Effect of Delayed Targeted Intraoperative Radiotherapy vs Whole-Breast Radiotherapy on Local Recurrence and Survival: Long-term Results from the TARGIT-A Randomized Clinical Trial in Early Breast Cancer

Jayant S. Vaidya, Max Bulsara, Christobel Saunders, Henrik Flyger, Jeffrey S. Tobias, Tammy Corica, Samuele Massarut, Frederik Wenz, Steffi Pigorsch, Michael Alvarado, Michael Douek, Wolfgang Eiermann, Chris Brew-Graves, Norman Williams, Ingrid Potyka, Nicholas Roberts, Marcelle Bernstein, Douglas Brown, Elena Sperk, Siobhan LawsMarc Sütterlin, Steinar Lundgren, Dennis Holmes, Lorenzo Vinante, Fernando Bozza, Montserrat Pazos, Magali Le Blanc-Onfroy, Günther Gruber, Wojciech Polkowski, Konstantin J. Dedes, Marcus Niewald, Jens Blohmer, David McCready, Richard Hoefer, Pond Kelemen, Gloria Petralia, Mary Falzon, Michael Baum, David Joseph

Research output: Contribution to journalArticlepeer-review


Importance: Conventional adjuvant radiotherapy for breast cancer given daily for several weeks is onerous and expensive. Some patients may be obliged to choose a mastectomy instead, and some may forgo radiotherapy altogether. We proposed a clinical trial to test whether radiotherapy could be safely limited to the tumor bed. Objective: To determine whether delayed second-procedure targeted intraoperative radiotherapy (TARGIT-IORT) is noninferior to whole-breast external beam radiotherapy (EBRT) in terms of local control. Design, Setting, and Participants: In this prospective, randomized (1:1 ratio) noninferiority trial, 1153 patients aged 45 years or older with invasive ductal breast carcinoma smaller than 3.5 cm treated with breast conservation were enrolled from 28 centers in 9 countries. Data were locked in on July 3, 2019. Interventions: The TARGIT-A trial was started in March 2000; patients were randomized after needle biopsy to receive TARGIT-IORT immediately after lumpectomy under the same anesthetic vs EBRT and results have been shown to be noninferior. A parallel study, described in this article, was initiated in 2004; patients who had their cancer excised were randomly allocated using separate randomization tables to receive EBRT or delayed TARGIT-IORT given as a second procedure by reopening the lumpectomy wound. Main Outcomes and Measures: A noninferiority margin for local recurrence rate of 2.5% at 5 years, and long-term survival outcomes. Results: Overall, 581 women (mean [SD] age, 63 [7] years) were randomized to delayed TARGIT-IORT and 572 patients (mean [SD] age, 63 [8] years) were randomized to EBRT. Sixty patients (5%) had tumors larger than 2 cm, or had positive nodes and only 32 (2.7%) were younger than 50 years. Delayed TARGIT-IORT was not noninferior to EBRT. The local recurrence rates at 5-year complete follow-up were: delayed TARGIT-IORT vs EBRT (23/581 [3.96%] vs 6/572 [1.05%], respectively; difference, 2.91%; upper 90% CI, 4.4%). With long-term follow-up (median [IQR], 9.0 [7.5-10.5] years), there was no statistically significant difference in local recurrence-free survival (HR, 0.75; 95% CI, 0.57-1.003; P =.052), mastectomy-free survival (HR, 0.88; 95% CI, 0.65-1.18; P =.38), distant disease-free survival (HR, 1.00; 95% CI, 0.72-1.39; P =.98), or overall survival (HR, 0.96; 95% CI, 0.68-1.35; P =.80). Conclusions and Relevance: These long-term data show that despite an increase in the number of local recurrences with delayed TARGIT-IORT, there was no statistically significant decrease in mastectomy-free survival, distant disease-free survival, or overall survival. Trial Registration: ISRCTN34086741, ClinicalTrials.gov Identifier: NCT00983684.

Original languageEnglish
Article numbere200249
Number of pages10
JournalJAMA oncology
Issue number7
Publication statusPublished - Jul 2020

ASJC Scopus subject areas

  • Oncology
  • Cancer Research


Dive into the research topics of 'Effect of Delayed Targeted Intraoperative Radiotherapy vs Whole-Breast Radiotherapy on Local Recurrence and Survival: Long-term Results from the TARGIT-A Randomized Clinical Trial in Early Breast Cancer'. Together they form a unique fingerprint.

Cite this