Approach

Despite EM being a frequently mild and self-limiting disease, a thorough history and clinical examination are essential to determine if there are any triggers that can be avoided in the future.[41]​ More serious mucocutaneous skin diseases such as Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) also need to be ruled out.[3]

  • EM minor has cutaneous manifestations only, while EM major has cutaneous manifestations and involves 1 or more mucosal sites.

  • The lesions of both EM minor and EM major involve <10% total body surface area.

History

A detailed history of recent infection, prior recurrences, and new drug administration is necessary. Causes should be carefully investigated before any patient is labelled as having an idiopathic cause.[2]​​[41]

  • Infections: the most commonly associated infections are herpes simplex virus (HSV) infection and infections caused by Mycoplasma pneumoniae.[39] Other infectious agents reported to trigger EM include cytomegalovirus, Epstein-Barr virus, SARS-CoV-2, hepatitis B virus, hepatitis C virus, influenza virus, HIV, herpes zoster, gardnerella, histoplasmosis (with concomitant erythema nodosum), coccidioidomycosis, orf (a disease of sheep and goats caused by a parapox virus that can be transmitted to humans), and syphilis.[2][7][10]​​​[11][12][13][14]​​​​​​[15][16]​​[37][38]​​​​​​​​ ​

  • Drugs: associated medications include certain antibiotics; docetaxel or paclitaxel; immune checkpoint inhibitors; sorafenib; tumour necrosis factor (TNF)-alpha inhibitors; antimalarials; hydroxychloroquine; lenalidomide; methotrexate; anticonvulsants; statins; bisphosphonates; non-steroidal anti-inflammatory drugs (NSAIDs); metamizole; oral contraceptives; imiquimod; lidocaine; triclocarban; barium contrast.[1][18][19][20][21][22][23][24][25]​​[26][27]​ However, this list is not exhaustive and you should check your local drug information source. Photo-distributed lesions have been noted with phenylbutazone, triclocarban, paclitaxel, and statins.[22]​​​[23]

  • Vaccines and allergens: vaccines against organisms such as hepatitis B, smallpox, varicella, meningococcus, human papillomavirus, and SARS-CoV-2 have been known to elicit EM.[14]​[28][40]​​ Allergic response to contact allergens, such as tattoos, can also trigger EM.​[29]

The lesions typically appear a few days after the trigger. Some lesions will initially clearly resemble target lesions; others will evolve from small erythematous plaques. The lesions have a rapid onset and usually increase in number over 4 to 7 days. They can cause general discomfort but are not itchy until they start to heal. Oral mucosal involvement may be particularly painful for the patient, and in more severe cases causes a reduced ability to take fluids and food.

Clinical examination

EM presents with typical target lesions (annular erythematous rings with an outer erythematous zone and central blister sandwiching a zone of normal skin tone) and atypical targetoid papules (no central blistering). The clinical pattern of the lesions is the most important diagnostic tool, with the characteristic target lesions commonly found in a symmetrical distribution on the extremities. Targetoid lesions are more common centripetally.[2][3][4][5]​​ In the presence of target lesions, targetoid lesions corroborate the diagnosis of EM, when they occur rapidly. Mucosal membranes of the mouth, eyes, nose, and genitalia should also be examined for mucosal erosions seen in EM major. Some typical target lesions and minimal mucosal disease is the finding most suggestive of EM, especially in the setting of HSV infection or Mycoplasma pneumoniae infection.

A more general physical examination should also be carried out to identify any possible infectious cause. HSV infection is characterised by clustered vesicles on an erythematous base. A red tympanic membrane strongly suggests Mycoplasma pneumoniae, as do rhonchi, rales, and/or wheezes.[Figure caption and citation for the preceding image starts]: Palmar target lesionsFrom the personal collection of Nanette Silverberg, MD; used with permission [Citation ends].com.bmj.content.model.Caption@714dbd5[Figure caption and citation for the preceding image starts]: Target and targetoid lesionsFrom the personal collection of Nanette Silverberg, MD; used with permission [Citation ends].com.bmj.content.model.Caption@64b12d29[Figure caption and citation for the preceding image starts]: Target lesions on the face and mucosal erosions with crusting due to HSV-1 recurrenceFrom the personal collection of Nanette Silverberg, MD; used with permission [Citation ends].com.bmj.content.model.Caption@3d9869de

Laboratory evaluation

Most cases of EM can be diagnosed by history and clinical examination alone, and no further investigations are needed. However, if there is diagnostic uncertainty after clinical examination, a biopsy for haematoxylin and eosin staining can be performed. If the result of this biopsy is not conclusive, a biopsy for immunofluorescence can also be carried out.

If the cause of EM is not apparent from clinical examination, laboratory tests are performed in order to try to establish a cause. Since the most common infections to trigger EM are HSV and mycoplasma, the initial tests are full blood count, electrolytes, HSV serology, cold agglutinins, M pneumoniae titres, and/or chest x-ray (depending on the clinical state of the patient). If these tests are negative, then tests for other less common infectious causes are carried out. Herpes zoster-associated EM can be differentiated from generalised extension of herpes by rapid polymerase chain reaction.[12]

HSV serology can also be useful if there have been recurrent episodes of EM but there are no specific HSV lesions found. Anti-desmoplakin antibodies have been noted in patients with recurrent EM.[36]

Distinguishing Stevens-Johnson syndrome and toxic epidermal necrolysis from EM

Other more severe reaction patterns should be ruled out, including those of SJS and TEN. Severe limitation of oral intake and pain with voiding are more common with these diseases, but can be seen with EM major.[2][3][4][5][9][42] SJS affects <10% total body surface area and tends to have extensive oral and genital mucosal involvement. Often, an offending drug exposure is identified. TEN demonstrates extensive denudation of the skin, generally more than 30%. If it is difficult to differentiate a suspected case of EM from SJS or TEN, 2 tests can be performed:

  • The Asboe-Hansen sign of TEN is typified by the physical enlargement of blisters with direct pressure applied to the top of the blister, demonstrating basal keratinocyte necrosis. SJS and TEN also have the Nikolsky's sign of skin slough on touch, not present in EM.[2]

  • Biopsy and fresh frozen tissue assessment can demonstrate necrotic keratinocytes in SJS and TEN. Monocytic infiltrates and red blood cells are more typical on histopathology of EM; there is no epidermal necrosis and the inflammatory infiltrate is within the dermis.[2][3][4][5][32][42] ​​

    See Stevens-Johnson syndrome and toxic epidermal necrolysis (Diagnosis Approach).

Use of this content is subject to our disclaimer