Livedoid vasculopathy in hemoglobinopathy-associated chronic leg ulcers

Abstract

Chronic leg ulceration is a debilitating manifestation of hemoglobinopathies, and best management is uncertain. Livedoid vasculopathy (LV) is a cutaneous non-inflammatory thrombotic vasculopathy treated with anticoagulation that has been identified in hemoglobinopathy-associated chronic leg ulceration. However, most patients with hemoglobinopathy-associated ulcers do not undergo workup for secondary causes, and the prevalence and relevance of LV is unclear. Outcomes of secondary workup were examined retrospectively in this study. 108 patients with hemoglobinopathy-associated chronic leg ulcers were identified. 15% of patients underwent skin biopsy, and 97% of biopsies showed non-specific findings. Two patients had LV and neither responded to anticoagulants. Livedoid vasculopathy is a rare cause of ulceration in hemoglobin gene disorders and the benefit of anticoagulation in these cases is unclear.

Introduction

Chronic leg ulceration in hemoglobinopathies remains poorly understood, difficult to treat, and a significant burden to patients [1]. In sickle cell disease (SCD), ulcer formation is likely due to vaso-occlusion by sickled erythrocytes resulting in cutaneous ischemia; however, most individuals with SCD do not develop ulcers [2]. Therefore, additional factors, such as microtrauma, venous stasis, inflammation, endothelial dysfunction due to chronic hemolysis, and comorbid conditions such as diabetes mellitus likely contribute [3]. The involvement of microvascular thrombosis in leg ulcer formation has been suggested by a case series demonstrating thrombotic vasculopathy in skin biopsies of patients with SCD [2].

Livedoid vasculopathy (LV) is a painful, ulcerative, waxing and waning, non-vasculitic, cutaneous thrombotic vasculopathy typically treated with anticoagulation [4]. Diagnosis of LV is predicated on a combination of skin biopsy findings and characteristic clinical features including atrophie blanche (white, stellate scarring) and violaceous patches (retiform purpura) over the bilateral lower extremities [5]. One case of LV has been reported in association with sickle cell disease [6]. However, patients with SCD or thalassemia-associated ulcers rarely undergo skin biopsy due to concerns regarding the risks of exacerbating pain and wound healing [7, 8]. It is also unknown how often biopsies yield actionable results or whether anticoagulation is beneficial in these patients [3].

Having encountered patients with LV and hemoglobinopathy in our practice, we set out to identify the frequency and outcomes of laboratory, skin biopsy, and vascular imaging workup for secondary causes of hemoglobinopathy-associated ulcers including LV.

Methods

The study was approved by the Institutional Review Board at our institution. The Carolina Data Warehouse for Health, an electronic medical record data repository of more than 4.5 million patients at a large, public academic health system, was queried between April 2014 and December 2022 for patients with diagnostic codes associated with hemoglobinopathy and chronic leg ulceration. Chronic leg ulcers were defined as a break in the skin below the knee persisting for more than four weeks [9]. Charts were manually reviewed to verify diagnoses and extract comorbidity data such as diabetes mellitus. Any available duplex ultrasound studies evaluating for deep vein thrombosis (DVT), chronic venous insufficiency (CVI), and peripheral artery disease (PAD) were also reviewed. Laboratory evaluation performed specifically to workup systemic causes of ulceration was recorded. This included vasculitis studies (anti-nuclear antibody, anti-neutrophil cytoplasmic antibody, complement C3 & C4, rheumatoid factor, cryoglobulins), hypercoagulability panel (antiphospholipid antibodies, protein C, protein S, and antithrombin activity, factor V Leiden and prothrombin gene mutation, serum protein electrophoresis, cold agglutinins) and infectious workup (hepatitis, syphilis, and HIV serologies). Individuals with sickle cell trait or a history of stem cell transplant were excluded. Cases of LV were identified according to our previously published diagnostic approach; that is, painful, relapsing ulcers over the bilateral lower extremities accompanied by retiform purpura and atrophie blanche with evidence of thrombotic vasculopathy and hyaline thickening of vessel walls on skin biopsy without an inflammatory infiltrate [10]. A case of LV in delta beta thalassemia was referred to us and not identified through the database query. This case was detailed below but not included in the outcome tally (Case 2).

Results

One-hundred and eight patients with hemoglobinopathy and chronic leg ulceration were identified between 2014 and 2022. Comorbidity data for all cohorts was outlined in Table 1.

Table 1.

Secondary workup of hemoglobinopathy associated leg ulceration.

	Sickle cell disease	Thalassemia
n	89	19
Female	33 (37%)	11 (58%)
Diabetes	10 (11%)	14 (74%)
US CVI: Positive/Tested	9/14 (10%/16%)	4/5 (21%/26%)
US PAD Positive/Tested	5/21 (6%/24%)	3/4 (16%/21%)
DVT history	32 (36%)	5 (26%)
Skin biopsy	11 (12%)	2 (11%)

Abbreviations: US = ultrasonography, CVI = chronic venous insufficiency, PAD = peripheral artery disease, DVT = deep vein thrombosis

Eighty-nine patients had SCD, with 75 (84%) homozygous (HbSS) sickle cell anemia genotype. Thirty-three (37%) were female. Seventy-six (85%) of these patients had been treated with hydroxyurea and mean fetal hemoglobin (HbF) percentage was 9%. Eleven patients (11%) had undergone skin biopsy. Mean age at first ulcer biopsy was 49 years (standard deviation 10.7 years). Eight biopsies showed non-specific ulcerative changes, two showed evidence of stasis dermatitis and one was consistent with LV (Case 1). Only one patient had undergone laboratory evaluation for systemic causes of ulceration which was negative for antiphospholipid antibodies, antinuclear antibodies and rheumatoid factor with normal protein C and protein S levels.

Nineteen patients had thalassemia. Two of these patients underwent skin biopsy at ages 51 and 53 respectively. Both biopsies showed venous stasis. No patients underwent laboratory vasculitis, hypercoagulability or infectious studies specifically to workup the leg ulceration.

Case 1

A 58-year-old female with sickle cell disease (HbSS) on hydroxyurea and a history of relatively unprovoked left lower extremity deep vein thrombosis (DVT) developed a painful ulcer over the left ankle one year after a three-month course of therapeutic anticoagulation for her DVT. Her wound was initially treated with compression therapy and healed over six months while continuing hydroxyurea. This was followed by multiple recurrences over the bilateral lower extremities (Fig. 1). Skin biopsy at the ulcer margin performed at age 62 revealed LV. Apixaban was initiated at 2.5 mg twice daily for long term DVT prophylaxis and ulcer management; however, she continued to have recurrences and discontinued anticoagulation after a year. Voxelotor, a hemoglobin S polymerization inhibitor, was then added to hydroxyurea, leading to healing of ulcers over a month without any recurrences for over a year. As of publication, voxelotor is no longer available on the market.

Fig. 1.

Left: healing ulcer with surrounding hyperpigmentation and atrophie blanche (Case 1). Right: deep ulceration over the medial malleolus with surrounding retiform purpura (Case 2)

Baseline hemoglobin was 9 g/dL. Mean HbF was 24.5% (Reference 0-1.9%), lactate dehydrogenase (LDH) was 684 U/L (Reference 120–246 U/L), and absolute reticulocyte count was 89.2*10^9/L (Reference 23–100*10^9/L). She has not undergone a thrombophilia workup. A chronic venous insufficiency study of the bilateral lower extremities was unremarkable.

Case 2

A 28-year-old man with non-transfusion-dependent delta beta thalassemia intermedia developed a palpable purpuric eruption over the bilateral legs which then ulcerated (Fig. 1). No immediate improvement in leg ulcers was seen after one week of systemic prednisone therapy, but lesions healed after six weeks of weekly debridement and compression wraps. Skin biopsy at the ulcer margin after a fifth flare was consistent with LV. Daily aspirin 325 mg and rivaroxaban 20 mg were initiated, but he experienced two recurrences regardless, and aspirin and rivaroxaban were then discontinued. A trial of monthly partial red cell exchange transfusion was started during an ulcer free period and he did not develop recurrent ulcers over six months.

Baseline hemoglobin was 10 g/dL, HbF was > 99%, and there was evidence for chronic hemolysis on labs. Hypercoagulability workup was negative for Factor V Leiden and Prothrombin G20210A mutation. Anticardiolipin and anti-beta-2-glycoprotein-I IgG/IgM were within normal limits on two occasions and lupus anticoagulant testing was negative. Antinuclear antibody, antineutrophil cytoplasmic antibodies, cryoglobulins, anti-double stranded DNA, and rheumatoid factor were negative. Duplex ultrasound examination did not show evidence of venous reflux.

Discussion

There is insufficient evidence to support routine skin biopsy in patients with hemoglobinopathy-associated chronic leg ulceration. Few such patients underwent skin biopsy (15%) in this study and most biopsies showed non-specific findings or stasis changes (97%). Only two patients had evidence of livedoid vasculopathy. While multiple biopsies may be required to make the diagnosis of LV due to patchy vascular involvement, skin biopsy is not without risk [5]. Hemoglobinopathy patients may be particularly prone to pain and poor wound healing at biopsy sites. Skin biopsy may be considered in patients with treatment refractory ulcers or in those with lesions, such as retiform purpura, that are atypical for venous stasis ulcers, as findings such as vasculitis or thrombotic vasculopathy would have implications for management. Biopsy may also be considered in individuals in which there is concern for malignancy or malignant transformation in consultation with dermatology.

The role of antithrombotic therapy in LV associated with hemoglobinopathies is also unclear. Sickle cell disease and thalassemia are thrombophilic, in part due to the vascular effects of chronic hemolysis [11, 12]. Hemoglobinopathy patients with chronic ulcers may be particularly hypercoagulable. In our study, 37% of patients with sickle cell disease and chronic leg ulcers had a history of DVT, and, consistent with another published series, this is higher than the prevalence reported in the general sickle cell population [2, 13]. Whether ulcers might be related to thrombotic vasculopathy or post-thrombotic syndrome in these patients is unclear. Patients with ulcers may also be sicker, less mobile, and more prone to thrombotic complications. One case of LV associated with sickle cell disease that did not recur following initiation of apixaban has been reported [6]. Six-week therapy using antithrombin concentrate was found to be helpful in a small series of patients with sickle cell-beta thalassemia and refractory ulcers [14]. Anticoagulation is the most commonly prescribed monotherapy for LV, and anticoagulants and antiplatelet agents are considered first-line treatment options for the condition [4, 15]. While anticoagulation was not clearly beneficial in the two cases reported here, these numbers are too small to draw definitive conclusions. Additionally, the choice and optimal dosing of anticoagulant and antiplatelet therapy for LV has not been determined and no randomized trials comparing treatments or dosing levels have been conducted. Apixaban dosing was not provided in the prior published report of LV in sickle cell disease [6]. A systematic review of LV therapy showed similar effectiveness of rivaroxaban at both prophylactic and treatment doses [4, 16]. Increased effectiveness of higher doses of aspirin has been reported, though this was based on comparisons across limited case series [17].

It is notable that most patients with sickle cell disease had received treatment with hydroxyurea. Hydroxyurea has been associated with ulcer formation with similar histologic appearance to livedoid vasculopathy in patients with myeloproliferative neoplasms possibly due to the promotion of megaloblastic, non-deformable red blood cells leading to microvascular occlusion [18, 19]. However, hydroxyurea has not clearly been linked to leg ulcers in sickle cell disease [3, 20]. This may be due to its promotion of oxygen carrying capacity, as well as decreased hemolysis, leukocyte count and inflammatory cytokines [3].

Conclusion

Livedoid vasculopathy contributes to few cases of hemoglobinopathy-associated leg ulceration, but its true prevalence in this population is unknown as patients rarely undergo skin biopsy. The decision to pursue skin biopsy should be performed with caution given the risks of pain and poor wound healing, and be driven by a patient’s symptoms, ulcer appearance, and treatment refractoriness. It is not known whether anticoagulation has any benefit in ulcer prevention, and it is noteworthy that the two patients in our series had recurrent skin ulcers despite anticoagulation.

Abbreviations

SCD

Sickle cell disease

LV

Livedoid vasculopathy

HbF

Fetal hemoglobin

DVT

Deep vein thrombosis

LDH

Lactate dehydrogenase

Data availability

No datasets were generated or analysed during the current study.