Abstract

Imatinib is remarkably effective as single-agent treatment of newly diagnosed chronic myeloid leukemia (CML) in chronic phase (CP) but most of the published data are based on a single multicenter study (acronym IRIS) carried out under the supervision of the manufacturer. We report here an independent validation of the IRIS results. At our institution from June 2000 and August 2006 207 consecutive newly diagnosed CML-CP patients were started on imatinib 400mg within 6 months from diagnosis, of whom 17 were included in the IRIS study. Their median age was 46.3 years. Their Sokal risk score was ‘low’ in 59 (28.5%) patients, ‘intermediate’ in 73 (35.3%) and ‘high’ in 54 (26.1%). The median follow-up was 33 months (range 12–85). By intention to treat analysis the estimated overall survival (OS) and progression-free survival (PFS, defined by progression to accelerated or blastic phase) at 48 months were respectively 90.5% (95CI, 82.1–95.2%) and 90.2% (95CI, 84.5–94.0%) (data censored at the time of allografting). 34 (16%) patients discontinued imatinib at some point while still in CP and received a second generation TK inhibitor or an allograft. The cumulative incidence of complete cytogenetic response (CCyR) and major molecular response (MMolR) were 59.1% (95CI, 52.0–65.9%) and 11.9% (95CI, 8.0–17.4%) by 12 months, and 86.7% (95CI, 79.4–91.7%) and 58.4% (95CI, 47.6–68.5%) by 48 months respectively (data censored at the time of changing to another treatment). The cumulative incidence of loss of CCyR at 48 months was 17.5% (95CI, 9.9–29.1%). The imatinib dosage was increased in 70 (33.8%) patients to 600 mg/day and in 19 (11.6%) patients the dose was further increased to 800 mg/day during follow-up; in 5(2%) patients the dose of imatinib was directly escalated to 800mg/day. The only pre-therapy factor found to be predictive for PFS was the Sokal (and also Euro) score. A 12-month landmark analysis showed that patients who achieved a CCyR had significantly better OS and PFS at 48 months (OS - 95.2% vs 87.1%, p=0.026; PFS - 97.8% vs 87.3%, p=0.007 respectively), while the achievement of a MMolR by 18 months had no significant impact on OS or PFS. Nevertheless patients in CCyR who achieved a MMolR at any time during the follow-up were significantly less likely to lose their CCyR than CCyR patients who maintained higher BCR-ABL transcript levels (RR=0.08, p=0.0003). No plateau was seen either in the OS or in the PFS curves; however when analysis was restricted to patients who were on CCyR at 12 months the OS and PFS curves did achieve plateaux. Five patients (2.4%) achieved a complete molecular response (CMolR) during the follow up. When restricting the analysis to the 31 patients who had at least 4 years of follow-up only 2 patients (6.5%) had a sustained CMolR while on imatinib 400mg/day. In summary our findings confirm the results reported in the recent IRIS update, notably the high rates of CCyR and high rates OS and PFS at 48 months. Moreover, our results suggest that achieving a CCyR before 12 months is strongly associated with better OS, which accords with the recommendations published recently by the European LeukemiaNet. We showed moreover that once a patient has achieved CCyR, a further reduction in the transcript levels to MMolR is still clinically relevant, although the time-frame is not yet clear.

Author notes

Disclosure: No relevant conflicts of interest to declare.