Effect of Previous Cancer History on Survival of Patients with Different Subtypes of Breast Cancer.

Patients with a previous cancer history (PCA) are routinely excluded from most clinical trials, which may limit the accuracy and universality of clinical trials. We aimed to explore the association between PCA and survival of patients with different molecular subtypes of breast cancer. Patients diagnosed with breast cancer from the Surveillance, Epidemiology, and End Results (SEER) database between 2010 and 2015 were included in this retrospective cohort study. The primary outcome was overall survival (OS), which was calculated from date of diagnosis to date of death or censor date during this period. The relationship between PCA and OS of patients with different molecular subtypes of breast cancer was analyzed by the Kaplan-Meier curves and multivariate Cox proportional-hazards model. A total of 35,640 primary breast cancer patients were included, and 2,038 (5.72%) patients had a PCA. Female genital system cancer (491 cases, 24.09%) was the largest proportion type of previous cancer, and HER2-positive (24,754 cases, 69.46%) breast cancer was the most common subtype. Patients with previous female genital/endocrine system cancer history and other cancers history were associated with a poorer OS in overall patients, and in patients with triple-negative and HER2-positive subtypes (P < 0.05). In patients with Luminal A and Luminal B subtypes, previous other cancers history was related to poor OS (P < 0.05), while female genital/endocrine system cancer history may not influence the OS (P > 0.05). Subgroup analyses presented that PCA was related to poor OS in patients aged 40-64 years and ≥65 years (P <0.05), while prognosis in patients aged 18-40 years may not be influenced by PCA (P > 0.05). The impact of PCA on the prognosis of breast cancer patients was related to molecular type, patient age, and type of PCA. In clinical trials of breast cancer, the exclusion criteria for PCA patients may be modified according to the above variables.

1. Introduction

Breast cancer (BC) has become the leading female malignancy worldwide, and about two-thirds of patients with early non-metastatic BC can be cured [1]. In 2020, 2.26 million women became new BC patients and approximately 684,996 patients died of BC [2]. The number of cancer survivors has been reported on the rise [3], which may lead to a rise in the number of patients with multiple primary cancers [4, 5]. Previous studies have indicated that patients with a previous cancer history (PCA) account for approximately 4% to 14% of all BC patients [6-8].

Patients with PCA are usually listed as exclusion criteria for study populations in the cancer clinical trials [9], which occur in approximately 77% of BC studies and 80% of lung cancer studies [10, 11]. Only 3%-5% of these patients were enrolled in trials each year [12, 13]. The exclusion of these patients with PCA may limit the accuracy and universality of clinical trials [14, 15]. The Clinical Trial Eligibility Working Group recommended that patients should not be excluded based solely on previous cancer in clinical trials [13, 16]. Several studies found that the impact of PCA on the clinical outcomes of cancer patients may be related to the type of tumor [6, 10, 17, 18]. However, few studies have reported the impact of PCA on the survival of BC patients. Furthermore, previous studies have demonstrated that the molecular type of BC is an important factor influencing the survival of BC patients [19-21]. Only a recent study showed that PCA was a risk factor for survival in patients with advanced BC [6], but they did not conduct further analyses by molecular type of BC. Exploring the impact of PCA on the prognosis of patients with different molecular types of BC may help improve the accuracy of BC clinical trials.

Therefore, the purpose of this study was to analyze the relationship between PCA and survival of patients with different molecular subtypes of BC. Furthermore, the impact of different types of PCA on the survival of patients was also analyzed.

2. Methods

2.1. Data Source and Populations

The analysis data of this retrospective cohort study were obtained from the Surveillance, Epidemiology, and End Results (SEER) database (2010 to 2015). The SEER database was established by the National Cancer Institute of the United States (US) to achieve cancer prevention, diagnosis, and treatment by collecting, analyzing, and disseminating cancer-related data. The database covers about 28% of the US population [22], and data include demographic characteristics (e.g., age, sex, and race) and tumor characteristics (e.g., year of diagnosis, primary tumor site, histology, behavior, and stage) were collected. Since the data on the human epidermal growth factor receptor 2 (HER2) molecular type of BC in the SEER database only included after 2010, the data from 2010 to 2015 were utilized for analysis. The International Classification of Diseases for Oncology, 3rd edition (ICD-O-3) codes (C50.0-C50.6, C50.8, and C50.9) were utilized to identify BC patients. Exclusion criteria: (1) age <18 years at diagnosis; (2) previous BC history; (3) patients with incomplete data such as survival data, follow-up data, and molecular types of BC. All data usage in this study was in accordance with the data-use agreements of SEER database. Anonymized patient data from the SEER database were used in this study, and interventions on patients were not involved. Therefore, this study was granted an ethical exemption by the Ethics Committee of The Second Affiliated Hospital of Shantou University Medical College.

2.2. Data Collection

Demographic and clinicopathological data were extracted, including age (18-40, 40-65, and ≥65 years), race (blacks, Hispanics, whites, and others), marital status (married, separated/divorced, single, widowed, and unknown), estrogen receptor (ER) status (negative and positive), pathology grade (I, II, III, IV, and unknown), progesterone receptor (PR) status (negative, positive, and unknown), T stage (T1, T2, T3, T4, unknown), N stage (N1, N2, N3, N4, and unknown), radiation (no or yes), chemotherapy (no or yes), surgery (no or yes), regional nodes positive (no or yes), molecular subtypes (triple negative, HER2 positive, Luminal A, and Luminal B), survival status (alive and dead), and survival months. The primary outcome was overall survival (OS), which was calculated from date of diagnosis to date of death (between 2010 and 2015) or censor date.

The American Joint Committee on Cancer (AJCC) (6th edition) staging system was utilized to determine the T stage and N stage [23]. The classification criteria for molecular subtypes of BC were based on the criteria in 2011 [24]. PCA was identified by the SEER sequence number, which contained information on all primary reportable tumors in the patient. For example, the sequence number “00” means that the patient has only one primary cancer. If the patient is diagnosed with the second reportable tumor, the sequence number of the first tumor is changed from “00” to “01,” the sequence number of the second cancer is “02,” and so on.

2.3. Statistical Analysis

The Kolmogorov-Smirnov test was utilized to assess the normality of the data. Measurement data were described by mean ± standard deviation (SD) or median and interquartile range [M (Q1, Q3)], and the t-test or Mann–Whitney U rank-sum test was utilized to compare differences between groups. Categorical data were expressed by the numbers and proportions [n (%)], and the chi-square test was utilized to compare differences between groups. Univariate analysis was utilized to analyze the differences between patients with and without a PCA. The Kaplan-Meier (K-M) curves and Cox proportional-hazards model were utilized to determine the effect of PCA on survival in patients with different subtypes of BC. Hazard ratio (HR) and 95% confidence interval (CI) were utilized for data measurement. Model 1 was a univariate analysis model; model 2 was an age-adjusted model; model 3 was a multivariate analysis model that adjusted for age, race, marital status, grade, ER (or not), PR (or not), T stage, N stage, chemotherapy, radiation, surgery, and regional nodes positive.

The SAS 9.4 software (SAS Institute Inc., Cary, NC, USA) was utilized to complete the univariate and multivariate analyses, and KM curves were completed by the R 4.20 software (R Foundation for Statistical Computing, Vienna, Austria). P < 0.05 was considered to be statistically significant.

3. Results

3.1. Characteristics of Patients

A total of 44,335 patients with primary BC were extracted from the SEER database (2010 to 2015). Of these patients, 5,741 patients with a previous BC history and 2,954 patients with incomplete BC molecular subtypes data were excluded, 35,640 patients including 2,038 (5.72%) with PCA and 33,602 (94.28%) without were enrolled in this study. Table 1 presents the characteristics of included patients. In terms of BC subtypes, 3,853 (10.81%) patients were triple negative, 24,754 (69.46%) patients were HER2 positive, 2,406 (6.75%) patients were Luminal A, and 4,627 (12.98%) patients were Luminal B. Among the characteristics of patients, more patients were 40-65 years (57.02%), whites (65.36%), married (50.64%), II-III grade (71.88%), ER-positive (81.38%), and PR-positive (70.02%). Among all included patients, 29,387 (82.46%) patients were alive and 6,253 (17.54%) patients died. The median survival time of all patients was 27.00 (11.00, 47.00) months. Among the 2,038 patients with a PCA, female genital system cancer (24.09%), digestive system cancer (20.71%), and respiratory system cancer (11.29%) accounted for a higher proportion of patients (Figure 1).

Table 1

Characteristics of all included patients.

Variables	Total (n =35640)	Prior cancer history		Statistics	P
		No (n =33602)	Yes (n =2038)
Breast subtypes, n (%)				χ 2 = 54.815	<0.001
Triple negative	3853 (10.81)	3658 (10.89)	195 (9.57)
HER2 positive	24754 (69.46)	23197 (69.03)	1557 (76.40)
Luminal A	2406 (6.75)	2318 (6.90)	88 (4.32)
Luminal B	4627 (12.98)	4429 (13.18)	198 (9.72)
Age, years, n (%)				χ 2 = 696.118	<0.001
18-40	2778 (7.79)	2734 (8.14)	44 (2.16)
40-65	20322 (57.02)	19590 (58.30)	732 (35.92)
≥65	12540 (35.19)	11278 (33.56)	1262 (61.92)
Race, n (%)				χ 2 = 136.469	<0.001
Blacks	4442 (12.46)	4251 (12.65)	191 (9.37)
Hispanics	4354 (12.22)	4195 (12.48)	159 (7.80)
Whites	23294 (65.36)	21722 (64.64)	1572 (77.13)
Others	3550 (9.96)	3434 (10.22)	116 (5.69)
Marital, n (%)				χ 2 = 187.292	<0.001
Married	18048 (50.64)	17132 (50.99)	916 (44.95)
Separated/divorced	4038 (11.33)	3812 (11.34)	226 (11.09)
Single	6023 (16.90)	5766 (17.16)	257 (12.61)
Widowed	4741 (13.30)	4275 (12.72)	466 (22.87)
Unknown	2790 (7.83)	2617 (7.79)	173 (8.49)
Grade, n (%)				χ 2 = 45.520	<0.001
I	5749 (16.13)	5368 (15.98)	381 (18.69)
II	14079 (39.50)	13250 (39.43)	829 (40.68)
III	11541 (32.38)	11002 (32.74)	539 (26.45)
IV	155 (0.43)	150 (0.45)	5 (0.25)
Unknown	4116 (11.55)	3832 (11.40)	284 (13.94)
ER, n (%)				χ 2 = 26.880	<0.001
Negative	6636 (18.62)	6345 (18.88)	291 (14.28)
Positive	29004 (81.38)	27257 (81.12)	1747 (85.72)
PR, n (%)				χ 2 = 3.949	0.139
Negative	10514 (29.50)	9952 (29.62)	562 (27.58)
Positive	24955 (70.02)	23490 (69.91)	1465 (71.88)
Unknown	171 (0.48)	160 (0.48)	11 (0.54)
T stage, n (%)				χ 2 = 80.042	<0.001
T1	14455 (40.56)	13533 (40.27)	922 (45.24)
T2	9833 (27.59)	9253 (27.54)	580 (28.46)
T3	3304 (9.27)	3177 (9.45)	127 (6.23)
T4	4054 (11.37)	3914 (11.65)	140 (6.87)
Unknown	3994 (11.21)	3725 (11.09)	269 (13.20)
N stage, n (%)				χ 2 = 94.441	<0.001
N1	17970 (50.42)	16750 (49.85)	1220 (59.86)
N2	10705 (30.04)	10231 (30.45)	474 (23.26)
N3	3011 (8.45)	2879 (8.57)	132 (6.48)
N4	2370 (6.65)	2268 (6.75)	102 (5.00)
Unknown	3994 (11.21)	3725 (11.09)	269 (13.20)
Radiation, n (%)				χ 2 = 37.418	<0.001
No	23356 (65.53)	21893 (65.15)	1463 (71.79)
Yes	12284 (34.47)	11709 (34.85)	575 (28.21)
Chemotherapy, n (%)				χ 2 = 220.064	<0.001
No	19145 (53.72)	17726 (52.75)	1419 (69.63)
Yes	16495 (46.28)	15876 (47.25)	619 (30.37)
Surgery, n (%)				χ 2 = 3.579	0.059
No	6725 (18.87)	6308 (18.77)	417 (20.46)
Yes	28915 (81.13)	27294 (81.23)	1621 (79.54)
Regional nodes positive, n (%)				χ 2 = 0.038	0.845
No	15236 (42.75)	14369 (42.76)	867 (42.54)
Yes	20404 (57.25)	19233 (57.24)	1171 (57.46)
Status, n (%)				χ 2 = 111.988	<0.001
Alive	29387 (82.46)	27883 (82.98)	1504 (73.80)
Dead	6253 (17.54)	5719 (17.02)	534 (26.20)
Survival months, M (Q1, Q3)	27.00 (11.00,47.00)	27.00 (11.00,47.00)	25.00 (10.00,43.00)	Z = -3.893	<0.001

Note: HER2: human epidermal growth factor receptor 2; ER: estrogen receptor; PR: progesterone receptor.

Figure 1

Distributions of previous cancer types in patients with breast cancer.

Significant differences were found in breast subtype, age, race, marital status, grade, ER status, T stage, N stage, radiation, chemotherapy, survival status, and survival months among patients with or without PCA (all P < 0.001) (Table 1).

3.2. Relationship between PCA and Survival of Patients with Different BC Subtypes

To investigate the relationship between different types of PCA and the prognosis of patients with different BC subtypes, the types of PCA were grouped into previous female genital/endocrine system cancer history and previous other cancers history. The K-M curves of the relationship between different types of PCA and the prognosis of BC patients are presented in Figure 2. The results indicated that a worse prognosis was found in patients with PCA than those without (P < 0.0001). BC patients with a previous female genital/endocrine system cancer history (HR =1.27; 95% CI, 1.08-1.51) and previous other cancers history (HR =1.81; 95% CI, 1.63-1.51) were related to a poor prognosis. After adjusting for all confounders, patients with a previous female genital/endocrine system cancer history (HR =1.38; 95% CI, 1.16-1.63) and previous other cancers history (HR =1.55; 95% CI, 1.40-1.72) were still related to poor OS. In terms of BC subtypes, previous female genital/endocrine system cancer history (HR =1.70; 95% CI, 1.18-2.43) and previous other cancers history (HR =1.49; 95% CI, 1.14-1.96) were related to worse survival in patients with triple-negative subtype. Among HER2-positive subtype patients, patients with previous female genital/endocrine system cancer history (HR =1.32; 95% CI, 1.08-1.64) and previous other cancers history (HR =1.48; 95% CI, 1.31-1.68) were also associated with a poor prognosis. However, previous other cancers history was only related to poor OS in patients with Luminal A (HR =1.64; 95% CI, 1.06-2.55) and Luminal B (HR =2.50; 95% CI, 1.83-3.43) subtypes, while previous female genital/endocrine system cancer history may not influence the prognosis significantly (all P > 0.05) (Table 2). The K-M curves of the effect of PCA on the survival of patients with different BC subtypes are demonstrated in Figure 3.

Figure 2

The Kaplan-Meier (K-M) curves of the relationship between different types of previous cancer history and the survival of breast cancer patients.

Table 2

Impact of prior cancer history on the survival of patients with different breast cancer subtypes.

Populations	Samples	Subgroups	Model 1		Mode 2		Model 3∗
Total	35640		HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
	33602	No	Ref		Ref		Ref
	672	Prior female genital/endocrine system cancer history	1.27 (1.08-1.51)	0.005	1.07 (0.91-1.27)	0.412	1.38 (1.16-1.63)	<0.001
	1366	Prior other cancers history	1.81 (1.63-2.00)	<0.001	1.41 (1.27-1.56)	<0.001	1.55 (1.40-1.72)	<0.001
Triple negative	3568	No	Ref		Ref		Ref
	73	Prior female genital/endocrine system cancer history	1.31 (0.92-1.87)	0.137	1.22 (0.86-1.75)	0.267	1.70 (1.18-2.43)	0.004
	122	Prior other cancers history	1.62 (1.25-2.11)	<0.001	1.31 (1.01-1.72)	0.045	1.49 (1.14-1.96)	0.004
HER2 positive	23197	No	Ref		Ref		Ref
	500	Prior female genital/endocrine system cancer history	1.35 (1.10-1.66)	0.005	1.10 (0.89-1.36)	0.372	1.32 (1.08-1.64)	0.008
	1057	Prior other cancers history	1.90 (1.67-2.15)	<0.001	1.45 (1.28-1.65)	<0.001	1.48 (1.31-1.68)	<0.001
Luminal A	2318	No	Ref		Ref		Ref
	35	Prior female genital/endocrine system cancer history	0.81 (0.36-1.81)	0.603	0.65 (0.29-1.45)	0.294	0.88 (0.39-1.99)	0.758
	53	Prior other cancers history	2.00 (1.30-3.06)	0.002	1.47 (0.96-2.26)	0.081	1.64 (1.06-2.55)	0.027
Luminal B	4429	No	Ref		Ref		Ref
	64	Prior female genital/endocrine system cancer history	1.22 (0.69-2.16)	0.497	1.04 (0.59-1.84)	0.889	1.49 (0.84-2.67)	0.175
	134	Prior other cancers history	2.50 (1.85-3.38)	<0.001	1.85 (1.37-2.51)	<0.001	2.50 (1.83-3.43)	<0.001
Age: 18-40 years	2734	No	Ref		—		Ref
	13	Prior female genital/endocrine system cancer history	—		—	—	—	—
	31	Prior other cancers history	0.91 (0.29-2.84)	0.871	—	—	2.00 (0.62-6.41)	0.245
Age: 40-44 years	19590	No	Ref				Ref
	299	Prior female genital/endocrine system cancer history	1.42 (1.08-1.87)	0.012	—	—	1.84 (1.40-2.43)	<0.001
	433	Prior other cancers history	1.66 (1.35-2.05)	<0.001	—	—	2.19 (1.77-2.72)	<0.001
Age: ≥65 years	11278	No	Ref		—		Ref
	360	Prior female genital/endocrine system cancer history	0.95 (0.77-1.17)	0.611	—	—	1.18 (0.95-1.46)	0.132
	902	Prior other cancers history	1.35 (1.19-1.51)	<0.001	—	—	1.39 (1.23-1.57)	<0.001
Race: blacks	4251	No	Ref		Ref		Ref
	54	Prior female genital/endocrine system cancer history	0.97 (0.56-1.68)	0.919	0.87 (0.50-1.51)	0.624	1.28 (0.74-2.23)	0.381
	137	Prior other cancers history	1.45 (1.08-1.94)	0.014	1.25 (0.93-1.68)	0.149	1.41 (1.04-1.92)	0.027
Race: Hispanics	4195	No	Ref		Ref		Ref
	73	Prior female genital/endocrine system cancer history	1.38 (0.78-2.45)	0.268	1.25 (0.71-2.22)	0.440	1.91 (1.07-3.41)	0.030
	86	Prior other cancers history	2.29 (1.55-3.39)	<0.001	2.02 (1.36-3.00)	<0.001	2.12 (1.41-3.18)	<0.001
Race: others	3434	No	Ref		Ref		Ref
	55	Prior female genital/endocrine system cancer history	1.47 (0.79-2.75)	0.229	1.35 (0.72-2.53)	0.347	1.98 (1.04-3.75)	0.037
	61	Prior other cancers history	1.71 (0.99-2.97)	0.057	1.41 (0.81-2.45)	0.229	1.55 (0.88-2.72)	0.130
Race: whites	21722	No	Ref		Ref		Ref
	490	Prior female genital/endocrine system cancer history	1.33 (1.10-1.62)	0.004	1.09 (0.90-1.32)	0.393	1.32 (1.08-1.60)	0.006
	1082	Prior other cancers history	1.84 (1.64-2.06)	<0.001	1.39 (1.23-1.56)	<0.001	1.54 (1.36-1.73)	<0.001

Note: model 1, univariate analysis model; model 2, age-adjusted model; model 3, multivariate analysis model that adjusted for age, race, marital status, grade, ER (or not), PR (or not), T stage, N stage, chemotherapy, radiation, surgery, and regional nodes positive; ∗model 3 did not adjust for ER and PR when performing an analysis on populations with triple negative, HER2 positive, Luminal A, and Luminal B; HR: hazard ratio; 95% CI: 95% confidence interval.

Figure 3

The K-M curves of the impact of previous cancer history on the survival of patients with different breast cancer subtypes. (a) Triple-negative subtype; (b) HER-positive subtype; (c) Luminal A subtype; (d) Luminal B subtype.

Stratified analyses were conducted according to patient age and race. In terms of age, previous other cancers history was related to poor prognosis in patients aged 40-64 years (HR =2.19; 95% CI, 1.77-2.72) and ≥65 years (HR =1.39; 95% CI, 1.23-1.57). However, only a previous female genital/endocrine system cancer history (HR =2.50; 95% CI, 1.83-3.43) was a risk factor for OS in patients aged 40-64 years. In addition, the prognosis of patients aged 18-40 years may not be influenced by PCA (P > 0.05). Among patients of different races, except for previous female genital/endocrine system cancer may not influence the prognosis of blacks and previous other cancers history may not affect the prognosis of other races (P > 0.05), both previous female genital/endocrine system cancer history and previous other cancers history were risk factors for OS of different races (P < 0.05) (Table 2). In addition, the K-M curves of PCA for survival in patients with different ages and races were displayed in Figures 4 and 5, respectively.

Figure 4

The K-M curves of the association between previous cancer history and the survival of patients with different ages. (a) Age 18-40 years; (b) age 40-65 years; (c) age ≥65 years.

Figure 5

The K-M curves of the relationship between previous cancer history and the survival of patients with different races. (a) Blacks; (b) Hispanics; (c) whites; (d) others.

4. Discussion

The association between PCA and survival of patients with different BC subtypes was analyzed. There were 5.72% of BC patients who had a PCA, and HER2-positive (69.46%) BC was the most common subtype. The results displayed that a PCA was associated with poor OS in BC patients. In terms of BC subtypes, PCA may be related to poor OS in patients with triple-negative and HER2-positive subtypes, while the survival of patients with Luminal A and B subtypes may not be influenced by previous female genital/endocrine system cancer history. Furthermore, stratified analyses showed that the prognosis of patients aged 18-40 years may not be influenced by PCA, and previous female genital/endocrine system cancer history may also not influence the prognosis of patients aged ≥65 years.

In clinical practice, patients with PCA are routinely excluded as previous cancers may affect the prognostic outcomes. Previous studies have assessed the relationship between PCA and the prognosis of cancer patients [17, 18, 25, 26]. These studies demonstrated that the relationship between PCA and the prognosis of patients was related to the type of cancer. Laccetti et al. found that the prognosis of lung cancer patients may not be affected by PCA, independent of the stage and type of previous cancer [17, 18]. Wen et al. suggested that the clinical prognosis of most gastric cancer patients may be independent of PCA [25]. However, Zhu et al. showed that a PCA was linked to poorer OS in larynx cancer patients [26]. The study conducted by Lin et al. indicated that the poor prognosis in patients with advanced BC was affected by PCA [6], which was similar to our results. The study of Lin et al. was mainly about the impact of the diagnosis time and location of previous cancer on the survival of BC patients, while our study was to assess the influence of PCA on the prognosis of patients with different molecular subtypes. Furthermore, our study analyzed the impact of different types of PCA on the survival of patients.

Our results demonstrated that a PCA was related to poorer OS of patients with triple-negative and HER2-positive subtypes. Ren et al. found that the mortality of patients was linked to BC subtypes, specifically HER2-positive patients had the highest mortality, followed by the triple-negative, Luminal A, and Luminal B subtypes [27]. Furthermore, BC patients with the HER2-positive subtype had the highest number of genetic mutations compared with other subtypes [28]. Our results found that the HER2-positive subtype accounted for the largest proportion in BC patients with PCA. PCA was related to poor prognosis in patients with HER2-positive subtype, and clinicians may need to pay more attention to the treatment and management of these patients. Furthermore, we found that the survival of patients with Luminal A and Luminal B subtypes may not be influenced by previous female genital/endocrine system cancer history. One possible explanation was that endocrine therapy in BC patients with Luminal A and Luminal B subtypes may reduce the impact of a previous female genital/endocrine system cancer history on the prognosis of patients. Because endocrine therapy has become an essential treatment for patients with ER-positive early BC [29]. Our results also indicated that prior other cancers history was related to poorer survival in patients with Luminal A and Luminal B subtypes. This may be related to many factors, such as the type and treatment methods of previous cancer, and the specific explanation may require further study.

Subgroup analyses presented that the prognosis of BC patients aged 18-40 years may not be influenced by PCA. Age at diagnosis is commonly considered to correlate with prognosis in BC patients [30]. Young age (<40 years) has been identified as an independent risk factor associated with poorer prognosis in BC patients in several studies [31, 32]. However, the association of age with BC mortality is not a simple linear correlation, with women aged 45 to 55 having the lowest risk of dying from BC [33, 34]. In our study, PCA was related to poorer OS in patients aged 40-64 years, whereas OS in patients aged 18-40 years may not be influenced by PCA. This could be potentially explained that young age patients diagnosed with previous cancer were more frequently involved in the healthcare system (e.g., regular follow-up examinations), which led to the early diagnosis of BC. Other possible explanations included individualized patient biology and treatment responsiveness [18].

This study filled the gap in the relationship between PCA and prognosis of BC patients with different molecular subtypes. However, the study also had some limitations that should be considered. First, some confounders such as genetic mutations, prior cancer occurred, stage of prior cancer, and type and dose of treatment drugs may affect the survival, but they are not available due to the limitations of the SEER database. Second, compared to the entire SEER database, the data sample we included in the analysis was small as data on BC molecular subtypes were not complete until after 2010. Third, the sample size of some patients with previous cancer was too small to analyze the relationship between a certain PCA and the survival of BC patients. Fourth, we did not analyze the relationship between PCA and BC-specific survival.

5. Conclusion

This study explored the association between PCA and the survival of patients with different molecular subtypes of BC. PCA was associated with poorer survival of patients with triple-negative and HER2-positive subtypes, and the prognosis of patients with Luminal A and Luminal B subtypes may not be influenced by previous female genital/endocrine system cancer history. In BC clinical trials, the exclusion criteria for patients with PCA should be modified according to the BC type, age, and type of PCA rather than directly excluding patients with a history of cancer. Such processing can obtain more accurate clinical trial results.