Metronomic dose decitabine therapy for TP53 complex karyotype myeloid neoplasms

Abstract

Abstract Introduction: Hypomethylating agent and venetoclax (VEN)–based therapy has transformed the treatment landscape for older adults with AML, offering improved remission rates and survival compared to traditional chemotherapy. However, patients with TP53 mutations (TP53mut) and complex karyotype (CK), continue to experience dismal outcomes characterized by primary resistance, early relapse, and high therapy-related toxicity. In this population, the cytotoxic effect of standard-dose azacitidine and VEN is poorly tolerated, often leads to treatment discontinuation or early mortality. Low-dose, metronomic decitabine (LoDec, 0.1–0.2 mg/m²/WEEK) offers a non-cytotoxic alternative that depletes DNMT1 and retains epigenetic-modifying activity. LoDec has demonstrated efficacy in relapsed AML and in combination with VEN for de novo disease, but its role in the frontline treatment of TP53mut CK MDS/AML has not been defined. Here, we evaluate the safety and efficacy of LoDec/VEN in this high-risk population. Methods: We retrospectively reviewed all patients with newly diagnosed TP53mut CK myeloid neoplasms (MDS and AML) treated at Johns Hopkins between 2018 and 2024.TP53mut were identified by targeted next-generation sequencing, and CK was defined as the presence of ≥3 cytogenetic abnormalities. Patients were grouped by frontline therapy: LoDec plus VEN (400 mg, one dose per WEEK), standard azacitidine (Aza, 75 mg/m² × 7 days) plus VEN (400 mg/day), or standard decitabine (StDec, 20 mg/m² × 5 days). We collected baseline demographic and clinical data at diagnosis and evaluated transfusion independence (TFI, ≥4 weeks without RBC or platelet transfusions), overall survival (OS), 90-day mortality (90dM), and hospitalization burden (total inpatient days, proportion of days hospitalized over total days survived). Analyses were performed using Kaplan-Meier estimates for OS, ANOVA for comparison of continuous variables, and chi-square tests for categorical variables. Results: We identified 94 patients with TP53mut CK myeloid neoplasms who received the above therapies: 20 received first-line LoDec/VEN (4 MDS, 5 AML/MDS, 11 AML), 57 received Aza/VEN (2 MDS, 10 AML/MDS, 45 AML) and 17 received StDec (1 MDS, 8 AML/MDS, 8 AML). Patients in the LoDec/Ven group were similar in age (mean 73 years, p=0.23), gender (50% male, p=0.24), and ECOG (mean 2.05, p=0.2) when compared to those in the Aza/VEN (mean age 69, 70% male, mean ECOG 1.63) and StDec (mean age 68, 70% male, mean ECOG 2.00) groups. Patients in the LoDec/Ven group had on average higher Charlson co-morbidity index (CCI) scores compared to the other groups (LoDec/Ven 7.9, Aza/Ven 6.3, StDec 6.4, p<0.05). A notable feature at presentation for TP53mut CK myeloid neoplasm was the occurrence of “disease-related fever,” defined as T > 38C without clinical, laboratory, or radiographic evidence of infection suggesting a potential association between TP53mut CK disease biology and inflammatory cytokine production. This clinical feature was similarly distributed (p=0.3) in all treatment groups (15% LoDec/VEN, 26% Aza/VEN, 54% StDec). Median OS for patients treated with LoDec/VEN was 203 days, significantly different from Aza/VEN (90 days) and StDec (371 days) (p=0.01). LoDec/Ven group had significantly lower 90dM (15%, p<0.01) compared to StDec (23.5%) and Aza/VEN (51.7%). These results underscore the poor outcomes in this high-risk population and the relative tolerability of LoDec/VEN. Hospitalization burden was significantly lower in the LoDec/Ven cohort, with patients spending in average 19.7 days (p<0.01) in the hospital (15% of days survived, p<0.01), versus 31.9 days (40%) with Aza/VEN and 50 days (40%) with StDec. TFI was achieved in 47% of LoDec/VEN-treated patients, 47% of the StDec patients, and 30% of the Aza/VEN patients (p=0.28). Conclusions: Our data suggest that LoDec/Ven is a reasonable frontline approach for TP53mut CK myeloid neoplasms. Currently, many providers favor Aza/VEN or StDec for patients deemed fit to proceed to BMT. However, only 6/74 patients treated with these regimens ultimately underwent transplant, and none survived beyond 14 months. In contrast, those treated with LoDec/Ven had higher CCI at diagnosis. Nevertheless, many achieved TFI but were not properly assessed for depth of response, raising the possibility that some may have achieved remission levels sufficient for more intensive consolidation. A prospective clinical trial is underway to assess this strategy.