Risk of second primary cancer among patients with early operable breast cancer registered or randomised in Danish Breast Cancer cooperative Group (DBCG) protocols of the 77, 82 and 89 programmes during 1977–2001

Abstract

Breast cancer survivors have increased risks of developing second primary cancers due to shared etiology, life style factors but also to primary breast cancer treatment. Among 53 418 patients registered by the population based Danish Breast Cancer Cooperative Group (DBCG) during 1977–2001, 31 818 patients were treated and followed according to guidelines of DBCG. In addition to surgery 23% received tamoxifen, 23% chemotherapy and 35% radiotherapy as treatment for primary breast cancer. Second primary cancers were identified by linkage to the population based Danish Cancer Register. Cancer incidence rates of the Danish population were used for calculation of standardized incidence ratios (SIRs). Time at risk was from diagnosis of breast cancer+1 year until death or through 2002. Risk for all second primary cancers combined was increased, SIR=1.04 (95% confidence interval 0.99–1.08). Sites with significantly elevated risks included corpus uteri (SIR=1.23), ovary (1.39), soft tissues (2.24), acute leukaemia (2.02), and sites potentially inducible by breast cancer radiotherapy combined (1.11). For irradiated patients compared to non-irradiated the risk was increased for all sited combined, radiotherapy-related sites, colon and soft tissues. Tamoxifen treated had, compared to non-treated, elevated risk for cancer of corpus uteri (SIR=1.83 vs 1.04). Patients given adjuvant chemotherapy had, compared to those not, elevated risks for all sites combined (SIR=1.24 vs 1.01) and for ovary (2.16 vs 1.24). Risk for cancer of the lung, uterus and ovary was analysed using multivariate Poisson regression. For lung cancer the risk was related to radiotherapy and time since diagnosis, the relative risk for lung cancer being 1.33 (95% CI 1.00–1.77) (irradiated vs non-irradiated). Ovary cancer risk was inversely related to age at diagnosis but not to treatment and corpus uteri cancer risk related to tamoxifen treatment, relative risk 1.57. The findings are in accordance to other population based studies.

Breast cancer is the leading cause of cancer morbidity and the second leading cause of cancer mortality among Danish women [1]. The incidence of breast cancer has been steadily increasing for longer than 50 years, but the mortality has remained stable as the prognosis has improved [2]. This has been attributed to earlier diagnosis and more widespread use of pre- or postoperative adjuvant treatments such as radiotherapy, chemotherapy and hormonal treatment which has been demonstrated to improve disease free and overall survival [3], [4].

With improved survival more breast cancer patients are at risk of having a new primary cancer diagnosed. The incidence of new primary cancers in breast cancer patients has been studied in case-series, case-control studies and cohort studies using population based cancer register-based data. The advantage of case series and case-control studies is that they often include precise information about demographic and treatment data, while weaknesses are a limited number of patients often carefully selected and with a short follow-up time. Cancer register data based cohort studies are population based and can include huge numbers of unselected patients with often a very long follow-up. However, such studies have rarely precise information about primary treatment for breast cancer. In the largest cohort studies among patients with breast cancer, significantly increased risks for second cancers at several sites were demonstrated compared to that of the general population [5–7]. The increased risk of second primary cancers has been ascribed to shared etiology including genetic susceptibility, environmental and lifestyle factors but also to treatment effects.

In this study we examined the incidence of second primary cancers among patients registered, nationwide and population-based, treated and followed-up according to the guidelines or to randomised trials conducted by the Danish Breast Cancer Cooperative Group (DBCG) during 1977–2001.The incidence of new primary cancer among the patients in the present cohort has been reported earlier as minor parts of cancer registry based international studies [5], [6]. These studies did not include data on treatment of early breast cancer. The present study includes data on initial postoperative treatment enabling examination of the relationship between initial treatment and subsequent risk of new primary cancers.

Material and methods

Patients in Denmark with early breast cancer (i.e. operable breast cancer without distant metastases at the time of diagnosis) have since 1977 been treated according to the guidelines of DBCG. Nearly all cases diagnosed with breast cancer in the country have been reported to the DBCG. Patients have been stratified into low and high-risk groups according to estimated risk of recurrence. Patients in the low risk group did not receive systemic adjuvant therapy while patients with high risk received pre-defined systemic adjuvant therapy in the framework of randomised clinical trials or outside clinical trials according to guidelines or to patient preference. The definition of high and low risk has changed gradually over time to include a higher fraction in the high risk group so that more and more patients have received adjuvant systemic treatment. Thus adjuvant chemotherapy was administered to an increasing number of patients over time. Up to 1989 it was CMF or CMF-like regimens while since 1989 many patients received anthracycline-containing chemotherapy. Hormonal treatment consisted mostly of tamoxifen administered for one year until mid 1990′s where the treatment was prolonged to 5 years. Radiotherapy has been given postmastectomy to almost all high-risk patients (node positive, large tumour) during 1977–1982. During 1982–1989 postmastectomy radiotherapy or control was administered by randomisation for patients participating in the randomised trials while patients outside trials usually received postmastectomy radiotherapy. After 1989 postmastectomy radiotherapy was recommended only to selected patient groups (young age, many positive lymph nodes) until 1999 where the majority of node positive patients undergoing mastectomy received radiotherapy. After breast conserving surgery postoperative radiotherapy has been recommended to all patients. Postoperative radiotherapy has usually included chest wall or residual breast and, for node positive, regional lymph nodes including internal mammary, (since 1982) periclavicular and axillary lymph nodes. Patients are followed-up for disease status by DBCG until recurrence, death or 10 years after diagnosis. The DBCG database does not hold information about therapy for metastatic or locally advanced breast cancer. Patients are followed-up life long for vital status by register linkage [9].

The Danish Cancer Register has since 1943 received reports about all diagnosed cases of cancer including secondary or higher order primary cancers from hospital departments, pathology departments and physicians. The reports include information about age, sex, date of birth and diagnosis, and site of the cancer. The Danish Cancer Register is considered to be complete and precise [9].

From the DBCG database information including birth date, date of diagnosis of breast cancer, postoperative adjuvant treatment (chemotherapy yes/no, tamoxifen yes/no, radiotherapy yes/no) and vital status was extracted for female patients with breast cancer. Only patients who were treated and followed in the framework of a treatment protocol were included because only for such patients the information about therapy is considered to be valid. Eligibility criteria for treatment in a DBCG-protocol have changed over time, e.g. the upper age limit has been increased. For the cases where information about treatment was missing, it was assumed that the patient received the therapy prescribed in the treatment protocol. The patients were linked to the Danish Cancer Register. Patients with a previous malignancy (excluding non-melanoma skin cancer) were excluded. Likewise, patients from the DBCG database not having a diagnosis of breast cancer in the Danish Cancer Register were excluded. For the remaining patients, second primary cancers (excluding non-melanoma skin cancer) were identified, and information about date of diagnosis and site of the second primary cancer was extracted. Second primary breast cancer following a first primary breast cancer was not considered as an outcome as registration hereof has not been consistent over the years.

Standardised incidence ratios (SIR′s) were calculated as the observed numbers of site specific second cancers relative to the expected numbers which were estimated by applying the appropriate numbers of person-years at risk to incidence rates of the general population specified by site, gender, age (5-years intervals) and calendar period (5-years intervals). Cancer sites were classified by groups of ICD-10 codes as in NORDCAN [10]. Time at risk was from date of diagnosis of breast cancer as registered in the DBCG+one year until first occurrence of second primary cancer, death, emigration or December 31, 2002. Ninety five percent confidence intervals (CI) were constructed assuming the expected numbers to be Poisson-distributed. Standard likelihood ratio tests comparing Poisson regression models were used to test for heterogeneity across subgroups of follow-up interval, age, and treatment regimes in separate models. Tests for trend were applied where appropriate. Level of significance was set to 0.05. SIR‘s were calculated site-specific, for all cancer sites combined, and for sites potentially inducible by breast cancer radiotherapy based on the literature (salivary glands, esophagus, lung, pleura, thyroid, bone, soft tissue, acute leukaemia) [11–23] or suspected (stomach).

For selected cancer sites, lung, corpus uteri, and ovary, the risk of second cancer after 5 years from diagnosis of breast cancer was analysed using multiplicative Poisson models including the variables histology (ductal, lobular, other), time since diagnosis, age at diagnosis, and adjuvant treatment (chemotherapy, tamoxifen, radiotherapy).

Results

From the database of the DBCG 53 418 patients were registered in the DBCG 77, 82 and 89 programmes. Of these 19 435 were not treated according to a protocol (due to high age 9 799 patients, surgery not according to DBCG guidelines 1 870, prior malignancy 1 584, metastases at time of diagnosis 1 267, bilateral or inoperable disease 1 150, misclassification 1 091, comorbidity 956, no patient consent 755, and other reasons 963. Note that criteria for non-eligibility have changed over time). In 1999 new treatment programmes were launched but some patients registered during 2000–2001 were treated according to the treatment programmes DBCG 77, 82 and 89. They were included in this study resulting in follow-up of 33 983 patients. For 2 165 of these patients, survival after breast cancer diagnosis was shorter than one year or there were discrepancies between information in the DBCG database and the cancer register, leaving a total of 31 818 eligible patients. Characteristics of the patients are given in Table I. The median follow-up time was 7.8 years (range 1–25 years). The median age at breast cancer diagnosis was 56 years (range 19–89). Twenty three percent of patients received adjuvant tamoxifen, 23% received adjuvant chemotherapy, and 35% received postoperative radiotherapy as adjuvant treatment after surgery.

Table I.  Characteristics of 31 818 1-year survivors of breast cancer in Denmark 1977–2001 treated according to Danish Breast Cancer Cooperative Group guidelines and followed-up for second primary cancers.

	No. of pts	Person-years at risk	No. of pts with second primary cancers
All	31 818	256 563	1 993
Age at breast cancer diagnosis
< 50	9 936	89 652	420
50–59	9 483	77 416	604
60–69	9 427	71 226	762
70–79	2 809	17 385	192
80–89	163	884	15
Calendar year of breast cancer diagnosis
1977–1981	4 600	53 183	468
1982–1986	5 909	64 591	528
1987–1991	6 916	63 228	468
1992–1996	9 542	59 220	421
1997–2001	4 851	16 341	108
Years of follow-up after breast cancer
1–9		195 390	1 370
10–19		57 249	568
20 +		3 923	55
Adjuvant tamoxifen
No	24 614	209 098	1 539
yes	7 204	47 465	454
Adjuvant chemotherapy
no	24 380	201 077	1 664
yes	7 438	55 486	329
Postoperative radiotherapy
no	20 633	171 927	1 332
yes	11 185	84 636	661

Among 1-year survivors of breast cancer a total of 1 993 secondary primary cancers were observed (Table II). The SIR for all sites combined was 1.04 (95% CI 0.99–1.08). Sites with significantly increased SIR included corpus uteri (SIR = 1.23), ovary (1.39), soft tissues (2.24), acute leukaemia (2.02), and the group of sites potentially inducible by breast cancer radiotherapy combined (1.11). Borderline significantly elevated SIR′s were observed for mouth and rectum (1.15). Notably, risk for lung cancer was not increased (1.01). Significant trends towards increasing risk with increasing time since diagnosis of breast cancer (suggestive of a treatment effect) were seen for radiation-related sites combined and lung. Sites with risks significantly lower than expected were liver (0.28), nose (0.39), and cervix uteri (0.69). For cervix uteri and ovary a non-significant inverse relation between risk and time since breast cancer diagnosis was seen.

Table II.  Number of second primary cancers with standardized incidence ratios among 1-year survivors of breast cancer in Denmark 1977–2001, total and specified for second cancer site and for years since diagnosis of breast cancer.

	Follow-up interval since breast cancer diagnosis
	All			1–9			10–19			20–
	N^1	SIR^2	(95% CI)^3	N	SIR	(95% CI)	N	SIR	(95% CI)	N	SIR	(95% CI)
Lip	3	0.7	(0.0–1.5)	2	0.7	(0.0–1.7)	1	0.8	(0.0–2.3)	0	0.0	(0.0–0.0)
Tongue	8	1.3	(0.4–2.2)	3	0.7	(0.0–1.5)	5	2.9	(0.4–5.5)	0	0.0	(0.0–0.0)*^4
Salivary glands	2	0.5	(0.0–1.2)	0	0.0	(0.0–0.0)	1	0.9	(0.0–2.8)	1	10.1	(0.0–29.9)
Mouth	22	1.6	(1.0–2.3)	16	1.7	(0.9–2.6)	5	1.3	(0.2–2.5)	1	3.0	(0.0–8.9)
Pharynx	14	1.3	(0.6–1.9)	10	1.2	(0.5–2.0)	3	1.1	(0.0–2.3)	1	5.1	(0.0–15.0)
Oesophagus	20	1.0	(0.6–1.5)	13	1.0	(0.5–1.6)	7	1.2	(0.3–2.1)	0	0.0	(0.0–0.0)
Stomach	57	1.3	(0.9–1.6)	41	1.3	(0.9–1.7)	14	1.2	(0.6–1.8)	2	1.9	(0.0–4.6)
Small intestine	5	0.9	(0.1–1.6)	2	0.5	(0.0–1.2)	3	1.8	(0.0–3.9)	0	0.0	(0.0–0.0)
Colon	233	1.0	(0.9–1.2)	156	1.0	(0.9–1.2)	73	1.1	(0.8–1.3)	4	0.6	(0.0–1.3)
Rectum	144	1.2	(1.0–1.3)	99	1.2	(0.9–1.4)	41	1.1	(0.8–1.5)	4	1.2	(0.0–2.5)
Liver	6	0.3	(0.1–0.5)	3	0.2	(0.0–0.4)	3	0.5	(0.0–1.0)	0	0.0	(0.0–0.0)
Gallbladder	21	0.9	(0.5–1.2)	13	0.8	(0.4–1.2)	8	1.2	(0.4–2.1)	0	0.0	(0.0–0.0)
Pancreas	82	1.1	(0.8–1.3)	58	1.1	(0.8–1.4)	22	1.0	(0.6–1.4)	2	1.0	(0.0–2.3)
Nose, sinuses	2	0.4	(0.0–0.9)	2	0.6	(0.0–1.4)	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)
Larynx	10	0.9	(0.4–1.5)	8	1.0	(0.3–1.7)	2	0.8	(0.0–1.9)	0	0.0	(0.0–0.0)
Lung	323	1.0	(0.9–1.1)	185	0.8	(0.7–0.9)	117	1.3	(1.1–1.6)	21	2.9	(1.6–4.1)*
Pleura	5	1.4	(0.2–2.7)	3	1.2	(0.0–2.6)	1	1.1	(0.0–3.1)	1	13.1	(0.0–38.9)
Cervix uteri	50	0.7	(0.5–0.9)	40	0.7	(0.5–0.9)	8	0.5	(0.2–0.9)	2	2.2	(0.0–5.2)
Corpus uteri	183	1.2	(1.1–1.4)	141	1.3	(1.1–1.5)	42	1.1	(0.8–1.5)	0	0.0	(0.0–0.0)
Uterus, other	9	1.3	(0.5–2.2)	7	1.5	(0.4–2.6)	2	1.1	(0.0–2.5)	0	0.0	(0.0–0.0)
Ovary, uterine adnexa	181	1.4	(1.2–1.6)	142	1.5	(1.2–1.7)	38	1.2	(0.8–1.6)	1	0.4	(0.0–1.2)
Other female genital org	25	1.1	(0.7–1.5)	17	1.1	(0.6–1.6)	8	1.2	(0.4–2.0)	0	0.0	(0.0–0.0)
Kidney	44	0.9	(0.7–1.2)	37	1.1	(0.8–1.5)	7	0.5	(0.1–1.0)	0	0.0	(0.0–0.0)
Bladder etc.	109	1.1	(0.9–1.3)	72	1.0	(0.8–1.3)	30	1.0	(0.7–1.4)	7	2.7	(0.7–4.8)
Melanoma of skin	76	1.0	(0.8–1.2)	59	1.0	(0.8–1.3)	17	0.9	(0.5–1.3)	0	0.0	(0.0–0.0)
Eye	8	1.3	(0.4–2.2)	6	1.3	(0.3–2.4)	1	0.6	(0.0–1.8)	1	7.5	(0.0–22.3)
Brain, nervous system	69	0.9	(0.7–1.1)	45	0.8	(0.6–1.0)	22	1.1	(0.6–1.5)	2	1.2	(0.0–2.8)
Thyroid	11	0.9	(0.4–1.4)	10	1.1	(0.4–1.8)	1	0.3	(0.0–1.0)	0	0.0	(0.0–0.0)
Bone	2	0.8	(0.0–2.0)	0	0.0	(0.0–0.0)	2	3.3	(0.0–7.9)	0	0.0	(0.0–0.0)
Soft tissues	23	2.2	(1.3–3.2)	15	2.1	(1.0–3.2)	8	2.8	(0.9–4.7)	0	0.0	(0.0–0.0)
Non-Hodgkins lymphoma	76	1.2	(0.9–1.4)	53	1.2	(0.9–1.5)	22	1.2	(0.7–1.7)	1	0.6	(0.0–1.9)
Hodgkins disease	5	1.0	(0.1–1.9)	4	1.1	(0.0–2.2)	1	0.9	(0.0–2.6)	0	0.0	(0.0–0.0)
Multiple myeloma	26	1.0	(0.6–1.3)	17	0.9	(0.5–1.3)	9	1.2	(0.4–1.9)	0	0.0	(0.0–0.0)
Acute leukaemia	42	2.0	(1.4–2.6)	30	2.1	(1.3–2.8)	11	1.9	(0.8–3.1)	1	2.0	(0.0–5.9)
Other leukaemia	34	1.0	(0.7–1.4)	21	0.9	(0.5–1.4)	12	1.2	(0.5–1.8)	1	1.1	(0.0–3.2)
RT-related sites^5	485	1.1	(1.0–1.2)	297	1.0	(0.9–1.1)	162	1.3	(1.1–1.6)	26	2.6	(1.6–3.6)*
All sites	1,993	1.0	(1.0–1.1)	1,370	1.0	(1.0–1.1)	568	1.1	(1.0–1.2)	55	1.2	(0.9–1.5)

1. Observed number.

2. Standardised incidence ratio.

3. 95% confidence interval.

4. *significant test for trend.

5. Salivary glands, oesophagus, stomach, lung, pleura, thyroid, bone, soft tissue and acute leukaemia.

Risk for second primary cancers combined was highest among patients diagnosed with breast cancer below 50 years of age (SIR = 1.35) with a significant trend of a decreasing risk with increasing age (Table III). The significant inverse relation between age and risk was seen for radiation-related sites combined, esophagus, stomach, lung, ovary, bladder, and acute leukaemia.

Table III.  Number of second primary cancers with standardized incidence ratios among 1-year survivors of breast cancer in Denmark 1977–2001 specified for second cancer site and for age at diagnosis.

	Age at breast cancer diagnosis
	<50			50–59			60–69			70–89
	N^1	SIR^2	(95% CI)^3	N	SIR	(95% CI)	N	SIR	(95% CI)	N	SIR	(95% CI)
Lip	0	0.0	(0.0–0.0)	1	0.9	(0.0–2.6)	1	0.5	(0.0–1.6)	1	1.8	(0.0–5.4)
Tongue	1	0.8	(0.0–2.3)	4	2.1	(0.0–4.1)	3	1.3	(0.0–2.9)	0	0.0	(0.0–0.0)
Salivary glands	0	0.0	(0.0–0.0)	1	1.0	(0.0–2.9)	1	0.7	(0.0–1.9)	0	0.0	(0.0–0.0)
Mouth	3	1.3	(0.0–2.7)	4	1.0	(0.0–2.0)	13	2.5	(1.1–3.9)	2	1.2	(0.0–2.8)
Pharynx	3	1.1	(0.0–2.3)	4	1.0	(0.0–2.0)	6	1.6	(0.3–3.0)	1	1.4	(0.0–4.1)
Oesophagus	7	2.7	(0.7–4.7)	4*	0.7	(0.0–1.4)	8	1.0	(0.3–1.6)	1	0.4	(0.0–1.0)*
Stomach	15	2.9	(1.4–4.3)	15	1.3	(0.7–2.0)	20	1.1	(0.6–1.5)	7	0.7	(0.2–1.3)*
Small intestine	1	1.2	(0.0–3.7)	2	1.2	(0.0–2.9)	2	0.8	(0.0–2.0)	0	0.0	(0.0–0.0)
Colon	24	1.0	(0.6–1.4)	67	1.2	(0.9–1.4)	114	1.1	(0.9–1.3)	28	0.7	(0.5–1.0)
Rectum	23	1.4	(0.8–1.9)	42	1.2	(0.8–1.5)	59	1.1	(0.8–1.4)	20	1.1	(0.6–1.5)
Liver	1	0.4	(0.0–1.1)	1	0.2	(0.0–0.5)	3	0.3	(0.0–0.7)	1	0.3	(0.0–0.8)
Gallbladder	0	0.0	(0.0–0.0)	8	1.4	(0.4–2.3)	9	0.8	(0.3–1.3)	4	0.8	(0.0–1.6)
Pancreas	10	1.2	(0.4–1.9)	18	0.9	(0.5–1.3)	41	1.2	(0.8–1.5)	13	1.0	(0.5–1.6)
Nose, sinuses	0	0.0	(0.0–0.0)	1	0.7	(0.0–1.9)	1	0.4	(0.0–1.3)	0	0.0	(0.0–0.0)
Larynx	3	1.4	(0.0–3.0)	3	0.8	(0.0–1.6)	4	1.0	(0.0–2.0)	0	0.0	(0.0–0.0)
Lung	78	1.5	(1.2–1.9)	106	1.0	(0.8–1.2)	122	0.9	(0.8–1.1)	17	0.6	(0.3–0.9)*
Pleura	2	4.8	(0.0–12)	0	0.0	(0.0–0.0)	2	1.3	(0.0–3.1)	1	2.0	(0.0–5.8)
Cervix uteri	20	0.9	(0.5–1.3)	13	0.6	(0.3–0.9)	12	0.5	(0.2–0.8)	5	0.9	(0.1–1.7)
Corpus uteri	26	1.1	(0.7–1.5)	55	1.1	(0.8–1.4)	81	1.3	(1.0–1.6)	21	1.6	(0.9–2.2)
Uterus, other	1	0.8	(0.0–2.4)	2	1.0	(0.0–2.3)	4	1.5	(0.0–3.0)	2	2.5	(0.0–5.8)
Ovary, uterine adnexa	67	2.4	(1.8–3.0)	67	1.6	(1.2–2.0)	39	0.8	(0.6–1.1)	8	0.7	(0.2–1.1)*
Other female genital org	0	0.0	(0.0–0.0)	11	1.9	(0.8–3.0)	10	1.0	(0.4–1.7)	4	1.0	(0.0–2.0)
Kidney	4	0.7	(0.0–1.3)	13	1.0	(0.4–1.5)	18	0.9	(0.5–1.3)	9	1.4	(0.5–2.3)
Bladder etc.	29	2.3	(1.4–3.1)	26	0.9	(0.5–1.2)	42	0.9	(0.6–1.2)	12	0.9	(0.4–1.3)*
Melanoma of skin	23	1.0	(0.6–1.3)	25	1.1	(0.6–1.5)	24	1.0	(0.6–1.4)	4	0.6	(0.0–1.3)
Eye	3	2.3	(0.0–4.8)	2	1.0	(0.0–2.4)	3	1.3	(0.0–2.7)	0	0.0	(0.0–0.0)
Brain, nervous system	17	0.9	(0.5–1.3)	26	1.0	(0.6–1.4)	19	0.7	(0.4–1.0)	7	1.0	(0.3–1.8)
Thyroid	7	1.9	(0.5–3.2)	3	0.9	(0.0–1.9)	0	0.0	(0.0–0.0)	1	0.7	(0.0–2.2)
Bone	1	1.7	(0.0–5.0)	1	1.4	(0.0–4.2)	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)
Soft tissues	7	3.5	(0.9–6.0)	4	1.4	(0.0–2.7)	11	2.7	(1.1–4.3)	1	0.8	(0.0–2.4)
Non-Hodgkins lymphoma	12	1.0	(0.4–1.6)	23	1.2	(0.7–1.7)	33	1.2	(0.8–1.7)	8	1.0	(0.3–1.7)
Hodgkins disease	1	0.8	(0.0–2.5)	4	3.1	(0.1–6.1)	0	0.0	(0.0–0.0)	0	0.0	(0.0–0.0)
Multiple myeloma	2	0.6	(0.0–1.5)	5	0.7	(0.1–1.2)	12	1.0	(0.4–1.5)	7	1.8	(0.5–3.1)
Acute leukaemia	13	3.8	(1.7–5.8)	14	2.5	(1.2–3.7)	7	0.8	(0.2–1.4)	8	2.8	(0.9–4.7)*
Other leukaemia	4	1.0	(0.0–2.0)	13	1.5	(0.7–2.3)	14	0.9	(0.4–1.4)	3	0.5	(0.0–1.1)
RT-related sites^5	130	1.9	(1.6–2.2)	148	1.1	(0.9–1.2)	171	0.9	(0.8–1.1)	36	0.7	(0.5–1.0)*
All sites	420	1.3	(1.2–1.5)	604	1.1	(1.0–1.1)	762	1.0	(0.9–1.0)	207	0.9	(0.7–1.0)*

1. Observed number.

2. Standardised incidence ratio.

3. 95% confidence interval.

4. *significant test for trend.

5. Salivary glands, oesophagus, stomach, lung, pleura, thyroid, bone, soft tissue and acute leukaemia.

Significantly increased risks for second primary cancers were observed in patients given postoperative radiotherapy compared to patients who did not receive radiotherapy (Table IV) for all sites combined, radiotherapy-related sites, colon, and soft tissues. Among patients treated with tamoxifen the risk was marginally increased compared to those not treated with tamoxifen for stomach, kidney, melanoma of skin, and other leukaemia, but especially, as expected, for corpus uteri (SIR = 1.83 vs 1.04, statistically significant). For ovary, other female genital organs, and lung the risk was significantly lower among patients given tamoxifen compared to not-treated. Risk of second primary cancer was statistically significantly higher among patients given adjuvant chemotherapy vs. those who were not for all sites combined (SIR = 1.24 vs. 1.01), radiotherapy-related sites, esophagus, stomach, ovary, and bladder.

Table IV.  Number of second primary cancers with standardized incidence ratios among 1-year survivors of breast cancer in Denmark 1977–2001 specified for second cancer site and for administered postoperative adjuvant radiotherapy, tamoxifen and chemotherapy.

	Radiotherapy						Tamoxifen						Chemotherapy
	yes			no			yes			no			yes			no
	N^1	SIR^2	(95% CI)^3	N	SIR	(95% CI)	N	SIR	(95% CI)	N	SIR	(95% CI)	N	SIR	(95% CI)	N	SIR	(95% CI)
Lip	2	1.6	(0.0–3.9)	1	0.3	(0.0–1.0)	0	0.0	(0.0–0.0)	3	1.0	(0.0–2.0)	1	1.9	(0.0–5.6)	2	0.6	(0.0–1.3)
Tongue	1	0.5	(0.0–1.6)	7	1.6	(0.4–2.8)	3	2.2	(0.0–4.6)	5	1.0	(0.1–1.9)	2	2.1	(0.0–4.9)	6	1.1	(0.2–2.1)
Salivary glands	1	0.9	(0.0–2.6)	1	0.4	(0.0–1.1)	0	0.0	(0.0–0.0)	2	0.7	(0.0–1.6)	0	0.0	(0.0–0.0)	2	0.6	(0.0–1.5)
Mouth	4	1.0	(0.0–2.0)	18	1.9	(1.0–2.8)	6	2.0	(0.4–3.5)	16	1.6	(0.8–2.3)	4	2.1	(0.0–4.1)	18	1.6	(0.8–2.3)
Pharynx	7	2.0	(0.5–3.6)	7	0.9	(0.2–1.6)	2	0.8	(0.0–2.0)	12	1.4	(0.6–2.2)	2	1.0	(0.0–2.4)	12	1.3	(0.6–2.1)
Oesophagus	9	1.6	(0.6–2.7)	11	0.8	(0.3–1.3)	4	0.9	(0.0–1.7)	16	1.1	(0.6–1.6)	6	2.5	(0.5–4.5)	14	0.8	(0.4–1.3)*
Stomach	19	1.4	(0.8–2.0)	38	1.2	(0.8–1.6)	19	1.8	(1.0–2.6)	38	1.1	(0.8–1.5)	12	2.4	(1.0–3.8)	45	1.1	(0.8–1.4)*
Small intestine	0	0.0	(0.0–0.0)	5	1.2	(0.2–2.3)	1	0.7	(0.0–2.2)	4	0.9	(0.0–1.8)	2	2.8	(0.0–6.6)	3	0.6	(0.0–1.3)
Colon	85	1.3	(1.0–1.6)	148	0.9	(0.8–1.1)*^4	53	1.0	(0.7–1.2)	180	1.1	(0.9–1.2)	25	1.0	(0.6–1.4)	208	1.0	(0.9–1.2)
Rectum	39	1.0	(0.7–1.4)	105	1.2	(1.0–1.4)	30	1.0	(0.6–1.4)	114	1.2	(1.0–1.4)	22	1.4	(0.8–2.0)	122	1.1	(0.9–1.3)
Liver	1	0.2	(0.0–0.5)	5	0.3	(0.0–0.6)	0	0.0	(0.0–0.0)	6	0.4	(0.1–0.7)	0	0.0	(0.0–0.0)	6	0.3	(0.1–0.6)
Gallbladder	4	0.6	(0.0–1.1)	17	1.0	(0.5–1.5)	5	0.8	(0.1–1.6)	16	0.9	(0.5–1.3)	2	0.9	(0.0–2.1)	19	0.9	(0.5–1.3)
Pancreas	24	1.1	(0.6–1.5)	58	1.1	(0.8–1.4)	17	0.9	(0.5–1.3)	65	1.1	(0.9–1.4)	10	1.2	(0.4–1.9)	72	1.1	(0.8–1.3)
Nose, sinuses	0	0.0	(0.0–0.0)	2	0.5	(0.0–1.3)	0	0.0	(0.0–0.0)	2	0.5	(0.0–1.2)	0	0.0	(0.0–0.0)	2	0.4	(0.0–1.1)
Larynx	6	1.8	(0.4–3.3)	4	0.5	(0.0–1.1)	2	0.8	(0.0–2.0)	8	1.0	(0.3–1.6)	3	1.7	(0.0–3.6)	7	0.8	(0.2–1.4)
Lung	110	1.2	(1.0–1.4)	213	0.9	(0.8–1.1)	63	0.8	(0.6–1.0)	260	1.1	(0.9–1.2)*	54	1.2	(0.9–1.5)	269	1.0	(0.9–1.1)
Pleura	2	1.9	(0.0–4.6)	3	1.2	(0.0–2.6)	0	0.0	(0.0–0.0)	5	1.9	(0.2–3.5)	2	4.7	(0.0–11.3)	3	1.0	(0.0–2.1)
Cervix uteri	23	1.0	(0.6–1.4)	27	0.6	(0.3–0.8)	12	0.8	(0.4–1.3)	38	0.7	(0.4–0.9)	15	1.0	(0.5–1.6)	35	0.6	(0.4–0.8)
Corpus uteri	55	1.2	(0.9–1.6)	128	1.2	(1.0–1.4)	65	1.8	(1.4–2.3)	118	1.0	(0.9–1.2)*	23	1.1	(0.6–1.5)	160	1.3	(1.1–1.4)
Uterus, other	4	2.0	(0.0–3.9)	5	1.1	(0.1–2.0)	4	2.6	(0.1–5.2)	5	1.0	(0.1–1.8)	0	0.0	(0.0–0.0)	9	1.6	(0.5–2.6)
Ovary, uterine adnexa	65	1.6	(1.2–2.0)	116	1.3	(1.1–1.5)	27	0.9	(0.6–1.3)	154	1.5	(1.3–1.8)*	46	2.2	(1.5–2.8)	135	1.2	(1.0–1.5)*
Other female genital org	9	1.3	(0.4–2.1)	16	1.0	(0.5–1.5)	2	0.4	(0.0–0.9)	23	1.3	(0.8–1.9)*	3	1.1	(0.0–2.2)	22	1.1	(0.6–1.6)
Kidney	18	1.3	(0.7–1.9)	26	0.8	(0.5–1.1)	14	1.2	(0.6–1.9)	30	0.8	(0.5–1.2)	7	1.2	(0.3–2.1)	37	0.9	(0.6–1.2)
Bladder etc.	37	1.2	(0.8–1.6)	72	1.0	(0.8–1.2)	25	1.0	(0.6–1.4)	84	1.1	(0.9–1.3)	23	1.8	(1.1–2.6)	86	1.0	(0.8–1.2)*
Melanoma of skin	24	1.0	(0.6–1.3)	52	1.0	(0.7–1.2)	20	1.3	(0.7–1.8)	56	0.9	(0.7–1.1)	6	0.4	(0.1–0.7)	70	1.1	(0.9–1.4)*
Eye	2	1.0	(0.0–2.5)	6	1.4	(0.3–2.5)	1	0.7	(0.0–2.1)	7	1.4	(0.4–2.5)	1	1.0	(0.0–2.9)	7	1.3	(0.3–2.3)
Brain, nervous system	25	1.0	(0.6–1.4)	44	0.8	(0.6–1.0)	16	0.9	(0.5–1.4)	53	0.8	(0.6–1.1)	7	0.5	(0.1–0.9)	62	0.9	(0.7–1.2)
Thyroid	5	1.2	(0.2–2.3)	6	0.7	(0.1–1.3)	0	0.0	(0.0–0.0)	11	1.1	(0.4–1.7)	4	1.7	(0.0–3.3)	7	0.7	(0.2–1.2)
Bone	1	1.3	(0.0–3.9)	1	0.6	(0.0–1.8)	0	0.0	(0.0–0.0)	2	1.1	(0.0–2.5)	1	2.4	(0.0–7.2)	1	0.5	(0.0–1.5)
Soft tissues	12	3.9	(1.7–6.0)	11	1.5	(0.6–2.5)*	7	3.0	(0.8–5.2)	16	2.0	(1.0–3.0)	6	3.9	(0.8–7.0)	17	2.0	(1.0–2.9)
Non-Hodgkins lymphoma	17	0.9	(0.5–1.3)	59	1.3	(1.0–1.6)	15	1.0	(0.5–1.5)	61	1.2	(0.9–1.5)	8	0.8	(0.3–1.4)	68	1.2	(0.9–1.5)
Hodgkins disease	2	1.3	(0.0–3.1)	3	0.9	(0.0–1.9)	0	0.0	(0.0–0.0)	5	1.3	(0.2–2.5)	1	1.2	(0.0–3.5)	4	1.0	(0.0–2.0)
Multiple myeloma	1	0.1	(0.0–0.4)	25	1.3	(0.8–1.8)*	3	0.4	(0.0–1.0)	23	1.1	(0.7–1.6)	2	0.6	(0.0–1.5)	24	1.0	(0.6–1.4)
Acute leukaemia	15	2.4	(1.2–3.6)	27	1.9	(1.2–2.6)	9	1.8	(0.6–3.1)	33	2.1	(1.4–2.8)	9	3.2	(1.1–5.3)	33	1.8	(1.2–2.5)
Other leukaemia	8	0.8	(0.3–1.4)	26	1.1	(0.7–1.5)	12	1.5	(0.6–2.3)	22	0.9	(0.5–1.2)	7	1.8	(0.5–3.2)	27	0.9	(0.6–1.3)
RT-related sites^5	174	1.3	(1.1–1.6)	311	1.0	(0.9–1.1)*	102	1.0	(0.8–1.2)	383	1.2	(1.0–1.3)	94	1.5	(1.2–1.8)	391	1.0	(0.9–1.1)*
All sites	661	1.2	(1.1–1.2)	1332	1.0	(0.9–1.0)*	454	1.0	(0.9–1.1)	1,539	1.0	(1.0–1.1)	329	1.2	(1.1–1.4)	1664	1.0	(1.0–1.1)*

1. Observed number.

2. Standardised incidence ratio.

3. 95% confidence interval.

4. * significant test for heterogeneity.

5. Salivary glands, oesophagus, stomach, lung, pleura, thyroid, bone, soft tissue and acute leukaemia.

For the sites with relatively many observed cases and a priori hypothetical relation to treatment or shared etiology, i.e. lung, corpus uteri, and ovary, the SIR was analyzed by Poisson regression with the explaining factors: years since diagnosis, age at diagnosis, and adjuvant treatment. Table V shows for lung cancer a significant relation between risk and time since diagnosis, the risk relative to 5–9 years after diagnosis increasing to 2.99 more than 20 years after diagnosis. No significant relation was seen for chemotherapy but the risk was borderline significantly increased for radiotherapy with risk relative to no radiotherapy of 1.33. For cancer of corpus uteri a significantly increased risk of 1.57 was seen with tamoxifen. For ovary the risk was significantly related to young age at diagnosis of breast cancer (1.62 for age <50 years relative to age 50–59). The increased risk for second primary cancer of ovaries associated with chemotherapy suggested by the univariate analysis in Table IV was not confirmed in the multivariate analysis, where the relative risk was 1.28 (95% CI 0.79–2.07). Likewise, the protective effect of tamoxifen on the risk of second ovarian cancer and lung cancer suggested in Table IV, could not be confirmed in the multivariate analysis where the risk for tamoxifen treated relative to not treated was for ovary 0.78 (not significant) and for lung 0.90 (not significant).

Table V.  Risk of second primary cancer of lung, corpus uteri and ovary in 5-year survivors of breast cancer in Denmark 1977–2001 in relation to time since diagnosis of breast cancer, age at diagnosis and adjuvant treatment. The adjusted relative risk is normalized to the category with most observations and given with 95% confidence intervals.

	Lung		Corpus uteri		Ovary
		p < 0.01		p = 0.30		p = 0.14
Time since diagnosis, years
5–9	1		1		1
10–14	1.47 (1.08–2.00)		0.74 (0.48–1.13)		0.76 (0.48–1.19)
15–19	1.53 (1.04–2.24)		0.75 (0.41–1.37)		0.55 (0.28–1.08)
20–25	2.99 (1.83–4.90)
		p = 0.33		p = 0.84		p < 0.01
Age at diagnosis
< 50	1.29 (0.91–1.82)		1.15 (0.68–1.96)		1.62 (1.02–2.57)
50–59	1		1		1
60–69	1.07 (0.77–1.48)		0.93 (0.60–1.43)		0.48 (0.26–0.87)
70–79	0.70 (0.32–1.54)		1.19 (0.57–2.46)		0.30 (0.07–1.26)
		p = 0.05		p = 0.78		p = 0.23
Radiotherapy
Yes	1.33 (1.00–1.77)		0.94 (0.63–1.41)		1.31 (0.85–2.01)
No	1		1		1
		p = 0.57		p = 0.04		p = 0.43
Tamoxifen
Yes	0.90 (0.63–1.29)		1.57 (1.02–2.40)		0.78 (0.41–1.47)
No	1		1		1
		p = 0.58		p = 0.57		p = 0.33
Chemotherapy
Yes	1.11 (0.77–1.58)		0.85 (0.50–1.47)		1.28 (0.79–2.07)
No	1		1		1

Discussion

Among patients diagnosed with early breast cancer approximately 80% are offered adjuvant systemic treatment and/or postoperative radiotherapy [8] as these treatments significantly improve disease free and overall survival [3], [4]. However, it can be assumed that a substantial number of patients will receive adjuvant treatment unfoundedly due to our inability to precisely identify those patients who will gain from the therapy. At present, when selecting patients for therapy using classic patho-anatomical prognostic factors, the average reduction of the risk of recurrence induced by adjuvant therapy is relatively low, in absolute terms often only a few percent [3], [4]. In that context it is crucial that the toxicity of the treatment, including carcinogenicity, is minimal.

In large international cancer register based studies of breast cancer patients excessive numbers of second cancers at several sites have been described [5], [6] (where more than 95% of the patients in the present study were included) and [7]. Thus, Brown et al. [6] found among 376 825 1-year survivors of breast cancer patients from cancer registers in the Scandinavian countries that the risk of second solid cancers excluding contralateral breast was 1.15 and the risk of cancers potentially related to radiation therapy was 1.34 (p < 0.05). However, they also calculated that the risk of being diagnosed with non-hematological, non-breast cancer among breast cancer patients 20 years after diagnosis in absolute terms was in fact lower than that of the general population except for patients below 40 years at diagnosis of breast cancer. Mellemkjær et al. [5] found among 525 527 patients from cancer registers around the world that the risk of a second primary cancer excluding contralateral breast was 1.25 (p < 0.05) with significantly increased risks for almost all sites except liver, gall bladder, cervix uteri, brain, Hodgkin's disease, and multiple myeloma. In an US cancer register study [7] the risk for a second new primary cancer, excluding contralateral breast, among 322 863 patients who survived longer than 2 months after diagnosis of primary breast cancer was increased, 1.01 (p < 0.05). Among the patients who received radiotherapy as initial treatment (approximately one third of the patients) the risk was even higher, 1.11 (p < 0.05)

In the present study, the risk for second cancers was only marginally and not significantly increased for all sites combined (SIR = 1.04), i.e. not as high as in the non-US cancer register studies. This may be related to differences in the follow-up time. Sites with significantly decreased risks included nose and sinuses (very few cases) and liver. The latter has been observed in other studies as well [5–7] and has been ascribed to misclassification of metastases in the liver. We found that the risk for second cancers at sites potentially inducible by breast cancer radiation was increased and related to increasing time since diagnosis of breast cancer. The risk was furthermore demonstrated to be significantly higher among patients given postoperative radiotherapy indicating that this treatment modality carries a significant risk of inducing second cancers. Interestingly, the risk for second primary cancer at the sites mouth, rectum, and non-Hodgkin's lymphoma was significantly increased among patients who did not receive radiotherapy but not among patients who received radiotherapy. Similar observations have not been made in the US cancer register study [7] and it may be a chance finding.

The increased risk for second cancers among patients treated with radiation was observed for several sites including esophagus, larynx, pharynx, thyroid, stomach, bone, soft tissue, and leukaemia, but numerically lung probably contributed most. The SIR for lung cancer was not increased overall, but the SIR for patients given radiotherapy was borderline significantly higher than for those not irradiated (p = 0.05) and in multivariate analysis taking into account other factors, the relation to radiotherapy was demonstrated, patients given radiotherapy having a 33% higher risk of second lung cancer compared to not treated (p = 0.05). In support of this view is also the observed relation with increasing time since the radiotherapy in accordance with classic observations on radiation carcinogenesis. Similar observations have also been made in several other studies such as case series, randomized trials [3], case-control studies and cancer register studies [5], [6], [23]. Thus, in a US cancer register study it was noted that for patients given radiotherapy for early breast cancer during 1973–1982, the ratio of the risks for death from ipsilateral vs. contralateral lung cancer increased significantly over follow-up time to reach 2.71 (p < 0.05) after 14 years [23]. It underlines the need to improve radiotherapy techniques to avoid radiation to pulmonary tissues and to improve the selection of patients for radiotherapy.

The incidence of second cancer of the corpus uteri was significantly increased in our study as it has been in several others. We demonstrated a clear relation with risk for cancer of corpus uteri and adjuvant tamoxifen treatment with relative risk 1.57 in multivariate analysis. Such increased risks have been observed in other studies as well. Thus in patients participating in breast cancer prevention trials, tamoxifen increased the risk by 2.4 times [24]. The reason why the increase in this risk of this study was more modest could be related to the relatively short treatment time with tamoxifen (only one year) of this study. Besides tamoxifen, shared etiology, most probably endocrine factors, between breast and corpus uteri cancer could explain the increased risk. Thus, in [5] the reverse risk, i.e. risk of second breast cancer among patients diagnosed with corpus uteri cancer, was significantly increased, 1.35 in favour of shared etiology as part of the excess incidence. Tamoxifen has not consistently been linked with increased risks for other second cancers.

Risk of second cancer of the ovary was significantly increased. It was demonstrated to be related to young age at diagnosis and in the multivariate model (adjusted for age) not to any treatment modality. The frequency of heritable breast cancer is by far highest among very young breast cancer patients and it is assumed that shared genetic susceptibility including aberrations of the BRCA1 and BRCA2 genes is involved [25].

The present study examines only the incidence of new primary cancers among the subset of Danish breast cancer patients diagnosed 1977–2001 who were treated according to a treatment protocol or programme of the Danish Breast Cancer Cooperative Group (60% of all patients). Information about treatment among patients treated or followed outside treatment protocols or programmes was considered to be to imprecise. However, the reasons for non-eligibility given in the results section, in our opinion render it unlikely that the exclusion of these patients from the analysis have changed the overall results of this study.

In conclusion, this population based study of patients with early breast cancer demonstrated a modest increase in risk of second cancers, most notably in potentially radiation-related sites, and especially for lung cancer where the risk was increased by 33% in irradiated patients compared to non-irradiated patients. Risk for cancer of corpus uteri was increased in patients given tamoxifen and risk of ovarian cancer was increased in very young patients.