Advertisement

Risk of second non-breast cancer after radiotherapy for breast cancer: A systematic review and meta-analysis of 762,468 patients

Published:November 07, 2014DOI:https://doi.org/10.1016/j.radonc.2014.10.004

      Abstract

      Background and purpose: Radiotherapy for breast cancer both decreases loco-regional recurrence rates and improves overall survival. However, radiotherapy has also been associated with increased second cancer risk at exposed sites. In this meta-analysis, we estimated the risk of second non-breast cancers after radiotherapy for breast cancer. Material and methods: The databases Medline/Pubmed, Cochrane, Embase and Cinahl were systematically searched, for cohort studies on second cancer after radiotherapy for breast cancer, from inception to August 1st 2013. Included studies were to report the relative risk (RR) of second cancers comparing irradiated female breast cancer patients to unirradiated patients. Primary endpoints were all second non-breast-cancers and second cancers of the lung, esophagus, thyroid and second sarcomas. RRs were pooled using random-effects meta-analysis. Results: Thirteen studies comprising 762,468 breast cancer patients were included in the meta-analysis. Five or more years after breast cancer diagnosis radiotherapy was significantly associated with an increased risk of second non-breast cancer RR 1.12 (95% confidence interval [CI] 1.06–1.19), second cancer of the lung RR 1.39 (95% CI 1.28–1.51), esophagus RR 1.53 (95% CI 1.01–2.31) and second sarcomas RR 2.53 (95% CI 1.74–3.70). The risk increased over time, and was highest 15 or more years after breast cancer diagnosis, for second lung RR 1.66 (95% CI 1.36–2.01) and second esophagus cancer RR 2.17 (95% CI 1.11–4.25). There was no significant association between radiotherapy and second thyroid cancer. Conclusions: Radiotherapy for breast cancer is significantly associated with increased risks of second non-breast cancer, overall and in organs adjacent to the previous treatment fields. Despite a relative small absolute risk, the growing number of long-time survivors after breast cancer warrants the need for normal tissue sparing radiotherapy techniques.
      Breast cancer continues to be the most prevalent cancer among women in the Western world [

      GLOBOCAN. Available from: URL: http://globocan.iarc.fr; 2014.

      ] with a cumulative probability of more than 10% in North America and several Western European countries [
      • Forouzanfar M.H.
      • Foreman K.J.
      • Delossantos A.M.
      • et al.
      Breast and cervical cancer in 187 countries between 1980 and 2010: a systematic analysis.
      ]. In the last twenty years advances in breast cancer treatment, have led to prolonged survival-times and the majority of breast cancer patients are now cured of the disease and are becoming long-time survivors. Adjuvant radiotherapy plays an essential role in modern breast cancer treatment and has been shown to both decrease loco-regional recurrence rates and improve overall survival [
      • Overgaard M.
      • Hansen P.S.
      • Overgaard J.
      • et al.
      Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial.
      ,
      • Overgaard M.
      • Jensen M.B.
      • Overgaard J.
      • et al.
      Postoperative radiotherapy in high-risk postmenopausal breast-cancer patients given adjuvant tamoxifen: Danish Breast Cancer Cooperative Group DBCG 82c randomised trial.
      ,
      • Clarke M.
      • Collins R.
      • Darby S.
      • et al.
      Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials.
      ,
      • Darby S.
      • McGale P.
      • Correa C.
      • et al.
      Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials.
      ,
      Ebctcg Early Breast Cancer Trialists’ Collaborative Group
      Effect of radiotherapy after mastectomy and axillary surgery on 10-year recurrence and 20-year breast cancer mortality: meta-analysis of individual patient data for 8135 women in 22 randomised trials.
      ]. The changes in surgical procedures of early breast cancer have furthermore lead to an increased number of patients receiving breast conserving surgery, thus the majority of patients today will receive radiotherapy as part of their adjuvant treatment.
      During the last decade there has been an increased awareness as to radiation induced second cancers [
      • Tubiana M.
      Can we reduce the incidence of second primary malignancies occurring after radiotherapy? A critical review.
      ], and both large-scale cancer registry based studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ] and smaller single institutional studies [
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Rubino C.
      • de V.F.
      • Diallo I
      • Shamsaldin A
      • Le MG
      Increased risk of second cancers following breast cancer: role of the initial treatment.
      ] have shown, that radiotherapy for breast cancer is associated with an increased risk of second cancer. Smaller studies have however due to limited population size and observation time at times resulted in ambiguous results. The 2005 Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) overview of clinical trials is the only former published study with pooled risk estimates of second cancers after radiotherapy for breast cancer [
      • Clarke M.
      • Collins R.
      • Darby S.
      • et al.
      Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials.
      ]. Therefore, to obtain more knowledge on the risk of second cancers after radiotherapy for breast cancer we conducted the following systematic review and meta-analysis.
      Two larger cohort studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ] have previously reported that second cancers after radiotherapy for breast cancer particularly develop in organs adjacent to the previous treatment fields i.e. the organs that receive the highest dose of radiation exposure. The objective of the current analysis was therefore, to examine the risk of second cancers of the lung, esophagus, and soft tissues together with the total risk estimate of all second non-breast cancers, comparing female breast cancer patients treated with radiotherapy to those not treated with radiotherapy. Second thyroid cancer was also included as an outcome, due to its potential close location to the former treatment fields. The risk of second cancer by time since breast cancer diagnosis was additionally examined.

      Materials and methods

      Study eligibility

      The following study was conducted applying the PRISMA [

      PRISMA. Transparent reporting of systematic reviews and META-analyses. Available from: URL: <http://www.prima-statement.org/statement.htm>, 2014.

      ] guidelines for a systematic review and meta-analysis. Only cohort studies examining the risk of second non-breast cancers, comparing primary early breast cancer patients treated with radiotherapy with those patients not allocated to radiotherapy were eligible for inclusion. This analysis was restricted to the following predefined second cancer sites: lung, esophagus, thyroid, soft tissues and the combined group of all second non-breast cancers. The included studies had to report the relative risk (RR) or equivalents together with the corresponding 95% confidence interval (CI). Only original published data were considered i.e. no overviews or reviews. The following type of studies were excluded: studies on patients treated with protons or light ions, studies based on cancer mortality data, studies on patients exposed during childhood, and studies where the number of patients treated with radiotherapy was determined by approximation.

      Search strategy and selection criteria

      The Medline/Pubmed, Cochrane, Embase and Cinahl databases were systematically searched for eligible studies, from inception to August 1st 2013, using the following Medical Subject Headings (MeSH) ‘breast neoplasm’ and ‘radiotherapy’. The terms were combined with ‘neoplasm’, ‘second primary’, ‘radiation-induced’ and ‘comparative study’ using the Boolean operator ‘and’. The search strategy included no language restriction. The reference lists of the included studies were additionally carefully reviewed in the search for additional studies.

      Definition of outcomes

      The primary endpoints were all second non-breast cancers, second cancers of the lung, esophagus, thyroid, and second sarcomas. The primary metrics were the pooled RRs of second cancer comparing breast cancer patients treated with radiotherapy to those who were not. To evaluate the risk over time the RRs were pooled, both in a total risk estimate, and for the time intervals ⩾5, ⩾10 and ⩾15 years following breast cancer diagnosis. The interval from primary breast cancer diagnosis to the beginning of the observation time varied between zero and five years throughout the different studies. Therefore, in the meta-analysis we defined a ‘total pooled RR’ estimate to only included studies where the observation-time was counted from ⩽2 years after breast cancer diagnosis and onward. We did this, as we judged these RRs to be sufficiently similar for pooling of data. All other risk estimates were pooled according to the appropriate second cancer latency periods: ⩾5, ⩾10 and ⩾15 years following breast cancer diagnosis, respectively. This was done regardless of the total observation time in the studies.

      Data extraction and statistical analysis

      The RRs with the corresponding 95% CI were extracted directly from each included study. For studies reporting both unadjusted and adjusted risk estimates the most adjusted estimate was included. Risk ratios were pooled using the DerSimonion and Laird random-effects model to obtain RRs with associated 95% CI [
      Meta-analysis in Stata: an updated collection from the Stata Journal.
      ]. Each study was weighted by the inverse of its variance. The standard errors (SE) were derived from the reported 95% CI, by applying the equation SE = ln (upper limit of 95% CI/lower limit of 95% CI)/(1.96 × 2). In the event of studies reporting zero observations of second cancers the study was omitted from the final analyses, as no risk estimate was reported. I2 statistics were used to quantify heterogeneity between study results, where I2 is the percentage of total variation across studies due to heterogeneity rather than chance. Statistically significant heterogeneity was defined as I2 statistic greater than 50%. All analyses were carried out using STATA IC 11.2. (Statistical Software, TX, USA). The pooled risk estimates according to time after breast cancer diagnoses (⩾5, ⩾10, and ⩾15 years) were, if possible, extracted directly from the included studies. However, if not originally presented, we derived them from the risk estimates 1–4, 5–9, 10–14, and ⩾15 years after breast cancer diagnosis, if presented in the studies. These risk estimates were pooled, as described above. The pooled risk estimates and the corresponding forest plots for the 5-year increments: 1–4, 5–9, 10–14 years following breast cancer diagnosis for second lung, esophagus and thyroid cancer are presented in the appendix as appendix Supplementary Figs. 1–3. The risk according to type of breast cancer surgery was additionally estimated.

      Results

      Description of the studies

      The systematic search indentified 812 potentially relevant references (Fig. 1). Of these, 99 studies qualified for full text review of which 80 did not meet the eligibility criteria of this study. A second careful examination of the remaining 19 potential candidates led to the exclusion of further six studies: One study as only the standardized incidence rate ratios were reported [
      • Levi F.
      • Randimbison L.
      • Te V.C.
      • La V.C.
      Cancer risk after radiotherapy for breast cancer.
      ] and further five studies due to overlapping cohorts: two on second lung cancer [
      • Andersson M.
      • Jensen M.B.
      • Engholm G.
      • Henrik S.H.
      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.
      ,
      • Neugut A.I.
      • Robinson E.
      • Lee W.C.
      • Murray T.
      • Karwoski K.
      • Kutcher G.J.
      Lung cancer after radiation therapy for breast cancer.
      ] one on second lung and thyroid cancer [
      • Yu G.P.
      • Schantz S.P.
      • Neugut A.I.
      • Zhang Z.F.
      Incidences and trends of second cancers in female breast cancer patients: a fixed inception cohort-based analysis (United States).
      ], and two on second sarcomas [
      • Huang J.
      • Mackillop W.J.
      Increased risk of soft tissue sarcoma after radiotherapy in women with breast carcinoma.
      ,
      • Kirova Y.M.
      • Vilcoq J.R.
      • Asselain B.
      • Sastre-Garau X.
      • Fourquet A.
      Radiation-induced sarcomas after radiotherapy for breast carcinoma: a large-scale single-institution review.
      ]. The descriptive characteristics of the five excluded studies are listed in the appendix Supplementary Table 1.
      Figure thumbnail gr1
      Fig. 1Consort diagram of the study selection. SIR: Standardized incidence ratio. If the studies referred to the same population or in the case of overlapping cohorts the largest study and/or with the most relevant information was included in the meta-analysis.
      A total of 13 original studies were included in the final analysis [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Rubino C.
      • de V.F.
      • Diallo I
      • Shamsaldin A
      • Le MG
      Increased risk of second cancers following breast cancer: role of the initial treatment.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ,
      • Huang J.
      • Walker R.
      • Groome P.G.
      • Shelley W.
      • Mackillop W.J.
      Risk of thyroid carcinoma in a female population after radiotherapy for breast carcinoma.
      ,
      • Mery C.M.
      • George S.
      • Bertagnolli M.M.
      • Raut C.P.
      Secondary sarcomas after radiotherapy for breast cancer: sustained risk and poor survival.
      ,
      • Rutqvist L.E.
      • Johansson H.
      Long-term follow-up of the Stockholm randomized trials of postoperative radiation therapy versus adjuvant chemotherapy among ‘high risk’ pre- and postmenopausal breast cancer patients.
      ,
      • Zablotska L.B.
      • Neugut A.I.
      Lung carcinoma after radiation therapy in women treated with lumpectomy or mastectomy for primary breast carcinoma.
      ,
      • Zablotska L.B.
      • Chak A.
      • Das A.
      • Neugut A.I.
      Increased risk of squamous cell esophageal cancer after adjuvant radiation therapy for primary breast cancer.
      ,
      • Zhang W.
      • Becciolini A.
      • Biggeri A.
      • Pacini P.
      • Muirhead C.R.
      Second malignancies in breast cancer patients following radiotherapy: a study in Florence, Italy.
      ], nine population based cohort studies, three single institution cohort studies and one regional prospective study. Of the 13 included studies, several reported risk estimates according to various second cancer sites [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Rubino C.
      • de V.F.
      • Diallo I
      • Shamsaldin A
      • Le MG
      Increased risk of second cancers following breast cancer: role of the initial treatment.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ] occasionally leading to the fact that studies despite the primary exclusion of overlapping cohorts would refer to the same population and, or had overlapping cohorts. In such case, the largest study was included in the meta-analysis.
      All of the included studies happened to be based on European or North American populations of female breast cancer patients treated in the period between 1954 and 2007. The cumulative sample size including all 13 studies was 1,644,488 women. After eliminating overlapping cohorts the analysis included a total of 762,468 breast cancer patients, of which 321,143 (42%) received radiotherapy whereas 441,325 (58%) did not. The age at primary breast cancer diagnosis was: mean 59 years, based on 5 studies and median 57 years, based on 2 studies. 6 studies reported no exact age at breast cancer diagnosis [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ,
      • Huang J.
      • Walker R.
      • Groome P.G.
      • Shelley W.
      • Mackillop W.J.
      Risk of thyroid carcinoma in a female population after radiotherapy for breast carcinoma.
      ,
      • Rutqvist L.E.
      • Johansson H.
      Long-term follow-up of the Stockholm randomized trials of postoperative radiation therapy versus adjuvant chemotherapy among ‘high risk’ pre- and postmenopausal breast cancer patients.
      ]. The reported mean and median follow-up time was 8 years based on 11 studies. Two studies provided no follow-up time [
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Mery C.M.
      • George S.
      • Bertagnolli M.M.
      • Raut C.P.
      Secondary sarcomas after radiotherapy for breast cancer: sustained risk and poor survival.
      ]. The descriptive characteristics of the included studies are listed in Table 1. The reported RRs with corresponding 95% CI of second cancers used in the meta-analysis according to cancer site along with the number of included patients and observed second cancers are listed in the appendix Supplementary Table 2.
      Table 1Clinical and baseline characteristics for the 13 studies included in the meta-analysis.
      Author, year & countryTreatment yearsData sourceNumber of patientsMean/median age at BC years (range)Mean/median Follow-up (years) & Inclusion from (years)CommentsSurgery and radiation dose (Gray)
      Andersson (1991)
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      Denmark
      1977–1982DBCG & The Danish Cancer Registry2692NR, includes only postmenopausal womenMedian 7.9

      Inclusion from BC diagnosis
      Mastectomy

      Dose: 36–41 Gy in 20–22 fractions
      Berrington de Gonzalez (2010)
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      USA
      1973–2000The SEER Registries182,057NR (<40–74)Mean 13

      Inclusion from 5 y after BC diagnosis
      Mastectomy & BCS

      No information on dose
      Grantzau (2013)
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      Denmark
      1982–2007DBCG & The Danish Cancer Registry46,176RT Median 55 (21–93) Non-RT Median 58 (21–96)Median RT 5/Non-RT 8

      Inclusion from 1 y after BC diagnosis
      Mastectomy & BCS

      Dose: 48–50 Gy in 24–25 fractions ± boost to tumor bed of 10–24 Gy
      Huang (2001)
      • Huang J.
      • Walker R.
      • Groome P.G.
      • Shelley W.
      • Mackillop W.J.
      Risk of thyroid carcinoma in a female population after radiotherapy for breast carcinoma.
      USA
      1973–1993The SEER Registries194,798NRMedian 6.8

      Inclusion from 2 y after BC diagnosis
      No information on surgery or radiation dose
      Kirova (2007)
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      France
      1981–1997Review of Institutional Records.16,705Median 55Median 10.5

      Inclusion from NR
      Mastectomy & BCS

      Mean dose 54.3 Gy ± boost to tumor bed of mean 16.4 Gy
      Mery (2009)
      • Mery C.M.
      • George S.
      • Bertagnolli M.M.
      • Raut C.P.
      Secondary sarcomas after radiotherapy for breast cancer: sustained risk and poor survival.
      USA
      1973–2003The SEER registries563,155Mean 60NR

      Inclusion from 1 y after BC diagnosis
      Includes first invasive breast cancer and DCISPartial/radical mastectomy or no surgery

      No information on dose
      Roychoudhuri (2004)
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      UK
      1961–2000The Thamsen Cancer Registry64,782NR (up to 85)NR

      Inclusion from BC diagnosis. RRs used in meta-analysis based on estimates from 1 y after BC diagnosis
      Women treated with adjuvant CTX were excluded in the studyNo information on surgery or radiation dose
      Rubino (2000)
      • Rubino C.
      • de V.F.
      • Diallo I
      • Shamsaldin A
      • Le MG
      Increased risk of second cancers following breast cancer: role of the initial treatment.
      France
      1973–1992Review of Institutional Medical Records4416Mean 55 (21–91)Mean 9.5

      Inclusion from 1 y after BC diagnosis
      18% of the RT treated received ovarian castration onlyMastectomy & BCS

      No information radiation dose
      Rutqvist (2006)
      • Rutqvist L.E.
      • Johansson H.
      Long-term follow-up of the Stockholm randomized trials of postoperative radiation therapy versus adjuvant chemotherapy among ‘high risk’ pre- and postmenopausal breast cancer patients.
      Sweden
      1976–1990Review of Medical Records & the Swedish Cancer Registry1226NRMedian 18.4

      Inclusion from BC diagnosis
      HRs estimated separately for Pre- & Postmenopausal womenMastectomy

      Dose: 46 Gy, 2 Gy per fraction over 4.5 weeks
      Schaapveld (2008)
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      Netherlands
      1989–20032 Dutch Cancer Registries CCCN & CCS58,068NRMedian 5.4

      Inclusion from NR
      HRs estimated separately for the age groups <50 y/>50 y at BC diagnosisMastectomy & BCS

      No information on dose
      Zablotska (2003)
      • Zablotska L.B.
      • Neugut A.I.
      Lung carcinoma after radiation therapy in women treated with lumpectomy or mastectomy for primary breast carcinoma.
      USA
      1973–1998The SEER Registries260,541Mastectomy mean RT 57/Non-RT 62

      BCS Mean RT 59/Non-RT 67
      Mastectomy mean RT 8.3/Non-RT 8.7

      BCS RT 5.6/Non-RT 4·8

      Inclusion from 6 mo after BC diagnosis
      HRs estimated separately for Mastectomy and BCSMastectomy & BCS

      No information on dose
      Zablotska (2005)
      • Zablotska L.B.
      • Chak A.
      • Das A.
      • Neugut A.I.
      Increased risk of squamous cell esophageal cancer after adjuvant radiation therapy for primary breast cancer.
      USA
      1973–2000The SEER Registries244,624Mastectomy mean RT 57/Non-RT 61

      BCS Mean RT 59/Non-RT 66
      Mastectomy mean RT 9.1/Non-RT 9.8

      BCS mean RT 7.3/Non-RT 6.1

      Inclusion from 6 mo after BC diagnosis
      HRs estimated separately for Mastectomy and BCSMastectomy & BCS

      No information on dose
      Zhang (2011)
      • Zhang W.
      • Becciolini A.
      • Biggeri A.
      • Pacini P.
      • Muirhead C.R.
      Second malignancies in breast cancer patients following radiotherapy: a study in Florence, Italy.
      Italy
      1965–1994Hospital Records from 1984 including the Tuscany Cancer Registry5,248Mean 55Mean 8 Inclusion from 2 y after BC diagnosisNo information on surgery

      Total dose up to 60 Gy in 2 Gy per fraction.
      Abbreviations: BC: Breast cancer; DBCG: Danish Breast Cancer Cooperative Group; NR: Not reported; SEER: Surveillance, Epidemiology, and End Results Program, BCS: Breast conserving surgery; RT: Radiotherapy; DCIS: Ductal carcinoma in situ; CTX: Chemotherapy; CCCN: Comprehensive Cancer Centers North-Netherlands, CCS: Cancer Center South; y: years; mo: months.

      All second non-breast cancers

      Seven studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Rubino C.
      • de V.F.
      • Diallo I
      • Shamsaldin A
      • Le MG
      Increased risk of second cancers following breast cancer: role of the initial treatment.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ,
      • Rutqvist L.E.
      • Johansson H.
      Long-term follow-up of the Stockholm randomized trials of postoperative radiation therapy versus adjuvant chemotherapy among ‘high risk’ pre- and postmenopausal breast cancer patients.
      ,
      • Zhang W.
      • Becciolini A.
      • Biggeri A.
      • Pacini P.
      • Muirhead C.R.
      Second malignancies in breast cancer patients following radiotherapy: a study in Florence, Italy.
      ] reported the endpoint of all second non-breast cancers. These studies included a total of 299,883 patients, of which 44% received radiotherapy for breast cancer whereas 56% did not. The treatment years ranged between 1954 and 2007. In the total pooled RR estimate, radiotherapy was significantly associated with an increased risk of second non-breast cancers with a RR of 1.22 (95% CI 1.06–1.41), based on five studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Rubino C.
      • de V.F.
      • Diallo I
      • Shamsaldin A
      • Le MG
      Increased risk of second cancers following breast cancer: role of the initial treatment.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ,
      • Rutqvist L.E.
      • Johansson H.
      Long-term follow-up of the Stockholm randomized trials of postoperative radiation therapy versus adjuvant chemotherapy among ‘high risk’ pre- and postmenopausal breast cancer patients.
      ] including 5465 incidences of second cancers (Fig. 2). The risk remained significantly elevated ⩾5 years after breast cancer diagnosis with a RR of 1.12 (95% CI 1.06–1.19), but non-significantly ⩾10 years after breast cancer diagnosis RR 1.20 (95% CI 0.88–1.64). These estimates were based on 10,517 (not reported in one study [
      • Zhang W.
      • Becciolini A.
      • Biggeri A.
      • Pacini P.
      • Muirhead C.R.
      Second malignancies in breast cancer patients following radiotherapy: a study in Florence, Italy.
      ]) and 790 second cancers, respectively. The total pooled RR estimate showed significant heterogeneity between study results, I2 = 66%.
      Figure thumbnail gr2
      Fig. 2Second non-breast cancer: Random effects meta-analysis of the relative risk (RR) of second non-breast cancer according to follow-up since primary breast cancer diagnosis among radiotherapy treated breast cancer patients compared to those not treated with radiotherapy. Pooled relative risks for second non-breast cancer after first breast cancer, comparing patients treated with radiotherapy to those not treated with radiotherapy. The RR is reported as a total pooled RR estimate and according to time since breast cancer diagnosis. Solid squares represent the RR for each study with corresponding 95% confidence intervals. The Diamonds represents the pooled estimate. Each study is weighted (%) by the inverse of its variance. In the two studies by Berrington de Gonzalez.et al.
      [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ]
      and Grantzau et al.
      [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ]
      risk estimates were solely based on second solid non-breast cancers, in the study by Zhang el al.
      [
      • Zhang W.
      • Becciolini A.
      • Biggeri A.
      • Pacini P.
      • Muirhead C.R.
      Second malignancies in breast cancer patients following radiotherapy: a study in Florence, Italy.
      ]
      non-melanoma skin cancers but not second leukemia were included as an outcome and in the study by Schaapveld et al.
      [
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ]
      meningiomas were not included as an outcome. Abbreviations: RT: Radiotherapy, ES: Estimate, y: years, Tot: Total, NR: Not reported. Data on second non-breast cancer by time since breast cancer diagnosis in the study by Grantzau et al.
      [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ]
      are previously unpublished.

      Risk of second cancer by organ site

      Second lung cancer

      Seven studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ,
      • Zablotska L.B.
      • Neugut A.I.
      Lung carcinoma after radiation therapy in women treated with lumpectomy or mastectomy for primary breast carcinoma.
      ] reported risk estimates on second lung cancer. They included a total of 631,021 patients, of which 40% received radiotherapy, whereas 60% did not. The patients were treated between 1961 and 2007. The total pooled RR estimate yielded a significant association between radiotherapy and second lung cancer with a RR of 1.23 (95% CI 1.07–1.43), based on six studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Andersson M.
      • Storm H.H.
      • Mouridsen H.T.
      Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
      ,
      • Zablotska L.B.
      • Neugut A.I.
      Lung carcinoma after radiation therapy in women treated with lumpectomy or mastectomy for primary breast carcinoma.
      ] and 4016 incidences of second lung cancer (Fig. 3). The risk of second lung cancer increased gradually by time following breast cancer diagnosis with pooled estimates of 1.39 (95% CI 1.28–1.51), 1.59 (95% CI 1.39–1.81), and 1.66 (95% CI 1.36–2.01) for the latency periods ⩾5, ⩾10 and ⩾15 years after breast cancer diagnosis, respectively. These risk estimates were based on 2899, 1128, and 512 second lung cancers, respectively.
      Figure thumbnail gr3
      Fig. 3Second lung cancer: Random effects meta-analysis of the relative risk (RR) of second lung cancer according to follow-up since primary breast cancer diagnosis among radiotherapy treated breast cancer patients compared to those not treated with radiotherapy. Pooled relative risks for second lung cancer after first breast cancer, comparing patients treated with radiotherapy to those not treated with radiotherapy. The RR is reported as a total pooled RR estimate and according to time since breast cancer diagnosis. Solid squares represent the RR for each study with corresponding 95% confidence intervals. The Diamonds represent the pooled estimate. Each study is weighted (%) by the inverse of its variance. Abbreviations: RT: Radiotherapy, ES: Estimate, Mast: Mastectomy, BCS: Breast Conserving Surgery, y: years, Tot: Total, NR: Not reported. Data on second lung cancer by time since breast cancer diagnosis in the study by Grantzau et al.
      [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ]
      are previously unpublished.

      Second esophageal cancer

      Four studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Zablotska L.B.
      • Chak A.
      • Das A.
      • Neugut A.I.
      Increased risk of squamous cell esophageal cancer after adjuvant radiation therapy for primary breast cancer.
      ] reported the endpoint second esophagus cancer (Fig. 4). The studies included a total of 413,650 patients, of which 39% were treated with radiotherapy, whereas 61% was not. The treatment years ranged between 1961 and 2007. The risk of second esophagus was non-significantly increased in the total pooled RR estimate, with a RR of 1.17 (95% CI 0.89–1.54). Five or more years following breast cancer diagnosis, radiotherapy however became significantly associated with second esophagus cancer and furthermore increased by each 5 year increment in time. Thus 5, 10 and 15 or more years after breast cancer diagnosis the RRs were 1.53 (95% CI 1.01–2.31), based on 255 second cancers, 1.56 (95% CI 1.03–2.38) based on 147 second cancers and 2.17 (95% CI 1.11–4.25) based on 46 second cancers, respectively.
      Figure thumbnail gr4
      Fig. 4Second esophagus cancer: Random effects meta-analysis of the relative risk (RR) of second esophageal cancer according to follow-up since primary breast cancer diagnosis among radiotherapy treated breast cancer patients compared to those not treated with radiotherapy. Pooled relative risks for second esophagus cancer after first breast cancer, comparing patients treated with radiotherapy to those not treated with radiotherapy. The RR is reported as a total pooled RR estimate and according to time since breast cancer diagnosis. Solid squares represent the RR for each study with corresponding 95% confidence intervals. The Diamonds represent the pooled estimate. Each study is weighted (%) by the inverse of its variance. Abbreviations: RT: Radiotherapy, ES: Estimate, Mast: Mastectomy, BCS: Breast Conserving Surgery, y: years, Tot: Total, NR: Not reported. Data on second esophagus cancer by time since treatment in the study by Grantzau et al.
      [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ]
      are previously unpublished.

      Second thyroid cancer

      For the endpoint second thyroid cancer four studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Roychoudhuri R.
      • Evans H.
      • Robinson D.
      • Moller H.
      Radiation-induced malignancies following radiotherapy for breast cancer.
      ,
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Huang J.
      • Walker R.
      • Groome P.G.
      • Shelley W.
      • Mackillop W.J.
      Risk of thyroid carcinoma in a female population after radiotherapy for breast carcinoma.
      ] were included, comprising a total of 322,461 breast cancer patients (Fig. 5). Patients were treated in the years between 1961 and 2007 and radiotherapy was given to 37% of the patients, whereas 63% received no radiotherapy. There was no significant association between radiotherapy and second thyroid cancer neither in the total pooled RR estimate, RR 1.05 (95% CI 0.78–1.43) based on 188 second cancers (unknown number in one study [
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ]), nor over time. The RRs for the latency periods ⩾5, ⩾10 and ⩾15 years following breast cancer diagnosis were 0.96 (95% CI 0.59–1.57), 1.53 (95% CI 0.69–3.39), and 2.21 (95% CI 0.64–7.61), respectively based 113, 42, and 15 incidences of second thyroid cancers, respectively.
      Figure thumbnail gr5
      Fig. 5Second thyroid cancer: Random effects meta-analysis of the relative risk (RR) of second thyroid cancer according to follow-up since primary breast cancer diagnosis among radiotherapy treated breast cancer patients compared to those not treated with radiotherapy. Pooled relative risks for second thyroid cancer after first breast cancer, comparing patients treated with radiotherapy to those not treated with radiotherapy. The RR is reported as a total pooled RR estimate and according to time since breast cancer diagnosis. Solid squares represent the RR for each study with corresponding 95% confidence intervals. The Diamonds represent the pooled estimate. Each study is weighted (%) by the inverse of its variance. Abbreviations: RT: Radiotherapy, ES: Estimate, NR: Not reported. Data on second thyroid cancer by time since treatment in the study by Grantzau et al.
      [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ]
      are previously unpublished.

      Second sarcomas

      Four studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Schaapveld M.
      • Visser O.
      • Louwman M.J.
      • et al.
      Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
      ,
      • Kirova Y.M.
      • Gambotti L.
      • De R.Y.
      • Vilcoq J.R.
      • Asselain B.
      • Fourquet A.
      Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
      ,
      • Mery C.M.
      • George S.
      • Bertagnolli M.M.
      • Raut C.P.
      Secondary sarcomas after radiotherapy for breast cancer: sustained risk and poor survival.
      ] on second sarcomas were included in the analysis for the total pooled RR estimate. The studies comprised a total of 684,104 patients treated between 1973 and 2007 (Fig. 6). Radiotherapy was given to 41% of the patients whereas 59% received no radiotherapy. In the total pooled RR estimate the association between radiotherapy and second sarcomas was significantly and more than two fold increased with a RR of 2.41 (95% CI 1.41–4.13), based on 1048 incidence cases. The risk remained significantly elevated ⩾5 years after breast cancer diagnosis with a RR of 2.53 (95% CI 1.74–3.70) based on 124 incidences of second sarcomas. The reported risk estimates were based on all second sarcomas that were observed in the cohorts, regardless of the anatomical location of these second cancers. Two studies, based on data from the SEER database, also estimated the risk specifically for angiosarcomas with reported RRs of 7.63 (95% CI 4.9–11.9) [
      • Mery C.M.
      • George S.
      • Bertagnolli M.M.
      • Raut C.P.
      Secondary sarcomas after radiotherapy for breast cancer: sustained risk and poor survival.
      ] among +1-year survivors and 13.7 (95% CI 4.0–95.6) [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ] among +5-year survivors of breast cancer (see appendix Supplementary Table 2).
      Figure thumbnail gr6
      Fig. 6Second sarcomas: Random effects meta-analysis of the relative risk (RR) of second sarcomas according to follow-up since primary breast cancer diagnosis among radiotherapy treated breast cancer patients compared to those not treated with radiotherapy. Pooled relative risks for second cancer of the soft tissues after first breast cancer, comparing patients treated with radiotherapy to those not treated with radiotherapy. The RR is reported as a total pooled RR estimate and according to time since breast cancer diagnosis. Solid squares represent the RR for each study with corresponding 95% confidence intervals. The Diamonds represent the pooled estimate. Each study is weighted (%) by the inverse of its variance. Abbreviations: RT: Radiotherapy, ES: Estimate, Mast: Mastectomy, BCS: Breast Conserving Surgery, y: years, Tot: Total, NR: Not reported. Data on second cancers of the soft tissues by time since treatment in the study by Grantzau et al.
      [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ]
      are previously unpublished data.

      Second cancer risk by breast cancer surgery type

      In the study by Berrington de Gonzalez et al. [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ], the risk according to breast cancer surgery type, was estimated for the combined group of second cancers potentially receiving the highest radiation dose, i.e. the lung, esophagus, pleura, bone, and soft tissues. We also estimated the risk according to surgery type in the Danish cohort study [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ] by applying the same combined group of second cancers as Berrington de Gonzalez et al. [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ] For patients diagnosed with one of these second cancers five or more years after breast cancer diagnosis the pooled RRs including both studies [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ,
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ] were significantly increased both after a mastectomy RR 1.64 (95% CI 1.42–1.89) and after a lumpectomy RR 1.29 (95% CI 1.16–1.43). See appendix Supplementary Fig. 4.

      Discussion

      In this meta-analysis of female early breast cancer patients, radiotherapy was significantly associated with and increased risk of second non-breast cancers. The strength of this study is its large size, including more than 700,000 patients. The study was furthermore conducted flowering the PRISMA [

      PRISMA. Transparent reporting of systematic reviews and META-analyses. Available from: URL: <http://www.prima-statement.org/statement.htm>, 2014.

      ] guidelines of systematic reviews and meta-analysis providing a reliable and comprehensive assessment of second cancer risk after radiotherapy for early breast cancer. For patients with a latency time of at least five years, from breast cancer diagnosis to second cancer diagnosis, radiotherapy was significantly associated with an increased risk of second non-breast cancers overall and specifically for second cancers of the lung, esophagus, and second sarcomas. For these second cancers the risk increased gradually by each five year increment in time since breast cancer diagnosis. There was no significant association between radiotherapy and second thyroid cancer.
      The 2005 EBCTCG overview of randomized trials is currently the only other study reporting pooled estimates on second cancer risk after radiotherapy for early breast cancer [
      • Clarke M.
      • Collins R.
      • Darby S.
      • et al.
      Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials.
      ]. This overview was based on trials that began in the mid 1970s and included a total of 31,800 female breast cancer patients and 1351 second cancers. The annual risk ratios comparing irradiated women with unirradiated women were significantly increased for second lung cancer (1.61, p < 0.001) and second sarcoma (2.34, p = 0.03), but only borderline significant for second esophagus cancer (2.06, p = 0.05). There was no excess risk of second thyroid cancer 0.69 (p = 0.4). The results were in line with our total pooled estimates. The above results of the EBCTCG overview were not included in this meta-analysis due to some degree of overlap between patients of the overview and the included cohorts of this study and furthermore as estimates in the EBCTCG were reported as annual risk ratios.
      We found no significant association between radiotherapy for breast cancer and second thyroid cancer, neither in the total estimate, nor over time since breast cancer diagnosis. Increased risks of second thyroid cancer have predominantly been reported in patients exposed to radiotherapy during childhood [
      • de V.F.
      • Hardiman C.
      • Shamsaldin A.
      • et al.
      Thyroid carcinomas after irradiation for a first cancer during childhood.
      ,
      • Veiga L.H.
      • Lubin J.H.
      • Anderson H.
      • et al.
      A pooled analysis of thyroid cancer incidence following radiotherapy for childhood cancer.
      ]. In a pooled analysis of 275,263 childhood cancer survivors, radiotherapy was significantly associated with an increased risk of second thyroid cancer and furthermore increased with decreasing age at exposure [
      • Veiga L.H.
      • Lubin J.H.
      • Anderson H.
      • et al.
      A pooled analysis of thyroid cancer incidence following radiotherapy for childhood cancer.
      ]. It therefore seems that mainly children carry a risk of radiation induced thyroid cancers and if any risk persists among adult breast cancer patients, it is too low to be detectable in a cohort of over 300,000 women.
      In the two studies by Zablotska et al. [
      • Zablotska L.B.
      • Neugut A.I.
      Lung carcinoma after radiation therapy in women treated with lumpectomy or mastectomy for primary breast carcinoma.
      ,
      • Zablotska L.B.
      • Chak A.
      • Das A.
      • Neugut A.I.
      Increased risk of squamous cell esophageal cancer after adjuvant radiation therapy for primary breast cancer.
      ], included in the current analysis, the risk of second lung and esophagus cancer was reported separately for patients treated with either post-mastectomy or post-lumpectomy radiotherapy. The studies were based on data from the SEER database and covered the treatment years 1973–1998/2000. In both studies only radiotherapy after mastectomy was significantly associated with increased risks. Notably though, as breast conserving surgery first became common during the mid 1990s [
      • Shank B.
      • Moughan J.
      • Owen J.
      • Wilson F.
      • Hanks G.E.
      The 1993–94 patterns of care process survey for breast irradiation after breast-conserving surgery-comparison with the 1992 standard for breast conservation treatment. The Patterns of Care Study, American College of Radiology.
      ], the time at risk is much shorter for patients treated with a lumpectomy compared with patients treated with a mastectomy. It is therefore possible, that the time at risk was too short in the lumpectomy group to show significantly increased risks. Conversely, Berrington de Gonzalez et al. [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ] found a significantly increased risk of second cancers in patients treated with radiotherapy after both lumpectomy and mastectomy, based on the same cohort and follow-up period, see appendix Supplementary Fig. 4. The risk estimates were given for the combined group of second cancers of the lung, esophagus, pleura, bone, and soft tissues and the comparison group included all unirradiated patients regardless of the type of breast cancer surgery. Using this combined group of second cancers, we further showed in the pooled estimates, including the study by Berrington de Gonzalez et al. [
      • Berrington de Gonzalez A.
      • Curtis R.E.
      • Gilbert E.
      • et al.
      Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
      ] and the study by Grantzau et al. [
      • Grantzau T.
      • Mellemkjaer L.
      • Overgaard J.
      Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
      ], that there is evidence of a significantly increased risk after both post-lumpectomy and post-mastectomy radiotherapy. It therefore appears that radiotherapy to the conserved breast, like radiotherapy after mastectomy, carries a risk of second cancer induction.
      As this meta-analysis was based on summary results and not on individual patient data, potential methodological limitations in our analysis are primarily inherited from the design of the included studies, such as, duration of observation-time and study populations. Studies were further not standardized as to radiation treatment techniques or systemic adjuvant treatment. However, all included studies came from European and North American institutions, thus the included cohorts were fairly compatible according to applied diagnostic criteria’s, adjuvant treatments and western lifestyle and therefore judged sufficiently similar to allow pooling of data. Information on smoking status, alcohol consumption, BMI, genetic factors, and other potential confounders that could affect second cancer risks was not available in any of the studies. It is well known that smoking is the leading cause of lung cancer [
      • Pesch B.
      • Kendzia B.
      • Gustavsson P.
      • et al.
      Cigarette smoking and lung cancer–relative risk estimates for the major histological types from a pooled analysis of case-control studies.
      ] and in the combination with heavy alcohol consumption the primary cause of squamous-cell carcinoma of the esophagus [
      • Castellsague X.
      • Munoz N.
      • De S.E.
      • et al.
      Independent and joint effects of tobacco smoking and alcohol drinking on the risk of esophageal cancer in men and women.
      ]. Increased risks of esophagus cancer have also been linked to obesity [
      • Polednak A.P.
      Estimating the number of U.S. incident cancers attributable to obesity and the impact on temporal trends in incidence rates for obesity-related cancers.
      ]. However, it is typically not possible to adjust for these confounders in large cohort studies. Not all second tumors in the studies were histopathologically confirmed and it is possible that some of these cancers were misclassified metastatic breast cancers. Though, if of significance this would probably primarily affect the total pooled risk estimates and would diminish over time following breast cancer diagnosis.
      The level of heterogeneity for each analysis was often either low or moderate. As the overall study period covered more than 50 years and included patients from several countries, some differences between studies in regard to radiotherapy techniques will inevitably exist, and would explain some of the observed in-between study heterogeneity. Substantial heterogeneity was however only observed in the total pooled RR estimate of all second non-breast cancers. This was most likely due to, that not all studies reported the outcome for all second non-breast cancer uniformly (see Fig. 2. footnote). It should also be noted, that only few studies reported data on second cancers developing ten or more years after breast cancer diagnosis, hence the pooled risk estimates were only based on two to three studies. However, with increasing time since breast cancer diagnosis the risk estimates became remarkably similar.
      The dose–response relation for second solid cancer induction after radiotherapy for breast cancer is currently only examined in few epidemiological studies. We have recently shown, that second lung cancer after radiotherapy for breast cancer is associated with the delivered dose to the lung [
      • Grantzau T.
      • Thomsen M.S.
      • Vaeth M.
      • Overgaard J.
      Risk of second primary lung cancer in women after radiotherapy for breast cancer.
      ]. The rate of lung cancer increased linearly with 8·5% per delivered Gray to the lung (95% CI 3.1% to 23.3%) among female ⩾5 year survivors. In a similar study, on second esophageal cancer, Morton et al. [
      • Morton L.M.
      • Gilbert E.S.
      • Hall P.
      • et al.
      Risk of treatment-related esophageal cancer among breast cancer survivors.
      ] showed, that the excess odds ratio of second esophageal cancer also increased linearly with 9% per delivered Gray to the tumor location (95% CI 4% to 16%) among ⩾5 year survivors of female breast cancer. Together, these epidemiological data indicate, that any dose reduction to the surrounding normal tissue would result in a beneficial risk reduction of second cancers in organs adjacent to the radiation fields.
      During the last 50 years breast cancer treatment has changed considerably. This includes changes in surgical procedures, more extensive use of systemic treatment, and changes in radiotherapy techniques including the change from two dimensional (2D) treatment planning to the currently used CT-guided techniques. Mastectomy has, since the mid 90’s, largely been replaced by breast conserving surgery resulting in an increasing number of patients receiving adjuvant radiotherapy. Whereas orthovoltage X-ray and Co60 have now been abandoned, in favor of photon and electron beams, the total breast dose has remained relatively constant. Over the last ten years though, an increasing number of new radiation therapy techniques, such as intensity modulated radiation therapy (IMRT), volumetric modulated arc therapy (VMAT) and intraoperative radiotherapy (IORT) have rapidly entered clinical practice. Due to insufficient follow-up time, no cohort studies have yet estimated the risk of second cancer using these more novel treatment techniques, and it therefore remains uncertain what risk these methods carry. Thus it is important to remember, that risk estimations related to these techniques are only based on theoretical models. While IMRT and VMAT have advantages in providing better target coverage and dose conformity as to conventional three-dimensional conformal radiotherapy (3D-CRT), these newer techniques have been shown to have greater out-of-beam doses [
      • Hall E.J.
      • Wuu C.S.
      Radiation-induced second cancers: the impact of 3D-CRT and IMRT.
      ,
      • Lee B.
      • Lee S.
      • Sung J.
      • Yoon M.
      Radiotherapy-induced secondary cancer risk for breast cancer: 3D conformal therapy versus IMRT versus VMAT.
      ]. Compared with conventional 3D-CRT, IMRT both uses more treatment fields and these modulated fields furthermore produce more radiation leakage, combined leading to a larger volume of normal tissue exposed to low dose radiation. Due to these circumstances, it has been estimated by Hall et al., that IMRT would almost double the incidence of second cancers compared to conventional 3D-CRT [
      • Hall E.J.
      • Wuu C.S.
      Radiation-induced second cancers: the impact of 3D-CRT and IMRT.
      ].
      Accelerated partial breast irradiation (APBI) given as intraoperative radiotherapy (IORT), is another technique that recently has emerged as a new treatment strategy after lumpectomy, to low risk breast cancer patients. Compared to conventional whole breast irradiation IORT reduces the treatment time, the irradiation field and the maximal doses to the surrounding normal tissues [
      • Aziz M.H.
      • Schneider F.
      • Clausen S.
      • et al.
      Can the risk of secondary cancer induction after breast conserving therapy be reduced using intraoperative radiotherapy (IORT) with low-energy x-rays?.
      ]. The 5-year results of the IORT trials ELOIT and TARGIT-A have just been released [
      • Veronesi U.
      • Orecchia R.
      • Maisonneuve P.
      • et al.
      Intraoperative radiotherapy versus external radiotherapy for early breast cancer (ELIOT): a randomised controlled equivalence trial.
      ,
      • Vaidya J.S.
      • Wenz F.
      • Bulsara M.
      • et al.
      Risk-adapted targeted intraoperative radiotherapy versus whole-breast radiotherapy for breast cancer: 5-year results for local control and overall survival from the TARGIT - a randomised trial.
      ]. Both studies found a significantly higher local recurrence rate after IORT compared with whole breast irradiation. The TARGIT-A investigators however, argued that IORT reduced both cardiovascular and second cancer mortality compared to whole breast irradiation [
      • Vaidya J.S.
      • Wenz F.
      • Bulsara M.
      • et al.
      Risk-adapted targeted intraoperative radiotherapy versus whole-breast radiotherapy for breast cancer: 5-year results for local control and overall survival from the TARGIT - a randomised trial.
      ]. With a median follow-up time of only five years these conclusions are however premature and IORT remains an experimental approach in early breast cancer treatment.
      In conclusion, radiotherapy for breast cancer is associated with a small but significantly increased risk of second cancers of the lung, esophagus, and soft tissues. There was significant association between radiotherapy and second thyroid cancer. Adjuvant radiotherapy continues to be an essential part of breast cancer treatment and the benefits in regard to local tumor control and consequently improved overall survival by far outweigh the risk of getting a radiation induced second cancer 10–15 years after treatment. Thus, the emerging challenge in breast cancer treatment is to optimize current and evolving radiotherapy techniques by reducing the normal tissue dose without compromising the target, and to improve the identification of patients that are most, or least, likely to benefit from adjuvant radiotherapy.

      Conflict of interest statement

      The authors declare no conflict of interest.

      Funding

      CIRRO ( The Lundbeck Foundation Centre for Interventional Research in Radiation Oncology ), Aarhus University ; Faculty of Health Sciences, the Danish Medical Research Council , the Danish Cancer Society and from the European Atomic Energy Community’s Seventh Framework Program (7/2007–2013) under grant agreement No. 231965 ; ALLEGRO.

      Appendix A. Supplementary data

      Reference

      1. GLOBOCAN. Available from: URL: http://globocan.iarc.fr; 2014.

        • Forouzanfar M.H.
        • Foreman K.J.
        • Delossantos A.M.
        • et al.
        Breast and cervical cancer in 187 countries between 1980 and 2010: a systematic analysis.
        Lancet. 2011; 378: 1461-1484
        • Overgaard M.
        • Hansen P.S.
        • Overgaard J.
        • et al.
        Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial.
        N Engl J Med. 1997; 337: 949-955
        • Overgaard M.
        • Jensen M.B.
        • Overgaard J.
        • et al.
        Postoperative radiotherapy in high-risk postmenopausal breast-cancer patients given adjuvant tamoxifen: Danish Breast Cancer Cooperative Group DBCG 82c randomised trial.
        Lancet. 1999; 353: 1641-1648
        • Clarke M.
        • Collins R.
        • Darby S.
        • et al.
        Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials.
        Lancet. 2005; 366: 2087-2106
        • Darby S.
        • McGale P.
        • Correa C.
        • et al.
        Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials.
        Lancet. 2011; 378: 1707-1716
        • Ebctcg Early Breast Cancer Trialists’ Collaborative Group
        Effect of radiotherapy after mastectomy and axillary surgery on 10-year recurrence and 20-year breast cancer mortality: meta-analysis of individual patient data for 8135 women in 22 randomised trials.
        Lancet. 2014; 383: 2127-2135
        • Tubiana M.
        Can we reduce the incidence of second primary malignancies occurring after radiotherapy? A critical review.
        Radiother Oncol. 2009; 91: 4-15
        • Grantzau T.
        • Mellemkjaer L.
        • Overgaard J.
        Second primary cancers after adjuvant radiotherapy in early breast cancer patients: a national population based study under the Danish Breast Cancer Cooperative Group (DBCG).
        Radiother Oncol. 2013; 106: 42-49
        • Berrington de Gonzalez A.
        • Curtis R.E.
        • Gilbert E.
        • et al.
        Second solid cancers after radiotherapy for breast cancer in SEER cancer registries.
        Br J Cancer. 2010; 102: 220-226
        • Roychoudhuri R.
        • Evans H.
        • Robinson D.
        • Moller H.
        Radiation-induced malignancies following radiotherapy for breast cancer.
        Br J Cancer. 2004; 91: 868-872
        • Schaapveld M.
        • Visser O.
        • Louwman M.J.
        • et al.
        Risk of new primary nonbreast cancers after breast cancer treatment: a Dutch population-based study.
        J Clin Oncol. 2008; 26: 1239-1246
        • Kirova Y.M.
        • Gambotti L.
        • De R.Y.
        • Vilcoq J.R.
        • Asselain B.
        • Fourquet A.
        Risk of second malignancies after adjuvant radiotherapy for breast cancer: a large-scale, single-institution review.
        Int J Radiat Oncol Biol Phys. 2007; 68: 359-363
        • Rubino C.
        • de V.F.
        • Diallo I
        • Shamsaldin A
        • Le MG
        Increased risk of second cancers following breast cancer: role of the initial treatment.
        Breast Cancer Res Treat. 2000; 61: 183-195
      2. PRISMA. Transparent reporting of systematic reviews and META-analyses. Available from: URL: <http://www.prima-statement.org/statement.htm>, 2014.

      3. Meta-analysis in Stata: an updated collection from the Stata Journal.
        Stata Press Publication, 4905 Lakeway Drive, College Station, texas 77845, US2009
        • Levi F.
        • Randimbison L.
        • Te V.C.
        • La V.C.
        Cancer risk after radiotherapy for breast cancer.
        Br J Cancer. 2006; 95: 390-392
        • Andersson M.
        • Jensen M.B.
        • Engholm G.
        • Henrik S.H.
        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.
        Acta Oncol. 2008; 47: 755-764
        • Neugut A.I.
        • Robinson E.
        • Lee W.C.
        • Murray T.
        • Karwoski K.
        • Kutcher G.J.
        Lung cancer after radiation therapy for breast cancer.
        Cancer. 1993; 71: 3054-3057
        • Yu G.P.
        • Schantz S.P.
        • Neugut A.I.
        • Zhang Z.F.
        Incidences and trends of second cancers in female breast cancer patients: a fixed inception cohort-based analysis (United States).
        Cancer Causes Control. 2006; 17: 411-420
        • Huang J.
        • Mackillop W.J.
        Increased risk of soft tissue sarcoma after radiotherapy in women with breast carcinoma.
        Cancer. 2001; 92: 172-180
        • Kirova Y.M.
        • Vilcoq J.R.
        • Asselain B.
        • Sastre-Garau X.
        • Fourquet A.
        Radiation-induced sarcomas after radiotherapy for breast carcinoma: a large-scale single-institution review.
        Cancer. 2005; 104: 856-863
        • Andersson M.
        • Storm H.H.
        • Mouridsen H.T.
        Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer.
        J Natl Cancer Inst. 1991; 83: 1013-1017
        • Huang J.
        • Walker R.
        • Groome P.G.
        • Shelley W.
        • Mackillop W.J.
        Risk of thyroid carcinoma in a female population after radiotherapy for breast carcinoma.
        Cancer. 2001; 92: 1411-1418
        • Mery C.M.
        • George S.
        • Bertagnolli M.M.
        • Raut C.P.
        Secondary sarcomas after radiotherapy for breast cancer: sustained risk and poor survival.
        Cancer. 2009; 115: 4055-4063
        • Rutqvist L.E.
        • Johansson H.
        Long-term follow-up of the Stockholm randomized trials of postoperative radiation therapy versus adjuvant chemotherapy among ‘high risk’ pre- and postmenopausal breast cancer patients.
        Acta Oncol. 2006; 45: 517-527
        • Zablotska L.B.
        • Neugut A.I.
        Lung carcinoma after radiation therapy in women treated with lumpectomy or mastectomy for primary breast carcinoma.
        Cancer. 2003; 97: 1404-1411
        • Zablotska L.B.
        • Chak A.
        • Das A.
        • Neugut A.I.
        Increased risk of squamous cell esophageal cancer after adjuvant radiation therapy for primary breast cancer.
        Am J Epidemiol. 2005; 161: 330-337
        • Zhang W.
        • Becciolini A.
        • Biggeri A.
        • Pacini P.
        • Muirhead C.R.
        Second malignancies in breast cancer patients following radiotherapy: a study in Florence, Italy.
        Breast Cancer Res. 2011; 13: R38
        • de V.F.
        • Hardiman C.
        • Shamsaldin A.
        • et al.
        Thyroid carcinomas after irradiation for a first cancer during childhood.
        Arch Intern Med. 1999; 159: 2713-2719
        • Veiga L.H.
        • Lubin J.H.
        • Anderson H.
        • et al.
        A pooled analysis of thyroid cancer incidence following radiotherapy for childhood cancer.
        Radiat Res. 2012; 178: 365-376
        • Shank B.
        • Moughan J.
        • Owen J.
        • Wilson F.
        • Hanks G.E.
        The 1993–94 patterns of care process survey for breast irradiation after breast-conserving surgery-comparison with the 1992 standard for breast conservation treatment. The Patterns of Care Study, American College of Radiology.
        Int J Radiat Oncol Biol Phys. 2000; 48: 1291-1299
        • Pesch B.
        • Kendzia B.
        • Gustavsson P.
        • et al.
        Cigarette smoking and lung cancer–relative risk estimates for the major histological types from a pooled analysis of case-control studies.
        Int J Cancer. 2012; 131: 1210-1219
        • Castellsague X.
        • Munoz N.
        • De S.E.
        • et al.
        Independent and joint effects of tobacco smoking and alcohol drinking on the risk of esophageal cancer in men and women.
        Int J Cancer. 1999; 82: 657-664
        • Polednak A.P.
        Estimating the number of U.S. incident cancers attributable to obesity and the impact on temporal trends in incidence rates for obesity-related cancers.
        Cancer Detect Prev. 2008; 32: 190-199
        • Grantzau T.
        • Thomsen M.S.
        • Vaeth M.
        • Overgaard J.
        Risk of second primary lung cancer in women after radiotherapy for breast cancer.
        Radiother Oncol. 2014; 111: 366-373
        • Morton L.M.
        • Gilbert E.S.
        • Hall P.
        • et al.
        Risk of treatment-related esophageal cancer among breast cancer survivors.
        Ann Oncol. 2012; 23: 3081-3091
        • Hall E.J.
        • Wuu C.S.
        Radiation-induced second cancers: the impact of 3D-CRT and IMRT.
        Int J Radiat Oncol Biol Phys. 2003; 56: 83-88
        • Lee B.
        • Lee S.
        • Sung J.
        • Yoon M.
        Radiotherapy-induced secondary cancer risk for breast cancer: 3D conformal therapy versus IMRT versus VMAT.
        J Radiol Prot. 2014; 34: 325-331
        • Aziz M.H.
        • Schneider F.
        • Clausen S.
        • et al.
        Can the risk of secondary cancer induction after breast conserving therapy be reduced using intraoperative radiotherapy (IORT) with low-energy x-rays?.
        Radiat Oncol. 2011; 6: 174
        • Veronesi U.
        • Orecchia R.
        • Maisonneuve P.
        • et al.
        Intraoperative radiotherapy versus external radiotherapy for early breast cancer (ELIOT): a randomised controlled equivalence trial.
        Lancet Oncol. 2013; 14: 1269-1277
        • Vaidya J.S.
        • Wenz F.
        • Bulsara M.
        • et al.
        Risk-adapted targeted intraoperative radiotherapy versus whole-breast radiotherapy for breast cancer: 5-year results for local control and overall survival from the TARGIT - a randomised trial.
        Lancet. 2014; 383: 603-613