Skin Eruptions in Acute Hemorrhagic Edema of Young Children: Systematic Review of the Literature

Acute hemorrhagic edema of young children, also termed cockade purpura with edema or Finkelstein--Seidlmayer disease, is a skin-limited small-vessel leukocytoclastic vasculitis, which characteristically occurs after a simple, mostly viral, febrile illness [1-4]. It affects infants 4 weeks to 2 years of age and spontaneously remits within 3 weeks with little propensity toward relapses [2-5].

Although vascular immunoglobulin A deposits are not detected in the majority of cases [6], hemorrhagic edema is often deemed to be the infantile variant of immunoglobulin A vasculitis [1-5]. The latter condition presents with a palpable rash mainly affecting the lower extremities and the buttocks [3, 4]. Available narrative reviews on hemorrhagic edema state that affected children develop, over 48 h or less, extensive and symmetrically distributed annular or nummular eruptions (Fig. 1) and edema [2-4]. However, type, distribution, and evolution of the rash over the disease course have never been systematically investigated. To address this issue, we employed the data contained in the Acute Hemorrhagic Edema BIbliographic Database (AHEBID).

AHEBID [5-7] progressively integrates all the original articles on acute hemorrhagic edema of young children published after the initial communication in 1913. For this purpose, the bibliography search engines Excerpta Medica, the National Library of Medicine databases, and Google scholar are screened every second month by two of us for “acute hemorrhagic edema,” “cockade purpura and edema,” and “infantile Henoch-Schönlein purpura” without any language restriction. AHEBID also includes the medical literature on this condition collected by some of us in the early eighties of the last century.

The first systematic literature search was made in January 2019. As of December 31, 2020, the database incorporated 301 original reports (letters, case reports, or full-length articles) published since 1970, which addressed 488 individually documented cases (Fig. 2).

AHEBID [7] is available on request (email: [email protected]). Furthermore, the raw data extracted for the current analysis may be found on the data repository Zenodo as “Skin eruptions in acute hemorrhagic edema of young children: systematic review of the literature.” In all cases, the diagnosis of hemorrhagic edema made in the original reports was reviewed using three classic predefined criteria: raised annular or nummular eruptions and skin edema in a not-ill-appearing child [2-4]. The clinical diagnosis was supported by a biopsy disclosing a nongranulomatous neutrophil infiltration into small-vessel walls with karyorrhexis in 241 cases.

For the present analysis, we included exclusively cases with documented type, distribution, or evolution of skin eruptions. Since numerous and sometimes confusing descriptive terms are used to denote skin eruptions, we developed the glossary reported in Table 1 [8] to define annular or nummular eruptions, edema, petechiae, purpuric eruptions, ecchymoses, macules, simple or papular wheals, and scratch marks. The distinctive signs of inflammatory skin edema, including tenderness (or pain), induration, and absent indentation after pressure were specifically addressed.

A predesigned form was employed to record information about demographics of cases, and type, distribution, or evolution of eruptions. Two authors independently extracted the data in a nonblinded fashion. Uncertainties were resolved through team discussions and consensus.

Categorical variables are presented as frequency, continuous variables as median and interquartile range. For statistics, the two-tailed χ² test and the Mann-Whitney-Wilcoxon test were used [9]. p values <0.05 were considered significant.

The skin eruptions were poorly documented in 105 (22%) cases, 74 boys and 31 girls, 12 (9–20) months of age. The key features of the rash were accurately documented in the remaining 383 (78%) cases (261 boys and 122 girls, 11 [8–18] months of age), mostly (n = 321) with the support of photographs. Patients without and with accurately documented skin eruptions did not significantly differ with respect to sex and age.

The skin rash was bilateral and symmetrically distributed in 380 and unilateral in the remaining 3 cases. The skin eruptions reported in the 383 cases appear in Table 2. Annular or nummular eruptions and edema, two inclusion criteria, were reported in all cases. Annular eruptions in a strict sense, usually targetoid, were reported in 98% of cases and were very often associated with polycyclic (slightly less than every second case) or arciform (approximately every fourth case) eruptions. Nummular eruptions were also reported in more than 90% of cases. Tenderness or pain (n = 103), induration (n = 22), and absent indentation after pressure (n = 32) characterized skin edema. Purpuric eruptions and ecchymoses (and, occasionally, also pete-chiae) were reported in the vast majority of cases. Macules and simple or papular wheals were described in a minority of cases. Scratch marks were rarely observed.

The distribution of annular or nummular eruptions, edema, and macules is presented in Table 3. The relative distribution of annular and nummular eruptions was very similar. The following body regions were affected, in decreasing order of frequency, by the two mentioned eruptions: legs and feet, arms and hands, face and ears, trunk, and genitals.

With the exception of feet, which were affected in about three-quarter of cases, the same distribution was reported for edema.

No detailed information about the evolution of round skin eruptions was provided for 407 cases: 280 boys and 127 girls 11 (9–17) months of age. These data were extracted for the remaining 81 cases: 55 boys and 26 girls 11 (9–17) months of age. Patients without and with documented evolution of eruptions did not significantly differ with respect to sex and age.

In the mentioned 81 cases, a papular wheal, a macule, or a simple wheal often preceded the classic nummular or annular eruption (Fig. 3). A nummular eruption often evolved into an annular targetoid or, less frequently, into an annular nontargetoid eruption. Finally, both the evolution of an annular nontargetoid into an annular targetoid eruption and, less frequently, the opposite were also reported.

Acute hemorrhagic edema of young children is typically confined to the skin. Textbooks and reviews state that extensive edema and round (either nummular or annular) eruptions characterize this condition [1-4]. This careful review points out that the symmetrically distributed rash of this vasculitis is more heterogeneous than so far believed and may be summarized and discussed as follows.

Tenderness (or pain), induration, and absent indentation after pressure characterize skin edema in most children with hemorrhagic edema. In our experience, warmth is a further feature. Terms such as acute inflammatory skin edema or, more rarely, noninfectious pseudo-cellulitis have been used to denote this rare but peculiar form of skin edema [10].

The vast majority of children present with both nummular and annular eruptions. In addition to annular eruptions in a strict sense, many children present with polycyclic or arciform eruptions. On the other hand, annular eruptions in a broad sense may be either nontargetoid (i.e., with only two zones of color) or targetoid (i.e., with three zones of color).

Purpuric eruptions and ecchymoses are reported in the majority of cases (sometimes in association with petechiae). Macules and both simple and papular wheals may also occur. Scratch marks, specific signs associated with pruritus, are uncommon.

The initial eruption may be either a simple or papular wheal or a macule that evolves into a nummular or, less frequently, directly an annular eruption. Nummular eruptions successively evolve into annular ones. Finally, both the evolution of an annular nontargetoid into a classic annular targetoid eruption and, less frequently, the opposite may also occur.

Nontargetoid and especially targetoid annular eruptions are the classic hallmark of erythema multiforme [11, 12]. However, recent reports suggest that no more than one-quarter of children with the mentioned skin eruption suffer from erythema multiforme [13, 14]. In this age group, acute ordinary urticaria or urticaria multiforme and acute hemorrhagic edema underlie, in the vast majority of cases, targetoid-like eruptions [13-15]. Since erythema multiforme is rare before the age of 4 years, hemorrhagic edema accounts for targetoid-like eruptions in many children 3 years or less of age. The inappropriate term “infantile erythema multiforme” has been sometimes utilized in the past to denote hemorrhagic edema of young children [14]. Syphilis and paintball injuries are further rare causes of targetoid or nontargetoid annular eruptions in childhood [13-15].

Dispute exists as to whether immunoglobulin A vasculitis and acute hemorrhagic edema of young children are different clinical features of the same disease or separate diseases. Some believe that hemorrhagic edema is the infantile variant of immunoglobulin A vasculitis. This statement is sustained by a report documenting the concomitant appearance of hemorrhagic edema and immunoglobulin A vasculitis in first-degree relatives [16]. Cases also exist of children with findings overlapping between the two vasculitides [17]. On the other hand, there are sufficient data to consider hemorrhagic edema a distinct entity, including the failure to detect depositions of immunoglobulin A in most biopsy specimens and the skin-limited involvement. The present systematic analysis, which documents type and distribution of skin eruptions in hemorrhagic edema, further supports this assumption. A report recently released by some of us identified a truncating variant in the HCK gene in 4 subjects affected by hemorrhagic edema [18]. This gene is located on chromosome 20 and encodes for a protein-tyrosine kinase that is predominantly expressed in hemopoietic cells. Furthermore, aberrant activity of the HCK gene has been noted in inflammatory and cancer diseases. Further investigations are warranted to determine if and how the HCK gene is involved not only in hemorrhagic edema but also in immunoglobulin A vasculitis.

Skin vasculitides, including, among others, immunoglobulin A vasculitis, have been associated with the severe acute respiratory syndrome coronavirus 2 [19]. To the best of our knowledge, no cases of hemorrhagic edema have been so far reported in patients affected by this virus. In 2017, acute hemorrhagic edema was observed in an infant with an upper respiratory infection caused by coronavirus NL63 [20].

The most important limitation of the present analysis comes from the small number of published articles on children affected by hemorrhagic edema, which were published over 50 years. A second limitation is that the analysis incorporated exclusively data from case reports. Finally, because the diagnosis of hemorrhagic edema is mostly made on clinical grounds, diagnostic accuracy depends on experience and skills of the examiner.

This is the first study which carefully characterizes type, distribution, and evolution of skin eruptions in hemorrhagic edema of young children. These data help physicians to rapidly and noninvasively make the clinical diagnosis of this apparently alarming but very benign vasculitis.

This study carefully characterizes type, distribution, and evolution of eruptions in acute hemorrhagic edema.

Verbal informed consent to publish the pictures was obtained from the parents of the children prior to the publication. This consent procedure was reviewed and approved by the Ethics Committee of Southern Switzerland (approval number 3653).

The authors have no conflicts of interest to declare.

This research received no external funding.

P.O.R., M.G.B., S.A.G.L., and G.P.M. developed the Acute Hemorrhagic Edema Bibliographic Database; G.B., P.O.R., M.G.B., and G.P.M. conceptualized the study design; G.B. and P.O.R. extracted the data; G.B., C.L., and S.A.G.L. analyzed the data; G.B., M.G.B., and G.P.M. wrote the manuscript; all authors read and commented on the final version of the manuscript.