Specific Identification of High Risk Disease Using Molecular Profiling By Mymap (Myeloma MLPA and translocation PCR) of 1,036 Cases

Introduction

Identifying molecular high risk myeloma remains a diagnostic challenge. We previously reported co-segregation of >1 adverse lesion [t(4;14), t(14;16), t(14;20), gain(1q), del(17p)] by iFISH to specifically characterise a group of high risk patients (Boyd et al., Leukemia 2012). However, implementation of this approach is difficult using FISH because of its technical limitations. We recently developed and validated a novel high-throughput all-molecular testing strategy against FISH (MyMaP- Myeloma MLPA and translocation PCR; Kaiser MF et al., Leukemia 2013; Boyle EM et al., Gen Chrom Canc 2015). Here, we molecularly characterised 1,036 patients from the NCRI Myeloma XI trial using MyMaP and validated the co-segregation approach.

Materials, Methods and Patients

Recurrent translocations and copy number changes were assayed for 1,036 patients enrolled in the NCRI Myeloma XI (NCT01554852) trial using CD138+ selected bone marrow myeloma cells taken at diagnosis. The trial included an intensive therapy arm for younger and fitter and a non-intense treatment arm for elderly and frail patients. Analysis was performed using MyMaP, which comprises TC-classification based multiplex qRT-PCR and multiplex ligation-dependent probe amplification (MLPA; MRC Holland). Median follow up for the analysis was 24 months.

Results

Adverse translocations [t(4;14), t(14;16), t(14;20)] were present in 18.2% of cases, del(17p) in 9.3%, gain(1q) in 34.5% and del(1p32) in 9.4% of cases. All adverse lesions were associated with significantly shorter PFS and OS by univariate analysis (P<0.05 for all).

Of the 1,036 analysed cases, 13.5% carried >1 adverse lesion, 33.9% had one isolated adverse lesion and 52.6% had no adverse lesion. Presence of >1, 1 or no adverse lesion was associated with a median PFS of 17.0, 23.9 and 30.6 months (P=3.0x10^-9) and OS at 24 months of 67.9%, 75.0% and 86.0% (P=1.8x10^-7), respectively.

Del(1p) was associated with shorter PFS and OS for the intensive, but not for the non-intensive therapy arm and was independent of the co-segregation model by multivariate analysis regarding OS (P=0.006). We thus included del(1p) as an additional adverse lesion in the model for younger patients. The groups with >1 (19.4% of cases), 1 (31.1%) and no adverse lesions (49.5%) were characterised by median PFS of 19.4, 29.4 and 39.1 months (P=1.2x10^-10) and median 24-months survival of 73.8%, 86.4% and 91.5% (P=1.4x10^-6), respectively. Hazard Ratio for >1 adverse lesion was 3.0 (95% CI 2.1-4.1) for PFS and 3.8 (95% CI 2.2-6.5) for OS.

By multivariate analysis, co-segregation of adverse lesions was independent of ISS for PFS /OS in the entire group of 1,036 cases and in the intensive treatment arm. We integrated adverse lesions and ISS into a combined model defining High Risk (>1 adv les + ISS 2 or 3; 1 adv les + ISS 3) and Low Risk (no adv les + ISS 1 or 2; 1 adv les + ISS 1) and the remainder as Intermediate Risk. The High Risk, Intermediate Risk and Low Risk groups of the total cohort included 11.2%, 41.2% and 41.6% of cases with median PFS of 15.8, 19.8 and 35.2 months (P<2.2x10^-16) and median OS at 24 months of 62.9%, 73.7%, and 90.7% (P=4.0x10^-14), respectively.  Integration of ISS into the model for younger patients resulted in highly specific identification of a High Risk group (15.6% of cases) with HR 3.8 (CI 2.6-5.4) for PFS and 6.2 (CI 3.3-11.6) for OS.

Conclusions

Co-segregation analysis of adverse genetic lesions is a specific molecular risk stratification tool which has now been validated in two large independent trials including a real-world population of all age groups (UK MRC Myeloma IX; NCRI Myeloma XI; total 1,905 patients). MyMaP is a validated all-molecular analysis approach that makes the otherwise technically challenging assessment of multiple genetic regions by FISH accessible using standard laboratory equipment without bioinformatics requirements.