In the DCAG group, the univariate analysis determined that the following were associated with OS and EFS (Table 2, Figure 2C and D): ECOG score ≥3 (OS, HR 7.66, 95% CI 1.39–42.16, P=0.019; EFS, HR 6.04, 95% CI 1.08–33.89, P=0.041); NR after the second induction therapy (OS, HR 4.12, 95% CI 1.10–15.40, P=0.035; EFS, HR 4.34, 95% CI 1.14–16.45, P=0.031); and FLT3-ITD (OS, HR 4.69, 95% CI 1.89–11.69, P=0.001; EFS, HR 3.80, 95% CI 1.49–9.73, P=0.005). Also in the DCAG group, the multivariate analysis (Table 2) showed that the following were significantly associated with shorter OS: ECOG score ≥3 (HR 66.75, 95% CI 1.49–298.53, P=0.030), and FLT3-ITD (HR 20.08, 95% CI 1.18–342.10, P=0.038).

Prognostic significance of clinical features and gene mutations in patients receiving IA induction chemotherapy

In patients receiving IA induction chemotherapy, the log-rank test indicated that NR after the first (OS, P=0.0004; EFS, P=0.001) or second induction therapy (OS, P=0.0003; EFS, P=0.005) defined a set of subgroups with poor OS and EFS (Figure S1). The univariate analysis showed that the following were significantly associated with OS and EFS (Table 3Figure 3A and B): NR after the first induction therapy (OS, HR 4.53, 95% CI 2.03–10.10, P=0.0004; EFS, HR 3.97, 95% CI 1.79–8.80, P=0.012) or second induction therapy (OS, HR 11.33, 95% CI 5.13–25.03, P=0.0003; EFS, HR 13.02, 95% CI 5.87–28.86, P=0.001); NCCN poor-risk status (OS, HR 3.37, 95% CI 1.35–8.41, P=0.009; EFS, HR 3.53, 95% CI 1.41–8.82, P=0.007); and no extramedullary infiltration (OS, HR 0.35, 95% CI 0.17–0.71, P=0.004; EFS, HR 0.32, 95% CI 0.16–0.65, P=0.002). In these patients, the univariate analysis also showed that a ECOG score ≥3 (HR 3.99, 95% CI 1.00–15.84, P=0.050) was significantly associated with OS. The multivariate analysis showed that no extramedullary infiltration (HR 0.07, 95% CI 0.02–0.32, P=0.001) and an ECOG score ≥3 (HR 14.49, 95% CI 2.59–83.33, P=0.009) were significantly associated with shorter OS. Significantly associated with EFS were no extramedullary infiltration (HR 0.15, 95% CI 0.05–0.05, P=0.009).

Continue Reading

(To view a larger version of Table 3, click here.)

(To view a larger version of Figure 3, click here.)

Clinical features of patients with mutations in DNA methylation

In this study, the highest rate of mutations were class I (62/161, 38.5%), such as FLT3-ITD, KIT, NRAS, KRAS, and PTPN11 (protein tyrosine phosphatase, non-receptor type 11). The next most frequent were epigenetic modification mutations (60/161, 37.3%) including DNMT3A, TET2, IDH1/2, ASXL1, DOT1L. The third and fourth most frequent mutations were class II (42/161, 26.1%; NPM1, CEBPA, RUNX1, GATA2, and ETV6) and tumor suppressor mutations (26/161, 16.1%; WT1, PHF6, and TP53). Also, spliceosome genes, cohesion complex genes, and NOTCH family mutations were identified in 12 (7.5%), 17 (10.6%) and 1 (0.6%) patient, respectively (Figure S2). Spliceosome genes included U2AF1, SRSF2, and SF3B1/2, and cohesion complex genes were STAG2, RAD21, SMC1A, and SMC3.

Gene mutations related to DNA methylation (TET2, DNMT3A, and IDH1/IDH2) are among the most frequently identified in AML, and demethylating agents are effectively used in treating AML (6–8). Therefore, we analyzed the clinical features of the patients harboring these mutations. Of the 161 patients with follow-up data, 37 carried altogether 43 mutations in TET2 or DNMT3A/DNMT3B, with or without IDH1/IDH2. DNMT3A/IDH1 co-mutations were found in 4 patients, DNMT3A/TET2 co-mutations in one patient, and DNMT3A/IDH2 co-mutation in one patient. Thirty-one patients carried only one DNA methylation-related mutation. Compared with the 124 patients without DNA methylation-related mutations (age 40.9 years, favorable risk status 19.83%, ≥2 mutations 59.7%), the 37 patients with these mutations were significantly older with progressive diseases (53.5 years; P=0.01), with lower favorable risk status (2.86%; P=0.027), and were more likely to have ≥2 mutations (91.9%; P=0.0004). However, the 2 groups were similar in EFS and OS (P=0.36 and P=0.47, respectively).

In the DCAG group specifically, there were no significant differences in OS or EFS (P=0.57 and P=0.48, respectively) between patients without DNA methylation-related mutations and those with such mutations (TET2, DNMT3A, and IDH1/IDH2).

Hematopoietic toxicity and treatment-related death

Grade 4 neutropenia and thrombocytopenia were universal in the study population (Table S4).

In the DCAG group specifically, after the first cycle of induction chemotherapy the median recovery times of neutrophils and platelets were 15.8 and 13 days, respectively. Platelet recovery (≥20×109/L) typically preceded WBC count recovery after induction chemotherapy for AML. The pace of platelet recovery was generally brisk, with a median of 13 days for the platelet level to rise higher than 20×109/L. The most common grade-3 or grade-4 adverse events were related to myelosuppression. No patients in the DCAG group died during the induction therapy and no subject required to transfer to intensive care.

READ FULL ARTICLE Curated publisher From Dovepress