Adjuvant irradiation for breast cancer

The use of postoperative adjuvant irradiation in breast conservation and after mastectomy is increasing. However, concerns exist about the risks of cardiac morbidity and mortality induced by radiation. These risks are worrying because three randomised trials have suggested that locoregional irradiation after mastectomy, when combined with appropriate systemic therapy, improves local control and survival compared with systemic therapy alone in both premenopausal and postmenopausal women at high risk of local relapse.^1–3 In 1995 the Early Breast Cancer Trialists' Collaborative Group contributed a valuable overview of 36 randomised trials of surgery and radiotherapy (including 28 405 women for whom data on mortality were available).⁴ There was no difference in overall survival between the irradiated and non-irradiated groups after mastectomy or breast conserving surgery. Although radiotherapy did reduce deaths from breast cancer, it was associated with an increased risk of death from other causes.

These other deaths were mainly cardiovascular. Patients aged 60 or older at the time of randomisation were particularly at risk of death from causes other than breast cancer. Data were not included on the laterality of the primary tumour or on radiotherapy technique. Both are important and would ideally be included in future overviews. More of the heart is likely to be included in irradiation of left sided cancers. Non-randomised data from the surveillance, epidemiology, and end-results (SEER) study showed that the risk of fatal myocardial infarction was associated with irradiation of left sided tumours in women younger than 60 but not in women older than 60.⁵ Patients with affected axillary nodes had a higher relative risk of dying of a fatal myocardial infarction. This is probably due to the increased use of peripheral lymphatic irradiation in the node positive group. Regional nodal irradiation for left sided tumours is more likely to irradiate part of the heart.

An analysis of cardiac toxicity was carried out in a Swedish trial, in which no systemic therapy was given.⁶ The dose to the heart was estimated using three dimensional reconstruction in a small number of patients, and the risk of fatal myocardial infarction was found to be three times higher in those patients with the largest volumes of the heart irradiated for left sided tumours than in controls. Studies using thallium-201 scintigraphy in a small series of women with stage 1-111 breast cancer who underwent breast or chest wall irradiation for left sided tumours showed that half of the patients showed new scintigraphic cardiac defects, indicating regional hypoperfusion.⁷ These probably reflect radiation induced damage to the microcirculation.

Irradiation of the internal mammary nodes is an additional risk factor for cardiac morbidity. In the Danish trials the peripheral lymphatics, including the internal mammary chain, were irradiated in addition to the chest wall.³ It is unclear whether irradiation of all these areas is critical to optimal locoregional control. The results of the EORTC 2292 trial should clarify the impact of internal mammary irradiation on locoregional control and morbidity. Interestingly, the internal mammary nodes were irradiated in the Danish trials but using a technique that delivered a relatively small dose to the heart. This may well account for the significantly better overall survival—9-10% at 10 years—seen in the Danish trials.¹,³ This notion is supported by the long term follow up data, which show no increase in cardiac morbidity or mortality in either premenopausal or postmenopausal groups.⁸ All patients were treated with a uniform technique and to an international standard dose and fractionation regimen.

Many of the older radiotherapy trials included in the Oxford overview used techniques which delivered higher cardiac doses and higher doses per fraction of radiotherapy.⁴ These techniques would not now be considered acceptable. Both factors may have contributed to the excess cardiac mortality seen in the Manchester and Oslo 11 trials.⁸,⁹ By contrast the later Stockholm trial⁶ showed a survival advantage for adjuvant irradiation, albeit of borderline statistical significance.¹⁰ Current evidence therefore suggests that modern radiotherapy is associated with lower cardiac morbidity. Over time the impact of modern radiotherapy techniques may be reflected in a net survival gain as modern trials of radiotherapy accrue longer term follow up.

How should the dose to the heart be minimised during locoregional radiotherapy? One possibility is to adopt the radiotherapy technique in the Danish trial, which made use of electron beams with limited penetration beyond the chest wall. Shielding the heart completely from megavoltage beams is difficult using conventional radiotherapy planning simulators and older linear accelerators. A recent survey of clinical oncologists in the United Kingdom reflects this, showing that 35% of them undertook shielding of the heart.¹¹ However, the increasing replacement of old linear accelerators with newer machines with the greater technical capability to shield critical structures using multileaf collimators should help. For accurate measurement of irradiated cardiac volume, three dimensional computed tomography planning facilities are needed with a sufficiently wide aperture to allow planning in the same position as for radiotherapy. Computed tomography planning in selected patients may also enable the dose to the heart to be reduced.

With the expansion in the use of adjuvant chemotherapy regimens containing potentially cardiotoxic anthracyclines and of adjuvant locoregional irradiation, long term cardiac morbidity and mortality may increase. At least for postmastectomy irradiation for women at high risk of locoregional recurrence, there is evidence that the survival benefits may be comparable to those of adjuvant therapy. A national consensus on breast radiotherapy planning technique would help in minimising cardiac morbidity while maximising locoregional control and survival.