52O Resistance to imatinib induced by treatment interruption in advanced GIST: Long-term outcome of the randomized BFR14 study

The long-term impact of tyrosine kinase inhibitor (TKI) interruption on resistance and survival in patients with advanced GIST is unclear. We report here on the long-term outcome of patients with advanced GIST stopping imatinib in the BFR14 randomized study. Patients with advanced GIST (any mutation) aged >=18 with a PS 0-3, adequate renal and hepatic function were eligible in the BFR14 study. Patients in complete or partial response, or stable disease at 3 time periods (1 year, 3 years, 5 years) were proposed for randomisation between treatment discontinuation and restart at progression (STOP arm) or continuation of the treatment until progression (CONT arm). Randomization at 1, 3 and 5 years took place from Feb 2003 to Mar 2004, May 2005 to May 2007, and Nov 2007 to Jul 2010 respectively. The primary objective was to compare progression free survival (PFS) in the two groups. Secondary endpoints included overall survival (OS) and time to imatinib resistance (TTIR), defined as the time between randomisation and progression on imatinib (i.e., after restart in the stop arm). Median follow-up after 1, 3, and 5 years randomisations were 234, 200 and 164 months respectively. Among the 434 patients included in the BFR14 trial, respectively 58 were randomized in the STOP and CONT arms after 1 year, 50 after 3 years, and 27 after 5 years. Two were lost for follow-up. PFS was significantly improved in the CONT vs STOP arms in the three randomization periods. (p<0.001). The TTIR was significantly shorter for patients randomized in the STOP vs CONT arms after 3-years (66 vs 127 months, p=0.009), and after 5-years of imatinib (58.6 vs NR, p=0.03), while no significant difference was found for those randomised after 1 year. Survival after the randomization at 1 year was not significantly different but median OS were 66 vs 127 months in the STOP and CONT arms. Survival after randomization at 3 years was significantly shorter in the STOP arm after 10 year (logrank p=0.005, 15 year survival 5% vs 42%). Survival after randomization at 5 years was not significantly different. In this randomized study, matinib interruption in non-progressing GIST patients was associated with faster resistance to imatinib and shorter OS in the long-term.