University of Southern California Dissertations and Theses

An analysis of disease-free survival and overall survival in inflammatory breast cancer

(USC Thesis Other)

An analysis of disease-free survival and overall survival in inflammatory breast cancer

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content AN ANALYSIS OF DISEASE-FREE SURVIVAL AND OVERALL SURVIVAL IN
INFLAMMATORY BREAST CANCER

by

Tian Wang

____________________________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATISTICS/ EPIDEMIOLOGY)

May 2016

Copyright 2016 Tian Wang

ii
ACKNOWLEDGEMENT

I would like to thank my professor, Dr. Richard Sposto for his help and support in guiding me
through my thesis. He has always been patient and helpful with my questions, and also brings
humor to our meetings. I am immensely indebted to him for his help.

Additionally I would like to thank Dr. Julie E. Lang, Lily Tung, Elizabeth Pan, Dr. Melanie
Crutchfield and USC Norris Cancer Center, Keck Hospital of USC, and Los Angeles County
medical center for providing me the chance of this research and offering me help with the data.

I would also like to acknowledge my thesis committee members, Professor Wendy Mack and
Professor Todd Alonzo, for taking the time out of their schedule to review my thesis.

Last but not least, I would like to thank my family for being with me through thick and thin and
for always helping me achieve my full potential.

iii
TABLE OF CONTENTS

ACKNOWLEDGEMENT .............................................................................................................. II
LIST OF FIGURES ...................................................................................................................... IV
LIST OF TABLES ......................................................................................................................... V
ABSTRACT .................................................................................................................................. VI
1. INTRODUCTION ...................................................................................................................... 1
2. PATIENTS AND METHODS .................................................................................................... 4
3. RESULTS ................................................................................................................................... 7
4. DISCUSSION ........................................................................................................................... 11
5. CONCLUSION ......................................................................................................................... 16
6. REFERENCES ......................................................................................................................... 27

iv
LIST OF FIGURES

Figure 1. Disease-free survival for Stage III patients by IBC status.. .......................................... 17
Figure 2. Overall survival for Stage III patients by IBC status.. .................................................. 17
Figure 3. Locoregional control survival for Stage III patients by IBC status. .............................. 18
Figure 4. Overall survival for Stage IV patients by IBC status .................................................... 18
Figure 5. Progression-free survival for Stage IV patients by IBC status. ..................................... 19
Figure 6. Overall survival for all patients by IBC status .............................................................. 19
Figure 7. Disease-free survival for all patients by IBC status ...................................................... 20
Figure 8. Cumulative Incidence curves of different competing risks for non-IBC patients. ........ 21
Figure 9. Cumulative Incidence curves of different competing risks for IBC patients. ............... 21

v
LIST OF TABLES

Table 1. Patient Characteristics for non-IBC and IBC patients .................................................... 22
Table 2. Survival analyses statistics .............................................................................................. 23
Table 3. 2-year cumulative incidence rates of different competing risks for non-IBC and IBC .. 24
Table 4. Univariate and Multivariate analyses for disease-free survival ...................................... 25
Table 5. Univariate and Multivariate analyses for overall survival .............................................. 26

vi
ABSTRACT

OBJECTIVE: To examine if Inflammatory Breast Cancer (IBC) have worse disease-free and
overall survival than non-Inflammatory Breast Cancer (non-IBC) patients.

METHODS: Overall survival (OS) and disease-free survival (DFS) and locoregional control
survival (LRC) were compared between IBC and non-Inflammatory Breast Cancer (non-IBC)
Stage III patients. OS and progression-free survival (PFS) were compared between IBC and non-
IBC Stage IV patients. Univariate analysis was first done by OS and DFS analyses with Kaplan
Meier method for all IBC and non-IBC patients. Cox proportional hazards regression modeling
was then used to obtain hazard ratios and likelihood ratio p-values. Cumulative incidence
analysis was used to estimate and compare rates of different competing risks for IBC and non-
IBC patients. Multivariate analyses through Cox proportional hazard regression modeling and
likelihood ratio tests were performed to determine the important prognostic factors for overall
survival and disease-free survival.

RESULTS: For Stage III patients, OS, DFS and LRC were highly significantly different between
IBC and non-IBC patients (p=0.0032, p=0.0005 and p=0.0012, respectively). Race, tumor
histology, estrogen receptor status, and tumor size were significantly associated with both OS
and DFS in multivariate analysis. African American race, invasive ductal histology, ER-negative
receptor status, and larger tumor size were associated with poorer survival.

vii
CONCLUSION: IBC is an aggressive breast cancer with high recurrence rate and poor clinical
prognosis. Stage III IBC patients have a poorer survival outcome than Stage III non-IBC patients.
Further studies need to be done on the prognostic factors for IBC patients.

1
1. Introduction
Inflammatory Breast Cancer (IBC) is a rare and very aggressive form of breast cancer, which
develops rapidly and has early metastasis. It also has a high probability of recurrence and a poor
clinical prognosis (Lang, Wecsler et al. 2014). Most IBCs are invasive ductal carcinomas, which
are caused by cancer cells blocking the lymph vessels in the skin leading to a red and swollen
appearance of the breast. IBC cases account for approximately 1-5% of breast cancer diagnosed
in the United States. IBC is diagnosed at either stage III or stage IV depending on whether the
cancer cells affect only nearby lymph nodes or other tissues as well. IBCs tend to be diagnosed
among younger aged women compared to other type of breast cancers, and is more commonly
diagnosed among African American women than non-Hispanic White women (National Cancer
Institute, 2015). Symptoms of IBC are mainly redness (erythema) and swelling (edema)
involving more than one-third of the breast. The breast might become harder and there is pitting
and ridging of the breast skin, which make the breast look like an orange peel. Sometimes it also
involves inversion of the nipple. The swelling of the skin can make one breast look larger and
feel heavier than the other, and the breast can also be tender and warm. IBC patients usually feel
the breast to be painful and itchy (National Cancer Institute, 2015).

The main treatment for IBC is multimodal therapy, which includes systematic chemotherapy,
surgery and radiation therapy. Multimodal therapy has been shown to result in a better response
and longer survival. Treatment for Stage III IBC patients is a little different from Stage IV IBC
patients. For Stage III patients, the tumor has not spread outside the breast or nearby lymph
nodes, so the common treatment is to use chemotherapy to shrink the tumor first, followed by
surgery to remove the tumor. Radiation therapy usually follows thereafter. Sometimes more

2
chemotherapy is given after radiation. Chemotherapy before surgery is called neoadjuvant
therapy, which is mainly used to shrink the tumor, so the tumor can be easily removed from the
breast. Currently, anthracycline-based and taxane-based chemotherapy are usually used for
neoadjuvant therapy; it has been shown that these two chemotherapies, used subsequent to each
other, provide a better treatment outcome than each of them used individually (Cancer.org, 2015).
For HER2-positive patients, trastuzumab (Herceptin) might be given as well, and sometimes
with the drug pertuzumab (Perjeta). However, Herceptin cannot be used at the same time with an
anthracycline-based chemotherapy, because Herceptin might cause heart problems when using
together with an anthracycline. Stage IV IBC patients are usually given systemic therapy, which
may include chemotherapy, hormonal therapy, and/or targeted therapy with a drug that targets
HER2 (for HER2 positive cancers) (Cancer.org, 2015).

Previous studies on the management and treatment of IBC were restricted to single-center series,
and the results have been limited by inconsistent response criteria and small sample sizes. This
study is a retrospective cohort study on patients from the USC Norris Comprehensive Cancer
Center, Keck Hospital of USC and Los Angeles County medical center, who were given a
variety of treatments for IBC between 2006-2013. The IBC cohort is compared to stage III and
stage IV non-IBC patients to optimally test hypothesis that IBC has worse survival (Lang,
Wecsler et al. 2014).

The aims of the study are as follow:
Aim1: To determine if IBC has poorer survival than non-IBC. Locoregional control (LC),
disease-free survival (DFS), and overall survival (OS) are compared between Stage III non-IBC

3
and IBC patients. For Stage IV IBC patients, we will determine OS and progression-free
survival (PFS) rates. OS and DFS for all non-IBC and IBC patients were also compared (Lang,
Wecsler et al. 2014).

Aim2: To compare the rates of different types of events between non-IBC and IBC patients.
Competing risks include: death, local recurrence, regional recurrence and distant metastasis for
IBC and non-IBC patients.

Aim3: To identify prognostic factors that correspond to better or worse outcomes in IBC patients
(Lang, Wecsler et al. 2014). Factors to evaluate include demographic information, biomarker
status, and extent of disease.

4
2. Patients and Methods
2.1 Enrollment
The study population was patients diagnosed with IBC or non-IBC and treated at LAC-USC
medical center, Norris Center Hospital, and/or Keck Hospital starting from January 1
st
, 2006 to
December 31
st
, 2013. Patients who were first diagnosed or received initial treatment from the
above hospitals were included. Patients who were diagnosed or first treated in other hospitals and
then came to the above hospitals were excluded.
All patients included were identified through the electronic medical record system at USC Keck
Medical Center and LAC-USC (Affinity, Cerner, and Montage). Data were entered into a secure
REDCap database maintained by the Southern California Clinical and Translational Science
Institute (SC-CTSI). Only approved study personnel could have access to patient identifiers and
enter data from clinical sources (Lang, Wecsler et al. 2014). After the data were fully collected
and entered, the data were exported into a dataset that was devoid of any personal identifiers.
The study was approved by the University of Southern California Institutional Review Board.

2.2 Statistical Methods
2.2.1 Patient characteristics
Characteristics of patients were compared according to IBC status using the chi-square test.

2.2.2 Survival Analysis
Survival analysis for Stage III and Stage IV patients was performed. Overall survival, disease-
free survival, and locoregional control survival were analyzed for Stage III patients and overall
survival and progression-free survival were calculated for Stage IV patients. All survival

5
endpoints were compared between IBC and non-IBC patients. Univariate analysis of all patients
(Stage III and Stage IV combined) by IBC status for overall survival and disease-free survival
were performed. For all survival analyses, the starting point for follow-up was the date of
diagnosis. The endpoint of the survival analyses were 1) overall survival: time from date of
diagnosis to date of death; 2) Stage III disease-free survival: minimum time from date of
diagnosis to date of local recurrence, date of regional recurrence, date of distant metastasis, or
date of death; 3) Stage III locoregional recurrence free survival: minimum time from date of
local recurrence or date of regional recurrence; 4) progression-free survival: minimum time from
date of diagnosis to date of local recurrence, date of regional recurrence, date of distant
metastasis, date of progression, or date of death. Patients who did not experience any event were
censored at the end of follow-up. All survival analyses were visualized using Kaplan-Meier
survival curves (Kaplan and Meier, 1958). The Kaplan-Meier method provides a nonparametric
estimate of the survival for events of interest. For all patients and Stage III patients, 5-year
survival rates with 95% Confidence Intervals were obtained. For Stage IV patients, 4-year
overall survival rates and 2-year progression-free survival rates were obtained. 5-year survival
rates for Stage IV patients were not available because all patients either had events or have been
censored by the time of year 5. Survival curves were compared for statistical differences between
IBC and non-IBC groups using log-rank tests. Unadjusted hazard ratios were estimated and
statistically compared with Cox proportional hazards regression.

2.2.3 Cumulative incidence
To calculate the cumulative incidence, death, disease progression, local recurrence, regional
recurrence and distant metastatic recurrence were considered events of interest. The probability

6
of each event of interest was calculated using nonparametric estimation of cumulative incidence,
taking into account informative censoring due to competing risks. Cumulative incidence
estimation was calculated first by getting the Kaplan-Meier estimate for all competing risks
including progression, local recurrence, regional recurrence and distant metastatic recurrence.
The conditional probability of each specific event was then calculated by the joint probability of
this event and the probability of remaining free of other events (Cox and Oakes 1984).
Cumulative incidence graphs were plotted with a competing risks approach by year since
diagnosis for IBC and non-IBC patients.

2.2.4 Identification of prognostic factors
Cox proportional hazards regression modeling was used to identify risk factors for disease-free
and overall survival among all breast cancer patients and to generate a prognostic model. This
method was used because it can model the effect of several variables on survival rates under the
proportional hazards assumption (Walters, 2009). Both univariate and multivariate analysis were
performed for the risk factors listed in Table 1. For multivariate analysis, likelihood ratio tests
were used for model selection. Model selection was performed as follow: 1) univariate analysis
identified potential prognostic factors, all variables had p<0.25 were included in a multivariate
model, 2) variables were removed from the model one by one in reverse order of significance,
until all remaining variables were significant at p<0.05 or marginally larger than 0.05. Statistical
interactions for IBC status with each individual risk factor were tested for both overall survival
and disease-free survival.
All statistical evaluations were performed with the STATA/SE 12.0 software package. P<0.05
(two-tailed) was considered statistically significant.

7
3. Results
3.1 Descriptive data
A total of 235 Stage III (n=167) and Stage IV (n=68) breast cancer patients were included in the
study. 167 (141 Stage III, 26 Stage IV) patients had non-inflammatory breast cancer (non-IBC)
and 68 (56 Stage III, 12 Stage IV) had inflammatory breast cancer (IBC). 50% of the patients
were over 50 years old and 50% were under 50 years old. Approximately 84% of the patients
were Stage III patients and the other 16% patients were Stage IV patients. 53% of the patients
were Hispanic, 26% of the patients were non-Hispanic White, 8% were Black and 13% of the
patients were Asian/Pacific Islanders. 37% of patients had received neoadjuvant chemotherapy.
Of all the patients, 66% were estrogen receptor positive (ER-positive), 51% were progesterone
receptor positive (PR-positive), and 24% were HER2 receptor positive. Using chi-square tests to
compare variables between IBC and non-IBC patients, estrogen receptor and progesterone
receptor status were found to significantly differ between IBC and non-IBC patients. Results are
shown in Table 1.

3.2 Univariate survival comparisons of IBC to non-IBC patients
Survival rates, hazard ratios of IBC comparing to non-IBC and log-rank test p-values are listed in
Table 2. Figures 1 to 7 show the Kaplan-Meier curves for each survival outcome. For Stage III
patients, the outcomes of OS (p=0.0032), DFS (p=0.0005) and LRC (p=0.0012) all significantly
differed between IBC and non-IBC patients, with IBC patients having poorer survival outcomes.
The 5-year overall survival rate for non-IBC patients was 0.81 (95% CI: 0.68, 0.89), and for IBC
patients was 0.64 (95% CI: 0.43, 0.78) (Figure 2) (hazard ratio= 2.86, 95% CI: 1.38, 5.94). The
5-year disease-free survival rate for non-IBC patients was 0.59 (95% CI: 0.47, 0.70), and for IBC

8
patients was 0.36 (95% CI: 0.20, 0.52) (Figure 1) (hazard ratio= 2.27, 95% CI: 1.41, 3.66). The
5-year locoregional recurrence free survival rate for non-IBC patients was 0.87 (95% CI: 0.76,
0.93), and for IBC patients was 0.66 (95% CI: 0.45, 0.80) (Figure 3) (hazard ratio= 3.59, 95%
CI: 1.57, 8.18). For the small number of 38 Stage IV patients, the hazard ratio for overall
survival was 1.01 (95% CI: 0.19, 5.20, Figure 4) and for progression-free survival was 0.98
(95% CI: 0.40, 2.37, Figure 5). Neither overall survival rates (p=0.99) nor progression-free
survival rates (p=0.96) were statistically different between non-IBC and IBC patients. For all
patients (Stage III and Stage IV combined), the hazard ratios for overall survival and disease-free
survival were 2.43 (95% CI: 1.26, 4.70, Figure 6) and 1.95 (95% CI: 1.29, 2.95, Figure 7). Both
overall survival (p=0.0064) and disease-free survival (p=0.0012) significantly differed in non-
IBC and IBC patients.

3.3 Cumulative Incidence Analysis
Results of 2-year cumulative incidence analysis are shown in Table 3; cumulative incidence
curves for non-IBC and IBC patients are shown in Figure 8 and Figure 9. Table 3 includes the 2-
year cumulative incidence rates and total number of events for all competing risks (death,
progression, local recurrence, regional recurrence and distant metastatic recurrence) for IBC and
non-IBC patients. For both IBC and non-IBC patients, distant metastatic recurrence had the
highest cumulative incidence rates among all the competing events. Cumulative incidence rates
for distant metastasis, progression, regional recurrence and death for IBC patients were all higher
than non-IBC patient rates (Figure 8, Figure 9). The 2-year total cumulative incidence rate of all
failures for IBC patients is 0.50, higher than the 2-year total cumulative incidence rate of all
failures for non-IBC patients (0.26). The 2-year distant metastatic cumulative incidence rate for

9
IBC patients is 0.34 and for non-IBC is 0.14. The distribution of failure events between IBC and
non-IBC patients did not differ (p=0.42).

3.4 Multivariate analyses for prognostic factors
Results of Cox proportional hazards modeling to identify prognostic factors analyses for disease-
free and overall survival are shown in Table 4 and Table 5. The tables display hazard ratios, 95%
confidence intervals, and likelihood ratio p-values for all variables in the same model and for
variables included in the final model after model selection. The results (Table 4) indicate that
IBC patients had a 1.95 (95% CI: 1.29, 2.95) times higher chance of recurrence (local, regional,
or metastatic), progression or death; after adjusting for other variables, the hazard ratio was 2.02
(95% CI: 1.26, 3.22). IBC status was significantly associated with disease-free survival rate
before (p=0.0021) and after (p=0.0041) adjusting for other variables. Compared to non-Hispanic
White women, Black women had a 1.80 (95% CI: 0.88, 3.69; p=0.0006) times higher risk of
experiencing another event (including death, local recurrence, regional recurrence distant
metastasis and progression); after adjusting for other variables, the hazard ratio was 1.76 (95%
CI: 0.81, 3.80; p=0.033). For neoadjuvant chemotherapy, the chance of experiencing another
event was 1.60 (95% CI: 0.91, 2.79; p<0.0001) times higher than women who did not receive
neoadjuvant chemotherapy; after adjusting for other variables, the hazard ratio was 1.34 (95% CI:
0.74, 2.42; p=0.0012). After the model selection process, race, neoadjuvant chemotherapy,
estrogen receptor status, IBC, and size of tumor were statistically significant prognostic factors
for disease-free survival rates (p=0.0220, p<0.0001, p=0.0012, p=0.0039 and p=0.0447
respectively), and tumor histology was marginally significant with disease-free survival rates
(p=0.0863).

10

The results (Table 5) indicate that IBC patients had a 2.43 (95%CI: 1.26, 4.70) times higher
chance of death compared to non-IBC patients; after adjusting for other risk factors, the hazard
ratio was 3.45 (95% CI: 1.56, 7.65). IBC status was significantly associated with overall survival
rate before (p=0.0105) and after (p=0.0024) adjusting for other variables. For tumor histology,
lobular breast cancer has better overall survival rates compared to invasive ductal breast cancer.
Hazard ratio is 0.35 (95% CI: 0.08, 1.48; p=0.3108) before adjusting for other factors and 0.23
(95% CI: 0.05, 1.07; p=0.0832) after adjusting for other factors. After the model selection
process, radiation status, estrogen receptor status, HER2 receptor and IBC status were
statistically significant prognostic factors for overall survival rates (p=0.0022, p=0.0010, 0.0103
and p=0.0020 respectively), and race and tumor histology were marginally significant with
overall survival rates (p=0.0563 and p=0.0574 respectively).

After testing statistical interaction terms, no risk factors were found to significantly interact with
IBC status for both overall survival and disease-free survival.

11
4. Discussion
Inflammatory breast cancer is considered one of the worst types of breast cancers, which has
“rapid tumor progression, early metastasis, and poor survival” (Smoot, Koch et al. 2006). The
current treatment for IBC includes multimodal therapy, which is a combination of radiation
therapy, chemotherapy, surgery, and hormone therapy. Through multiple studies, multimodal
therapy has been found to be better than single-modality or dual-modality therapy in terms of
survival. After prognostic factor modeling, race, menopausal status and tumor diameter were
considered important for overall and disease-free survival (Smoot, Koch et al. 2006). Another
possible prognostic factor for IBC survival is the sequence in which the treatment is conducted
(i.e. the order of treatment); however there is still debate for this prognostic factor (Smoot, Koch
et al. 2006).

In our study, in order to examine if IBC patients have worse survival than non-IBC patients,
several survival analyses were performed. Hazard ratios and survival rates were obtained to
compare the survival between IBC and non-IBC patients. For the survival analysis, Stage III and
Stage IV patients were separated. In addition, cumulative incidence rates for competing risks
including death, progression, local recurrence, regional recurrence and distant metastasis were all
analyzed to estimate the recurrence rates. This analysis was conducted to determine which type
of recurrence most patients get, and to compare if IBC patients have different types of recurrence
compared to non-IBC patients. Cox regression models were used to determine what prognostic
factors were related to both disease-free survival and overall survival. In the univariate and
multivariate analysis, IBC status was included as a risk factor to determine if IBC patients have
worse overall survival and disease-free survival. In addition to the factors that previous studies

12
identified, our study considered other factors such as clinical stage, tumor histology, neoadjuvant
chemotherapy, and radiation treatment. To determine if any risk factor differentially impacted
survival in IBC versus non-IBC patients, statistical interaction terms were tested between IBC
status and every risk factor variables.

In a previous study among 128 patients with non-metastatic disease, the 5-year overall survival
rate for IBC was 42% and disease-free survival was 21% (Smoot, Koch et al. 2006). Our study
found that the 5-year survival rate for IBC is 63% (95% CI: 44%, 77%), and the 5-year overall
survival rate is 30% (95% CI: 17%, 46%). The overall survival rate is 11% and disease-free
survival is 9% higher than the Smoot et al paper. This could be due to the quality of treatment. In
the Smoot et al paper, patients were treated between 1985 and 2003, an earlier period than our
study. Treatment for IBC could have improved over this period. This could also be due to the
difference of sample sizes of the two studies. Smoot et al paper included 128 IBC patients while
our study has 68 IBC cases.

Based on our results, compared to non-IBC patients, IBC patients have a relatively worse
survival. In the survival analysis, the analysis for Stage III and Stage IV patients were separated.
For Stage III patients, IBC had a much worse result in overall survival, disease-free survival and
locoregional survival compared to non-IBC patients. However, for Stage IV patients, the hazard
ratios of overall survival and progression-free survival were not statistically different for IBC
and non-IBC (p=0.99 and p=0.96 respectively). These results could due to a combination of two
reasons. One reason is that there were only 38 Stage IV patients. The sample size was too small
to detect a difference between IBC and non-IBC. The other is that Stage IV breast cancer is

13
invasive and affects not only breast and nearby lymph nodes, but also organs like lungs, brain,
and bones. In general, Stage IV patients already have the worst outcome of all breast cancer
stages, so both IBC and non-IBC patients would have very low survival rates. Thus, no
conclusion could be drawn from the results of Stage IV patients.

In the multivariate analysis, we tested IBC status as a risk factor for survival among all 235
breast cancer patients. After adjusting for other risk factors, the hazard ratio of IBC patients was
3.45 (HR 95% CI: 1.56, 7.65) times higher than non-IBC patients for overall survival. For DFS
analysis, the hazard ratio was 2.02 (HR 95% CI: 1.26, 3.22). To summarize, both of the survival
analysis and multivariate analysis show that IBC patients tend to have worse survival, which is
consistent with the results found by previous studies. No variables were found to be statistically
interacting with IBC status for overall and disease-free survival, meaning that no risk factors
assessed in this study had a differential impact on survival outcomes in IBC and non-IBC
patients. Through the multivariate analysis, we found that in terms of disease-free survival, the
following factors were significant: race, tumor histology, neoadjuvant chemotherapy, estrogen
receptor status, IBC, and tumor size. For overall survival, tumor histology, radiation therapy,
estrogen receptor, IBC status and tumor size were significant factors. ER status and race were
also found to be important in previous studies. Panades et al’s study of 485 patients found that
ER status was independently associated with breast cancer-specific survival (p=0.02); the hazard
ratio for ER-negative was 1.94 (95% CI: 1.08, 2.48) (Panades, Olivotto et al. 2005). In order to
compare this with our results, the hazard ratio for ER-negative was converted to ER-positive
hazard ratio, which is 0.52 (95% CI: 0.40, 0.93). In our study, ER status was statistically
associated with overall survival and disease-free survival both independently and within the final

14
multivariate model. The hazard ratio in the univariate disease-free analysis for ER-positive was
0.59 (95% CI: 0.39, 0.88) and in the multivariate analysis was 0.44 (95% CI: 0.27, 0.72). The
hazard ratio in the univariate overall analysis for ER-positive was 0.41 (95% CI: 0.21, 0.79) and
in multivariate analysis was 0.26 (95% CI: 0.12, 0.59). Similar with Panades et al’s results, our
findings indicate that ER-negative patients have a higher risk of recurrence and of mortality. For
race, Schinkel et al pointed out that despite IBC status, black women had significantly worse
survival than non-Hispanic White women. In univariate analysis for IBC patients, the hazard
ratio for Black women was 1.32 (95% CI: 1.21, 1.44) compared to Non-Hispanic White women
(Schinkel, Zahm et al. 2014). This is consistent with our results, which showed African
American women had hazard ratios of 3.58 (95% CI: 1.14, 11.23) and 1.72 (95% CI: 0.82, 3.65)
compared with non-Hispanic White women for overall and disease-free survival respectively.
For tumor histology, we found that most patients had invasive ductal, and these patients had the
worst survival among all types.

Another important aspect of the study is the treatment of inflammatory breast cancer. Combined
multimodal therapy was used to increase the survival of IBC patients and quality of life.
However, in terms of sequence of treatment, there still remains ambiguity. Panades et al reported
that neither early radiation therapy nor late radiation therapy has any significant influence on the
locoregional relapse-free survival or breast cancer-specific survival. They also found that there is
no statistically significant impact on survival when mastectomy was done before or after
chemotherapy (Panades, Olivotto et al. 2005). However, in the Liauw et al paper, up-front
mastectomy was found to be more beneficial to cause-specific survival in the multivariate
analysis. This was compared between 18 patients who had up-front mastectomy and 43 patients

15
who had neoadjuvant chemotherapy (Liauw, Benda et al. 2004). Similar with the results reported
in Liauw et al’s paper, we found that in the multivariate analysis, neoadjuvant chemotherapy is
associated with a higher risk of recurrence and higher risk of mortality; the hazard ratios are 1.60
(95% CI: 0.91, 2.79) and 1.11 (95% CI: 0.47, 2.62), respectively. Since neoadjuvant
chemotherapy had 60 observations missing, univariate analysis was only done for 175 patients
who had neoadjuvant chemotherapy recorded. The results show the same effect of neoadjuvant
chemotherapy on disease-free and overall survival. In other words, based on our results, up-front
surgery would be more beneficial. It is conceivable that treating tumor with neoadjuvant
chemotherapy first allows the tumor to grow or have further metastasis during the chemotherapy
treatment, so that using mastectomy to remove the tumor first would be more effective (Liauw,
Benda et al. 2004). Worse outcomes of neoadjuvant chemotherapy could be because patients
who received chemotherapy first had worse tumor characteristics than patients who had up-front
surgery, since chemotherapy is mainly used to shrink the tumor so it is easy to remove the tumor
afterwards. Given that different literature has not shown a uniform conclusion, further research
needs to be done in this aspect. Figuring out the most effective sequence of the multimodal
treatment is key for IBC patients, as this will help prolong not only their life, but also improve
their quality of life.

16
5. Conclusion
IBC patients have worse survival compared to non-IBC patients. Cumulative incidence analysis
shows no obvious difference in the types of recurrence between IBC and non-IBC patients. From
multivariate analysis, factors including race, tumor histology, estrogen receptor and tumor size
were significantly associated with survival. Black women, invasive ductal patients, ER-negative
and larger tumor size was associated with poorer survival. However, multimodal treatment for
IBC is lacking in the field, and needs to be further investigated.

17

Figure 1. Disease-free survival for Stage III patients by IBC status. Hazard ratio for IBC patients
compared to non-IBC patients is 2.27 (95% CI: 1.41, 3.66) (log-rank test p-value=0.0005).

Figure 2. Overall survival for Stage III patients by IBC status. Hazard ratio for IBC patients
compared to non-IBC patients is 2.86 (95% CI: 1.38, 5.94) (log-rank test p-value=0.0032).

18

Figure 3. Locoregional control survival for Stage III patients by IBC status. Hazard ratio for IBC
patients compared to non-IBC patients is 3.59 (95% CI: 1.57, 8.18) (log-rank test p-
value=0.0012).

Figure 4. Overall survival for Stage IV patients by IBC status. Hazard ratio for IBC patients
compared to non-IBC patients is 1.01 (95% CI: 0.19, 5.20) (log-rank test p-value=0.9939).

19

Figure 5. Progression-free survival for Stage IV patients by IBC status. Hazard ratio for IBC
patients compared to non-IBC patients is 0.98 (95% CI: 0.40, 2.37 ) (log-rank test p-
value=0.9621).

Figure 6. Overall survival for all patients by IBC status. Hazard ratio for IBC patients compared
to non-IBC patients is 2.43 (95% CI: 1.26, 4.70) (log-rank test p-value=0.0064).

20

Figure 7. Disease-free survival for all patients by IBC status. Hazard ratio for IBC patients
compared to non-IBC patients is 1.95 (95% CI: 1.29, 2.95) (log-rank test p-value=0.0012).

21

Figure 8. Cumulative Incidence curves of different competing risks (including death,
progression, local recurrence, regional recurrence and distant metastasis) for non-IBC patients.

Figure 9. Cumulative Incidence curves of different competing risks (including death,
progression, local recurrence, regional recurrence and distant metastasis) for IBC patients.

22
Table 1. Patient Characteristics for non-IBC and IBC patients

Variables Categories Non-IBC
N=167
IBC
N=68
Total
N=235
P-value
Age <50 years old 86 (52) 31 (46) 117 (50) 0.41
>50 years old 81 (48) 37 (54) 118 (50)
Race Non-Hispanic 44 (26) 17 (25) 61 (26) 0.51
Hispanic 92 (55) 33 (49) 125 (53)
Black 11 (7) 8 (12) 19 (8)
Asian/Pacific
Islander
20 (12) 10 (15) 30 (13)
Clinical Stage

Stage III 141 (84) 56 (82) 197 (84) 0.70
Stage IV 26 (16) 12 (18) 38 (16)
Tumor Histology Invasive ductal 126 (75) 49 (72) 175 (74) 0.72
Lobular 19 (11) 9 (13) 28 (12)
Other invasive 5 (3) 4 (6) 9 (4)
N/A 17 (10) 6 (9) 23 (10)
Neoadjuvant
chemotherapy
No 69 (41) 20 (29) 89 (38) 0.21
Yes 59 (35) 27 (40) 86 (37)
N/A 39 (23) 21 (31) 60 (26)
Mastectomy No 66 (40) 28 (41) 94 (40) 0.81
Yes 101 (60) 40 (59) 141 (60)
Radiation No 47 (28) 18 (26) 65 (28) 0.80
Yes 120 (72) 50 (74) 170 (72)
Estrogen receptor Negative 47 (28) 32 (47) 79 (34) 0.005
Positive 120 (72) 36 (53) 156 (66)
Progesterone receptor Negative 74 (44) 40 (59) 114 (49) 0.044
Positive 93 (56) 28 (41) 121 (51)
HER2 receptor Negative 129 (77) 49 (72) 178 (76) 0.40
Positive 38 (23) 19 (28) 57 (24)
Size Smaller than 4.1cm 71 (43) 30 (44) 101 (43) 0.27
Larger than 4.1cm 71 (43) 33 (49) 104 (44)
N/A 25 (14) 5 (7) 30 (13)

23
Table 2. Survival analyses statistics

Patients Survival analysis
type
Non-IBC 5-year
survival rate
(95% CI)
IBC 5-year survival
rate
(95% CI)
HR (95% CI)** P-value*

Stage III
Overall Survival 0.81 (0.68, 0.89) 0.64 (0.43, 0.78) 2.86 (1.38, 5.94) 0.0032
Disease-free 0.59 (0.47, 0.70) 0.36 (0.20, 0.52) 2.27 (1.41, 3.66) 0.0005
Locoregional 0.87 (0.76, 0.93) 0.66 (0.45, 0.80) 3.59 (1.57, 8.18) 0.0012

Stage IV
Non-IBC 4-year
survival rate (95%
CI)
IBC 4-year survival
rate (95% CI)

Overall survival 0.62 (0.28, 0.84) 0.58 (0.08, 0.89) 1.01 (0.19, 5.20) 0.99
Non-IBC 2-year
survival rate (95%
CI)
IBC 2-year survival
rate (95% CI)

Progression-free 0.43 (0.23, 0.62) 0.34 (0.08, 0.63) 0.98 (0.40, 2.37) 0.96

All
Non-IBC 5-year
survival rate (95%
CI)
IBC 5-year survival
rate (95% CI)

Overall Survival 0.79 (0.67, 0.87) 0.63 (0.44, 0.77) 2.43 (1.26, 4.70) 0.0064
Disease-free 0.52 (0.41, 0.62) 0.30 (0.17, 0.46) 1.95 (1.29, 2.95) 0.0012
*P-values are obtained from log-rank test
**Hazard Ratios (HR) are obtained from Cox regression analysis

24
Table 3. 2-year cumulative incidence rates of different competing risks (including death,
progression, local recurrence, regional recurrence and distant metastasis) for non-IBC and IBC
(P-value obtained from Chi-square test of event distribution between non-IBC and IBC patients)

Type of failure Non-IBC IBC
2-year
Cumulative
Incidence Rate
(95% CI)
2-year
number of total
events
2-year
Cumulative
Incidence Rate
(95% CI)
2-year
number of total
events
Distant metastasis 0.14 (0.09, 0.20) 19 0.34 (0.22, 0.46) 20
Progression 0.09 (0.05, 0.14) 13 0.10 (0.04, 0.19) 6
Local recurrence 0.01 (0, 0.03) 1 0 0
Regional recurrence 0.01 (0, 0.03) 1 0.03 (0.01, 0.10) 2
Death 0.01 (0, 0.03) 1 0.03 (0.01, 0.09) 2
All failures (total) 0.26 (0.18, 0.31) 35 0.50 (0.38, 0.63) 30
Chi-square p-value 0.42

25
Table 4. Univariate and Multivariate analyses for disease-free survival

N
Disease-free survival
univariate
Disease-free
survival
multivariate
Disease-free
survival
multivariate

Hazard Ratio
(95% CI)

p-value
(LR)
Hazard Ratio
(95% CI)
p-value
(LR)
Hazard Ratio
(95% CI)
p-value
(LR)
Age <50 years old
>=50 years old
117
118
1
0.89 (0.60, 1.33)
0.58 1
0.63 (0.38, 1.02)
0.0603
Race Non-Hispanic
Hispanic
Black
Asian/Pacific
Islander
61
125
19
30
1
0.60 (0.38, 0.95)
1.80 (0.88, 3.69)
0.29 (0.12, 0.69)
0.0006 1
0.68 (0.40, 1.16)
1.76 (0.81, 3.80)
0.40 (0.16, 1.02)
0.0334 1
0.76 (0.46, 1.26)
1.72 (0.82, 3.65)
0.35 (0.14, 0.88)
0.0220
Clinical Stage Stage III
Stage IV
197
38
1
3.77 (2.36, 6.01)
<0.0001 1
1.24 (0.65, 2.35)
0.52
Tumor
histology
Invasive ductal
Lobular
Other Invasive
N/A
175
28
9
23
1
0.71 (0.37, 1.38)
0.48 (0.12, 1.94)
0.98 (0.53, 1.82)
0.51 1
0.57 (0.28, 1.15)
0.24 (0.06, 1.03)
1.14 (0.60, 2.18)
0.0477 1
0.66 (0.33, 1.31)
0.26 (0.06, 1.08)
1.09 (0.58, 2.07)
0.0863
Neoadjuvant
chemotherapy
No
Yes
N/A
89
86
60
1
1.60 (0.91, 2.79)
6.34 (3.72, 10.80)
<0.0001 1
1.34 (0.74, 2.42)
3.77 (1.85, 7.69)
0.0012 1
1.29 (0.72, 2.31)
4.45 (2.46, 8.04)
<0.0001
Mastectomy No
Yes
94
141
1
0.64 (0.43, 0.95)
0.0297 1
1.12 (0.67, 1.89)
0.66
Received
radiation
No
Yes
65
170
1
0.34 (0.22, 0.52)
<0.0001 1
0.62 (0.34, 1.13)
0.12
Estrogen
receptor
Negative
Positive
79
156
1
0.59 (0.39, 0.88)
0.0118 1
0.48 (0.25, 0.93)
0.0269 1
0.44 (0.27, 0.72)
0.0012
Progesterone
receptor
Negative
Positive
114
121
1
0.63 (0.42, 0.95)
0.0253 1
0.75 (0.40, 1.39)
0.36
HER2
receptor
Negative
Positive
178
57
1
1.14 (0.73, 1.79)
0.56 1
0.69 (0.41, 1.16)
0.15
IBC status No
Yes
167
68
1
1.95 (1.29, 2.95)
0.0021 1
2.02 (1.26, 3.22)
0.0041 1
1.99 (1.26, 3.16)
0.0039
Size (mean) <4.1cm
>=4.1cm
N/A
101
104
30
1
1.26 (0.79, 1.99)
3.15 (1.84, 5.40)
0.0003 1
0.83 (0.49, 1.42)
2.28 (1.15, 4.51)
0.0156 1
0.89 (0.54, 1.48)
1.94 (1.05, 3.59)
0.0447

26
Table 5. Univariate and Multivariate analyses for overall survival

N
Overall survival
univariate
Overall survival
multivariate
Overall survival
multivariate

Hazard Ratio
(95% CI)

p-value
(LR)
Hazard Ratio
(95% CI)
p-value
(LR)
Hazard Ratio
(95% CI)
p-value
(LR)
Age <50 years old
>=50 years old
117
118
1
0.68 (0.35, 1.33)
0.25 1
0.69 (0.32, 1.52)
0.36
Race Non-Hispanic
Hispanic
Black
Asian/Pacific
Islander
61
125
19
30
1
0.88 (0.40, 1.95)
3.12 (1.03, 9.41)
0.60 (0.16, 2.20)
0.13 1
0.78 (0.30, 2.04)
3.13 (0.87, 11.27)
0.64 (0.15, 2.81)
0.19 1
0.77 (0.33, 1.77)
3.58 (1.14, 11.23)
0.44 (0.11, 1.73)
0.0563
Clinical Stage Stage III
Stage IV
197
38
1
1.54 (0.68, 3.53)
0.32 1
0.60 (0.21, 1.76)
0.35
Tumor
histology
Invasive ductal
Lobular
Other Invasive
N/A
175
28
9
23
1
0.35 (0.08, 1.48)
0.61 (0.08, 4.49)
0.56 (0.17, 1.84)
0.31 1
0.23 (0.05, 1.07)
0.24 (0.03, 1.91)
0.89 (0.26, 3.11)
0.0832 1
0.22 (0.05, 0.98)
0.23 (0.03, 1.82)
0.82 (0.24, 2.80)
0.0574
Neoadjuvant
chemotherapy
No
Yes
N/A
89
86
60
1
1.11 (0.47, 2.62)
2.59 (1.16, 5.76)
0.0375 1
0.93 (0.37, 2.32)
1.59 (0.51, 4.99)
0.65
Mastectomy No
Yes
94
141
1
0.76 (0.39, 1.47)
0.41 1
0.98 (0.41, 2.33)
0.97
Received
radiation
No
Yes
65
170
1
0.39 (0.20, 0.76)
0.0079 1
0.36 (0.16, 0.85)
0.0202 1
0.31 (0.15, 0.64)
0.0022
Estrogen
receptor
Negative
Positive
79
156
1
0.41 (0.21, 0.79)
0.0087 1
0.43 (0.14, 1.37)
0.14 1
0.26 (0.12, 0.59)
0.0010
Progesterone
receptor
Negative
Positive
114
121
1
0.45 (0.23, 0.88)
0.0174 1
0.57 (0.19, 1.69)
0.32
HER2
receptor
Negative
Positive
178
57
1
0.67 (0.29, 1.53)
0.32 1
0.27 (0.10, 0.71)
0.0039 1
0.33 (0.14, 0.82)
0.0103
Ibc status No
Yes
167
68
1
2.43 (1.26, 4.70)
0.0105 1
3.45 (1.56, 7.65)
0.0024 1
3.33 (1.56, 7.08)
0.0020
Size (mean) <4.1cm
>=4.1cm
N/A
101
104
30
1
1.74 (0.81, 3.74)
2.12 (0.81, 5.59)
0.22 1
1.35 (0.57, 3.22)
2.05 (0.60, 6.96)
0.52

27
6. References

Cancer.org,. (2015) How is inflammatory breast cancer treated?. URL
http://www.cancer.org/cancer/breastcancer/moreinformation/inflammatorybreastcancer/infl
ammatory-breast-cancer-inflammatory-br-ca-treatment [accessed 23 September 2015]

Cox, D. R. and D. Oakes (1984). Analysis of survival data. London ;
New York, Chapman and Hall.

Kaplan, E. and Meier, P. (1958) Nonparametric Estimation from Incomplete Observations.
Journal of the American Statistical Association, 53, 457.

Lang, J. E., et al. (2014). Inflammatory Breast Cancer IRB proposal.

Liauw, S. L., et al. (2004). "Inflammatory breast carcinoma: outcomes with trimodality
therapy for nonmetastatic disease." Cancer 100(5): 920-928.

National Cancer Institute,. (2015) Inflammatory Breast Cancer. URL
http://www.cancer.gov/types/breast/ibc-fact-sheet [accessed 23 September 2015]

Panades, M., et al. (2005). "Evolving treatment strategies for inflammatory breast
cancer: a population-based survival analysis." J Clin Oncol 23(9): 1941-1950.

Schinkel, J. K., et al. (2014). "Racial/ethnic differences in breast cancer survival
by inflammatory status and hormonal receptor status: an analysis of
the Surveillance, Epidemiology, and End Results data." Cancer Causes
Control 25(8): 959-968.

Smoot, R. L., et al. (2006). "A single-center experience with inflammatory breast
cancer 1985-2003." Arch Surg 141(6): 567-572; discussion 572-563.

Walters, S. (2009) What Is...? Series, 2nd ed. Hayward Medical Communications.

Abstract (if available)

Abstract OBJECTIVE: To examine if Inflammatory Breast Cancer (IBC) have worse disease-free and overall survival than non-Inflammatory Breast Cancer (non-IBC) patients. ❧ METHODS: Overall survival (OS) and disease-free survival (DFS) and locoregional control survival (LRC) were compared between IBC and non-Inflammatory Breast Cancer (non-IBC) Stage III patients. OS and progression-free survival (PFS) were compared between IBC and non-IBC Stage IV patients. Univariate analysis was first done by OS and DFS analyses with Kaplan Meier method for all IBC and non-IBC patients. Cox proportional hazards regression modeling was then used to obtain hazard ratios and likelihood ratio p-values. Cumulative incidence analysis was used to estimate and compare rates of different competing risks for IBC and non-IBC patients. Multivariate analyses through Cox proportional hazard regression modeling and likelihood ratio tests were performed to determine the important prognostic factors for overall survival and disease-free survival. ❧ RESULTS: For Stage III patients, OS, DFS and LRC were highly significantly different between IBC and non-IBC patients (p=0.0032, p=0.0005 and p=0.0012, respectively). Race, tumor histology, estrogen receptor status, and tumor size were significantly associated with both OS and DFS in multivariate analysis. African American race, invasive ductal histology, ER-negative receptor status, and larger tumor size were associated with poorer survival. ❧ CONCLUSION: IBC is an aggressive breast cancer with high recurrence rate and poor clinical prognosis. Stage III IBC patients have a poorer survival outcome than Stage III non-IBC patients. Further studies need to be done on the prognostic factors for IBC patients.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Air pollution and breast cancer survival in California teachers: using address histories and individual-level data

PDF

Predictive factors of breast cancer survival: a population-based study

PDF

Survival of children and adolescents with low-risk non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated with surgery only: an analysis of 234 patients from the Children’s Oncology Group stud...

PDF

Risk factors and survival outcome in childhood alveolar soft part sarcoma among patients in the Children’s Oncology Group (COG) Phase 3 study ARST0332

PDF

Disparities in colorectal cancer survival among Latinos in California

PDF

Construction of a surgical survival prediction model of stage IV NSCLC patients-based on seer database

PDF

Lymph node metastases in non-rhabdomyosarcoma soft tissue sarcoma

PDF

Incidence and survival rates of the three major histologies of renal cell carcinoma

PDF

An assessment of necrosis grading in childhood osteosarcoma: the effect of initial treatment on prognostic significance

PDF

A novel risk-based treatment strategy evaluated in pediatric head and neck non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) patients: a survival analysis from the Children's Oncology Group study...

PDF

Pathogenic variants in cancer predisposition genes and risk of non-breast multiple primary cancers in breast cancer patients

PDF

The effect of renal function on toxicity of E7389 (eribulin) among patients with bladder cancer

PDF

Instability of heart rate and rating of perceived exertion during high-intensity interval training in breast cancer patients undergoing anthracycline chemotherapy

PDF

The impact of the COVID-19 pandemic on cancer care delivery

PDF

Screening and association testing of coding variation in steroid hormone coactivator and corepressor genes in relationship with breast cancer risk in multiple populations

PDF

The role of heritability and genetic variation in cancer and cancer survival

PDF

Body size and the risk of prostate cancer in the multiethnic cohort

PDF

Arm lymphedema in a multi-ethnic cohort of female breast cancer survivors

PDF

Carboplatin and vincristine chemotherapy for progressive low grade gliomas in pediatric patients with or without neurofibromatosis type 1 (NF1)

PDF

Disparities in gallbladder, intra-hepatic bile duct, and other biliary cancers among multi-ethnic populations: a California Cancer Registry study

Asset Metadata

Creator Wang, Tian (author)

Core Title An analysis of disease-free survival and overall survival in inflammatory breast cancer

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biostatistics and Epidemiology

Publication Date 03/31/2016

Defense Date 03/31/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag inflammatory breast cancer,neoadjuvant chemotherapy,OAI-PMH Harvest,survival analysis

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sposto, Richard (committee chair), Alonzo, Todd (committee member), Mack, Wendy (committee member)

Creator Email tiffanytianwang@gmail.com,wang362@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-223132

Unique identifier UC11278050

Identifier etd-WangTian-4217.pdf (filename),usctheses-c40-223132 (legacy record id)

Legacy Identifier etd-WangTian-4217.pdf

Dmrecord 223132

Document Type Thesis

Format application/pdf (imt)

Rights Wang, Tian

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA