Risk of cardiac death among cancer survivors in the United States; a SEER database analysis

Abstract

Purpose: Population-based data on the risk of cardiac death among cancer survivors are needed. This scenario was evaluated in cancer survivors (>5 years) registered within the Surveillance, Epidemiology and End Results (SEER) database.
Methods: The SEER database was queried using the SEER*Stat program to determine the frequency of cardiac death compared to other causes of death and to determine heart disease-specific and cancer-specific survival rates in each of the 10 most common cancers in men and women in the SEER database (including colorectal, lung, breast, prostate, uterine corpus, urinary bladder, kidney, thyroid cancers, cutaneous melanoma and non-Hodgkin lymphoma). Associations were ascertained with Chi-squared tests and multivariate analyses were conducted through Cox-proportional hazard models.
Results: For the frequency and survival sessions 2513716 patients (diagnosed between 1973 and 2008) were assessed, while for the case listing session 2146496 patients (diagnosed between 1988 and 2008) were assessed. For survivors of all 10 cancers combined, the 10-year cancer specific survival rate was 91.4% while 10-year heart disease-specific survival rate was 94.6%. For cancer-specific survival rates, the highest rates were related to thyroid cancer survivors while the lowest rates were related to lung cancer survivors. For heart disease-specific survival rates, the highest rates were related to thyroid cancer survivors while the lowest rates were related to both lung cancer survivors and urinary bladder cancer survivors. The following factors were associated with a higher likelihood of cardiac death: male gender, old age at diagnosis, black race and local treatment with radiotherapy rather than surgery (P<0.0001 for all parameters).
Conclusion: Among cancer survivors (>5 years), cardiac death is a significant cause of death and there is a wide variability among different cancers in the relative importance of cardiac death vs. cancer-related death

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=iery20
Download by: [Omar Abdel-Rahman] Date: 18 June 2017, At: 08:48
Expert Review of Anticancer Therapy
ISSN: 1473-7140 (Print) 1744-8328 (Online) Journal homepage: http://www.tandfonline.com/loi/iery20
Risk of cardiac death among cancer survivors in
the United States; a SEER database analysis
Omar Abdel-Rahman
To cite this article: Omar Abdel-Rahman (2017): Risk of cardiac death among cancer survivors
in the United States; a SEER database analysis, Expert Review of Anticancer Therapy, DOI:
10.1080/14737140.2017.1344099
To link to this article: http://dx.doi.org/10.1080/14737140.2017.1344099
Accepted author version posted online: 16
Jun 2017.
Submit your article to this journal
View related articles
View Crossmark data

Accepted Manuscript
Publisher: Taylor & Francis
Journal: Expert Review of Anticancer Therapy
DOI: 10.1080/14737140.2017.1344099
Original Research
Risk of cardiac death among cancer survivors in the United States; a SEER
database analysis
Omar Abdel-Rahman
Clinical Oncology Department, Faculty of Medicine, Ain Shams University, Cairo,
Egypt.
Corresponding author:
Omar Abdel-Rahman (omar.abdelrhman@med.asu.edu.eg)

Accepted Manuscript
Abstract:
Purpose: Population-based data on the risk of cardiac death among cancer survivors
are needed. This scenario was evaluated in cancer survivors (>5 years) registered
within the Surveillance, Epidemiology and End Results (SEER) database.
Methods: The SEER database was queried using the SEER*Stat program to determine
the frequency of cardiac death compared to other causes of death and to determine
heart disease-specific and cancer-specific survival rates in each of the 10 most common
cancers in men and women in the SEER database (including colorectal, lung, breast,
prostate, uterine corpus, urinary bladder, kidney, thyroid cancers, cutaneous melanoma
and non-Hodgkin lymphoma). Associations were ascertained with Chi-squared tests
and multivariate analyses were conducted through Cox-proportional hazard models.
Results: For the frequency and survival sessions 2513716 patients (diagnosed between
1973 and 2008) were assessed, while for the case listing session 2146496 patients
(diagnosed between 1988 and 2008) were assessed. For survivors of all 10 cancers
combined, the 10-year cancer specific survival rate was 91.4% while 10-year heart
disease-specific survival rate was 94.6%. For cancer-specific survival rates, the highest
rates were related to thyroid cancer survivors while the lowest rates were related to lung
cancer survivors. For heart disease-specific survival rates, the highest rates were
related to thyroid cancer survivors while the lowest rates were related to both lung
cancer survivors and urinary bladder cancer survivors. The following factors were
associated with a higher likelihood of cardiac death: male gender, old age at diagnosis,
black race and local treatment with radiotherapy rather than surgery (P<0.0001 for all
parameters).
Conclusion: Among cancer survivors (>5 years), cardiac death is a significant cause of
death and there is a wide variability among different cancers in the relative importance
of cardiac death vs. cancer-related death.
Keywords: Cancer; cardiac disease; cardiac death; lung cancer; breast cancer.

Accepted Manuscript
1. Introduction:
Cancer and cardiac disease are the two global epidemics of our age. Cancer has
claimed the lives of approximately 8.2 million humans worldwide in 2012 [1]. Moreover,
the number of new cases of cancer is expected to increase by approximately 70% in the
coming 2 decades [2]. On the other hand, cardiovascular disease is the number one
cause of death worldwide; with more people dying from it than from any other cause of
death and it has claimed the lives of 17.5 million humans worldwide in 2012 [3].
Cancer management intersects with cardiac health at many points. Etiologically, many
factors predispose to both cancer and heart disease (e.g. tobacco smoking, obesity and
excess alcohol consumption) [4-7]. Thus, many patients with cancer are already at a
higher risk for cardiac disease because of background medical conditions. Moreover,
many anti-cancer treatments (including chemotherapy, radiotherapy and targeted
therapy) may cause cardiac injury as a side effect [8]. This may contribute further to the
risk of cardiac events in cancer patients. The majority of treatment-associated adverse
cardiac effects develop latently years after administration of anticancer treatment [9].
Thus, long-term cancer survivors are at a particularly high risk for developing cardiac
disease and cardiac death [10]. This risk of course varies according to the co-
morbidities, administered anti-cancer treatments as well as underlying cancer diagnosis.
Currently, there are scarce population-based data describing the real weight of cardiac
death in long term cancer survivors compared to other competing causes of death (both
cancer- and non cancer-related). These data are fundamental in guiding the cardiac-
sparing interventions while treating our patients as well as in planning appropriate follow
up strategies for the long-term survivors.
The objective of this analysis was to review the risk of cardiac death among cancer
survivors (>5 years since the initial diagnosis) who were diagnosed initially with any of
the 10 most common cancers in women and men in the Surveillance, Epidemiology and
End Results (SEER) database.

Accepted Manuscript
2. Methodology:
The data for this study were extracted from the SEER-18 registry of the US National
Cancer Institute[11]. Data were retrieved utilizing the SEER*Stat software Version 8.3.2.
2.1. Selection of the study cohort:
To identify eligible patients diagnosed with one the 10 cancers, the following parameters
were selected in the selection tab of the SEER*Stat:
(Site and Morphology.Site recode ICD-O-3/WHO 2008) = ' Colon and Rectum','
Lung and Bronchus',' Melanoma of the Skin',' Breast',' Corpus Uteri',' Prostate','
Urinary Bladder',' Kidney and Renal Pelvis',' Thyroid',' Non-Hodgkin Lymphoma' )
The included patients were further restricted to those surviving more than 5 years (60
months).
Cardiac death was considered when “diseases of the heart” was mentioned as the
cause of death in cause-specific death classification. Cancer registries within the SEER
database use algorithms to process causes of death from death certificate in order to
identify a single, disease-specific, cause of death. However, in some cases,
misattribution may occur in some cases (e.g. a death may be attributed to the site of
metastasis rather than the primary site).
The above parameters have been utilized in the three types of assessments (known as
sessions within the SEER*Stat); namely: frequency session, survival session and case
listing session. For frequency and survival sessions, the search was limited to SEER
database (1973-2008) ; while for case listing session, the search was limited to SEER
database (1988-2008) because AJCC staging information were not available before
1988.
2.2. Data collection:
For the case listing session, covariates extracted for each case included site of the
primary tumor, age at diagnosis, gender, race, stage, laterality (if applicable), surgery or
radiotherapy and cause-specific death classification. Data regarding systemic therapy
and co-morbidities were not available in the SEER database.

Accepted Manuscript
2.3. Statistical analysis:
In this study, baseline clinico-pathological features for patients who did or did not
develop cardiac death were evaluated and compared. Chi-Squared test was used to
compare the features of patients who did or did not develop cardiac death. Cox
proportional hazards model was used to conduct multivariate analyses of factors
affecting cardiac-specific mortality among different evaluated cancers (factors evaluated
include age diagnosis, race, AJCC stage at diagnosis, local treatment modality and
laterality if applicable). Statistical significance was declared with a two-sided p value <
0.05.
Survival sessions in the SEER*Stat were utilized to calculate heart disease-specific
survival rates and cancer specific survival rates for each of the evaluated cancers
across different years following the initial diagnosis. For the survival analysis, it was
restricted to cases in the research database, with known age and who were actively
followed. The following cases were excluded from the survival analysis: Death
certificate-only and autopsy-only diagnoses, cases with unknown/missing cause of
death and alive cases with no survival time.
All of the statistical analyses were conducted with SPSS Statistics 20.0 (IBM, NY) as
well as the Seer*stat program; Microsoft excel was also utilized to produce figures
regarding the frequency of cardiac death, cancer-specific survival rate and heart
disease-specific survival rate among the evaluated cancers. The study cutoff is the date
of submission of most recent SEER data (December 2013).
3. Results:
3.1. Frequency session results:
The vital status of all evaluated patients in the frequency session (evaluating 2513716
patients diagnosed 1973-2008) has been assessed up to the study cutoff; and among
them, 70% were alive, 6% died because of cardiac disease, 7% died because of the
primary cancer and 17% died because of other causes of death (among which
approximately 6% died of second primary cancer). Stratification of the cause of death
according to the duration since diagnosis revealed that with more time since the initial
diagnosis, the role of cardiac death increased, while the role of cancer-specific death

Accepted Manuscript
decreased (figure-1). An additional detailed analysis of vital status as well as
stratification of the cause of death according to duration since diagnosis has been
conducted for each of the evaluated 10 cancers (supplementary figures 1-10). This
analysis showed wide variation among the evaluated cancers; while in some cancers
(lung cancer) the role of cancer-related death is far more important among cancer
survivors even up to 19 years post diagnosis, in other cancers (urinary bladder cancer)
the role of cardiac death is more significant starting from 5 years post diagnosis.
3.2. Survival session results:
The cancer-specific survival rate and heart disease-specific survival rate of all patients
(2513716 patients diagnosed 1973-2008) have been evaluated stratifying the outcomes
according to the site of the primary tumor and according to the duration since the year
of diagnosis (figures 2a-2b). For survivors of the evaluated 10 cancers combined, 10-
year cancer-specific survival rate was: 91.4% and 15-year cancer–specific survival rate
was: 85.5%. On the other hand for survivors of the evaluated 10 cancers, 10-year heart
disease-specific survival rate was: 94.6% and 15-year heart disease-specific survival
rate was: 88.4%.
For cancer-specific survival rate, the highest rates were related to thyroid cancer
survivors; while the lowest rates were related to lung cancer survivors. For heart
disease-specific survival rate, the highest rates were related to thyroid cancer survivors;
while the lowest rates were related to both lung cancer survivors and urinary bladder
cancer survivors.
3.3. Case listing session results:
C.1. Clinico-pathological characteristics of patients with or without cardiac death:
A total of 2146496 cancer survivors diagnosed (1988-2008) were identified and included
in this session. Table-1 summarizes the clinico-pathological characteristics of cancer
survivors who did or did not develop cardiac death. Overall, the following characteristics
were associated with a higher likelihood of cardiac death, including: male gender, old
age at diagnosis, black race, and local treatment with radiotherapy rather than surgery
(P<0.0001 for all parameters). The reason for the comparison displayed in table-1 is to

Accepted Manuscript
showcase the baseline characteristics of patients who developed a cardiac death
compared to the overall patient population.
C.2. Multivariate analysis of factors affecting cardiac-specific mortality:
An additional multivariate analysis assessing the factors affecting cardiac-specific
mortality among different cancers is detailed in supplementary tables 1-10. Overall,
factors associated with worse cardiac-specific mortality among different cancers
included older age at diagnosis, male gender, black race and local treatment with
surgery (P<0.05 for all parameters). Local treatment with radiotherapy was associated
with worse cardiac-specific mortality in lung cancer survivors only. It has however to be
acknowledged that treatment data in SEER are available only for the first course of
therapy.
4. Discussion:
The most notable findings of this study are: (1) Cardiac death is a more important cause
of death than cancer itself 10 years following the diagnosis of cancer; (2) there is a wide
variability among different cancers in the relative importance of cardiac death vs.
cancer-related death among cancer survivors; and (3) five years should not be taken as
the universal limit for cancer cure because cancer-related death continue to occur
beyond this limit.
In this study, patients at highest risk for the development of cardiac death are those with
a higher baseline risk for cardiac disease. This analysis showed that male patients are
more likely to experience cardiac death and in multivariate analysis they have a worse
cardiac-specific mortality; this is in line with data showing a higher baseline risk of
cardiovascular disease among men compared to women[12,13]. This analysis also
showed that African Americans are more likely to suffer from cardiac deaths compared
to other race categories. This is explained by data suggesting a higher background
cardiac risk among African Americans [14]. Additionally, this analysis showed that
elderly people are more prone to cardiac death events. This is expected given the
higher probability of co-morbidities in this age group which may predispose to a higher
cardiac events risk.

Accepted Manuscript
Moreover, patients who were treated with surgery generally have a lower risk of cardiac
death compared to those who were not. This may be explained by the fact that cardiac
co-morbidity is an important reason for exclusion from radical cancer surgery; thus, the
association of surgery with lower cardiac death risk may actually be a form of selection
bias. Lung cancer patients receiving radiotherapy have a shorter time to cardiac death
in multivariate analysis, this is in line with previous population-based data suggesting
enhanced cardiac risk following thoracic radiotherapy for lung cancer as well as for
other thoracic cancers [15,16]. The same finding was not replicated with breast cancer;
probably because of the fact that both right and left sided breast cancer patients were
collectively grouped in the analysis which may have ameliorated the risk expected
essentially with left-sided tumors[17]. For non Hodgkin lymphoma survivors, the
administration of radiotherapy was associated with better cardiac-specific mortality in
multivariate analysis. This is in line with previous data and may be explained by the
increased exposure to anthracyclines in patients not treated with radiotherapy [18].
Looking again at the heart disease-specific survival rates of different cancers (Figure 2-
b), it is interesting to note that the lower rates are probably linked to tobacco-related
cancers (e.g. lung cancer, urinary bladder cancer and kidney/renal pelvis cancer). This
underlines the important role of background smoking history in assessing the cardiac
risk of cancer survivors.
Looking further at the cancer-specific survival curves of different cancers (Figure 2-a) as
well as distribution of causes of death according to the duration since diagnosis, it is
interesting to note that cancer-related death continued to occur without a plateau even
19 years after diagnosis in the majority of cancers. This finding challenges the common
rule of thumb of considering 5 years survival as the arbitrary cutoff after which the
patient may be considered cured from cancer. This finding is also in line with previously
published population-based data [19].
The results of this analysis should be interpreted with caution given the absence of
information about other potential risk factors for cardiac disease such as smoking,
alcohol consumption, and co-morbidities in the SEER database. Moreover, details of
potentially cardiotoxic systemic therapy as well as technical details/doses of
radiotherapy were not available. Additionally, the exact cardiac disease/events which

Accepted Manuscript
led to death were not mentioned which may compromise our understanding of the true
correlation between cancer diagnosis/treatment and cardiac death.
The staging of patients included in the current analysis includes 2 different editions of
the AJCC system (3rd edition for those diagnosed from 1988-2003; and 6th edition for
those diagnosed from 2004-2008). This may have partly confounded the assessment of
the impact of the stage on survival.
Another important caveat for the interpretation of these results is the dependence of the
current analysis on the cause of death recorded within death certificates and /or hospital
records. Although the SEER database itself has a rigorous quality assurance program,
similar quality cannot be guaranteed of every medical registrar who recorded the cause
of death for each patient, particularly in the presence of some data casting doubt about
the level of accuracy of the cause of death reported in death certificates [20]. It is also
thought that a certifier is more likely to invoke cancer as a cause of death if it was
recently diagnosed, even if the patient died of something else [21,22]. However, this is
less of a concern here given the fact that all evaluated patients have been diagnosed
since 5 years at least.
It has also to be noted that given the study cutoff of 2013, some recently diagnosed
patients have been censored and did not have sufficient follow up time. This caveat is
relevant when interpreting different supplementary figures for each tumor site.
Moreover, statements made in the results should take into account the fact that the
analysis was conditional on surviving at least 5 years.
The practical implications of the current analysis indicate the need to: (1) pay more
attention to potentially cardiotoxic therapies that are administered to our patients; (2)
tailor follow up strategies of cancer survivors not only according to the risks of cancer
relapse but also according to the risk of cardiac events.
5. Conclusions:
Within the stated limitations of the above analysis, cardiac death is a more important
cause of death than cancer itself 10 years following the diagnosis of cancer and there is
a wide variability among different cancers in the relative importance of cardiac death vs.
cancer-related death among cancer survivors. These data are supportive of a more

Accepted Manuscript
meticulous cardiac-sparing treatment planning as well as tailored follow up strategies for
cardiac disease among cancer survivors.
Funding
This manuscript has not received any funding.
Declaration of interest
The author has no relevant affiliations or financial involvement with any organization or
entity with a financial interest in or financial conflict with the subject matter or materials
discussed in the manuscript. This includes employment, consultancies, honoraria, stock
ownership or options, expert testimony, grants or patents received or pending, or
royalties.

Accepted Manuscript
References
Reference annotations
* Of interest
** Of considerable interest
1. http://publications.iarc.fr/Non-Series-Publications/World-Cancer-Reports/World-
Cancer-Report-2014. last accessed on 17/11/2016
2. http://www.who.int/mediacentre/factsheets/fs297/en/. last accessed on
17/11/2016
3. http://www.who.int/nmh/publications/ncd-status-report-2014/en/ last accessed on
17/11/2016
4. Simmons VN, Pineiro B, Hooper MW et al. Tobacco-Related Health Disparities
Across the Cancer Care Continuum. Cancer Control, 23, 434-41 (2016)
5. Carroll AJ, Labarthe DR, Huffman MD et al. Global tobacco prevention and
control in relation to a cardiovascular health promotion and disease prevention
framework: A narrative review. Prev Med, 93, 189-97 (2016)
6. O'Keefe JH, Bhatti SK, Bajwa A et al. Alcohol and cardiovascular health: the
dose makes the poison...or the remedy. Mayo Clin Proc, 89, 382-93 (2014)
7. Simon TG, Corey KE, Chung RT et al. Cardiovascular Risk Reduction in Patients
with Nonalcoholic Fatty Liver Disease: The Potential Role of Ezetimibe. Dig Dis
Sci, 61, 3425-35 (2016)
8. Ahmed SH, Moussa Sherif DE, Fouad Y et al. Principles of a risk evaluation and
mitigation strategy (REMS) for breast cancer patients receiving potentially
cardiotoxic adjuvant treatments. Expert Opin Drug Saf, 15, 911-23 (2016)
9. Abdel-Rahman O, Alorabi M. Use of angiotensin-converting enzyme inhibitors in
the prophylaxis of anthracycline or trastuzumab-related cardiac dysfunction:
preclinical and clinical considerations. Expert Rev Anticancer Ther, 15, 829-37
(2015)
10. Abdel-Rahman O. Causes of death in long-term lung cancer survivors: a SEER
database analysis. Curr Med Res Opin, 1-16 (2017)
11. *Surveillance, Epidemiology and End Results Program. About the SEER
Program. Available at: http://seer.cancer.gov/about. Accessed 25th June 2016.
This is the website for accessing SEER database.
12. **Maas A, Appelman YEA. Gender differences in coronary heart disease. Neth
Heart J, 18, 598-602 (2010). This study evaluated gender differences in coronary
heart disease.
13. Rivera MP. Lung cancer in women: differences in epidemiology, biology,
histology, and treatment outcomes. Semin Respir Crit Care Med, 34, 792-801
(2013)
14. Jolly S, Vittinghoff E, Chattopadhyay A et al. Higher cardiovascular disease
prevalence and mortality among younger blacks compared to whites. Am J Med,
123, 811-8 (2010)

Accepted Manuscript
15. Lin SH, Zhang N, Godby J et al. Radiation modality use and cardiopulmonary
mortality risk in elderly patients with esophageal cancer. Cancer, 122, 917-28
(2016)
16. Park HS, Harder EM, Mancini BR et al. Central versus Peripheral Tumor
Location: Influence on Survival, Local Control, and Toxicity Following
Stereotactic Body Radiotherapy for Primary Non-Small-Cell Lung Cancer. J
Thorac Oncol, 10, 832-7 (2015)
17. Ye JC, Yan W, Christos P et al. Second cancer, breast cancer, and cardiac
mortality in stage T1aN0 breast cancer patients with or without external beam
radiation therapy: a national registry study. Clin Breast Cancer, 15, 54-9 (2015)
18. *Pugh TJ, Ballonoff A, Rusthoven KE et al. Cardiac mortality in patients with
stage I and II diffuse large B-cell lymphoma treated with and without radiation: a
surveillance, epidemiology, and end-results analysis. Int J Radiat Oncol Biol
Phys, 76, 845-9 (2010). This study evaluated cardiac mortality in large cell
lymphomas.
19. Hubbard MO, Fu P, Margevicius S et al. Five-year survival does not equal cure in
non&#x2013;small cell lung cancer: A Surveillance, Epidemiology, and End
Results&#x2013;based analysis of variables affecting 10- to 18-year survival.
The Journal of Thoracic and Cardiovascular Surgery, 143, 1307-13)
20. Halanych JH, Shuaib F, Parmar G et al. Agreement on Cause of Death Between
Proxies, Death Certificates, and Clinician Adjudicators in the Reasons for
Geographic and Racial Differences in Stroke (REGARDS) Study. Am J
Epidemiol, 173, 1319-26 (2011)
21. Asnaes S, Frederiksen V, Fenger C. The value of the hospital autopsy. A study
of causes and modes of death estimated before and after autopsy. Forensic Sci
Int, 21, 23-32 (1983)
22. Masocco M, Kodra Y, Vichi M et al. Mortality associated with neurofibromatosis
type 1: a study based on Italian death certificates (1995-2006). Orphanet J Rare
Dis, 6, 11 (2011)

Accepted Manuscript
Table-1: baseline characteristics of cancer survivors included in the analysis (N.= 2146496) (Case listing session was limited to patients diagnosed
between 1988-2008):
Parameter Patients with cardiac death (N.= 138557) Patients without cardiac death (N.= 2007939) P value
Cancer type
Colorectal
Lung
Breast
Prostate
Endometrium
Melanoma
Thyroid
Urinary bladder
Kidney/renal
pelvis
Non Hodgkin
lymphoma
23812 (17.2%)
6053 (4.4%)
27955 (20.2%)
46886 (33.8%)
5536 (4%)
5576 (4%)
1121 (0.8%)
11809 (8.5%)
4205 (3%)
5604 (4.1%)
240880 (12.1%)
80737 (4%)
538794 (26.8%)
544999 (27.1%)
91077 (4.5%)
145030 (7.2%)
80566 (4%)
114190 (5.7%)
71344 (3.6%)
100322 (5%)
<0.0001
Age
< 40 years
40-70 years
>70 years
402 (0.3%)
41699 (30.1%)
96456 (69.6%)
121431 (6%)
1279681 (63.7%)
606827 (30.2%)
<0.0001
Gender
Male
Female
81252 (58.6%)
57305 (41.4%)
984226 (49%)
1023713 (51%)
<0.0001
Race
White
Black
Others
119775 (86.4%)
11850 (8.6%)
6923 (5%)
1692658 (84.3%)
171214 (8.5%)
144067 (7.1%)
<0.0001
Histology
Adenocarcinoma
SCC
Others
7907 (57.2%)
2487 (1.8%)
56863 (41%)
1008195 (50.2%)
38548 (1.9%)
957196 (47.9%)
<0.0001
AJCC stage
0
2596 (1.9%)
42562 (2.1%)
<0.0001

Accepted Manuscript
I
II
III
Unknown
38731 (28%)
26099 (18.8%)
10304 (7.4%)
60827 (43.9%)
626314 (31.2%)
521318 (26%)
176776 (8.8%)
640969 (31.9%)
Surgery
Yes
No
Unknown
64831 (46.8%)
73345 (52.9%)
381 (0.3%)
1063282 (53%)
937828 (46.7%)
6829 (0.3%)
<0.0001
Radiation
therapy
Yes
No
Unknown
30696 (22.2%)
103022 (74.4%)
4839 (3.5%)
499621 (24.9%)
1378133(68.6%)
130185 (6.5%)
<0.0001

Accepted Manuscript
Figure
-
cutoff.
(
time fr
o
-
-
1: (a) Pie
c
(
b) Bar ch
a
o
m diagno
s
c
hart of th
e
a
rt of the v
i
s
is.
e
vital stat
u
i
tal status
o
u
s of canc
e
o
f cancer
s
e
r survivor
s
s
urvivors (
1
s
(1973-2
0
1
973-2008
0
08) at the
) stratified
study
by the

Accepted Manuscript
Figure
2: (a) Can
c
c
er-specifi
c
c
survival
r
r
ates for the 10 canc
e
e
rs evalua
t
t
ed in the
a
a
nalysis.

Accepted Manuscript
Figure
analysi
2: (b) Hea
r
s.
r
t disease-
-
specific s
u
u
rvival rate
s
s
for the 1
0
0
cancers
e
e
valuated in the

Accepted Manuscript
S. figure
-
S. figure
-
from dia
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
-
1a: Pie char
-
1b: Bar cha
gnosis; y-ax
i
Vit
a
Alive
Death
5-7
years
7-9
year
t of the vital
rt of the vit
a
i
s: number o
f
9%
8%
a
l statu
s
due to prima
r
s
9-11
years
11-
1
yea
r
status of co
l
a
l status of c
o
f
patients/
x
-
6
21%
s
of col
o
surviv
o
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
l
orectal can
c
o
lorectal can
-
axis: years
a
6
2%
o
rectal
c
o
rs
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
c
er survivors
cer survivor
s
a
fter diagnos
c
ancer
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
(1973-2008
)
s
(1973-200
8
is.
t
h
o primary
e
s of death
)
at the stud
y
8
) stratified
b
y
cutoff.
b
y the time

Accepted Manuscript
S. figure
-
S. figure
-
diagnosi
0
2000
4000
6000
8000
10000
12000
14000
16000
-
2a: Pie char
-
2b: Bar cha
s; y-axis: nu
m
Vital
s
Alive
Death
5-7
years
7-9
year
t of the vital
rt of the vit
a
m
ber of pati
e
1
9
2
s
tatus o
due to prima
r
s
9-11
years
11-
1
yea
r
status of lu
n
a
l status of l
u
e
nts/ x-axis:
y
7%
9
%
2
4%
f lung c
a
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
n
g cancer su
r
u
ng cancer s
u
y
ears after
d
50%
a
ncer s
u
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
r
vivors (197
3
u
rvivors (19
7
d
iagnosis.
u
rvivor
s
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
3
-2008) at th
e
7
3-2008) stra
s
t
h
o primary
e
s of death
e study cuto
f
tified by the
ff.
time from

Accepted Manuscript
S. figure
-
S. figure
-
from dia
0
1000
0
2000
0
3000
0
4000
0
5000
0
6000
0
7000
0
8000
0
9000
0
10000
0
-
3a: Pie char
-
3b: Bar cha
gnosis; y-ax
i
Vital s
t
Alive
Death
0
0
0
0
0
0
0
0
0
0
0
5-7
years
7-
9
yea
t of the vital
rt of the vit
a
i
s: number o
f
5%
8%
t
atus of
due to prima
r
9
rs
9-11
years
11-
1
yea
status of br
e
a
l status of b
r
f
patients/
x
-
14%
breast
c
C
a
r
y cancer O
t
1
3
rs
13-15
years
15-
1
yea
e
ast cancer
s
r
east cancer
-
axis: years
a
73%
c
ancer
s
a
rdiac death
t
her causes o
f
1
7
rs
17-19
years
s
urvivors (19
survivors (1
9
a
fter diagnos
s
urvivo
r
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
73-2008) at
t
9
73-2008) st
is.
r
s
t
h
o primary
e
s of death
t
he study cu
t
ratified by t
h
t
off.
h
e time

Accepted Manuscript
S. figure
-
S. figure
-
from dia
0
2000
0
4000
0
6000
0
8000
0
10000
0
12000
0
-
4a: Pie char
-
4b: Bar cha
gnosis; y-ax
i
Vit
Alive
Death
0
0
0
0
0
0
0
5-7
years
7-
9
yea
t of the vital
rt of the vit
a
i
s: number o
f
8%
7%
al statu
due to prima
r
9
rs
9-11
years
11-
1
yea
status of pr
o
a
l status of p
r
f
patients/
x
-
6
18%
s of pr
o
surviv
o
C
a
r
y cancer O
t
1
3
rs
13-15
years
15-
1
yea
o
state cance
r
ostate canc
e
-
axis: years
a
6
7%
o
state c
a
o
rs
a
rdiac death
t
her causes o
f
1
7
rs
17-19
years
r survivors (
1
e
r survivors
(
a
fter diagnos
a
ncer
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
1
973-2008)
a
(
1973-2008)
is.
t
h
o primary
e
s of death
a
t the study
c
stratified b
y
c
utoff.
y
the time

Accepted Manuscript
S. figure
-
cutoff.
S. figure
-
time fro
m
0
5000
10000
15000
20000
25000
30000
-
5a: Pie char
-
5b: Bar cha
m
diagnosis;
Vital
s
Alive
Death
5-7
years
7-9
year
t of the vital
rt of the vit
a
y-axis: num
b
4%
3%
s
tatus o
f
due to prima
r
s
9-11
years
11-
1
yea
r
status of m
a
a
l status of
m
b
er of patie
n
82
%
11%
f
malig
n
surviv
o
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
a
lignant mel
a
m
alignant me
n
ts/
x
-axis: y
e
%
n
ant m
e
o
rs
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
a
noma survi
v
lanoma surv
e
ars after di
a
e
lanom
a
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
v
ors (1973-2
0
ivors (1973-
2
a
gnosis.
a
t
h
o primary
e
s of death
008) at the
s
2
008) stratif
i
s
tudy
i
ed by the

Accepted Manuscript
S. figure
-
cutoff.
S. figure
-
the tim
e
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
-
6a: Pie char
-
6b: Bar cha
e
from diagn
o
V
Alive
Death
5-7
years
7-9
year
t of the vital
rt of the vit
a
o
sis; y-axis:
n
5%
11%
1
V
ital sta
t
lymp
h
due to prima
r
s
9-11
years
11-
1
yea
r
status of no
a
l status of n
o
n
umber of p
a
7
1
4%
t
us of n
o
h
oma s
u
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
n Hodgkin l
y
o
n Hodgkin l
a
tients/
x
-a
x
i
s
0%
o
n hod
g
u
rvivor
s
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
y
mphoma su
r
ymphoma s
u
s
: years afte
r
g
kin
s
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
r
vivors (197
3
u
rvivors (19
7
r
diagnosis.
t
h
o primary
e
s of death
3
-2008) at th
7
3-2008) str
a
e study
a
tified by

Accepted Manuscript
S. figure
-
cutoff.
S. figure
-
the tim
e
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
-
7a: Pie char
-
7b: Bar cha
e
from diagn
o
Vital s
t
Alive
Death
5-7
years
7-9
year
t of the vital
rt of the vit
a
o
sis; y-axis:
n
9%
6%
t
atus o
f
due to prima
r
s
9-11
years
11-
1
yea
r
status of ur
i
a
l status of u
r
n
umber of p
a
6
24%
f
urinar
y
surviv
o
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
i
nary bladde
r
r
inary bladd
e
a
tients/
x
-axi
s
6
1%
y
bladd
e
o
rs
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
r
cancer sur
v
e
r cancer su
r
s
: years afte
r
e
r canc
e
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
v
ivors (1973-
r
vivors (197
3
r
diagnosis.
e
r
t
h
o primary
e
s of death
2008) at the
3
-2008) strat
i
study
i
fied by

Accepted Manuscript
S. figure
-
S. figure
-
from dia
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
-
8a: Pie char
-
8b: Bar cha
gnosis; y-ax
i
Vital s
t
Alive
Death
5-7
years
7-9
year
t of the vital
rt of the vit
a
i
s: number o
f
6%
7%
t
atus of
due to prima
r
s
9-11
years
11-
1
yea
r
status of ki
d
a
l status of ki
f
patients/
x
-
15%
kidney
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
d
ney cancer
s
dney cancer
-
axis: years
a
72%
cancer
s
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
s
urvivors (1
9
survivors (1
a
fter diagnos
s
urvivo
r
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
9
73-2008) at
973-2008) s
t
is.
r
s
t
h
o primary
e
s of death
the study cu
t
ratified by t
h
toff.
h
e time

Accepted Manuscript
S. figure
-
S. figure
-
from dia
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
-
9a: Pie char
-
9b: Bar cha
gnosis; y-ax
i
Alive
5-7
years
7-9
year
t of the vital
rt of the vit
a
i
s: number o
f
Vital st
a
Cardiac d
e
s
9-11
years
11-
1
yea
r
status of th
y
a
l status of t
h
f
patients/
x
-
1%
2
%
a
tus of
t
e
ath De
a
1
3
r
s
13-15
years
15-
1
yea
r
y
roid cancer
h
yroid cance
r
-
axis: years
a
9
2
%
5%
t
hyroid
a
th due to pri
m
1
7
r
s
17-19
years
survivors (1
9
r
survivors (
1
a
fter diagnos
2
%
cancer
m
ary cancer
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
9
73-2008) a
t
1
973-2008) s
is.
survivo
Other ca
u
t
h
o primary
e
s of death
t
the study c
u
s
tratified by
t
rs
u
ses of death
u
toff.
t
he time

Accepted Manuscript
S. figure
-
cutoff.
S. figure
-
the tim
e
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
-
10a: Pie ch
a
-
10b: Bar c
h
e
from diagn
o
Vital
s
Alive
Death
5-7
years
7-9
year
a
rt of the vit
a
h
art of the vi
t
o
sis; y-axis:
n
6%
3%
1
s
tatus o
f
due to prima
r
s
9-11
years
11-
1
yea
r
a
l status of u
t
al status of
u
n
umber of p
a
7
6
1
5%
f
uterin
e
surviv
o
C
a
r
y cancer O
t
1
3
r
s
13-15
years
15-
1
yea
r
terine corpu
u
terine corp
u
a
tients/
x
-axi
s
6
%
e
corpu
o
rs
a
rdiac death
t
her causes o
f
1
7
r
s
17-19
years
s cancer sur
v
u
s cancer su
r
s
: years afte
r
s cance
f
death
Alive
Cardiac dea
t
Death due t
o
cancer
Other caus
e
v
ivors (1973
-
rvivors (197
3
r
diagnosis.
r
t
h
o primary
e
s of death
-
2008) at th
e
3
-2008) stra
t
e
study
t
ified by

Accepted Manuscript
Supplementary table-1: Multivariate analysis of factors affecting cardiac-specific mortality among
colorectal cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.095 (0.066-0.135) <0.0001
>70 6.092 (5.906-6.284) <0.0001
Race
White
Ref. (1.00)
Black 1.096 (1.019-1.122) 0.006
Others 0.664 (0.628-0.702) <0.0001
Gender
Male
Ref. (1.00)
Female 0.794 (0.774-0.815) <0.0001
Surgery
Yes
Ref. (1.00)
No 1.057 (1.009-1.107) 0.019
Radiation
yes
Ref. (1.00)
No 1.250 (1.187-1.317) <0.0001
AJCC stage
Stage o
Ref. (1.00)
Stage I 1.025 (0.942-1.117) 0.563
Stage II 1.086 (1.001-1.178) 0.047
Stage III 1.188 (1.094-1.290) <0.0001
Supplementary table-2: Multivariate analysis of factors affecting cardiac-specific mortality among lung
cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.090 (0.050-0.163) <0.0001
>70 3.461 (3.282-3.650) <0.0001

Accepted Manuscript
Race
White
Ref. (1.00)
Black 1.264 (1.161-1.377) <0.0001
Others 0.608 (0.535-0.691) <0.0001
Gender
Male
Ref. (1.00)
Female 0.660 (0.627-0.695) <0.0001
Surgery
Yes
Ref. (1.00)
No 1.390 (1.308-1.477) <0.0001
Radiation
yes
Ref. (1.00)
No 0.771 (0.718-0.827) <0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.955 (0.875-1.042) 0.298
Stage I 1.084 (0.996-1.181) 0.063
Laterality
Left
Ref. (1.00)
Right 0.901 (0.856-0.949) <0.0001
Supplementary table-3: Multivariate analysis of factors affecting cardiac-specific mortality among breast
cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.132 (0.107-0.163) <0.0001
>70 9.519 (9.248-9.798) <0.0001
Race
White
Ref. (1.00)
Black 1.365 (1.302-1.431) <0.0001
Others 0.607 (0.568-0.649) <0.0001
Gender
Male
Ref. (1.00)
Female 0.702 (0.612-0.804) <0.0001
Surgery
Yes
Ref. (1.00)

Accepted Manuscript
No 1.050 (1.014-1.088) 0.006
Radiation
yes
Ref. (1.00)
No 1.477 (1.425-1.531) <0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.721 (0.677-0.768) <0.0001
Stage I 0.773 (0.726-0.824) <0.0001
Laterality
Left
Ref. (1.00)
Right 0.987 (0.963-1.012) 0.318
Supplementary table-4: Multivariate analysis of factors affecting cardiac-specific mortality among
cutaneous malignant melanoma survivors(1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.070 (0.052-0.094) <0.0001
>70 11.004 (10.366-11.681) <0.0001
Race
White
Ref. (1.00)
Black 1.453 (1.037-2.036) 0.030
Others 0.833 (0.585-1.186) 0.311
Gender
Male
Ref. (1.00)
Female 0.684 (0.648-0.723) <0.0001
Surgery
Yes
Ref. (1.00)
No 1.050 (0.986-1.119) 0.130
Radiation
yes
Ref. (1.00)
No 0.942 (0.665-1.335) 0.737
Year of diagnosis
2000-2008
Ref. (1.00)
1988-2000 1.182 (1.108-1.261) <0.0001
AJCC stage

Accepted Manuscript
Stage III
Ref. (1.00)
Stage II 0.900 (0.792-1.022) 0.103
Stage I 1.026 (0.910-1.158) 0.672
Supplementary table-5: Multivariate analysis of factors affecting cardiac-specific mortality among
uterine corpus cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.237 (0.158-0.355) <0.0001
>70 6.559 (6.166-6.978) <0.0001
Race
White
Ref. (1.00)
Black 1.397 (1.221-1.598) <0.0001
Others 0.663 (0.568-0.775) <0.0001
Surgery
Yes
Ref. (1.00)
No 1.938 (1.599-2.350) <0.0001
Radiation
yes
Ref. (1.00)
No 0. 986 (0.915-1.062) 0. 712
Year of diagnosis
2000-2008
Ref. (1.00)
1988-2000 1.269 (1.182-1.362) <0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.947 (0.897-1.000) 0.049
Stage I 1.114 (1.022-1.214) 0.014
Supplementary table-6: Multivariate analysis of factors affecting cardiac-specific mortality among
prostate cancer survivors (1988-2008):

Accepted Manuscript
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.354 (0.050-2.510) 0.298
>70 3.445 (3.307-3.588) <0.0001
Race
White
Ref. (1.00)
Black 1.250 (1.181-1.323) <0.0001
Others 0.725 (0.661-0.796) <0.0001
Surgery
Yes
Ref. (1.00)
No 2.828 (2.680-2.983) <0.0001
Radiation
yes
Ref. (1.00)
No 1.280 (1.221-1.341) <0.0001
Year of diagnosis
2000-2008
Ref. (1.00)
1988-2000 1.555 (1.482-1.632) <0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.938 (0.869-1.012) 0.098
Stage I 0.907 (0.868-0.947) <0.0001
Supplementary table-7: Multivariate analysis of factors affecting cardiac-specific mortality among
urinary bladder cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0. 065 (0.050-0.157) 0.101
>70 5.151 (4.940-5.371) <0.0001
Race
White
Ref. (1.00)

Accepted Manuscript
Black 1.102 (1.000-1.213) 0.050
Others 0.676 (0.609-0.752) <0.0001
Surgery
Yes
Ref. (1.00)
No 1.215 (1.110-1.329) <0.0001
Radiation
yes
Ref. (1.00)
No 0.694 (0.607-0.793) <0.0001
Year of diagnosis
2000-2008
Ref. (1.00)
1988-2000 1.242 (1.193-1.294) <0.0001
Gender
Male
Ref. (1.00)
Female 0.753 (0.720-0.786) <0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.894 (0.779-1.027) 0.113
Stage I 0.936 (0.815-1.075) 0.350
Supplementary table-8: Multivariate analysis of factors affecting cardiac-specific mortality among kidney
cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.113 (0.080-0.161) <0.0001
>70 4.722 (4.434-5.030) <0.0001
Race
White
Ref. (1.00)
Black 1.381 (1.254-1.522) <0.0001
Others 0.759 (0.650-0.886) <0.0001
Gender
Male
Ref. (1.00)
Female 0.821 (0.771-0.874) <0.0001
Surgery
Yes
Ref. (1.00)
No 1.292 (1.191-1.402) <0.0001

Accepted Manuscript
Radiation
yes
Ref. (1.00)
No 1.294 (0.883-1.896) 0.186
Year of diagnosis
2000-2008
Ref. (1.00)
1988-2000
1.231 (1.143-1.326)
<0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.914 (0.756-0.932) 0.353
Stage I 0.839 (0.756-0.932) 0.001
Laterality
Left
Ref. (1.00)
Right 0.994 (0.935-1.056) 0.842
Supplementary table-9: Multivariate analysis of factors affecting cardiac-specific mortality among
thyroid cancer survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.299 (0.126-0.708) 0.006
>70 12.096 (8.227-17.784) <0.0001
Race
White
Ref. (1.00)
Black 2.641 (1.593-4.377) <0.0001
Others 0.587 (0.239-1.445) 0.247
Gender
Male
Ref. (1.00)
Female 0.529 (0.364-0.768) 0.001
Surgery
Yes
Ref. (1.00)
No 1.126 (0.979-1.294) 0.095
Radiation
yes
Ref. (1.00)
No 0.816 (0.609-1.093) 0.173
Year of diagnosis
2000-2008
Ref. (1.00)

Accepted Manuscript
1988-2000
1.071 (0.924-1.241)
0.361
AJCC stage
Stage III
Ref. (1.00)
Stage II 0.942 (0.664-1.336) 0.738
Stage I 0.895 (0.713-1.124) 0.340
Supplementary table-10: Multivariate analysis of factors affecting cardiac-specific mortality among non
Hodgkin lymphoma survivors (1988-2008):
Variable Hazard ratio (95% CI)
P value
Age
40-69 years
Ref. (1.00)
<40 years 0.128 (0.101-0.162) <0.0001
>70 6.513 (6.148-6.899) <0.0001
Race
White
Ref. (1.00)
Black 1.216 (1.084-1.363) 0.001
Others 0.737 (0.651-0.836) <0.0001
Gender
Male
Ref. (1.00)
Female 0.813 (0.771-0.857) <0.0001
Surgery
Yes
Ref. (1.00)
No 0.989 (0.685-1.427) 0.951
Radiation
yes
Ref. (1.00)
No 1.066 (1.000-1.137) 0.049
Year of diagnosis
2000-2008
Ref. (1.00)
1988-2000
1.164 (1.095-1.238)
<0.0001
AJCC stage
Stage III
Ref. (1.00)
Stage II 1.134 (0.881-1.460) 0.329

Accepted Manuscript
Stage I 1.097 (0.882-1.364) 0.504