1 Introduction
The most common malignant tumor in childhood, acute lymphoblastic leukemia (ALL) is the abnormal proliferation of primary and juvenile lymphocytes in the bone marrow, inhibits normal hematopoiesis in the bone marrow, and can extensively infiltrate organs, such as the liver, spleen, and lymph nodes. Over the past few decades, the 5-year survival rates of patients with ALL have been over 90% due to increased understanding of the pathogenesis of ALL and optimized chemotherapy regimens [
1]. The poor prognosis of ALL in children is not only related to the malignancy itself but also to the complications of chemotherapy toxicity during treatment.
Venous thromboembolism (VTE), which comprises deep venous thrombosis and pulmonary embolism (PE), is a severe and common complication in pediatric ALL. VTE is affected by many factors, and its formation is closely related to the presence of a tumor. Moreover, VTE is an aggressive and therapeutic clinical marker of tumors and is related to increased mortality [
2], may have a negative impact on scheduled ALL chemotherapy, and cause poor prognosis in children. However, no clinical report on a large sample regarding the incidence, outcome, and prognosis of children with ALL with VTE has been published.
This study aimed to determine the prevalence, clinical features, and outcome of VTE in a pediatric ALL population. In addition, we explored the risk factors for VTE development in a large cohort of children with ALL.
2 Materials and methods
2.1 Study population
This study included 7640 children (aged < 18 years old; 4521 males (59.18%) and 3119 females (40.82%)) with newly diagnosed ALL from 20 hospitals in China between January 2015 and December 2019. Mature B-ALL, down syndrome, mixed phenotype leukemia, second tumor, congenital immunodeficiency disease, and metabolic disease were excluded. All patients were treated according to the Chinese Children’s Cancer Group-ALL-2015 (CCCG-ALL-2015) regimen.
2.2 ALL diagnosis and risk classification
Patients were diagnosed with ALL according to morphology combined with immunology, cytogenetics, and molecular biology criteria. Immunophenotyping and minimal residual disease were measured by flow cytometry. Reverse-transcription polymerase chain reaction was performed for the detection of fusion genes, such as BCR/ABL1.
Patients with ALL were stratified (Table S1) into three risk groups after induction treatment: low risk, intermediate risk, and high risk. All patients received risk-stratified and minimal residual disease-directed therapy.
2.3 Chemotherapy treatment
The CCCG-ALL-2015 protocol is based on St. Jude Children’s Research Hospital Total XV study with slight modification [
3]. It comprises window, induction (weeks 1–7), consolidation (weeks 8–15), maintenance I (weeks 16–34), and maintenance II chemotherapy (weeks 35–126). All patients received the same window treatment containing dexamethasone (6 mg/m
2/day, 4 days).
2.4 Study design
The CCCG-ALL-2015 study is a multicenter clinical study involving 20 hospitals in China. The study prospectively collected information on ALL patients, including demographics, ALL-related data, chemotherapy effects, laboratory data, treatment complications and prognosis. This study was approved by the ethical review committee of each participating center. All written informed consent was obtained from a parent or guardian for participants aged < 18 years.
VTE cases were reported as adverse events. VTE diagnosed at every center must be reported to the CCCG-ALL-2015 registry system. VTE was reported in the CCCG-ALL-2015 database from date of ALL diagnosis to September 2020. The median follow-up time in our research was 3.3 years (range: 0.8–5.8 years). At the follow-up deadline, we checked again case by case for VTE diagnosis based on imaging, coagulation indicators, and clinical features. The CCCG-ALL-2015 study clearly stipulates that clinical diagnosis of thrombosis associated with drugs or catheterization needs to be registered as an adverse reaction. Detailed clinical data were collected retrospectively for each patient from participating hospitals, including demographics, diagnosis date, chemotherapy phase, treatment, laboratory data, and outcome. VTE treatment includes observation, anticoagulation, and catheter removal, but no treatment suggestions are clearly specified, so each central treatment varies. Normal coagulation function results were as follows: prothrombin time, 12–14.5 s; activated partial thromboplastin time, 32–45 s; fibrinogen degradation product, < 5 μg/mL; D-dimer, < 0.5 μg/mL; antithrombin-III, 80%–120%.
We regularly cared for and checked the intravenous catheters. Children with ALL who had a venous catheter were routinely screened for thrombosis before catheter placement, every 3 months after catheterization, and before catheter removal by ultrasonography, including limb vein ultrasound and cerebral vein Doppler ultrasound, or by cranial magnetic resonance imaging (MRI). When cerebral vein Doppler ultrasound showed abnormalities, computed tomography (CT) or MRI was performed for further examination. Children with ALL and without a catheter were checked for thrombosis every 3 months after ALL diagnosis by ultrasonography. In addition, imaging was performed immediately when patients developed thrombus-related symptoms. When thrombus-related symptoms appeared after the initial thrombosis recovered, imaging examination was conducted to check whether the thrombus had recurred.
VTE was defined as a deep venous thrombosis in the upper extremities (axillary, subclavian, or jugular veins) or lower extremities (femoral, popliteal, superficial femoral, or iliac veins) or a thrombosis in the pulmonary vasculature. Doppler ultrasonography, venography, pulmonary ventilation-perfusion scan, CT or CT angiography, and MRI or MR angiography were used in diagnosing VTE. VTE was considered catheter related when it occurred in the vein in which the catheter was inserted. When any symptom of VTE were observed in patients despite that imaging had not been performed based on those symptoms, VTE was classified as “symptomatic” [
4]. Residual thrombosis was defined as an undissolved thrombus remaining after 3 months of conventional anticoagulant therapy.
Septicemia was defined in patients with positive blood culture or septic shock. It was also reported as an adverse event, which must be reported to the CCCG-ALL-2015 registry system from every center when patients were diagnosed.
2.5 Statistical analysis
SPSS software version 23.0 (IBM Corporation, Armonk, USA) was used for all statistical analyses. Categorical data were expressed as percentages. Gender, age, immunophenotyping, risk groups, mediastinal mass, and BCR/ABL were tested for any difference in the incidence of VTE. Between-group differences were evaluated using the chi-square test or Fisher’s exact test. Continuous data were expressed as mean ± standard deviation or median (Q1, Q3). The t test or Mann–Whitney U test was used to compare groups. A univariate analysis was performed to analyze risk factors for VTE, and multiple regression analysis was performed for those with significance. Results were presented as odds ratios (ORs) with 95% confidence intervals (CIs). P values < 0.05 were considered significant.
3 Results
3.1 Prevalence of VTE in children with ALL
A total of 159 VTE events were reported in 7640 children with ALL at an incidence rate of 2.08% and incidence density of 62.1 per 10 000 person-years. In our cohort, VTE occurred within 3–454 days of ALL diagnosis, with a median time to event of 44 days. Among the VTE patients, 70.44% (n = 112) and 98.74% (n = 157) developed embolism within the first 3 and 12 months after ALL diagnosis, respectively. None of these patients had previously experienced VTE.
Fig.1 presents occurrence of VTE in different chemotherapy stages.
3.2 Risk factors for VTE development
Ninety-three patients (58.49%) with VTE had received PEG-asparaginase (PEG-ASP) therapy less than 21 days before the VTE occurred. VTE was associated with corticosteroids (prednisone < 2 weeks or dexamethasone < 5 days before VTE development) in 89 patients (55.97%). Approximately 33.33% of cases (n = 53) occurred during combined PEG-ASP and corticosteroid therapy.
The clinical characteristics of the ALL and VTE patients are shown in Tab.1.
Univariate correlation analysis found that the following factors are related to VTE: age (12–18 years; P = 0.027), T-ALL (P = 0.015), and mediastinal mass (P = 0.032). Gender, clinical risk, BCR/ABL1, and WBC count were not associated with VTE.
Multiple regression analysis of the associated factors from the univariate analyses identified T-ALL as an independent risk factor for VTE (OR 1.74, 95% CI 1.08–2.80, P = 0.022; Tab.2).
Considering infant ALL is not the same as the common pediatric ALL, we further calculated the incidence of VTE in infants ( < 1 year of age) in this study and compared it with that in other age groups. A total of 134 (1.75%, 134/7640) infants had ALL in the study, and four of them developed VTE. Although the incidence of VTE in infants was not significantly different from that in patients aged 1–12 years (2.99% (4/134) vs. 1.96% (137/6977), P = 0.598) and 12–18 years (2.99% (4/134) vs. 3.4% (18/529), P = 1.0), it was quite high.
Septicemia is the most common adverse event associated with ALL treatment. After comparing the incidence of septicemia between patients with and without VTE before thrombosis development, we found that septicemia can evidently promote the occurrence of VTE (1.83% (118/6436) vs. 3.41% (41/1204), P < 0.001).
3.3 Venous catheter and VTE
Among all VTEs, 75.47% (n = 120) were catheter-related thrombosis (CRT) and 24.53% (n = 39) were non-CRT. The incidence rates of CRT and non-CRT in children with ALL were 1.57% and 0.51%, respectively. CRT patients had a higher incidence of septicemia before thrombosis than non-CRT patients (30% vs. 12.82%, P = 0.033).
A total of 155 children had venous catheters in the VTE group. We documented the types of catheters and proportions of CRT and non-CRT among different catheters in Tab.3. Owing to the relatively small number of Port, Hickman, and Broviac catheters, we only analyzed the occurrence of different types of VTE in patients with peripherally inserted central catheter (PICC) and central venous catheter (CVC). The proportion of CRT in patients with CVC (94.29%) was higher than that of PICC (71.82%, P = 0.006).
3.4 VTE location and symptoms
The location of thromboses in ALL patients is shown in Fig.2. Of the 159 VTE cases, 108 cases (67.92%) were in the upper extremities and included 107 cases (99.07%) accompanied by CRT. One patient (0.63%) had a PE associated with lower extremity deep venous thrombosis.
Of the 28 patients with cerebral thrombosis, 20 (71.43%) developed cerebral sinus venous thrombosis (CSVT) and 8 (28.57%) developed cerebral venous thrombosis. The sites of CSVT were the upper sagittal sinus (n = 10), transverse sinus (n = 8), sigmoid sinus (n = 6), straight sinus (n = 2), and torcular herophili (n = 1). Multiple venous sinus involvement was frequent (70%, 14/20).
The incidence of symptomatic VTE in ALL children in this group was 1.22% (93/7640). Approximately 58.49% (93/159) developed symptomatic VTE, and 41.51% (66/159) were asymptomatic. Symptomatic VTE usually presented one or more symptoms. The most common clinical symptom of symptomatic thrombosis in this group was extremity swelling (n = 55), and other symptoms included neurological abnormalities (n = 26), local cannulation site or systemic infection (n = 17), extremity pain, skin color changes (n = 13), and catheter malfunction (n = 2). The clinical condition of children with cerebral thrombosis was usually serious. Of the patients with cerebral thrombosis, 92.86% (n = 26) had obvious symptoms, including convulsion (n = 14), headache (n = 13), coma or confusion (n = 5), vomiting (n = 2), fatigue (n = 2 cases), and paresis (n = 1). Nine patients entered the pediatric intensive care unit for treatment. One patient with PE had obvious tachypnea and tachycardia at diagnosis. Staphylococcus aureus was found in the blood culture samples of three patients with systemic infection in symptomatic thrombi, and Staphylococcus epidermis was found in catheter exudation of one patient with skin infection.
The clinical data of patients with symptomatic and asymptomatic VTE were compared, and the clinical characteristics of patients with symptomatic VTE were found (Tab.4). Tab.4 presents major symptomatic VTE in patients aged ≥12 and < 18 years (P = 0.023), non-CRT patients (P < 0.001), and patients with cerebral thrombosis (P < 0.001).
All patients underwent a coagulation function test at the time of VTE (Table S2).
3.5 VTE treatment and recurrence
Nearly all VTE patients (93.08%, 148/159) were treated with anticoagulants, but 6.92% (11/159) did not receive any treatment and included five who had asymptomatic thrombosis. Most patients (87.16%, 129/148) were treated with low-molecular-weight heparin (LMWH), and four patients commenced unfractionated heparin (UFH). Fourteen patients were started on urokinase (UK) but continued with LMWH or UFH. One patient underwent surgery for thrombus removal. The median duration of LMWH/UFH treatment was 23 days (range: 5–312 days), and the thrombus remission rate after 3 months was 86.47% (115/133). Comprehensive treatment (UK + LMWH/UFH) was 27 days (range: 4–129 days), and the remission rate was 85.71% (12/14). The remission rate in patients who received no treatment was 81.82% (9/11). No obvious difference was observed, and 49 patients removed their catheters after treatment.
Bleeding events were reported in 6 of 148 patients (4.05%): nasal hemorrhage occurred in three patients, cerebral hemorrhage in two, and intraperitoneal hemorrhage in one. All cases of bleeding occurred during LMWH therapy within 1 month after the start of anticoagulation therapy.
Eight patients (5.03%) had VTE that recurred, and the incidence density of recurrent VTE was 2.1 per 100 person-years. In addition, we compared the VTE recurrence rates of patients with different antithrombotic therapies and found that all the patients with recurrent thrombosis were initially treated with heparin alone. Interestingly, patients who did not receive VTE therapy did not develop recurrent thrombosis. The mean recurrence time was 72 ± 35 days (range: 22–133 days).
All patients with recurrent VTE were < 9 years old and had no mediastinal mass. The median duration of initial anticoagulation therapy in patients with recurrent VTE was 7 days, which was significantly shorter than that in patients who did not have recurrent VTE (P = 0.026). Some factors associated with VTE recurrence are presented in Tab.5.
Non-ultrasound-guided venous cannulation and residual thrombus resulted in a high incidence of recurrent thrombosis in patients with VTE. Whether the catheter was removed or not had no effect on the recurrence of VTE.
3.6 VTE impact on ALL chemotherapy
Up to the follow-up deadline, 153 (96.23%) patients with VTE were alive, and 6 (3.77%) died; 13 (8.18%) had ALL that recurred. None of the deaths were attributed to VTE.
In 11 patients (6.92%), VTE occurrence led to changes in the chemotherapy scheme, such as delayed or premature cessation of ASP therapy (n = 6, 54.55%), ASP type alteration (n = 2, 18.18%), and reduction in chemotherapy drug dosage (n = 3, 27.27%). Finally, one VTE patient had relapsed ALL, but no death was attributed to VTE. Moreover, no neurological sequela was reported in patients with cerebral thrombosis after treatment.
4 Discussion
VTE is rare in a healthy pediatric population (only 0.07–0.14 per 10 000 children). However, the rate can increase 100–1000 times in hospitalized children [
5]. Tumor is one of the most common primary diseases in pediatric VTE, accounting for one-fifth of children’s thrombotic events. ALL is not only the most common tumor type diagnosed in childhood but also the most common primary malignant tumor disease in children with VTE [
6].
Significant difference in VTE incidence reported by medical centers are due to variations in patient age, study design, ALL chemotherapy protocol, and geographic ethnicity [
7]. The incidence rate of VTE in children with ALL was 1.1%–36.7% [
8]. Caruso
et al. [
8] reported that the median time of VTE occurrence is 2–3 months. The Nordic Society of Pediatric Hematology and Oncology showed that the cumulative 5-year incidence of VTE was 6.1% in 1038 children with ALL (aged 1–18 years), children aged 15–17 years had the highest incidence, and prevalence in children < 8 years old was the lowest [
9]. A report of DCOG ALL-10 study showed that 7.6% of 788 ALL children developed VTE but only included symptomatic VTE. The reported incidence of symptomatic VTE in children with ALL ranged from 3% to 14% [
2]. In our study, we identified 7640 children with ALL, the largest group reported to date; 159 (2.08%) of these children had VTE, which was within the range of VTE incidence reported previously. However, the incidence rate of VTE and symptomatic cases in our study was lower than the above published rates. We considered the existence of asymptomatic thrombi because screening was possibly incomplete and untimely. The median time to VTE development in our study was 44 days, which is shorter than that previously reported possibly because of discrepancies among the physical characteristics of patients and intensities of treatment.
The risk of VTE within the first year after ALL diagnosis is the highest and is particularly high within the initial months but declines with time after diagnosis. In a 3-year follow-up survey of 300 children with ALL treated at the University of Texas MD Anderson Cancer Center, 33% of VTE events occurred within 1 month of diagnosis and 80.6% within 12 months [
10]. Supporting these findings, our result showed that 98.74% of VTE events occurred in the first 12 months and may be related to chemotherapy drugs and tumor cell status at different times.
Thrombosis in children with ALL can be attributed to many factors [
11]. Significantly high risk in children with tumor not only is due to tumor load, vein catheters, and comorbidities but also is closely related to chemotherapy [
12]. ASP and corticosteroid administration were the primary chemotherapy protocols in our cohort. VTE frequently coincided with ASP and corticosteroid administration, and most VTEs in the ALL group were related to ASP and/or corticosteroids therapy [
13]. Most patients (56.6%) in our cohort experienced VTE during induction chemotherapy due to the frequent use of corticosteroids and ASP during the induction period. After a large number of leukemia cells were destroyed, numerous necrotic substances and procoagulant factors promoted thrombosis
in vivo. This result is in agreement with the Dutch childhood oncology group where 59.3% of VTE events occurred during induction treatment [
14].
Some analyses have demonstrated a relatively protective effect against VTE in young children compared with older children, increased levels of several natural anticoagulants, and decreased levels of some critical procoagulant factors after ASP [
14]. The present study showed that patients aged ≥12 years had a high incident rate of VTE, indicating that older age is a risk factor for VTE and confirming other studies. Giordano [
15] described children with T-ALL as having higher thrombin levels. Similarly, the T-ALL subtype appeared to be associated with VTE in our analysis. Multivariable logistic regression analysis showed that T-ALL is an independent risk factor for VTE and results in a 1.74-fold increase in risk of VTE occurrence. Increase in chemotherapy intensity and older age may have contributed to thrombosis in patients with T-ALL. The incidence was higher in patients with mediastinal masses, suggesting that VTE onset is associated with tumor load. As described above, a high incidence of VTE was observed in patients with T-ALL and mediastinal masses, and T-ALL and mediastinal masses were more common in adolescent with ALL. These results further explained VTE susceptibility in older children.
Multiple adverse events are associated with ALL treatment, and the incidence of septicemia in the CCCG-ALL-2015 study was quite high [
16]. We found a significant trend in which a large number of patients with septicemia experience VTE, which highlights the influence of inflammation in thrombosis. The association between inflammation and thrombus may be related to complement activation, increased tissue factor activity, and endogenous coagulation pathway activation after vascular endothelial cells are injured [
17,
18]. Thus, infection should be considered as an indication for VTE prophylaxis.
Our study demonstrated that the proportion of CRT in children with CVC was significantly higher than in children with PICC. Therefore, more attention should be paid to thrombosis prophylaxis care in CVC. Septicemia not only promotes thrombosis formation but also significantly promotes thrombosis development after catheter placement.
Furthermore, our study showed that most VTE events occurred in the upper extremities and were associated with indwelling catheters. Thrombosis sites are mostly similar in adult and pediatric patients, but the proportion of PE in adults is higher than that in children, and the proportion of cerebral thrombosis is higher in children than in adults [
19]. Our cohort had only one PE case. Adults have a higher number of PE because they usually have comorbidities (e.g., pulmonary disease). Children may be prone to cerebral thrombosis due to cerebral vascular dysplasia.
In our cohort, the incidence of symptomatic thrombosis was significantly higher in adolescence than in children aged < 12 years because older children are more likely to express their feelings timely and accurately. Young children who may not be able to verbalize symptoms associated with VTE and for whom clinical signs may be subtle. Thrombus-associated symptoms mostly occurred in children with non-CRT possibly because most CRT events were mural thrombi. The condition of children with cerebral thrombosis is usually severe [
9], which is in agreement with the significantly higher incidence of symptomatic VTE than in other locations.
Although no evidence-based studies have been performed for VTE therapy in children with hematological malignancies, treatment of VTE primarily involves therapeutic anticoagulation with LMWH and thrombolysis with UK [
20]. Our major patients had received LMWH or UK + LMWH therapy, but we did not find any prominent advantages. Most patients in this group had thrombosis of CRT and upper limb, and the median treatment duration for VTE was short in a group of patients possibly due to location and good response to thrombosis treatment. The incidence of bleeding associated with anticoagulation varies among patient populations and among studies but is estimated at 0%–5% [
21], and 4.08% of all the VTE events in our cohort resulted in bleeding shortly after the administration of LMWH therapy. These observations highlight the need to pay attention to the safety of LMWH use in pediatric populations.
In some cases, such as VTE associated with malignancy, recurrence risk is sufficiently high [
12]. Premature cessation of anticoagulation is associated with an unacceptably high risk of recurrent VTE. However, continuing anticoagulation is associated with increased major bleeding events. Consequently, the anticoagulation treatment period should be considered carefully. We found strong associations among residual thrombus, venous cannulation technology, and VTE recurrence. Residual thrombosis is an important cause of VTE recurrence [
22]. The presence of residual thrombosis causes persistent damage to the vascular endothelium, which causes secondary thrombus formation. In general, when the medical staff finds residual thrombosis in a patient after anticoagulant treatment, exploration and analysis of the specific situation is conducted for the formulation of a treatment plan in line with the patient’s current situation and prevention of thrombus recurrence. Repeated venous puncture can aggravate vascular damage and promote the recurrence of VTE. Intravenous catheterization complemented with imaging reduces the risk of VTE recurrence. Whether or not a catheter is removed does not alter the risk of recurrent thrombosis and can thus remain after VTE therapy if children with chemotherapy need it.
ASP is an essential component of pediatric ALL therapy. Premature cessation of ASP not only decreases the remission rate of ALL but also results in leukemia prone to relapse [
7]. Although VTE influenced the course of ASP chemotherapy in some patients, and we found that VTE did not significantly affect ALL outcomes. We can prolong the follow-up time to further explore the role of VTE in the prognosis of children with ALL in China and investigate patients’ survival state.
With a long follow-up period, this study has the largest sample size of children with ALL to date. The results of this study can reflect the incidence of VTE in children with ALL in China and establish a database of thrombosis in these children. However, our study still has some limitations. For example, the under-reporting of adverse reactions and the presence of asymptomatic thrombus may lead to a low incidence of VTE, and the actual incidence of VTE in Chinese children with VTE may be higher.
By leveraging our findings, we can assess the risk of VTE occurrence and recurrence in children with ALL and optimize prophylactic strategies to reduce VTE incidence in the pediatric population.
PDF
(576KB)
