Jiménez et al. Biomarker Research (2021) 9:37 https://doi.org/10.1186/s40364-021-00290-z RESEARCH Open Access Immunological and genetic kinetics from diagnosis to clinical progression in chronic lymphocytic leukemia Isabel Jiménez1,2, Bárbara Tazón-Vega2,3, Pau Abrisqueta2,3, Juan C. Nieto1,2, Sabela Bobillo2,3, Carles Palacio-García2,3, Júlia Carabia1,2, Rafael Valdés-Mas4, Magdalena Munuera1,2, Lluís Puigdefàbregas1,2, Genís Parra5,6, Anna Esteve-Codina5,6, Clara Franco-Jarava7, Gloria Iacoboni2,3, María José Terol8, José Antonio García-Marco9, Marta Crespo1,2† and Francesc Bosch2,3*† Abstract Background: Mechanisms driving the progression of chronic lymphocytic leukemia (CLL) from its early stages are not fully understood. The acquisition of molecular changes at the time of progression has been observed in a small fraction of patients, suggesting that CLL progression is not mainly driven by dynamic clonal evolution. In order to shed light on mechanisms that lead to CLL progression, we investigated longitudinal changes in both the genetic and immunological scenarios. Methods: We performed genetic and immunological longitudinal analysis using paired primary samples from untreated CLL patients that underwent clinical progression (sampling at diagnosis and progression) and from patients with stable disease (sampling at diagnosis and at long-term asymptomatic follow-up). Results: Molecular analysis showed limited and non-recurrent molecular changes at progression, indicating that clonal evolution is not the main driver of clinical progression. Our analysis of the immune kinetics found an increasingly dysfunctional CD8+ T cell compartment in progressing patients that was not observed in those patients that remained asymptomatic. Specifically, terminally exhausted effector CD8+ T cells (T-betdim/−EomeshiPD1hi) accumulated, while the the co-expression of inhibitory receptors (PD1, CD244 and CD160) increased, along with an altered gene expression profile in T cells only in those patients that progressed. In addition, malignant cells from patients at clinical progression showed enhanced capacity to induce exhaustion-related markers in CD8+ T cells ex vivo mainly through a mechanism dependent on soluble factors including IL-10. Conclusions: Altogether, we demonstrate that the interaction with the immune microenvironment plays a key role in clinical progression in CLL, thereby providing a rationale for the use of early immunotherapeutic intervention. Keywords: CLL, Immune evasion, Clinical progression, T cell exhaustion * Correspondence: fbosch@vhio.net † Marta Crespo and Francesc Bosch contributed equally to this work. 2 Department de Medicina, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain 3 Servei d’Hematologia, Vall d’Hebron Hospital Universitari, Experimental Hematology, Vall d’Hebron Institute of Oncology (VHIO), Vall d’Hebron Barcelona Hospital Campus, Passeig Vall d’Hebron 119-129, 08035 Barcelona, Spain Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Jiménez et al. Biomarker Research (2021) 9:37 Page 2 of 14 Background progression did not experience significant changes over Identification of patients that are at high risk of progres- time in their CD8+ T cell exhaustion status and the cap- sion to advanced clinical stages needing treatment in acity of malignant cells to induce T cell exhaustion and chronic lymphocytic leukemia (CLL) is expedited by the to secrete IL-10 was lower compared to malignant cells use of different prognostic parameters including the from progressing patients and maintained over time.. mutational status of the IGHV genes or its surrogates These results could contribute to lay the foundations for like ZAP-70 expression [1, 2], chromosomal aberrations the clinical testing of early therapeutic interventions on and gene mutations (i.e., TP53 or NOTCH1), [3, 4] or the immune system to prevent or delay progression of prognostic scores like the CLL-IPI [5]. Despite the great the disease. advances in the understanding of the biology of this dis- ease over the last decade, the underlying mechanisms Methods that drive clinical progression from early stages have not Primary samples been fully deciphered yet.. Longitudinal genetic studies Thirty eight patients diagnosed with CLL were enrolled performed at the time of diagnosis and at progression in the study. Peripheral blood mononuclear cells before treatment show non-recurrent and limited (PBMCs) were isolated by Ficoll density gradient and changes in around half of the patients [6–15]. In agree- cryopreserved. Plasma was obtained from EDTA blood ment, changes in the gene expression profile of leukemic and stored at − 80 °C. Samples were collected at two cells are also infrequent at progression [16]. These find- time points: diagnosis and progression before treatment ings indicate that CLL progression from early stages is or long-term asymptomatic follow-up. For most of the not mainly driven by genetic evolution, which highlights experiments, only a subgroup of the patients is repre- the potential role of the leukemic microenvironment in sented due to availability of samples. For co-culture as- the evolution of the disease. says, PBMCs from age-matched healthy donors were Besides the generalized immune dysfunction observed used (HD; n = 17; 64 years old). A written informed con- in patients diagnosed with CLL [17], circulating T cells sent was obtained from all individuals in accordance from these patients exhibit an increased proportion of with the declaration of Helsinki and the study was ap- dysfunctional and exhausted antigen-experienced cells proved by the local clinical research ethics committee. characterized by high expression of inhibitory receptors, impaired proliferation and cytotoxic activity as well as a Whole-exome sequencing (WES) and data processing defective immune synapse with leukemic cells, [18–22] Libraries for WES were sequenced on HiSeq2500 all indicating defective anti-tumor T cell responses. (Illumina, San Diego, CA, USA) with a read length of Moreover, the immunosuppressing microenvironment 100 bp paired-end. Raw data was processed using the prevalent in CLL is also boosted by the expansion of Real Time Analysis software (RTA 1.18.66.3, Illumina) myeloid-derived suppressor cells (MDSCs) [23] and the to generate FASTQ sequence files, which were processed production of IL-10 by CLL cells [24], among other using the bioinformatics software HD Genome One factors. (DREAMgenics, Oviedo, Spain). These observations led us to hypothesize that a The detection of copy number variations (CNVs) was disease-induced T cell exhaustion accumulating from performed by a modified version of the exome2cnv algo- diagnosis could translate into an escape from immune rithm [25]. For tumor samples, the algorithm employed surveillance driving CLL from early asymptomatic stages a pool of all control samples as background. The cancer to clinical progression. To identify mechanisms of cell fraction (CCF) and the 95% confidence interval (CI) clinical progression we studied the specific changes that were computed using the R package Palimpsest [26]. See occur in patients that progress compared to those that Supplementary Methods for detailed information. remain asymptomatic along time. In detail, we per- formed genetic and immunological longitudinal analysis RNA sequencing (RNA-Seq) and data processing using paired primary samples from untreated CLL Libraries for RNA-Seq were sequenced on HiSeq2500 patients that experienced clinical progression (sampling (Illumina) with a read length of 100 bp paired-end. at diagnosis and at progression) and from patients with Reads were mapped against the human reference gen- stable disease (sampling at diagnosis and at long-term ome (GENCODE release 28) and genes were quantified asymptomatic follow-up). Our results show that CLL as detailed in Supplementary Methods. cells exhibit limited and non-recurrent genetic changes at progression while CD8+ T cells display increased ex- Flow cytometry analysis haustion features, potentially induced by IL-10 secreted Cryopreserved PBMCs were thawed and stained with by malignant B cells, and a differential transcriptome at monoclonal antibodies (mAbs, Supplementary Table S1) progression. In contrast, patients without evidence of as detailed in Supplementary Methods. Jiménez et al. Biomarker Research (2021) 9:37 Page 3 of 14 For the assessment of intracellular IL-10 produced by (median time to progression (TTP): 29 months, range 5– CLL cells, PBMCs were co-cultured for 48 h with 96 months), we collected serial paired samples at diagno- UE6E7T-2 cells, CD40L (Peprotech, Rocky Hill, NJ, sis and at the time of progression before treatment. USA) and TLR9L (CpG ODN2006, Invivogen, San Definition of progression and requirement for treatment Diego, CA, USA) as we previously described [27] and were established following the international workshop then cells were stimulated with Leukocyte Activation on CLL (IWCLL) criteria [28]. As a control group, we Cocktail (BD Biosciences, Franklin Lakes, NJ, USA) for collected serial samples at diagnosis and at follow-up 5 h prior to staining. from 13 patients that had stable disease (median age: 66 Cells were acquired by a Navios™ cytometer (Beckman years, range 47–81 years; median time to second sam- Coulter, Brea, CA, USA). Data were analyzed using the pling: 39 months, range 30–77 months). For this group, FlowJo v10 software (TreeStar, Ashland, OR, USA) and the median follow-up was 77 months (range 41–101 the Cytobank platform (Santa Clara, CA, USA). months) and only one patient (CLL46) progressed (19 months after second sampling). Clinical characteristics Determination of IL-10 in plasma are summarized in Table 1 and detailed in Supplemen- Concentrations of plasmatic IL-10 were measured using tary Table S2. the Simple Plex™ Assay for the detection of human IL-10 For the purpose of analyzing potential genetic evolu- (R&D Systems, Minneapolis, MN, USA) on Ella™ Auto- tion related to clinical progression, we performed longi- mated ELISA Platform following the manufacturer’s tudinal WES in paired samples from 12 patients at instructions. diagnosis and progression. With a mean read depth of 110X, the limit of detection was set at a minimum B and T lymphocytes co-cultures coverage of 20 reads and a minimum of 0.1 variant allele After negative selection, B-CLL cells or B cells from HD frequency (VAF). In accordance with previous WES were co-cultured with T cells from CLL or HD at 1:2 studies in CLL, mutation rate at both time points was and 1:10 T to B cell ratios as detailed in Supplementary low and consisted of a mean ± SEM (standard error of Methods. the mean) of 12.2 ± 3.3 (range 6–17) somatic single nu- cleotide variants (SNVs) and insertions and deletions Statistical analysis (indels) per exome. (Supplementary Table S3). We then Comparisons were performed using the Wilcoxon screened for clonal shifts from diagnosis to progression matched-pairs rank test or the Mann-Whitney U test in by calculating significant changes in the CCF of the unpaired samples. Differences were considered statisti- alterations detected in each patient [29] (see supplemen- cally significant if P < 0.05. All the statistical analyses tary methods for detailed information about CCF calcu- were carried out and graphed using the GraphPad Prism lation). A significant change was determined if the 95% version 6.0 software. confidence intervals of the CCF in the diagnosis and progression sample did not overlap [29]. We found that Results 50% of patients showed significant changes at progres- CLL cells show limited and non-recurrent genetic changes sion affecting the CCF of at least one alteration. How- at clinical progression ever, the remaining 50% of patients exhibited clonal In a series of 25 patients (median age: 63 years, range stability at clinical progression (Fig. 1a and b). At diag- 40–82 years) that experienced clinical progression nosis, mutations in CLL driver genes [29, 30] were found Table 1 Clinical characteristics of progressing and non-progressing patients Progressed Non-progressed Gender M (68%) M (69%) BINET/RAI stage at diagnosis A0 (72%) A0 (92%) IGHV Status UM (56%) UM (8%) Median Range Median Range Age at diagnosis 63 40–82 66 47–81 TTP (months) 29 5–96 – – Follow-up without progression (months) – – 77 41–101 Time to second sampling (months) 29 5–96 39 30–77 9 Lymphocytes·10 /L at diagnosis 12.20 3.3-65.8 10.40 3.8-31.2 Lymphocytes·109/L at second sampling 73.05 2.3-287.1 22.40 5.2-85.3 Jiménez et al. Biomarker Research (2021) 9:37 Page 4 of 14 Fig. 1 Longitudinal analysis of the CCF of SNVs and indels from paired B-CLL cells at diagnosis and progression before treatment. a Percentage of SNVs and indels with significantly increased (red),decreased (blue) or stable (grey) CCF from diagnosis to progression. Patients harboring mutations in CLL driver genes are marked with an asterisk. Absolute numbers of SNVs and indels detected per patient are detailed in italics inside bars. b Comparison of the CCF with 95% CI for each alteration detected per patient (n = 12) between diagnosis and progression. Significantly increased (red lines), decreased (blue lines) or stable CCF (grey lines) are shown. CLL driver genes are plotted with bold lines and labeled with gene name in 9 out of 12 (75%) patients (mean ± SEM of 1.4 ± 1.2 changes over time affecting these drivers were observed drivers per patient) (Fig. 1a). Among these, only one pa- (Supplementary Table S4). tient (CLL51) showed increased size of drivers CCF at Regarding changes in CNVs by WES, at diagnosis, progression: two variants in NFKBIE and ATM genes CNVs were detected in 10 out of 12 (83%) patients with rised at progression (Fig. 1b, in bold red). Also in this a mean ± SEM of 4.0 ± 4.1 (range 1–12) CNVs per patient, two additional variants from the same genes had patient (Supplementary Table S5 and Supplementary fixed CCF between time points (Fig. 1b, in bold black). Fig. S1). Seven out of 12 (58%) patients had recurrent Furthermore, one patient (CLL31) among those without CNVs associated with CLL, including del(13q), tri(12), alterations in driver genes acquired a mutation at del(11q) and del(17p)(3), but all remained stable over progression affecting the gene TENM1 (CCF = 0.31) time. Nonetheless, we did observe acquisition of del(8p) (Fig. 1b), not previously associated with CLL. and del(15p) with a CCF of 72 and 44%, respectively, at We also analyzed the mutational status of 9 CLL progression in the same patient that showed increased driver genes (TP53, BIRC3, ATM, NOTCH1, SF3B1, CCF in driver genes (CLL51). XPO1, MYD88, FBXW7 and POT1) by targeted next For patients without clinical progression, we analyzed generation sequencing in the same patients. With a the panel of CLL driver genes in paired B-CLL cells at mean read depth of 2000X, the limit of detection was set diagnosis and asymptomatic follow-up (n = 9 patients). at a minimum coverage of 100 reads and a minimum of We identified mutations at diagnosis in 4 out of 9 (44%) 0.05 VAF. Despite high depth sequencing, no additional non-progressing patients. One of them (CLL23) displayed Jiménez et al. Biomarker Research (2021) 9:37 Page 5 of 14 increased VAF in one ATM variant, and another one at the time of progression as denoted by higher co- (CLL47) showed reduced VAF in one mutation affecting expression of PD1 with CD244 or CD160 (Fig. 2d and e). FBXW7 at second sampling (Supplementary Table S4). Altogether, we observed a longitudinal increase from Altogether, these results indicate that adquisition or diagnosis to progression in antigen-experienced effector expansion of genetic alterations is not linked to naturally memory CD8+ T cell subsets together with an increased occurring clinical progression, as previously reported by co-expression of inhibitory receptors. On the contrary, no others [6–12, 14, 15]. significant changes were observed over time in patients that did not progress. Effector memory CD8+ T cells co-expressing inhibitory receptors accumulate at CLL progression while they Terminally exhausted CD8+ T cells (T- remain steady over time in patients with stable disease betdim/−EomeshiPD1hi) accumulate at CLL progression Prior studies in patients diagnosed with CLL have T-bet and Eomesodermin (Eomes) are two T-box tran- shown an increase in defective circulating CD8+ T scription factors that regulate the differentiation cells displaying a terminally differentiated phenotype process of CD8+ T cells after antigen encounter and co- compared to aged-matched healthy donors [18–21, operate in the maintenance of long-term immunity 31, 32]. However, how these CD8+ T cells potentially [36]. Several studies have pointed out that they both evolve from diagnosis to clinical progression using have roles in CD8+ T cell exhaustion. T-bet represses longitudinal samples has not yet been studied. To in- PD1 expression and other inhibitory receptors [37] vestigate this, we analyzed the immunophenotype of while Eomes, in contrast, is related to high expression T cells in paired PBMC samples from our two co- of multiple inhibitory receptors, including PD1 [38, 39]. horts: 19 patients were analyzed at time of diagnosis Accordingly, differential expression of both transcrip- and progression before treatment and 10 patients tion factors combined with moderate or high PD1 levels were analyzed at time of diagnosis and a second time defines two distinctly exhausted CD8+ T cell subsets: point during follow-up without progression (from the progenitor (T-bethiEomesdim/−PD1mid) and the here referred to as “non-progression”). Firstly, the terminal progeny (T-betdim/−EomeshiPD1hi) [38] (Sup- CD4/CD8 ratio at second sampling was significantly plementary Fig. S3). Moreover, higher expression of decreased over time only in patients that progressed, Eomes in CD8+ T cells has been observed in patients in whom effector memory CD8+ T cells (TEM: diagnosed with CLL compared to HD and its function CCR7−CD45RA−) were the sole expanded T cell subset at is essential for proper T cell expansion in CLL mouse progression, whereas no significant changes in CD8+ T models [40]. Since we observed that CD8+ T cells from cells maturation subsets were observed in non-progressors CLL patients gained features of a more severe exhaus- (Supplementary Fig. S2a and b). tion degree at progression, we hypothesized that the PD1 is expressed in chronically stimulated CD8+ T terminal progeny would also be increased at progression. cells and has a relevant role in T cell exhaustion Indeed, we found that the CD8+ progenitor subset [33, 34]. It also has been described as preferentially remained stable over time in both progressors and non- expressed in effector memory CD8+ T cells, which progressors (Fig. 3a and c), while the terminally exhausted are increased in CLL compared to healthy donors CD8+ subpopulation (T-betdim/−EomeshiPD1hi) was sig- [18, 21]. Here, we observed an accumulation of PD1- nificantly increased only in progressing patients (Fig. 3b expressing CD8+ T cells (Supplementary Fig. S2c) and an and c). These findings confirm that CD8+ T cells at pro- enrichment in PD1+ TEM and PD1+ TEMRA CD8+ subsets. gression exhibit a severe terminal exhaustion condition This increase in exhausted CD8+ T cells was not observed likely impairing their ability to control the growth of at non-progression (Fig. 2a to c). To further investigate malignant cells. whether other features of T cell exhaustion and impaired effector function in CD8+ T cells were also enhanced at T cells acquire a distinct transcriptional profile at CLL CLL progression, we measured the co-expression of PD1 progression with other inhibitory receptors, namely CD244 and In order to broadly characterize the alterations that CD160 [33, 35]. In line with our previous findings, we ob- occur over time in T cells from patients diagnosed served that CD8+ T cells co-expressing PD1 with CD244 with CLL cells related to clinical progression, we per- or CD160 significantly rose only at progression, although formed paired RNA-Seq analysis on isolated T cells the fold-change increase in CD8+ T cells expressing (mean purity of 92%): 13 patients were analyzed at PD1+CD160+ was not significantly higher than the one time of diagnosis and progression and 6 patients were observed in non-progressing patients (Supplementary Fig. analyzed at diagnosis and at follow-up (non-progres- S2d to f). In addition, the expanded TEM CD8+ subset in sion). After selecting uniquely mapped reads, the un- progressing patients gained features of severe exhaustion supervised hierarchical clustering analysis of paired Jiménez et al. Biomarker Research (2021) 9:37 Page 6 of 14 Fig. 2 PD1+ effector memory CD8+ T cells subsets and co-expression of inhibitory receptors in progressing and non-progressing CLL patients. a Absolute numbers of PD1+ EM and b PD1+ EMRA CD8+ T cells in progressing (left, n = 18) and non-progressing patients (middle, n = 10) at diagnosis and progression or non-progression. Fold change between time points of PD1+ EM and PD1+ EMRA CD8+ cells in progressing and non-progressing patients (right). c Representative viSNE plots of T cell differentiation subsets (naïve: CCR7+CD45RA+; central memory, CM: CCR7+CD45RA−; effector memory, EM: CCR7−CD45RA− and EM CD45RA+, EMRA: CCR7−CD45RA+) and PD1 expression in CD4+ and CD8+ T cells at the two time points from one representative patient from each group. d Percentages of EM CD8+ cells co-expressing CD244 and PD1 (n = 12) and e CD160 and PD1 (n = 10). Graphs show mean ± SEM or paired values. (*P < 0.05; **P < 0.01; ***P < 0.001; Wilcoxon matched paired test or Mann-Whitney test) Jiménez et al. Biomarker Research (2021) 9:37 Page 7 of 14 Fig. 3 Progenitor (T-bethiEomesdim/−PD1mid) and terminally (T-betdim/−EomeshiPD1hi) exhausted CD8+ T cells in progressing and non-progressing CLL patients. a Percentages of T-bethiEomesdim/− out of PD1midCD8+ T cells and b T-betdim/−Eomeshi out of PD1hiCD8+ T cells in progressing (left, n = 12) and non-progressing patients (middle, n = 9) at diagnosis and progression or non-progression. Fold change between time points of both subsets comparing progressing and non-progressing patients (right). c Representative viSNE plots of PD1, T-bet and Eomes expression in CD4+ and CD8+ T cells at the two time points from one representative patient from each group. Graphs show mean ± SEM or paired values. (*P < 0.05; **P < 0.01; ***P < 0.001; Wilcoxon matched paired test or Mann-Whitney test) samples from progressing patients defined two main expressed in T cells from non-progressed patients at clusters: one corresponding exclusively to T cells at the time of follow-up after supervised analysis (all progression and another cluster including T cells at genes padj< 0.05 are detailed in Supplementary Table diagnosis plus two samples at progression, highlighting S6). Moreover, those 3 differentially expressed genes that the transcriptional profile of T cells at progression found in non-progressing patients were also found at was clearly distinct from that of T cells at diagnosis progression. Briefly, the transcriptional profile of T (Fig. 4). A total of 80 genes (including protein coding cells at progression suggests lower mobility and differ- and lncRNA transcripts) were significantly up or entiation capacity as well as an impairment in mito- downregulated in T cells from diagnosis to progres- chondrial oxidative phosphorylation (Supplementary sion, while in contrast only 3 genes were differentially Table S7), essential processes for the maintenance of T Jiménez et al. Biomarker Research (2021) 9:37 Page 8 of 14 Fig. 4 RNA-Seq of T cells from progressing CLL patients. Heatmap showing the unsupervised hierarchical clustering of the top-50 differentially expressed genes from paired sorted T cells at diagnosis and progression (n = 13) cell effector functions [41]. Additionally, genes related upregulation of genes associated with immune re- to fatty acids and amino acids catabolism and glucose sponse and exhaustion [42–44]. Collectively, these re- transporters were upregulated, while lower expression sults point towards an impaired cytoskeleton of genes related to the synthesis of cellular compo- formation, mitochondrial metabolism and immune nents and RNA processing mechanisms were identified dysregulation, consistent with the exhausted and at progression, suggesting a potentially dysregulated T dysfunctional status of T cells that is aggravated at cell metabolism. T cells at progression also showed CLL progression. Jiménez et al. Biomarker Research (2021) 9:37 Page 9 of 14 CLL cells from patients at clinical progression have that secretion of soluble factors lead to upregulation of enhanced capacity to induce PD1 expression in CD8+ T exhaustion markers. In this regard, CLL cells are known cells via soluble factors including IL-10 to exhibit features of regulatory B cells, such as IL-10 The observed increased exhaustion at progression com- production, [24] which has been found to have immuno- pared to patients that had not progressed could be suppressive function in CLL [45, 46], be related to mainly caused by a potentially increased interaction with shorter overall survival [47] and to correlate with the ex- malignant cells due to higher tumoral load; alternatively, pression of PD1 in tumor infiltrating lymphocytes in malignant cells at progression could be intrinsically solid tumors [48]. Based on this, we initially measured more capable of inducing exhaustion in T cells. In order the levels of IL-10 in paired plasma samples from CLL to gain insight into the functional mechanisms that patients and found that plasmatic IL-10 significantly in- could trigger said increased exhaustion of CD8+ T cells, creased at progression while remained stable over time we compared the effect on T cells of malignant B-CLL in non-progressing patients (Fig. 5e). Next, to investigate cells obtained from patients at clinical progression vs. whether malignant cells acquire an increased capacity to malignant cells from patients at follow-up of stable dis- produce IL-10 at progression and, therefore, higher plas- ease (non-progression). To do that we firstly co-cultured matic levels are not simply reflecting the higher tumoral T cells from patients with CLL (T-CLL) with increasing load at progression, we assessed the production of IL-10 concentrations of either healthy B cells (B-HD) or au- by CLL cells in vitro after microenvironmental stimuli tologous B-CLL cells obtained from patients at progres- [27]. We detected an increased percentage of IL-10+ sion or from patients at follow-up (non-progression). CLL cells only in malignant cells obtained at progression After 7 days of co-culture, we found that PD1 expression (Fig. 5f and Supplementary Fig. S4d), indicating that was increased in CD8+ T cells in presence of progressed leukemic cells from patients at clinical progression have B-CLL cells at any leukemic cell to T cell ratio, while B- indeed increased potential to create an immunosuppres- HD cells did not induce changes in PD1 expression sive microenvironment. Moreover, the induction of PD1 (Fig. 5a). Moreover, B-CLL cells from patients that did expression in vitro in CD8+ T cells was partially blocked not progress were only able to induce PD1 expression in after IL-10 neutralization (Fig. 5g). Finally, we analyzed autologous CD8+ T cells at the highest B to T cell ratio the potential accumulation of MDSCs as an additional tested, suggesting that there are intrinsic characteristics source of plasmatic IL-10, and we observed [49] an accu- in malignant B cells from patients that are in need of mulation of these immunosupresive cells over time in all treatment that can contribute to the observed T cell ex- CLL patients regardless of their clinical evolution, haustion. The same results were obtained when looking although the degree of increment was higher in progres- at PD1 and CD244 co-expression (supplementary Fig. sing patients (Supplementary Fig. S4e). These observa- S4a). Therefore, and in order to further compare the tions suggest that at the time of progression not only capacity to induce exhaustion of malignant B cells from there is an increased load of leukemic cells, but these patients at progression versus malignant cells from non- also have a higher IL-10 capacity production, which can progressing patients, we co-cultured T cells from healthy contribute to the increased induction of CD8+ T cell donors (T-HD) with both types of malignant cells. In exhaustion. this setting, only progressed B-CLL cells were capable of inducing PD1 expression in CD8+ T cells whereas non- Discussion progressed B-CLL cells were not (Fig. 5c), while they The biological processes that lead to clinical progression showed no differences in the induction of PD1 and from early asymptomatic stages in patients diagnosed CD244 co-expression. Thus, although we cannot com- with CLL are not well understood and, consequently, pletely rule out that the observed enhanced exhaustion the pathogenesis behind the natural history of this at progression is merely caused by increased tumoral disease remains unclear. In this regard, longitudinal load, these results indicate that malignant cells at pro- studies from diagnosis to clinical progression are essen- gression are intrinsically more capable of inducing PD-1 tial to elucidate these mechanisms. To the best of our expression in both autologous or HD-derived T cells. knowledge, this is the first comprehensive longitudinal To investigate whether this T cell state induced by analysis of the genetic and immunological processes CLL cells occurs through a cell-to-cell mediated mech- driving disease progression in CLL. The longitudinal anism or, otherwise, is mediated by soluble factors, T- molecular study herein presented showed limited and CLL and B-CLL co-cultures were performed with non-recurrent molecular changes in CLL cells at pro- transwells. We observed that the induction of PD1 and gression, indicating that genetic clonal evolution is not CD244 in CD8+ T cells was equivalent when there was the main driver of clinical progression, like previously no contact between autologous T cells and leukemic observed in other series [6–15]. Conversely, longitudinal cells (Fig. 5d and Supplementary Fig. S4c, suggesting studies focused on potential changes in the immune Jiménez et al. Biomarker Research (2021) 9:37 Page 10 of 14 Fig. 5 (See legend on next page.) Jiménez et al. Biomarker Research (2021) 9:37 Page 11 of 14 (See figure on previous page.) Fig. 5 PD1 expression in CD8+ T cells after co-culture with B-HD cells or B-CLL cells and contribution of soluble factors to PD1 expression. a Percentages of PD1+ CD8+ T cells from progressing CLL patients after stimulation with anti-CD3 and anti-CD28 for 7 days (grey dots) and in presence of B-HD cells (yellow dots) or B-CLL cells at the time of progression (red dots) at the indicated T:B ratios (n = 10). b Percentages of PD1+CD8+ T cells from non-progressing CLL patients after stimulation with anti-CD3 and anti-CD28 for 7 days (grey dots) and in presence of B- HD cells (yellow dots) or B-CLL cells at asymptomatic follow-up (blue dots) at the indicated T:B ratios (n = 7). c Percentages of PD1+CD8+ T cells from age-matched healthy donors (T-HD) after stimulation with anti-CD3 and anti-CD28 for 7 days (grey dots) and in presence of B-CLL cells at progression (bold red dots) or B-CLL cells at asymptomatic follow-up (bold blue dots) at the indicated T:B ratios (n = 6). d Percentages of PD1+CD8+ T cells from CLL patients after stimulation with anti-CD3 and anti-CD28 for 7 days and in contact with B-CLL cells or separated by transwell inserts at 1:10 T:B ratio (n = 14). e Concentration (pg/ml) of IL-10 in plasma from progressing (left, n = 19) and non-progressing patients (middle, n = 10). Increment of IL-10 in plasma between time points comparing progressing and non-progressing patients (right). f Percentage of IL-10+ CLL cells in progressing (left, n = 7) and non-progressing patients (middle, n = 5) at the two time points after co-culture of paired PBMCs with bone marrow stromal UE6E7T-2 cells, CD40L and TLR9L for 48 h. Increment of IL-10+ CLL cells between time points in progressing and non- progressing patients (right). g Percentage of PD1+CD8+ T cells from CLL after 7-day co-culture with progressing B-CLL cells at 1:10 T:B ratio, following the protocol previously described, and after adding anti-human IL-10 neutralizing antibody (n = 11). Graphs show mean ± SEM or paired values (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; Wilcoxon matched paired test or Mann-Whitney test) microenvironment have not yet been conducted. Our inducing exhaustion in both autologous and HD-derived analysis on the immune kinetics found an increasingly CD8+ T cells compared to malignant CLL cells from dysfunctional CD8+ T cell compartment in progressing patients with stable disease, as well as increased capacity patients that was not observed in those patients that to secrete IL-10 upon microenvironmental stimuli. In remained asymptomatic. Moreover, we showed that sol- addition, we found that PD1 upregulation depended uble factors, such as IL-10 produced by CLL cells, play a mainly on soluble factors. Therefore, although the accu- role in CD8+ T cell exhaustion and in the progression of mulation of exhausted CD8+ T cells in progressing the disease. patients might be merely the consequence of the higher An altered anti-tumor immune response is evidenced tumoral load characterizing progression in the majority in CLL by diverse factors affecting mainly T cells of patients (see Supplementary Table S2 for clinical [18–21, 31, 32]. Recent studies using the Eμ-TCL1 data), our data showing increased in vitro capacity to in- mouse model indicate that CD8+ T cells can delay duce T cell exhaustion and to produce IL-10 are com- CLL progression at the same time that their expres- patible with an scenario where accumulating malignant sion of inhibitory receptors progressively increases cells also acquire higher immunosuppressive properties [50]. Also, although only studied in 3 cases, in spon- along the course of the disease only in patients that will taneously regressing CLL the expression of PD1 in eventually progress. These cells would then promote en- CD8+ T cells also decreased when malignant cells gagement in a positive feed-back system further increas- declined [51]. Accordingly, our longitudinal immune ing CD8+ T cell exhaustion and ultimately facilitating analysis in CLL patients showed that effector memory the transition to clinical progression. In this regard, CD8+ T cell subsets expressing PD1 accumulated specific- Gonnord et al. recently described that CD8+ T cells from ally at clinical progression. In addition, EM CD8+ T cells untreated CLL patients that will need therapy within 6 co-expressing PD1 with CD244 and CD160 and terminally months after analysis display an unique signature which exhausted (T-betdim/−EomeshiPD1hi) CD8+ T cells were is not correlated with the time that CD8+ T cells have significantly increased at progression, denoting severe dys- been exposed to CLL cells [52], indicating again that the function in CD8+ T cells over time. Importantly, these exhaustion of T cells is not a mere product of increased changes were not observed in non-progressing patients, exposure to malignant cells, either in time or in tumoral supporting the primary role of immune dysfunction in load. Further studies trying to definitively figure out the CLL progression. mechanisms behind the coevolution of the immune sys- Broad analysis of transcriptomic changes also revealed tem and malignant cells in CLL are definitively worth significantly different expression profiles in T cells at performing. progression compared to diagnosis while only minor dif- Our results provide pre-clinical evidence to support ferences were found in non-progressing patients along the design of clinical studies aimed at improving anti- time. Although the low quality of the RNA we obtained tumoral T cell responses. This may be particularly ad- hampered a broader analysis, we were able to identify vantageous in the context of treatment with autologous changes in the transcriptome specifically associated with chimeric antigen receptor (CAR)- T cells derived from clinical progression when using only uniquely mapped patients in an advanced disease stage, which display low reads to avoid potential artifacts due to low input. cytotoxic activity and achieve low response rates [53]. In this study, we also identified that CLL cells obtained Notably, immunomodulatory effects of ibrutinib therapy at the time of progression had enhanced capacity of are able to recover T cell function and subsequently Jiménez et al. Biomarker Research (2021) 9:37 Page 12 of 14 increase the response rate to CAR-T cell therapy non-progressing patients (n = 10) at diagnosis and progression or non- [54, 55]. Moreover, these results lay the foundations progression. b Absolute numbers of CD8+ T cell differentiation subsets for the clinical testing of immunotherapy in early (naive: CCR7+CD45RA+; central memory, CM: CCR7+CD45RA−; effector memory, EM: CCR7−CD45RA− and EM CD45RA+, EMRA: CCR7−CD45RA+) stages of the disease to prevent or delay clinical pro- in progressing (n = 19) and non-progressing patients (n = 10) at diagnosis gression. In addition, the identification of signs of and progression or non-progression. c Absolute numbers of PD1+CD8+ T increased immunosuppression could be exploited in cells in progressing (left, n = 19) and non-progressing patients (middle, n = 10) at diagnosis and progression or non-progression. Fold change of the future to improve early detection of patients that PD1+CD8+ T cells between time points comparing progressing and non- will likely progress. progressing patients (right). d Percentage of PD1+CD244+ CD8+ T cells in progressing (left, n = 12) and non-progressing patients (middle, n = 9) at diagnosis and progression or non-progression. Fold change of Conclusions PD1+CD244+CD8+ T cells between time points comparing progressing Collectively, our findings indicate that at clinical pro- and non-progressing patients (right). e Percentage of PD1+ 160+ CD8+ T cells in progressing (left, n = 12) and non-progressing patients (middle, gression CLL cells exhibit limited genetic changes from n = 9) at diagnosis and progression or non-progression. Fold change of diagnosis, while CD8+ T cells show increased exhaustion PD1+CD160+CD8+ T cells between time points comparing progressing that can be induced by IL-10 secreted by malignant B and non-progressing patients (right). f Density plots of PD1, CD160 and CD244 coexpression in CD8+ T cells in representative patients at diagno- cells. In contrast, patients that remain stable over time sis and progression and at diagnosis and non-progression. Graphs show did not experience significant changes in their genetic or mean ± SEM or paired values (*P < 0.05; **P < 0.01; ***P < 0.001; Wilcoxon T cell compartments. Also, malignant cells from patients matched paired test or Mann-Whitney test). Supplementary Figure S3. Flow cytometric analysis of progenitor and terminal CD8+ subsets. Gating at clinical progression showed enhanced capacity to in- strategy followed for the identification of T-bethiEomesdim/−PD1mid and T- duce exhaustion-related features in CD8+ T cells ex vivo. betdim/−EomeshiPD1hi CD8+ populations. Supplementary Figure S4. These results provide evidence to support the explor- Co-expression of PD1 and CD244 in CD8+ T cells after co-culture with B- CLL cells. MDSCs in progressing and non-progressing CLL patients. a Per- ation of immunotherapeutic interventions in early stages centages of PD1+CD244+ cells out of CD8+ T cells from progressing (left) aiming at avoiding or delaying clinical progression of the and non-progressing (right) CLL patients after stimulation with anti-CD3 disease. Additionally, further analysis of both immuno- and anti-CD28 for 7 days (grey dots) and in presence of B-HD cells (yel- low dots) or B-CLL cells at the time of progression (red dots, n = 10) or logical and malignant B-CLL cell-intrinsic factors in lar- asymptomatic follow-up (blue dots, n = 7) at the indicated T:B ratios. b ger cohorts of patients will help us better identify those Percentages of PD1+CD244+ cells out of CD8+ T cells from healthy age- patients that are more likely to progress shortly after matched donors (T-HD) after stimulation with anti-CD3 and anti-CD28 for 7 days (grey dots) and in presence of B-CLL cells at progression (bold red diagnosis. dots) or B-CLL cells at asymptomatic follow-up (bold blue dots) at the in- dicated T:B ratios. c Percentages of CD8+ T cells from CLL patients co- Abbreviations expressing PD1 and CD244 after stimulation with anti-CD3 and anti-CD28 CAR: chimeric antigen receptor; CCF: cancer cell fraction; CI: confidence for 7 days and in contact with B-CLL cells or separated by transwell in- interval; CLL: chronic lymphocytic leukemia; CNV: copy number variation; serts at 1:10 T:B ratio for 7 days (n = 14).d Dot plots of IL-10+ B cells gated IWCLL: international workshop on CLL; MDSC: myeloid-derived suppressor on CD19+CD5+ cells after 5 h of leukocyte stimulation (PIB), or brefeldin A cells; PBMCs: peripheral blood mononuclear cells; RNA-seq: RNA sequencing; (BFA) as control, from one representative patient from the progressed SEM: standard error of the mean; TTP: time to progression; VAF: variant allele and non-progressed groups respectively. e Percentage of MDSCs fraction; WES: whole exome sequencing (CD14+HLA-DRlow/−) out of CD14+ cells in progressing (left, n = 17) and non-progressing patients (middle, n = 10) at diagnosis and progression or non-progression. Increment of MDSCs between time points comparing Supplementary Information progressing and non-progressing patients (right). Graphs show mean ± The online version contains supplementary material available at https://doi. SEM or paired values (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; org/10.1186/s40364-021-00290-z. Wilcoxon matched paired test or Mann-Whitney test). Additional file 1: Supplementary Table S2. Detailed clinical Acknowledgements characteristics from CLL patients. Supplementary Table S3. SNVs and The authors thank the patients and healthy donors for their blood donations. Indels analysis by WES in progressing CLL patients. Supplementary The authors also thank the Cellex Foundation for providing research facilities Table S4. CLL driver gene analysis (TP53, BIRC3, ATM, NOTCH1, SF3B1, and equipment and the CERCA Programme / Generalitat de Catalunya for XPO1, MYD88, FBXW7 and POT1) by NGS in progressing and non- institutional support. progressing CLL patients. Supplementary Table S5. CNVs analysis by WES in progressing CLL patients. Supplementary Table S6. Differen- tially expressed genes in T cells from CLL patients. Authors’s contributions I.J., M.C. and F.B. designed reseach; I.J., B.T-V., J.C.N., J.C., M.M., L.P. and C.F-J. Additional file 2: Supplementary Table S1. Monoclonal antibodies performed research; I.J., B.T-V., M.C. and F.B. analyzed and interpreted data; (mAbs). Supplementary Table S7. Highlighted dysregulated genes in P.A., S.B., J.C.N., C.P-G., R.V-M., A.E-C., G.P. and G.I. contributed to data analysis; T CLL cells at progression. Supplementary Figure S1. Longitudinal P.A., S.B., G.I., M.J.T. and J.A.G-M. provided patient samples; M.C. and F.B. analysis of the CCF of CNVs from paired B-CLL cells at diagnosis and pro- supervised the work; I.J., M.C. and F.B. wrote the manuscript; all authors gression before treatment. Comparison of the CCF with 95% CI for each reviewed the manuscript. The author(s) read and approved the final CNV detected per patient (n = 10) between diagnosis and progression. manusript. Significantly increased (red lines) and stable CCF (grey lines) are shown. Recurrent CNVs in CLL (del(13q), del(11q), del(17p) and +[12]) are plotted with bold lines and labeled with CNV name: stable CCF (bold black) is Funding shown. Supplementary Figure S2. CD8+ T cell differentiation subsets This work was supported by the Instituto de Salud Carlos III, Fondo de Investigaciones Sanitarias (PI17/00950, M.C., PI18/01392, P.A. and PI17/00943, and PD1 expression in CD8+ T cells from progressing and non- progressing CLL patients. a CD4/CD8 ratio in progressing (n = 19) and F.B.) and co-financed by the European Regional Development Fund (ERDF) and Fundación Asociación Española Contra el Cáncer (M.C. and P.A.), Gilead Jiménez et al. Biomarker Research (2021) 9:37 Page 13 of 14 Fellowships (GLD16/00144, GLD18/00047, F.B.) and Fundació la Marató de 6. Stilgenbauer S, Sander S, Bullinger L, Benner A, Leupolt E, Winkler D, et al. TV3 (201905–30-31 F.B). S.B. is the recipient of a postdoctoral fellowship from Clonal evolution in chronic lymphocytic leukemia: acquisition of high-risk Fundación Alfonso Martin Escudero. R.V-M. is supported by a Torres Quevedo genomic aberrations associated with unmutated VH, resistance to therapy, fellowship from the Spanish Ministry of Science and Innovation (PTQ-16- and short survival. Haematologica. 2007;92(9):1242–5. https://doi.org/1 08623). A.E-C. is funded by ISCIII/MINECO (PT17/0009/0019) which is co- 0.3324/haematol.10720. funded by FEDER. M.C. holds a contract from Ministerio de Ciencia, 7. Gunnarsson R, Mansouri L, Isaksson A, Goransson H, Cahill N, Jansson M, Innovación y Universidades (RYC-2012-2018). et al. Array-based genomic screening at diagnosis and during follow-up in chronic lymphocytic leukemia. Haematologica. 2011;96(8):1161–9. https:// Availability of data and materials doi.org/10.3324/haematol.2010.039768. For original data, please contact the corresponding author. WES and RNA- 8. Landau DA, Carter SL, Stojanov P, McKenna A, Stevenson K, Lawrence Seq data generated during the current study are deposited at EGA and GEO MS, et al. Evolution and impact of subclonal mutations in chronic under accession numbers EGAS00001004116 and GSE141787, respectively. lymphocytic leukemia. Cell. 2013;152(4):714–26. https://doi.org/10.1016/j. cell.2013.01.019. 9. Nadeu F, Delgado J, Royo C, Baumann T, Stankovic T, Pinyol M, et al. Clinical Declarations impact of clonal and subclonal TP53, SF3B1, BIRC3, NOTCH1, and ATM mutations in chronic lymphocytic leukemia. Blood. 2016;127(17):2122–30. Ethics approval and consent to participate https://doi.org/10.1182/blood-2015-07-659144. A written informed consent was obtained from all individuals in accordance 10. Rose-Zerilli MJJ, Gibson J, Wang J, Tapper W, Davis Z, Parker H, et al. with the declaration of Helsinki and the study was approved by the clinical Longitudinal copy number, whole exome and targeted deep sequencing of research ethics committee at Vall d’Hebron Hospital Universitari. “good risk” IGHV-mutated CLL patients with progressive disease. Leukemia. 2016;30(6):1301–10. https://doi.org/10.1038/leu.2016.10. Consent for publication 11. Smith EN, Ghia EM, DeBoever CM, Rassenti LZ, Jepsen K, Yoon K-A, et al. Not applicable. Genetic and epigenetic profiling of CLL disease progression reveals limited somatic evolution and suggests a relationship to memory-cell development. Competing interests Blood Cancer J. 2015;5(4):e303. https://doi.org/10.1038/bcj.2015.14. P.A. has received honoraria from Janssen, Roche, Celgene and AbbVie. G.I. 12. Hernández-Sánchez M, Kotaskova J, Rodríguez AE, Radova L, Tamborero D, has received honoraria from Celgene, Novartis, Roche, AbbVie, Janssen and Abáigar M, et al. CLL cells cumulate genetic aberrations prior to the first Gilead. J.A.G-M. has received research funding and honoraria from Roche, therapy even in outwardly inactive disease phase. Leukemia. 2019;33(2):518– Janssen and AbbVie. M.C. has received research funding from, Pharmacyclics, 58. Roche, ArQule and AstraZeneca. F.B. has received research funding and 13. Gruber M, Bozic I, Leshchiner I, Livitz D, Stevenson K, Rassenti L, et al. honoraria from Roche, Celgene, Takeda, AstraZeneca, Novartis, AbbVie and Growth dynamics in naturally progressing chronic lymphocytic leukaemia. Janssen. All remaining authors have declared no conflicts of interest. Nature. 2019;570(7762):474–9. https://doi.org/10.1038/s41586-019-1252-x. 14. Leeksma AC, Taylor J, Wu B, Gardner JR, He J, Nahas M, et al. Clonal diversity Author details predicts adverse outcome in chronic lymphocytic leukemia. Leukemia. 2019; 1 Experimental Hematology, Vall d’Hebron Institute of Oncology (VHIO), Vall 33(2):390–402. https://doi.org/10.1038/s41375-018-0215-9. d’Hebron Barcelona Hospital Campus, C/Natzaret 115-117, 08035 Barcelona, 15. Ramassone A, D’Argenio A, Veronese A, Basti A, Soliman SHA, Volinia S, Spain. 2Department de Medicina, Universitat Autònoma de Barcelona, 08193 et al. Genetic dynamics in untreated CLL patients with either stable or Bellaterra, Spain. 3Servei d’Hematologia, Vall d’Hebron Hospital Universitari, progressive disease: a longitudinal study. J Hematol Oncol. 2019;12(1):114. Experimental Hematology, Vall d’Hebron Institute of Oncology (VHIO), Vall https://doi.org/10.1186/s13045-019-0802-x. d’Hebron Barcelona Hospital Campus, Passeig Vall d’Hebron 119-129, 08035 16. Fernandez V, Jares P, Salaverria I, Gine E, Bea S, Aymerich M, et al. Gene Barcelona, Spain. 4DREAMgenics, 33011 Oviedo, Spain. 5Centre for Genomic expression profile and genomic changes in disease progression of early- Regulation, Barcelona Institute of Science and Technology, 08003 Barcelona, stage chronic lymphocytic leukemia. Haematologica. 2008;93(1):132–6. Spain. 6Universitat Pompeu Fabra, 08002 Barcelona, Spain. 7Servei https://doi.org/10.3324/haematol.11694. d’Immunologia, Vall d’Hebron Institut de Recerca (VHIR), Vall d’Hebron 17. Riches JC, Gribben JG. Immunomodulation and immune reconstitution in Hospital Universitari, Vall d’Hebron Barcelona Hospital Campus, Passeig Vall chronic lymphocytic leukemia. Semin Hematol. 2014;51(3):228–34. https:// d’Hebron 119-129, 08035 Barcelona, Spain. 8Department of Hematology, doi.org/10.1053/j.seminhematol.2014.05.006. Clínic University Hospital, INCLIVA Biomedical Research Institute, 46010 18. Riches JC, Davies JK, McClanahan F, Fatah R, Iqbal S, Agrawal S, et al. T cells Valencia, Spain. 9Department of Hematology, Puerta de Hierro University from CLL patients exhibit features of T-cell exhaustion but retain capacity Hospital, 28222, Majadahonda, Madrid, Spain. for cytokine production. Blood. 2013;121(9):1612–21. https://doi.org/10.1182/ blood-2012-09-457531. Received: 2 December 2020 Accepted: 26 April 2021 19. Ramsay AG, Johnson AJ, Lee AM, Gorgün G, Le Dieu R, Blum W, et al. Chronic lymphocytic leukemia T cells show impaired immunological synapse formation that can be reversed with an immunomodulating drug. J References Clin Invest. 2008;118(7):2427–37. https://doi.org/10.1172/JCI35017. 1. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmutated Ig 20. Ramsay AG, Clear AJ, Fatah R, Gribben JG. Multiple inhibitory ligands induce V(H) genes are associated with a more aggressive form of chronic impaired T-cell immunologic synapse function in chronic lymphocytic lymphocytic leukemia. Blood. 1999;94(6):1848–54. https://doi.org/10.1182/ leukemia that can be blocked with lenalidomide: establishing a reversible blood.V94.6.1848. immune evasion mechanism in human cancer. Blood. 2012;120(7):1412–21. 2. Crespo M, Bosch F, Villamor N, Bellosillo B, Colomer D, Rozman M, et al. https://doi.org/10.1182/blood-2012-02-411678. ZAP-70 expression as a surrogate for immunoglobulin-variable-region 21. Palma M, Gentilcore G, Heimersson K, Mozaffari F, Näsman-Glaser B, Young mutations in chronic lymphocytic leukemia. N Engl J Med. 2003;348(18): E, et al. T cells in chronic lymphocytic leukemia display dysregulated 1764–75. https://doi.org/10.1056/NEJMoa023143. expression of immune checkpoints and activation markers. Haematologica. 3. Döhner H, Stilgenbauer S, Benner A, Leupolt E, Kröber A, Bullinger L, et al. 2017;102(3):562–72. https://doi.org/10.3324/haematol.2016.151100. Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J 22. Nunes C, Wong R, Mason M, Fegan C, Man S, Pepper C. Expansion of a CD8(+ Med. 2000;343(26):1910–6. https://doi.org/10.1056/NEJM200012283432602. )PD-1(+) replicative senescence phenotype in early stage CLL patients is 4. Bosch F, Dalla-Favera R. Chronic lymphocytic leukaemia: from genetics to associated with inverted CD4:CD8 ratios and disease progression. Clin Cancer treatment. Nat Rev Clin Oncol. 2019;16(11):684–701. Res. 2012;18(3):678–87. https://doi.org/10.1158/1078-0432.CCR-11-2630. 5. International CLL-IPI working group. An international prognostic index for 23. Jitschin R, Braun M, Buttner M, Dettmer-Wilde K, Bricks J, Berger J, et al. CLL- patients with chronic lymphocytic leukaemia (CLL-IPI): a metaanalysis of cells induce IDOhi CD14+HLA-DRlo myeloid-derived suppressor cells that individual patient data. Lancet Oncol. 2016;17(6):779–90. https://doi.org/10.1 inhibit T-cell responses and promote TRegs. Blood. 2014;124(5):750–60. 016/S1470-2045(16)30029-8. Epub 2016 May 13. https://doi.org/10.1182/blood-2013-12-546416. Jiménez et al. Biomarker Research (2021) 9:37 Page 14 of 14 24. DiLillo DJ, Weinberg JB, Yoshizaki A, Horikawa M, Bryant JM, Iwata Y, et al. Differentiation Program Initiated Early during Tumorigenesis. Immunity. Chronic lymphocytic leukemia and regulatory B cells share IL-10 2016;45(2):389–401. competence and immunosuppressive function. Leukemia. 2013;27(1):170– 45. Alhakeem SS, McKenna MK, Oben KZ, Noothi SK, Rivas JR, Hildebrandt GC, 82. https://doi.org/10.1038/leu.2012.165. et al. Chronic lymphocytic leukemia-derived IL-10 suppresses antitumor 25. Valdés-Mas R, Bea S, Puente DA, López-Otín C, Puente XS. Estimation of immunity. J Immunol. 2018;200(12):4180–9. https://doi.org/10.4049/ copy number alterations from exome sequencing data. PLoS One. 2012; jimmunol.1800241. 7(12):e51422. https://doi.org/10.1371/journal.pone.0051422. 46. Drennan S, D’Avola A, Gao Y, Weigel C, Chrysostomou E, Steele AJ, et al. IL- 26. Shinde J, Bayard Q, Imbeaud S, Hirsch TZ, Liu F, Renault V, et al. Palimpsest: 10 production by CLL cells is enhanced in the anergic IGHV mutated subset an R package for studying mutational and structural variant signatures and associates with reduced DNA methylation of the IL10 locus. Leukemia. along clonal evolution in cancer. Bioinformatics. 2018;34(19):3380–1. 2017;31(8):1686–94. https://doi.org/10.1038/leu.2016.356. 27. Purroy N, Abrisqueta P, Carabia J, Carpio C, Calpe E, Palacio C, et al. 47. Fayad L, Keating MJ, Reuben JM, O’Brien S, Lee BN, Lerner S, et al. Targeting the proliferative and chemoresistant compartment in chronic Interleukin-6 and interleukin-10 levels in chronic lymphocytic leukemia: lymphocytic leukemia by inhibiting survivin protein. Leukemia. 2014;28(10): correlation with phenotypic characteristics and outcome. Blood. 2001;97(1): 1993–2004. https://doi.org/10.1038/leu.2014.96. 256–63. https://doi.org/10.1182/blood.V97.1.256. 28. Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero G, Döhner H, 48. Thommen DS, Koelzer VH, Herzig P, Roller A, Trefny M, Dimeloe S, et al. A et al. iwCLL guidelines for diagnosis, indications for treatment, response transcriptionally and functionally distinct PD-1 + CD8 + T cell pool with assessment, and supportive management of CLL. Blood. 2018;131(25):2745– predictive potential in non-small-cell lung cancer treated with PD-1 60. https://doi.org/10.1182/blood-2017-09-806398. blockade. Nat Med. 2018;24(7):994–1004. https://doi.org/10.1038/s41591-01 29. Landau DA, Tausch E, Taylor-Weiner AN, Stewart C, Reiter JG, Bahlo J, et al. 8-0057-z. Mutations driving CLL and their evolution in progression and relapse. 49. Azzaoui I, Uhel F, Rossille D, Pangault C, Dulong J, Priol JL, et al. T-cell defect Nature. 2015;526(7574):525–30. https://doi.org/10.1038/nature15395. in diffuse large B-cell lymphomas involves expansion of myeloid-derived 30. Gonzalez-Perez A, Perez-Llamas C, Deu-Pons J, Tamborero D, Schroeder MP, suppressor cells. Blood. 2016;128(8):1081–92. https://doi.org/10.1182/blood-2 Jene-Sanz A, et al. IntOGen-mutations identifies cancer drivers across tumor 015-08-662783. types. Nat Methods. 2013;10(11):1081–2. https://doi.org/10.1038/nmeth.2642. 50. Hanna BS, Roessner PM, Yazdanparast H, Colomer D, Campo E, Kugler S, 31. Mackus WJM, Frakking FNJ, Grummels A, Gamadia LE, De Bree GJ, Hamann et al. Control of chronic lymphocytic leukemia development by clonally- D, et al. Expansion of CMV-specific CD8+CD45RA+CD27- T cells in B-cell expanded CD8+ T-cells that undergo functional exhaustion in secondary chronic lymphocytic leukemia. Blood. 2003;102(3):1057–63. https://doi.org/1 lymphoid tissues. Leukemia. 2019;33(3):625–37. 0.1182/blood-2003-01-0182. 51. Kwok M, Oldreive C, Rawstron AC, Goel A, Papatzikas G, Jones RE, et al. 32. Brusa D, Serra S, Coscia M, Rossi D, D’Arena G, Laurenti L, et al. The PD-1/ Integrative analysis of spontaneous CLL regression highlights genetic and PD-L1 axis contributes to T-cell dysfunction in chronic lymphocytic microenvironmental interdependency in CLL. Blood. 2020;135(6):411–28. leukemia. Haematologica. 2013;98(6):953–63. https://doi.org/10.3324/haema 52. Gonnord P, Costa M, Abreu A, Peres M, Ysebaert L, Gadat S, et al. tol.2012.077537. Multiparametric analysis of CD8+ T cell compartment phenotype in chronic 33. Gros A, Robbins PF, Yao X, Li YF, Turcotte S, Tran E, et al. PD-1 identifies the lymphocytic leukemia reveals a signature associated with progression patient-specific CD8(+) tumor-reactive repertoire infiltrating human tumors. toward therapy. Oncoimmunology. 2019;8(4):e1570774. Available from: J Clin Invest. 2014;124(5):2246–59. https://doi.org/10.1172/JCI73639. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6422371/. 34. Ahmadzadeh M, Johnson LA, Heemskerk B, Wunderlich JR, Dudley ME, 53. Porter DL, Hwang W-T, Frey NV, Lacey SF, Shaw PA, Loren AW, et al. White DE, et al. Tumor antigen-specific CD8 T cells infiltrating the tumor Chimeric antigen receptor T cells persist and induce sustained remissions in express high levels of PD-1 and are functionally impaired. Blood. 2009; relapsed refractory chronic lymphocytic leukemia. Sci Transl Med. 2015; 114(8):1537–44. https://doi.org/10.1182/blood-2008-12-195792. 7(303):303ra139. 35. Blackburn SD, Shin H, Haining WN, Zou T, Workman CJ, Polley A, et al. 54. Gill SI, Vides V, Frey NV, Metzger S, O’Brien M, Hexner E, et al. Coregulation of CD8+ T cell exhaustion by multiple inhibitory receptors Prospective clinical trial of anti-CD19 CAR T cells in combination with during chronic viral infection. Nat Immunol. 2009;10(1):29–37. https://doi. Ibrutinib for the treatment of chronic lymphocytic leukemia shows a org/10.1038/ni.1679. high response rate. Blood. 2018;132(Suppl 1):298–8. https://doi.org/10.11 36. Intlekofer AM, Takemoto N, Wherry EJ, Longworth SA, Northrup JT, Palanivel VR, 82/blood-2018-99-115418. et al. Effector and memory CD8+ T cell fate coupled by T-bet and 55. Gauthier J, Hirayama AV, Hay KA, Li D, Lymp J, Sheih A, et al. Comparison of eomesodermin. Nat Immunol. 2005;6(12):1236–44. https://doi.org/10.1038/ni1268. efficacy and toxicity of CD19-specific chimeric antigen receptor T-cells alone or in combination with Ibrutinib for relapsed and/or refractory CLL. Blood. 37. Kao C, Oestreich KJ, Paley MA, Crawford A, Angelosanto JM, Ali MA, et al. 2018;132(Suppl 1):299–9. https://doi.org/10.1182/blood-2018-99-111061. Transcription factor T-bet represses expression of the inhibitory receptor PD-1 and sustains virus-specific CD8+ T cell responses during chronic infection. Nat Immunol. 2011;12(7):663–71. https://doi.org/10.1038/ni.2046. Publisher’s Note 38. Paley MA, Kroy DC, Odorizzi PM, Johnnidis JB, Dolfi DV, Barnett BE, et al. Springer Nature remains neutral with regard to jurisdictional claims in Progenitor and terminal subsets of CD8+ T cells cooperate to contain published maps and institutional affiliations. chronic viral infection. Science. 2012;338(6111):1220–5. https://doi.org/1 0.1126/science.1229620. 39. Pauken KE, Wherry EJ. Overcoming T cell exhaustion in infection and cancer. Trends Immunol. 2015;36(4):265–76. https://doi.org/10.1016/j.it.2015. 02.008. 40. Llaó-Cid L, Roessner PM, Chapaprieta V, Öztürk S, Roider T, Bordas M, et al. EOMES is essential for antitumor activity of CD8 + T cells in chronic lymphocytic leukemia. Leukemia. 2021;17:1–11. 41. Thommen DS, Schumacher TN. T Cell Dysfunction in Cancer. Cancer Cell. 2018;33(4):547–62. 42. Best JA, Blair DA, Knell J, Yang E, Mayya V, Doedens A, et al. Transcriptional insights into the CD8(+) T cell response to infection and memory T cell formation. Nat Immunol. 2013;14(4):404–12. https://doi.org/10.1038/ni.2536. 43. Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P, et al. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature. 2011;472(7344):481–5. https://doi.org/10.1038/na ture09907. 44. Schietinger A, Philip M, Krisnawan VE, Chiu EY, Delrow JJ, Basom RS, et al. Tumor-Specific T Cell Dysfunction Is a Dynamic Antigen-Driven