Introduction

Leukemia is the most common childhood cancer, representing 28.8% of cancer cases worldwide1. CNS complications that appear in the course of ALL treatment occur in appro−ximately 10% of patients in retrospective cohort groups2,3,4. The majority of these complications appear during the first 6 months, with 37% to 56.5% occurring in the induction phase2,3,4. and may be caused by leukemia itself and/or by cancer therapies, including chemotherapy6,7,8,9,10,11. Recent studies report that age 10 years and older, a high-risk phenotype, T-cell ALL, and CNS leukemia are associated with a higher risk of CNS toxicity3,5. Based on the current literature, the most common CNS symptoms that appear in children during or shortly after ALL treatment include seizures, disturbances of consciousness, constant headaches, hemiplegia, hallucinations, and visual impairment12,13. The most common causes of those disorders include posterior reversible encephalopathy syndrome (PRES)14,15, methotrexate-related stroke-like syndrome (SLS), cerebral venous sinus thrombosis16,17,18,19, other toxic leukoencephalopathies (LE)20, CNS bleeding, including stroke21,22., and neuroinfections2,3,23. Chemotherapy drugs known to cause CNS complications and toxicity in ALL are MTX, glucocorticosteroids, vincristine, L-asparaginase (L-ASP), and cytarabine24,25. Methotrexate-derived neurotoxicity occurs in 3–7% of ALL pediatric patients26. Methotrexate metabolites are known to inhibit ribonucleotide 5-aminoimidazole-4-carboxamide formyltransferase (ATIC), thereby promoting the release of adenosine, which, by binding to adenosine receptors, modifies neuronal responses in the CNS27. The most frequently reported CNS complications associated with the use of glucocorticoids include mental and cognitive disorders.

Approximately 20% of patients develop depression, mania, and psychosis28., which require pharmacological treatment, and 75% report psychiatric symptoms that resolve after therapy discontinuation29. The incidence of neurotoxic consequences of vincristine use is not precisely known because of variability among study groups. The most significant side effect of vincristine therapy is neurotoxicity, primarily peripheral neuropathy30. According to several studies, up to 33% of children suffer from CNS disorders during L-ASP treatment31. The mechanism of L-ASP neurotoxicity derives from the breakdown of aspartate and glutamine into aspartic and glutamic acids and increased ammonia levels32. Excessive stimulation of the N-methyl-D-aspartate (NMDA) receptor can lead to cell death in CNS neurons33. Notably, the rate of severe CNS adverse events caused by cytarabine, such as somnolence, confusion, and, more infrequently, seizures, cerebral dysfunction, and acute cerebellar syndrome, is around 12% of all patients34, possibly due to wide-ranging loss of Purkinje cells in the cerebellum35.

The study’s objective was to analyze the cause, number, and type of cerebrovascular complications developed by children with ALL hospitalized for anti-cancer treatment in the Department of Pediatric Hematology, Oncology, and Transplantology of the University Children’s Hospital in Lublin from 2014 to 2023.

Index case

A 4-year-old girl assigned to a standard risk group of pre-B-ALL common ( +) developed PRES syndrome during chemotherapy induction phase. In the treatment according to the current protocol, the patient received vincristine, daunorubicin, PEG-Lasparaginase (PEG-ASP), methotrexate, and prednisone. On the 28th day of treatment, the girl developed a seizure, facial muscle spasm near the corner of the mouth on the left side, eye turn and dilated pupils without reaction to light. Laboratory test results revealed anemia (HGB 9 g/dl vs reference range 11.5–15.5 g/dl) and thrombocytopenia (26 × 103/µL vs normal: 140–410  × 103/µL). Moreover, an elevated procalcitonin level reaching 1.11 ng/mL (normal: < 0.5 ng/mL) and a C-reactive protein (CRP) level of 6.71 mg/dL (reference range 0–0.5 mg/dL) were detected. The following abnormalities were found in the coagulogram:

PT 26 s (normal: 12–16 s), ATIII 68% (reference range 75–150%), D-dimer 959 ng/ml (normal < 500 ng/ml). The radiological report from imaging studies showed features of PRES syndrome (Figs. 1 and 2). Cerebralfluid examination revealed pleocytosis (25 cells/ul). The neurological examination showed neck stiffness, weakened tendonreflexes, and decreased muscle tone in the upper and lower limbs. On the EEG examination, abnormal recording with periods of generalized paroxysmal, sometimes exacerbated, slow delta-theta waves (2–4 Hz to 200–250 uV) against a background of significantly slowed, with features of disorganization, baseline delta-theta waves (2–4-7 Hz to 80–150 uV). The patient communicated with gestures, reported headache, had vomiting reflexes, nausea. Treatment included diazepam, mannitol, and amlodipinum. Additionally, the girl developed profound bone marrow aplasia, leukopenia, neutropenia, thrombocytopenia, toxic liver damage, polyneuropathy, and oral mucositis, so she required broad-spectrum antibiotic therapy, blood transfusion, parenteral nutrition and rehabilitation. Interruption of anticancer treatment lasted 24 days. Currently, the patient is under the constant care of a hematologist, cardiologist, nephrologist, neurologist and speech therapist, due to heart rhythm disturbances, hypertension, epilepsy and speech disorders. The patient does not show any hematooncological changes and remains in oncological remission.

Fig. 1
figure 1

LA. Bilateral cortical/subcortical areas, mainly in the parietal/occipital (predominantly left side) and temporal (predominantly right side) regions, and partly in the thalamus, with increased IS in DF and T2-weighted images. Imaging of the lesions in correlation with the clinical picture suggests changes in the course of PRES. B. On follow-up examination: partial regression of the described lesions. In the right parietal lobe, a small, approximately 4 mm in diameter, signal area not seen in the previous examination.

Full size image
Fig. 2
figure 2

CT Brain. Areas of decreased attenuation within the white matter in the medial part and posterior horn of the left lateral ventricle. Slight attenuation elevation of the cerebellar hemispheres. Ventricular system slightly dilated, symmetrical. Frontal horn width 8–9 mm.

Full size image

Results of the study

The clinical information about children with ALL and cerebrovascular complications was summarized in Table 1. Patients’ laboratory tests were summarised in Table 2. From 2014 to 2023 (10 years) 144 young patients with ALL were hospitalized in the Department of Pediatric Hematology, Oncology, and Transplantology of the University Children’s Hospital in Lublin, Poland, among whom 12,5% (n = 18 patients) developed cerebrovascular toxicities during treatment. 58.8% of patients were treated with AIEOP 2017 and 41.2% with ALLIC 2009. All patients were hospitalized at the time when their symptoms first appeared. Cerebrovascular complications were equally distributed in girls and in boys. In our analysis, there were no cases of individuals having more than one neurological incident throughout the course of therapy. The patients’ ages averaged 7.1 years and the median distribution was 6.3 years, with a range of results from 0.6 to 15.3 years. Half of the group were patients from 2.9 to 10.8 years old. Five cases, representing 29% of the group, were found to have genetic aberrations, of which four patients had ETV6::RUNX1, and one patient KMT2A::AFF1. One case, representing 5.88% of the study group, had a family history of cystic kidney disease, which was found in the patient’s sister. Chronic diseases were present in 44% of the subjects, including SIADH and mild intellectual disability (n = 2 each ), as well as epilepsy, visual impairment, and ataxia-telangiectasia syndrome (n = 1 each). Among the chronic illnesses present in subjects, no central nervous system (CNS) related diseases or hypercoagulable states were identified.B-ALL cases represented 61.11% (n = 11) of the study group, and T-ALL patients represented 38.59% (n = 7). According to the AIEOP-BFM ALL 2017 POLAND protocol, patients are classified into standard risk (SR), intermediate risk (IR), or high risk (HR) groups based on their early response to induction therapy and the presence of high-risk genetic abnormalities. In our cohort, the SR group accounted for 27.78% of the patients, the IR group included 22.22%, and the HR group comprised 38.89%. The remaining 11.11% of patients were not assigned to a risk group at the time of CNS involvement occurrence. Most of the complications (83.33%) occured during treatment induction. The most commonly indicated symptom was seizures, occurring in half the patients. Other symptoms included impaired consciousness (39% of the subjects), petechiae (22%), limb tremors (11%), headache (11%), vomiting (6%), and nausea (6%). In a group of six (33.33%) patients, an increase in inflammatory markers, i.e.C-reactive protein and procalcitonin, was observed. Three patients had elevated protein C levels, with concentrations of 77 mg/L, 82 mg/L and 90 mg/L. Ischaemic lesions (44.44%) were most frequently observed on MRI; in a group representing 40% of the patients, lesions were found on head CT, and 20% had no lesions. The results of the imaging studies are shown in Table 1. The results of laboratory tests are shown in Table 2.

Table 1 Characteristics of children with ALL and cerebrovascular abnormalities.
Full size table
Table 2 Children’s results of laboratory tests.
Full size table

In a group representing 11.11% (n = 2) of the patients, changes were found on ultrasound of the carotid vessels and in 55.56% (n = 10) no changes were found. Treatment interruption time averaged 14.0 days and the median was 18.5 days, with a range of 0 to 30 days. No significant (p < 0.05) associations were found between gender and the presence of individual symptoms. Neither any significant (p < 0.05) associations were observed between risk groups and the incidence of individual symptoms.

There were no significant (p < 0.05) associations between risk groups and radiological changes. Nor any significant (p < 0.05) associations were found between the diagnosis and the occurrence of individual symptoms.

Discussion

Cerebrovascular complications during treatment of acute lymphoblastic leukemia (ALL) pose significant clinical challenges, often requiring modifications to chemotherapy protocols that may directly affect patient outcomes. In our 10-year retrospective analysis, 12.5% of pediatric ALL patients exhibited clinical signs of cerebrovascular involvement, a figure consistent with the reported incidence range of 3–18.4%. These complications frequently necessitate interruption or alteration of therapy, highlighting the need for adaptive strategies that balance toxicity management with the preservation of treatment efficacy The clinical relevance of these complications stems from their potential to disrupt curative therapy. Despite the protective role of the blood–brain barrier, leukemic cells can infiltrate the central nervous system (CNS) by accumulating at microvascular branching points, leading to cerebrovascular lesions. This pathophysiological mechanism underlies many of the neurological sequelae observed during ALL therapy36,37,38,39,40,41,42. Methotrexate-induced leukoencephalopathy—a well-documented neurotoxic effect—may require replacement with intrathecal cytarabine to limit CNS damage while maintaining antileukemic efficacy43,44,45,46,47. Posterior reversible encephalopathy syndrome (PRES), often associated with agents such as methotrexate, L-asparaginase,

vincristine, and corticosteroids48,49,50,51,52, may paradoxically require reintroduction of corticosteroids to control vasogenic edema and associated neurological symptoms52,53,54,55,56.

Cerebral thrombosis linked to L-asparaginase—reported in 2.1–15% of cases, depending on the study8,57,58,59,60—presents a particular therapeutic dilemma. Although thrombosis frequently necessitates temporary cessation of asparaginase, withholding this agent has been correlated with worse outcomes, particularly in adolescents and young adults (AYA)61. In such cases, substitution with alternative formulations, such as Erwinia chrysanthemi–derived asparaginase, offers a viable approach that allows therapy to continue without compromising efficacy. These examples underscore the importance of maintaining dose intensity while managing toxicity to optimise survival. In our cohort, glucocorticosteroids, vincristine, and methotrexate were the chemotherapeutics most commonly associated with CNS toxicity, consistent with previous literature24,62,63,64,65,66. Methotrexate toxicity often presents with stroke-like symptoms such as hemiparesis or altered consciousness, typically within 13 days post-administration67,68,69. Vincristine, known for inducing neuropathy involving peripheral, autonomic, and occasionally central nervous systems, can necessitate dose reductions that may impair treatment continuity and quality of life. Unfortunately, no reliable clinical predictors of vincristine-induced neuropathy have been identified thus far. L-asparaginase–related CNS complications manifest in various forms, including cerebral hemorrhage (2.1%), thrombosis (15%), and thromboembolic events (4.2%). Sinus thrombosis is most frequently observed during remission induction, although it can occur at later treatment stages. Prolonged seizures have also been documented as a presenting symptom of asparaginase-associated cerebral complications. PRES occurs more commonly in ALL patients (55%) than in those with AML (9%) and is typically characterised by seizures and altered mental status64,65,66. These manifestations highlight the importance of vigilant neurological monitoring during high-risk phases of treatment.

Cerebrovascular complications in patients treated for acute lymphoblastic leukemia (ALL) result from a complex interplay of cancer-associated hypercoagulability, treatment-induced vascular injury, and hematological abnormalities such as hyperleukocytosis and thrombocytopenia. Acute management involves rapid correction of hematological parameters, including platelet transfusions and coagulation factor replacement, to reduce the risk of hemorrhagic and ischemic events. However, anticoagulant or antiplatelet therapies must be used cautiously due to bleeding risks inherent in thrombocytopenic patients. Monitoring minimal residual disease (MRD) is essential not only for relapse risk assessment but also for guiding therapy intensity. Persistent MRD often necessitates intensified chemotherapy, which may prolong exposure to neurotoxic and vasculotoxic agents, thereby indirectly increasing cerebrovascular risk. Therefore, strategies aimed at rapid MRD eradication could reduce cerebrovascular complications by shortening intensive treatment duration. Advances in radiotherapy, such as intensity-modulated radiotherapy (IMRT), allow precise targeting of leukemic sites while sparing healthy brain tissue and vasculature, thus mitigating late radiation-induced vasculopathy characterized by endothelial damage and accelerated atherosclerosis Additionally, hematopoietic stem cell transplantation (HSCT) and hematopoietic growth factors (HGFs) influence endothelial function and blood counts, potentially increasing cerebrovascular risks, which necessitates vigilant neurological monitoring during and after transplantation. There is potential to explore endothelial-protective agents and optimised management of comorbidities to preserve vascular health in ALL patients. Pharmacogenetic profiling may further enable personalised treatment adjustments to minimise neurotoxicity. An integrated, multidisciplinary approach combining hematological monitoring, MRD-guided therapy, advanced radiotherapy techniques, and neuroprotective strategies is critical to improving outcomes and quality of life in ALL patients at risk of cerebrovascular complications5,70,71,72,73. Although computed tomography (CT) remains the first-line imaging modality due to its accessibility, its limited sensitivity, missing up to 75% of hemorrhages—demonstrates the superiority of magnetic resonance imaging (MRI) in detecting ischemic changes and subtle lesions74,75,76,77,78. Age is a known modifier of risk, with children over 10 years more susceptible to CNS toxicity. Notably, however, 72.2% of affected patients in our study were aged 10 years. B-cell ALL (B-ALL) predominated (61.11%), in contrast to findings from some previous reports. While genetic alterations such as the ETV6::RUNX1 fusion were identified, no specific mutations definitively linked to increased CNS toxicity were observed79,80,81. Approximately 80% of acute CNS complications resolve with appropriate intervention, yet mortality remains around 15%. Survivors may experience long-term neurocognitive deficits, including reduced processing speed and impaired attention, which are thought to correlate with caudate nucleus volume depletion82,83. These findings support the need for early and continuous neuropsychological assessment throughout treatment.

This study has several limitations, including its retrospective, single-centre design and relatively small sample size, which may limit the generalizability of the findings. Additionally, the often nonspecific clinical presentation of CNS complications necessitated a strong reliance on neuroimaging for diagnosis. Future multicenter, prospective studies incorporating standardised imaging protocols and biomarker evaluation are needed to better stratify risk and refine therapeutic approaches.

In conclusion, cerebrovascular complications during ALL treatment significantly affect therapy continuation and overall prognosis. Timely recognition and multidisciplinary management are essential. Therapeutic adaptations—such as replacing intrathecal methotrexate with cytarabine in cases of leukoencephalopathy, reintroducing corticosteroids for PRES management, and utilising alternative asparaginase formulations in thrombotic events—can effectively mitigate toxicity while preserving therapeutic efficacy. Close collaboration between hematologists, neurologists, and radiologists is critical to optimising care and improving patient outcomes.

Methods

The study was conducted as a retrospective single-center study. The study group included patients from 0 to 18 years old with acute lymphoblastic leukemia hospitalized in the Department of Pediatric Hematology, Oncology, and Transplantology of the University Children’s Hospital in Lublin. Due to retrospective nature of the study, informed consent was waived by the Bioethics Committee of the Medical University of Lublin (KE-0254/208/10/2022). Statistical analyses were carried out using the statistical 1.8.4.152 PQStat package version. Correlation coefficients of biochemical parameters between the sexes were estimated by means of Pearson’s linear correlation coefficients and Spearman’s monotonic correlation coefficients. The relationship between the sex and symptoms was analyzed using Fisher’s exact test. The association of risk groups with symptoms and radiological changes was analyzed with Fisher’s exact test. Also the association of diagnosis with symptoms and treatment protocol was analyzed with the use of Fisher’s exact test. Similarly, Fisher’s exact test was used to analyze the association of the treatment protocol with symptoms and radiological changes . A test probability p < 0.05 was considered significant and it was judged as highly significant at p < 0.01. This study obtained approval of the Bioethics Committee of the Medical University of Lublin, KE-0254/208/10/2022. In this study, diagnostic methods were applied in accordance with the current Polish standards for the diagnostic management of acute lymphoblastic leukemia in children, as recommended by the Polish Society of Pediatric Oncology and Hematology. Genetic diagnostics primarily relied on conventional cytogenetic analysis, which allowed for the detection of chromosomal abnormalities. Additionally, molecular techniques such as polymerase chain reaction (PCR) and fluorescence in situ hybridization (FISH) were employed to identify characteristic translocations and genetic mutations relevant for prognosis and treatment decisions. All testing was performed following national guidelines to ensure consistency, reliability, and appropriate clinical interpretation of the results.