HeidelbergCement (United Kingdom)

Abstract Introduction: In June 2010, we published the results of the TARGIT-A trial (1) that compared “one-size fits” all radiotherapy (whole breast radiotherapy-EBRT group) with risk -adjusted radiotherapy (TARGIT- group: single dose TARGeted Intraoperative radioTherapy with additional whole breast radiotherapy if adverse prognostic factors were found). These results showed that the risk of local recurrence in the TARGIT group was non-inferior to that of the EBRT group (the difference between the two arms was 0.25% at 4 years). We now report the results analyzed after further follow up of the total trial cohort without unblinding. Method: As often repeated, unplanned analysis comparing two groups in a randomized trial can result in an increased alpha-spent as well as carry the risk of a false positive result, we have remained blinded to the further recurrences in the trial according to allocated treatment, since the original publication in the Lancet. Instead we proposed and performed a blinded analysis of the local recurrence rate for the whole cohort. We plotted the Kaplan Meier plots and compared the estimated 4-year recurrences. Results: Amongst the 2232 patients randomized, there were 13 recurrences at the time of Lancet publication and since then, we have had 8 additional recurrences. The number of patients who have completed at least 4 years of follow up has increased from 420 to 717. We found that the 4-year Kaplan Meier estimate of local recurrence was 1.08% (95% CI 0.59 −1.96) at the time of the Lancet publication and it is 1.09% (95% CI 0.65 - 1.85) now. Conclusion: We found that the overall 4-year recurrence rates of the TARGIT-A trial have remained stable with a longer follow up and therefore; it is statistically implausible, that one particular arm has a significantly higher local recurrence.