Are You Confident of the Diagnosis?

What you should be alert for in the history
Characteristic findings on physical examination
Expected results of diagnostic studies
Diagnosis confirmation

Bullous drug eruptions are best conceptualized as a heterogeneous group of diseases. These may be classified more precisely based on their integral clinical and histopathologic features. The following drug eruptions fall under this heading:

–Drug-induced pemphigus (DIP)

–Drug-induced bullous pemphigoid (DIBP)

Continue Reading

–Drug-induced linear IgA disease (LAD)

–Lichen planus pemphigoides (LPP)

–Bullous fixed drug eruption (FDE)

–Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN). This topic is covered on its own Web page, see Toxic Epidermal Necrolysis and Stevens-Johnson Syndrome.

Drug-induced pemphigus

Clincally, DIP may present either as pemphigus vulgaris or pemphigus foliaceus. Thiol drugs are responsible for most of the p foliaceus cases. In these patients, small, flaccid bullae eventuate into scaly or exfoliative plaques. Mucous membranes are spared. The diagnosis may not be obvious initially. The other causative drugs (see Etiology) induce a p vulgaris variant. Here, patients present with oral and pharyngeal erosions and the skin may also be involved. The interval between drug initiation and onset of pemphigus may be weeks to months and has been reported as being up to one year.

The differential diagnosis is broad and includes idiopathic pemphigus variants and the other bullous drug eruptions mentioned here.

Occasionally, widespread p vulgaris may mimic TEN clinically. Both will have a positive Nikolsky sign. The correct diagnosis may be made on histopathologic evaluation of the epidermis (so-called “jelly-roll”) or of a skin biopsy along with immunofluorescence studies. Drug-induced p foliaceus variants may mimic seborrheic dermatitis, which may lead to a delay in diagnosis.

Drug-induced bullous pemphigoid

The clinical presentation of DIBP resembles bullous pemphigoid (BP) in the majority of cases. The eruption is typically pruritic. Urticarial plaques may precede the tense bullae, which erupt on these plaques or separate from them. Mucosal and facial involvement may be greater in DIBP than the idiopathic variant. A targetoid appearance on palms and soles has also been reported (Figure 1).

Figure 1.

Bullous drug eruption. This patient was found to have linear IgA bullous dermatosis secondary to therapy with vancomycin. (Courtesy of Bryan Anderson, MD)

Histopathologically and on immunofluorescence, the findings resemble BP. Initially, there is eosinophilic spongiosis and lining up of eosinophils along the dermo-epidermal junction. A subepidermal split is accompanied by a mixed, eosinophil-predominant infiltrate. On direct immunofluoresence, there is linear IgG and C3 at the basement membrane zone (BMZ). Salt-split skin shows positive immunofluorescence in the roof of the bulla. Circulating autoantibodies to BPA-1 (230kd; type XVII collagen) and BPA-2 (180kd) are also found.

The differential diagnosis includes idiopathic BP and other subepidermal autoimmune bullous diseases (that present with tense bullae, such as linear IgA disease and epidermolysis bullosa acquisita) should be considered. When the urticarial phase, but no bullae, is present, the differential diagnosis includes urticarial vasculitis, Sweet’s syndrome, Well’s cellulitis, or non-pigmenting FDE.

Resolution of the eruption may occur rapidly after withdrawal of the offending agent. However, in some cases, referred to as “drug-triggered BP,” the eruption will be ongoing months after discontinuation of the offending drug.

Drug-induced LAD

The majority of cases of LAD are idiopathic. However, drug-induced cases are well recognized. The onset of the eruption is 1 to 15 days after initiation of the offending drug. Clinically, there are tense vesicles that coalesce to form bullae. These may be arranged in an annular or linear grouping. There may be a background of erythema or associated urticarial plaques. Lesions favor trunk and extremities and palms, soles and mucous membranes may be involved, but mucous membrane involvement is thought to occur at a lower rate than in idiopathic LAD. Two patients with morbilliform eruptions without blistering have been described. The classic findings of LAD were seen on direct immunofluorescence specimens in these patients.

Histopathologically there is subepidermal blister formation with a neutrophil-predominant inflammatory infiltrate. The diagnosis is clinched with direct immunofluoresence studies on perilesional normal skin, where linear, homogeneous bands of IgA are seen at the basement membrane zone.

As in the above diagnoses, drug-induced LAD must be distinguished from the idiopathic variant so as to be able to impact prognosis. Other subepidermal autoimmune bullous diseases that present with tense bullae, such as BP and epidermolysis bullosa acquisita, should be considered. Drug-induced linear IgA disease has also been reported to mimic erythema multiforme and TEN.

Bulla formation ceases within 24 to 72 hours, and resolution is usually complete within 2 to 7 weeks of discontinuation of the inciting agent.

Lichen planus pemphigoides

This is a rare disease where features of lichen planus (LP) and bullous pemphigoid (BP) are present. LPP may be drug-induced in some cases (see under Etiology for a comprehensive list). Lichenoid papules and plaques and bullae that arise within them or within normal skin are seen. Systemic corticosteroids are the treatment of choice for LPP. Topical steroids, dapsone, tetracycline and nicotinamide, isotretinoin, and immunosuppressive drugs have also been used.

Histopathologically, the findings of LP and BP are seen. The diagnosis is clinched by the presence of IgG and C3 at the dermoepidermal junction on direct immunofluorescence. Circulating autoantibodies to BPA-1 and BPA-2 are also found.

The differential diagnosis includes the bullous variant of lichen planus. In these cases, severe interface dermatitis leads to separation of the epidermis and clinical bullae. In bullous LP, bullae are confined to lichenoid plaques, whereas in LPP, bullae are seen to arise independently from these plaques. A thorough drug history is needed to differentiate drug-induced LPP from the idiopathic variant.

On cessation of drug therapy, abatement of the eruption occurs over weeks to months.

Bullous fixed drug eruption

This is a distinctive drug eruption that is characterized by reddish or dusky round macules or erythematous, edematous round plaques. Bullae and erosions (especially in areas of trauma) may result. If bullae are prominent, the designation bullous FDE is given. FDE may take days, or up to 2 weeks from initial drug exposure to manifest. On rechallenge, symptoms and signs develop within minutes to hours.

If widespread, bullous FDE may mimic TEN. A clinical clue to the diagnosis of bullous FDE in this case is the round edges of the bullae (as opposed to the irregular, jagged edges in TEN). Histopathologic differentiation may be required.

Histopathologically, there is a vacuolar interface dermatitis that leads to separation of the epidermis and a subepidermal bulla. There is associated pigment incontinence and melanophages in the dermis. Individual apoptotic keratinocytes are seen in the epidermis. In the dermis, upper dermal edema and a mixed inflammatory infiltrate, often with prominent eosinophils, is seen. The infiltrate occurs in the superficial and mid-dermis.

The diagnosis of these eruptions can only be made by obtaining a thorough drug history in all cases of newly diagnosed blistering disease. The subtype of bullous eruption is diagnosed by considering the typical clinical, histopathologic and immunofluorescence studies for each. The generic term “bullous drug eruption” should be avoided as a diagnosis.

Histopathologically, there are no specific features that set drug-induced pemphigus apart from the idiopathic variants. In p foliaceus-like DIP, there is acantholysis in the upper epidermis (in or adjacent to the granular layer). The roof of the bulla is commonly not visualized and acantholyis may be subtle. In p vulgaris-like DIP, there is suprabasal acantholysis. In both types, eosinophilic spongiosis may be seen. The diagnosis is clinched by direct immunofluoresence where IgG is seen intercellularly within the epidermis. C3, IgM and IgA are seen less frequently.

Who is at Risk for Developing this Disease?

P vulgaris has been associated with the presence of HLA DRB1*0402, DRB1*1401, DQB1*0302, DRB1*14 and DQB1*0503. DRB1*0402 was identified in all 6 DIP patients in one study. There is no gender predominance and the age of patients has ranged from 30 to 83.

No associated HLA subtypes have been reported with DIBP, drug-induced LAD or LPP to date. FDE is linked to the presence of HLA-B22 and HLA-Cw1.

Patients with drug-induced LAD tend to be older than those in the idiopathic group. The risk of vancomycin- induced LAD has been found not to be dose-dependent and its severity does not correlate with serum vancomycin levels.

Drug-induced and non-drug-induced LPP occurs in a younger age group than BP. Typically patients in the 4th or 5th decade of life are affected, although pediatric cases have also been reported.

Fixed drug eruptions accounted for 3% of 97 cases of drug eruptions seen in hospitalized patients in one study. Most patients with FDE are adults between the ages of 20 and 40 years, but children, and even infants have been reported to develop it. There is a trend towards male predominance.

What is the Cause of the Disease?

DIP: The first drug to be identified as causing pemphigus was penicillamine. Since then, many other drugs have been implicated. These may be broken down into 3 groups, according to the chemical structure of the drug.

1. Thiol or SH drugs (those containing a sulfhydryl radical): penicillamine, captopril, gold sodium thiomalate, penicillin and piroxicam.

2. Phenol drugs: cefadroxil, rifampin, levodopa, aspirin.

3. Non-thiol non-phenol drugs: a miscellaneous group including calcium channel blockers, angiotensin-convert­ing enzyme (ACE) inhibitors, non-steroidal anti-inflammatory drugs (NSAIDS), interferon-α and interleukin-2.

DIBP. Furosemide is the most commonly reported drug culprit, but antimicrobials (penicillins, terbinafine), sulphasalazine, PUVA therapy and more recently adalimumab have also been implicated.

Drug-induced LAD. Vancomycin is the most commonly reported culprit, however, other antibiotics such as ampicillin and trimethoprim/sulphamethoxazole as well as lithium, ACE inhibitors, diclofenac, furosemide, G-CSF and phenytoin may also be inciting agents.

Drug-induced LPP. Drugs that have been implicated in the causation of LPP include captopril, cinnarizine, ramipril, simvastatin, PUVA, and antituberculous medications.

Bullous FDE. Rifampin, paracetamol, metronidazole, paclitaxel, erythromycin, and ibuprofen have all been reported to cause bullous FDE.


DIP. On a pathophysiologic level, acantholysis has been induced by thiol drugs in vitro. While the majority of thiol-induced cases show circulating and tissue-bound antibodies, a small subset of cases do not demonstrate these findings. It has been hypothesized that biochemically-induced acantholysis exposes antigenic determinants that then induce auto-antibody formation in these cases.

A primary immunologic pathomechanism seems to be operable in ACE inhibitor-induced cases; 52% of 63 patients on ACE-inhibitors, who had no clinical DIP, demonstrated a positive indirect immunofluorescence with antibodies directed to the cytoplasm of keratinocytes in one study. All 48 controls showed no such findings. Autoantibodies to desmoglein 1 and 3 were found in 8 of 10 patients with DIP.

Thiol-induced DIP has been reported to remit spontaneously in half of cases as opposed to only 15% of non-thiol-induced DIP.

While the pathomechanism of DIBP has not been completely elucidated, it is proposed that drug haptens bind to and alter antigenic determinants in the BMZ, thereby inducing autoantibodies. In the case of PUVA, ultraviolet light is thought to injure and thereby alter antigenic determinants in the BMZ. Circulating antibodies to BPAg-1 and 2 are present as in idiopathic BP.

In LAD, IgA autoantibodies are directed against a 97 kDa ectodomain of BPAg-2, known as LABD97. Autoantibodies to LABD97, BPAg-1 and BPAg-2 have been found in 2 drug-induced cases.

In drug-induced LPP, it is thought that drug-induced damage to the basement membrane from the lichenoid interface dermatitis exposes antigenic determinants and stimulates an autoimmune response that initiates BP.

In bullous FDE, it is thought that the inciting agent bind to basal keratinocytes or occasionally melanocytes, as a hapten. Intercellular adhesion molecule-1 (ICAM1) is released, which upregulates CD8+T lymphocytes. CD8+ T cells are subsequently activated and give rise to a cellular cytotoxic response.

The Fas ligand (FasL) pathway has been shown to play a major role in the resultant apoptosis of keratinocytes. Cytokines, such as interferon-g and tumor necrosis factor- α, augment the tissue damage. Interleukin-15 expression from keratinocytes acts in the formation of drug-specific CD8+ memory T cells so that when re-exposure to drug occurs, a more rapid and localized response develops.

CD4+ T lymphocytes also play a role in that they produce interleukin-10, which assists in a dampening down of the CD8+ – initiated response at the time of resolution of the lesions.

Systemic Implications and Complications

Systemic involvement does not typically occur in these patients. In widespread cases of drug-induced pemphigus vulgaris, or in those patients with severe oral disease who cannot eat, hospitalization and attention to fluid and electrolyte balance as well as caloric intake is needed.

Treatment Options

Table 1. Medical treatments for bullous drug eruptions

Table 1.
All eruptions Drug-induced pemphigus vulgaris Drug-induced bullous pemphigoid Drug-induced lichen planus pemphigoides and bullous FDE
Stop offending drugPotent topical steroids Orally: mucosal protectants (eg, milk of magnesia); topical anesthetic agents (eg, viscous lidocaine); potent topical steroids; intralesional triamcinolone Systemically: oral corticosteroids, mycophenolate mofetil, intravenous immunoglobulin (IVIG), rituximab Oral corticosteroids,tetracycline and nicotinamide, dapsone, mycophenolate mofetil, azathioprine Oral corticosteroids

Optimal Therapeutic Approach for this Disease

Of prime importance for the prognosis of these patients is to consider a drug cause in the pathogenesis of these auto-immune blistering diseases. A meticulous drug history and knowledge of the causative drugs is needed. The offending drug should be discontinued promptly. This may lead to resolution within weeks in some patients. Complete resolution may be prolonged up to months or a year in some cases of LAD, DIBP and DIP, especially if thiol-induced.

Potent or ultrapotent topical steroids (bethamethasone diproprionate 0.05% or clobetasol proprionate 0.05%, or equivalent) are first-line treatment for symptom relief. This is usually adequate therapy for drug-induced LAD, drug-induced LPP and bullous FDE, which will resolve within days to weeks after drug withdrawal.

For drug-induced pemphigus, withdrawal of the drug does not lead to immediate abatement in most cases. Therefore, standard therapeutic regimens for pemphigus vulgaris need to be considered. One such regimen is oral prednisone at a dose of 1mg/kg/day in concert with mycophenolate mofetil (starting at 1g twice daily and increasing by 500mg/month to a maximum of 3g/day). Rituximab with or without IVIG may be considered for disease that is not controlled or worsening on this regimen.

Orally, mucosal protectants (eg, milk of magnesia), topical anesthetic agents (eg, viscous lidocaine) and potent topical steroids (clobetaol proprionate 0.05% ointment) and magic mouthwash (usually a combination of diphenhydramine, viscous lidocaine and aluminum hydroxide) may be used. Intralesional triamcinolone injections at a concentration of 10mg/cc can be given into erosions on the buccal mucosa. For p.foliaceus variants, oral corticosteroid treatment may be sufficient to control disease.

If there are no systemic contraindications, a short tapering course of oral corticosteroids (eg, prednisone starting at 0.5 to 1mg/kg/day, depending on extent of disease) may be used in drug-induced BP. Tetracycline (500mg 4 times daily) and nicotinamide (500mg 3 times daily) may be used as first-line oral treatment in the more elderly and for those in whom oral corticosteroids are contraindicated. Dapsone may be used as a steroid-sparing agent in neutrophil-rich histopathologic subtypes. Mycophenolate mofetil and azathioprine have also been successfully employed as steroid-sparing agents in BP.

For severe drug-induced LPP or generalized bullous FDE, a short course of oral corticosteroids at a dose of 0.5 to 1mg/kg/day may be given.

Patient Management

As stated above, it is crucial to bear medications in mind as a cause of these bullous diseases. The prognosis should be discussed with the patient and his or her family. The above therapies may be introduced based on the severity of disease and may be tapered according to response.

The side effects of oral corticosteroids and/or immunosuppressive agents should be discussed in depth with the patient and his or her family. The decision to begin such therapies should be made after consideration of the patient’s comorbidities and after full consideration of the risk/benefit ratio.

Unusual Clinical Scenarios to Consider in Patient Management

Rarely, generalized bullous FDE and drug-induced LAD have mimicked TEN. In such cases, monitoring in an intensive care unit setting is recommended until an accurate diagnosis is reached.

What is the Evidence?

Armstrong, AW, Fazeli, A, Yeh, SW, Mackool, BT, Liu, V. “Vancomycin-induced linear IgA disease manifesting as bullous erythema multiforme”. J Cutan Pathol. vol. 31. 2004. pp. 393-7.

Billet, SE, Kortuem, KR, Gibson, LE, El-Azhary, R. “A morbilliform variant of vancomycin-induced linear IgA bullous dermatosis”. Arch Dermatol. vol. 144. 2008. pp. 774-8.

Waldman, MA, Black, DR, Callen, JP. “Vancomycin-induced linear IgA bullous disease presenting as toxic epidermal necrolysis”. Clin Exp Dermatol. vol. 29. 2004. pp. 633-6. (These three papers detail three different morphologic variants of drug-induced linear IgA disease.)

Brenner, S, Bialy-Golan, A, Ruocco, V. “Drug-Induced Pemphigus”. Clin Dermatol. vol. 16. 1998. pp. 393-397. (A comprehensive review article on DIP.)

Cozzani, E, Rosa, GM, Drosera, M, Intra, C, Barsotti, A, Parodi, A. “ACE inhibitors can induce circulating antibodies directed to antigens of the superficial epidermal cells”. Arch Dermatol Res. 2010 Jun 20. (A clinical study showing that autoantibodies to epidermal cells without clinical manifestations are associated with ACE inhibitor therapy.)

Harting, MS, Hsu, S. “Lichen planus pemphigoides: a case report and review of the literature”. Dermatol Online J. vol. 12. 2006. pp. 10(A useful article detailing the clinical manifestations of LPP and providing a comprehensive literature review.)

Kirtschig, G, Middleton, P, Bennett, C, Murrell, DF, Wojnarowska, F, Khumalo, NP. “Interventions for bullous pemphigoid”. Cochrane Database Syst Rev. vol. 10. 2010. pp. CD002292( A systematic review of randomized controlled trials for the treatment of bullous pemphigoid.)

Landau, M, Brenner, S. “Histopathologic findings in drug-induced pemphigus”. Am J Dermatopathol. vol. 19. 1997. pp. 411-4. (A clinical study where five dermatologists examined specimens of drug-induced and non-drug-induced pemphigus, and found no differences between the two groups.)

Lee, JJ, Downham, TF. “Furosemide-induced bullous pemphigoid: case report and review of literature”. J Drugs Dermatol. vol. 5. 2006. pp. 562-4. (A review of drug-induced BP.)

Strowd, LC, Taylor, SL, Jorizzo, JL, Namazi, MR. “Therapeutic ladder for pemphigus vulgaris: Emphasis on achieving complete remission”. J Am Acad Dermatol. 2010 Jul 13. (The response of 18 patients to a therapeutic regimen used at the authors’ institution for the treatment of pemphigus vulgaris.)