Real-world insights into ATRA-based maintenance therapy and prognostic markers of relapse in acute promyelocytic leukemia Free

Real-world insights into ATRA-based maintenance therapy and prognostic markers of relapse in acute promyelocytic leukemia

Background:

Acute promyelocytic leukemia (APL) is highly curable; however, relapse remains a clinically significant challenge. Although maintenance therapy was historically thought to be preventative of relapse, very long-term results of trials such as AIDA-0493 reported that addition of a maintenance does not provide any survival advantage in measurable residual disease (MRD) negative patients. We aimed to evaluate the long-term clinical impact of ATRA-based maintenance therapy and other prognostic factors of relapse in a multicenter real-world cohort of Korean APL patients treated with ATRA+Idarubicin (AIDA) protocol.

Methods:

We retrospectively analyzed 286 APL patients who achieved complete remission following AIDA-based chemotherapy from 2002–2024 across five South Korean tertiary centers. Propensity score (PS) matching was performed to analyze survival differences between maintenance therapy and observation groups, along with other prognostic factors. The PS was estimated using a multivariable logistic regression model incorporating age, sex, and Sanz risk score as covariates. Patients were matched in a 1:3 ratio using nearest-neighbor matching without replacement. Covariate balance between groups was assessed using standardized mean differences and no substantial residual imbalance was observed. Survival outcomes were estimated using the Kaplan–Meier method, and differences between groups were assessed with the log-rank test. Since PS matching was performed, we performed univariate analysis to compare outcomes between matched groups. Key prognostic variables including maintenance therapy, Sanz risk, MRD, FLT3 mutation status, and sex were assessed.

Results:

252 patients (88.1%) received maintenance following consolidation, whereas 34 patients (11.9%) did not. Among those who received maintenance, 210 patients (73.4%) were treated with ATRA alone, while 42 patients (14.7%) received ATRA in combination with low-dose chemotherapy (6-mercaptopurine and methotrexate). PS matching of 1:3 matched 34 patients who did not receive any maintenance therapy and 102 patients who did. The comparison of baseline characteristics between patients who only underwent observation and who underwent maintenance therapy did not show any differences between the two groups.

After a median follow-up of 76.82 months, the 10-year relapse-free survival (RFS) was 85.2% in total patients. RFS performed in the PS matched cohort showed non-significant outcomes in observation vs maintenance group (75.3% vs 81.9%, p=0.193). Cumulative incidence of relapse treating death as a competing event, was not significantly different between the two groups (18% vs 24%, p=0.45). RFS based on post consolidation MRD during follow up was significantly lower in the positive group (26.8% vs 87.4%, p<0.001). RFS based on FLT3 mutation status showed worst prognosis in FLT3-ITD patients, followed by FLT3-TKD and no mutation (75.1% vs 78.6% vs 92.5%, p=0.013). Male patients showed poorer RFS than female patients (75.7% vs 93.9%, p<0.001).

Moreover, in the univariate analysis of the PS matched cohort, ATRA maintenance therapy did not significantly decrease relapse risk (HR = 0.56; 95% CI: 0.23–1.37; p = 0.2). In contrast, post consolidation MRD showed increased risk (HR = 20.16; 95% CI: 7.62–53.32; p < 0.001) and FLT3-ITD mutation status also showed increased risk of recurrence (HR = 4.11; 95% CI: 1.1–15.31; p=0.035). Sanz risk was borderline significant (HR = 2.17; 95% CI: 0.94–4.89; p=0.07) and male patients showed increased risk of recurrence than females (HR = 5.96; 95% CI: 1.4–25.46; p = 0.016).

Conclusion:

Post-consolidation MRD positivity emerged as the most significant predictor of relapse in APL, while ATRA-based maintenance therapy did not confer a significant long-term benefit in relapse prevention. In addition, FLT3 mutation status may also serve as potential prognostic markers for relapse. These findings support the integration of molecular profiling and MRD-guided approaches to tailor individualized post-remission strategies in APL. Although current treatment protocols recommend ATO+ATRA without maintenance for low-risk patients, our results remain clinically relevant in settings where access to ATO is limited or unavailable, and where healthcare resources constrain the use of long-term maintenance therapies.