Lost Cases Saved: Digital PCR Identifies Common Hodgkin Lymphoma Mutations from Ultra Low-Yield CT-DNA Samples Unsuitable for the Next-Generation Sequencing

Background: Cell-free DNA (cfDNA) has been established as a source of DNA for genotyping and minimal residual disease (MRD) monitoring in classic Hodgkin lymphoma (cHL) patients. However, the concentration of tumor-specific circulating tumor DNA (ctDNA) varies greatly among specimens and individual patients. Next-generation sequencing (NGS), the gold standard for cfDNA analysis, requires approximately 50-100 ng of DNA to ensure adequate sensitivity and coverage for detecting mutations and other genomic alterations. Many samples do not meet the minimum ctDNA concentration required for proper NGS analysis (library preparation). We have shown that digital PCR (dPCR) is a superior alternative for MRD target mutations detection in cHL (Prochazka et al., Blood 2022; 140 (Suppl 1): 6417-6418), which is faster, less expensive, and spares ctDNA quantity. The validity of the dPCR at very low concentration ctDNA samples has not been well described.

Aim: To perform a multiplex digital PCR MRD target mutation screening in low-concentration ctDNA samples not suitable for NGS analysis.

Methods: CfDNA was extracted from peripheral blood plasma using the QiaAmp Circulating Nucleic Acid Kit (Qiagen) at the time of diagnosis or relapse. For library preparation, we used SureSelect XT HS2 technology (Agilent Technologies) based on “target enrichment” with molecular barcodes. Sequencing was performed on a NovaSeq6000 (Illumina). Data were analyzed with SureCall software (Agilent Technologies) with a sensitivity of 1.0% variant allele frequency (VAF). The detected variants were annotated using COSMIC, dbSNP, Ensembl, and ClinVar. Samples with cfDNA below 40 ng were deemed unsuitable for analysis using our NGS panel.

For dPCR experiments, we used the Naica® system 6-color digital PCR (Stilla Technologies Inc.) with Sapphire chips. In total, we used three screening assays: one for a hotspot mutation in the XPO1 gene p.(E571K) and two multiplex dPCR assays for mutations in the STAT6 gene p.(N417Y) and p.(D419A/G/H/N/Y). Overall, our approach allows us to analyze seven missense mutations by dPCR.

To ensure the validity of our assays, we used genomic DNA (gDNA) from a blood donor as a negative control and gDNA from the lymphoma cell line L-1236 for XPO1 p. (E571K), and synthetic DNA fragments (gBlock™; Integrated DNA Technologies, Inc.) for STAT6 mutations as positive controls. Finally, to verify the functionality of our multiplex STAT6dPCR assays, we simultaneously analyzed ctDNA from two cases with confirmed missense mutations in the STAT6 gene by NGS.

Results: We analyzed cfDNA in 105 patients with HL enrolled in the prospective Czech Hodgkin Study Group database (NCT06263530), of whom 46 (44%) had low-yield cfDNA quantity below 40 ng. Two nodular lymphocyte-predominant HLs and five PCR reaction failures were excluded and 39 cases (40 at diagnosis, four at relapse) were analyzed by dPCR. The median age of the pts was 43 yrs, with clinical stages as follows: I in four (10%), II in twenty-two (56%), III in nine (23%) and IV in four (10%) of the pts, respectively; German Hodgkin Study Group (GHSG) stages were: early in twelve (31%), intermediate in eleven (28%), and advanced in sixteen (41%) of cases. Histology subtypes were: lymphocyte-rich in nine (23%), mixed-cellularity in fourteen (36%), and nodular sclerosis in sixteen (41%) of the pts. The median concentration of cfDNA was 0.65 ng/µL (range: 0.29-1.23 ng/µL), and the median amount of cfDNA used for each dPCR run was 7.8 ng (range: 3.5-14.8 ng). Mutations were detected in 10/39 (26%) of the patients: five in the XPO1gene p.(E571K) with a median VAF of 2.93% (range: 0.44-4.46%) and five in the STAT6 gene (one p.(N417Y), two p.(D419H), and two p.(D419G)) with a median VAF of 1.82% (range: 0.59-3.50%). The samples with/wo detected mutations showed similar concentrations of cfDNA (median 0.66 and 0.70 ng/µL in Mut+ vs Mut- cases, respectively, p=0.62).

Conclusion: We have developed a rapid and sensitive multiplex dPCR assay capable of identifying target mutations in a substantial number of samples with very low ctDNA content. Additionally, a more complex STAT6 multiplex assay is being developed, which includes hotspot mutations (N421K) and new probe designs to overcome the issue of detecting compound mutations in the STAT6 gene. This new assay will be presented at the conference.

Acknowledgement: Supported by AZV NU22-03-00182, MH CZ - DRO (FNOl, 00098892), and IGA_LF_2024_001.

Papajik:AstraZeneca: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: support for attending meetings; Janssen-Cilag: Honoraria, Membership on an entity's Board of Directors or advisory committees; AbbVie: Honoraria, Membership on an entity's Board of Directors or advisory committees. Bachanova:Citius: Research Funding; Incyte: Research Funding. Prochazka:Swixx: Consultancy, Speakers Bureau; Novartis: Consultancy; Eli Lilly: Consultancy; Abbvie: Consultancy.