The Clonal Architecture of CXCR4 mutations in Waldenstrom's Macroglobulinemia Shows Highly Variable Subclonal Distribution, and Multiple Mutations within Individual Patients Indicative of Targeted Genomic Instability

Background: Whole genome sequencing (WGS) identified activating CXCR4^WHIM somatic mutations in nearly 30% of patients with Waldenstrom's Macroglobulinemia (WM) (Blood 123(11):1637-46). Both nonsense and frameshift CXCR4^WHIM mutations occur in WM, with over 30 different types of mutations described within the regulatory carboxyl-terminal domain of CXCR4. CXCR4^WHIM mutations almost always occur with activating MYD88 mutations, and impact both disease presentation and treatment outcome (Blood 123(18):2791–6; NEJM 372(15):1430–40.). The clonal architecture of CXCR4^WHIM mutations relative to MYD88 mutations and their role in disease evolution remains to be clarified.

Methods: We used Sanger sequencing and highly sensitive AS-PCR assays that we developed for the most common CXCR4^WHIM mutations (S338X C>A and C>G) to evaluate for CXCR4^WHIM mutations. In conjunction with an AS-PCR MYD88^L265P assay that we previously developed (Leukemia 28(8):1698–707), we also profiled tumor samples for MYD88^L265P and CXCR4^S338X mutations in 164 WM, 12 IgM MGUS, 20 MZL, 32 CLL, 14 MM, 7 non-IGM MGUS patients, and 32 healthy donors. Next generation transcriptome sequencing data was also performed for validation in select cases.

Results: AS-PCR detected CXCR4^S338X mutations in WM and IgM MGUS patients not revealed by Sanger sequencing. By combined AS-PCR and Sanger sequencing, CXCR4^WHIM mutations were identified in 44/102 (43%), 21/62 (34%), 2/12 (17%), and 1/20 (5%) untreated WM, previously treated WM, IgM MGUS and MZL patients, respectively, but not in CLL, MM, non-IGM MGUS patients or healthy donors. Cancer cell fraction analysis in WM and IgM MGUS patients showed CXCR4^S338X mutations were primarily subclonal, with highly variable clonal distribution (median 35.1%, range 1.2%-97.5%; Figure 1). Sanger sequencing identified 3 patients with multiple CXCR4 mutations, which were shown to be compound heterozygous by TA cloning and sequencing of at least 40 clones. The addition of AS-PCR to the Sanger sequencing results also revealed multiple CXCR4^WHIM mutations in many individual patients that included homozygous and compound heterozygous mutations that were validated by next generation sequencing that offered a median of 5,819 (range 5,217-12,235) reads that overlapped the mutated loci.

Conclusions: Taken together, these findings show that CXCR4^WHIM mutations are more common in WM patients than previously revealed by WGS or Sanger sequencing. Moreover, CXCR4 mutations are primarily subclonal supporting their acquisition after MYD88^L265P in WM oncogenesis. The exclusive finding of frameshift and nonsense but not missense variants within the carboxyl-terminal domain of CXCR4 suggests that significant selection pressures exist for activating mutations within the WM clone. Lastly, multiple CXCR4^WHIM mutations are common in WM patients indicative of targeted genomic instability within the carboxyl-terminal domain of CXCR4.