Resistance to Bruton's tyrosine kinase inhibitors (BTKi) remains a major therapeutic challenge in B-cell malignancies, limiting treatment durability. Here, we identify ferroptosis suppression as a central mechanism of BTKi resistance in mantle cell lymphoma (MCL). Aberrant BRG1 activity protects cells from BTKi-induced ferroptosis by restricting reactive oxygen species (ROS) and labile iron. Mechanistically, BRG1 promotes resistance through both BTK-dependent survival signaling and a BTK-independent transcriptional program. The latter is mediated by BRG1-driven induction of MEF2B, which upregulates NDUFA4L2 to inhibit mitochondrial respiration, thereby blocking mitochondria-dependent ferroptosis. Pharmacologic inhibition of BRG1 disrupts these programs, restoring ferroptotic sensitivity and synergizing with BTKi across resistant MCL models. Together, these findings establish BRG1 as a central regulator of therapy resistance and provide a rationale for co-targeting BRG1 and BTK as a therapeutic strategy for B-cell malignancies.

Mantle cell lymphoma (MCL) is an uncommon lymphoid malignancy, accounting for 2.5-6% of all non-Hodgkin lymphomas. Covalent Bruton tyrosine kinase inhibitors (cBTKi), such as ibrutinib, zanubrutinib, and acalabrutinib, are mainstays of treatment relapsed MCL, and their use is emerging in the front-line setting. However, progression on cBTKi treatment is common, and these patients often experience more aggressive disease and a poor clinical outcome. Management of cBTKi-refractory MCL is non-standard and challenging. Pirtobrutinib, a highly selective, first-in-class reversible BTKi, has demonstrated clinical activity in relapsed/refractory (R/R) MCL, but responses only occur in about half of patients and the response duration is limited. Identifying relapse-prone features in MCL patients and improving BTKi-based therapeutic approaches is a critical area of research with significant unmet need. In this context, through an unbiased CRISPR/Cas9 screening approach, we discovered a significant dependency of cBTKi-resistant, both TP53 wild type and mutant MCL cells, on SMARCA4, the common catalytic subunit of the mammalian BAF nucleosome remodeling complex. Notably, components of the BAF complex are mutated in at least 10% of MCL cases, with SMARCA4 mutations implicated in resistance to cBTKi combination treatment. Stable expression of SMARCA4 T910M mutant, a recurrent mutation in MCL, led to increased resistance to cBTKi in various MCL cell lines, highlighting the importance of SMARCA4 in MCL maintenance. We found that pharmacological inhibition of SMARCA4 by FHD-286, an allosteric SMARCA2/4 ATPase inhibitor, which is a clinical stage asset owned by Foghorn Therapeutics and currently under investigation in AML in combination with decitabine and cytarabine, potently induced cytotoxicity in MCL cells at low nanomolar ranges. Remarkably, MCL cells expressing the SMARCA4 T910M mutation exhibited even greater sensitivity to FHD-286 compared to those expressing wild-type SMARCA4. These findings suggest that targeting SMARCA4 may represent a novel therapeutic vulnerability in cBTKi-resistant MCL. Mechanistically, our precision run-on (PRO)-seq combined with CUT&RUN analysis of cBTKi-resistant MCL cells revealed that SMARCA4 inhibition suppresses the transcription of a set of genes in the B cell receptor (BCR) signaling pathway (i.e., CD79A/B, SYK, BTK, RAC2, CARD11). We investigated whether maintaining active BCR signaling is a mechanism through which SMARCA4 supports the survival of MCL cells. By comparing cBTKi-resistant and sensitive isogenic MCL cell line pairs, we identified NR4A1 gene expression as a marker of activated BCR signaling, reflecting resistance to cBTKi. Both genetic and pharmacological targeting of SMARCA4 transcriptionally suppressed NR4A1 expression in cBTKi-resistant MCL cells, supporting its role in BCR signaling. Consistently, combining FHD-286 with pirtobrutinib led to a substantial synergistic cytotoxic effect in MCL cell lines and primary MCL cells. Altogether, our findings suggest that SMARCA4 plays a tumor maintenance role in MCL. Targeting SMARCA4 can be further exploited as a therapeutic strategy in cBTKi-resistant MCL and may be even more effective in SMARCA4-mutated MCL.

BACKGROUND: Initial treatment for mantle cell lymphoma (MCL) continues to evolve with the optimization of targeted agents. We previously reported durable efficacy of the lenalidomide and rituximab (LR) regimen as frontline MCL therapy (NEJM 2015:373:1835; Blood 2018:132:2016). We hypothesized that the addition of acalabrutinib (A), a next-generation BTK inhibitor, to LR (ALR) would accelerate molecular CR, allowing response-adapted and time-limited treatment. We also explored feasibility of the obinutuzumab-based triple combination of ALO. We report the mature finding of the triplet ALR, as well as preliminary data on ALO. METHODS: The study included induction and maintenance, with acalabrutinib provided at 100 mg BID continuously. Lenalidomide was administered at 15 mg during induction (cycles 1-12), with dose escalation to 20 mg if tolerated, then dose reduced to 15 mg during maintenance. For ALR (n=24), rituximab was administered weekly x 4 during cycle 1, then once every other cycle throughout induction and maintenance. For ALO (n=10), obinutuzumab was administered weekly x 3 during cycle 1, monthly for cycles 2-6, then once every other cycle for study duration. Acalabrutinib and lenalidomide could be discontinued after 24 cycles of treatment for subjects achieving MRD-undetectable CR. All treatment could be discontinued after 36 cycles. The primary objective was MRD-undetectable (<10-6) CR rate at the conclusion of ALR induction therapy. Secondary objectives included safety, response rates by Lugano criteria, survival. Exploratory objectives included peripheral blood MRD (PB-MRD) and cfDNA. PB-MRD was measured with AdaptiveBiotech ClonoSeq assay. Tumor genomic mutation profile and cfDNA were assessed by CAPP-Seq NGS with bioinformatic pipeline for variant calling. RESULTS: 24 subjects were enrolled for ALR, 10 for ALO, and the study met its accrual. At study entry, median age was 64 years (range 35-82), and the M:F ratio was 4:1. All patients had stage III/IV disease, 38% had elevated LDH, and 91% had bone marrow involvement. MIPI scores were 38% low-, 32% intermediate-, and 29% high-risk. Ki67 index was <30% in 65% subjects. TP53 mutations were detected in 29% of subjects. Treatment was associated with expected side effects. Grade 3-4 hematologic toxicities included asymptomatic neutropenia (38%), thrombocytopenia (4%) and anemia (4%). Grade 3-4 non-hematologic toxicities included rash (42%), fatigue (4%), nausea (4%), and vomiting (4%). Routine infections were mostly grade 1/2; COVID-19 infection was common (88%) due to the Omicron wave. For ALR, all 24 (100%) patients completed induction and moved onto maintenance. The ORR was 100% (90%CI=88%-100%) and CR 83% (90%CI=66%-94%) after 12 cycles of induction. PB-MRD was undetectable (<10-6) in 50% patients (12 of 24) after 6 cycles, 67% (16 of 24) after 12 cycles, and 77% (17 of 22) after 24 cycles. 83% patients (5 of 6) with TP53 mutations achieved molecular CR after 12 cycles. Three patients progressed while on maintenance therapy, including 2 with TP53 mutations. 16 (67%) patients in molecular remission discontinued all study treatment including at least Acala-Len x24 cycles and R maintenance x36 cycles, and stayed in remission. At a median follow-up of 41 months (range 37-51), the 3-yr OS and PFS for ALR were 95.2% (95%CI=86%-100%) and 87.5% (95%CI=74%-100%), respectively. MIPI score, Ki67 index, TP53 mutations, or MRD status had no impact on survivals. For ALO (n=10), both ORR and CR were 90% (90%CI=61%-100%) following induction. PB-MRD was undetectable (<10-6) in 89% patients (8 of 9) after 6 cycles, 100% (9 of 9) after 12 cycles. At a median follow-up of 17 months (range 13-19), the 1-yr OS and PFS were both 100% (95%CI=100%-100%). Biomarker study with longitudinal cfDNA analysis in ALR revealed clonal evolution of pre-existing as well as new TP53 mutation subclones at time of relapse as possible resistance mechanisms. CONCLUSIONS: This study demonstrates that the ALR regimen has high rates of durable complete and molecular responses, and is feasible as a time-limited initial therapy for patients with both TP53 wild type and mutant MCL. The ALO regimen is feasible with encouraging initial safety and efficacy. MRD and cfDNA analyses provide real-time and non-invasive monitoring of molecular response and mutational evolution, which warrants further evaluation in response-adapted strategy (ClinicalTrials.gov - NCT03863184).

This comprehensive study delves into the intricate dynamics of CDK4/6 inhibitors (abemaciclib, ribociclib, palbociclib) efficacy in the context of pediatric brain tumors, with a specific focus on establishing the temporal dependencies of their effectiveness. Beyond elucidating the time-dependent aspects, the research also explores into the feasibility and safety considerations associated with a metronomic intrathecal delivery with Palbociclib. This involves a thorough examination of pharmacokinetics and an analysis of liver and renal function through a panel assessment in vivo. By methodically exploring the optimal dosing within the brain after administration of Palbociclib via a systemic or intrathecal administration; the study aims to provide nuanced insights into enhancing therapeutic outcomes for pediatric brain tumors. Ultimately, we also provide a mechanistic understanding of the inhibition of CDK4/6 over time using RNAseq analysis; and an assessment of the efficacy of the proposed metronomic intrathecal delivery strategy in a preclinical mouse model of leptomeningeal ATRT. This multifaceted study not only expands our understanding of the temporal dynamics of CDK4/6 inhibitors but also lays the foundation for informed decision-making regarding the potential of this innovative therapeutic delivery method in the realm of pediatric brain tumor treatment.

The immunomodulatory drugs (IMiDs) lenalidomide (Len) and pomalidomide (Pom) are standard of care for multiple myeloma (MM). Cereblon (CRBN), a component of the CRL4CRBN E3 ligase, is required for IMiD's anti-myeloma activity. Binding of IMiD to CRBN promotes the recruitment of transcription factors IKZF1 and IKZF3 to CRL4CRBN for ubiquitination and degradation, leading to loss of IRF4, an IKZF1/3 target essential for survival of myeloma cells. MEIS2 was identified by protein array as an endogenous substrate that competes with IMiD for binding to the same domain on CRBN. However, whether MEIS2 functions as an inhibitor of CRBN and its relevance in IMiD therapy has not been investigated. To address this, first we discovered that high MEIS2 RNA expression in myeloma cells before therapy was tightly associated with poorer overall survival (OS) (p-value = 0.0001) in the CoMMpass Bor-Lenalidomide-Dex clinical trial (N=272). Conversely, high CRBN RNA expression correlated with improved OS (p-value =0.0077). Next, we demonstrated that MEIS2 protein was differentially expressed in primary MM cells but absent in normal plasma cells (PC)s, B cells and lymphoma cells. In contrast, the expression of CRBN protein varied in PCs, and was significantly greater in treatment-naïve MM cells than in IMiD refractory MM cells. This resulted in a ~3-fold increase in the MEIS2 to CRBN ratio in IMiD refractory patients, implicating a role for MEIS2 in inhibiting CRBN and promoting clinical resistance to IMiD therapy. In addition, high expression of core genes that promote cell cycle progression through early G1 (CDK4 and CDK6) and late G1 (E2F1, CDK2, CCNA2) also tightly correlated with poorer OS. This is in line with dysregulation of G1-S progression in relapsed-refractory MM we previously reported (Ely et al, 2005) and suggests that inhibition of CDK4/6 may reprogram MEIS2 for enhanced IMiD killing. Indeed, induction of prolonged early G1 arrest by CDK4/6 inhibition with palbociclib 1) enhanced Len and Pom killing in model MM cell lines; 2) reduced the MEIS2 protein by transcriptional repression and increased the CRBN protein by post-transcriptional regulation in cooperation with Len; 3) this led a > 5 fold reduction in the MEIS2/CRBN protein ratio, exacerbated loss of IKZF1/3 and IRF4, and enhanced apoptosis (caspase- cleavage). Importantly. these findings were validated in freshly isolated primary bone marrow myeloma cells (BMMCs) (n=27) in a stromal co-culture, and the IMiD response ex vivo mimicked the clinical response to the first post-biopsy Len or Pom therapy in patients independent of prior therapy. Thus, induction of early G1 arrest by CDK4/6 inhibition reprograms primary MM cells for IMiD killing by reducing the MEIS2 to CRBN protein ratio through opposing regulation of MEIS2 and CRBN. Loss and gain of function studies demonstrated that MEIS2 inhibits IMiD killing by attenuating the induction of interferon downstream of the IKZF1/3-IRF4 axis, more prominently in early G1 arrest induced by CDK4/6 inhibition in MM cell lines as well as primary BMMCs. Further biochemical analyses revealed that CDK4/6 inhibition rapidly accelerated the displacement of MEIS2 from CRBN within one hour of Len treatment, concurrent with enhanced ubiquitination and degradation of IKZF1 and IKZF3 in both MM cell lines and primary BMMCs. In summary, we have demonstrated for the first time that MEIS2, an endogenous CRBN substrate, is a negative regulator of IMiD therapy in MM: 1) High expression of MEIS2 is a prognostic indicator for poor OS in IMiD therapy, in opposition to improved OS with high CRBN expression; 2) MEIS2 inhibits IMiD killing of MM cells by attenuating the induction of interferon; 3) inhibition of CDK4/6 enhances IMiD killing by rapidly displacing MEIS2 from CRBN through enhanced Len recruitment of IKZF1/3 and by time-dependent opposing regulation of MEIS2 and CRBN synthesis. Selective expression of MEIS2 in MM cells but not in normal PCs, B cells or lymphoma cells further suggest MEIS2 as a druggable target in MM. In this study, we provide the first evidence that CDK4/6 inhibition reverses MEIS2 inhibition of CRBN that has a potential to enhance the clinical response to IMiD. The prospect that through cell cycle control CDK4/6 inhibition may also enhance the anti-MM activity of CELMoDs, the newly developed IMiDs such as iberdomide, in relapsed-refractory setting is exciting and warrants future investigation.

Drug resistance remains a formidable challenge in MCL, largely due to unrestrained proliferation of MCL cells driven by aberrant Cyclin D1 and CDK4 expression. In preclinical studies, inhibition of CDK4/6 not only induces early G1 cell cycle arrest but also reprograms MCL cells for therapy vulnerability, suggesting that CDK4/6 inhibition may deepen and prolong the clinical response to BTK inhibition. We tested this hypothesis in a phase I clinical trial for recurrent MCL and defined the resistant mechanism by integrated longitudinal genomic analysis of individual patients. Palbociclib (CDK4/6i) was administered on days 1-21 of a 28-day cycle; ibrutinib (BTKi) was given continuously. For longitudinal analysis, sequential tissue and blood specimens from 27 evaluable patients were collected before and during therapy, and on progression. Single cell RNA-seq (scRNA-seq) analysis of PBMCs or the monocytic fractions from bone marrow and lymph node (53 samples) was performed using a MCL-specific in-house reference library. The data were then integrated with whole transcriptome (WTS) and whole exome sequencing (WES) of MCL cells isolated from the same specimens, flowcytometry, immunoblotting, IHC, and CBC /differential to deduce the number of total MCL and immune cells in each specimen. CDK4/6 Inhibition appears to deepen and prolong the BTKi response, with a CR rate of 42% and 5 patients (2 CR and 3 PR) remained on therapy for ~ 10 years. Longitudinal scRNA-seq (210,000 cells) revealed that MCL cells comprise 4 major transcriptomically distinct clusters with cluster 1 (C1) resembling quiescent normal B cells. RNA velocity analysis further indicated that both C2 and C3 are in late G1, having evaded CDK4/6 inhibition; but only C2 progresses to the proliferating C4. C3 is terminally arrested in late G1 and long-lived with elevated BCL2 and IRF4 expression. In durable CR, all MCL clusters were depleted as expected. Primary resistance and progression on therapy correlated with a marked expansion of C2 or C3 MCL cells. Integrated longitudinal scRNA-seq analysis with WES and WTS revealed that C3 resistance was due to RB1+/del CDKN2A+/del (hemizygous loss of RB1 and CDKN2A), whereas C2 resistance was driven by CDK4amp-RB1+/del-CDKN2A+/del. Thus, compound copy number variation (cCNV) drives differential resistance to CDK4/6i and BTKi, reinforcing the critical importance of controlling the CDK4/6-RB-CDKN2A axis in MCL therapy. Durable CR was also associated with maintenance of CD8+ T cells and CD8+ T effector memory (TEM) cells. Conversely, progression on therapy was accompanied by rapid depletion of CD8+TEMs and proliferating (Ki67+) CD8+ T cell expressing all exhausting markers including LAG3,TIGIT and TIM3. Collectively, our data suggest that CD8+ T cell surveillance cooperates with MCL cell intrinsic cCNV to discriminate durable clinical response from resistance to targeting CDK4/6 and BTK. Moreover, IL10 was produced in bone marrow MCL cells before therapy regardless of the subsequent clinical response, but downregulated in early G1-arrested in vivo in a responding patient. Plasma IL10 was selectively elevated in MCL patients on progression from CR, along with sustained IL10 synthesis in myeloid cells. These novel findings suggest that IL10 promotes resistance to CDK4/6i and BTKi and a new strategy to overcome resistance. Indeed, exogenous IL10 completely abolished the killing of MCL cells from a PR patient by CDK4/6i+ BTKi and CDK4/6i+ PI3Ki ex vivo. IL10 also abrogated cCNV-guided killing of MCL cells C3 - and C2-resistant patients by BTKi+BCL2i, and C3-resistant patients by CDK2i+BCL2i ex vivo. All therapies were restored by disrupting IL10 signaling. In summary, by integrated longitudinal scRNA-seq analysis of a hypothesis-driven clinical trial, we have provided the first evidence in humans that 1) CDK4/6 inhibition deepens and prolongs the clinical response to BTKi; 2) cCNV of CDK4-RB1-CDKN2A drives differential expansion of resistant C2 MCL cells in transit to proliferating C4 MCL cells or long-lived non-proliferating C3 MCL cells; 3) CD8+ T cell maintenance and surveillance cooperates with MCL cell intrinsic cCNV to discriminate durable response from resistance; and 4) IL10 is a key mediator of MCL-Immune cell interaction that promotes resistance. These discoveries have profound implications for genome-guided strategy to overcome drug resistance in MCL.

PDF file - 53K, Validation of gene expression data using QPCR compared to microarray values (data reported as fold change).

PDF file - 60K, Kaplan-Meier estimates of (A)PFS and (B)OS in patients in the ITT population (N-38).

PDF file - 72K, Most common all-grade and grade ≥3 AEs during DoVeD induction and bortezomib-based stem cell mobilization.

PDF file - 59K, Responses rates after DoVeD induction, bortezomib- or non-bortezomib-based mobilization, and SCT.

Detection of BTK C481S mutation in MCL patients at relapse from a durable but not transient ibrutinib response or no response.

<p>Clinical characteristics of familial myeloma probands who were studied with germline whole exome sequencing.</p>

Supplementary Methods and Supplementary Figure Legends from Cell-Cycle Reprogramming for PI3K Inhibition Overrides a Relapse-Specific C481S BTK Mutation Revealed by Longitudinal Functional Genomics in Mantle Cell Lymphoma

Significant non-synonymous damaging SNVs detected in CDS by WTS in Pt 1 serial samples.

Key Points • PRMT5 inhibition induces a FOXO1 driven, proapoptotic program in MCL.• Combination therapy with PRMT5 and BCL-2 inhibition is synergistic in preclinical MCL models, including those with ibrutinib resistance.

NIK and p52 of the alternative NF-kappaB pathway are expressed in JEKO-1 MCL cells and not modulated by targeting of CDK4 with PD 0332991 in combination with ibrutinib.

<p>Supplementary Figure S2. PD 0332991 synchronization enhanced cytotoxic killing of U937 cells by Ara-C.</p>

Supplementary Figures 1-2, Table 1 from A Novel Orally Active Small Molecule Potently Induces G1 Arrest in Primary Myeloma Cells and Prevents Tumor Growth by Specific Inhibition of Cyclin-Dependent Kinase 4/6