The Impact of Maintenance Lenalidomide on the Mutational Status of the Myeloma Clone at Relapse in the NCRI Myeloma XI Trial for Newly Diagnosed Multiple Myeloma Patients (NDMM)

Introduction

With a multifactorial mechanism of action and excellent PFS associated with prolonged exposure, lenalidomide (len) is an attractive candidate for maintenance therapy. Len exerts its action by interaction with cereblon (CRBN) which forms a ubiquitin ligase complex with cullin-4A (CUL4), damaged DNA binding protein 1 (DDB1) and regulator of cullins 1 (ROC1). Downstream effects are mediated via Ikaros, Aiolos, MYC, IRF4, basigin (BSG) and solute carrier family 16 member 1 (SLC16A1).

The impact of selective pressure on MM clonal architecture and mutational load has not been assessed. Although not a DNA damaging agent there is an apparent effect of maintenance len increasing the risk of second cancers and a suggestion that it could select for aggressive clones in high risk disease. We addressed the hypothesis that len may increase the rate of mutation at relapse by performing whole exome sequencing (WES) on 70 paired presentation/relapse samples from patients enrolled to the Myeloma XI trial (MXI), 35 of whom received maintenance Len and 35 not.

Methods

WES was performed to a median depth of 125x on 70 presentation/relapse pairs from patients enrolled to the MXI trial. MXI is a phase III study comparing thalidomide, len and bortezomib induction combinations and len vs observation maintenance treatment in both transplant eligible (TE) and transplant non-eligible (TNE) NDMM patients.

We selected patients who had completed induction +/- ASCT and been randomised to receive maintenance therapy with len or observation. All patients had disease progression determined by IMWG criteria at the time of the relapse sample.

Of the 70 patients, 30 were enrolled in the TE pathway and 40 in the TNE pathway. The median time to relapse following maintenance randomisation was 323 days (296 len vs 325 observation). 35 patients (50%) achieved a CR as their best response, 26 (37%) a VGPR and 9 (13%) a PR. The median age was 66 and 69 for those receiving len and those being observed respectively. High risk disease status was confirmed in 33 (47%) patients at presentation (≥ 1of t(4;14), t(14;16), t(14;20), +1q, -17p, -1p).

Results

The median number of non-silent mutations (NSM) found at presentation and relapse was 37 and 41 respectively (p=0.25). In patients receiving len maintenance the median number of NSM at presentation was 37 vs 34 at relapse (p=0.69). In those being observed the median number of NSM at presentation was 42 vs 52 at relapse (p=0.21). Mutations in genes important in myeloma pathogenesis seen in more than one patient at presentation included KRAS (16), NRAS (14), DIS3 (6), HIST1H1E (2), RB1 (2), EGR1 (2), TP53 (2) and FAM46C (2). These were seen in a total of 37 (53%) patients. One patient had both an NRAS and KRAS mutation. At relapse 7 patients lost mutations (NRAS (3), KRAS (3), DIS3 (1)) and 6 patients gained mutations (KRAS (2), NRAS (2), TP53 (1), FAM46C(1)).

Paired presentation/relapse copy number (CN) data (MLPA) was available for 38 patients (54%). At relapse there was evidence of a change in CN status with 5 (13%) patients gaining CN changes associated with high risk (gain 1q (4), del 17p (1).

Six patients (9%) were found to have mutations in genes associated with len action; CRBN (1), IRF4 (1), DDB1 (2), SLC16A1 (2). No mutations were found in Ikaros, Aiolos, ROC1, CUL4 or BSG.

The CRBN mutation was found at relapse only, in a patient who had achieved a CR and undergone 232 days of len maintenance. The IRF4 mutation was seen at presentation and relapse in a patient who achieved CR and received 754 days of len prior to relapse. Both patients with DDB1 mutations received len induction, ASCT, achieved CR and were randomised to observation. In one patient the mutation was seen at presentation and relapse whilst in the other only at relapse. Both patients with mutations in SLC16A1 were treated with len induction and ASCT to CR. In one patient, randomised to observation the mutation was seen at both time points and they relapsed after 156 days. The other, with the mutation present at presentation only was randomised to len maintenance and relapsed after 256 days.

Conclusions

This is the largest study comparing the genetics of presentation/relapse myeloma in a len treated population. Overall, the number of mutations at presentation vs relapse remained stable. We show that len does not affect the mutational load at relapse but may select for mutations conferring len resistance although at present further analysis is required to confirm this.