
--0.82 (0.38, 1.75)
--Income
46 (8.8)
19 (15.7)
50 (31.5)
< $20,000
--1.00
--1.00
--42 (7.4)
28 (22.6)
50 (40.7)
$20,000-$39,999
--0.78 (0.50, 1.20)
--1.46 (0.79, 2.71)
--45 (7.1)
30 (24.6)
36 (27.7)
$40,000-$69,999
--0.79 (0.51, 1.23)
--2.11 (1.11, 4.02)
--35 (5.7)
25 (18.8)
30 (28.9)
≥$70,000
--0.61 (0.37, 1.02)
--1.68 (0.84, 3.37)
--1 row percentage
2 HRs from Cox proportional hazards model including race, age at diagnosis (continuous), stage, BMI, smoking status, and adjuvant therapy. Variables were coded as presented in
3 HRs additionally adjusted for income and education
NE=not estimable; NHW=Non-Hispanic White, NHB=Non-Hispanic Black. Missing values were coded as a separate category for each variable and included in the mo
1.00
2.32 (1.38, 3.90)
5.95 (4.15, 8.52)
37 (12.6)
12 (22.6)
58 (39.5)
1.00
1.00
2.33 (1.17, 4.65) 2.24 (1.11, 4.51)
4.34 (2.72, 6.92) 4.21 (2.60, 6.80)
41 (26.0)
7 (29.2)
55 (57.3)
1.00
1.00
0.94 (0.41, 2.16) 0.89 (0.38, 2.12)
2.90 (1.88, 4.47) 2.76 (1.77, 4.31)
15 (15.3)
6 (25.0)
15 (41.7)
1.00
1.00
1.74 (0.61, 5.0) 1.65 (0.58, 4.65)
4.13 (1.84, 9.24) 5.00 (2.02, 12.31)
1.00
0.80 (0.51, 1.24)
0.70 (0.47, 1.05)
23 (21.1)
23 (21.1)
47 (21.3)
1.00
1.00
0.82 (0.46, 1.49) 0.84 (0.46, 1.52)
0.93 (0.55, 1.57) 0.94 (0.55, 1.60)
19 (36.5)
29 (38.7)
43 (35.5)
1.00
1.00
1.09 (0.60, 2.00) 1.10 (0.59, 2.06)
0.84 (0.46, 1.53) 0.78 (0.42, 1.46)
6 (17.1)
12 (25.0)
15 (23.8)
1.00
1.44 (0.50, 4.09)
1.64 (0.61, 4.47)
1.00
1.74 (0.58, 5.18)
1.86 (0.65, 5.36)
1.00
0.55 (0.27, 1.15)
0.65 (0.46, 0.92)
72 (24.0)
9 (27.3)
23 (15.4)
1.00
1.00
1.27 (0.62, 2.60) 1.31 (0.63, 2.70)
0.55 (0.34, 0.89) 1.55 (0.33, 0.89)
66 (36.1)
6 (35.3)
26 (38.2)
1.00
1.00
1.39 (0.58, 3.34) 1.31 (0.53, 3.27)
1.11 (0.69, 1.76) 1.00 (0.62, 1.61)
20 (19.8)
4 (28.6)
11 (27.5)
1.00
0.95 (0.30, 3.04)
1.53 (0.72, 3.28)
1.00
1.10 (0.33, 3.69)
1.92 (0.80, 4.58)
1.00
0.56 (0.37, 0.84)
1.11 (0.67, 1.83)
0.59, 0.39, 0.89)
25 (15.2)
39 (35.1)
16 (21.1)
27 (19.3)
1.00
1.00
1.43 (0.81, 2.51) 1.48 (0.82, 2.64)
1.00 (0.51, 1.96) 0.96 (0.49, 1.91)
0.66 (0.36, 1.19) 0.64 (0.35, 1.16)
24 (29.3)
34 (41.5)
11 (30.6)
33 (43.4)
1.00
1.00
1.05 (0.60, 1.86) 1.14 (0.64, 2.06)
0.77 (0.37, 1.60) 0.74 (0.35, 1.57)
1.02 (0.56, 1.84) 1.08 (0.58, 1.98)
10 (17.9)
10 (27.0)
6 (20.7)
10 (27.8)
1.00
1.45 (0.54, 3.88)
0.98 (0.32, 2.98)
1.00 (0.37, 2.71)
1.00
1.35 (0.48, 3.78)
1.13 (0.35, 3.67)
1.10 (0.38, 3.19)
1.00
1.02 (0.66, 1.59)
1.05 (0.67, 1.66)
0.72 (0.42, 1.26)
18 (36.0)
33 (19.9)
32 (22.9)
24 (17.7)
---------
1.00
0.80 (0.41, 1.53)
0.85 (0.44, 1.65)
0.79 (0.37, 1.69)
17 (46.0)
37 (35.2)
24 (36.9)
25 (35.2)
---------
1.00
0.62 (0.33, 1.17)
0.67 (0.33, 1.36)
0.60 (0.28, 1.31)
6 (22.2)
13 (25.0)
8 (18.2)
9 (26.5)
---------
1.00
1.12 (0.38, 3.34)
0.64 (0.19, 2.17)
1.11 (0.31, 4.00)
1.00
1.36 (0.89, 2.08)
0.83 (0.51, 1.35)
0.98 (0.56, 1.71)
33 (30.3)
22 (19.6)
20 (18.5)
21 (20.4)
---------
1.00
0.75 (0.42, 1.36)
0.79 (0.42, 1.48)
1.10 (0.55, 2.20)
23 (36.5)
20 (31.3)
29 (48.3)
12 (26.1)
---------
1.00
0.84 (0.43, 1.63)
1.52 (0.81, 2.85)
0.86 (0.38, 1.97)
8 (22.2)
8 (18.6)
10 (32.3)
6 (20.7)
---------
1.00
0.91 (0.30, 2.77)
2.17 (0.69, 6.80)
1.33 (0.33, 5.41)
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Page 23 of 25
International Journal of Cancer
Table 3. Multivariable hazard ratios (HRs) and 95% confidence intervals (CIs) for associations between race and risk of recurrence stratified by stage in the GOG 210 trial, 2003-2011
Characteristic
recurrences (%)1
Stage I
Stage II
Stage III
(recurrences=343, N=3,484)
(recurrences=78, N=361)
(recurrences=332, N=853)
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
recurrences (%)1
HR (95% CI)2
HR (95% CI)3
Race
Non-Hispanic White
283 (9.6)
1.00
1.00
58 (21.1)
1.00
1.00
250 (36.4)
1.00
1.00
Non-Hispanic Black
48 (13.9)
1.57 (1.14, 2.16)
1.46 (1.05, 2.03)
17 (29.3)
1.37 (0.70, 2.68)
1.56 (0.77, 3.17)
63 (48.8)
1.23 (0.92, 1.65)
1.25 (0.92, 1.69)
Hispanic
12 (6.3)
0.81 (0.46, 1.46)
0.73 (0.41, 1.33)
3 (10.7)
0.77 (0.23, 2.59)
1.16 (0.32, 4.19)
19 (51.4)
1.79 (1.11, 2.88)
1.78 (1.10, 2.89)
Age
Histology
Low-grade endometrioid
135 (6.1)
1.00
1.00
21 (13.5)
1.00
1.00
37 (16.5)
1.00
1.00
High-grade endometrioid
60 (15.5)
2.49 (1.81, 3.42)
2.50 (1.81, 3.43)
10 (23.8)
2.49 (1.14, 5.44)
2.23 (1.01, 4.93)
49 (38.3)
3.08 (2.00, 4.73)
3.15 (2.04, 4.85)
Serous
55 (17.2)
2.55 (1.79, 3.63)
2.53 (1.78, 3.60)
22 (35.5)
3.52 (1.74, 7.12)
3.85 (1.85, 8.02)
118 (53.2)
4.28 (2.89, 6.33)
4.27 (2.88, 6.34)
Mixed cell
37 (12.6)
1.88 (1.28, 2.75)
1.89 (1.29, 2.77)
12 (22.6)
2.05 (0.95, 4.43)
2.08 (0.97, 4.48)
58 (39.5)
3.18 (2.09, 4.84)
3.21 (2.11, 4.91)
Carcinosarcoma
41 (26.0)
4.24 (2.88, 6.24)
4.30 (2.92, 6.32)
7 (29.2)
2.70 (1.06, 6.92)
2.86 (1.11, 7.39)
55 (57.3)
5.58 (3.62, 8.61)
5.67 (3.67, 8.76)
Clear cell
15 (15.3)
2.28 (1.31, 3.95)
2.28 (1.31, 3.97)
6 (25.0)
1.92 (0.73, 5.00)
1.94 (0.73, 5.15)
15 (41.7)
3.57 (1.94, 6.58)
3.66 (1.98, 6.77)
2
BMI 1 year ago, kg/m
Normal (< 25 kg/m )
73 (10.4)
1.00
1.00
19 (28.4)
1.00
1.00
61 (35.9)
1.00
1.00
2
Overweight (25-30 kg/m )
89 (12.6)
1.05 (0.77, 1.43)
1.01 (0.74, 1.39)
19 (26.0)
0.80 (0.41, 1.58)
0.84 (0.42, 1.66)
72 (34.5)
0.82 (0.58, 1.16)
0.82 (0.58, 1.16)
2
Obese (≥ 30 kg/m )
151 (9.0)
0.88 (0.66, 1.17)
0.81 (0.60, 1.08)
30 (17.9)
0.55 (0.30, 1.01)
0.58 (0.31, 1.07)
157 (41.8)
1.12 (0.83, 1.52)
1.13 (0.83, 1.55)
2
Smoking status
232 (10.4)
1.00
1.00
49 (21.0)
1.00
1.00
215 (40.3)
1.00
1.00
Current
13 (6.1)
0.66 (0.38, 1.16)
0.62 (0.35, 1.10)
3 (13.6)
0.79 (0.23, 2.68)
1.00 (0.29, 3.43)
30 (46.2)
1.48 (0.99, 2.19)
1.50 (1.00, 2.23)
Former
Adjuvant treatment
87 (9.1)
0.87 (0.68, 1.11)
0.89 (0.69, 1.13)
21 (21.2)
0.89 (0.52, 1.50)
0.81 (0.47, 1.39)
79 (34.5)
0.79 (0.61, 1.03)
0.78 (0.60, 1.02)
Never
None
188 (7.8)
1.00
1.00
14 (16.3)
1.00
1.00
47 (33.6)
1.00
1.00
Chemotherapy
45 (19.5)
1.35 (0.94, 1.93)
1.37 (0.96, 1.97)
6 (20.7)
0.96 (0.35, 2.61)
0.93 (0.34, 2.56)
154 (47.7)
1.05 (0.75, 1.47)
1.05 (0.75, 1.48)
Radiation
74 (12.9)
1.16 (0.88, 1.54)
1.15 (0.87, 1.52)
39 (25.5)
1.33 (0.70, 2.53)
1.25 (0.64, 2.41)
25 (32.1)
0.77 (0.47, 1.25)
0.76 (0.46, 1.25)
Chemotherapy and radiation
34 (12.9)
0.93 (0.63, 1.37)
0.94 (0.63, 1.39)
19 (20.7)
0.60 (0.29, 1.27)
0.56 (0.26, 1.19)
104 (35.1)
0.75 (0.53, 1.06)
0.74 (0.52, 1.05)
Less than high school
48 (13.2)
---
1.00
10 (20.8)
---
1.00
51 (44.7)
---
1.00
High school/GED
116 (10.7)
---
0.85 (0.59, 1.21)
28 (23.0)
---
1.24 (0.55, 2.81)
105 (38.0)
---
0.93 (0.65, 1.33)
Some college/technical school
92 (9.2)
---
0.81 (0.55, 1.18)
25 (22.9)
---
1.25 (0.53, 2.94)
89 (39.6)
---
1.05 (0.72, 1.52)
College graduate/beyond
85 (8.3)
---
0.82 (0.54, 1.22)
15 (18.3)
---
0.80 (0.31, 2.06)
85 (36.5)
---
0.87 (0.57, 1.31)
< $20,000
84 (11.7)
---
1.00
17 (17.7)
---
1.00
78 (39.0)
---
1.00
$20,000-$39,999
75 (10.1)
---
0.86 (0.62, 1.20)
22 (24.2)
---
1.53 (0.77, 3.04)
73 (37.1)
---
0.99 (0.70, 1.39)
$40,000-$69,999
84 (10.1)
---
1.02 (0.73, 1.43)
12 (14.6)
---
1.03 (0.45, 2.35)
74 (42.1)
---
1.14 (0.80, 1.62)
≥$70,000
1 row percentage
52 (6.5)
---
0.64 (0.42, 0.95)
19 (30.2)
---
2.86 (1.27, 6.44)
58 (34.9)
---
1.20 (0.79, 1.82)
Education
Income
2 HRs from Cox proportional hazards model including race, age at diagnosis (continuous), histology, BMI, smoking status, and adjuvant therapy. Variables were coded as presented in table.
3 HRs additionally adjusted for income and education.
NE=not estimable
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
International Journal of Cancer
Page 24 of 25
Figure 1. EC recurrence according to race and ethnicity in A) Low-grade endometrioid, B) Highgrade endometrioid, C) Serous, D) Mixed cell E) Carcinosarcoma, and F) Clear cell
A.
B.
1.00
log-rank p=0.15
0.00
0.00
log-rank p=0.02
0
1000
2000
3000
time to recurrence (days)
4000
0
Number at risk (events)
NHW
NHB
Hispanic
High-grade endometrioid
Kaplan-Meier estimates
0.25
0.50
0.75
Kaplan-Meier estimates
0.25
0.50
0.75
1.00
Low-grade endometrioid
2264
201
145
(109)
(17)
(4)
1000
2000
3000
time to recurrence (days)
4000
Number at risk (events)
1880
134
93
(51)
(4)
(3)
1266
49
38
(4)
(1)
(0)
124
8
4
(0)
(0)
(0)
0
0
0
NHW
NHB
Hispanic
468
60
28
(85)
(15)
(1)
291
32
19
(14)
(0)
(1)
153
17
3
(3)
(0)
(0)
17
5
1
(0)
(0)
(0)
0
0
0
D.
C.
1.00
log-rank p=0.12
0.00
0.00
log-rank p=0.98
0
1000
2000
3000
time to recurrence (days)
4000
0
Number at risk (events)
NHW
NHB
Hispanic
Mixed cell
Kaplan-Meier estimates
0.25
0.50
0.75
Kaplan-Meier estimates
0.25
0.50
0.75
1.00
Serous
437
130
36
(117)
(33)
(10)
1000
2000
3000
time to recurrence (days)
4000
Number at risk (events)
227
61
17
(23)
(8)
(1)
89
13
3
(3)
(0)
(0)
10
1
0
(0)
(0)
(0)
0
0
0
NHW
NHB
Hispanic
415
61
17
(73)
(15)
(3)
280
32
11
(12)
(4)
(0)
140
8
4
(0)
(0)
(0)
20
0
0
(0)
(0)
(0)
0
0
0
F.
E.
1.00
log-rank p=0.97
0.00
0.00
log-rank p=0.06
0
1000
2000
3000
time to recurrence (days)
4000
Number at risk (events)
NHW
NHB
Hispanic
Clear cell
Kaplan-Meier estimates
0.25
0.50
0.75
Kaplan-Meier estimates
0.25
0.50
0.75
1.00
Carcinosarcoma
205
54
19
(63)
(25)
(9)
0
1000
2000
time to recurrence (days)
Number at risk
96
16
6
(4)
(0)
(0)
40
6
1
(1)
(1)
(0)
1
0
0
(0)
(0)
(0)
NHW
0
0
0
NHW
NHB
Hispanic
NHB
122
26
10
3000
(events)
(26)
(4)
(2)
72
13
2
Hispanic
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
(3)
(1)
(0)
26
4
0
(0)
(0)
(0)
1
1
0
Page 25 of 25
International Journal of Cancer
Figure 2. EC recurrence according to race and ethnicity in A) Stage I, B) Stage II, and C) Stage III
A.
Kaplan-Meier estimates
0.25
0.50
0.75
1.00
Stage I
0.00
log-rank p=0.001
0
1000
Number at risk (events)
NHW
NHB
Hispanic
2949
345
190
(212)
(39)
(10)
2299
211
120
4000
3000
2000
time to recurrence (days)
(62)
(9)
(2)
B.
1421
75
43
(9)
(0)
(0)
141
15
5
(0)
(0)
(0)
0
0
0
Kaplan-Meier estimates
0.25
0.50
0.75
1.00
Stage II
0.00
log-rank p=0.11
0
1000
Number at risk (events)
NHW
NHB
Hispanic
275
58
28
(43)
(14)
(2)
189
32
17
2000
3000
time to recurrence (days)
(14)
(1)
(1)
C.
112
9
4
(1)
(2)
(0)
8
0
0
4000
(0)
(0)
(0)
0
0
0
Kaplan-Meier estimates
0.25
0.50
0.75
1.00
Stage III
0.00
log-rank p<0.001
0
1000
Number at risk (events)
NHW
NHB
Hispanic
687
129
37
(218)
(56)
(17)
NHW
358
45
11
2000
3000
time to recurrence (days)
(31)
(7)
(2)
181
13
2
NHB
(1)
(0)
(0)
24
0
0
4000
(0)
(0)
(0)
0
0
0
Hispanic
John Wiley & Sons, Inc.
This article is protected by copyright. All rights reserved.
Документ
Категория
Без категории
Просмотров
5
Размер файла
855 Кб
Теги
31127, ijc
1/--страниц
Пожаловаться на содержимое документа