Triple-Negative Breast Cancer is Associated with Higher Risk of Local Recurrence after 3D-Conformal External Beam Accelerated Partial Breast Irradiation (3D-APBI)

Reviewer: Samuel Swisher-McClure, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 2, 2010

Authors: I. M. Pashtan, M. Ancukiewicz, J. Y.Wo, A. E. Hirsch, B. L. Smith, S. N. Powell, A. Recht, A. G. Taghian.
Affiliations: Harvard Radiation Oncology Program, Boston, MA. Massachusetts General Hospital, Boston, MA. Boston Medical Center, Boston, MA. Memorial Sloan-Kettering Cancer Center, New York, NY. Beth Israel Deaconess Medical Center, Boston, MA.


  • APBI has become an increasingly utilized component of breast conserving therapy.
  • ASTRO has previously published consensus guidelines regarding selection of appropriate patients for APBI
  • ASTRO consensus guidelines (Smith et al. IJROBP, 2009) state that suitable patients for APBI include patients who meet the following criteria:
    • Age ? 60 years
    • Tumor size ? 2.0 cm that is unicentric, removed with negative margins > 2mm, Estrogen receptor positive, and without evidence of LVSI
    • Patients must have no evidence of Lymph node metastases as assessed by Sentinel Lymph Node Biopsy.
    • APBI is not recommended for patients with DCIS or for patients who have received neoadjuvant chemotherapy outside of the context of a clinical trial.
  • However, there is limited data regarding the association of biologic sub-type with clinical outcomes in patients receiving APBI.
  • Data from the American Society of Breast Surgeons Mammosite Registry Trial (Shaitelman et al. Cancer, 2010) found that negative estrogen receptor (ER) status was significantly associated with risk of ipsilateral breast tumor recurrence after APBI for invasive breast cancer (Odds Ratio 4.01, p = 0.0003).
  • Retrospective analysis of a series of patients receiving breast conservation therapy at the Dana-Farber Cancer Center (Nguyen. JCO, 2008) also found that patients with triple-negative invasive breast cancer have a higher risk of local recurrence at 5 years after whole breast irradiation (WBI) compared to patients with ER positive tumors ( 7% vs. 1.5%, Adjusted Hazards Ratio = 7.1; 95% CI 1.6-31; P = .009)
  • Here, we evaluate whether biologic subtype, as determined by estrogen, progesterone, and HER-2 receptors, is associated with local recurrence (LR) following 3D-APBI.


  • This study was a sub-set analysis of an IRB approved prospective clinical trial of dose-escalation in APBI.
    • Patients enrolled in this trial received one of three doses of radiation: 32 Gy, 36 Gy, or 40 Gy delivered in 4 Gy/BID fractions.
  • From October 2003 through November 2005, 99 patients with Stage I breast cancer were treated on the first dose-step of trial (32 Gy in 4 Gy/BID fractions).
  • The analysis included only the cohort of patients receiving 32 Gy because this was the group of patients with the longest term follow-up data available.
  • Eligibility criteria included:
    • Infiltrating ductal histology or special subtype
    • pT1N0
    • Absence of extensive intraductal component (EIC) and lymphovascular space invasion LVI
    • Margins > 2 mm
  • Radiation techniques included:
    • 64% of patients were treated with mixed photons and electrons, usually with 2 mini-tangents and an en-face electron field delivering about 20% of the dose.
    • 16% of patients were treated with photons alone.
    • 20% of patients were treated with protons.
  • 7 patients received chemotherapy and 65 hormonal therapy.
  • Ipsilateral breast tumor recurrences were categorized as "true" vs. "elsewhere recurrence"
    • Elsewhere recurrence was defined as occurring in a separate quadrant and at least 3 cm from the primary tumor bed.
    • True recurrence occurred within the same quadrant and within 3 cm of the tumor bed.
    • The location of the recurrence was evaluated independently by a blinded radiologist.
  • Median FU was 66.6 months (range, 1.6-74.3 months).


  • Characteristics of the Study Population:
    • The median patient age was 61 years (range, 40-80 years)
    • The median tumor size was 0.8 cm
    • 91% had ductal carcinomas
    • 89% of tumors were ER+
    • 10% had a triple-negative phenotype (75% of these patients received systemic chemotherapy in addition to breast irradiation).
    • 1% of patients had an ER/PR-, Her2 + tumor.
    • 47% were histologic grade 1, 42% grade 2, and 10% grade 3
  • Higher tumor grades correlated with triple-negative phenotype (Kendall's tau = 0.43, p< 0.0001).
  • Five patients had LR at a median of 47.5 months after APBI (range, 29.5-55.3 months), for a 5-year actuarial LR rate of 6% (95% CI, 0.6-11%).
  • All of these local recurrences were categorized as "elsewhere recurrences".
  • Three of these occurred in patients with triple-negative disease (1 same quadrant recurrence, 2 elsewhere), for a 5-year actuarial LR rate of 32.5% (95% CI, 0-57%).
  • 2 occurred in non-triple-negative patients (both elsewhere), for a 5-year actuarial LR rate of 3% (95% CI, 0-7%) [p = 0.0001, log-rank test].
  • The hazard ratio for LR with triple-negative phenotype was 15.4 (95% CI, 2.5-91).
  • All 3 of the triple negative patients who developed LR had previously received systemic chemotherapy.
  • On multivariate analysis, triple-negative phenotype was the only predictor of LR, with borderline statistical significance (p = 0.052, log-rank test) after adjusting for tumor grade.
  • Effect of tumor grade on risk of LR was not significant (p = 0.33, log-rank test) after adjusting for triple-negative status.
  • All LR patients were successfully salvaged with mastectomy.

Author's Conclusions

  • In this study, patients with triple-negative breast cancer had a significantly higher LR rate than patients with other receptor subtypes following 3D-APBI to 32 Gy.
  • Caution should be used in deciding whether or not to treat triple negative breast cancer patients with 3D-APBI outside of the setting of a clinical protocol.
  • Additional prospective studies are needed to assess the long-term efficacy of this approach, and whether patients with triple-negative disease might be better treated with whole-breast irradiation.

Clinical Implications

  • This study was a subset analysis of a prospective clinical trial of dose escalation in 3D-APBI as part of breast conservation therapy for early stage breast cancer.
  • The authors found that triple negative breast cancer patients treated with APBI had a significantly elevated 5 yr LR rate of 32.5% compared to 3% in non-triple negative patients.
  • This rate of LR seen with APBI is inferior to the results observed by Nguyen et al. in triple negative patients treated with WBI (7% at 5 years). The observation in the current study that the majority of failures were not within or near the tumor bed supports the utility of WBI in this population.
  • The published ASTRO consensus guidelines indicate that only women with ER + tumors are deemed "suitable" for APBI outside of the context of a clinical trial.
  • This study provides prospective clinical data supporting the recommendation that triple negative breast cancer patients are unsuitable for APBI unless as part of a clinical trial.
  • Limitations of the study include:
    • The study had a limited sample size of only 99 patients, 10 of which were triple negative.
    • There were also a small number of events with only 5 patients developing LR with the current follow-up.
  • APBI is currently being evaluated in the ongoing NSABP B-39/RTOG 0413 trial which is a phase 3 randomized trial comparing APBI to WBI for women with early stage breast cancer undergoing breast conservation therapy.
  • The results of that study should better inform the selection of appropriate patients for APBI.

Stay informed with the latest information from OncoLink!   Subscribe to OncoLink eNews
View our newsletter archives