Table IV.
Outcomes of de novo AML patients who at CR cleared DNMT3A R882 and all other mutations, patients who retained only a DNMT3A R882 mutation and patients who retained a DNMT3A R882 mutation and had one or more other mutations.
| End point |
DNMT3A CR Group 1* (n = 14) |
DNMT3A CR Group 2a* (n = 30) |
DNMT3A CR Group 2b* (n = 12) |
P† |
P‡ Gl vs. G2a |
HR (95% CI) Gl vs. G2a |
P‡ Gl vs. G2b |
HR (95% CI) Gl vs. G2b |
P‡ G2a vs. G2b |
HR (95% CI) G2a vs. G2b |
|---|---|---|---|---|---|---|---|---|---|---|
| Complete remission, n (%) | 14 (100) | 27 (100) | 15 (100) | – | – | – | – | – | – | – |
| Disease-free survival | ||||||||||
| Median, years | 0·6 | 1·6 | 0·6 | 0·06 | 0·35 | 1·63 (0·63–4·21) | 0·99 | 0·68 (0·24–1·95) | 0·06 | 0·42 (0·17–1·05) |
| % disease-free at 3 years (95% CI) | 29 (9–52) | 40 (23–57) | 17 (3–41) | |||||||
| % disease-free at 5 years (95% CI) | 29 (9–52) | 40 (23–57) | 8 (0–31) | |||||||
| Overall survival | ||||||||||
| Median, years | 1·3 | 3·8 | 1·3 | 014 | 0·40 | 1·67 (0·61–4·53) | 0·99 | 0·78 (0·26–2·35) | 014 | 0·47 (0·18–1·23) |
| % alive at 3 years (95% CI) | 36 (13–59) | 57 (37–72) | 33 (10–59) | |||||||
| % alive at 5 years (95% CI) | 36 (13–59) | 46 (27–63) | 25 (6–50) |
AML, acute myeloid leukaemia; CI, confidence interval; CR, complete remission; HR, hazard ratio; n, number.
Group 1 (G1) is defined as patients whose DNMT3A R882 mutation cleared below the VAF cut-off of 3% in their remission sample and who had no other AML mutation. Group 2a (G2a) is defined as patients with a DNMT3A R882 mutation with a VAF >3% in their remission sample but who cleared all other AML mutations. Group 2b (G2b) is defined as patients with a DNMT3A R882 mutation with a VAF >3% in their remission sample and persistent or newly acquired other AML mutations.
Overall P-value pertaining to comparison of all three patients groups.
Pairwise comparisons are adjusted from multiple comparisons using the Bonferroni method (Westfall et al, 1999).