Background: The effectiveness of accelerated partial breast irradiation (APBI), especially concerning intraoperative radiotherapy (IORT), is still controversial. In 2013 we reported (Veronesi U, Orecchia R, Maisonneuve P, et al. Lancet Oncol 2013) the 5-year results of the ELIOT (ELectron IntraOperative Therapy) trial, an equivalence randomised phase III study, registered with ClinicalTrials.gov, number NCT01849133. This study was done at a single institution, the European Institute of Oncology (Milan, Italy). Women aged 48-75 years with early breast cancer, a maximum tumour diameter of up to 2.5 cm, and suitable for breast-conserving surgery (BCS) were randomly assigned in a 1:1 ratio to receive either whole-breast external radiotherapy (WBI) or intraoperative radiotherapy (IORT) with electrons. Study coordinators, clinicians, and patients were aware of the assignment. Patients in the IORT group received one single dose of 21 Gy to the tumour bed during surgery. Those in the WBI group received 50 Gy in 25 fractions of 2 Gy each, followed by a boost of 10 Gy in 5 fractions. The primary endpoint was occurrence of ipsilateral breast tumour recurrences (IBTR). After a medium follow-up of 5.8 years, 35 patients in the IORT arm and 4 patients in the WBI arm had an IBTR (p<0.0001). The 5-year event rate for IBRT was 4.4% in the IORT group and 0.4% in the WBI group (hazard ratio 9.3). During the same period, 34 women allocated to IORT and 31 to WBI died (p=0.59). Five-year overall survival was 96.8% in the IORT arm and 96.9% in the WBI arm. In patients with data available (n=464 for IORT; n=412 for WBI) we noted significantly fewer skin side-effects in women in the IORT arm than in those in the WBI arm (p=0.0002).
Aims: Hereafter we reported the long-term outcomes, in terms of local (IBTR) and regional recurrences and overall survival (OS), of the ELIOT trial.
Methods: Between November 2000 and December 2007, 1305 women were randomised using a random permuted block design, stratified for clinical tumour size, to receive WBI (n=654) or IORT with electrons (n=651). The main analysis was by intention to treat, but also a per-protocol analysis was performed. Cumulative incidence of local, loco-regional, distant events and OS were presented at 5-, 10-, and 15 years of follow-up.
Results: After a medium follow-up of 12.4 years, 86 (6.6%) patients developed IBTR, with 70 (10.7%) cases in the IORT cohort and 16 (2.4%) in the WBI cohort (P<0.0001) (hazard ratio 4.62). The 5-, 10-, and 15-year event rates for IBRT in the IORT arm were 4.2%, 8.1%, and 12.6%, respectively, and in the WBI arm 0.5%, 1.1% and 2.4%. Most patients in the ELIOT arm who developed an IBTR received a second course of surgery, with 35 (50%) of the 70 patients who repeated BCS, and 24 (34.3%) underwent total mastectomy. Thirteen women developed axillary of other regional lymph node metastasis, of which 11 (1.7%) in the ELIOT arm and two (0.3%) in the WBI arm. A lower number of contralateral breast cancer was recorded in the ELIOT arm (18 cases) than in the WBI arm (27 cases), but the difference was not statistically significant (p=0.42). We analysed factors associated with IBTR, and defined a group of women (75 in the ELIOT arm and 66 in the WBI arm) at “very low risk” of IBTR, consisting in women with three concomitant characteristics (tumour size <1cm, grade 1, and Luminal A [Ki-67<14%]), each associated with an IBTR rate <10% after ELIOT at 15-year. The incidence of IBTR in this subgroup was extremely low with either modality. Similar results were found in the per-protocol analysis. No statistical difference was also observed for the development of distant metastasis, 46 in the ELIOT arm and 54 in the WBI arm (p=0.60). Ten- and 15-year OS were 90.7% and 83.4% in the IORT group, and 92.7% and 82.4% in the WBI group, respectively. Similar results were found in the per-protocol analysis. No further data about long-term toxicity were collected, and earlier effects were reported in the previous published analysis.
Conclusions: This is the study with the longest follow-up among all the APBI and IORT trials. Long-term results confirmed the higher rate of IBTR in the IORT arm compared to WBI arm, but, of great importance, without any impact on distant metastasis and survival rates. We conducted this trial over the period 2000-2007, when the use of APBI, and particularly by IORT, was quite pioneering, and very few selection criteria were adopted. This allowed to identify a small group of patients at “very low risk” who fared well with IORT. For other women a careful assessment of risks and benefits is required in the context of a fully informed and shared decision making. Finally, the results of this study stressed the importance of the proper selection of women candidate to IORT.