Elsevier

The Lancet Oncology

Volume 6, Issue 8, August 2005, Pages 557-565
The Lancet Oncology

Fast track — Articles
Long-term mortality from heart disease and lung cancer after radiotherapy for early breast cancer: prospective cohort study of about 300 000 women in US SEER cancer registries

https://doi.org/10.1016/S1470-2045(05)70251-5Get rights and content

Summary

Background

Radiotherapy for early breast cancer can decrease breast cancer mortality but increase other mortality, mainly from heart disease and lung cancer. The mean cardiac dose from irradiation of a left-sided breast cancer can be two or three times that for a right-sided breast cancer. The mean ipsilateral (ie, on the same side as the breast cancer) lung dose can also be two or three times the mean contralateral lung dose. Particularly during the 1970s, when typical heart and lung exposures were greater than now, the laterality of an irradiated breast cancer could measurably affect cardiac mortality and mortality from cancer of the right or the left lung decades later. This study aimed to assess the hazards in the general US population from routine cancer-registry and death-certificate data.

Methods

We analysed data for 308 861 US women with early breast cancer of known laterality (left-sided or right-sided) who were registered in the US Surveillance Epidemiology and End Results (SEER) cancer registries during 1973–2001 and followed prospectively for cause-specific mortality until Jan 1, 2002.

Findings

115 165 (37%) received radiotherapy. Among those who did not, tumour laterality was of little relevance to subsequent mortality. For women diagnosed during 1973–82 and irradiated, the cardiac mortality ratio (left versus right tumour laterality) was 1·20 (95% CI 1·04–1·38) less than 10 years afterwards, 1·42 (1·11–1·82) 10–14 years afterwards, and 1·58 (1·29–1·95) after 15 years or more (trend: 2p=0·03). For women diagnosed during 1983–92 and irradiated, the cardiac mortality ratio was 1·04 (0·91–1·18) less than 10 years afterwards and 1·27 (0·99–1·63) 10 or more years afterwards. For women diagnosed during 1993–2001 and irradiated the cardiac mortality ratio was 0·96 (0·82–1·12), with none yet followed for 10 years. Among women irradiated for breast cancer who subsequently developed an ipsilateral or contralateral lung cancer, the lung cancer mortality ratio (ipsilateral versus contralateral) for women diagnosed during 1973–82 and irradiated was 1·17 (0·62–2·19), 2·00 (1·00–4·00), and 2·71 (1·65–4·48), respectively, less than 10 years, 10–14 years, and 15 or more years afterwards (trend: 2p=0·04). For women irradiated after 1982 there is, as yet, little information on lung cancer risks more than 10 years afterwards.

Interpretation

US breast cancer radiotherapy regimens of the 1970s and early 1980s appreciably increased mortality from heart disease and lung cancer 10–20 years afterwards with, as yet, little direct evidence on the hazards after more than 20 years. Since the early 1980s, improvements in radiotherapy planning should have reduced such risks, but the long-term hazards in the general populations of various countries still need to be monitored directly.

Introduction

Although radiotherapy for early breast cancer can reduce the risk of death from the disease several years later, it usually involves some irradiation of the heart or lungs, and some of the breast cancer radiotherapy regimens used during previous decades could increase the risks of heart disease and lung cancer many years after treatment.1, 2, 3, 4 Regimens that incidentally irradiate the heart will generally do so to a greater extent when used to treat left-sided breast cancer than when used to treat right-sided breast cancer. Likewise, those that incidentally irradiate the lung will generally affect the ipsilateral lung (ie, on the same side as the breast cancer) more than the contralateral lung. These differences in exposure can be substantial: for example, the mean dose to the heart or to one of the lungs may differ by a factor of two or three, depending on the laterality (ie, left or right) of the breast tumour.5, 6 Some indication of the extent to which the radiotherapy regimens of the 1970s or early 1980s were causing mortality from heart disease and lung cancer 10–20 years later in the general population of patients with breast cancer can, therefore, be obtained from routine cancer-registry data that recorded the use or otherwise of radiotherapy for any breast cancers, the laterality of those breast cancers and of any subsequent primary lung cancers, and the certified causes of any deaths during the first and, particularly, subsequent decades after treatment.

The US Surveillance Epidemiology and End Results (SEER) cancer registries have, since 1973, covered in this way a substantial proportion of the US population. Increases in heart disease7, 8 and lung cancer9, 10, 11 after radiotherapy for breast cancer have been reported in the SEER data, and the 20-year hazards of the breast cancer radiotherapy regimens used in the 1970s and early 1980s are now becoming clear. However, it is as yet less clear how much risk there will be from the regimens used since the early 1980s, when the majority of women with early breast cancer started to receive breast-conserving surgery (BCS) plus radiotherapy, often just to the breast.

Although cardiac doses and lung doses from radiotherapy have decreased since the 1970s,12, 13 some breast cancer radiotherapy techniques used during the 1980s would still have involved mean doses to the heart and ipsilateral lung of several Gray (Gy), so some eventual risk might still be expected. Atomic bomb survivors exposed to single doses of up to 4 Gy had dose-related excess risks decades later of about 17% per Gy for heart disease and 90% per Gy for lung cancer.14, 15 Likewise, in patients given fractionated radiotherapy for peptic ulcer disease with mean cardiac doses of up to 4 Gy there was a dose-related increase in late cardiac mortality.16

In atomic bomb survivors and in patients with peptic ulcers, the excess risks did not become clear until more than 10 years after exposure. Thus, a possible explanation for the lack of any definite hazard from the US regimens of the 1980s in previous analyses of SEER data (some of which truncated data after only 12 years to compare the 1980s with the 1970s without bias)8 is that the follow-up is not yet long enough.17 We have therefore reanalysed the risks of heart disease and lung cancer in the latest data from the SEER cancer registries, with particular emphasis on the periods 10–14 years and 15 or more years after exposure.

Section snippets

Materials

The SEER programme of the US National Cancer Institute began registering cancer incidence for about 10% of the US population in 1973. Since then its coverage has increased,18 and it currently registers cancer incidence and subsequent cause-specific mortality in 26% of the US population. Women were potentially eligible for the present study if they were registered in SEER during 1973–2001 with local or regional breast cancer of known laterality diagnosed at age 20–79 years, and had no previous

Results

Of 318 293 potentially eligible women, 6859 (2%) had radiotherapy status unknown, 654 (<1%) had received only brachytherapy implants or radiation of unknown type, 1919 (<1%) had no information on survival, and 308 861 (97%) were included. Of the included women, 115 165 (37%) were recorded as having received external-beam radiotherapy as part of the initial treatment for breast cancer. Of those known to have died before 2002, the certified cause was available for 26 216 (96%) of the 27 400

Discussion

We have shown that, in women recorded in the US SEER cancer registries as having been diagnosed with breast cancer during 1973–82 and irradiated, mortality from heart disease was increased among women with left-sided breast tumours compared with women with right-sided breast tumours. In these same women, mortality from cancer of the ipsilateral lung was increased compared with mortality from cancer of the contralateral lung. For both heart disease and ipsilateral lung cancer these increases,

References (31)

  • J Cuzick et al.

    Cause-specific mortality in long-term survivors of breast cancer who participated in trials of radiotherapy

    J Clin Oncol

    (1994)
  • Favourable and unfavourable effects on long-term survival of radiotherapy for early breast cancer: an overview of the randomised trials

    Lancet

    (2000)
  • Effects of radiotherapy and surgery in early breast cancer: an overview of the randomized trials

    N Engl J Med

    (1995)
  • M Deutsch et al.

    The incidence of lung carcinoma after surgery for breast carcinoma with and without postoperative radiotherapy

    Cancer

    (2003)
  • M Sautter-Bihl et al.

    Radiotherapy of internal mammary lymph nodes in breast cancer

    Strahlenther Onkol

    (2002)
  • Cited by (0)

    View full text