Frequent PML-RAR α Mutations in Relapse Patients on Acute Promyelocytic Leukemia (APL) Intergroup Phase III Trial C9710.

Mutations in the ligand binding domain (LBD) of the RARα-region of PML-RARα were detected in 33% of tested patients who relapsed after treatment with all-trans retinoic acid (ATRA) on the first North American intergroup APL trial INT0129 (E2491). We hypothesized that the incidence of these mutations at the time of first morphological relapse would be reduced on the successor trial C9710, since the new treatment regimen (concurrent ATRA and chemotherapy for induction, first consolidation with two cycles of arsenic trioxide [ATO] in 50% of patients, and 2 subsequent cycles of consolidation with ATRA and daunorubicin) might prevent the early selection of ATRA-resistant PML-RARα mutant subclones. Procedures for making this assessment from low-density bone marrow (BM) and peripheral blood (PB) cells are well-established. In patients found to have a PML-RARα mutation, pre-relapse BM and PB samples collected for monitoring minimal residual disease (MRD) are also being tested for the presence of the mutation, using a mutation-specific real-time PCR assay, in order to assess the dynamics of mutant subclone emergence. Reconstruction experiments in which first-round, allele-specific PCR products from relapse mutant cells were serially diluted in PCR product from non-mutant APL cells documented a capacity to detect 1 mutant template among 10³ to 10⁴ non-mutant templates. 9/18 (50%) relapse patients, representing about half of all C9710 relapses to date, tested positive for PML-RARα LBD mutations by DNA sequence analysis, usually with virtual replacement of non-mutant PML-RARα. 9/10 mutations (1 patient had a double mutation) were missense, 2 repeats of previously reported mutations (Arg276Trp, Gly289Arg) and 5 novel mutations (Leu224Pro, Lys238Glu, Ile273Phe, Arg276Gln x2, Gly289Glu x2); 1 was an in-frame 3-codon deletion (Δ412–414). Serial monitoring of post-consolidation samples for the relapse mutation in 3 patients did not detect mutations until or just prior to clinical relapse. The median time (months) to relapse from post-consolidation therapy assessment for MRD was 9.5 (range, 4–36) for mutant patients and 9 (range, 1.5–16) for non-mutant patients. Two of the serially monitored patients relapsed 7 and 24 months after the termination of 12-month maintenance therapy with ATRA ± methotrexate (MTX) and 6-mercaptopurine (6MP). This, together with the lack of a detectable mutant subclone at the conclusion of maintenance therapy, indicates that proximate ATRA selection was not involved in the emergence of the predominant PML-RARα mutant subclone in these 2 patients. Overall, these results support two conclusions:

1. contrary to our pre-study hypothesis, the incidence of PML-RARα mutations was not reduced in patients who relapsed on protocol C9710, but, since the patient assignment to consolidation therapy with ATO and/or maintenance therapy with MTX/6MP has not been disclosed, a difference between the randomized treatment groups is possible and

2. the late emergence of PML-RARα mutant subclones suggests that the mutations provide an intrinsic APL cell growth/survival advantage that likely contributes to the probability of disease recurrence.