Keywords
Epidermal necrolysis toxic - herpes-associated erythema multiforme - herpes simplex
virus - Stevens–Johnson syndrome
Introduction
Erythema multiforme (EM) is an acute, recurrent, self-limited, inflammatory mucocutaneous
disease that manifests on the skin and often on the oral mucosa. The skin lesions
may take several forms such as macules, papules, vesicles, bullae, and hence the term
“multiforme”. The classic skin lesion consists of central blister or necrosis with
concentric rings of variable color around it named typical “target” or “iris” lesion
that is pathognomonic of EM. Exact etiology is not known, consider to be a hypersensitivity
reaction, and the most common inciting factors are infection, particularly with herpes
simplex virus (HSV) or drug reaction to nonsteroidal anti-inflammatory drug (NSAIDs)
or anticonvulsants. Adverse drug-related skin reactions that have a frequency above
1% are urticaria, angioedema, photosensitivity, fixed drug eruptions, EM, Stevens–Johnson
syndrome (SJS), and toxic epidermal necrolysis (TEN). Historically, fulminant forms
of EM were labeled SJS and TEN (Lyell disease). However, more recent data suggest
that EM is etiopathologically distinct from those two latter conditions, and they
are discussed separately.[1]
Epidemiology
Although it can happen at any age, EM minor is more common in patients among 20 and
40 years, in spite of more than 20% of the cases affect children after 3 years old
and adolescents.[2] Recurrences occurs in 37% of the cases, and they usually happen in the spring and
autumn season, with clinical severity increase of the attacks.[2],[3],[4] In agreement with Farthing et al.,[3] EM minor may be recurrent and the oral cavity is frequently affected. The prevalence
of oral EM minor varies from 35% to 65% among patients with skin lesions. However,
in patients where EM minor was diagnosed by oral lesions, incidence of skin lesions
ranged from 25% to 33%.[5] An interdisciplinary study reported that 70% of cutaneous recurrent EM minor patients
had an oral involvement, comprising of multiple, large, shallow, extremely painful,
and debilitating ulcers, with entire oral mucosa affected in over 20%.[3] The oral lesions have predilection for the vermilion border of the lips and the
buccal mucosa, mostly sparing the gingiva.[5]
Etiology and Pathogenesis
Etiology and Pathogenesis
Even though many factors may be involved in the EM, frequently the basic cause of
the disease is unknown. In contrast to Skin EM, which is mainly caused by systemic
drugs (principally anticonvulsants, sulfonamides, NSAIDs, and antibiotics) and HSV
infection, the etiological agents remain obscure in many oral EM cases. Many studies
based on cohorts with cutaneous involvement found a relation between EM and HSV infection
which has not always established in studies of stomatological cohorts. HSV DNA has
been demonstrated in cutaneous and oral lesions, but the role of HSV in the etiology
of oral EM remains uncertain. Some studies show that recurrent EM precipitated by
sun exposure is associated with HSV infection in 65%–70% of cases and is known as
herpes-associated EM (HAEM), both by history of 1–3 weeks before onset of EM, seropositivity
for HSV antibodies, and identification of HSV antigens. Approximately 87% had concurrent
Recurrent herpes labialis (RHL). Using polymerase chain reaction (PCR) techniques,
HSV gene products have been identified in 71%–81% of cases of recurrent EM. For nonrecurrent
EM, this falls to 27%. It is postulated that HSV antigens incite a T-cell-mediated
delayed-type hypersensitivity reaction that generates Interferon –α, with the amplified
immune system recruiting more T-cells to the region. Cytotoxic T-cells, natural killer
cells, and/or cytokines destroy the epithelial cells.[1] On the other hand, drug-induced EM was a mechanistically definite condition in which
keratinocytes were positive for tumor necrosis factor-ALFA, a sign of toxic injury.
These findings furnish the mechanistic support for prior clinical and histopathology
observations that these are separate conditions.[6]
Other viral, bacterial, fungal, and protozoal infections and medications may also
play a role. Because it is a hypersensitive reaction, HSV is not cultured from lesions.[1] A genetic predisposition to EM minor may be of importance, as suggested by the familial
tendency that has been documented. Certain Human leukocyte antigen (HLA) phenotypes
may predispose the host to develop this disease in response to an extent of stimuli.
HLA-B62 is found in a high proportion in patients with recurrent EM minor and also
in patients with recurrent HSV infection.[7]
Even though some rare cases of EM minor can be idiopathic,[2] several etiological factors can be associated with its development. Some medicines
or topic contactants, food allergy, hepatitis B virus, HSV, and Epstein-Barr virus
infections, coxsackievirus infections, mumps, streptococcal and mycoplasma pneumonia
(Eaton agent) infections, coccidioidomycosis, Candida, histoplasma, Yersinia, radiation
(mainly the ultraviolet [UV]), dermatomyositis, leprae, diseases as lupus erythematosus,
bowel disease, Wegener's granulomatosis, renal carcinoma, physical agents (Koebner
phenomenon), and acute alcoholism are mentioned as etiological factors. A recent report
also has proposed that rare cases of EM may be induced by cytomegalovirus infection.
There is some evidence that sufferers have a defect in delayed-type hypersensitivity
and a reduced lymphocyte response. The pathogenesis of EM minor may involve an immune
complex-mediated vasculitis.[8],[9]
Herpes-Associated Erythema Multiforme
Herpes-Associated Erythema Multiforme
The literature has suggested a strong association between HSV and EM, especially recurrent
EM.[10]
Investigations associating HSV (1 or 2) as an etiological factor of EM minor were
earlier described in the decades of 30 and 40 of the previous century.[2] The HAEM is a recurrent disease that can be precipitated by sun exposure and does
not progress to SJS.[11] Even in the absence of a clear clinical history of HSV infection, subclinical HSV
is likely the precipitating factor, as confirmed by the PCR analysis of HSV. Before
PCR studies were performed, it was evaluated that 15%–65% of EM are secondary to HSV
infection and that a significant proportion of idiopathic EM was related to subclinical
HSV infection. PCR studies actually have been able to detect HSV DNA in 36%–75% of
EM.[1] Suggesting an explanation for the physiopathology of these lesions, some authors
hypothesized that HSV is engulfed by macrophages at the site of the HSV lesion that
precedes HAEM development. These phagocyte cells are nonpermissive for HSV replication,
resulting in a degradation of the viral DNA and dissemination of fragments to peripheral
skin.[12],[13] HSV DNA fragments with a whole DNA polymerase gene (Pol) are deposited at different anatomical skin sites where Pol is expressed. Activated T-cells are recruited to the site of Pol expression resulting in an inflammatory cascade.[14] The skin from HAEM lesions was positive for the viral Pol gene in 86% of acute lesions. However, it was not seen in uninvolved skin, adjacent
to the HAEM lesions. The skin of healed HAEM from 1 to 3 months were shows PCR positive
for the viral pol gene.[13]
In agreement with Imafuku et al.,[15] the viral DNA is cleared from the skin within 1–1.5 months of HSV lesion resolution,
whereas HAEM lesional skin is still positive 1–3 months after healing. Still in agreement
with these authors, the positive HSV DNA was detected in keratinocytes, germinative
cells, and epithelial cells from the outer root sheath of the hair follicle and in
the epithelial cover for sensory nerve endings. Using in situ reverse transcription
PCR, these authors also observed the RNA signal in keratinocytes within the basal
and spinous epidermal layers with a distribution similar to that of the viral DNA.
This signal was cytoplasmic, probably reflecting the RNA function in translation.
The Pol RNA was observed in acute but not healed HAEM lesional skin that was positive only
for Pol DNA. Therefore, the HAEM lesion development is related with Pol gene expression.[15] A few years ago, an international group of investigators started a large case–control
study, the Severe Cutaneous Adverse Reactions (SCAR) study, to determine the risk
factors for EM, SJS, and TEN.[16] The SCAR study was a multinational case–control study conducted through extensive
surveillance networks of about 1800 hospital departments and 120 million inhabitants
of France, Germany, Italy, and Portugal from February 1, 1989, to July 31, 1995. The
results of this study on a large number of patients confirm that EM on the one hand
and SJS and TEN on the other hand behave as dissimilar disorders, occurring in patients
with different demographic characteristics, presenting with different clinical patterns
and with different risk factors.[16]
Therefore, in the current knowledge, the EM spectrum, which includes EM minor usually
associated or not to HSV or others infections, can be separate from the spectrum of
SJS (EM major) and of TEN [17] that frequently are associated with drug exposition.
Few authors consider the SJS as a subclass of the EM major;[17] however, such classification was not considered in our work due to the small number
of authors with this concept. Most of the researches link SJS as a synonym of the
EM major.[2] In agreement with the current literature in the EM minor, the skin or mucous surfaces
or both simultaneously can be affected. Nevertheless, only one mucous membrane is
affected, usually the oral mucosa, and none additional systemic involvement is present.
This revision study considers the EM minor like a distinct entity from SJS and TEN,
could be associated or not to HSV.
Clinical Characteristics
Lesions of the EM minor can be persistent (continuous), cyclical (acute and self-limiting),
or recurrent; the cyclical and recurrent occur mainly in the HAEM.[18] The condition can begin with nonspecific prodromal symptoms such as headache, malaise,
and fever. Symptoms last from 3 to 10 days, after which an inflammatory process yields
the pathognomonic target or “iris” lesion. The EM minor skin lesions usually caused
by herpes simplex are predominantly raised and distributed on the extremities and/or
the face, with mucosal erosions involving one or several sites. On the other hand,
lesions that are widespread flat atypical targets or macules plus blisters were mostly
drug induced.[19] In the HAEM, HSV lesions can herald the appearance of target lesions by 2–17 days.[10] Mainly in cases of primary HSV infection, there are frequently systemic signs and
symptoms preceding the lesions, and the oral ulcers are typically much smaller.[19] The EM minor lesions in HAEM can reach about 200 or more, evolve over 24–48 h, and
are usually fixed and symmetrically distributed for about a week. These lesions also
attacking more that one of the mucous surfaces could also happen simultaneously with
the cutaneous involvement. In the other EM minor-induced lesions, the target lesions
typically appear on the cutaneous surfaces, including palms, soles, and extensor aspect
of extremities and less often on the face and neck. The lesions begin as erythematous
papules, expanding 2–3 cm in diameter with a dusky purple center, a pale middle zone,
and an erythematous border. Burning, pruritus, as well as central blistering or crusting
may occur.[10] However, these lesions also may occur in one or more rarely in several mucosal surfaces.[20] When the mucous surfaces are affected, the oral mucous membrane is commonly the
most affected, being present in 25%–50% of all EM minor patients.[19] Hemorrhagic crusting of the lips and ulceration predominantly of the nonkeratinized
mucosa characterize oral lesions. When it affects the lips, it results in erosions
or serum-hemorrhagic crusts, with pathognomonic blood-stained crusting of erosions
on swollen lips, hindering the phonation, the feeding, and limiting the oral movement.[2] The intraoral lesions attack mainly in the anterior part, being the tongue and the
buccal mucous membrane, the more involved places.[21] Although any place can be affected, the hard palate [21] and the gum are usually preserved (only 16% of the patients).[2]
Other mucous membranes that can be affected, mainly in the HAEM cases, are the eyes,
nose, genitalia, esophagus, and respiratory tract.[19] The ocular lesions are of particular concern because they can result in scarring
and progressive blindness.[2]
Histological Findings
The characteristic histopathological change of EM minor is epidermal cell death, which
is termed “satellite cell necrosis,” mimicking apoptotic cell death. Among some apoptosis
inducers, the perforin, a pore-making granule from natural killer cells, has been
suggested.[22]
Another apoptotic mechanism that can also be related is the altered expression of
apoptotic regulatory proteins. The intense expression of Bcl-2 protein by the inflammatory
cells in EM minor supports a role for this protein in the maintenance or persistence
of the infiltrate in submucosa. An altered or increased expression of Fas antigen
throughout the epithelium in correlation with the inflammatory cell infiltrates has
been reported in many skin diseases including EM minor.[11] Some apoptosis inducers (i.e., viral infections and glucocorticoids) are common
causative agents of EM minor. Epidermal cell death is also a characteristic feature
of SJS and TEN. However, compared with SJS and TEN, apoptosis was far less in EM minor,
maybe imply a better prognosis.[22] In the early lesions <24 h old, direct immunofluorescence showed an unspecific granular
deposition of IgM, IgG, or C3 in the blood vessel walls of the upper dermis. The transient
production of immune complexes plays an important role in the pathogenesis of this
disease.[23] It has been proposed that the ulcerative inflammatory lesions of the EM minor may
be the result of ischemic necrosis of epithelium as a consequence of immune-mediated
vasculitis.[11]
Treatment
Before any treatment is prescribed, possible underlying causes, such as medications,
diet, infections, or systemic diseases should be determined and eliminated.[10] The prophylactic and therapeutic use of acyclovir, in cases of HAEM, is a common
practice.[5] HSV lesions can precede the appearance of target lesions by 2–17 days, and intermittent
therapy with acyclovir at a dosage of 200 mg twice daily for 5 days, beginning at
the first aura of HSV infection (i.e., local tingling and burning), can prevent and
minimize the symptoms of EM.[10],[24] In patients who have recurrent EM associated with HSV, suppressive treatment using
acyclovir (400 mg twice a day for 6 months) has also been effective in preventing
recurrence. Newer generation antiherpes drugs such as valacyclovir hydrochloride and
famciclovir are also useful in both intermittent and suppressive therapy.[2],[10] The acyclovir administration at the onset of clinical symptoms does not prevent
the EM episode. It is possible that, by the time clinical symptoms are recognized,
sufficient viral replication has already happened to induce a host response to the
virus.[10] Therefore, once onset the earliest symptoms, there is no effective treatment.[25] In addition, because EM is self-limited, symptomatic therapy with antiseptics, antihistamines,
and analgesics is recommended.[10] The oral psoralen plus UV-A (PUVA) therapy has proven to be an equally effective
treatment and it is anticipated that it can be used as a long-term maintenance therapy
without undue concern for adverse effects. The oral PUVA therapy consists of methoxsalen
and exposure of the hands and feet or the whole body to UV-A radiation using a regular
schedule of three treatments each week. With the remission of the lesions, the treatment
may be decreased to weekly exposures for maintenance in some months. However, generalized
exacerbation of the eruption may be triggered by PUVA therapy since EM can occur as
a photodermatosis.[26] It has been showed that childhood HAEM may be unresponsive to therapy with oral
acyclovir. In this case, corticosteroids should be considered as a mode of treatment.[5] However, some authors believe that treatment with corticosteroids is not indicated
in HAEM. Even though systemic corticosteroid therapy is frequently used to treat recurrent
EM and it may partially suppress the disease, it may also make HAEM episodes more
frequent, prolong the duration of attacks, and is associated with side effects.[10],[25],[27] The use of topical and systemic corticosteroids, though, is debatable. The antimalarials
(mepacrine or hydroxychloroquine) have been shown to be occasionally useful when acyclovir
treatment failed [27] and azathioprine can be used as a last resort to suppress an acute attack in patients
with severe disease who do not respond to the other measures.[8],[10],[28] However, it is recommended as second-line treatment due to its side effects.[27] If this treatment fails, mycophenolate mofetil can be tried. It has been shown to
be an effective and relatively safe immunosuppressive agent in recurrent EM; however,
its use is limited by its high cost.[29]
Controversies
There were no particular etiologic agent, no particular clinical feature, no specific
investigations, and no particular treatment pertaining to specific etiology of the
lesion. If this mystery is solved, this can result in proper diagnosis and treatment
of this enigmatic lesion.
Conclusion
In spite of several factors implicated, the exact etiology of EM minor is still unresolved,
and although several attempts have been made, no specific criteria exist for its diagnosis.
The specific pathogenic mechanisms, as well as the multifactorial development hypothesis
of the lesions, are still being investigated. The treatment, except for the symptomatic
therapy with antiseptics, analgesics, and antibiotics, is still being adapted for
prophylaxis, control, and elimination of the possible related underlying causes.
