Ingmar Glauche

Chronic myeloid leukemia is controlled by tyrosine kinase inhibitors, yet only about half of well-responding patients who stop therapy remain in long-term remission, and currently it cannot be predicted who will relapse. Mechanistic models reproduce both outcomes but rely on many parameters and longitudinal immune measurements, limiting clinical use. Here we derive a minimal two-equation model in which the immune compartment is eliminated by a quasi-steady-state argument and retained implicitly as a tumor load-dependent kill rate. This reduction lowers the parameter count from fourteen to eight, requires no direct immune measurements, and preserves the two key features of immune-leukemia interaction: a switch-like per-capita kill rate and a bell-shaped immune window peaking at deep remission. A complete analytical treatment shows that a patient’s long-term outcome depends on a single quantity, the leukemic fitness, defined as the ratio of leukemic proliferation to maximal immune-mediated killing. Values above one render the cure steady state unstable, whereas values below one make it stable and therefore attainable; the rate-limiting step is depletion of the quiescent stem cell reservoir below a separatrix, reproducing the clinical link between treatment duration and remission level with durable remission after cessation. Fitted to 75 patients of the DESTINY trial, the minimal model matches the fit quality of its more detailed predecessors. Leukemic fitness estimated from time courses up to treatment cessation, including the dose-reduction period, is associated with outcome, with discrimination improving further when six months of post-cessation data are added. Guided by this single parameter, we design two personalized strategies: extended full-dose therapy for low-fitness patients who have not yet crossed the separatrix, and maintenance dosing for high-fitness patients. Leukemic fitness is thus a clinical predictor that can be estimated from routine monitoring data, and a minimally parametrized model is enough to turn patient trajectories into personalized treatment decisions. Author Summary: Most people living with chronic myeloid leukemia today take a targeted drug that keeps the disease under control, and many who respond well would like to stop treatment. Clinical trials show that about half of them stay healthy without the drug, while the other half relapse and must resume therapy. Today there is no reliable way to tell in advance who belongs to which group. To address this, we built a deliberately simple mathematical model that uses only the routine blood tests already collected during monitoring. From these data, the model weighs how aggressive the leukemia is against how well the immune system keeps it in check and combines the two into a single number we call the leukemic fitness. Using data from 75 patients in a dose-reduction trial, we show that this number separates patients who are likely to stay in remission from those likely to relapse. It also points to personalized options: some patients predicted to relapse could still reach remission if treated a little longer, while for others we can estimate a lower, tailored maintenance dose. In this way, the leukemic fitness could help doctors make better treatment decisions and improve the patient’s quality of life.

Background and Introduction Imatinib (IMA) revolutionised the treatment of chronic myeloid leukaemia (CML). Therapeutic drug monitoring (TDM) of IMA is recommended to improve treatment outcomes in adults, but guidance for reliable interpretation of plasma concentrations in children is limited. Using real-world paediatric data, we aimed to develop a practical algorithm with recommendations for an IMA TDM procedure in paediatric oncology. Methods Imatinib plasma concentrations and clinical data were collected from patients enrolled in the multicentre CML-PAED-II study and subsequent registry. We evaluated the method of exponential extrapolation for estimating interpretable IMA trough concentrations and used linear mixed effects models to assess plasma concentration predictors. Associations between IMA trough concentrations and molecular response were analysed, and a paediatric-specific target threshold was evaluated. Findings were integrated into an algorithm for clinical application of IMA TDM. Results In total, 269 IMA trough concentrations from 72 patients and BCR::ABL1/ABL1 transcript ratios from 50 patients were included. Using a correction tool, we obtained estimates of trough concentration by exponential extrapolation. Administered IMA dose was the main predictor of plasma concentrations. Higher trough concentrations tended to correlate with faster initial molecular response. The adult-derived target threshold of 1000 ng/mL proved applicable in paediatric practice. The developed algorithm outlines clinical decision points and consequences, which are key elements of existing intensified clinical pharmacological/pharmaceutical care programmes during oral antitumour therapy. Conclusion The algorithm supports the use of TDM during paediatric IMA treatment, enabling early identification of exposure below the targeted concentration and variations in treatment efficacy to further improve treatment outcomes.

Acute myeloid leukemia (AML) is a severe disorder of the blood forming system which is commonly treated with cytotoxic drugs. It is still incompletely understood how the dynamic interaction between leukemia and the stem cell-niche influences AML progression, treatment outcome and resistance formation. Mathematical models provide a framework to quantitatively investigate different assumptions about these regulations, especially with respect to the regulation of stem cell proliferation, and can help to identify potential clinical targets. To this end we structurally compare two recently published, complementary models of AML progression and treatment. Based on an analytical reformulation of the models we highlight a fundamental difference in how the models describe the regulation of proliferation. We assess the models’ suitability to describe disease dynamics and treatment outcomes by fitting each model to 275 clinical remission time courses and respective therapy cycles. Furthermore, we perform a systematic screening with univariate alterations for relevant parameters to evaluate their influence regarding velocity and depth of remission during induction therapy as well as the length of remission thereafter. The results align with the increasing consensus that therapies targeting niche-interactions mechanisms could potentially contribute to improve treatment outcomes in AML. Our derivation of a unified model integrates both aspects of proliferation regulation and provides the basis for further investigation of the extent to which these mechanisms are used. We conclude that the regulation of normal and leukemic stem cell proliferation can be coupled to niche-dependent and niche-independent mechanisms. While both assumptions allow to sufficiently describe clinically available time course data, their complementary roles in leukemia progression but also in normal and stress hematopoiesis need to be further delineated.

Hematopoietic stem cells (HSCs) maintain lifelong production of blood by balancing self-renewal and differentiation. However, certain aspects of their divisional dynamics, namely the role of quiescence and the intrinsic heterogeneity of the HSC pool, are not completely understood. High-resolution clonal tracking provides a powerful resource to investigate such dynamics as the data captures patterns of clonal persistence, dilution and late clonal emergence. Here, we apply mechanistic mathematical modeling to longitudinal clonal data from non-human primates to explore structural requirements that underlie the observed dynamical patterns. We show that models treating HSCs as a single, homogeneous population can explain the gradual loss of clonal diversity, but fail to reproduce clone size distributions and the long-term persistence of small and late-appearing clones. To address this, we propose a stochastic, two-compartment model in which HSCs transition reversibly between an actively cycling state and a quiescent, potentially niche-bound state. Compared to the simpler one-compartment model, this advanced framework provides a substantially improved fit for different metrics, consistently captures clone size distributions and explains the delayed activation and sustained coexistence of small and large clones. These results provide quantitative evidence that heterogeneity within the HSC pool, particularly the existence of a reversible quiescent state, is critical to account for clonal aspects of long-term hematopoiesis. Our findings highlight how clonal data can uncover underlying regulatory mechanisms and supports a central role for niche-mediated HSC quiescence in maintaining stable and diverse blood production over time. Highlights: Clonal data offers a robust approach to explore hematopoietic stem cell dynamics Homogeneous HSC population cannot reproduce observed long-term clonal patterns Niche-mediated quiescence improves model fit across different clonality metrics Model results suggest reversible transition between HSC quiescence and activation

The ability of metastatic cancer cells to invade distant tissues requires them to cross a variety of tissue boundaries, each posing distinct structural and biochemical challenges. In particular, the boundary between dense, extracellular matrix (ECM)‐rich tumor tissue and surrounding stromal tissue is associated with phenotypic changes in MDA‐MB‐231 breast cancer cells following transmigration, including increased invasiveness and aggressiveness. It remains unclear whether this transition arises from selective, permissive filtering of pre‐existing subpopulations, such as cancer stem cells, or an instructive response of the entire cell population. Here, by combining single‐cell migration analysis, heterogeneity analysis of cell proliferation, and computational modeling, it is demonstrated that tumor‐tissue boundaries act as instructive interfaces. Using an established 3D fibrillar collagen I matrix model of interfaces, it is shown that all cells can transmigrate the interface with no evidence of selective filtering of subpopulations. Proliferation heterogeneity remains unchanged between transmigrated and non‐transmigrated cells, further supporting an instructive mechanism. Simulations confirm that the interface instructively modulates cell behavior. These results indicate that tissue boundaries can reprogram cancer cell phenotypes, representing a potentially targetable mechanism in metastatic progression.

Background: Despite the excellent survival for chronic myeloid leukemia (CML) patients treated with tyrosine kinase inhibitors (TKI), most CML patients report poorer quality of life due to TKI side-effects. In the DESTINY trial ( NCT01804985 ), CML patients halved their TKI dose for 12 months prior to TKI cessation, with 24 months of follow-up off therapy. Enrolled patients (n=174) were in stable deep molecular remission (DMR; BCR::ABL1 ^IS IS IS burden at TKI de-escalation/cessation is predictive for recurrence. As shown previously, BCR::ABL1 kinetics during dose de-escalation are predictive for molecular recurrence after TKI stop. However, it has not been determined if BCR::ABL1 RT-qPCR in bone marrow (BM) improves risk prediction compared to PB. Objectives: We compared the predictive value of BCR::ABL1 ^IS measurements at the time of dose de-escalation between PB and BM with respect to molecular recurrence, including assessing if the inclusion of a BM measurement at time of dose de-escalation improved the predictive power of BCR::ABL1 kinetics during de-escalation. Methods: With written informed consent, PB and BM samples were collected and BCR::ABL1 ^IS was measured using RT-qPCR. Of 107 patients evaluated, 42 had molecular recurrence and 65 maintained TFR. To describe BCR::ABL1 kinetics during dose de-escalation, we applied linear regression. Estimated individual intercept and slope parameters were used within a logistic regression model to check for their predictive value for molecular recurrence. Results: The median log10[BCR::ABL1 ^IS ] at time of dose de-escalation was higher in patients who experienced recurrence in both BM and PB (BM: −2.69 for recurrence vs. −4 for TFR; PB: −2.78 for recurrence vs. −2.98 for TFR). Based on samples with detectable log10[BCR::ABL1 ^IS ], transcript levels were moderately correlated between PB and BM (Spearman’s r=0.6). Using logistic regression, log10[BCR::ABL1 ^IS ] values from either BM or PB taken at time of de-escalation were similarly predictive of TFR success at 36 months post de-escalation (BM: OR=1.8; PB: OR=2.6). Moreover, using intercept and slope as characteristics of the treatment kinetics during dose de-escalation jointly as independent, PB-based predictors (OR _intercept =13.8, OR _slope =1.5) led to a similar conclusion and allowed for classification of patients into three risk groups: low, high, unclear. We identified 54 patients (50%) with low risk, 20 patients (19%) with high risk, and 33 patients (31%) with unclear risk. Considering TFR success, patients in the low-risk group show 87% MRFS 24 months after TKI stop, whereas all high-risk patients had molecular recurrence within 16 months of trial entry. Conclusion: No difference in the predictive power of a single measurement taken prior to dose de-escalation using either PB or BM samples could be shown. However, predictive power was enhanced by determination of slope and intercept of sequential PB BCR::ABL1 samples obtained during TKI de-escalation. Overall, BM evaluation prior to treatment cessation was not more predictive than PB of TFR success, however using BCR::ABL1 kinetics in PB during TKI de-escalation enabled classification of patients for risk of molecular recurrence.