Abstract: Background: CD20 expression is a controversial issue regarding response prediction to anti-CD20 therapy in chronic lymphocytic leukemia (CLL). Methods: Median fluorescence intensities (MFIs) of standard fluorescence beads from the daily calibration of flow cytometers according to EuroFlow protocols were used to establish a normalization approach to study CD20 expression on CLL cells. CD20 MFI was retrospectively assessed prior to and during treatment from flow cytometric measurements of peripheral blood in patients with different depths of molecular response in the four phase-II CLL2-BXX trials (BIG; BAG; BIO; BCG; N = 194) administering either Obinutuzumab or Ofatumumab in combination with targeted agents. Results: No significant difference was observed between the normalized and measured MFIs of CD19 and CD20 on CLL cells. During treatment, CD20 expression levels on CLL cells did not significantly differ between the four investigated different treatment schemes, but a strong molecular response to Ofatumumab seemed to correlate with higher CD20 expression prior to therapy. Conclusions: Standardized staining and instrument monitoring enable a robust assessment of longitudinal biological variations of marker expression based on MFI values. Obinutuzumab showed a higher proportion of patients with a strong MRD response independent from initial CD20 expression, whereas high pre-therapeutic CD20 expression levels seem to correlate with a profound response to Ofatumumab.

Abstract: Background: High throughput sequencing of immunoglobuline heavy chain gene rearrangements (IGH) recently demonstrated that B-cell precursor ALL (BCP-ALL) is a highly oligoclonal disease in children (Theunissen, Hematologica, 2017), however, little is known about the degree of oligoclonality in adults with BCP-ALL. Furthermore, no data exist on potential changes of the subclonal composition during therapy. Therefore, we aimed to monitor the clonal architecture of the leukemic blasts before and after Rituximab containing induction treatment in the context of the current German Multicentric Acute Lymphoblastic Leukemia (GMALL) 08/2013 trial. Materials & Methods: We identified and studied complete and incomplete immunoglobulin (IG) heavy chain rearrangements in 100ng of bone marrow (BM) DNA at diagnosis (dx) and 500ng BM DNA after Rituximab-containing induction I in 19 patients with BCP-ALL. We employed IGH-VJ FR1 and IGH-DJ NGS assays developed within the EuroClonality-NGS Consortium (www.euroclonalityngs.org), performing next generation sequencing on Illumina MiSeq. We analysed data with the ARResT/Interrogate bioinformatics platform (Bystry, Bioinformatics, 2017), which is also able to isolate and use the DN-J stem of V/DJ junctions to link clonally related rearrangements. Clones with abundance ≥5 %, and all subclones carrying the same DN-J stem as the dominant clones were considered as leukemia-associated and studied further. Employing the sequence of the DN-J stem, we could also link related complete and incomplete rearrangements, even though those are amplified in two separate PCRs. Results: At dx, 18/19 patients carried at least 1 IGH rearrangement with abundance ≥ 5%, 10 patients carried 2 or more. In all 18 patients with dominant markers detected at dx, subclones with abundance 〈 5% carrying the same DN-J stem as the dominant clone were present, indicating oligoclonality and clonal evolution. On average, 53.1 (range 0-295) (sub)clones per DN-J stem were detected at dx, and 32.7 (range 0-238) after treatment. Next, we compared the kinetics of all (sub)clones with abundance ≥1% in at least one of the time-points. Of the 18 patients, 6 only had subclones with abundance 〈 1%, and we did not investigate the kinetics of such subclones. In another 11 patients (Fig. 1, Pt. 1-11), all sub(clones) had the same kinetics, with no clone gaining predominance over time. In 1 patient (Fig. 1, Pt. 12), 3 (sub)clones which were present at dx disappeared and a new subclone appeared after treatment. This patient had a pro-B immunophenotype, where oligoclonality and clonal instability are well known phenomena (Szczepanski, Leukemia, 2001). Conclusions: It has recently been shown that ALL is a highly oligoclonal disease in children (Theunissen, Hematologica, 2017), and our study extends this finding to adults with BCP-ALL. We furthermore demonstrate that the subclonal composition remains stable in the majority of patient during induction chemoimmunotherapy. The fact that the response to treatment is generally consistent among different (sub)clones has important implications for MRD quantification as it reassures the usage of the dominant clonal IG gene rearrangement for MRD monitoring in ALL. However, also significant changes of the clonal composition may occur in BCP-ALL as exemplified in one patient of our cohort. Further investigations are necessary to elucidate factors that influence subclonal heterogeneity in response to treatment. Disclosures Viardot: Pfizer: Consultancy, Honoraria; Gilead Kite: Consultancy, Honoraria; Roche: Consultancy, Honoraria; Amgen: Consultancy; BMS: Consultancy, Honoraria. Kneba:AbbVie: Consultancy, Honoraria; Roche: Consultancy, Honoraria. Goekbuget:Pfizer: Consultancy, Other: Travel support, Research Funding; Novartis: Consultancy, Other: Travel support, Research Funding; Kite / Gilead: Consultancy; Amgen: Consultancy, Other: Travel support, Research Funding; Celgene: Consultancy. Brüggemann:Incyte: Consultancy; PRMA: Consultancy; Regeneron: Research Funding; Affimed: Research Funding; Pfizer: Speakers Bureau; Amgen: Consultancy, Research Funding, Speakers Bureau; Roche: Speakers Bureau.

Abstract: Introduction: Treatment with Rituximab has improved the outcome for patients with B-cell precursor ALL (B-ALL) and is conventionally restricted to patients with CD20 expression on leukemic blasts above a level of 20%. However, it is not known whether this arbitrary cutoff is biologically meaningful, or if also patients with lower CD20 expression levels profit from Rituximab treatment. In addition, CD20 expression levels might differ between different compartments and can change during early treatment. Therefore, in the current German Multicentric Acute Lymphoblastic Leukemia (GMALL) 08/2013 trial Rituximab is applied to all BCR-ABL1-negative B-ALL patients during induction and early consolidation irrespective of the initial blast CD20 expression. In this study, CD20 expression on blasts at diagnosis and after a 5-day dexamethasone prephase is thoroughly characterized and correlated to MRD response after induction I including one dose of Rituximab to evaluate its impact on response to Rituximab. Methods: Comparative quantification of CD20 expression levels was performed at diagnosis (blood and bone marrow (BM)) and at the end of the prephase (blood) and correlated to minimal residual disease (MRD) response after early Rituximab treatment. Samples were subjected to the EuroFlow standardized stain, lyse and wash and instrument setting procedures (van Dongen et al., 2012) and stained with a 3-tube, 8-color panel, containing the markers CD45, CD20, CD10, CD66c, CD3, CD19, CD71, CD9, CD13+33, CD34, CD22, CD11a, and CD38. CD20 median fluorescence intensities (CD20-MFI) as well as percentages of CD20+ B-ALL blasts among all blasts (%CD20+) were measured. MRD was quantified at day 22 after induction I using real-time quantitative PCR of clonal immune gene rearrangements. MRD response after induction was defined as MRD decrease to a level below 10-4, MRD persistence as detection of quantifiable MRD ≥10-4. Results: FCM results of a total of 166 samples of 84 B-ALL patients were evaluated. CD20 expression of circulating blasts significantly increased after a 5-day prephase in common/pre-B-ALL (c/pre-B) but not in pro-B-ALL in a cohort of 36 paired peripheral blood samples (Fig. 1 A-B, n=8 pro-B-ALL, paired t-test of CD20-MFI and %CD20: p=.1016 and p=.1660, Fig. 2 A-B, n=28 c/pre-B-ALL, p=.0029 and p=0.0060). Interestingly, the increase of CD20 expression on B-ALL blasts contrasted that on mature normal B-cells, where a decrease occurred in pro-B-ALL and c/pre-B-ALL samples at day 6 (paired t-test: p=.0485 and p 〈 .0001, respectively) (Fig.1A, 2A). In 24 paired blood/BM pre-treatment samples the percentage of CD20+ blasts was significantly higher in blood in c/pre-B-ALL (n=19: paired t-test: p=.0009), but not in pro-B-ALL (n=5: paired t-test: p=.095) (Fig. 1C, 2C). The significance levels were not consistently reached for the comparison of CD20-MFI. Nevertheless, we show that in 3/24 (12.5%) patients the percentage of CD20+ B-ALL blasts in BM did not reach the arbitrary cutoff of 20% whereas the matching diagnostic blood did. Finally, CD20 expression levels from 25 BM and 30 pre-treatment blood samples and 38 blood samples after dexamethasone prephase from 54 patients were assessed (in total 93 samples). In further calculations we correlated MRD response values after one dose of Rituximab with the CD20 expression levels on blasts. In detail, the highest measured percentage value of %CD20 per patient in the group of MRD responders was compared to the highest measured %CD20 per patient in the group of MRD persisters. CD20 expression levels were significantly lower in 33 MRD persisters compared to 21 MRD responders (Fig 3, p=.0336). The difference was mainly attributable to a high frequency of MRD persisters in pro-B-ALL compared to c/pre-B-ALL (MRD persistence in pro-B-ALL 8/9 (89%) compared to 25/45 (56%) in c/pre-B-ALL). Conclusion: We measured significant differences in CD20 expression levels on leukemic blasts (i) between pre-treatment blood and BM and (ii) between pre-treatment blood and blood after prephase in patients with c/pre-B-ALL challenging the conventional eligibility criteria for the CD20 targeted treatment. MRD responders tended to have higher CD20 expression levels compared to MRD persisters. Extension of the patient cohort is ongoing to confirm this observation. Furthermore, we will analyze the impact of 3 rituximab applications after induction phase II. Disclosures Ritgen: abbvie: Research Funding; Roche: Honoraria, Research Funding. Viardot:Gilead Kite: Consultancy, Honoraria; Roche: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Amgen: Consultancy. Kneba:Roche: Consultancy, Honoraria; AbbVie: Consultancy, Honoraria. Goekbuget:Kite / Gilead: Consultancy; Celgene: Consultancy; Novartis: Consultancy, Other: Travel support, Research Funding; Pfizer: Consultancy, Other: Travel support, Research Funding; Amgen: Consultancy, Other: Travel support, Research Funding. Brüggemann:Incyte: Consultancy; Amgen: Consultancy, Research Funding, Speakers Bureau; PRMA: Consultancy; Affimed: Research Funding; Regeneron: Research Funding; Pfizer: Speakers Bureau; Roche: Speakers Bureau.

Abstract: Introduction Rituximab (R) administration results in significant outcome improvement in B cell precursor acute lymphoblastic leukemia (B-ALL) patients (pts), but is usually restricted to pts with ≥20% CD20+ leukemic blasts. Yet, this arbitrary cut-off is not proven biologically sensible. Moreover, CD20 expression might differ between blood (pb) and bone marrow (bm) and varies under prednisone during early treatment. In the present GMALL08/2013 trial R is administered to all BCR-ABL1-negative B-ALL pts irrespective of the initial leukemic CD20 expression. We assessed the initial and post-prephase CD20 expression in GMALL08/2013 pts and correlated the values with MRD response after Induction I (Ind I) and Consolidation I (Cons I). A historical B-ALL GMALL07/2003 cohort without R treatment and with available CD20 expression and MRD data was used to evaluate for R-unrelated effects. Methods Comparative immunophenotypic quantification of CD20 expression in 207 B-ALL pts was performed at diagnosis (pb d0 and/or bm d0) and/or (a/o) after a 5-day dexamethasone- and cyclophosphamide-containing prephase (pb d6) under EuroFlow standardized procedures. CD20 median fluorescence intensities (CD20-MFI) and percentages of CD20+ B-ALL blasts/all blasts (%CD20+ BL) were assessed. Minimal residual disease (MRD) was determined after Ind I (after 1x R) and Cons I (after 4x R) by quantitative PCR for clone-specific immune gene rearrangements to stratify pts as molecular complete response (MolCR, MRD negativity, assay sensitivity at least 1x10 -4), molecular intermediate response (MolIR, MRD positive non-quantifiable or positive & lt;1x10 -4) and molecular failure (MolFAIL, MRD ≥1x10 -4). Results bm d0/pb d0. In 91 paired bm d0/pb d0 samples %CD20+ BL as well as the CD20-MFI were significantly higher in pb in common/pre-B ALL (c/pre-B ALL) (n=76: paired t-test: p & lt;.0001 and p & lt;.0001) and in normal mature B cells (CD20-MFI paired t-test; pro-B/ c/pre-B p=0.026/p & lt;.0001), but not in pro-B ALL (n=15: paired t-test: p=.81 and p=.76) (Fig. 1A, 2A). Notably, in 6/76 (7.9%) c/pre-B ALL pts the %CD20+ BL in bm d0 was lower and in corresponding pb d0 higher than the arbitrary cut-off of 20%. pb d0/pb d6. CD20 expression of circulating blasts significantly increased after a 5-day prephase in c/pre-B ALL but not in pro-B ALL in 106 paired pb d0/pb d6 samples (paired t-test of CD20-MFI and %CD20+ BL; n=20 pro-B ALL, p=.09 and p=.25; n=86 c/pre-B ALL, p & lt;.0001 and p & lt;.0001). Normal mature B cells presented with the opposite effect (CD20-MFI paired t-test; pro-B/ c/pre-B: p=0.005/p & lt;.0001) (Fig. 1B, 2B). Notably, the %CD20+ BL in pb d0 was lower and in corresponding pb d6 higher than the arbitrary cut-off of 20% in 2/20 (10.0%) pro-B ALL and 12/86 (13.9%) c/pre-B ALL pts. Molecular response under R. The values of %CD20+ BL were correlated with MRD response after Ind I and Cons I (Fig. 3). Since the CD20 expression in the present cohort was shown to be significantly modulated in a drug- and compartment-dependent manner, we used the highest measured value of %CD20+ BL out of the three available values per patient (bm d0, pb d0 a/o pb d6). In the historical cohort (n=145) one value per patient for %CD20+ BL was available. Due to low CD20 expression pro-B ALL did not show any differences in the %CD20+ BL among the risk groups in the present (Ind I n=23, Cons I n=20) and the historical (Ind I n=11; Cons I n=11) cohort. The differences in %CD20+ BL in relation to molecular response were significant in c/pre-B ALL between MolCR and MolFAIL after Ind I (n=127) and Cons I (n=120) (Mann-Whitney test: p=.0002 and p=.0028) and of lower significance between MolIR and MolFAIL after Ind I (p=.013) and between MolCR and MolIR after Cons I (p=.029) in the present cohort. Within the historical cohort (Ind I n=145, Cons I n=143) no significant differences were observed. Conclusions Leukemic CD20 expression was modulated between compartments (bm d0/pb d0) and showed a significant increase in a drug-dependent manner in c/pre-B ALL (pb d0/pb d6) probably in response to dexamethasone. The results might challenge the conventional eligibility criteria for CD20 targeted treatment in c/pre-B ALL. MRD persisters showed lower initial CD20 expression compared to MRD responders in the present cohort consistently receiving R, but not in the historical cohort without R treatment. Accordingly, R seems to improve the early MRD response predominantly in pts with higher CD20 expression. Supported by DJCLS Figure 1 Figure 1. Disclosures Szczepanowski: Amgen: Speakers Bureau. Trautmann: Amgen: Speakers Bureau. Ritgen: Roche: Consultancy, Other: Travel support, Research Funding; Abbvie: Consultancy, Other: Travel support, Research Funding; Chugai: Consultancy; MSD: Consultancy, Other: Travel support; Celgene: Other: Travel support. Nachtkamp: Celgene: Other: Travel Support; bsh medical: Speakers Bureau; Jazz: Speakers Bureau. Viardot: Kite/Gilead: Honoraria, Membership on an entity's Board of Directors or advisory committees; Amgen: Membership on an entity's Board of Directors or advisory committees; Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees; F. Hoffmann-La Roche Ltd: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; University Hospital of Ulm: Current Employment. Baldus: Jazz: Honoraria; Celgene/BMS: Honoraria; Amgen: Honoraria; Novartis: Honoraria. Schwartz: Morphosys: Research Funding; Gilead: Other: Travel grants, Speakers Bureau; Pfizer: Honoraria, Speakers Bureau; Amgen: Membership on an entity's Board of Directors or advisory committees, Other: Travel grants, Speakers Bureau; Novartis: Speakers Bureau; Jazz Pharmaceuticals: Other: Travel grants, Speakers Bureau; BTG International Inc: Membership on an entity's Board of Directors or advisory committees; MSD Sharp & Dohme: Membership on an entity's Board of Directors or advisory committees; Basilea: Other: Travel grants. Goekbuget: Pfizer: Consultancy, Other: Research funding for institution; Amgen: Consultancy, Other: Invited talks for company sponsored symposia (with honoraria); Research funding for institution; Astra Zeneca: Other: Invited talk for company sponsored symposia (with honor); Gilead/Kite: Consultancy; Novartis: Consultancy, Other: Research funding for Institution; Jazz Pharmaceuticals: Other: Research funding for institution; Incyte: Other: Research funding for Institution; Cellestia: Consultancy; Erytech: Consultancy; Morphosys: Consultancy; Servier: Consultancy, Other; Abbvie: Other. Brüggemann: Amgen: Other: Advisory Board, Travel support, Research Funding, Speakers Bureau; Incyte: Other: Advisory Board; Janssen: Speakers Bureau. OffLabel Disclosure: Rituximab administration to patients with CD20-negative ( & lt;20% CD20+ blasts/all blasts) BCR-ABL-negative B-ALL.

Abstract: Detection of patient- and tumor-specific clonally rearranged immune receptor genes using real-time quantitative (RQ)-PCR is an accepted method in the field of precision medicine for hematologic malignancies. As individual primers are needed for each patient and leukemic clone, establishing performance specifications for the method faces unique challenges. Results for series of diagnostic assays for CLL and ALL patients demonstrate that the analytic performance of the method is not dependent on patients’ disease characteristics. The calibration range is linear between 10 -1 and 10 -5 for 90% of all assays. The detection limit of the current standardized approach is between 1.8 and 4.8 cells among 100,000 leukocytes. RQ-PCR has about 90% overall agreement to flow cytometry and next generation sequencing as orthogonal methods. Accuracy and precision across different labs, and above and below the clinically applied cutoffs for minimal/measurable residual disease (MRD) demonstrate the robustness of the technique. The here reported comprehensive, IVD-guided analytical validation provides evidence that the personalized diagnostic methodology generates robust, reproducible and specific MRD data when standardized protocols for data generation and evaluation are used. Our approach may also serve as a guiding example of how to accomplish analytical validation of personalized in-house diagnostics under the European IVD Regulation.

Abstract: The standardized EuroFlow protocol, including CD19 as primary B‐cell marker, enables highly sensitive and reliable minimal residual disease (MRD) assessment in B‐cell precursor acute lymphoblastic leukaemia (BCP‐ALL) patients treated with chemotherapy. We developed and validated an alternative gating strategy allowing reliable MRD analysis in BCP‐ALL patients treated with CD19‐targeting therapies. Concordant data were obtained in 92% of targeted therapy patients who remained CD19‐positive, whereas this was 81% in patients that became (partially) CD19‐negative. Nevertheless, in both groups median MRD values showed excellent correlation with the original MRD data, indicating that, despite higher interlaboratory variation, the overall MRD analysis was correct.

Abstract: Introduction High-dose chemotherapy and stem cell transplant (SCT) remains a standard of care in medically fit patients (pts) with newly diagnosed (ND) MM. Induction triplets with at least one of the newer compounds are recommended. Bortezomib (V), lenalidomide (R) and dexamethasone (D; VRD) ranks among the most effective regimens and VRD/SCT was superior to VRD alone in an RCT. In a phase 2 study, we demonstrated RAD induction (lenalidomide 25 mg d1-21; Adriamycin 9 mg/m2 iv d1-4; dexamethasone 40 mg d 1-4 and 17-20 every 4 weeks) followed by SCT to be safe and effective (Knop et al., Leukemia 2017). Therefore, we decided to compare RAD versus (vs) VRD (lenalidomide 25 mg, d1-14; subcutaneous bortezomib 1.3 mg/m2 d 1, 4, 8, 11; dexamethasone 20 mg d 1+2, 4+5, 8+9, 11+12 every 3 weeks) induction (3 cycles each) in an RCT. MethodsThe current study was set up according to a double 2x2-factorial design to enrol transplant-eligible pts up to 65 years. The post-induction (PI) complete response (CR) rate as per IMWG criteria was the efficacy co-primary endpoint. We hypothesized the CR rate following RAD to be non-inferior to VRD which was estimated to be 20%. The study was powered to confirm non-inferiority of RAD at a margin of 10% with a one-sided alpha level of .05. Cytogenetic characterization was performed by fluorescence in situ hybridization (FISH) from CD138-enriched plasma cells. Minimal residual disease (MRD) was assessed by second-generation eight-color flow cytometry (FC; EuroFlow protocol). Bone marrow (BM) samples from baseline and defined restaging time points were analyzed for an acquisition of ⩾107cells/sample. In a subgroup of 103 pts, we evaluated the applicability of comprehensive immunoglobulin (Ig) amplicon next generation sequencing (NGS) to detect molecular MRD markers and to compare the results with FC. NGS-based marker screening was performed in baseline BM. Sequencing libraries were prepared using 2-step PCR employing multiplex primer sets for IGH V-D-J (FR1, FR2 and FR3), IGH D-J and IGK loci (V-J and KDe). For MRD detection, we used 1-step library preparation with the same primer sets. Results476 pts with a median age of 55 (range, 32-65) years were randomized between 05/2012 and 06/2016 and 469 received at least one dose of study drug. High-risk (HR) FISH abnormalities comprised del17p (11.3% of pts); t(4;14) (11.7%); and t(14;16) (4.5%). 232 pts were randomized to receive RAD, and 237 to VRD, respectively. 90.5% of RAD vs 93.7% of VRD pts completed all 3 cycles. PI CR rate was 13.5% (95% CI, 9.4%-18.7%) with RAD vs 13.4% (95% CI, 9.3-18.5) with VRD, (P=.971). Rates of ≥VGPR were 40.6% (50% CI, 34.2%-47.3%) with RAD vs 48.9% (95% CI, 42.3-55.6%) with VRD (P=.076). In pts with HR cytogenetics, rates of ≥VGPR were 43.3% (RAD) vs. 59.3% (VRD; P=.096). From 317 pts with paired samples, 33/151 (21.9%) of RAD vs 45/166 (27.1%) of VRD pts were FC MRD negative (P=.169) following induction at a median sensitivity level of 6.73x10-6. 197/239 positive pts (82.4%%) had MRD levels above 0.01%, and 42 (17.6%), between 0.0001 and 0.01%. Flow MRD negativity as per IMWG MRD criteria (Kumar et al, Lancet Oncol 2016) was seen in 8/151 (5.2%) pts with RAD vs 6/166 (3.6%) with VRD (P=.27). The remainder of pts did not (yet) fulfil IMWG CR for various reasons. NGS marker screening identified at least 1 Ig marker in 98/103 evaluable patients. To date, 47/98 pts were analyzed for NGS MRD following induction. Four out of 47 (8.5%) subjects were sequencing negative (3/4 post-VRD) with all of them also being IMWG FC MRD negative. One VRD patient died during induction for a mortality rate of 0.2 %. 62.1% of RAD vs 55.3% of VRD pts experienced at least one serious adverse event (SAE; p=.16). SAEs with relationship to study drugs of at least °3 severity occurred in 26.3% (RAD) vs 23.6% (VRD) pts (p=.523). ConclusionsTo the best of our knowledge, this is the first RCT to compare two R-based triplets in SCT-eligible pts. The co-primary efficacy endpoint was met with identical PI CR rates of around 13% for RAD and VRD, respectively. However, a trend emerges to favor VRD over RAD in terms of at least VGPR including HR FISH subjects. Analysis of MRD by multicolor FC showed 5% of pts to be already IMWG flow MRD-negative. Results for all 98 pts evaluable for NGS MRD will be presented. As of yet, too few progression events have occurred to estimate the second co-primary endpoint, 3-year progression-free survival. Longitudinal response and MRD analysis are ongoing. Disclosures Langer: Celgene: Consultancy. Mügge:Celgene: Honoraria, Research Funding; Janssen: Honoraria; Novartis: Honoraria; Bristol Myers Squibb: Honoraria; Amgen: Honoraria. Blau:Amgen: Other: Advisory board; BMS: Other: Advisory board; Novartis: Other: Advisory boards; Takeda: Other: Advisory board; Janssen: Other: Advisory board, Research Funding; Celgene: Other: Advisory board, Research Funding. Rollig:Janssen: Research Funding; Bayer: Research Funding. Dechow:AMGEN: Consultancy; Celgene: Honoraria. Gramatzki:Affimed: Research Funding. Brümmendorf:Merck: Consultancy; Janssen: Consultancy; Takeda: Consultancy; Pfizer: Consultancy, Research Funding; Novartis: Consultancy, Research Funding. Schmidt:Gilead: Honoraria, Other: Travel Grants; Takeda: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Travel Grants; Celgene: Honoraria. Knop:Bristol Myers Squibb: Consultancy, Honoraria, Research Funding; Janssen: Consultancy, Honoraria, Research Funding; Takeda: Consultancy, Honoraria, Research Funding; Celgene: Consultancy, Honoraria, Research Funding; Amgen: Consultancy, Honoraria, Research Funding.

Abstract: The malignant transformation leading to a maturation arrest in B-cell precursor acute lymphoblastic leukemia (BCP-ALL) occurs early in B-cell development, in a pro-B or pre-B cell, when somatic recombination of variable (V), diversity (D), and joining (J) segment immunoglobulin (IG) genes and the B-cell rescue mechanism of V H replacement might be ongoing or fully active, driving clonal evolution. In this study of newly diagnosed BCP-ALL, we sought to understand the mechanistic details of oligoclonal composition of the leukemia at diagnosis, clonal evolution during follow-up, and clonal distribution in different hematopoietic compartments. Methods Utilizing high-throughput sequencing assays and bespoke bioinformatics we identified BCP-ALL-derived clonally-related IGH sequences by their shared ‘DNJ-stem’. Results We introduce the concept of ‘marker DNJ-stem’ to cover the entirety of, even lowly abundant, clonally-related family members. In a cohort of 280 adult patients with BCP-ALL, IGH clonal evolution at diagnosis was identified in one-third of patients. The phenomenon was linked to contemporaneous recombinant and editing activity driven by aberrant ongoing D H /V H -DJ H recombination and V H replacement, and we share insights and examples for both. Furthermore, in a subset of 167 patients with molecular subtype allocation, high prevalence and high degree of clonal evolution driven by ongoing D H /V H -DJ H recombination were associated with the presence of KMT2A gene rearrangements, while V H replacements occurred more frequently in Ph-like and DUX4 BCP-ALL. Analysis of 46 matched diagnostic bone marrow and peripheral blood samples showed a comparable clonal and clonotypic distribution in both hematopoietic compartments, but the clonotypic composition markedly changed in longitudinal follow-up analysis in select cases. Thus, finally, we present cases where the specific dynamics of clonal evolution have implications for both the initial marker identification and the MRD monitoring in follow-up samples. Discussion Consequently, we suggest to follow the marker DNJ-stem (capturing all family members) rather than specific clonotypes as the MRD target, as well as to follow both VDJ H and DJ H family members since their respective kinetics are not always parallel. Our study further highlights the intricacy, importance, and present and future challenges of IGH clonal evolution in BCP-ALL.