Importance of Minimal Residual Disease in the Era of Targeted Therapies in Chronic Lymphocytic Leukemia

Background: The level of minimal residual disease (MRD), often binarized into detectable or undetectable MRD according to certain thresholds, is strongly associated with long-term outcomes after chemo- and chemoimmunotherapy. Summary: Driven by our improved understanding of the biology of chronic lymphocytic leukemia (CLL), the recent decade has shown a shift from chemotherapy-based regimens to regimens based on targeted agents that exploit distinct biological vulnerabilities of CLL. These targeted agents can be broadly classified into inhibitors of Bruton tyrosine kinase (BTK) and BCL2 as well as CD20-directed antibodies. Depending on which agent and which combination of agents is used, the levels or status of MRD can have varying clinical relevance. This has implications on the prognosis after therapy as well as on possible strategies to guide treatment duration and intensity. Key Messages: This review summarizes the main discoveries related to MRD in the context of targeted therapies. Furthermore, it provides an overview on current hurdles and caveats related to the implementation of MRD in regular clinical care and summarize open research questions that need to be addressed with future clinical studies.


Introduction
Assessing the depth and quality of remission during or after cancer therapy is a key aspect of medical oncology.Historically, this was mostly based on basic laboratory or imaging methods that allowed to estimate how much of disease was still detectable in terms of lymph node size or blood count, compared with the patient's status before therapy.While this allows for an estimate of how complete a remission might be, the association with long-term clinical outcomes was not always strong.This was partly due to the arbitrary nature of thresholds defining a complete response (CR) or partial response (PR) [1,2].In hematology, one of the first methods to quantify the depth of remission during therapy was introduced in chronic myeloid leukemia in the context of targeted tyrosine kinase inhibition with imatinib.By quantifying the amount of BCR-ABL transcripts, it became possible to identify patients with the deepest remissions, who would eventually be able to cease therapy for a certain time while maintaining the remissions [3,4].
In chronic lymphocytic leukemia (CLL), the first approach to quantify minimal residual disease (MRD) was based on the distinct immunophenotype of circulating CLL cells.This feature could be leveraged by applying multi-color flow cytometry with antibody panels that captured cluster of differentiation (CD) markers like CD19 and CD5 on the surface of CLL cells and by that provide an estimate on the number of circulating malignant lymphocytes.Using this method, several studies showed in the context of chemo-and chemoimmunotherapy that patients with MRD levels below 10 −4 (i.e., detecting one CLL cell per 10,000 leukocytes) had similarly good outcomes regardless of whether they had attained a PR or CR by conventional iwCLL criteria.It also became evident that adding a CD20 antibody to chemotherapy reduced the levels of MRD more effectively than chemotherapy alone.Increased remission depth was also associated with longer time until disease recurrence, thereby suggested that the end of treatment MRD status is a particularly useful surrogate to indicate quality and durability of remission [5,6].
Over the recent decade, our understanding of the biology and pathophysiology of CLL has substantially improved, which allowed for new targeted agents that leverage distinct vulnerabilities of CLL.Inhibitors of B-cell receptor signaling, in particular, Bruton tyrosine kinase (BTK) inhibitors, have demonstrated substantial efficacy in patients with CLL [7].Likewise, BH3 mimetics such venetoclax have shown high single-agent activity that was associated with substantial reduction of MRD levels [8].Given the distinct mechanisms of action, the prognostic impact of MRD had to be revisited in the context of targeted therapies: While several studies demonstrated that MRD levels did not correlate with clinical outcomes in the context of continuous BTK inhibition, treatment with BCL2 inhibitors lead to unprecedented deep remissions that were strongly linked with PFS and even OS [9].With continuous BTK inhibition, a deepening of remissions is observed with longterm intake, as observed with an increase of MRD and clinical responses [10].However, the depth of remission does not significantly correlate with survival, as observed in the ECOG1912 study, demonstrating the distinct mechanism of action of continuous BTK inhibition that does not require MRD reduction to achieve long-term disease control [11].
The iwCLL guidelines specifically recommend the measurement of MRD in clinical studies that aimed to maximize the depth of remissions and suggest the cut-off of 10 −4 to report the fraction of patients with undetectable MRD (uMRD) based on the intention-to-treat population [12].The guidelines so far do not recommend the use of MRD testing in routine clinical care, although the latest iteration of the guideline is from 2018 and substantial advances in our understanding of MRD have been made since.
This review aimed to summarize the current evidence on MRD in the context of targeted treatment of CLL.First, currently available MRD detection methods are explained, followed by a recapitulation of MRD data in the context of chemo-and chemoimmunotherapy.Subsequently, data from various prospective and retrospective studies of targeted agents are summarized, and an outlook on future challenges and strategies is provided.

MRD Detection Methods
Multi-parameter flow cytometry is currently the most widely used and standardized technique for MRD measurement in CLL (FLC).This is largely due to the European Research Initiative on CLL (ERIC), which established and is still updating consensus on material preparation, antibody combination, and analysis at a level of 10 −4 .Rawstron and colleagues led the development of the ERIC protocol, which suggests using an antibody panel targeting CD5, CD19, CD20, CD43, CD79b, and CD81 to detect CLL cells [13][14][15].The iwCLL guidelines from 2018 defined the threshold for uMRD at 10 −4 [12].It should be noted that this cut-off is to some extent arbitrary and motivated by the technical measurement limitations, rather than by validated cut-offs.Current commercial 6-color FLC can achieve sensitivity levels up to 10 −5 .It is conceivable that the standard threshold for uMRD will be lowered once more sensitive methods are created.
Polymerase chain reaction (PCR) of the immunoglobulin heavy chain variable gene is another method to detect MRD in CLL [16].Previously, this method was established as a semi-quantitative tool that allowed detection of MRD with a limited sensitivity of 10 −4 , while being technically straightforward as it uses consensus primers annealing to CLL clone-specific complementary determining region 3 (CDR3) [17].Quantitative results from allele-specific oligonucleotide PCR of the IGHV gene can be obtained by using real-time quantitative PCR (RT-qPCR).This "nested" approach provides for a high sensitivity up to 10 −6 but does not provide quantitative results [17].By using RT-qPCR, quantitative results from Importance of MRD in the Era of Targeted Therapies in CLL allele-specific oligonucleotide PCR of the IGHV gene can be derived.Similar to nested PCR, this requires CDR3 sequencing and RT-qPCR has a reduced sensitivity compared at approximately 10 −5 [18,19].This approach is therefore not widely used in routine clinical care, but due to its standardized approach it is frequently applied in clinical trials [20].
Next-generation sequencing (NGS) of rearranged IgH VDJ or DJ, IgK, and IgL receptor gene sequences, or translocated BCL1/IgH(J) and BCL2/IgH(J) receptor gene sequences is another approach for evaluating MRD based on DNA [21].Patient-specific primers are not necessary for this approach.However, to determine the disease-specific IgH sequence, a blood sample from before treatment is required.This approach also allows quasi-simultaneous determination of the IGHV mutational status.Acute lymphoblastic leukemia has been treated for some time using NGS-based MRD measurements, and targeted agents like blinatumomab have been approved for patients with MRD positivity after induction [22].Depending on a sufficient amount of isolated genomic DNA, sensitivity level of 10 −6 can be achieved with NGS-based MRD assessments.In practice, sufficient blood volume needs to be drawn to ensure high sensitivity levels; in the CLL14 study, where large scale longitudinal NGS-based MRD assessments were implemented, sufficient DNA amounts were available for >90% of patients, demonstrating the general feasibility of this approach [23].However, direct and indirect costs, e.g., due to more elaborate analysis pipelines, can be considerably higher with NGS methods compared with flow cytometry [24].[25] The most recent iwCLL guidelines revision takes into consideration the growing significance of MRD assessments, but does not usually suggest using MRD in regular clinical care [2].The guidelines do, however, recommend that MRD assessments should be considered in clinical studies.uMRD rates should be based on the intention-to-treat population, rather than, e.g., bestresponse-rates, in order to guarantee uniform presentation of study findings.Together, a number of triedand-true techniques have been created thus far to identify MRD using CLL-specific characteristics.Given the accessibility of flow cytometry in hematology, it is likely that flow-based MRD quantification will remain widely used in everyday clinical practice, especially as multi-color flow cytometry is enabling higher sensitivity.Likewise, as NGS methods are also being widely used in diagnostic hematology laboratories to define risk groups, NGS-based MRD quantification also yields promising results.

MRD in the Context of Chemo-and Chemoimmunotherapy
In 1992, Robertson et al. [26] conducted the first analysis of MRD in patients with CLL who were treated with fludarabine and prednisolone.Based on lowsensitivity two-color flow cytometry, 89% of patients with CRs, 51% of patients with nodal CRs, and 19% of patients with PRs had peripheral blood (PB) MRD levels below 10 −1 .Even with this somewhat rudimentary "MRD" technique, patients with CR and a deeper response had a much longer PFS than those with detectable MRD and CR.This was a significant breakthrough in characterizing treatment response and it set the stage for later improvements in MRD assessment in CLL.A significant development was the addition of CD20 antibody to chemotherapy and chemotherapy combinations for the management of CLL.Although no uMRD was seen in the study cohort of 100 patients, a phase 2 study found that adding rituximab to chlorambucil over 6 cycles increased progression-free survival in elderly patients with or without coexisting conditions to 23.5 months [27].Similarly, in the CLL11 study of the German CLL study group, which randomized patients with previously untreated CLL and coexisting conditions to receive chlorambucil, chlorambucil-rituximab, or chlorambucilobinutuzumab, reported no PB uMRD in the chlorambucil monotherapy arm, 3% in the chlorambucilrituximab, and 38% in the chlorambucil-obinutuzumab arm [28,29].In contrast to patients with uMRD, who had not yet achieved median PFS by the time of last follow-up, patients with detectable MRD at end of treatment (EoT) demonstrated a median PFS of 19.4 months.In the study's final analysis, patients who received chlorambucil-obinutuzumab had significantly longer OS than those who received chlorambucil alone or chlorambucil alone with rituximab, demonstrating that overall high rates of uMRD were associated with improved longterm outcomes [30].
The addition of rituximab to fludarabine and cyclophosphamide (FCR) was first established in young and fit patients with CLL in a pivotal phase 2 study from MD Anderson [31].This study confirmed feasibility and high long-term efficacy of chemoimmunotherapy in this group of previously untreated patients [32].At the EoT, uMRD by light chain reaction in bone marrow was reported in 106 of 246 (43%) of FCR-treated patients.This was also demonstrated in the randomized CLL8 study of the GCLLSG, where PB uMRD rates by FLC PB were seen in 90 of 408 (22%) FCR-treated patients and 51 of 409 (12%) FC-treated patients, respectively [33,34].Of note, MRD assessments were only available for a subgroup of the original ITT population; based on available EoT samples in the FCR arm, 90 (63%) of 143 patients had uMRD <10 −4 in PB [5].More recently, the ECOG1912 study reported 59% and the UK FLAIR study 75% PB uMRD rates their FCR arms, respectively [35,36].The study thereby demonstrated that the addition of a CD20 antibody induced deeper remissions than chemotherapy alone.Kovacs et al. [6] demonstrated that uMRD correlates better with PFS and even OS than conventional clinical response assessment in an effort to pool available response data from CLL8 and CLL10.Patients treated with chemo-or chemoimmunotherapy that resulted in uMRD at EoT had the longest PFS and OS, regardless of whether a CR or PR was observed.This result supports the high prognostic value of MRD as a trustworthy surrogate for a treatment regime's effectiveness.
Whether reaching MRD levels deeper than 10 −4 entails clinically tangible advantages is currently not fully clear.In the context of chemoimmunotherapy, Thompson and colleagues demonstrated that patients reaching PB uMRD levels of ≤10 −6 have a significantly longer PFS than patients having uMRD <10 −4 [37].This suggests that more sensitive MRD assays might allow improved prognostic discrimination.With targeted combination therapies like venetoclax-obinutuzumab in the CLL14 study (see below), no significant difference between patients with PB uMRD <10 −6 versus 10 −4 have been observed so far [38].However, possible confounders, such as PFS events due to comorbidities and unrelated to CLL, can potentially dilute clear associations between MRD and PFS and thus need to be considered when analyzing such data [39].

MRD in the Context of BTK Inhibitor Monotherapy
Ibrutinib, the first-in-class BTK inhibitor, was initially studied as a continuous monotherapy.It targets a variety of kinases but primarily inhibits the disrupted BCR pathway, which is a crucial part of the pathogenesis of CLL [40].Less than 10% of patients treated with ibrutinib achieve uMRD levels at any stage during such continuous therapy, according to studies in patients with treatmentnaive as well as relapsed/refractory CLL [36,41].However, due to the distinct mechanism of action of BTK inhibitor maintenance therapy, which suppresses clonal growth, effective disease control can be achieved despite detectable MRD.Long-term follow-up of patients treated with ibrutinib showed that the clinical response rates, particularly CRs, increased over time.Although the rate of uMRD was lower than in the CIT setting, only about 6% of PB patients with CLL had uMRD after 4 years of ibrutinib treatment.Additionally, there was no discernible difference in PFS between patients with and without uMRD during ibrutinib treatment [42].The ECOG1912 study randomized fit, previously untreated patients with CLL to either receiving FCR or ibrutinib and made the difference between the two paradigms very obvious: uMRD rates of 57% were observed after 6 cycles of FCR as opposed to uMRD rates of only 5% after 1 year on ibrutinib [11].Nevertheless, patients in the continuous ibrutinib monotherapy arm, with persisting detectable residual disease in virtually all patients, had a significantly longer PFS and even OS than patients in the FCR arm, despite the higher rate of uMRD after CIT.This finding suggests that the disease-modulating effects of continued ibrutinib therapy work independently of MRD depletion.Similar findings were reported for the next-generation BTK inhibitor acalabrutinib in the ELEVATE-TN study, in which elderly unfit patients with previously untreated CLL were given either chlorambucil-obinutuzumab, acalabrutinib-obinutuzumab, or acalabrutinib monotherapy [43,44].Although limited MRD assessment were carried out in this study, based on the intention-to-treat population, uMRD rate was 12% in the acalabrutinibobinutuzumab arm, 8% in the chlorambucilobinutuzumab arm, and 0.5% in the acalabrutinib monotherapy arm.Similar to acalabrutinib, the addition of obinutuzumab to ibrutinib also increased uMRD rates in PB to up to 35% in the iLLUMINATE study [45].Whether this improved uMRD rate might lead to additional clinical benefits is not yet clear, especially as patients in both ELEVATE-TN and iLLUMINATE continued treatment with the BTK inhibitor indefinitely.In summary, the prognostic value of MRD in the context of continuous BTK inhibition is very limited and thus MRD assessments so far do not have clinical implications.

MRD in the Context of Bcl2 Inhibitor Monotherapy or in Combination with CD20 Antibodies
Venetoclax, a Bcl-2 inhibitor, has so far demonstrated the best rate of MRD eradication when used alone and even more so when combined with other agents.After approximately a year of venetoclax use, studies reporting on venetoclax monotherapy in relapsed/refractory CLL showed uMRD rates of 26-36% in PB [46][47][48].When venetoclax was combined with other agents, the rate of uMRD increased even more.A phase 1b study for relapsed/refractory CLL showed the viability of venetoclax Importance of MRD in the Era of Targeted Therapies in CLL plus rituximab when uMRD rates in PB of 61% were found, ultimately enabling treatment discontinuation [49,50].The MURANO study also showed that adding the CD20 antibody rituximab to venetoclax increased uMRD rates at the end of combination therapy (i.e., after 6 cycles of rituximab plus venetoclax) to 62%, compared to just 13% in the comparator arm with bendamustine.This was true in patients with relapsed/refractory CLL (Table 1) [51].In the frontline setting, the CLL14 study demonstrated that the addition of obinutuzumab to venetoclax further increased uMRD rates with 76% of the patients having uMRD after venetoclax-obinutuzumab treatment compared to just 35% after chlorambucil-obinutuzumab therapy [52].In both the MURANO study and the CLL14 study, patients with uMRD had the longest PFS irrespective of CR or PR, suggesting that uMRD is one of the most sensitive surrogates for treatment efficacy also in the context of targeted agents [9, 53,54].

MRD in the Context of Bcl2 Inhibition plus BTK Inhibition
Since both Bcl2 inhibitors and BTK inhibitors target key vulnerabilities of CLL, combining agents venetoclax and ibrutinib was an obvious strategy to explore.First data were reported in 2019 from a monocentric phase 2 study of first-line ibrutinib plus venetoclax that demonstrated high best-MRD-responses of approximately 75%, although the rates in the intention-to-treat population after approx.One year of combination therapy was 55% [58].Similar rates were observed in the phase 2 CLARITY trial, which found an uMRD rate of 58% in patients with relapsed/refractory CLL [55].Of note, the UK FLAIR study, which among other research questions randomized patients to an MRD-guided ibrutinib plus venetoclax arm, showed that the rate of uMRD substantially increased after an additional year of combination therapy [60].Patients who received approx. 2 years of I + V had uMRD rates of 75%, thereby demonstrating that extended treatment with BCL2i plus BTKi can deepen remissions.Similar patterns were observed in several phase 2 studies on I + V in first-line and relapsed CLL [55,58,61,62].It is currently unclear how much this correlates with clinical endpoints and longer follow-up is required to understand whether patients derive any benefit from longer combination therapy.
The prognostic impact of uMRD at the end of combination therapy has become particularly evident from the randomized part of the CAPTIVATE study.This international phase 2 study first confirmed the efficacy of fixed-duration I + V in patients with previously untreated CLL [57].In a separate cohort, patients with uMRD (<10 −4 ) were randomized to either continuing on ibrutinib until progression or receive a placebo.The diseasefree and progression-free survival ultimately did not differ between patients on ibrutinib or on placebo, thereby demonstrating that patients do not benefit from continuing treatment if they reach uMRD [63].Notably, this pattern was also observed for patients with high-risk CLL.Conceptually, in order to understand the long-term benefits of maintenance treatment post combination therapy, several considerations have to be taken into account.First, the primary endpoint measure needs to be carefully considered since composite endpoints like PFS are likely to be extended by maintenance treatment compared to a placebo or watch and wait; however, the impact of second-line treatment on the long-term prognosis are neglected by this approach.Hence, the second PFS, i.e., PFS2, or even OS can be more insightful measures to compare maintenance versus no maintenance in the context of CLL [64].Second, given that patients with relapses after fixed-duration Ven-Obi or I + V are generally not refractory to BCL2 and BTK inhibition, the potential of re-treatment with the same agents can be an effective strategy to re-induce remissions with the same agents [65].Hence, future prospective studies exploring maintenance strategies need to consider OS or at least PFS2 as primary outcome measures.
In a follow-up analysis of the randomized GLOW study, in which patients with previously untreated CLL were randomized to receiving I + V or chlorambucilobinutuzumab, a strong interaction between IGHV status and EoT-MRD status was observed [56,66].Contrary to expectation, patients with unmutated IGHV had uMRD levels of approx.60%, whereas patients with mutated IGHV had uMRD levels of approx.40%; despite this, the PFS of patients with unmutated IGHV was shorter than patients with mutated IGHV [66].This observation suggests that in the context of BCL2i plus BTKi, the prognostic utility of MRD is dependent on the IGHV status since it this determines the kinetics of MRD response and regrowth.This is in line with previous observation with Ven-Obi that showed that MRD doubling time was strongly linked with presence of high-risk genomic features like IGHV or complex karyotype [9].

MRD in the Context of Triple-Targeted Therapy
Even though the rate of uMRD was significantly raised by the addition of CD20 antibodies or BTK inhibitors to venetoclax, some patients still had detectable MRD levels after 1 or 2 years of treatment.Numerous studies are examining whether a triple regimen of venetoclax, a BTK inhibitor, and a CD20 antibody raises uMRD rates even more.In a recent phase 2 study from Ohio State University, the rate of uMRD in BM was 67% in previously untreated CLL and 50% in relapsed/refractory CLL (MRD in PB not reported), which was lower than what has been observed with Ven-antibody doublet therapy [67].Similar triple combinations were tested in previously untreated CLL with TP53 aberrations in the CLL2-GIVe trial, which found 81% uMRD rates in PB [68].The phase 2 AVO study, which compares the effects of acalabrutinib, venetoclax, and obinutuzumab in previously untreated CLL [69], and the BOVen study, which compares the effects of zanubrutinib, venetoclax, and obinutuzumab [70], both found comparable rates of response and uMRD rates in PB of 82% and 92% after approximately 1 year of treatment.Overall, these phase 2 data do not definitively show that an upfront triple combination significantly raises uMRD rates over Venantibody doublets.Recently, the randomized CLL13-GAIA study, which contained a Ven-Obi arm and a Ven-Obi-Ibrutinib arm, reported uMRD rates between 87% with Ven-Obi and 92% with Ven-Obi-Ibrutinib, and comparable 3year PFS rates of 88 and 91%, respectively.Hence, the value of targeted triple therapy, both in terms of depth of remission and durability of response, remains unclear at this stage [71].

MRD in the Context of Immunotherapy
The use of allogeneic stem cell transplantation has been declining in the field of CLL, at least in areas where targeted agents are available, despite the fact that conventional cellular therapies like allogeneic BMT are still a mainstay in the treatment of some hematological malignancies like high-risk acute myeloid leukemia [72].Given that allogeneic stem cell transplantation has a curative potential for CLL, analysis of MRD response is particularly important.Patients with high-risk CLL were recruited in the CLL3X study, which reported PB uMRD 1 year after allogeneic stem cell transplantation in 27 (30%) of 90 transplanted patients [73,74].The CLL3X study also demonstrated that immunomodulation, either by tapering immunosuppressive drugs or donor leukocyte infusion, was responsible for achieving uMRD levels in 20% of the enrolled patients.Therefore, knowledge of uMRD can assist patients in the post-allogeneic transplant setting in making personalized treatment choices.Clinical trials are increasingly looking at anti-CD19 CAR T-cell therapy as a treatment choice for patients who relapsed on both BTK inhibitors and Bcl2 inhibitors.The first study on CAR T-cell therapy in CLL, and in hematology in general for that matter, reported CR with uMRD in 4 of 14 patients within 3 months of infusion [75].Again, uMRD status correlated with long-term disease control as no relapse has been reported in these patients thus far.Although the additional benefit of a BTK inhibitor is not yet known and the ideal timing of MRD assessment after CAR T-cell infusion has not yet been established, uMRD rates in PB between 75% and 90% have so far been reported [76,77].The follow-up of these studies is still relatively short and Importance of MRD in the Era of Targeted Therapies in CLL given that the overall reported PFS in studies like TRANSCEND had a median of 18 months, it is currently unclear how strong the post-infusion MRD status correlates with survival [78].

Using MRD to Guide Treatment Duration
In the age of chemoimmunotherapy, efforts to guide treatment duration based on MRD data have already been made.When uMRD was reached after 3 cycles in BM, Strati and colleagues published a retrospective analysis of patients treated with FCR in 2014 showing that patients who discontinued FCR treatment after 3 instead of 6 cycles, for example, due to toxicity, had a similar PFS and OS compared to patients treated with 6 cycles [79].This was further corroborated by the observation in the FCR300 study that patients with MRD ≤1% had a significantly longer PFS than patients with MRD >1% [80].This suggested that not all patients require the same duration of chemoimmunotherapy, though this has not yet been prospectively validated.
Different strategies to guide treatment by MRD status are presently being investigated in the context of targeted agents.Regarding stopping criteria, frequency, and sort of measurements, these ideas diverge.In a series of phase 2 studies, the GCLLSG assessed a method in which venetoclax, ibrutinib, or idelalisib are combined with obinutuzumab after an optional bendamustine debulking; for patients with uMRD in PB and CR over 2 subsequent visits (typically within 3 months), treatment was discontinued [81][82][83].One method was to use CR or uMRD after venetoclax-rituximab combination treatment as stopping criteria; so far, durable responses have been reported in relapsed/refractory disease [49,50].Other concepts consider the individual time until uMRD is reached to extrapolate the time until MRD might increase again; the UK CLL group tested this approach in the phase 2 CLARITY trial with venetoclax and ibrutinib in relapsed/refractory CLL [55].Patients who required 6 months to reach uMRD in BM discontinued treatment after overall 12 cycles, whereas patients who, for instance, required 12 months to reach uMRD were treated for 24 months in total.After 6 months of treatment, 24% of enrolled patients had uMRD and were therefore treated for overall 12 months; 58% of the patients had uMRD after 12 months and were therefore treated for 24 months.This concept is further investigated in the phase 3 FLAIR study which compares FCR, ibrutinib, ibrutinib plus rituximab, and ibrutinib plus venetoclax in previously untreated CLL [60,84].
In addition to adjusting treatment length on an individual basis, MRD may also be used to determine whether to increase treatment intensity on an individual basis.After 1 year of venetoclax monotherapy, 45% of all patients had uMRD in PB and BM and discontinued treatment, whereas 50% of the patients continued with venetoclax in addition to ibrutinib.Recently, Scarfo et al. [85] reported the initial findings from the phase 2 IM-PROVE study, in which patients who had started with venetoclax monotherapy and had not reached uMRD after 1 year of treatment received additional ibrutinib.The uMRD rate in PB and BM at month 24 was 84%, indicating that therapy intensification with a BTK inhibitor aids in MRD clearance.It should be noted that 53% of the patients had an MRD relapse during the most recent follow-up, despite no clinical disease development.This once more demonstrates how MRD dynamics evaluated by serial measurements over the course of follow-up may be able to improve the prognostic utility of individual MRD snapshots at particular timepoints.
The best way to use MRD as an instrument for clinical decision-making is still up for debate.Before the widespread use of MRD in everyday practice can be advised, a number of issues need to be solved.The first step is standardizing the MRD measurement technique.Flow cytometry-based MRD measurement for CLL is highly standardized and widely available in hematological laboratories, thanks to international collaborative initiatives like ERIC.The most practical and feasible cut-off is 10 −4 , but future research should consider the potential of higher sensitivity from NGS-based assays to determine whether a higher resolution is more informative.Second, the optimal MRD compartment that will serve as the foundation for treatment choices must be established.While many studies have confirmed that a CR or PR state is not very relevant when uMRD is achieved, the outcome of patients with discordant BM and PB MRD results is not entirely clear.Additionally, efforts are being made to measure MRD and retrieve cross-compartmental MRD information using cell-free DNA from plasma [86].This assay also opens the door to monitoring the emergence of new clones and the emergence of possible resistant mutations.Third, it is important to consider the quality of uMRD remissions.The fact that not all uMRD patients experience recurrence for the same length of time and that high-risk patients experience MRD relapse more quickly than others raises the possibility that clonal composition influences the stability of remission [86,87].In addition to these methodological issues, evidence from several randomized phase 3 studies suggests that BTK inhibitors can be used continuously to control disease Al-Sawaf without achieving uMRD [36,44,88].In contrast, MRDbased approaches may be more effective in preventing the emergence of resistance, improving tolerability, and ensuring long-term economic viability.Ongoing clinical studies like CLL17, which compares continuous ibrutinib to venetoclax-obinutuzumab and venetoclax-ibrutinib in previously untreated CLL, or MAJIC, which compares MRD-guided Ven-Obi to MRD-guided acalabrutinibvenetoclax will presumably answer this issue by comparing the different treatment paradigms to each other in a randomized, prospective way (NCT04608318, NCT05057494).

Conclusion
The EoT MRD status is one of the strongest prognostic features in the context of combination therapies.This was initially observed in the context of chemoimmunotherapy but has been largely confirmed also with targeted combination therapies.The incorporation of MRD assessments into regular clinical decision-making, beyond prognostication, will be a significant advancement.For current routine clinical care, flow cytometry-based MRD assessments with a 10 −4 cut-off from PB at the end of fixed-duration chemoimmunotherapy or targeted combination treatment should be considered to assess depth of remission.As demonstrated in this article, the MRD status can therefore be utilized as an additional powerful tool for reliable prognostication.
The available evidence presented in the review points to a number of potential scheduling approaches, including treatment schedules based on a single MRD assessment that determines whether or not therapy should continue, schedules based on multiple MRD assessments that direct stopping, continuing, or intensifying treatment, and schedules where MRD may be used only as a monitoring tool.To determine which patients have the deepest and most stable remissions as well as those who run the risk of either failing to achieve uMRD or losing MRD response, more prospective studies are required.In particular, randomized studies that compare MRD-guided treatment strategies to fixedduration, "MRD agnostic" strategies, are needed to disentangle patient-relevant benefits of MRD-guided treatment intensification or extension.Eventually, individualized treatment schedules that integrate MRD and individual genomic risk factors, will allow patients to undergo truly personalized and optimized therapy of CLL.