Don't Interrupt Imatinib in GIST Patients

ASCO-Interrupting imatinib (Gleevec) therapy in advanced gastrointestinal stromal tumor (GIST) patients who have had a long-lasting response results in a high risk of rapid progression, according to results of a clinical trial (BFR14) presented at the American Society of Clinical Oncology 43rd Annual Meeting (abstract 10005). That risk, said Axel Le Cesne, MD, speaking for the French Sarcoma Group, persists regardless of the quality of prior response to therapy. "The optimal duration of imatinib therapy in responding patients with advanced GIST is unknown," Dr. Le Cesne said.

The ongoing trial has enrolled 338 patients with advanced metastatic GIST who had received imatinib for 1 year. Initially, responders and those with stable disease were randomized to continued therapy with imatinib 400 mg (CONT) or to therapy interrupted until progressive disease followed by resumed dosing at 400 mg (STOP).

As previously reported, in 58 randomized patients, median progression-free survival (PFS) was 6.1 months among the STOP patients vs 29 months for those who had continued imatinib therapy (P < .0001). Dr. Le Cesne noted that 89% of STOP patients who progressed achieved tumor control after reintroduction of imatinib. Overall survival was similar between the groups.

3-Year Randomization

In the current study, patients whose response to 400 mg of imatinib has been sustained for 3 years are again randomized to CONT or STOP. In his ASCO talk, Dr. Le Cesne pre-sented an interim analysis of the first 50 patients.

At 1 year after this second randomization, there were 12 events in the 25 STOP patients (PFS 20.2%) and 1 event in the 24 CONT patients (PFS 91.7%, P = .0013). Progressive disease rates in the STOP arm were 47%, 64%, and 85% at 6 months, 9 months, and 1 year, respectively. There were no deaths. The randomization was stopped at this point, and STOP patients were offered imatinib.

Dr. Le Cesne pointed out that the 1-year PFS curves in the patients randomized to STOP after 1 year and after 3 years of imatinib response were similar at 25% (1 year) and 20.2% (3 year).

Prolonged Responses

Dr. Le Cesne and his colleagues also asked whether prolonged responding patients are different from the initial nonrandomized population (n = 276). They found female sex, small bowel primary site, and liver involvement only were more common in the patients still responding at 3 years (only female sex was significant).

In addition, mutational analysis conducted in 49% of patients at the 3-year randomization showed the exon 11 mutation to be present in 83% vs 8.5% each for exon 9 and wild type. In the exon 11 group, 90% of mutations were proximal. "Proximal mutations of exon 11 confers to the patient a high sensitivity to imatinib," he said.

Dr. Le Cesne concluded that imatinib interruption in responding advanced GIST patients after 3 years "results in a high rate of rapid progression whatever the type of response, including complete remissions. . . . Interruption is associated with reduced PFS and cannot be recommended outside of clinical trials."

He noted further that the impact of the reintroduction of imatinib at the same dose on tumor control is currently being investigated. The impact of imatinib interruption on overall survival after 3 years of treatment is yet unknown.

ASCO discussant Peter Reichardt, MD, Bad Saarow, Germany, commented on the similar PFS pattern in the 1- and 3-year STOP arms. "This clearly leads to the question of whether we already need to think about a third-generation adjuvant trial with treatment duration beyond 3 years," he said.

Dr. Le Cesne noted that, in fact, a new randomization of the BFR14 study has been proposed in nonprogressing patients at 5 years.