计算机模型提出减少儿童白血病化疗副作用的新方法
Computer models suggest ways to reduce side effects of chemotherapy in children
德国马格德堡奥托·冯·古里克大学的研究团队通过计算机模拟提出,在儿童急性髓系白血病化疗前使用G-CSF的新策略,可能减少副作用。该策略已在模拟中显示出潜在优势,但仍需临床验证。
A drug used to support blood cell production (hematopoiesis) could be more effective in a specific leukemia treatment if administered before chemotherapy rather than after it. This was the outcome of computer simulations that involved researchers from Otto von Guericke University Magdeburg. The finding resulted from a comparison of 323 treatment protocols by the research team led by mathematicians professor Dr. Sebastian Sager, Dr. Adrian Reimann and Anna Gebhard from the Institute of Mathematical Optimization.
In a second study, the team developed an approach to alleviating the significant fluctuations in key immune cells in children with leukemia within the model. Both results could help to design gentler treatments in the future but must first undergo clinical verification.
"Chemotherapy also impacts healthy cells," explains Sager. "This has particular consequences for the bone marrow, where new blood cells are produced. Slowing their production can result in a shortage of vital immune cells, meaning that infections can become life-threatening." The treatment must therefore satisfy two requirements: namely, fighting the leukemia while having as little adverse effect as possible on healthy hematopoiesis.
When does blood cell production need support?
The mathematicians' first study, published in PLOS One, examined one phase of the treatment of acute myeloid leukemia (AML), which was intended to consolidate therapy that had already been successful. The cancer drug Cytarabin was used. In addition, G-CSF growth factor can stimulate the formation of certain white blood cells.
The researchers used existing treatment data from 65 patients from Magdeburg and Ulm for their calculations. The data were processed so that individuals could not be directly identified. The data from Magdeburg were provided by the collaborating researchers, professors Dimitrios Mougiakakos and Enrico Schalk from the University Clinic for Hematology, Oncology and Cell Therapy in Magdeburg. On this basis, the researchers compared 323 different treatment procedures. In particular, they examined when and for how long G-CSF should be used.
The surprising finding was that in the computer model, the outcome of treatment protocols was more favorable when the supporting G-CSF treatment had already been completed before starting chemotherapy. It was particularly important that the treatment ended before the chemotherapy cycle.
"That was a surprise for us too," says Sager. Until now, G-CSF, if used at all, has usually been administered after chemotherapy in order to support the recovery of hematopoiesis. That being said, nobody from the patient group on which the study was based had received the newly examined prior treatment. The potential advantage it offers is therefore only a model prediction at present, which needs to be tested in clinical studies.
A trial run on a digital twin
This kind of comparison is made possible by digital twins. "They can be thought of as hematopoiesis trial runs tailored to individuals, with blood values and information on drug administration being entered into mathematical models. These calculate how certain blood cells might develop over the course of a treatment," explains Sager.
The team has been developing the model for many years. It is a major step forward that the researchers are now able to use it to conduct virtual studies and can also examine treatment protocols that have not yet been tried in the underlying patient data.
Unlike in clinical studies, the computer can be used to compare hundreds of variants under identical baseline conditions. For example, researchers can change the timing or duration of drug administration and calculate how it might affect hematopoiesis. This can be done considerably more quickly than in a clinical study and also makes it possible to examine treatment options that cannot be directly tested on humans for safety reasons.
"The digital twin represents selected processes, not a whole person," stresses Sager. This means that something that looks favorable on the computer can pave the way to a clinical trial but not replace it.
When blood cell production subsides
The second study, published in Quantitative Medicine, was concerned with the longer maintenance treatment in children with acute lymphoblastic leukemia, or ALL for short. Here the researchers looked at neutrophils, a group of white blood cells that are key to mounting a defense against infections. Their number can fluctuate significantly during treatment.
The researchers were able to show that such fluctuations can also come about through natural hematopoietic feedback mechanisms. In addition, certain treatment strategies can cause them to intensify. Phases when the number of immune cells falls significantly and the infection risk increases are especially problematic.
"In this case, there is a delay in the reaction of the hematopoietic system," explains Sager. "We might compare this to a shower whose temperature only changes a while after you turn the valve—if you only react to the water being too hot or too cold at any one time, you might adjust the temperature too drastically. In hematopoiesis too, delayed reactions and repeated interventions can intensify the fluctuations." The scientist clarifies that the analogy is a simplistic way of explaining the problem. The biological processes are considerably more complex.
The new STEADY treatment approach addresses this very issue. The acronym stands for "Stabilizing Treatment prEdicting Anc DYnamics." At the same time, the name is a play on the term "steady state." Blood values can also fluctuate as a result of this kind of condition.
Whereas at present, dosing decisions are primarily based on current blood test values, the procedure also takes into account the individual amount of time that new blood cells require for their development. It calculates in advance whether their number is likely to rise or fall in the coming two to three weeks. On this basis, dosing can be adjusted in a way that can mitigate these sharp fluctuations.
In simulations based on 452 patient treatment procedures, the new strategy led to smaller fluctuations and fewer instances of very low cell counts than the replicated comparison strategies. The treatment data used for this were provided by a hospital in Copenhagen. They were processed so that patients could not be directly identified. The data were made available to the research team by the collaborating researchers, professor Markus Metzler and Dr. Jacob Zierk from Erlangen Children's Hospital. The improvements in the model required neither a higher total drug dose nor more frequent dose adjustments.
Everyday family life shaped by two years of treatment
In Germany, according to information from the specialist "Children's Cancer Info" portal, every year between 550 and 600 children and adolescents under 18 years of age are diagnosed with ALL. The disease is especially prevalent in children between 1 and 5 years of age. Overall, the chance of long-term recovery is around 90%, but this varies between risk groups.
The treatment usually lasts for around two years and therefore consumes a large part of the early childhoods of children in affected families. It is therefore all the more important to reduce onerous side effects without putting the success of the treatment at risk.
However, the two studies have not yet shown whether the computer-calculated strategies can actually prevent severe infections or hospital stays.
The next critical step will therefore be to test the promising model predictions in clinical studies. Furthermore, technical and regulatory questions will need to be addressed before the procedures can be used regularly in hospitals. The foundations for this are already in place: the mathematical models and research software make it possible to evaluate individual blood results and derive predictions for different treatment protocols from them.
Publication details
Adrian-Manuel Reimann et al., Exploring innovative G-CSF schedules in AML cytarabine-based consolidation through a digital twin study of white blood cell recovery, PLOS One (2026). DOI: 10.1371/journal.pone.0329242
Sebastian Sager et al., From ANC Oscillations Towards an Improved Childhood ALL Maintenance Therapy, Quantitative Medicine (2026). DOI: 10.70534/qm-2026-0010
Journal information: PLoS ONE
Who's behind this story?
Lisa Lock
BA art history, MA material culture. Former museum editor, paramedic, and transplant coordinator. Editing for Science X since 2021. Full profile →
Andrew Zinin
Master's in physics with research experience. Long-time science news enthusiast. Plays key role in Science X's editorial success. Full profile →
Citation: Computer models suggest ways to reduce side effects of chemotherapy in children (2026, November 9) retrieved 9 November 2026 from https://medicalxpress.com/news/2026-11-ways-side-effects-chemotherapy-children.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
来源:Medical Xpress · medicalxpress.com