Discontinuation of HMA Therapy after Achieving Complete or Partial Response: Retrospective Analysis of Survival after Long-Term Follow up

Introduction: Hypomethylating agents (HMA), such as decitabine or 5-azacitidine, are the first-line treatment for higher-risk myelodysplastic syndromes (MDS) and are commonly used in refractory or elderly acute myeloid leukemia (AML) patients. There is anecdotal evidence that early discontinuation of HMA therapy is associated with relapse, but this has not been systematically assessed. In this analysis, we study the outcomes of patients with MDS and AML treated with HMA in whom therapy was interrupted while in response.

Methods: We included patients treated on 3 clinical trials of HMAs in which therapy discontinuation was considered after 12-24 months if a sustained response was achieved. We calculated progression-free survival (PFS) and overall survival (OS) after stopping therapy, and as a secondary objective, we explored clinical variables associated with outcome.

Statistical analysis was performed with SPSS v.22. PFS was defined as time from drug discontinuation until progression or death from any cause; OS was defined as time from drug discontinuation to death from any cause. Estimated PFS and OS curves were calculated by the Kaplan-Meier method, and the log rank test was used to identify variables with influence on survival data.

Results: From the initial 173 patients accrued in these trials, we identified 16 patients who had achieved complete or partial response (CR/PR) and who electively stopped treatment while response was maintained. Median age was 68 years (51-80). Diagnosis was AML in 10 patients (62%) and intermediate-2/high-risk MDS in 6 patients (38%); 81% of patients had not received any previous treatment.

Patients received a median of 6 courses (1-14) and all achieved either CR (n=15; 94%) or PR (n=1; 6%). The median number of cycles until response was 1 (1-4). Therapy was stopped after receiving the maximum courses of treatment scheduled on protocol for 7 patients (44%) and by patient decision for the other 9 (56%). There were no treatment discontinuations due to side effects. Patients who received their whole protocol treatments received significantly more courses of therapy (12 vs 6; p=0.001) and tended to have longer times to progression or relapse from off-treatment date (24 vs 5.78; p=0.08).

For the whole series, the estimated median OS and PFS were 15 months (95% confidence interval [CI]: 6-24) and 4 months (95% CI: 2-6), respectively. The main variable with impact on both OS and PFS was total number of courses of therapy. Patients who received more than 12 cycles of HMA showed significantly better median OS (20 months [95% CI: 12-27] vs 4 months [95% CI: 1-18]; p=0.043) and tended to have longer PFS (estimated PFS at 12 months: 50% vs 17%; p=0.062). Cytogenetics also had a significant influence on OS; those patients with high-risk cytogenetics had a poorer OS (estimated OS at 12 months: 33% vs 69%; p= 0.046).

At last follow up, 2 patients are still alive and receiving active treatment, with a median follow up of 104 months (92-117). Eleven patients (69%) relapsed or progressed after discontinuing treatment, with a median of 4 months (2-68) from last course of therapy, and 5 died from a different cause while in response. After relapse, 7 out of 11 patients received other treatments for MDS/AML, and 2 patients responded to those further treatments. HMA were reintroduced in 6 patients, and only 1 was sensitive to HMA after the first failure. In total, 14 patients (88%) died, and progressive disease (PD) was the most frequent cause (9/14, 65%).

Conclusion: When therapy was interrupted, loss of response was rapidly observed in most of patients, with a median of 4 months from end of treatment to progression or death. Additionally, although the number of patients was limited, we have identified two variables that could be associated with outcome: number of therapy courses administered before stopping therapy and cytogenetics. Based on these data and current information, discontinuation of HMA treatment should not be considered in the absence of any serious adverse event. In cases of discontinuation of treatment, number of courses received and cytogenetic risk seem to be important factors to predict outcome.