Over the last decade, circulating tumor DNA (ctDNA) has emerged as an important biomarker across a range of cancers, including lymphomas. The potential clinical applications of ctDNA technologies are broad, including tumor mutational genotyping, detection of molecular responses and minimal residual disease, as well as early detection of relapsing disease with specific attention to mechanisms of resistance. Numerous opportunities therefore exist to apply ctDNA methodologies to tailor therapy for individual patients. By applying targeted next-generation sequencing of ctDNA via Cancer Personalized Profiling by Deep Sequencing (CAPP-Seq) in patients with lymphoma, these opportunities for precision oncology largely fall into two groups—identification of specific targetable molecular features and quantitative assessment of disease burden and risk. Here we explore these approaches in patients with diffuse large B-cell lymphoma (DLBCL) undergoing first-line or salvage therapy. Utilizing CAPP-Seq from pretreatment plasma, mutational genotyping of diverse somatic alterations is possible, including single-nucleotide variants (SNVs), translocations, and copy number alterations. By combining these mutations through an integrated Bayesian framework, identification of specific disease subtypes is possible, including activated and germinal center B-cell molecular cell-of-origin subtypes. In addition, detection of emergent mutations during and after treatment can provide insights into mechanisms of therapeutic resistance, many of which are potentially targetable. In addition to these insights into disease biology, ctDNA can serve as a quantitative handle on disease burden. By measuring ctDNA disease burden in DLBCL patients from six centers throughout North America and Europe, we demonstrate the prognostic value of ctDNA levels prior to treatment. Additionally, we found that a molecular response to treatment, as detected by the change in ctDNA after one or two cycles of therapy, is highly prognostic for patient outcomes in both front-line and later lines of therapy. Moreover, these molecular response features can be combined with additional outcome predictors, such as the IPI and interim PET/CT scans, to build a dynamic risk model termed the Continuous Individualized Risk Index (CIRI). CIRI provides a personalized probability of likely patient outcomes that updates over time as more information becomes available. Outcome prediction with CIRI significantly improved on predictions using ctDNA molecular response, interim PET/CT scans, or the IPI alone. In summary, liquid biopsies through ctDNA afford numerous opportunities for personalized medicine, ranging from mutational profiling and molecular subtyping to disease quantification and response prediction. Further studies of novel clinical approaches are likely to change treatment paradigms in the near future.

Citation Format: David M. Kurtz. Approaches for personalized medicine in lymphoma through liquid biopsies [abstract]. In: Proceedings of the AACR Virtual Meeting: Advances in Malignant Lymphoma; 2020 Aug 17-19. Philadelphia (PA): AACR; Blood Cancer Discov 2020;1(3_Suppl):Abstract nr IA41.