Introduction:Somatic copy number alterations (SCNAs) constitute a major class of genomic aberrations known to have functional significance across cancers, including prognostic value in lymphoma. For example, diffuse large B-cell lymphoma (DLBCL) patients with tumors harboring SCNAs in MYC, BCL2, or TP53 tend to have inferior outcomes (Quesada, ASH 2016). However, detection of copy number alterations without invasive tissue biopsies remains difficult. We applied Cancer Personalized Profiling by Deep Sequencing (CAPP-Seq), a targeted high-throughput sequencing platform, to noninvasively detect and monitor genome-wide SCNAs in patients with DLBCL, primary mediastinal B-cell lymphoma (PMBCL), and classical Hodgkin lymphoma (cHL).

Methods: We profiled pre-treatment blood samples from 209 patients with diverse lymphomas including 168 DLBCL, 20 cHL, and 21 PMBCL, using a custom 314kb gene panel targeting recurrently altered regions in these cancers. Additionally, we sequenced 139 tumor biopsies to use as matched tissue gold standards for SCNA confirmation. In each sample, we interrogated the sequencing read distribution to summarize focal and broad SCNAs directly from plasma. All thresholds were tuned for 95% specificity, which were empirically verified in independent withheld healthy controls. Results were orthogonally validated through tissue karyotyping, fluorescence in situ hybridization (FISH), and whole genome sequencing (WGS) (Fig 1A and Fig 1B).

Results: We determined our limit of detection (LOD) for detecting SCNAs in plasma at ~1% circulating tumor fraction with 77% sensitivity; we observed higher sensitivity above 5% circulating tumor fraction (sensitivity = 95%). In our cohort, 70% of patients had over 1% circulating tumor burden with representation from all clinical stages. In DLBCL patients, we detected gains in MYC (12%), BCL2 (24%), BCL6 (19%), and PD-L1 (23%), as well as losses in TP53 (27%) and CDKN2A (19%). SCNAs detected in plasma were 85% concordant with those detected in matched tumor biopsies. Additionally, PD-L1 copy number gains were more frequently detected in both PMBCL (63%, p= 0.026) and cHL (71%, p= 0.012) than in DLBCL (23%), consistent with prior reports (Chapuy B, Blood 2016; Roemer MG, J. Clin. Oncol. 2016).

We also demonstrate that noninvasive detection of copy number gains in MYC and BCL2, as well as loss of TP53, are negatively prognostic for survival (p < 0.0001, p= 0.0013, p < 0.0001, respectively) (Fig 1C). Furthermore, we identified associations between DLBCL cell-of-origin classification and specific SCNAs, with non-GCB DLBCL enriched for BCL2, BCL6, and PD-L1 copy number gains (39% vs 17%; 31% vs 15%, and 39% vs 25%, respectively). Circulating tumor burden was not statistically different between GCB and non-GCB patients. TP53 losses were more frequently observed in relapsed/refractory patients compared to treatment naïve patients (22% vs 45%, p= 0.023).

We performed a preliminary analysis monitoring disease over time in three patients with clear SCNAs detected in plasma at diagnosis. Across longitudinal blood samples (n= 42), SCNA-based disease detection demonstrated 95% concordance with mutation-based disease detection, with a LOD down to ~0.1% circulating tumor burden. Our results suggest noninvasive SCNA profiling could provide an independent method for detecting residual disease, particularly in tumors characterized by structural alterations. Further validation and expansion of this monitoring method will be explored and presented.

Conclusions: Noninvasive identification of genome-wide SCNAs is feasible via targeted next generation sequencing. Our approach allows for both genotyping and monitoring of copy number gains and losses from plasma, without requiring matched germline or tissue specimens. This strategy not only allows for the noninvasive genotyping of tumor SCNAs, but also provides an orthogonal method for monitoring residual disease from liquid biopsies using next generation sequencing.

Figure 1: A) Matched tissue FISH staining confirms arm-level PD-L1 amplification (red) detected in plasma. D9Z1 (green) was used as an assay control. B) SCNAs detected in plasma were concordant with karyotype analysis of a matched tumor biopsy. C) Copy number alterations in MYC, BCL2, and TP53 are adversely prognostic for survival in R-CHOP treated DLBCL patients.


Hüttmann: Bristol-Myers Squibb, Takeda, Celgene, Roche: Honoraria; Gilead, Amgen: Other: Travel cost. Gaidano: Roche: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Gilead: Consultancy, Honoraria; AbbVie: Consultancy, Honoraria. Westin: Kite Pharma: Membership on an entity's Board of Directors or advisory committees; Apotex: Membership on an entity's Board of Directors or advisory committees; Novartis Pharmaceuticals Corporation: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees. Advani: Bayer Healthcare Pharmaceuticals: Research Funding; Gilead: Consultancy; Bristol-Myers Squibb: Consultancy, Research Funding; Celgene: Research Funding; Janssen: Research Funding; Genentech: Research Funding; Seattle Genetics: Research Funding; Nanostring: Consultancy; Spectrum: Consultancy; Millennium: Research Funding; FortySeven: Research Funding; Pharmacyclics: Research Funding; Regeneron: Research Funding; Agensys: Research Funding; Infinity: Research Funding; Juno Therapeutics: Consultancy; Kura: Research Funding; Merck: Research Funding; Pharmacyclics: Consultancy; Cell Medica: Research Funding; Sutro: Consultancy. Diehn: Varian Medical Systems: Research Funding; Roche: Consultancy; Novartis: Consultancy; Quanticel Pharmaceuticals: Consultancy. Alizadeh: Genentech: Consultancy; Celgene: Consultancy; Gilead: Consultancy; Roche: Consultancy; CiberMed: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.