To the Editor:

We read with great interest the review article on internal mammary nodes (IMNs) in breast cancer. Chen et al¹ should be commended on writing an eloquent and articulate summary on this controversial topic. The authors correctly suggested that decisions concerning whether to include the IMNs in radiation treatment fields should carefully consider the benefits of treatment against the potential risk of cardiac toxicity.

While randomized data on the survival benefit of regional nodal irradiation are pending, a common reason for not addressing the IMNs is the assumption that the incidence of involvement is relatively low. However, a modern study by Huang et al² analyzed the incidence of pathologically involved IMNs from 1,679 patients treated with an extended radical mastectomy with en bloc IMN dissection of the first through fourth intercostal spaces. Regardless of tumor location, the rate of IMN involvement was 41.5% for patients with more than six positive axillary lymph nodes, 28.1% with four to six positive axillary lymph nodes, and 15% with negative axillary lymph nodes. Furthermore, they reported a 47.5% incidence of pathologic positive IMNs among patients with four to six positive axillary lymph nodes and a medial tumor. Among axillary node-negative patients, the rate of pathologic positive IMNs was lower, but 3.1 times higher for medial tumors.

In 1999, Veronesi et al³ reported results from a randomized trial of patients treated with Halted mastectomy or extended radical mastectomy which included an IMN dissection. This did not show overall survival benefit; however, neither arm of the study allowed systemic chemotherapy or radiation. As Chen et al¹ observe, the trial was underpowered to find a survival difference due to small sample size and the prohibited use of systemic chemotherapy, further reducing its power to detect a difference in locoregional control. The outcome was overwhelmingly dismal with an overall survival of less than 30%. The trial did however find that pathologic involvement of axillary or IMNs were highly prognostic on survival curves.

Therefore, available data indicate the incidence of IMN involvement is not trivial and that patients with IMN involvement have inferior survival after surgery alone compared with patients without pathologic IMN disease. It is not surprising then to find data demonstrating superior results with the addition of radiation treatment to the IMN chain. Grabenbauer et al⁴ observed IMN radiation reduced distant metastases in patients with medial tumors with a hazard ratio of 1.6 compared with patients without IMN radiation (P = .02). Lê at al⁵ reported equivalent overall and metastases-free survival for 330 patients with medial tumors as for 492 patients with lateral tumors when IMN radiation was given for medial tumors.

Historically, IMN radiation has had associated cardiac morbidity as reported in the Early Breast Cancer Trialist Group study.⁶ In the trials of radiotherapy after mastectomy and axillary clearance in node-positive patients, the reduction in 15-year all-cause mortality was 4.4% (64.2% v 59.8%, 2P = .0009) compared with the 5.4% reduction in 15-year breast cancer mortality. At 20 years, the reduction in breast cancer mortality remains unchanged at 5.4% (66.4% v 61.0%), whereas that for all-cause mortality was 3.5% (72.3% versus 68.8%), indicating a continuing excess of non–breast cancer mortality long after treatment with the older radiotherapy regimens. In contrast, the Danish postmastectomy randomized trials used anterior-shaped electrons to treat the IMNs and found no increased morbidity or mortality of ischemic heart disease with this technique.⁷ The proportion of women who died from ischemic heart disease was 0.9% in the no-radiotherapy group compared with 0.8% in the radiotherapy group. Højris et al⁷ concluded that postmastectomy radiotherapy using their technique did not increase the actuarial risk of ischemic heart disease after 12 years.

The review by Chen et al¹ provided compelling evidence in favor of treating the IMNs, given that one weighs the benefit against the potential risk of cardiac toxicity. Although the article cited references that demonstrate no increase in cardiac mortality for left-sided versus right-sided breast cancers treated after 1980, specific reasons or changes in technique responsible for this improvement were not fully described. An important message that can be conveyed to Journal of Clinical Oncology readership is that radiation treatment planning and delivery techniques are available to safely include the IMNs at risk while avoiding coincidental cardiac irradiation.

This message is important in that many trials, such as the adjuvant trastuzumab studies, specifically exclude radiation of IMNs, for unwarranted fear of cardiovascular effects. At our institution, we routinely include the IMNs for all patients treated with postmastectomy radiation. Using computed tomography–based planning, we contour our target defined as the first three interspaces and use electron-based techniques that cover these nodes while sparing high dosages of radiation to the heart and coronary vasculature. Other investigators have summarized various techniques for treating the IMNs; however, none has included this split-electron technique.^6–10 The technique exploits the anatomic advantage created by the left ventricle lying near the chest wall inferiorly, typically below the IMNs that are at highest risk of involvement.

Because the IMNs at greatest risk lie in the first-third intercostal spaces, we will often use a higher-energy electron superiorly matched to a lower-energy electron. Figure 1 illustrates our split-electron beam technique using a higher-energy electron superiorly and a lower-energy electron inferiorly, both matched to chest wall tangents using 6 MV photons. Acceptable plans typically cover the IMN with the 90% isodose line and permit a small cold triangle of 35 Gy dose minimum. The cold match can be minimized by angling the electrons between 15 and 20 degrees depending on the slope of the chest wall tangents. This technique also has the advantage of irradiating little lung volume compared with techniques such as partially wide tangent fields. With evidence from the Early Breast Cancer Trialist's Collaborative Group that aggressive locoregional treatment translates to improved long-term survival, I expect more physicians will decide to treat the IMNs and hope this technique offers a rationale and safe option to do so.⁹

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The author(s) indicated no potential conflicts of interest.