Session Summaries from the ESH-iCMLf 28th Annual John Goldman Conference on CML: Biology and Therapy - Session 2

Scientific Session: Immunology and stem cells

Chair: Ravi Bhatia (Birmingham, AL)

This session explored how immune regulation, leukemic stem-cell biology and early treatment responses may shape long-term outcomes in CML. Across the keynote and three presentations, a common theme emerged: achieving deeper and more durable control of CML may require looking beyond BCR::ABL1 inhibition alone. The speakers examined the role of immune control in treatment-free remission (TFR), strategies to selectively eliminate persistent leukemic stem and progenitor cells (LSPCs), inflammatory signalling that influences the competitive fitness of leukemic versus normal stem cells, and single-cell profiling as a potential tool for predicting response and personalising TKI therapy.

Our key takeaways from the presentations:

Keynote presentation

Immune mechanisms in CML: where do we stand and where do we go now
Speaker: Agnes Yong (Perth)

“The immune milieu is a balance between effector and suppressor arms of the immune system.” - Agnes Yong

Key points:

• TFR determinants: Longer TKI treatment and longer duration of DMR, together with a deeper molecular response at treatment cessation, are associated with greater TFR success. Residual BCR::ABL1 positivity does not necessarily indicate disease capable of causing molecular relapse, as it may reflect expression in B lymphocytes. 
• Immune control: Successful TFR appears to involve a balance between immune effector and suppressor populations, with higher NK-cell numbers associated with TFR success and increased regulatory T cells and monocytic myeloid-derived suppressor cells associated with relapse. 
• Clonal extinction: A key unresolved question is the relative contribution of true clonal extinction versus ongoing immunological control in maintaining long-term TFR.
• Interferon: IFN-α combined with TKI can produce earlier and deeper molecular responses, but large clinical studies have not demonstrated an increase in TFR success. 
• Immunomodulation: Strategies to improve TFR should aim to enhance T-cell and NK-cell function, although previous approaches have limitations: lenalidomide proved too toxic and a suitable anti-TIM-3 agent is currently unavailable. 
• Next steps: Further immunomodulation remains an important avenue for improving TFR, with dendritic-cell vaccination now being developed and will form the group's next TFR study in Australia.

Presentation 1

Selective targeting of leukaemic stem cells in CML via dual-marker bispecific antibody immunotherapy
Speaker: Zsombor Kerekes (Glasgow)

“Yes, we can exploit dual antigen co-expression to selectively target LSPCs.” - Zsombor Kerekes

Key points:

• Dual targeting: Cooperative binding to CD34 and ITGA2B increases antibody binding avidity for double-positive CML LSPCs compared with cells expressing only one of the markers.
• Selectivity: Increased binding stability enables preferential binding of double-positive LSPCs over single-positive cells, with up to 28-fold increased binding.
• Immune recruitment: Fcγ receptor-dependent recruitment of effector immune cells translated selective antibody binding into preferential killing of double-positive LSPCs.
• In vivo activity: The approach selectively depleted double-positive target cells in vivo, while showing greater preservation of CD34 single-positive healthy HSPCs. 
• Proof of concept: The findings demonstrate that dual-antigen co-expression can be exploited to selectively target CML LSPCs, potentially overcoming limitations of monospecific LSC-directed antibodies.
• Future work: Key next steps are to demonstrate depletion of true long-term repopulating Ph+ LSCs, further improve double-positive binding selectivity through affinity engineering, and minimise single-positive bystander killing through dose optimisation.

Presentation 2

Signalling through TNFRSF1A enhances the competitive advantage of CML LSCs relative to normal HSCs during TKI treatment

Speaker: Ravi Bhatia (Birmingham, AL) on behalf of Tyler Freeland
“TNF receptor 1A signalling in CML cells is suppressing the competitive fitness of normal stem cells in the context of TKI treatment.” – Ravi Bhatia

Key points: 

• Stem-cell fitness: Persistence of BCR::ABL1-positive LSCs alone may not determine relapse; their functional fitness relative to residual normal HSCs may also influence outcomes after TKI cessation.
• TNF signalling: TNF/NF-κB signatures remain enriched during TKI treatment, with TNF signalling affecting both CML and normal stem-cell function.
• TNFRSF1A: Genetic deletion of TNFRSF1A produced different effects depending on treatment context: without TKI, TNFRSF1A-deficient CML HSCs showed increased competitive fitness.
• TKI response: During TKI treatment, however, TNFRSF1A knockout reduced the competitive fitness of CML HSCs while enhancing the competitive fitness of normal HSCs. 
• CML-driven suppression: The findings suggest that TNFRSF1A signalling within CML cells contributes to suppression of normal HSC competitive fitness during TKI therapy; deletion of the receptor in normal HSCs itself did not alter their competitive fitness.
• Future questions: Further work is needed to identify the responsible TNFRSF1A-expressing CML populations, define the mechanisms suppressing normal HSCs, determine the role of TNFRSF1B and validate the observations in human stem cells. 

Presentation 3

Early prediction of long-term molecular response in chronic myeloid leukemia using plasma drug concentration and single-cell immune and signalling profiles

Speaker: Stein-Erik Gullaksen (Bergen) 

“The end goal of all of this is to be able to provide personalised treatment optimisations to patients.” – Stein-Erik Gullaksen

Key points:

• Single-cell profiling: Deep single-cell immune and signalling profiling appears to have predictive value for long-term molecular response, supporting the possibility of identifying treatment response very early after starting a TKI.
• Pharmacokinetics: Plasma drug concentrations showed some association with earlier molecular responses, but overall pharmacokinetics appeared to have limited correlation with long-term outcomes. 
• Signalling: Treatment-induced signalling changes detected within the first days of therapy were associated with later molecular response, including changes identified across myeloid and CD34+ LSC-like populations.
• Detection: Side by side comparison of flow and mass cytometry showed a lower detection floor in the flow data. The speaker attributed this to differences in implementation, rather than to sensitivity of the platforms.
• Prediction: Combining early single-cell signalling data with predictive modelling showed potential to distinguish patients according to subsequent molecular response, supporting development of an early response-prediction strategy. 
• SWITCH: The next step is the SWITCH Phase Ia/II study, which will test whether rapid single-cell profiling and a pretrained prediction model can guide personalised TKI selection, including an early switch from imatinib to asciminib when indicated.