The Impact of gain1q on Mutational Structure and Clonal Evolution in a Uniformly Treated High-Risk Series of Patients at First Relapse

Introduction

In Multiple Myeloma (MM) the emergence of treatment resistant clones is a characteristic feature of relapse and this is particularly so for high-risk cases. A key driver event mediating progression, risk status and relapse is gain(1q) (1q+). We report on the impact of 1q+ on the genetic profile seen at first relapse in a uniformly treated, newly diagnosed series of 56 patients enrolled to the NCRI Myeloma XI Trial.

Methods

We included 56 high risk patients, defined as relapse within 30 months of maintenance randomisation (median 19 months, range 8-51). Of the 56 patients, 30 received lenalidomide maintenance and 26 were observed. Whole exome sequencing was conducted at presentation and relapse to a median depth of 122x for tumour samples and 58x for controls. Libraries were prepared using the SureSelectQXT sample prep kit and SureSelect Clinical Research Exome kit. MuTect was used to determine gene variants and SciClone clustering was undertaken to map mutation variant allele frequencies. MANTA was used to determine translocations and Sequenza for copy number aberrations. Clonal structure and mechanisms of clonal evolution were assessed using kernel density estimation of the cancer clonal fraction for all mutations. Wilcoxon matched-pairs signed rank tests (2-sided) were used to determine the significance between paired data sets, including mutational load. Fishers exact test was used to determine the difference between two nominal variables.

Results

We looked at mutational, structural and clonal evolution events in all patients based on 1q+ status at relapse. At diagnosis, 34% (19/56) patients had evidence of 1q+, increasing to 46% (26/56) at relapse, with all patients harbouring 1q+ at presentation having the lesion at relapse.

There was a significantly higher non-synonymous mutational load at relapse in patients with 1q+, 107 vs 126 (p=0.047), compared to those without 1q+, 36 vs 44 (p=0.140). Twenty two genes known to be significant in MM and mutations within the genes known to be important in IMiD mechanism of action were reviewed. Of the patients with 1q+, 92% (24/26) had at least one mutation during the course of the disease, compared to 77% in those without 1q+ (p=0.15).

The impact on tumour suppressor gene regions including deletions of chromosome 1p, 13, 14 and 17p was analyzed. Of the patients with 1q+, 77% (20/26) of patients had a deletion of one of these regions during the disease course, compared to 57% (17/30) of patients without 1q+ (p=0.16). At relapse a change in the profile of these lesions was noted in 23% (6/26) patients with 1q+, compared to 20% (6/30) patients without 1q+ (p=1).

Translocations involving MYC (t MYC) were also determined and found in 27% (7/26) of patients with 1q+ and 27% (8/30) of patients without (p=1). As with 1q+, t MYC was always preserved at relapse.

Mechanisms of evolution leading to relapse were established for all patients. Branching and linear evolution predominated, noted to be the mechanism leading to relapse in 88% (23/26) patients with 1q+ and 83% (25/30) without (p0.71). Stable evolution was noted in the remaining patients. 1q+ occurring as a new event at relapse was associated with branching or linear evolution in all patients (n=7), consistent with a change in clonal structure.

Conclusion

These data reveal that 1q+ is conserved throughout the disease course, suggesting it imparts a survival advantage and treatment resistant phenotype to the clone(s) containing it. The presence of 1q+ is associated with a significant increase in mutational load at relapse and a greater incidence of tumour suppressor gene structural deletions, mechanisms that may contribute to clonal evolution and therapeutic escape.