Leg Ulcers are Indicators of Systemic Dysfunction in Individuals with Sickle Cell Disease

Graphical Abstract

To the Editor:

Sickle cell disease (SCD) is an inherited genetic disorder that affects approximately 100,000 individuals in the United States and millions worldwide (1). There are a variety of clinical manifestations associated with SCD. Among them, leg ulcers are a recurrent and painful complication that affect quality of life and are associated with decreased life expectancy (2). Leg ulcers typically occur in 5–10% of individuals living with SCD globally; however, prevalence correlates with age and geographic region (2). Though their pathophysiology remains poorly defined, recently, it has been shown that leg ulceration in individuals living with SCD is associated with markers of hemolytic severity making chronic hemolysis a contributing factor to their occurrence (3).

Pulse pressure is defined as the difference between systolic and diastolic blood pressure. Elevated pulse pressure is predominately attributed to arterial stiffness, lower diastolic blood pressure, and microvascular damage (4, 5). It is also an independent predictor of poor cardiovascular outcomes and endothelial dysfunction and has been linked to hemolysis in individuals with SCD (5). Pulse pressure is associated with hemolytic biomarkers such as reticulocyte count and lactate dehydrogenase (LDH), and additional variables including hemoglobin oxygen saturation and creatinine (5).

To date, little attention has been devoted to pulse pressure in SCD, with no evidence of an association between pulse pressure and leg ulcers in individuals living with SCD. We set to examine pulse pressure, clinical, and laboratory parameters in a large, prospectively acquired, cohort of SCD adults with or without leg ulcers, in the United States.

Data was obtained from a prospective cross-sectional study (Insights into Microbiome and Environmental Contributions to Sickle Cell Disease and Leg Ulcers Study, NCT02156102), which examines the psychosocial, environmental, genetic, and clinical contributors to disease variation in individuals with SCD. Adults with all SCD genotypes (age 18 or older) were recruited between June 2014 and January 2020. All participants provided consent and received study compensation. Participants were enrolled either at the NIH Clinical Center in Bethesda, MD, or at the Montefiore Medical Center in Bronx, NY (IRB approval obtained at both sites). Medical history and clinical exams were completed at time of visit, with all participants seen at steady state, defined as being free of any vaso-occlusive events for at least 2 weeks. Individuals with active or healed leg ulcers were intentionally oversampled. Data analyses were performed using R (version 4.3.1), and statistical tests were carried out using the aov and cor.test functions from the base R stats package. Correlation data were analyzed using Spearman’s Correlation test.

The study population comprises 276 participants stratified by ulcer status—60 (21.7%) individuals with one or more active leg ulcers, 47 (17%) individuals with a history of leg ulcers but no active ulcer, and 169 (61.2%) individuals with neither. Stratification of our cohort was done to further examine if laboratory and clinical data significantly differed between participants with 1) active leg ulcers, 2) healed ulcers, or 3) who never experienced a leg ulcer. Participants’ mean age was 38.7 years old (range: 19–71), and the majority of participants were of Black or African American descent (N=255, 97%). Most participants had homozygous SCD or HbSS genotype (N=219, 79.3%), followed by 35 (12.7%) with HbSC, 13 (4.7%) with HbSB⁺, and 8 (2.9%) with HbSB⁰. At the time of data collection, 59.1% (163/276) of participants reported taking hydroxyurea: 61.7% (37/60) in the active leg ulcer group, 60.9% (28/46) in the ulcer history group, and 58% (98/169) in the no ulcer group.

Several differences in the clinical and laboratory parameters were detected between the three groups. Generalized SCD pain was assessed on a scale of 0–10 with 0 meaning no pain and 10 meaning the worst imaginable pain. Individuals with active ulcers reported significantly higher general pain scores than the two other groups with an average score of 3.36 for the active Ulcer group, 1.68 for Ulcer Hx group, and 1.93 for the No Ulcer group (p=0.001). In addition, individuals with active ulcers reported pain specific to their leg ulcers, with an average of 7.69 (range: 0 – 10, SD: 3), indicating that pain burden was higher for participants with active ulcers compared to the other two groups.

Average pulse pressures were significantly different across the three groups, with the active ulcer group having the widest average pulse pressure, 55.8 mmHG, and the no leg ulcer group having the lowest pulse pressure at 49.3 mmHG. Additionally, compared to the two other groups, individuals with active leg ulcers had significantly lower hematocrit (22.7 % – Active Ulcer vs. 26.1%– Ulcer Hx vs. 26.7% – No Ulcer Hx), lower hemoglobin (7.9 g/dL –Active Ulcer vs. 9.1 g/dL –Ulcer Hx vs. 9.3 g/dL –No Ulcer Hx), lower albumin (4.1 mg/L – Active Ulcer vs. 4.4 mg/L – Ulcer Hx vs. 4.4 mg/L – No Ulcer Hx), higher reticulocyte percentage (12.7 % --Active Ulcer vs. 9.4 % --Ulcer Hx vs. 8.6 % --No Ulcer Hx), higher reticulocyte count (293 –Active Ulcer vs. 254 –Ulcer Hx vs. 236 –No Ulcer Hx), higher erythrocyte sedimentation rate (ESR) (38.5 mm/hr – Active Ulcer vs. 22.9 Ulcer Hx vs. 19.9 No Ulcer Hx), and higher WBC (10.4 K/μl –Active Leg Ulcer vs. 8.2 K/μL—Hx of Leg Ulcer vs. 8.6 K/μL—No Leg Ulcer Hx). Platelet count, fetal hemoglobin, ferritin, protein S, creatinine, or C-reactive protein (CRP) were not significantly different. Though not significant, participants with active leg ulcers had a higher total bilirubin compared to the other participant groups (0.453 mg/dL--Active Ulcer vs. 0.443 mg/dL—Ulcer Hx vs. 0.406 mg/dL--No Ulcer Hx). Clinical and laboratory results are reported in Table 1.

Table 1 —

Clinical and Laboratory Parameters

Laboratory Results	Normal Ranges	Active Ulcer	Ulcer Hx	No Ulcer	p-valuea	Association with Pulse Pressureb
Pulse Pressure (mmHg)	30–40	55.8	50.7	49.3	0.001 **	-
Systolic Blood Pressure (mmHg)	≤ 120	122	117	118	0.237	-
Diastolic Blood Pressure (mmHg)	≤ 80	65.9	66.7	69	0.112	-
Albumin (g/dL)	3.5–5.2	4.12	4.42	4.40	<0.001 ***	−0.050
Antithrombin (%)	63–138	84.7	88.5	85.7	0.698	−0.040
Arterial Oxygen Saturation (%)	95–100	96.5	97.1	95.9	0.646	−0.020
Bilirubin Direct (mg/dL)	0.0–1.2	2.61	2.77	2.46	0.656	−0.013
Bilirubin Total (mg/dL)	0.0–0.3	0.453	0.443	0.40	0.813	−0.030
Creatinine (mg/dL)	0.67–1.17	0.832	0.926	0.886	0.850	0.193 ***
CRP (mg/L)	0–4.99	9.49	4.21	6.79	0.123	0.071
ESR (mm/hr)	0–25	38.5	22.9	20	<0.001 ***	0.152 *
Ferritin (μg/L)	30–400	1232	1291	955	0.494	−0.103
General Pain	0–10	3.36	1.68	1.93	0.001 **	0.025
Hematocrit (%)	40.1–51.0	22.7	26.1	26.7	<0.001 ***	−0.172 **
Hemoglobin (g/dL)	13.7–17.5	7.88	9.13	9.34	<0.001 ***	−0.189 **
Fetal Hemoglobin (%)	0.0 – 2.0	8.03	8.89	9.51	0.448	−0.004
LDH (U/L)	140–280	508	380	392	0.003 **	0.089
Platelet Count (103/μ)	161–347	385	342	330	0.116	−0.029
Protein C (%)	59–144	68.7	76.5	78.8	0.044 *	−0.030
Protein S (%)	55–134	50.4	54.4	58.2	0.107	−0.066
RBC (M/μL)	4.63–6.08	2.43	2.87	3.01	<0.001 ***	−0.095
Reticulocyte Count (10 9 /L)	26.0–95.0	293	254	236	0.044 *	0.026
Reticulocyte (%)	0.51–1.81	12.7	9.40	8.62	<0.001 ***	0.062
WBC (10 3 /μL)	4.23–9.07	10.4	8.23	8.61	<0.001 ***	0.015
Zinc (μg/dL)	66–110	70.8	72.1	73.0	0.616	0.055
Hydroxyurea	−	37/60 (63.8%)	28/46 (60.9%)	98/169 (59.4%)	0.841	-
Genotype:						0.461
HbSB+	-	1/60 (1.7%)	4/46 (8.7%)	8/169 (4.7%)	0.242	-
HbSB0	-	0/60 (0.0%)	0/46 (0.0%)	8/169 (4.7%)	0.076	-
HbSC	-	1/60 (1.7%)	5/46 (10.9%)	29/169 (17.2%)	0.007 **	-
HbSS	-	58/60 (96.7%)	37/46 (80.4%)	123/169 (72.8%)	<0.001 ***	-
Other	-	0/60 (0.0%)	0/46 (0.0%)	1/169 (0.6%)	0.732	-

^aP-values are from an ANOVA test of equal means across the three ulcer groups. For the categorical, measures hydroxyurea and genotype, p-values correspond to the likelihood ratio test for the ulcer group variable in a logistic regression model predicting the hydroxyurea or genotype indicator.

^*p < 0.05

^**p < 0.01

^***p < 0.001.

^bSpearman correlation is shown for all continuous laboratory results. For the categorical measures, hydroxyurea and genotype, the p-value for an ANOVA test for a difference in mean pulse pressure is shown.

Next, we analyzed the correlations between pulse pressure and each laboratory measure. Significant negative correlations were found between hematocrit and pulse pressure (p=0.004), and hemoglobin and pulse pressure (p=0.002). Significant positive correlations were found between ESR and pulse pressure (p=0.013), and creatinine and pulse pressure (p=0.001).

As previously stated, wider pulse pressure is an indication of cardiovascular and endothelial dysfunction attributing to arterial stiffness and microvascular damage (5). Our findings show that individuals with active leg ulcers have significantly wider pulse pressures compared to all other groups (p=0.002). Participants in the active leg ulcer group also showed significantly reduced albumin (p<0.001) alongside higher total bilirubin, suggesting impaired liver function and high hemolysis. Further indications of chronic hemolysis in participants with active leg ulcers were shown in their significantly higher reticulocyte percentages (p<0.001) and reticulocyte counts (p=0.044) compared to the other participant groups. Moreover, although all participants displayed some degree of anemia, those in the active leg ulcer group exhibited the lowest hemoglobin and hematocrit. Notably, inflammatory markers including WBC and ESR were higher in the active ulcer group, underscoring the presence of chronic inflammation or possible infection. The observed correlations provide further substantiation of the relationship between pulse pressure and chronic hemolysis and inflammation in individuals with active leg ulcers. The negative correlation between pulse pressure and hemoglobin, as well as hematocrit, underscores how a decline in hematocrit and hemoglobin, at least in part secondary to hemolysis, leads to a wider pulse pressure. In addition, the positive correlation between ESR and pulse pressure suggest that chronic inflammation in SCD may affect cardiovascular health status, as it has been shown in other disease states. Additionally, a noteworthy positive correlation between creatinine and pulse pressure demonstrates that as creatinine levels rise, so does pulse pressure, indicating its potential connection to kidney dysfunction. Considering these findings, we draw the conclusion that leg ulcers in individuals living with SCD, although a seemingly localized complication, represent evidence of systemic dysfunction.

Limitations for this study included its cross-sectional design, as there is no longitudinal data on our participants. Analysis would have been strengthened with multiple blood pressure readings; however, participants were only evaluated one time on this study. Nonetheless, the findings of our investigation of leg ulcers in individuals living with SCD may spur additional research that advances our understanding of the pathophysiology of leg ulcers and emphasizes their potential as a valuable window into the broader systemic processes at play.

The presence of a leg ulcer in an individual with SCD, whether it is an isolated episode or a recurrent one, is an indication to 1) perform a comprehensive clinical and laboratory evaluation to unmask occult cardiovascular disease and/or chronic kidney and/or liver disease and 2) augment SCD-directed interventions to prevent future vascular complications that have a high likelihood of occurring. As leg ulcers have been associated with increased risk of pulmonary hypertension and other cardiovascular complications; and pulse pressure is an easily obtainable measure of cardiovascular dysfunction and hemolysis, both can be used clinically to indicate systemic dysfunction in people living with SCD.

Data Availability

Data will be made available upon request.