WCRJ 2020; 7: e1673

THE ROLE OF AGE ON DISTANT

RECURRENCE AFTER BREAST CONSERVATIVE
THERAPY VS. MODIFIED RADICAL
MASTECTOMY AMONG IRANIAN PATIENTS
WITH EARLY STAGE OF BREAST CANCER
F. SHAMSI1, A. REZA BAGHESTANI2, A. AKHAVAN3, P. BASTANI4
1
Department of Biostatistics and Epidemiology, Abortion Research Centre, Yazd Reproductive Science
Institute, School of Public Health, Shahid Sadoughi University of Medical Sciences, Yazd, Iran
2
Physiotherapy Research Centre, Department of Biostatistics, School of Allied Medical Sciences, Shahid
Beheshti University of Medical Sciences, Tehran, Iran
3
Department of Radiolooncology, School of Medicine, Cancer Prevention Center, Seyyed Al-Shohada
Hospital, Isfahan University of Medical Sciences, Isfahan, Iran
4
Department of Surgery, Shahid Sadoughi University of Medical Sciences, Yazd, Iran

Abstract – Objective: Breast conserving therapy is highly recommended for the surgical manage-
ment of patients with early-stage of breast cancer. This study aimed to explore the effect of age on
developing distant metastasis following breast conserving therapy or modified radical mastectomy.
Patients and Methods: To this aim, medical records of 468 women diagnosed with T1-2, N0-1, and
M0 primary invasive breast cancer during 2005-2012 were selected. Patients were treated with either
breast conserving therapy or modified radical mastectomy with post-surgical radiation. Kaplan-Meier
method was used to estimate the breast cancer-free survival and distant metastatic-free survival rates.
Distant metastasis and prognostic factors were modeled by Cox proportional hazards. The Bayesian
approach was used to determine the po-tential threshold of patients’ age at operation and the cure
probabilities were estimated through two treat-ment types according to age change-point.
Results: There was no significant difference in Locoregional recurrence between two treat-
ments type (p=0.83), although a significantly greater recurrence of distant metastasis was observed
in the breast conserving therapy group (p<0.001). Considering metastatic-free survival, the age of
40 was estimated as the change-point of age at operation leading to a higher noticeable cure rate
in the modified radical mastectomy group.
Conclusions: Older patients with early-stage of breast cancer treated with modified radical
mastectomy had a significant decrease of distant metastasis compared with younger patients with
a threshold of 40 for age at operation.

KEYWORDS: Early-stage, Breast cancer, Breast conserving therapy, Modified radical mastectomy,
Age, Distant metastasis.

INTRODUCTION or without postoperative radiation for early-stage

of breast cancer. Although MRM as an invasive
Breast conserving therapy (BCT) including lumpec- treatment includes complete removal of the breast,
tomy following radiation therapy is known as an tumor, overlying skin, and axillary affected lymph
accepted alternative treatment to modified radical nodes, some reported increasing MT or MRM rates
mastectomy (MRM) or mastectomy (MT) with among all age groups with early-stage of breast can-

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

Corresponding Author: Ahmad Reza Baghestani, PhD; e-mail: Baghestani@sbmu.ac.ir 1

cer1-4. The overall, recurrence-free, or disease-free DM recurrences after completing treatment are
survival after conservative therapy and MRM or considered important. In a systematic meta-anal-
MT for the early-stage of breast cancer patients ysis including five population-based studies and
have been investigated in different clinical and one clinical trial comparing OS based on treatment
population-based studies5-11. Some studies indicat- types (BCT or MT) in young patients (≤40 years)
ed at-least equivalent overall survival (OS), distant with early-stage of breast cancer, the equivalent
metastasis-free survival (DMFS), breast cancer-spe- long-term results were reported26. As mentioned,
cific survival (BCSS) and local/regional recurrence on Korean patients aged ≤ 30 with early-stage of
(LR/RR) for the patients treated with BCT, MRM or cancer, no significant difference was observed in 5-
MT without considering postoperative radiation in and 10-year overall survival between BCT and MT
T1-2N-N+M07,9,11-13. Lumpectomy plus postsurgical groups. However, there was a noticeable lower LR
radiation is proposed as the first course of treatment rate in patients treated with MT in long-term fol-
in some observational investigations10,14-16. These low-up23. Considering various age groups in differ-
studies revealed noticeable higher OS, DMFS in the ent researches may lead to ambiguous inferences
BCT group compared to MT. However, some breast about the importance of age on cancer recurrences.
cancer patients prefer receiving MRM than BCT in Frequently, a proportion of treated cancer patients
Asian countries17. Socioeconomic status is the pre- may not experience the event of interest such as
dominant factor in choosing the more invasive sur- LR/RR, DM up to the end of long-term follow-up
gery. Selecting an appropriate treatment is strongly are regarded as cured or long-term survivors. Gen-
related to typical prognostic factors such as tumor erally, the stable plateau at the tail of Kaplan-Mei-
size, affected lymph nodes, hormone receptors sta- er curves is considered as empirical evidence for
tus (immunohistochemical staining status includ- the presence of a cured fraction. Cure models that
ing estrogen and progesterone receptor (ER/PR), handle the heterogeneity in treated patients are pre-
and epidermal growth factor receptor-2 (Her2))18. ferred in survival analysis for evaluating the actual
Besides, locoregional recurrence of breast cancer, prognostic effect of predictors27-30. This study aims
the occurrence of distant metastasis (DM), recon- to investigate the influence of age as an important
struction techniques, and potential adverse effects factor which may cause variation in oncological
of radiation therapy after surgery can influence the outcome in different surgical treatments of T1-
process of decision making19,20. 2N0-1M0 breast cancer patients. To this aim, the
In recent researches, unlike the results of a popu- mixture cure model was considered for modeling
lation-based study in American patients, age is con- survival data and the Bayesian approach was ap-
sidered as an influential factor for experiencing LR/ plied to detect the threshold of age at operation in
RR or DM which may change the overall survival the early stage of breast cancer patients31.
of breast cancer patients21-25. It was indicated in two
Asian studies that breast cancer-specific survival
and DMFS decreased in younger patients23,24. It was PATIENTS AND METHODS
shown in one population-based study in Dutch pa-
tients diagnosed with I-II stages of breast cancer that Participants
the LR rate decreased in the BCT group for women
younger than 40 years old21. Furthermore, the results All data were extracted from a historical-cohort
of another study in Netherland revealed that overall study database conducted at a university hospi-
survival decreased significantly with increasing age. tal in Iran during 2005-2012 32. Participants were
However, no age-specific differences were observed female admitted to Shahid Sadoughi Hospital in
in experiencing LR/RR and DM in breast cancer Yazd who had a primary ductal or lobular carci-
patients younger or older than 75 years old. Accord- noma without any previous occurrence of can-
ingly, it was suggested to conduct age-specific breast cer and were a candidate for breast-conserving
cancer studies to find the actual impact of age in a surgery with radiotherapy (BCT) or MRM. Pa-
heterogeneous group of patients22. tients were admitted to Shahid Ramadanzadeh
Breast cancer management for young patients Radiotherapy Center for post-surgical radiation.
with early-stage of cancer changed considerably The patients with no histological information,
into a multidisciplinary approach including less previous diagnosis of non-invasive breast cancer
radical surgery following radiation therapy, che- within five last years, any symptom of concurrent
motherapy, or hormonal therapy. Considering the malignancy or metastatic tumor, and those under-
situation of patients involved in different types of going chemotherapy or radiotherapy prior to sur-
surgery, it has been reported that younger patients gery were excluded from the study. The study was
are more susceptible to aggressive stages of breast limited to those patients with early stage of breast
cancer. Consequently, controlling the LR/RR and cancer (T1-2N0-1M0), who was divided as stage I

2
 AGE THRESHOLD EFFECT ON DISTANT METASTASIS.

and II according to the tumor size, lymph nodes with the Kaplan-Meier method, and the effect of
and metastases classification according to 7th edi- each variable on DM-free survival and OS curves
tion of the American Joint Committee on Cancer were determined using the log-rank test. Hazard
(AJCC) system. The study included 468 patients ratios (HRs) with 95% confidence intervals (CIs)
who underwent breast conserving surgery follow- were estimated for both DM-free survival and OS
ing radiation therapy or MRM with post-surgical for all factors using the unadjusted and adjusted
radiation. Those patients who were treated with Cox proportional hazard model. Also, the poten-
MRM without radiation therapy were excluded as tial threshold of the effective quantitative covari-
the patients commonly receive postsurgical radia- ates on DM-free survival was determined using
tion after MRM surgery. Demographic and clini- the Bayesian approach by considering the cure
copathologic factors including age of patients at proportion of patients using R software (https://
operation, grade, tumor size, lymph nodes status, cran.r-project.org/)31. p-value <0.05 was consid-
ER, PR and Her2 status, histology, presence of ered as statistically significant.
lymphovascular invasion, adjuvant treatment such
as chemotherapy or hormone therapy, and type
of surgery were measured for the primary tumor. RESULTS
Continuing tamoxifen as an adjuvant treatment to
10 years was recommended for patients33. Besides, In general, 468 patients ranged 22-81 years with a
the molecular subtypes of breast cancer were also median age of 47 years (Interquartile range (40-54
defined as Luminal A (ER+/PR+/HER2-/low Ki- years)) were included in this study, among which
67); Luminal B (ER+/PR+/HER2-/ high Ki-67); 261 patients (55.8%) underwent BCT and 207 pa-
HER2-overexpression (ER-/PR-/HER2+) and tri- tients (44.2%) underwent MRM with post-surgi-
ple-negative (ER-/PR-/HER2-) according to the cal radiation. The median follow-up time for DM
classification of immunohistochemistry (IHC) and death was 4.7 and 5 years, respectively. The
markers34, 35. The basal-like subtype was not iden- tumor baseline characteristics of patients accord-
tified because the basal marker (CK5/6) was not ing to received treatments are shown in Table 1.
recorded for patients36,37. The information was ex- As shown, among BCT group 58 (22%) distant
tracted from pathology reports and medical files metastasis and 13 (5%) LR/RR occurred. Results
during the follow-up. All patients provided written indicated patients in the BCT group had higher
informed consent. The approval of this study was mean age at operation, larger tumor size, more ax-
obtained from the Medical Ethical Committee of illary affected lymph nodes status, and more lym-
the Shahid Beheshti University of Medical Scienc- phovascular invasion (LVI) compared to those
es (IR.SBMU.RETECH.REC.1397.822), and the who underwent MRM with post-surgical radi-
Ethics Committee of the Shahid Sadoughi Univer- ation. However, the patients in the MRM group
sity of Medical Sciences gave confirmation. were more likely to have a higher value of the
ki-67 index. There was no significant difference
between the patients in both treatment groups re-
Follow-up and endpoints garding tumor type, tumor grading, ER-/PR- sta-
tus, Her2/neu-positive carcinoma status, adjuvant
Postoperative follow-up was performed until chemotherapy, and molecular subtype classifica-
March 20, 2017. Local recurrence, DM-free sur- tion.
vival situations, and survival time were followed After completing follow-up time, 87 patients
simultaneously. The primary and secondary end- died, among whom 18 (9%) and 69 (26%) died
points of this study were distant metastasis and from MRM and BCT group, respectively, while
death during the follow-up, respectively. Time-to- 381 patients were alive. Kaplan-Meier estimates
event was identified as the interval between the indicated a significant improvement in OS and
date of surgery and DM occurrence or death from DMFS curves for MRM over the BCT group, and
breast cancer, separately. In the absence of events Log-rank tests revealed a statistically significant
of interest at the last follow-up, the subjects were difference between the survival curves for both
censored. results, respectively (p-value<0.001) (Figures 1
(A) and (B)).
Using MRM with postsurgical radiation group
Statistical analysis as reference, univariable Cox model analysis in-
dicated that patients assigned to the BCT group
First, the χ2 test was used to compare the demo- (HR=3.09, 95% CI: 1.75-5.46) had a higher risk
graphic and clinicopathologic factors between of experiencing DM. As indicated in Table 2, in-
the two groups. Survival rates were estimated volving axillary lymph nodes increased the risk

3
TABLE 1. Clinical Characteristics of early stage of breast cancer Patients by surgery type.

Characteristics
BCT MRM with post-surgical p-value
n=261 (56%) radiation n=207 (44%)

(%) N (%)

Vital status <0.001

   Dead 69 (26) 18 (9)
   Alive 192 (74) 189 (91)
Distant metastasis <0.001
   Yes 58 (22) 15 (7)
   No 203 (78) 192 (93)
Locoregional recurrence 0.83
   Yes 13 (5) 9 (4)
   No 248 (95) 198 (96)
Age at surgery time (years), mean (SD) 48.92±10.96 45.83±10.51 0.002
Tumor size <0.001
   T1 <2 36 (14.9) 65 (32.2)
   T2 2-5 190 (78.5) 126 (62.4)
   T3 >5 16 (6.6) 11 (5.4)
Tumor type 0.35
   Ducal 197 (75.5) 155 (74.9)
   Lobular 24 (9.2) 13 (6.3)
   Other 40 (18.8) 39 (15.3)
Affected Lymph nodes <0.001
   N0 51 (19.5) 111 (53.6)
   N1 1-3 210 (80.5) 96 (46.4)
Tumor Grade 0.09
  I 27 (10.3) 14 (6.8)
   II 104 (39.8) 93 (44.9)
   III/IV 53 (20.3) 54 (26.1)
  Unknown 77 (29.5) 46 (22.2)
Lymphovascular invasion <0.001
   Yes 59 (22.6) 27 (13)
   No 112 (49.2) 132 (63.8)
   Unknown 90 (34.5) 48 (23.2)
Estrogen/progesterone receptors (ERPR) 0.16
   ER+PR+ 103 (39.5) 98 (47.3)
   ER-PR- 62 (23.8) 37 (17.9)
   unknown 96 (36.8) 72 (34.8)
Her2 status 0.22
   Negative 54 (20.7) 52 (25.1)
   Positive 122 (46.7) 102 (49.3)
   unknown 85 (32.6) 53 (25.6)
Ki67 0.002
   ≤15 41 (15.7) 56 (27.1)
   >15 41 (15.7) 40 (19.3)
   unknown 179 (68.6) 111 (53.6)
Adjuvant chemotherapy 0.42
   Yes 238 (91.2) 192 (93)
   No 23 (8.8) 14 (6.6)
   Unknown 0 1 (0.4)
Molecular subtype classification 0.8
   Luminal A 48 (34.5) 41 (30.6)
   Luminal B 83 (59.7) 84 (62.7)
   Triple negative 5 (3.6) 7 (5.2)
   HER2 positive 3 (2.2) 2 (1.5)

of DM (p-value=0.001). Patients were grouped in DM compared to those aged lower than 40 years
two according to age at the operation to determine old (Table 2). Other prognostic factors had a sim-
its effect more clearly. In this regard, patients who ilar effect on experiencing DM in both treatment
were ≥ 40 years old were more likely to develop groups.

4
 AGE THRESHOLD EFFECT ON DISTANT METASTASIS.

Fig. 1. A, Overall survival and (B) Distant relapse-free survival according to types of treatment in breast cancer patients.

TABLE 2. Univariable analysis of DM-free survival for early-stage of breast cancer patients.

Characteristics Univariable analysis

Time to distant metastasis

HR (95% CI) p-value

Treatment
   MRM with postsurgical radiation Ref
   BCT 3.09 (1.75-5.46) <0.001
Age at surgery time (years)
   ≤40 Ref
   >40 0.64 (0.4-1.05) 0.075
Affected lymph nodes
   N0 Ref
   N1 2.91 (1.53-5.54) 0.001
Tumor size
  T1 Ref
   T2 1.56 (0.81-2.98) 0.18
   T3 0.67 (0.15-3.03) 0.6
Grade
  I Ref
   II 1.78 (0.63-5.03) 0.27
   III/IV 2.05 (0.69-6.03) 0.19
Tumor Type
  Ductal Ref
   Lobular 1.03 (0.47-2.26) 0.93
   Other 0.43 (0.18-0.98) 0.046
LVI
  Yes Ref
   No 0.83 (0.44-1.57) 0.57
Estrogen/progesterone receptors (ERPR)
  ER+PR+ Ref
   ER-PR- 1.61 (0.9-2.88) 0.1
Her2 status
  Positive Ref
   Negative 1.48 (0.81-2.7) 0.2
Ki67
   ≤15 Ref 0.76
   >15 1.52 (0.46-2.86)
Adjuvant chemotherapy
  Yes Ref
   No 1.4 (0.67-2.93) 0.36
Molecular subtype classification
  Luminal A Ref
   Luminal B 0.97 (0.36-1.37) 0.3
   HER2 positive 2.59 (0.59-11.35) 0.2
   Triple negative 3.41 (1.23-9.45) 0.02

5
TABLE 3. Multivariable Analysis of DM-free survival for early-stage of breast cancer patients.

Characteristics Time to metastasis p-value

HR* (95% CI)
Model 1
Treatment
   MRM with postsurgical radiation Ref
   BCT 2.67 (1.44-4.96) 0.002
Age at surgery time (years)
   ≤40 Ref
   >40 0.6 (0.36-1.01) 0.046
Affected lymph nodes
   N0 Ref
   N1 2.17 (1.07-4.39) 0.03

Regarding the multivariable Cox regression The present study aimed to fit the correct mod-
analysis, the backward method was recognized as el to estimate the actual effect of significant co-
the most appropriate approach to assess the effect variates on time to DM occurrence utilizing an
of each adjusted variable and presenting the most appropriate approach for detecting the thresholds
fitted model. Based on the backward method, the of effective continuous variables. As shown in
established prognostic factors such as tumor size, Figure 1 (B), the patients in the MRM group after
nodal status, tumor grade, ER/PR status, Her2 sta- 50 months cured for developing DM at the end
tus, molecular subtype classification, and adjuvant of follow-up time and they may not die after 60
chemotherapy were considered in the model and months. Figures 2 (A) and (B), Figures 3 (A) and
the concluding model was achieved after removing (B) highlighted the difference between the cure
the non-significant variables (Table 3). Compared rates in both treatment groups by considering the
to the patients who received MRM followed by ra- patient’s age categories and nodal status, respec-
diation, those treated with BCT had a higher risk tively. A significant difference between cure rates
of occurring distant metastasis (HR=2.67, 95% CI demonstrated that the classification of these two
1.44-4.96) after adjusting to age and lymph node variables could help choose the appropriate treat-
status. Furthermore, the age at operation and nodal ment for patients with stage I and II of breast can-
status were considered as influential parameters in cer. Previous studies suggested different methods
experiencing DM (Table 3). to estimate a change-point of a quantitative vari-
The appropriateness of conducting a cure rate able38,39. The Bayesian approach was selected for
model was indicated in Figures 1 (A) and (B), detecting the threshold of patients’ age at operation
respectively, considering a stable region after through different latent distributions31. As shown
50 and 100 months for MRM and BCT groups. in Table 4, different Bayesian models were speci-

Fig. 2. A, Distant metastasis-free survival according to types of treatment in early-stage breast cancer patients aged ≤40 and
(B) in early-stage breast cancer patients aged >40.

6
 AGE THRESHOLD EFFECT ON DISTANT METASTASIS.

Fig. 3. A, Distant metastasis-free survival according to types of treatment in early-stage breast cancer patients with negative
nodule status and (B) with positive nodule status.

fied and the threshold of patients’ age at operation effect of treatment types, which was observed in
was estimated. Considering DIC values, model 3 Kaplan-Meier estimations (Figures 3 (A) and (B)).
had the best performance and the threshold of age
was estimated 40 years old as expected. Based on
model 3 in Table 4, the odds of cured patients in DISCUSSION
MRM with postsurgical radiation groups were 11
times higher than the BCT group for women at or Surgical management for the early stage of breast
younger than 40 years old compared to older pa- cancer has been gradually developed from radical
tients. Regarding the significant difference in cure surgery to sophisticated surgical procedures. this
rates between two types of surgery, the Bayesian sequence of instructions is determined for onco-
approach provided some shreds of evidence that logical safety and improvement of life quality. The
the age of patients at operation could modify the outcome of breast cancer patients is affected by

TABLE 4. Multivariable Analysis of DM-free survival for early-stage of breast cancer patients.

Characteristics b SE Exp (b) 95% HPD Interval

Model 1: Mixture cure model with threshold with exponential latency
Age at surgery time (years)≤ 35
MRM with postsurgical radiation BCT 0.4 0.32 1.49 (-0.18, 0.96)
ref
Age at surgery time (years)> 35
MRM with postsurgical radiation BCT 1.06 0.35 2.88 (0.37, 1.62)
ref
Model 2: Mixture cure model with threshold with log-logistic latency
Age at surgery time (years)≤ 34
MRM with postsurgical radiation BCT 0.87 0.41 2.38 (0.05, 1.53)
ref
Age at surgery time (years)> 34
MRM with postsurgical radiation BCT 1.41 0.29 4.09 (0.91, 1.94)
ref
Model 3: Mixture cure model with threshold with Weibull latency
Age at surgery time (years)≤ 40
MRM with postsurgical radiation BCT 2.4 0.62 11.02 (1.14, 3.53)
ref
Age at surgery time (years)> 40
MRM with postsurgical radiation BCT 1.53 0.83 4.62 (0.32, 3.29)
ref

SE, standard error of mean; HPD, highest posterior density.

7
the complex interactions between tumor biology, ed with MRM with postsurgical radiation, espe-
patients’ features, and received treatment. In the cially in patients aged lower than 40 years old,
early stage of breast cancer, the tumor features are confirms the results of previous meta-analysis
almost similar, and host conditions can influence study showing better OS for patients treated with
the treatment choice and outcome40. Although MT compared to breast conserving surgery with
some investigations indicated that younger and whole breast radiotherapy group in early breast
older patients may benefit from BCT or MRM cancer patients aged 40 years or younger26. Fur-
equally, age should be considered as a time-de- thermore, the detected threshold was exactly sim-
pendent covariate and effective factor in process ilar to what was reported from the Asian breast
of treatment decision-making that can enhance cancer study and the same rates were calculated
the surgeon’s knowledge to predict the treatment for OS24. However, other studies had shown no
outcome13,15. There are some reasons persuaded association between age and overall survival in
researchers to focus on long-term outcome of each different surgery types13,23. An explanation for
treatment in various age groups among breast this discrepancy could be attributed to more regu-
cancer patients such as high incidence of breast larly receiving postsurgical radiation in the MRM
cancer in young women, increasing rate of MT or group that caused a significant difference in cure
MRM rather than BCT in some nations2,4,41, fear probabilities between two treatments in younger
of cancer recurrence after treatment, postsurgical patients. Some studies reported an increased risk
treatment such as radiotherapy or chemotherapy, of local recurrence in the BCT group for young-
established oncology safety of BCT and improved er patients, while no difference was observed in
quality of life in younger patients with early-stage OS rate15,23,24. Based on the results of the present
of breast cancer23,42, reported at-least similar OS study, no significant difference was observed in
and DM-free survival after BCT or MT among locoregional recurrence between two surgery
young and old aged of patients with early-stage types in both age groups of patients, despite some
of breast cancer6,7,16,25. Besides, the shape of the differences in DM recurrence. In the BCT group,
relationship between patients’ age and cancer re- the risk of experiencing DM was 2 times higher
lapse or death, especially at 40-45 years of age, for patients younger than 40 years old.
may represent different cancer progression trends The use of Bayesian approach to detect the
according to age24, 43. threshold of a challenging covariate in DM-free
The present study focused on the change-point survival is considered as the main benefit in the
estimation of the patients’ age at operation con- present study. Moreover, examining different dis-
cerning treatment for early-stage of breast cancer. tributions for identifying the correct threshold of
Most of the previous studies evaluated the effect patient’s age at operation can be regarded as its
of age in the dichotomous method by using vari- flexibility. Based on the results of previous re-
ous definitions of youth13,22-24. In this observation- search, postsurgical radiation is an effective factor
al study, the effect of BCT and MRM on a cured for better OS in BCT group44. All patients in the
fraction of patients, considering DM occurrence, present study received postsurgical radiation after
were compared based on the estimated threshold each surgery for completing their treatment pro-
of patients’ age. Results indicated treatment types cess that helped to obtain a realistic view of treat-
and nodule status were significantly important, ment effects afforded by the two surgical types.
and the age at operation had a borderline effect on The selection bias and recording information er-
developing DM after adjusting to other clinico- rors before designing the study were considered as
pathological and histopathological features (Table some of the limitations in the present study, which
2). In line with the results of the previous studies, caused removing some cases. Furthermore, small
the cured proportion of treated early breast cancer DM occurrences in patients under the age of 40
patients increased for patients older than 40 years could be regarded as a reason for increasing comor-
old, as shown in Figures 2 (A) and (B). The Bayes- bidities. Long-term follow-up, at least 10 years or
ian method using a mixture cure rate model with higher can eliminate the limitation demonstrating
Weibull distribution as the latency part, determined the treatment effect more noticeable.
the threshold of patients’ age at operation at about
40 years old (Model 3 in Table 3)31. To summarize
the results, a significant difference was observed CONCLUSIONS
between the probabilities of not experiencing DM
in each surgery type for patients younger than 40 In conclusion, this study aimed to detect the cor-
years old compared to older ones (Table 3). rect threshold of patients’ age at operation for
The main finding of this study that OS and demonstrating the influence of each surgery type
DM-free survival was longer for all patients treat- after a 5-year follow-up. Based on the results,

8
 AGE THRESHOLD EFFECT ON DISTANT METASTASIS.

younger patients with early-stage of breast can-    9. Chen Z, Xu Y, Shu J, Xu N. Breast-conserving surgery
cer treated with BCT were inclined to a higher versus modified radical mastectomy in treatment of
early stage breast cancer: a retrospective study of 107
risk of experiencing DM compared to the MRM cases. J Cancer Res Ther 2015; 11: 29.
group. A definite threshold of age at operation 10. van Maaren MC, de Munck L, Jobsen JJ, Poortmans P,
was first determined to calculate the cure proba- de Bock GH, Siesling S, Strobbe LJ. Breast-conserving
bilities through two types of surgery. Patients’ age therapy versus mastectomy in T1-2N2 stage breast
can be regarded as an effective factor in the deci- cancer: a population-based study on 10-year overall,
relative, and distant metastasis-free survival in 3071
sion-making process for early stage patients that patients. Breast Cancer Res Treat 2016; 160: 511-521.
may increase the number of long-term survivors 11. van Maaren MC, de Munck L, de Bock GH, Jobsen JJ,
while an appropriate treatment should be selected van Dalen T, Linn SC, Poortmans P, Strobbe LJ, Siesling
with comprehensive consideration. S. 10 year survival after breast-conserving surgery
plus radiotherapy compared with mastectomy in early
breast cancer in the Netherlands: a population-based
study. Lancet Oncol 2016; 17: 1158-1170.
Acknowledgments: 12. Wang L, Ouyang T, Wang T, Xie Y, Fan Z, Lin B, Li J.
The authors would like to thank the Radiothera- Breast-conserving therapy and modified radical mastec-
py Center of Shahid Ramadanzadeh in Yazd for tomy for primary breast carcinoma: a matched compa-
providing the data. The authors would also like to rative study. Chinese J Cancer Res 2015; 27: 545.
13. Yu TJ, Liu YY, Hu X, Di GH. Survival following breast-
thank Prof. Binesh for her support and M. Heidari conserving therapy is equal to that following mastec-
for data collection. tomy in young women with early-stage invasive lobular
carcinoma. Eur J Surg Oncol 2018; 44: 1703-1707.
Conflict of interest: 14. Fisher S, Gao H, Yasui Y, Dabbs K, Winget M. Survival in
There are no conflicts of interest. stage I–III breast cancer patients by surgical treatment
in a publicly funded health care system. Ann Oncol
2015; 26: 1161-1169.
15. Christiansen P, Carstensen SL, Ejlertsen B, Kroman N,
Offersen B, Bodilsen A, Jensen M-B. Breast conserving
REFERENCES surgery versus mastectomy: overall and relative survi-
val—a population based study by the Danish Breast
  1. Bland KI, Chang HR, Copeland III EM. Modified radi- Cancer Cooperative Group (DBCG). Acta Oncol 2018;
cal mastectomy and simple mastectomy. The Breast : 57: 19-25.
Comprehensiv Management of Benign and Malignant 16. Lagendijk M, van Maaren MC, Saadatmand S, Strobbe
diseases 5th ed. Philadelphia: Elsevier; 2018; pp. 443- LJ, Poortmans PM, Koppert LB, Tilanus-Linthorst MM,
461. Siesling S. Breast conserving therapy and mastectomy
   2. Mahmood U, Hanlon AL, Koshy M, Buras R, Chumsri revisited: Breast cancer-specific survival and the in-
S, Tkaczuk KH, Cheston SB, Regine WF, Feigenberg SJ. fluence of prognostic factors in 129,692 patients. Int J
Increasing national mastectomy rates for the treatment Cancer 2018; 142: 165-175.
of early stage breast cancer. Ann Surg Oncol 2013; 20: 17. Liu JJ, Zhang S, Hao X, Xie J, Zhao J, Wang J, Liu L, Wang
1436-1443. PP, Zhang J. Breast-conserving therapy versus modified
  3. Kummerow KL, Du L, Penson DF, Shyr Y, Hooks MA. radical mastectomy: Socioeconomic status determines
Nationwide trends in mastectomy for early-stage breast who receives what—Results from case–control study in
cancer. JAMA Surg 2015; 150: 9-16. Tianjin, China. Cancer Epidemiol 2012; 36: 89-93.
   4. Dragun AE, Huang B, Tucker TC, Spanos WJ. Increasing 18. Billiar T, Andersen D, Hunter J, Brunicardi F, Dunn D,
mastectomy rates among all age groups for early stage Pollock RE, Matthews J. Schwartz’s principles of surge-
breast cancer: a 10-year study of surgical choice. Breast ry: McGraw-Hill Professional; 2009.
J 2012; 18: 318-325. 19. Lee W, Tan V, Choo H, Ong J, Krishnapriya R, Khong
   5. Correa C, McGale P, Taylor C, Wang Y, Clarke M, Davies S, Tan M, Sim Y, Tan B, Madhukumar P. Factors
C, Peto R, Bijker N, Solin L, Darby S. Overview of the influencing patient decision-making between simple
randomized trials of radiotherapy in ductal carcinoma mastectomy and surgical alternatives. BJS Open 2018;
in situ of the breast. J Natl Cancer Inst Monogr 2010; 3: 31-37.
2010: 162-177. 20. Gu J, Groot G, Holtslander L, Engler-Stringer R. Under-
   6. Litière S, Werutsky G, Fentiman IS, Rutgers E, Christia- standing Women’s Choice of Mastectomy Versus Breast
ens M-R, Van Limbergen E, Baaijens MH, Bogaerts J, Conserving Therapy in Early-Stage Breast Cancer. Clin
Bartelink H. Breast conserving therapy versus mastec- Med Insights Oncol 2017; 11: 1179554917691266.
tomy for stage I–II breast cancer: 20 year follow-up of 21. Van Laar C, Van der Sangen M, Poortmans P, Nieuwen-
the EORTC 10801 phase 3 randomised trial. Lancet huijzen G, Roukema J, Roumen R, Tjan-Heijnen V,
Oncol 2012; 13: 412-419. Voogd A. Local recurrence following breast-conserving
  7. Simone NL, Dan T, Shih J, Smith SL, Sciuto L, Lita E, treatment in women aged 40 years or younger: trends
Lippman ME, Glatstein E, Swain SM, Danforth DN. in risk and the impact on prognosis in a population-
Twenty-five year results of the national cancer institute based cohort of 1143 patients. Eur J Cancer 2013; 49:
randomized breast conservation trial. Breast Cancer Res 3093-3101.
Treat 2012; 132: 197-203. 22. Kiderlen M, van de Water W, Bastiaannet E, de Craen
   8. Agarwal S, Pappas L, Neumayer L, Kokeny K, Agarwal AJ, Westendorp RG, van de Velde CJ, Liefers G-J. Sur-
J. Effect of breast conservation therapy vs mastectomy vival and relapse free period of 2926 unselected older
on disease-specific survival for early-stage breast can- breast cancer patients: a FOCUS cohort study. Cancer
cer. JAMA Surg 2014; 149: 267-274. Epidemiol 2015; 39: 42-47.

9
23. Kang I, Kim JH, Park S, Hur H, Park HS, Kim SI, Cho 34. Al-Thoubaity FK. Molecular classification of breast can-
YU. Oncologic safety of breast-conserving surgery in cer: a retrospective cohort study. Ann Med Surg 2020;
breast cancer patients under the age of 35. Korean J 49: 44-48.
Clin Oncol 2017; 13: 32-38. 35. Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thürli-
24. Wong FY, Tham WY, Nei WL, Lim C, Miao H. Age exerts mann B, Senn HJ, Members P. Strategies for subtypes—
a continuous effect in the outcomes of Asian breast dealing with the diversity of breast cancer: highlights
cancer patients treated with breast-conserving therapy. of the St Gallen International Expert Consensus on
Cancer Commun 2018; 38: 39. the Primary Therapy of Early Breast Cancer 2011. Ann
25. Hwang ES, Lichtensztajn DY, Gomez SL, Fowble B, Oncol 2011; 22: 1736-1747.
Clarke CA. Survival after lumpectomy and mastectomy 36. Elesawy BH, Shawky AE, Arafa M. Immunohistochemi-
for early stage invasive breast cancer: the effect of stry-based subtyping of breast carcinoma in Egyptian
age and hormone receptor status. Cancer 2013; 119: women: a clinicopathologic study on 125 patients. Ann
1402-1411. Diagn Pathol 2014; 18: 21-26.
26. Vila J, Gandini S, Gentilini O. Overall survival according 37. Zhang M, Liu HX, Teng XD, Wang HB, Cui J, Jia SS, Gu
to type of surgery in young (≤ 40 years) early breast XY, Li ZG. The differences in CXCR4 protein expression
cancer patients: a systematic meta-analysis comparing are significant for the five molecular subtypes of breast
breast-conserving surgery versus mastectomy. Breast cancer. Ultrastruct Pathol 2012; 36: 381-386.
2015; 24: 175-181. 38. Othus M, Li Y, Tiwari R. Change point-cure models
27. Baghestani AR, Zayeri F, Akbari ME, Shojaee L, Kha- with application to estimating the change-point effect
dembashi N, Shahmirzalou P. Fitting cure rate model of age of diagnosis among prostate cancer patients. J
to breast cancer data of cancer research center. Asian Appl Stat 2012; 39: 901-911.
Pac J Cancer Prev 2015; 16: 7923-7927. 39. Zhao L, Feng D, Bellile EL, Taylor JM. Bayesian random
28. Baghestani AR, Moghaddam SS, Majd HA, Akbari ME, threshold estimation in a Cox proportional hazards cure
Nafissi N, Gohari K. Application of a non-mixture cure model. Stat Med 2014; 33: 650-661.
rate model for analyzing survival of patients with breast 40. Thewes B, Butow P, Bell ML, Beith J, Stuart-Harris R,
cancer. Asian Pac J Cancer Prev 2015; 16: 7359-7363. Grossi M, Capp A, Dalley D, Committee FSA. Fear of
29. Baghestani AR, Shahmirzalou P, Sayad S, Akbari ME, cancer recurrence in young women with a history of
Zayeri F. Comparison cure rate models by DIC criteria early-stage breast cancer: a cross-sectional study of
in Breast Cancer data. Asian Pac J Cancer Prev 2018; prevalence and association with health behaviours.
19: 1601-1606. Support Care Cancer 2012; 20: 2651-2659.
30. Hoseini M, Bahrampour A, Mirzaee M. Comparison 41. Gomez SL, Lichtensztajn D, Kurian AW, Telli ML, Chang ET,
of weibull and lognormal cure models with cox in the Keegan TH, Glaser SL, Clarke CA. Increasing mastectomy
survival analysis of breast cancer patients in Rafsanjan. rates for early-stage breast cancer? Population-based tren-
J Res Health Sci 2017; 17: 1-6. ds from California. Am J Clin Oncol 2010; 28: e155.
31. Shamsi F, Baghestani AR, Binesh F. Bayesian approach 42. van der Leest M, Evers L, van der Sangen MJ, Poort-
for cure models with a change-point based on cova- mans PM, van de Poll-Franse LV, Vulto AJ, Nieuwen-
riate threshold: application to breast cancer data. J huijzen GA, Brenninkmeijer SJ, Creemers GJ, Voogd
Biopharm Stat 2020; 30: 219-230. AC. The safety of breast-conserving therapy in patients
32. Binesh FA, Heidari M. comparison of local recurrence, with breast cancer aged≤ 40 years. Cancer 2007; 109:
distant metastases, survival levels of MRM and BCT in 1957-1964.
women with early-stage breast cancer [Unpublished 43. Natarajan L, Pu M, Parker BA, Thomson CA, Caan
manuscript]. Yazd: Shahid Sadoughi University of Me- BJ, Flatt SW, Madlensky L, Hajek RA, Al-Delaimy WK,
dical Sciences 2019. Saquib N. Time-varying effects of prognostic factors
33. Davies C, Pan H, Godwin J, Gray R, Arriagada R, Raina associated with disease-free survival in breast cancer.
V, Abraham M, Alencar VHM, Badran A, Bonfill X. Am J Epidemiol 2009; 169: 1463-1470.
Long-term effects of continuing adjuvant tamoxifen 44. Onitilo AA, Engel JM, Stankowski RV, Doi SA. Survival
to 10 years versus stopping at 5 years after diagnosis comparisons for breast conserving surgery and mastec-
of oestrogen receptor-positive breast cancer: ATLAS, a tomy revisited: community experience and the role of
randomised trial. Lancet 2013; 381: 805-816. radiation therapy. J Clin Med Res 2015; 13: 65-73.

10