AFFILIATIONS
1 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
2 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
3 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
4 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
5 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
6 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
7 Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
8 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
9 Hospital de la Princesa, Madrid, Spain
10 Hospital Clinic of Barcelona, Barcelona, Spain
11 Hospital Central de Asturias, Oviedo, Spain
12 Hospital Universitario de Salamanca, Salamanca, Spain
13 Institut Català d'Oncologia (ICO) Hospital Duran i Reynals, Hospital, L'Hospitalet, Spain
14 Universitary Hospital of Santiago de Compostela, Santiago de Compostela, Spain
15 Hospital Universitario 12 de Octubre, Madrid, Spain
16 Hospital Donostia, Donostia, Spain
17 Consejo Superior de Investigaciones Científicas (CSIC), Sevilla, Spain, Hospital Universitario Virgen del Rocío, Sevilla, Spain, Instituto de Biomedicina de Sevilla (IBIS), Sevilla, Spain, Universidad de Sevilla, Sevilla, Spain
18 Hospital Universitario Morales Meseguer, Murcia, Spain
19 Hospital Clínico Universitario INCLIVA, Valencia, Spain, University of Valencia, Valencia, Spain
20 Hospital del Mar Research Institute (IMIM), Barcelona, Spain, Hospital del Mar, Barcelona, Spain, Molecular Cytogenetics Laboratory, Barcelona, Spain
21 University Hospital of Salamanca, Salamanca, Spain
22 Hospital Clinico de Valencia, Valencia, Spain
23 Hospital Clinic of Barcelona, Barcelona, Spain
24 Hospital Universitario de la Santa Creu i Sant Pau, Barcelona, Spain
25 Hospital Universitario Infanta Leonor, Madrid, Spain
26 Marqués de Valdecilla University Hospital, Santander, Spain, Universidad de Cantabria, Santander, Spain
27 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
28 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
29 Vall d'Hebron Institute of Oncology (VHIO), Barcelona, Spain, Vall d'Hebron University Hospital (VHUH), Barcelona, Spain
ABSTRACT
BACKGROUND
Transformation to diffuse large B-cell lymphoma is the most common form of Richter’s transformation (RT) and is associated with poor prognosis. Novel treatment strategies including CD20xCD3 bispecific antibodies (BsAbs) have improved response rates, yet relapses occur. Phase II trial GELLC-9-Richter (NCT06926205) evaluates frontline mosunetuzumab (M)-CHOP induction followed by M monotherapy maintenance in RT patients. To study immune dynamics and identify biomarkers of response, we longitudinally assessed immune dynamic remodeling by proteomic and transcriptomic single-cell technologies.
METHODS
Patients with RT received induction of 6 cycles of M-CHOP. Responding patients who were not considered candidates for cellular therapy received 11 additional cycles of M monotherapy. The primary endpoint was the rate of metabolic complete response (CR) at the end of induction (EoI), assessed according to Lugano criteria. Blood was collected before treatment (BT), before infusion and 6h post-infusion during the M step-up phase (C1D8, C1D15, C2D1) and before each cycle up to C6. We analyzed samples using conventional flow cytometry (FlowCyt) at all time points (n=25), and a 42-color spectral FlowCyt panel (n=14) and paired scRNA/TCRseq (n=8) on cryopreserved PBMCs collected at BT, C2 and C6.
RESULTS
Conventional FlowCyt (n=25) revealed generalized lymphodepletion, including a rapid CLL cell decrease. Effector memory (EM) abundance (%) increased (p<0.01) while naïve % decreased (p<0.01), revealing effector differentiation. Activation markers in T-cells (CD38, HLA-DR, CD86, DNAM1) increased, consistent with BsAb-induced T-cell engagement. To assess T-cell margination, we compared pre and 6h post infusion during M step-up, which revealed recurrent T-cell depletion post-infusion and recovery by next cycle, indicating redistribution out of the circulation, potentially towards tumor sites. Th1 showed the strongest decrease post-infusion (p<0.01), while Treg increased (p<0.01), indicating preferential mobilization of antitumor subsets. To find determinants of response, we compared immune profiles of responders (R) (n=12: 11 CR, 1 PR) versus non-responders (NR) (n=5 PD) among patients with clinical assessment at EoI. Rs exhibited a distinct immune signature with: (1) Higher % of naïve CD4 BT (p<0.01) and across cycles (p<0.02), suggesting a broader repertoire may influence response; (2) markedly different checkpoint dynamics, with lower PD-1 on CD4 at BT (p<0.01) and during treatment (p<0.02) and a transient TIM-3 and LAG-3 upregulation during C1 followed by rapid downregulation, while levels maintained elevated on NRs throughout treatment; and (3) higher baseline NK % (p<0.01).
Spectral FlowCyt expanded the immunophenotype for 14 patients at BT, pre-C2 and pre-C6. Unsupervised analysis revealed CD8 TIM-3+ cluster increase (p<0.01) from BT to C2, then decrease at C6 (p<0.01), indicating checkpoint expression due to strong early activation, later resolved, correlating with conventional FC data. We found differential T-cell composition at BT in Rs (n=9: 8 CR, 1 PR) versus NRs (n=5 PD), with higher CD4 naïve (p<0.01) and lower CD8 EM TIGIT+ (p<0.01) and CD8 CD127+HVEM+LAG+ (p<0.01), highlighting baseline immune composition may have predictive value.
To study transcriptional remodeling and T-cell clonality evolution, we performed paired scRNA and TCRseq analysis in 8 patients at BT, pre-C2 and pre-C6. scRNAseq revealed decrease in B cells (p<0.01) and increase in monocytes (p=0.01) from BT to C6, and among T-cells CD4 naïve decrease (p=0.01) and CD8 central memory increase (p<0.01). TCRseq identified substantial expansion of CD4 clones (C2 n=16; C6 n=64) and CD8 (C2 n=73; C6 n=285), with CD8 EM showing the largest expansion (n=99), indicating selective enrichment of highest antitumor capacity subsets.
CONCLUSIONS
M-CHOP reshapes circulating immune composition in RT, driving B-cell depletion, dynamic T-cell redistribution with preferential antitumor mobilization, and effector expansion. Responders had distinct baseline and treatment-induced immune features, including increased NK and naïve T-cell subsets, and distinct activation and exhaustion profiles such as lower PD-1 and higher TIM-3 and LAG-3 peaks short after administering drugs at C1. These biomarkers of response to BsAb-based therapy provide insights into mechanisms underlying clinical efficacy. Ongoing full-cohort analysis will further delineate determinants of M-CHOP response.