Elevated VWF with normal ADAMTS13 in pediatric sickle cell disease: links to crises and ischemic stroke

Key Points

• •Elevated VWF with normal ADAMTS13 levels observed in VOC and ischemic stroke in pediatric SCD, accompanied with inflammation and hemolysis.
• •Hematological and endothelial markers may serve as alternatives to imaging for screening and detecting high-risk SCD complications.

Visual Abstract

Abstract

Sickle cell disease (SCD) remains a significant health care challenge associated with high burden or morbidity and mortality in sub-Saharan Africa, often resulting from vaso-occlusion and ischemic stroke. In under-resourced communities, inaccessibility of magnetic resonance imaging and computed tomography scanning complicates screening; therefore, there is a need for hematological assays as accessible cost-effective alternatives to bridge the gap. This cross-sectional study involving 80 pediatric patients with SCD and 32 healthy controls focused on the von Willebrand factor (VWF)–a disintegrin and metalloprotease with thrombospondin type 1 motif, member 13 (ADAMTS13) axis, considered to be involved in vaso-occlusion and ischemia in SCD. Using hematological assays, the study evaluated full blood count (FBC), VWF antigen (VWF:Ag), VWF activity (VWF:Act), and ADAMTS13 antigen levels in patients with SCD and healthy controls. The study revealed elevated neutrophil counts coupled with reduced red blood cell (RBC) counts, lower hemoglobin levels, elevated nucleated RBCs, and lower platelet counts in pediatric SCD phenotypes compared with those in healthy controls. VOC was associated with elevated VWF:Ag, normal VWF:Act, and normal ADAMTS13 antigen, whereas ischemic stroke cases showed no significant differences in VWF:Ag, VWF:Act, and ADAMTS13 antigen levels compared with those in the healthy controls. Compared with healthy controls, patients with SCD in steady state had elevated ADAMTS13 antigen levels. These findings indicate chronic inflammation and hemolysis, likely sustained by neutrophil-linked proinflammatory mediators, might be involved in the dysregulation of the VWF-ADAMTS13 axis, increasing the risk of vaso-occlusion. Therefore, FBC, VWF:Ag, VWF:Act, and ADAMTS13 antigen may be used for risk stratification and the early diagnosis of pediatric SCD complications.

Introduction

Sickle cell disease (SCD) is a major pediatric health problem in sub-Saharan Africa, with epidemiological studies identifying it as one of the leading causes of premature deaths.¹ Despite decades of research on this monogenic disorder linked to missense Glu6Val mutation in the β-globin gene, ∼50% to 90% of the affected children in sub-Saharan Africa are likely to die before their fifth birthday mostly because of undiagnosed and mismanaged cases.² In particular, vaso-occlusive crisis (VOC), ischemic stroke, and recurrent bacterial infections have emerged as leading complications in pediatric SCD, causing morbidity and mortality.^3,4 Therefore, the prevention and early management of these complications is key to improving prognosis and reversing the worrying trends of premature pediatric deaths in SCD.

However, there are persistent gaps in diagnosis of this disorder and its complications in less-resourced countries. For instance, in Kenya, regional variations in disease prevalence are evident, with the Western and Coast regions most affected, although access to newborn screening remains limited in these high-prevalence regions.⁵ Consequently, many SCD cases remain undiagnosed, whereas the diagnosed cases face stigma and inadequate care, particularly during crises and painful episodes. Moreover, misdiagnoses prevent affected children from receiving life-sustaining treatments including hydroxyurea therapy and penicillin V prescriptions that could protect them from complications.^3,5

The burden of managing both diagnosed and undiagnosed SCD cases in less-resourced countries and particularly Kenya lies in the under-resourced county and subcounty hospitals.^6,7 However, these facilities are underequipped to screen for both SCD and its complications. Currently, the diagnostic and screening technologies including echocardiogram, carotid ultrasound, magnetic resonance imaging, and computed tomography scans are expensive and inaccessible to the poor.⁸ In addition, personnel shortages hinder the adoption of imaging technologies for SCD management in resource-constrained rural hospitals. This increases the likelihood of the children with SCD dying from vaso-occlusion and ischemia, further pointing to the need for alternative screening and diagnostic tools that are more accessible and easier to implement in under-resourced hospitals.

The potential hematological markers that could be studied as accessible screening tools for SCD complications are a disintegrin and metalloprotease with thrombospondin type 1 motif, member 13 (ADAMTS-13) and von Willebrand factor (VWF). Under normal physiological conditions, ADAMTS13 is key in maintaining vascular patency through the targeted cleavage of VWF multimers.9, 10, 11 The binding of ADAMTS13 at the A2 domain of VWF prevents the formation of the ultralarge VWF that binds platelets at A1 to form the platelet plug.^12,13 Therefore, it is likely that the VWF-ADAMTS13 axis plays a key role in the pathophysiology of VOC and ischemic stroke. This study explores the mechanistic foundation of the VWF-ADAMTS13 axis in SCD complications and notes its potential diagnostic and screening value in the management of pediatric SCD. It was hypothesized that dysfunction in the VWF-ADAMTS13 axis may contribute to VOC and ischemic stroke in children with SCD. Although previous studies have documented the role of ADAMTS13 in the proteolytic prevention of VWF multimerization, the role of this axis in pediatric SCD complications remains poorly understood, necessitating further research on its prognostic value in resource-limited settings.^12,13

Patients and methods

This cross-sectional observational study was conducted at Kenyatta National Hospital (KNH) between January 2023 and June 2024, including 112 participants aged 2 to 14 years. Of the participants, 32 had SCD in steady state, 24 had SCD with VOC, 24 had SCD with ischemic stroke, and 32 were healthy controls, with the required sample size calculated based on the formula n = (z² × p × [1 − p])/d².¹⁴ According to the formula, n was the required sample size, z is the standard normal deviate (1.96), p is the prevalence of SCD in the population (3.2%), d is the margin of error (5%), and 1 − p represents the population without SCD (95.5%), giving a minimum sample size of 24 patients with SCD.¹⁵ Although the sample size used in this study was statistically adequate, it might be necessary to rely on larger sample sizes in future studies to validate differences between subgroups.

The inclusion of participants into the study groups was subject to having confirmed SCD diagnosis, receiving care at KNH, and parental consent and patient assent. Patients with SCD with known hematological disorders, comorbidities, and immunosuppressed status and those that had received transfusion within 3 months preceding the date of recruitment were excluded from participation.16, 17, 18, 19 The patients included in the study were classified into 1 of 3 SCD subgroups (steady state, VOC, or ischemic stroke) based on their diagnoses during inpatient or outpatient visits. In the hospitalized pediatric patients with SCD, cases of VOC were defined by acute painful episodes in the extremities, back, chest, and abdomen occurring in the absence of identifiable alternative etiologies and consistent with documented SCD-related pain episodes.²⁰ Ischemic stroke cases were defined by the sudden onset of focal neurological deficit persisting for >24 hours and confirmed by neuroimaging.⁴ These classifications were overseen by the resident hemato-oncologist.

Subject to the ethical guidelines of the University of Nairobi-Kenyatta National Hospital Ethics Review Committee (UoN-KNH ERC No. P569/06/2022), 4.0 mL of venous blood was collected from the participants and divided into 2 equal volumes of 2.0 mL in EDTA for full blood count (FBC) and 2.0 mL in sodium citrate for VWF and ADAMTS13 assays.²¹ The assay for the full hemogram was conducted immediately at the KNH hematology laboratory, whereas the citrated samples were centrifuged at 2000g for 15 minutes, and the supernatant was aliquoted and stored at −40.0°C.²¹ Assays for VWF antigen and activity were completed using respective HemosIL reagents on ACL Elite Pro, and assays for ADAMTS13 antigen were completed using Proteintech’s sandwich enzyme-linked immunosorbent assay kit as per the manufacturer’s guidelines. The findings from the assays across the 4 phenotypes (healthy controls, steady state, VOC, and ischemic stroke) were noted.

From the FBC results, the findings on white blood cell counts, neutrophil counts, and lymphocyte counts were noted, and neutrophil-to-lymphocyte ratio (NLR) calculated. In addition, red blood cell (RBC) count, hemoglobin (Hgb), hematocrit (HCT), mean corpuscular Hgb, mean corpuscular Hgb concentration, mean corpuscular volume (MCV), and nucleated RBC (NRBC) count were also noted. Normality tests were conducted on the assay findings, with all variables showing skewed distribution and high kurtosis; therefore, nonparametric tests were applied. The demographic characteristics of the study population were summarized, with categorical variables presented as percentages and continuous variables presented as medians and interquartile ranges. Using IBM SPSS Statistics version 29.0 and Stata MP version 15.0, the results from FBC, VWF antigen, VWF activity, and ADAMTS13 antigen were compared between the healthy controls and SCD phenotypes. The statistical significance of observed differences was determined using the Kruskal-Wallis H test and post hoc Mann-Whitney U test with Bonferroni adjustments. For additional details on the methodologies and findings, please see the supplemental Material.

Results

The study included 80 pediatric patients with SCD and 32 healthy controls. The median ages for the study participants were 6 years for the healthy controls, 6 years for those with SCD in steady state, 4 years for those with SCD with VOC, and 4 years for those with SCD with ischemic stroke. Of the children in the healthy control group, 58.3% were males and 41.7% were females. In the SCD groups, of those who had SCD in steady state, 36.0% were male and 64.0% were female; of those who had SCD with VOC, 53.3% were male and 46.7% were female; and of those who had SCD with ischemic stroke, 50% were male and 50% were female. Clinic attendance and hydroxyurea adherence varied across SCD phenotypes. Pediatric SCD cases in steady state were associated with higher clinical attendance, with a median of 3 scheduled clinic visits attended against the median of 2 clinic visits attended for both SCD with VOC and SCD with ischemic stroke. Similarly, the population with pediatric SCD in steady state had higher hydroxyurea prescription adherence than the populations with SCD with VOC and ischemic stroke (Table 1).

Table 1.

Patient demographics and clinical characteristics

Demographic characteristics	Healthy controls	Steady-state SCD	SCD with VOC	SCD with ischemic stroke
Sample size	32	32	24	24
Age, y
Median	6.00	6.00	4.00	4.00
Mean ± SD	5.90 ± 3.32	6.20 ± 3.38	3.80 ±1.32	5.44 ± 3.35
Sex, %
Males	58.3	36.00	53.30	50
Females	41.7	64.00	46.70	50
Area of residence, %
Rural	43.75	25.00	37.50	43.75
Urban	56.25	75.00	62.50	56.25
Income group, median∗	2.00	2.00	1.00	1.00
Clinic attendance, median†	0.00	3.00	2.00	2.00
Hydroxyurea prescription adherence, median‡	No usage	1.00	2.00	2.00

SD, standard deviation.

^∗Income group categories included low (1) and middle (2) income households.

^†Number of clinic visits with at least 4 routine clinic visits scheduled annually, once every quarter.

^‡Hydroxyurea prescription adherence was classified as always (1), sometimes (2), and rarely (3) adherent.

The analysis of the hematological and endothelial markers in healthy controls and SCD phenotypes revealed marked disparities consistent with SCD pathophysiology and the severity of the condition. For instance, all SCD groups reported significantly lower RBC, Hgb, and packed cell volume (PCV), with pediatric SCD patients with VOC being most affected, followed closely by SCD with ischemic stroke (Table 2). In steady state, RBC, Hgb, and PCV were equally lower than those in healthy controls but significantly higher than those in SCD with VOC and ischemic stroke. The red cell morphological indices, including mean corpuscular Hgb and mean corpuscular Hgb concentration, were significantly higher in SCD in steady state than in healthy controls (Table 2). In contrast, these red cell morphology indices were not significantly different in VOC and ischemic stroke compared with those in the healthy controls. In contrast, MCV was highest in the steady-state group but much lower in the VOC and ischemic stroke groups, whereas red cell distribution width–standard deviation (RDW-SD) was significantly elevated in the SCD with VOC group (Table 2). Finally, the analysis also confirmed the increased release of the premature RBCs by the bone marrow, with significantly higher percentages of NRBCs in SCD phenotypes than in healthy controls. However, the number of NRBCs was significantly lower in SCD with steady state and ischemic stroke than in SCD with VOC (Table 2).

Table 2.

Summary of findings on selected hematological and endothelial markers in healthy controls and SCD phenotypes

	Healthy controls, median (Q1-Q3)	Steady state, median (Q1-Q3)	VOC, median (Q1-Q3)	Ischemic stroke, median (Q1-Q3)
RBC, ×109/L	4720 (4430-5090)	2960∗ (2660-3320)	2820† (2540-3320)	3120‡ (2560-3480)
Hgb, g/dL	11.95 (10.15-13.25)	8.90∗ (8.20-10.15)	8.30† (7.00-8.50)	8.50‡ (7.85-9.90)
PCV, %	36.3 (32.75-40.10)	27.50∗ (25.05-30.20)	24.90† (20.70-26.00)	25.95‡ (23.70-28.80)
MCH, pg	25.2 (21.33-26.78)	30.40∗ (27.70-31.20)	26.00 (21.40-29.90)	27.30‡ (25.85-31.65)
MCHC, g/dL	32.3 (30.73-32.78)	32.60∗ (31.90-34.45)	32.60 (30.40-33.80)	33.95‡ (32.75-35.05)
MCV, fL	74.1 (67.78-81.53)	90.40∗ (78.20-95.70)	83.00 (68.5-91.00)	81.70‡ (76.35-94.18)
RDW-CV, %	17.7 (15.15-21.78)	18.40 (15.80-19.70)	21.50§ (19.80-25.90)	20.00 (18.45-22.00)
NRBC count, ×109/L	0.00 (0.00-0.00)	0.05∗ (0.00-0.12)	2.42†,§ (0.62-12.93)	0.18‡,|| (0.04-0.67)
PLT, ×109/L	427.50 (349.25-555.75)	273.00∗ (155.00-429.50)	384.00 (321.00-465.00)	305.50 (212.25-391.00)
VWF Ag, %	88.65 (46.03-105.25)	100 (77.55-111.50)	125.00† (111.50-133.00)	125 (68.00-132.00)
VWF Act, %	96.65 (82.83-133.75)	122 (113.50-157.00)	138.5 (96.10-179.00)	118 (72.75-158.50)
ADAMTS13 Ag, ng/mL	299.24 (224.41-396.36)	407.98∗ (382.90-478.79)	354.8 (280.76-497.03)	369.93 (324.80-419.71)
VWF Ag:VWF Act:ADAMTS13	0.0025 (0.0019-0.0035)	0.0019∗ (0.0013-0.0021)	0.0027 (0.0021-0.0035)	0.0025¶ (0.0020-0.0030)

Q1-Q3, interquartile range; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; PLT, platelet count; RDW-CV, red cell distribution width–coefficient of variation; Act, activity; Ag, antigen.

^∗Significant difference between healthy controls and SCD in steady state.

^†Significant difference between healthy controls and SCD with VOC.

^‡Significant difference between healthy controls and SCD with ischemic stroke.

^§Significant difference between SCD in steady state and SCD with VOC.

^‖Significant difference between SCD with VOC and SCD with ischemic stroke.

^¶Significant difference between SCD in steady state and SCD with ischemic stroke.

A scatterplot matrix for VWF antigen, VWF activity, and ADAMTS13 antigen revealed relatively skewed distribution across all the 3 variables (Figure 1), consistent with the findings of the Kolmogorov-Smirnoff test for normality. The matrix showed a strong positive correlation between VWF antigen and activity, with higher antigen levels corresponding to higher activity. However, the findings showed no correlation between ADAMTS13 antigen and either VWF antigen or VWF activity (Figure 1).

Figure 1.

Scatterplot matrix showing the distribution and pairwise relationships between VWF antigen, VWF activity, and ADAMTS13 antigen levels in the study cohort. The diagonal panels show histograms of individual variables, whereas the off-diagonal panels present bivariate scatterplots illustrating potential correlations. Act, activity; Ag, antigen.

The Kruskal-Wallis H test showed significant differences for white blood cell counts, neutrophil counts, and NLR, but the difference in lymphocyte counts between SCD phenotypes and healthy controls were insignificant (Figure 2). Neutrophil counts were elevated in SCD with VOC and ischemic stroke but remained within the normal range in steady state. Similarly, the Mann-Whitney U tests revealed significantly elevated NLRs in VOC and ischemic stroke compared to those in healthy controls and steady state (Figure 2). In contrast, no significant variations in lymphocyte counts were reported across any of the subgroups.

Figure 2.

Box and whisker plots of selected WBC differentials in healthy controls and SCD phenotypes. Significantly elevated WBC counts were recorded in vaso-occlusive crisis and ischemic stroke, with significantly elevated neutrophil counts and normal lymphocyte counts across the SCD phenotypes. The NLR was significantly higher in crisis and ischemia but normal in steady-state SCD. Dots (•) indicate outliers. “a” indicates a significant difference between healthy controls and SCD with VOC. “b” indicates a significant difference between healthy controls and SCD with ischemic stroke. “c” indicates a significant difference between SCD in steady state and SCD with VOC. “d” indicates a significant difference between SCD in steady state and SCD with ischemic stroke.

Further analysis of variations of selected endothelial markers in different SCD phenotypes revealed significantly elevated VWF antigen levels in SCD with VOC compared to healthy controls, with no significant differences reported for other phenotypes (Figure 3). In contrast, the differences in VWF activity between the healthy controls and SCD cohorts were insignificant. However, the study revealed slightly elevated VWF activity in the SCD groups, with VOC recording the highest activity (Figure 3). Conversely, ADAMTS13 antigen levels were significantly higher in SCD in steady state than in healthy controls. The enzyme levels were also higher in steady state but insignificantly different from the antigen levels in VOC and ischemic stroke (Figure 3). The Mann-Whitney U tests comparing the ratios of VWF antigen to VWF activity to ADAMTS13 antigen levels showed significant differences between SCD in steady state and SCD with ischemic stroke (P = .003) and between SCD with VOC and SCD with ischemic stroke (P = .004) (Figure 3).

Figure 3.

Measures of the VWF-ADAMTS13 axis across the SCD phenotypes. Significantly elevated VWF antigen with normal VWF activity in SCD with VOC, and significantly elevated ADAMTS13 antigen in SCD in steady state. An elevated VWF Ag:VWF Act:ADAMTS13 ratio was noted in ischemic stroke and VOC compared with that in steady state. Dots (•) indicate outliers. “a” indicates a significant difference between healthy controls and SCD in steady state. “b” indicates a significant difference between healthy controls and SCD with VOC. “c” indicates a significant difference between SCD in steady state and SCD with ischemic stroke. “d” indicates a significant difference between SCD with VOC and SCD with ischemic stroke. Act, activity; Ag, antigen.

Discussion

Inflammation is a key hallmark of VOC in SCD and is believed to play a role in ischemic stroke in this genetic condition. The activation of the inflammatory pathways in SCD complications (VOC and ischemic stroke) was supported by observed high NLRs across the healthy controls and SCD phenotypes. In particular, the elevated neutrophil counts reported in VOC and ischemic stroke suggest the activation of the inflammatory pathways in response to endovascular damage by the rigid sickled erythrocytes.²² At the site of injury, neutrophils activate key effector functions that include degranulation, release of the reactive oxygen species, phagocytosis, and the formation of the neutrophil extracellular traps (NETs).²³ In SCD, previous studies have linked free heme with an increased formation of NETs and a concurrent recruitment of neutrophils at the site of injury.²⁴ The formation of NETs, which capture platelets and sickled erythrocytes, is also linked to the pathogenesis of SCD and particularly the onset and progression of VOC.²⁵ In concurrence, this study showed that neutrophil counts were significantly elevated in VOC and ischemic stroke, supporting the findings of the previous studies on their key roles in SCD complications through the activation of both acute and chronic inflammatory cascades.²⁶ Furthermore, this study also showed significantly higher NLR in patients with SCD than in healthy controls. In particular, the levels were significantly elevated in SCD with VOC and ischemic stroke. This pointed to a likely activation of neutrophil-mediated innate immune responses to endothelial damage in SCD, potentially leading to inflammation, vaso-occlusion, and ischemic stroke.^27,28

Besides the elevated neutrophil counts, this study also showed notable disparities in the markers of anemia, particularly in VOC and ischemic stroke. Although the RBC counts and Hgb levels were significantly lower in all SCD phenotypes than in healthy controls, VOC and ischemic stroke in SCD were associated with lower levels than SCD in steady state. This is partly attributed to the sustained destruction of the sickled RBCs through inflammatory pathways mediated by the neutrophils.²² In addition, the study showed significantly lower PCV in SCD complications with lower Hgb levels.²⁹ Based on the findings from earlier studies, hemolysis might be associated with the release of free Hgb that may affect critical biochemical and physiological pathways through sustained chronic inflammation and oxidative stress.^30,31

In this study, the significantly lower levels of RBC indices such as Hgb and RBC counts, coupled with the observed elevated levels of NRBCs suggest active hemolysis that may lead to the release of free Hgb. This destruction of the RBCs could be inducing additional strain on the bone marrow to release more erythrocytes to replace the hemolyzed cells.³² Consequently, the bone marrow in SCD with VOC and ischemic stroke could be prematurely releasing the NRBCs into circulation.³³ These released RBCs also have significantly reduced MCV and may not effectively meet the demand for oxygen transport to the respiring tissues. Consequently, it predisposes the key tissues to hypoxia, which previous studies have linked to the increased production of proinflammatory cytokines.^34,35 This further leads to a vicious cycle that involves the sustained recruitment of neutrophils at the sites of inflammation and consequent destruction of the RBCs, thereby lowering the body’s capacity to transport respiratory gases efficiently. This is consistent with low RBC count, reduced Hgb concentration, and elevated NRBCs, suggesting hematopoietic stress in the patient population.³⁶

In addition to the observed nexus between the inflammatory pathways and hemolysis in SCD, this study also showed notable variations in the VWF-ADAMTS13 axis that is critical in hemostasis. The VWF antigen was generally higher in SCD phenotypes than in healthy controls, consistent with earlier studies.^37,38 However, the Mann-Whitney U tests revealed that although VWF antigen levels were high in SCD in steady state, it was not statistically different from the levels recorded in healthy controls. This could be attributed to lower-grade endothelial injuries in steady state, suppressing the activation of the inflammatory pathways. Consistently, compared to VOC and ischemic stroke phenotypes, the steady state was also associated with relatively lower neutrophil counts, confirming the active role of the neutrophils in exacerbating inflammatory responses and modulating VWF antigen levels through the inhibition of ADAMTS13 activity.³⁹ In contrast, there was a statistical difference in VWF antigen levels in VOC but not in ischemic stroke. This was contrary to the expected elevated levels of VWF antigen across both VOC and ischemic stroke phenotypes because of the active role of the glycoprotein in hemostasis.

Comparatively, VWF activity was elevated in SCD phenotypes compared to healthy controls (median = 96.65; Q1-Q3, 82.83-133.75), with the highest median activity recorded in VOC (median = 138.50; Q1-Q3, 96.10-179.00). Although these differences in VWF activity were not significantly different according to the Mann-Whitney U tests, they showed that the elevated activity in the SCD phenotypes indicated that the VWF-ADAMTS13 axis could be instrumental in onset and progression of the SCD complications.^40,41 Furthermore, some previous studies suggest that the regulation of VWF activity in SCD is independent of ADAMTS13 activity.⁴¹ In contrast, this study suggests that ADAMTS13 dysregulation and inhibition may be playing a significant role in the onset and progression of VOC and ischemic stroke in pediatric SCD.

For instance, the study on ADAMTS13 levels revealed higher ADAMTS13 antigen levels in SCD phenotypes than in healthy controls (median = 299.24; Q1-Q3, 224.41-396.36). The antigen levels were the highest and significantly different for SCD in steady state (median = 407.98; Q1-Q3, 382.90-478.79) compared with those in healthy controls, whereas they were high but insignificantly different between healthy controls and those with SCD with VOC (median = 354.80; Q1-Q3, 280.76-497.03) and ischemic stroke (median = 369.93; Q1-Q3, 324-419.71). These observations suggest that the biosynthesis and secretion of ADAMTS13 is upregulated in response to increased levels of VWF antigen, often in response to endothelial dysfunction.⁴² However, the effectiveness of the ADAMTS13 in maintaining endovascular patency appears suppressed based on the findings of this study. For instance, in VOC, this study revealed significantly elevated VWF antigen levels with normal ADAMTS13 antigen levels. This may signify an increased risk of vascular occlusion and stroke in patients with SCD.⁴³ The failure of the metalloprotease to cleave VWF multimers, as evident through elevated levels of VWF antigen in VOC, results in the formation of thrombi through the increased recruitment of platelets,⁴⁴ which is consistent with the observed patterns in platelet counts in this study.

A holistic analysis of the findings from the FBC, VWF, and ADAMTS13 assays reveal the active involvement of neutrophils in the onset and progression of SCD complications. Although originally believed to be active in acute inflammation, new studies have documented their involvement in chronic inflammatory conditions such as SCD.^45,46 For instance, they are instrumental in sustaining the vicious cycle of inflammation through the production of proinflammatory mediators and through hemolysis. Their tissue-destructive potential is also likely to increase the production of VWF, possibly explaining the elevated levels of the antigen in VOC. Without modulation, neutrophil activity may therefore increase the risk of vascular occlusion.

In addition, previous studies suggest that neutrophils adhering to inflamed microvascular tissue networks could sustain inflammation in SCD through positive feedback.⁴⁷ In such chronic inflammatory states, elevated levels of proinflammatory cytokines destabilize the VWF-ADAMTS13 axis.^10,48 For instance, interleukin-6 and tumor necrosis factor α are associated with increased release of ultralarge VWF, whereas interferon gamma, interleukin-4, and tumor necrosis factor α also inhibit the synthesis of ADAMTS13.⁴⁸ Consequently, the observed dysfunction in the VWF-ADAMTS13 axis in SCD could be attributed to the disruptive effects of inflammation and oxidative stress on both ADAMTS13 biosynthesis and its activity. The dysfunction in the axis was persistent even in the populations that reported average adherence to hydroxyurea prescriptions, consistent with the findings of an earlier study on an Irish cohort.¹⁹ Moreover, the competitive inhibition of the metalloprotease binding to the A2 domain of VWF by Hgb released from hemolyzed erythrocytes may further exacerbate the risk of thrombosis in the SCD cohort.⁴⁹ Thus, while the levels of ADAMTS13 in SCD phenotypes were normal to high, vaso-occlusion and ischemic stroke might have resulted from the competitive inhibition of ADAMTS13 by free Hgb.⁵⁰ However, a larger sample size focusing on these subgroups may be necessary in future comparative longitudinal studies. It is also necessary to further explore the role of Hgb to elucidate the influence of hemolysis on vaso-occlusion and ischemic stroke, and particularly its role in VWF-ADAMTS13 axis dysfunction. In addition, this study failed to assay the concentration of fetal Hgb that might also contribute to the dysfunction in the VWF-ADAMTS13 axis. Moreover, it is necessary to conduct a more rigorous exclusion of patients with underlying conditions because the current protocol relying on clinical records may have missed some of the comorbidities likely to affect the hematological and endothelial markers of interest.

Conclusions

In conclusion, this study showed that SCD phenotypes are associated with a complex interplay of anemic and inflammatory states that may modulate the VWF-ADAMTS13 axis. In SCD with VOC and ischemic stroke, there was significant surge in neutrophil counts coupled with low RBC and Hgb, signifying chronic inflammation and hemolysis of the sickled RBCs. Consequently, the proinflammatory state might have affected vascular patency, resulting in the observed dysfunction in the VWF-ADAMTS13 axis in VOC. In particular, elevated VWF antigen with normal ADAMTS13 antigen levels is likely indicative of VOC in pediatric SCD. This could be attributed to inhibition of the metalloprotease, making it unable to maintain vascular patency.

Conflict-of-interest disclosure: The authors declare no competing financial interests.