The optimal work up in non-Hodgkin lymphomas (NHLs) include morphological and immuno-phenotypic investigations as well as chromosome and molecular analysis. Among them, ‘next-generation sequencing’ (NGS) techniques have provided relevant additional data in diagnosis, prognosis and therapeutic management. Together with gene expression profiling (GEP) studies, NGS has revealed distinct clinicobiological entities and refined the lymphoma classification, especially in peripheral T-cell lymphoma . Although GEP results have strongly contributed to a better categorization and prognostication in lymphomas, they do not allow for optimization of front-line or relapse treatments, and fail to clearly demonstrate which patients could benefit from novel targeted therapies [2▪▪]. However, the ever-growing knowledge on lymphoma mutational landscape should offer the chance of capturing new tailored treatment possibilities [2▪▪]. Although NGS data in lymphomas still need further validation before being implemented in daily practise, we may reasonably consider that their clinical application is around the corner, at least partially . Accordingly, routine molecular laboratories should soon be able to provide reliable genomic information in a short time frame in order to remain clinically pertinent. In that context, the use of affordable technical approaches such as NGS panels that only target relevant genes instead of genome-wide sequencing is a pragmatic option. An optimal panel must include genes holding current diagnostic, prognostic and theranostic values for the different lymphoma subtypes while remaining sufficiently simple to comply with clinical requirements (good sensitivity, short reporting timelines and so on) and uniform application. For that purpose, a group of French experts has recently proposed two consensus panels; the first one composed of 33 genes for B cell lymphomas and the second one limited to 11 genes for T-cell lymphomas [4▪▪]. The authors highlighted that their panels need to evolve with regular updates according to advances in specialized literature. Their inspiring proposal is an excellent starting point for routine molecular laboratories seeking to implement NGS analysis in a clinical setting. In this study, we will review the mutational landscapes characterizing the B-cell and T-cell malignant lymphoma. For the sake of brevity, the discussion will be limited to genetic aberrations encountered in some B-cell and T-cell lymphoma and with seemingly clinical relevance in terms of diagnosis, risk stratification and therapy. More lymphomatous entities are included in Tables 1 and 2. The contribution of NGS to diagnostic, prognostic and personalized therapy in lymphomas as well as its use in liquid biopsies will also be discussed.
BIOMARKERS IDENTIFIED BY NEXT-GENERATION SEQUENCING IN DIFFERENT NON-HODGKIN'S LYMPHOMA SUBTYPES
Diffuse large B-cell lymphoma
Diffuse large B-cell lymphoma (DLBCL) is known to be a biologically and clinically heterogeneous disease. The cell of origin (‘COO’) classification (based on GEP) along with the double hit status (based on the presence of CMYC, BCL2 and/or BCL6 oncogene rearrangements) have undoubtedly optimized classification and prognostication in DLBCL . However, they have only modestly impacted the current therapeutic management of DLBCL. Also, variations in the patient outcome as well as in response to treatment still persist, even within seemingly well defined lymphoma subsets. This calls for a deeper knowledge of the extreme molecular heterogeneity that features DLBCL. Among the recent and interesting publications addressing that matter (see  for review), two seminal NGS studies deserve to be mentioned. On the basis of mutational profiles and genomic rearrangements, Schmitz et al.[6▪▪] identified four genetic subtypes relying on the cooccurrence of genetic aberrations: Group 1 ‘MCD’ characterized by MYD88/CD79B genes comutations and group 3 ‘N1’ (NOTCH1 mutations) most frequent among ABC lymphomas, group 2 ‘BN2’ (NOTCH2 mutations and BCL6 fusions) observed in similar proportions among ABC and GCB, and group 4 ‘EZB’ (EZH2 mutations and BCL2 translocations) present most often in GCB. BN2 and EZB subtypes were associated with better response to R-CHOP and favourable outcome, while MCD and N1 subgroups were not. Of note, MCD and BN2 subtypes depend on a chronic active BCR signalling pathway, opening the option for targeted treatments. By adding somatic copy number alterations among preexisting biomarkers, Chapuy et al.[7▪▪] defined five DLBCL subsets including cluster 1 (BCL6 and NOTCH2 comutations) and cluster 5 (chromosome 18q gain with BCL2 and MALT1 gene overexpression as well as CD79B and MYD88 mutations) that are both most often observed among ABC-DLBCL; cluster 2 (with biallelic TP53 inactivation, CDKN2A loss and genomic instability) as a COO-independent group; cluster 3 (BCL2 translocation, alterations of PTEN as well as of epigenetic mediators such as KMT2D, CREBBP and EZH2) and cluster 4 (alterations in BCR/PI3K, JAK/STAT and BRAF pathways) both subsets of GCB-DLBCL. Clusters 2, 3 and 5 represented poor risk groups, while clusters 1 and 4 were associated with favourable outcome. Of note, C1, C3 and C5, respectively, overlapped with the BN2, EZB and MCD groups of Schmitz et al.[6▪▪]. These two genomic studies provide more molecular homogeneity than the original COO classification and improve our understanding of the biology of DLBCL, with better prognostic and predictive information. Moreover, they pave the way for better usage of personalized therapy, as targeted treatments can only be given in the correct genomic context. Indeed, the incorporation of BTK inhibitors such as Ibrutinib to chemotherapy in DLBCL could be more beneficial when restricted to patients belonging to the MCD and cluster 5 genomic groups (activating MYD88 and CD79B genes comutations) [2▪▪,6▪▪,7▪▪]. This strategy would lead to more efficient use of drugs directed against MYD88 mutations, a strong independent parameter of adverse prognosis in DLBCL . Moreover, DLBCL displaying clusters 3 and 5 genomic profiles (with BCL2 overexpression/amplification in common) seem to be more sensitive to BCL2 inhibition, especially in combination with PI3K α/δ blockade .
A prognostic model called ‘m7-FLIPI’ that integrates mutation status of seven genes (EZH2, ARID1A, MEF2B, EP300, FOXO1, CREBBP and CARD11) with clinical risk parameters (such as the FLIPI) has been recently proposed . This promising approach might help to identify patients with an increased risk of failure to first-line immunochemotherapy . Interestingly, several genes composing this model are involved in epigenetic regulation and could have a potential theranostic impact. Also, a panel of four mutated genes (NOTCH2, DTX1, UBE2A and HIST1H1E) strongly related to a higher risk of DLBC transformation has recently been described, offering a useful molecular tool for the identification of patients with a shorter time to transformation . Despite their apparent clinical interest, these genomic models are still not used in daily routine.
Mantle cell lymphoma
A reliable diagnosis of mantle cell lymphoma (MCL) can be done by immunohistochemistry (CCND1 and/or SOX11 overexpression) or FISH (CCND1 rearrangement). The usefulness of NGS in MCL will be to identify biomarkers of poor prognosis (CDNK2A deletion, ATM, NOTCH1/2, NSD2 mutations among others) or with theranostic/predictive interest (BTK, PLCG2, BIRC3, SMARCA4 and MAP3K14 mutations) or both (TP53) [4▪▪,12,13]. Notably, TP53 aberrations have a strong predictive value in MCL, being frequently associated with resistance to current front-line therapies and demonstrate inferior response rates to ibrutinib . Moreover, stem cell transplant does not seem to abolish TP53 dismal prognostic value . These genes mutations are common in both nodal and nonnodal (leukemic) forms of MCL but are rather subclonal in the indolent leukemic variant, partially explaining their lack of prognostic significance in this subgroup of MCL [12,15]. As a watch-and-wait strategy is adopted at diagnosis in the majority of nnMCL, monitoring by NGS for expansion of TP53 mutated clones could serve as a surrogate of disease progression and indicate the need for therapeutic intervention. TP53 sequencing by NGS should currently be part of the diagnostic work up in MCL (particularly in blastoid cases) or when a treatment is planned, especially if an upfront transplant is anticipated. If positive, clinical trials should be an option . Other gene mutations are not currently used but could potentially be included in clinical practice.
Peripheral T-cell lymphoma
NGS and GEP studies have both brought robust diagnostic molecular signatures in peripheral T-cell lymphoma [20,21,22]. These signatures may serve as a diagnostic tool and/or be considered as potential targets for novel and more efficient therapies [23▪,24▪]. For the sake of space, NGS indications in T-cell lymphoma will be limited to the subtypes wherein this technology has brought a significant contribution in terms of diagnosis, prognosis and/or theranostics. More details are given in Table 2.
Angioimmunoblastic T-cell lymphoma and nodal peripheral T-cell lymphoma with follicular helper T-cell phenotype (TFH-PTCL)
A TET2/ DNMT3A/ RHOAG17V mutational profile tends to be confined to TFH-PTCL and angioimmunoblastic T-cell lymphoma (AITL), making the distinction between these two entities and the remaining PTCL-NOS subgroup possible [23▪,24▪]. Although the TET2 mutations are shared among PTCL entities, they are particularly frequent in AITL (up to 85%). The IDH2R172 mutation is highly specific of AITL and can serve as a diagnostic biomarker. It often cooccurs with TET2 mutations. The RHOAG17V is also frequently observed in AITL (as high as 70%) and exclusively when TET2 mutations are present [24▪]. Novel targeted treatments such as demethylating agents could be justified in T-cell lymphoma that exhibit epigenetic deregulation as major oncogenic alteration . Activating mutations affecting the TCR signalling pathway (CD28, PLCϒ1, CARD11, FYN and so on) are also described in TFH lymphomas but are currently not in use [24▪,26].
Other peripheral T-cell lymphomas with gene mutations of current or potential clinical interest
Due to their rarity, the genetic landscape in other PTCL is less known than the aforementioned T lymphoma subtypes molecular profiles. The mutational profiles deciphered in some of these T lymphoma subgroups could be of diagnostic or therapeutic use (see Table 2). Suffice to say that activating mutations in the JAK/STAT pathway are commonly observed among these T-cell lymphoma subtypes and could represent sensitive therapeutic targets to JAK/STAT inhibitors [24▪]. Notably, activation of the JAK/STAT pathway by a STAT3 oncogenic mutation leads to PD-L1 overexpression in NK/T-cell lymphoma . The combination of PD-1/PD-L1 blockade and STAT3 inhibitors might be a promising therapeutic approach in this PTCL subtypes . Gene mutations affecting the TCR pathway and, to a lesser extent, the NF-κB pathway have been identified in adult T-cell leukaemia/lymphoma (ATLL) [24▪]. Aside from the JAK/STAT pathway, the genetic profile in hepatosplenic T-cell lymphoma (HSTCL) is frequently characterized by mutations involving chromatin-modifying genes (SETD2, INO80 and ARID1B) . SETD2 gene mutation can also be observed in enteropathy associated T-lymphoma (EATL) [24▪]. It can thus allow to make the distinction between these last two entities and the remaining PTCL.
NEXT-GENERATION SEQUENCING AND PERSONALIZED TREATMENT APPROACH
As mentioned before, molecular information brought by GEP analyses  or double/triple hit (DHL/THL) status in DLBCL has not deeply impacted therapy decision-making [2▪▪]. Indeed, standard treatments such as R-CHOP or more intensified chemotherapeutic approaches remain the current choices in aggressive DLBCL or advanced stage disease [2▪▪] and use of targeted therapy guided by transcriptomic signatures along with standard chemotherapy did not always lead to the expected results [2▪▪,5▪▪]. In contrast, a more in-depth knowledge of the complex genomic background in lymphoma, as offered by NGS-based technologies, will help target specific vulnerabilities and promise a better use of tailor-made treatments. The success of personalized medicine will not solely rely on the identification of single druggable mutations but also on the precise identification of synergistic mutational cooccurrences and genomic aberrations inducing resistance to targeted therapy [2▪▪,5▪▪–7▪▪,9]. A recent NGS-based study performed on MCL patients has identified a distinct genome profile that separates responders to ibrutinib and venetoclax combinatory therapy from nonresponders [28▪]. The molecular distinction was based on inactivating mutations in the SWI-SNF chromatin-remodelling complex that lead to BCL-XL upregulation and subsequent resistance to the therapeutic combination [28▪]. Novel targeted treatments such as demethylating agents could be justified in T cell lymphomas harbouring TET2, DNMT3A and/or IDH2 mutations . The IDH2 mutational status may also serve as a guide for therapeutic strategies in AITL with the use of PI3K-mTOR inhibitors in IDH2 wild type cases, while alkylating agents and PARP inhibitors could have synergistic and efficient effects in mutated cases . However, this requires further validation before being implemented in daily practice. Simultaneous activations of multiple signalling pathways, as observed in cancer, might circumvent the inhibitory effects of a targeted monotherapy. Therefore, comprehensive knowledge of these pathways redundancies, along with the expansion of available therapeutic options, will probably raise the chance of treatment success through combination strategies that concurrently trigger several activated cancer-signalling mechanisms [2▪▪,5▪▪]. For that purpose, a computational system biology tool (called ‘DrugComboExplorer’) has been recently proposed to identify personalized drug combinations and more efficient treatment arms [29▪▪]. This tool combines multiomics data (i.e. DNA seq, gene copy data, RNA-seq data and so on) of an individual cancer patient with pharmacogenomics profiles of thousands of drugs and bioactive compounds from the NIH LINCS program (Library of Integrated Network-based cellular Signatures). Although this strategy undoubtedly needs further validation before being implemented in a clinical setting, it represents the ultimate goal of precision medicine.
NEXT-GENERATION SEQUENCING AND LIQUID BIOPSY: MOLECULAR DIAGNOSIS AND MINIMAL RESIDUAL DISEASE ASSESSMENT
Biopsies remain the best procedure for a reliable morphological diagnosis in lymphoma. Nevertheless, there are not always suitable to obtain a comprehensive tumour genomic assessment due to the spatial intratumour genetic heterogeneity observed in most cancers, including lymphomas [30,31]. In other words, the tissue sampling available for genetic studies will not necessarily reflect the genomic landscape composing the entire tumour tissue present in the patient. Moreover, the temporal tumour genetic heterogeneity – that is the emergence of genetic diversifications between the primary tumour and metastases or even between metastases themselves through the course of the disease – can substantially hamper the predictive power of biomarkers and affect the effectiveness of targeted therapies. The same is true for the identification of B or T-cell monoclonality in lymphoma at diagnosis and follow up. The association of NGS-based technologies and liquid biopsy (circulating tumour DNA, ctDNA) shows great promise to these issues. Indeed, ctDNA reflects the entire tumour genetics, and NGS simultaneously detects many tumour mutations or DNA sequences (IGH or TCR loci), overcoming the limitations of tracking single mutation or clone encountered with the used of conventional tools. The feasibility and reliability of NGS-based ctDNA genotyping for diagnosis and prognosis have been well demonstrated across several lymphoma subtypes [32–35]. The same is true for real-time monitoring of the disease evolution and therapy response, regardless of the nature of the target (gene mutation or B/T-cell monoclonality) [32–39]. NGS-based ctDNA analysis may distinguish genomic profiles among different lymphoma subsets indicating that ctDNA could be a noninvasive and feasible biomarker for diagnosis . Moreover, the ctDNA load strongly reflects tumour burden, as it appears to correlate significantly with LDH and IPI, as observed in DLBCL and NKTCL . NGS analyses on liquid biopsy represent an efficient complementary tool to PET scan imaging for the management of DLBCL, at diagnosis and during follow-up [33,40]. A recent MRD study on DLBCL patients treated with CAR-T cell therapy has even demonstrated better sensitivity and predictive value for progression to treatment than the PET scan . This study, among others, indicates that NGS coupled with liquid biopsy would represent an efficient platform for therapeutic efficacy measurement. Recently, a method called CIRI (Continuous Individualized Risk Index) has been proposed to dynamically evaluate individual outcome probabilities using risk predictors acquired over time . CIRI has been applied to monitoring various lymphoma cases while taking into account risk factors such as the IPI, COO classification, interim imaging (iPET), along with serial ctDNA measurements. This tool can provide risk assessments in real-time and throughout the disease course, enabling the development of adapted treatment and instantaneous therapy selection . Although further validations are needed, this new method would lead us closer to precision medicine, if suitable for routine use.
NGS analyses in lymphoma allow simultaneous detection of multiple genomic biomarkers that may be complementary to the morphological work up and represent valuable diagnostic and/or prognostic tools. With the expansion of targeted drugs, the use of NGS on liquid biopsies will be a major breakthrough not only towards tailor-made therapies at diagnosis but also towards a real-time and dynamic monitoring of tumour responses to treatment, with early detection of therapy-resistant clones potentially leading to protocole adjustments. Despite its apparent significant contribution to lymphoma management, NGS-based ctDNA analyses need further standardization and harmonization projects, especially at the technical level , before their implementation in routine clinical practice can become a reality.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING
Papers of particular interest, published within the annual period of review, have been highlighted as:
1. Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood 2016; 127:2375–2390.
2▪▪. Chiappella A, Crombie J, Guidetti A, et al. Are we ready to treat diffuse large b-cell and high-grade lymphoma according to major genetic subtypes? Hemasphere 2019; 3:e284.
3. Di Paolo A, Arrigoni E, Luci G, et al. Precision medicine
in lymphoma by innovative instrumental platforms. Front Oncol 2019; 9: article 1417.
4▪▪. Sujobert P, Le Bris Y, de Leval L, et al. The need for a consensus next-generation sequencing panel for mature lymphoid malignancies. HemaSphere 2019; 3:e169.
5▪▪. Coccaro N, Anelli L, Zagaria A, et al. Molecular complexity of diffuse large B-cell lymphoma: can it be a roadmap for precision medicine
? Cancers 2020; 185:1–20.
6▪▪. Schmitz R, Wright GW, Huang DW, et al. Genetics and pathogenesis of diffuse large B-cell lymphoma. New Engl J Med 2018; 378:1396–1407.
7▪▪. Chapuy B, Stewart C, Dunford AJ, et al. Molecular subtypes of diffuse large B cell lymphoma are associated with distinct pathogenic mechanisms and outcomes. Nat Med 2018; 24:679–690.
8. Vermaat JS, Somers SF, de Wreede LC, et al. MYD88 mutations identify a molecular subgroup of diffuse large B-cell lymphoma with an unfavourable prognosis. Haematologica 2020; 105:424–434.
9. Bojarczuk K, Wienand K, Ryan JA, et al. Targeted inhibition of PI3Kα/δ is synergistic with BCL-2 blockade in genetically defined subtypes of DLBCL. Blood 2019; 133:70–80.
10. Jurinovic V, Kridel R, Staiger AM, et al. Clinicogenetic risk models predict early progression of follicular lymphoma after first-line immunochemotherapy. Blood 2016; 128:1112–1120.
11. González-Rincón J, Méndez M, Gómez S, et al. Unraveling transformation of follicular lymphoma to diffuse large B-cell lymphoma. PLoS One 2019; 14:e0212813.
12. Clot G, Jares P, Giné E, et al. A gene signature that distinguishes conventional and leukemic nonnodal mantle cell lymphoma helps predict outcome. Blood 2018; 132:413–422.
13. Jain P, Wang M. Mantle cell lymphoma: 2019 update on the diagnosis, pathogenesis, prognostication, and management. Am J Hematol 2019; 94:710–725.
14. Condoluci A, Rossi D, Zucca E, Cavalli F. Toward a risk-tailored therapeutic policy in mantle cell lymphoma. Curr Oncol Rep 2018; 20:79.
15. Sakhadari A, Ok CY, Patel KP, et al. TP53 mutations are common in mantle cell lymphoma, including the indolent leukemic nonnodal variant. Ann Diagn Pathol 2019; 41:38–42.
16. Nakamura S, Ponzoni M. Marginal zone B-cell lymphoma: lessons from Western and Eastern diagnostic approaches. Pathology 2020; 52:15–29.
17. Tausch E, Beck P, Schlenk R. Prognostic and predictive role of gene mutations in chronic lymphocytic leukemia: results from the pivotal phase III study COMPLEMENT1. Haematologica 2020; [Epub ahead of print].
18. Cohen JA, Bomben R, Pozzo F, et al. An updated perspective on current prognostic and predictive biomarkers
in chronic lymphocytic leukemia in the context of chemoimmunotherapy and novel targeted therapy. Cancers (Basel) 2020; 12:pii: E894.
19. Treon SP, Xu L, Guerrera ML, et al. Genomic landscape of Waldenström macroglobulinemia and its impact on treatment strategies. J Clin Oncol 2020; 38:1198–1208.
20. Amador A, Greiner TC, Heavican TB, et al. Reproducing the molecular subclassification of peripheral T-cell lymphoma-NOS by immunohistochemistry. Blood 2019; 134:2159–2170.
21. Maura F, Agnelli L, Leongamornlert D, et al. Integration of transcriptional and mutational data simplifies the stratification of peripheral T-cell lymphoma. Am J Hematol 2019; 94:628–634.
22. Heavican TB, Bouska A, Yu J, et al. Genetic drivers of oncogenic pathways in molecular subgroups of peripheral T-cell lymphoma. Blood 2019; 133:1664–1676.
23▪. De Leval L. Approach to nodal-based T-cell lymphomas
. Pathology 2020; 52:78–99.
24▪. Iqbal J, Amador C, McKeithan TW, Chan WC. Querfeld C, et al. Molecular and genomic landscape of peripheral T-cell lymphoma. T-cell and NK-cell lymphomas
, cancer treatment and research. Basel: Springer Nature Switzerland AG; 2019. 31–59.
25. Lemonnier F, Dupuis J, Sujobert P, et al. Treatment with 5-azacytidine induces a sustained response in patients with angioimmunoblastic T-cell lymphoma. Blood 2018; 132:2305–2309.
26. Lone W, Alkhiniji A, Umakanthan JM, Iqbal J. Molecular insights into pathogenesis of Peripheral T cell lymphoma: a review. Curr Hematol Malig Rep 2018; 13:318–328.
27. Song TL, Nairismägi ML, Laurensia Y, et al. Oncogenic activation of the STAT3 pathway drives PD-L1 expression in natural killer/T-cell lymphoma. Blood 2018; 132:1146–1158.
28▪. Agarwal R, Chan YC, Tam CS, et al. Dynamic molecular monitoring reveals that SWI-SNF mutations mediate resistance to ibrutinib plus venetoclax in mantle cell lymphoma. Nat Med 2019; 25:119–129.
29▪▪. Huang L, Brunell D, Stephan C, et al. Driver network as a biomarker: systematic integration and network modeling of multiomics data to derive driver signaling pathways for drug combination prediction. Bioinformatics 2019; 35:3709–3717.
30. Araf S, Wang J, Korfi K, et al. Genomic profiling reveals spatial intra-tumor heterogeneity in follicular lymphoma. Leukemia 2018; 32:1261–1265.
31. Dubois S, Tesson B, Mareschal S, et al. Refining diffuse large B-cell lymphoma subgroups using integrated analysis of molecular profiles. EBioMedicine 2019; 48:58–69.
32. Scherer F, Kurtz DM, Newman AM, et al. Distinct biological subtypes and patterns of genome evolution in lymphoma revealed by circulating tumor DNA. Sci Transl Med 2016; 8:364ra155.
33. Condoluci A, Rossi D. The future of cell-free DNA testing to guide therapeutic decisions in B- cell lymphomas
. Curr Opin Hematol 2019; 26:281–287.
34. Sun P, Chen C, Xia Y, et al. Mutation profiling of malignant lymphoma by next-generation sequencing of circulating cell-free DNA. J Cancer 2019; 10:323–331.
35. Darrah JM, Herrera AF. Updates on circulating tumor DNA assessment in lymphoma. Curr Hematol Malig Rep 2018; 13:348–355.
36. Kurtz DM, Scherer F, Jin MC, et al. Circulating tumor DNA measurements as early outcome predictors in diffuse large B-cell lymphoma. J Clin Oncol 2018; 36:2845–2853.
37. Galimberti S, Genuardi E, Mazziotta F, et al. The minimal residual disease in non-Hodgkin's lymphomas
: from the laboratory to the clinical practice. Front Oncol 2019; 9:528.
38. Arzuaga-Mendez J, Prieto-Fernández E, Lopez-Lopez E, et al. Cell-free DNA as a biomarker in diffuse large B-cell lymphoma: a systematic review. Crit Rev Oncol Hematol 2019; 139:7–15.
39. Spina V, Bruscaggin A, Cuccaro A, et al. Circulating tumor DNA reveals genetics, clonal evolution, and residual disease in classical Hodgkin lymphoma. Blood 2018; 131:2413–2425.
40. Bohers E, Viailly PJ, Becker S, et al. Noninvasive monitoring of diffuse large B-cell lymphoma by cell-free DNA high-throughput targeted sequencing: analysis of a prospective cohort. Blood Cancer J 2018; 8:74.
41. Hossain NM, Dahiya S, Le R, et al. Circulating tumor DNA assessment in patient with diffuse large B-cell lymphoma following CAR T-cell therapy. Leuk Lymphoma 2019; 60:503–506.
42. Kurtz DM, Scherer F, Jin MC, et al. Dynamic risk profiling using serial tumor biomarkers for personalized outcome prediction. Cell 2019; 178:699–713.
43. Rossi D, Spina V, Bruscaggin A, Gaidano G. Liquid biopsy in lymphoma. Haematologica 2019; 104:648–652.