T‐cell immunomodulation occurs with different time kinetics during acalabrutinib and zanubrutinib therapy in chronic lymphocytic leukaemia

Chronic lymphocytic leukaemia (CLL) is characterized by profound immune dysregulation. In addition to impaired humoral immunity, patients with CLL exhibit abnormal T-cell subset distribution, shift towards antigen-experienced phenotypes, chronic activation and features of exhaustion, which contribute to impaired anti-tumour immunity and higher risk of infections. These immune alterations are increasingly recognized as a key part of the CLL disease phenotype, influencing both disease progression and response to therapy.1 Bruton tyrosine kinase (BTK) inhibitors (BTKi) have revolutionized CLL treatment. Ibrutinib, the first-in-class BTKi, induces durable disease control across treatment settings and has been shown to remodel the immune system. Ibrutinib reduces abnormal peripheral T-cell expansions, restores a more balanced CD4+/CD8+ ratio, decreases regulatory and Th1-skewed populations and partially reverses exhaustion marker expression.2, 3 The role of Interleukin-2-inducible T-cell kinase (ITK) inhibition has been suggested as a possible contributor, since ITK blockade can bias CD4+ T cells towards a Th1 phenotype, although the overall relevance of this mechanism remains uncertain.4 By contrast, the second-generation BTKi acalabrutinib and zanubrutinib were designed for higher kinase selectivity and show only minimal binding to ITK.5 Both drugs are at least as effective as ibrutinib; zanubrutinib recently demonstrated superior progression-free survival in the ALPINE trial, with 42.5-month follow-up, confirming a sustained advantage over ibrutinib, including in high-risk subgroups such as del(17p)/TP53-mutated patients.6 Importantly, second-generation BTKi are associated with lower rates of cardiovascular toxicity.5-7 Their immunological effects, however, are less defined. For zanubrutinib, available data are limited to short-term follow-up (6 months),8 underscoring the need for long-term prospective studies. We, therefore, performed longitudinal immune monitoring in patients receiving these agents to investigate their impact on T-cell reconstitution and exhaustion. We prospectively analysed 34 patients with CLL, of whom 18 received acalabrutinib and 16 received zanubrutinib. Clinical characteristics are summarized in Table S1. Peripheral blood samples were collected at baseline and at serial time points during therapy up to 24 months. Age- and sex-matched healthy donors (HD) served as controls. Flow cytometry quantified major T-cell subsets, helper subsets, memory differentiation states and exhaustion marker expression. Additional analyses included proliferative and resting fractions of CLL cells, plasma biomarker profiling by proximity extension assay (PEA) and BTK mutation screening. Detailed methodology is available in the Supporting Information: Methods. At 2 years of follow-up, most patients remained on therapy and achieved at least partial remission. In the acalabrutinib group, four patients had discontinued treatment due to progressive disease (n = 2) or intolerable toxicity (n = 2). In the zanubrutinib group, four had discontinued because of Richter transformation, secondary malignancy, COVID-19–related death or skin toxicity. Extended follow-up (median 50 months) revealed three additional progression events, all in relapsed/refractory (R/R) patients receiving zanubrutinib. Emergence of hotspot BTK mutations was observed in both cohorts, primarily in R/R patients (Figure S1). Total CD19+ B cells decreased significantly during therapy, with earlier reductions in the acalabrutinib cohort (by month 4) compared with the zanubrutinib cohort (by month 6), consistent with effective disease control in both groups (Figure 1A). Analysis of proliferative (PF) and resting (RF) CLL fractions showed no differences within each treatment group. When the two patient groups were pooled together to increase statistical power, a significant reduction of PF was observed (Figure S2). T-cell numbers were also abnormal at baseline (Figure 1B). Both CD4+ and CD8+ counts were elevated compared with HD, particularly striking for CD8+ cells: 3.9-fold higher in the acalabrutinib cohort and 8.5-fold higher in the zanubrutinib cohort (p < 0.0001). No significant differences were observed between treatment naïve (TN) and R/R patients within each treatment group. During therapy, both subsets decreased significantly. CD4+ T cells normalized by month 2 under acalabrutinib and by month 6 under zanubrutinib treatment. CD8+ T cells declined in both cohorts but, like CD19+ cells, remained elevated above HD levels at 2 years (Figure 1C,D). Persistent CD8+ lymphocytosis has also been reported with ibrutinib and may reflect chronic antigen stimulation.2, 9 Exhaustion marker expression was also abnormal at baseline. The frequency of PD-1+ and TIGIT+ CD4+ and CD8+ cells was significantly higher in patients compared with HD. Both subsets declined during therapy, earlier in the acalabrutinib cohort (by 2–3 months) compared with zanubrutinib (by 6 months). However, exhausted CD8+ cells remained above HD at 2 years, indicating incomplete restoration of immune competence (Figure 1E–H). This pattern is consistent with findings from prior ibrutinib studies, which also demonstrated partial but not complete reversal of exhaustion.2 Helper T-cell subsets showed distinct dynamics. Th1 cells were elevated in both cohorts at baseline. With acalabrutinib, levels declined by month 3 and reached HD levels by month 8. In contrast, declines did not occur until month 8 in the zanubrutinib cohort, and Th1 frequencies remained above HD at 2 years (Figure S3A). Th2 and Th17 remained largely within the HD range with only modest changes (Figure S3B,C). Regulatory T cells (Tregs) were expanded in both groups at baseline and decreased during therapy, normalizing by month 2 in the acalabrutinib group and by month 6 in the zanubrutinib group (Figure S3D). The decline in Tregs is consistent with previous observations under ibrutinib2, 10 and may contribute to enhanced immune surveillance. Memory subset remodelling was prominent (Figure 2). Both CD4+ and CD8+ central memory (CM) cells were elevated at baseline, consistent with chronic antigen exposure. Under therapy, CM cells decreased to HD levels more quickly in acalabrutinib-treated patients (by 3 and 12 months, respectively), whereas normalization was delayed in zanubrutinib-treated patients (CD8+ CM remained above HD at 24 months) (Figure 2B,F). CD4+ and CD8+ effector memory (EM) subsets, highly expanded at baseline, decreased progressively under both drugs and approached HD levels by 8 months to 1 year (Figure 2C,G). Terminal effector CD8+ T cells re-expressing CD45RA (TEMRA) were expanded foremost in the zanubrutinib group at baseline and normalized by 12 months (Figure 2D,H). Naïve CD4+ T cells, HD equivalent at baseline, declined below HD by month 6 in acalabrutinib and by 24 months in zanubrutinib group (Figure 2A). In contrast, naïve CD8+ cells remained stable throughout therapy (Figure 2E). These findings suggest that both drugs rebalance skewed differentiation, but prolonged therapy may reduce the naïve CD4+ compartment, potentially limiting regenerative capacity. Plasma biomarker profiling by PEA demonstrated a broad reduction of inflammatory mediators. Of 92 inflammatory biomarkers, 17 declined significantly in both cohorts, including the CLL-associated proteins CCL3, CCL4, CD5 and CD6. This pattern is consistent with prior ibrutinib observations, where CCL3 and CCL4 were identified as pharmacodynamic markers of effective B-cell receptor pathway inhibition.10, 11 Distinct drug-specific signatures were also identified: acalabrutinib uniquely modulated 15 biomarkers, including a transient increase in MMP1 and AXIN1 at 1 month, while zanubrutinib selectively altered 11 markers, notably upregulating MCP2 and MCP3 (Figure S4). These differences suggest partially overlapping but also distinct effects of the two inhibitors on the immune microenvironment. Together, our data demonstrate that both acalabrutinib and zanubrutinib induce marked immunological changes in CLL, including normalization of CD4+ subsets, rebalancing of memory T-cell populations and decreased exhaustion. The kinetics differed modestly, with acalabrutinib generally associated with faster normalization, particularly for CD4+ cells, Tregs and CM subsets. Zanubrutinib showed more delayed effects but converged over long-term follow-up. Importantly, CD8+ expansions and exhaustion persisted at 2 years, indicating that full immune reconstitution may not be achieved even with long-term BTKi monotherapy. These observations extend previous findings with ibrutinib to second-generation BTKi. Importantly, the similarities between all three BTKi support the notion that T-cell changes are largely driven by disease control rather than ITK inhibition. Nonetheless, the somewhat faster kinetics under acalabrutinib as well as the distinct protein signatures could reflect differences in pharmacology. Our findings have also implications for immunotherapy. The success of chimeric antigen receptor T-cell (CAR-T) therapy in CLL depends on the quality of autologous T cells, with higher frequencies of naïve and CM subsets predicting superior expansion and durable responses.12, 13 By reducing exhaustion and rebalancing differentiation subsets, acalabrutinib and zanubrutinib may enhance the quality of T cells for collection. In this context, recent data highlight that durable CAR-T responses in CLL depend not only on improved T-cell fitness but also on effective tumour debulking, as high disease burden has been shown to impair CAR-T expansion and persistence.14 BTK inhibition may therefore both lower tumour burden and restore T-cell balance, optimizing conditions for T-cell harvest. In conclusion, acalabrutinib and zanubrutinib both promote significant T-cell immunomodulation in CLL, consistent with many of the beneficial immune effects of ibrutinib while offering improved tolerability. Both agents reduce abnormal expansions, remodel helper and memory subsets and reduce exhaustion without full normalization, leaving overall immune restoration partly incomplete after 2 years of treatment. These effects are an important component of BTKi therapy and may help optimize infection control and integration with immunotherapies. In general, reductions and normalizations appeared earlier with acalabrutinib; however, it remains uncertain whether this reflects true drug-specific effects or differences in baseline T-cell numbers, sample processing and cohort heterogeneity. Further studies in larger cohorts are needed to confirm these observations and their significance. MP and AÖ supervised the sample collection. MLA collected the clinical data. KH and TAM designed and performed the flow cytometry analyses. MLA, OL and KH analysed the flow cytometry data. AB and QW prepared samples and analysed the PEA data. TdPS performed the PCR analyses. LP-P processed and analysed the PEA data. MLA, OL, AB and LP-P created figures. MP designed the research and supervised the data analysis. MLA, OL, AB, LP-P, TPS, AÖ, RZ, RM, RR, CIES and MP interpreted the results and wrote the manuscript. The authors thank the patients who consented to the use of their samples for this study and Ann Svensson for technical support. This work was supported by grants from the Swedish Blood Cancer Foundation, the Swedish Cancer Foundation, the Swedish Research Council, King Gustav V Jubilee Fund, the Stockholm County Council, the Cancer and Allergy Foundation and the Center for Innovative Medicine (CIMED) and by unrestricted research grants by BeOne Medicines and AstraZeneca (MP). RR received honoraria from AbbVie, AstraZeneca, Illumina, Janssen, Lilly and Roche. MP received unrestricted research grants from BeOne Medicines and AstraZeneca. The datasets and protocols generated and analysed during the current study are available from the corresponding author (Maria Andersson, [email protected]) upon reasonable request. Access will be granted for non-commercial academic purposes and in accordance with institutional and ethical guidelines. Data S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.