Cardiac Toxicity is Not Increased 25 Years After Treatment of Early Stage Breast Carcinoma With Mastectomy or Breast Conservation Therapy From the National Cancer Institute Randomized Trial

Reporter: Abigail T. Berman, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 30, 2012

Presenting Author: Charles B. Simone, II, MD
Presenting Author's Affiliation: Hospital of the University of Pennsylvania, Department of Radiation Oncology, Philadelphia, PA


  • The current options for the treatment of early-stage breast cancer are modified radical mastectomy or breast-conserving surgery plus radiotherapy, together known as breast conservation treatment.
  • The National Cancer Institute Breast Conservation Trial randomized 237 patients with stage I-II breast cancer from 1979-1986 to one of the following arms:
    • Arm 1: Modified radical mastectomy + axillary dissection (level I/II)
    • Arm 2: Lumpectomy + axillary dissection (level I/II) à RT
      • Radiation was delivered as follows: 45-50.4 Gy whole breast +/- regional nodes, 15-20 Gy boost (Ir-192 or electrons).
      • Of the 6 randomized trials in the 1970s-1980s that posed similar questions, the NCI trial was unique in that all patients underwent CT simulation with dose inhomogeneity corrections
    • Of note, all node positive patients (40%) received chemotherapy with adriamycin and cytoxan for 6-11 cycles. Tamoxifen was added for postmenopausal patients in 1985.
  • In the MRM patients, there were 63/116 deaths. In the BCT arm, there were 76/121 deaths, 43.8% vs. 37.9% alive at 25.7 yrs with p=0.38. Prior to 25 years, the curves were superimposable but then they began to separate. Thus, the question arose if the treatment toxicity might be causing separation of the curves (Simone NL, et al. Breast Cancer Res Treat. 2012;132(1):197-203.
  • There are conflicting data regarding cardiovascular complications following breast RT. Some studies report increased cardiac morbidity with left breast radiation. However, many of these positive studies used two-dimensional planning only, and patients were treated prior to the CT planning era.
  • This study conducted further investigation into the patients treated on the NCI Breast Conservation trial.

Materials and Methods

  • Approximately half of patients were willing to return to the NCI, and a total of 50/102 (26 BCT, 24 MRM) returned for comprehensive follow-up conducted as follows:
    • History: PMH, cardiac events, smoking
    • Exam: signs of heart failure, ischemia, arrhythmias
    • Cardiac labs: lipid panel, CRP, homocysteine, HbA1c, ProBNP, creatinine, cystatin-c
    • Radiation technique parameters assessed including
      • Central lung distance, field size, dose, boost
  • All patients underwent two forms of detailed cardiac imaging:
    • Cardiac MRI (3T) to evaluate anatomy and function
    • CT Angiogram to determine coronary arterial calcium (CAC) score and to evaluate for stenotic coronary artery disease


  • Patient characteristics compared between MRM and BCT patients as well as right breast versus left breast patients were well-balanced with regard to age at diagnosis, current age, BMI, smoking history, hormone replacement, therapy, hypertension, hyperlipidemia, chemotherapy, history of MI, history of heart failure, history of cardiac catheterization or CABG. There was a  non-significant trend towards a higher percentage of patients with diabetes in the MRM cohort than the BCT cohort (13 vs 4%).
  • Central lung distance was <2 cm in 11% (15% right vs. 9% left), 2-3 cm in 50% (53% right vs 46% left), and >3 cm in 39% (35% right vs. 45% left). There were no significant differences in CLD in right versus left.
  • Exam and laboratory findings were similar between MRM and BCT as well as right versus left with regard to diastolic BP, creatinine, total cholesterol, LDL cholesterol, HDL cholesterol, HbA1c, hs-CRP, proBNP, homocysteine, cystatin-c, Framingham Risk Score. There was a trend toward higher systolic BP in MRM than BCT patients (139 vs. 127). There was also a non-significant trend in pro-BNP in MRM versus BCT (256 vs. 176) and in right than left breast cancer patients (216 vs. 113).
  • Cardiac MRI revealed no significant differences between patients who underwent MRM or BCT and right versus left cancers with regard to the following endpoints: left atrial volme, RV end diastolic area, left ventricle mass, LV end-diastolic volume, LV end-systolic volume, stroke volume, LV ejection fraction, peak midwall strain, peak filling rate, time to PFR, diastolic volume recovery.
    • There was a significant difference in left ventricle mass (gm) in MRM vs. BCT (111 vs. 90.5%, p=0.02). Left ventricular mass difference was not significant after adjusting for systolic blood pressure.
    • There was also was a significant difference in time to PFR (sec) in MRM vs. BCT (647 vs. 487, p=0.02).
    • There were no patients with evidence of myocardial fibrosis, and there was 1 patient in each arm with pericardial thickening.
  • Cardiac CT angiogram revealed no significant differences between patients who underwent MRM or BCT and right versus left cancers with regard to the following endpoints: any visible plaque (>0), any significant stenosis (>50%), any severe stenosis (>70%), RCA segments visible plaque (>0), RCA segments significant stenosis (>50%), RCA segments severe stenosis (>70%), LM segments any stenosis, LAD segments visible plaque (>0), LAD segments significant stenosis (>50%), LAD segments severe stenosis (>70%), LCx segments visible plaque (>0), LCx segments significant stenosis (>50%), LCx segments severe stenosis (>70%), proximal segments visible plaque (>0), proximal segment significant stenosis (>50%), proximal segments severe stenosis (>70%).
    • Median CAC score was similar in both arms: BCT 25 (IQR 0, 86) vs. MRM 0 (IQR 0, 354), p=0.65.
    • There was no significant increase but there was a trend in the risk of visible atherosclerosis with chemotherapy (HR 2.4, 95% CI 0.94-6.32, p=0.07).

Author's Conclusions

  • This is the first study reporting comprehensive long-term cardiac outcomes after randomization for breast cancer therapy. This is reassuring data demonstrating that 25 years after breast radiation, cardiac toxicity does not seem to be responsible for slight decrease in patient survival in the BCT arm. In addition, there was no difference for left- vs. right-sided tumors.
  • This study was performed in the CT simulation and 3D planning era, therefore reflecting current practice.
  • Based on this study, in the era of 3D planning, patients with early-stage breast cancer treated with RT do not have a higher risk of long-term cardiac morbidity compared with patients having MRM.

Clinical Implications

  • This is a unique, important study that assesses multiple laboratory and radiographic aspects of cardiac risk over 25 years after treatment. The results show no increased risk of cardiac risk and they do not explain the slight increased survival seen in the MRM arm of the NCI study. This is reassuring for patients undergoing breast conservation treatment.
  • One potential limitation is that this study only assessed patients who were alive at 25 years. If they died of cardiac disease prior to this study being performed, they were not included in the analysis. This presents potential bias in the results.
  • It is important to note that this study was performed in the era of 3D planning. While CT simulation may decrease the risk of heart disease compared to older studies, heart dose and therefore risk is also highly dependent on patient anatomy and physician planning skill.
  • There have been multiple other studies examining cardiac toxicity that have shown an increased risk of coronary disease with RT.
    • The Uppsala University (Nilsson G, J Clin Oncol. 2012 Feb 1;30(4):380-6) found on coronary angiography data that the odds ratio for left-sided radiation for grade 3-5 stenosis LAD 4.4 and for grade 4-5 stenosis 7.2.
    • The UPENN study looked at patients screened with stress tests and cardiac catheterization and found a 10-year risk of CAD: right and left 7% (NS) but a 12-year risk of catheterization-detected abnormalities in the right 8% vs. left 59% (SS), again concluding that left-sided RT increases the risk of coronary damage.
    • The Duke study (Marks LB et al. Int J Radiat Oncol Biol Phys. 2005 Sep 1;63(1):214-23.) looked at technetium 99m sestamibi or tetrofosmin pre- and post-RT and found a 42% incidence of perfusion defects at 24 months.
    • Danish trials 82b/82c (Hojris I et al. Lancet. 1999 Oct 23;354(9188):1425-30.) found no increased risk of ischemic heart disease at 12 yrs.
  • None of these studies, however, had as extensive laboratory and radiographic data as the current study. In addition, they do not report 25 year data. They also do not uniformly use CT planning for the patients treated. For these reasons, the data presented here potentially reflect modern clinical practice the most accurately.