Uncommon Mutations in PML-RARα Associated with Poor Outcome after First Relapse in APL.

Background. Mutations in the RARα-region ligand binding domain (LBD) of PML-RARα were detected in 30 – 40% of patients (pts) tested after first relapse from ATRA-chemotherapy (CT) regimens in 2 North American Phase III trials. However, no follow-up data were reported to assess whether this affected subsequent outcome.

Methods. In the current study, 8 relapse pts were tested by previously-published techniques for PML-RARα LBD mutations after relapse from ATRA-CT regimens of the EAG. Treatment was according to the APL2000 protocol: randomization to induction with ATRA+DNR with AraC (arm-A) or no AraC (arm-B) followed by 2 consolidation courses with DNR with (A) or without (B) AraC and then 2 years maintenance with ATRA for 15 days every 3 months, as well as, 6MP and MTX (

Results.

Mutations were observed in the reported incidence range: 3/8 pts (37.5%). However, the findings were unusual, since 2/3 mutations were deletion mutations, while only 2/30 reported mutations were deletions, and since 2/3 pts harbored 2 mutant subclones, while only 1/30 such cases has been reported. Also, the nature of the mutations was unusual. In pt 7, the ΔK227∅S229/M227 mutation was atypically located in the amino-region of the LBD. In pt 4, a very atypical deletion/insertion (del/ins) mutation began in the hinge RARα-region at S157, extending 630 nt to R367, and was associated with a 547 nt insertion from the beginning of intron 4 that changed the translational reading frame, encoding a truncated protein lacking the entire LBD. Most notably, the 3 pts harboring these uncommon mutations either failed to achieve a second complete remission (CR) or suffered subsequent relapses, leading to death in 2 pts from resistant disease (pts 4 & 7), while 0/5 non-mutant pts relapsed (pt 5, death in CR2 was treatment-related).

Discussion. With such a small sample size, it is not possible to determine whether the uncommon mutations occurred by chance or related to some unrecognized element in the EAG pts. No readily apparent basis could be identified: the modest treatment dose-schedule differences from other Phase III studies using common therapeutic agents seems an improbable cause; one each of the mutant/non-mutant pts had an extra chromosome (trisomy 8); 2 of the non-mutant cases had treatment-related APL. Although further study is needed to determine whether the poor outcome in 3 pts was related to the uncommon nature of their mutations, these results suggest that testing for PML-RARα mutations at first relapse may be of prognostic and potential therapeutic value.