Drug reaction with eosinophilia and systemic symptoms (DRESS)

INTRODUCTION — Drug reaction with eosinophilia and systemic symptoms (DRESS) is a severe adverse drug reaction characterized by an extensive skin rash in association with visceral organ involvement, lymphadenopathy, eosinophilia, and atypical lymphocytosis. The clinical presentation is heterogeneous, and the disease course is typically prolonged. Despite the cessation of the offending drug, flares of disease may continue to occur. The latency between drug initiation and onset of disease is prolonged, typically between two to eight weeks. Reactivation of latent human herpesvirus infections is a common observed phenomenon [1-4].

DRESS will be reviewed in this topic. Other types of cutaneous drug reactions, drug fever, and drug allergy are discussed separately.

●(See "Drug eruptions".)

●(See "Exanthematous (maculopapular) drug eruption".)

●(See "Lichenoid drug eruption (drug-induced lichen planus)".)

●(See "Fixed drug eruption".)

●(See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Pathogenesis, clinical manifestations, and diagnosis".)

●(See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Management, prognosis, and long-term sequelae".)

●(See "Acute generalized exanthematous pustulosis (AGEP)".)

●(See "Drug hypersensitivity: Classification and clinical features".)

●(See "Drug fever".)

TERMINOLOGY — Drug reaction with eosinophilia and systemic symptoms (DRESS) is the most widely used term; drug-induced hypersensitivity syndrome (DIHS) is also frequently used [3]. Older terms were previously based on the causative drug (eg, anticonvulsant hypersensitivity syndrome, allopurinol hypersensitivity syndrome, dapsone syndrome).

EPIDEMIOLOGY — DRESS is estimated to occur in 0.9 to 2 per 100,000 patients per year [5,6]. In hospitalized patients, DRESS accounts for 10 to 20 percent of all cutaneous adverse drug reactions [7-9]. DRESS may occur in children, although the incidence is likely to be lower than in adults [10]. The risk of developing DRESS varies from drug to drug. For high-risk, antiseizure medications (table 1), the incidence of DRESS is estimated to be 1 in 1000 to 1 in 10,000 exposures [11].

ETIOLOGY AND RISK FACTORS

Drug exposure — A clear drug trigger can be identified in the majority of DRESS cases (approximately 80 percent) [4]. However, in the remaining 10 to 20 percent, the strength of drug causality is less clear, and in 2 percent of cases, no drug exposure is present [4]. A large proportion of cases (approximately 75 percent) are due to a few high-risk drugs (table 1). These include aromatic anticonvulsants (eg, carbamazepine, phenytoin, lamotrigine), allopurinol, sulfonamide-containing antibacterials, mexiletine, minocycline, and vancomycin [2,4,12]. In the pediatric population, antiepileptics and antibiotics account for the majority of cases of DRESS [10].

Evolving prescription practices also impact the incidence of DRESS due to a specific drug. As an example, the increased use of vancomycin in hospitalized patients resulted in vancomycin being the leading cause of DRESS in certain areas of the United States [13].

There is indirect evidence that the risk of DRESS for certain drugs is dose dependent. As an example, renal impairment has an additive effect to genetic predisposition on the risk of allopurinol-induced DRESS and other severe cutaneous adverse reactions [14]. Similarly, delayed drug clearance and accumulation have been found in cases of DRESS and Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN) induced by phenytoin [15] as well as in cases of DRESS induced by minocycline [16].

Pharmacogenetic susceptibility — Pharmacogenetic studies have found an association between DRESS risk and several human leukocyte antigen (HLA) haplotypes and genetic variants:

●HLA – Association of DRESS with polymorphisms in the HLA genes has been proven for certain medications in particular at-risk ethnic populations (table 2) [17-19]. HLA alleles encode for the structure of the major histocompatibility complex (MHC) and may influence the interaction between the drug or drug metabolite, T cell receptor, and MHC proteins on antigen-presenting cells.

●Genetic polymorphisms – Polymorphisms in genes encoding drug metabolizing enzymes (eg, cytochrome P [CYP] 450, N-acetyltransferase [NAT1, NAT2]) may result in accumulation of the drug or its active metabolites and increased risk of hypersensitivity reactions. As an example, Asian individuals who carry the CYP2C9*3 polymorphisms have an increased risk of phenytoin-induced severe cutaneous adverse reactions [15,16]. Similarly, the "slow acetylator" phenotype associated with variants in NAT genes was found to be a risk factor for sulfonamide-related hypersensitivity reactions [20,21]. However, such pharmacogenetic associations have been found only for a limited number of drugs and are specific to certain ethnic groups.

PATHOGENESIS — DRESS is considered a T cell-mediated hypersensitivity reaction. Although the exact pathogenesis is not fully understood, two main pathogenetic mechanisms are thought to be involved: a drug-specific immune response and the human Herpesviridae reactivation with a subsequent antiviral immune response.

Drug-specific immune response — The role of drug-specific immune response in the pathogenesis of DRESS has been proven based on the patch test positivity to some causative drugs as well as the in vitro demonstration of drug-specific CD4+ and CD8+ T cells that produce large amounts of tumor necrosis factor (TNF)-alpha and interferon (IFN)-gamma [22-26]. During the acute phase of the disease, there is an expansion of circulating activated T lymphocytes of both CD4+ and CD8+ subsets that harbor activation markers and a skewed repertoire of the antigen receptor. An expansion of regulatory T cells has also been demonstrated in the acute phase of DRESS [26-28].

A single-cell RNA sequencing study demonstrated enhanced Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling in activated T lymphocytes from the skin and blood of a patient with refractory DRESS and that circulating central memory CD4+ T cells were highly enriched in DNA from human herpesvirus (HHV) 6B [29]. This study suggests that the JAK-STAT pathway may have a role in the pathogenesis of DRESS and may be a potential therapeutic target.

Reactivation of Herpesviridae — Reactivation of viruses from the Herpesviridae family (eg, HHV-6, HHV-7, Epstein-Barr virus [EBV], cytomegalovirus [CMV]) is a known phenomenon associated with DRESS and occurs in up to 75 percent of patients [30-34]. HHV-6 is the most common virus in the family associated with DRESS, with reactivation reported in 16 to 60 percent of cases, depending on detection methodologies [4,30-32,35]. Multiple sequential reactivation of herpesviruses has been observed in up to 30 percent of cases [30]. Reactivation may occur in a manner similar to that observed in graft-versus-host disease [32].

However, the mechanisms and timing of viral reactivation in relation to the drug-specific immune response have not been clarified, and the role of virus reactivation in the pathogenesis of DRESS remains controversial. One hypothesis is that viral reactivation occurs as a result of an immunodeficiency state [18]. During the acute stage of DRESS, the population of T regulatory cells is expanded, while the number of B cells and the plasma levels of immunoglobulins are reduced, which may facilitate viral reactivation [26,36,37]. An alternative hypothesis is that certain drugs (eg, valproic acid, amoxicillin) may directly increase HHV-6 and CMV replication [38-40].

Since in humans latent viruses can be harbored by cells of the immune system (eg, T lymphocytes, monocytes/macrophages), the reactivation and release of viruses may be considered as an early marker of stimulation of these cells rather than the initiating event in the pathogenesis of DRESS [41].

CLINICAL PRESENTATION

Latency phase — The latency phase (from drug initiation to onset of reaction) typically ranges from two to eight weeks. However, the latency may be shorter in cases attributed to antibiotics, especially beta-lactam, and iodine contrast media [42,43]. The prodromal phase of DRESS is often characterized by nonspecific symptoms, such as fever, malaise, and lymphadenopathy.

Cutaneous and mucosal manifestations — The skin manifestations are the most obvious and are often the first clue to the diagnosis. The eruption starts as a maculopapular eruption that may progress to a coalescing erythema (picture 1A-C). Additional findings include purpura, infiltrated plaques, pustules, exfoliative dermatitis, and target-like lesions. Lesions are symmetrically distributed on the trunk and extremities. Facial edema is striking and present in the majority of cases (70 percent) [4,44]. Nearly 80 percent of patients have a >50 percent involvement of the body surface area (BSA). In rare instances (<3 percent), the rash may be mild or absent [4]. Pruritus may be an accompanying symptom.

Mucosal involvement can be seen in up to 50 percent of cases. However, it is typically mild, in contrast with Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN), and skin detachment is rarely seen.

Systemic symptoms and laboratory abnormalities — Systemic symptoms associated with DRESS include the following [3,4,35,44-48]:

●Fever ≥101.3°F or ≥38.5°C (75 to 90 percent)

●Lymphadenopathy (54 to 65 percent)

●Hematologic abnormalities:

•Eosinophilia >700/microL (82 to 95 percent)

•Leukocytosis (95 percent)

•Neutrophilia (78 percent)

•Lymphocytosis (25 to 52 percent)

•Monocytosis (69 percent)

•Atypical lymphocytes (35 to 67 percent)

●Symptoms and/or laboratory abnormalities related with any visceral involvement (90 percent)

Organ involvement — The extent and type of organ injury varies. Involvement of at least one internal organ occurs in approximately 90 percent of patients. Approximately 35 percent of patients may have two internal organs involved, and up to 20 percent of patients may have more than two-organ involvement [4]. Internal organ involvement may precede the development of cutaneous rash, and the clinical course of the cutaneous manifestations may not parallel that of visceral organs [4,49].

Liver — Liver injury is the most common visceral manifestation of DRESS, occurring in 53 to 90 percent of cases [4,44,46-48]. The majority of liver function test abnormalities are transient and generally mild [49,50]. Patterns of liver injury include cholestatic type (37 percent), hepatocellular form (19 percent), and mixed (27 percent). Up to 50 percent of cases may have elevations of liver enzymes more than 10 times the upper limit of normal [49]. Acute liver failure is rare and may require liver transplantation. In a series of 16 patients with DRESS and acute liver injury, nine patients recovered spontaneously and six underwent emergency liver transplantation [51]. (See "Acute liver failure in adults: Etiology, clinical manifestations, and diagnosis".)

Kidney — The spectrum of kidney injury associated with DRESS ranges from proteinuria to renal failure [4,44,46-48]. Acute interstitial nephritis occurs in 10 to 30 percent of DRESS cases, acute renal failure occurs in up to 8 percent of patients, and up to 3 percent of patients will require either short- or long-term dialysis [4,44,46,52,53]. Risk factors for the development of drug-induced kidney injury include older age and underlying renal or cardiovascular diseases [54]. Renal involvement may be more common in DRESS cases associated with allopurinol [55].

Pulmonary involvement — The symptoms of pulmonary involvement, which occurs in up to 30 percent of patients with DRESS, include shortness of breath and dry cough [4,44,46]. Chest radiographs or computed tomography (CT) scans may demonstrate interstitial infiltrates and pleural effusions. Respiratory complications may include acute interstitial pneumonitis, lymphocytic interstitial pneumonia, pleuritis, and acute respiratory distress syndrome [56-58].

Cardiac involvement — Cardiac involvement (2 to 20 percent of cases) is a severe complication and poor prognostic factor [4,46,59,60]. Clinical presentations include hypotension, tachycardia, dyspnea, and/or chest pain associated with left ventricular dysfunction and electrocardiogram (ECG) changes [59,60]. Cardiac involvement can be categorized into hypersensitivity myocarditis (which may be asymptomatic or associated with nonspecific symptoms and carries a better prognosis) and acute, necrotizing, eosinophilic myocarditis (which is the most severe form and is associated with over 50 percent mortality) [57,60]. Definitive diagnosis is based on endomyocardial biopsy.

Other organs — Central and peripheral nervous system involvement, including Bell's palsy, peripheral neuropathy, aseptic meningitis, cerebral vasculitis, and limbic encephalitis, has been reported in 2 to 8 percent of DRESS cases [2,4,46,52,61-64].

Gastrointestinal involvement is uncommon. It includes gastrointestinal bleeding, esophagitis, colitis, intestinal perforation, cholecystitis, and pancreatitis [44,52,53,65,66].

Rare complications include myositis, thyroid dysfunction, uveitis, and hemophagocytic syndrome [3,4,52,67-73].

Severe DRESS — Severe DRESS has been arbitrarily defined as DRESS cases with severe, life-threatening organ involvement that result in intensive care unit (ICU) admission or death [52]. No clear prognostic predictors on presentation have been determined. Reactivation of human herpesvirus (HHV) 6 and cytomegalovirus (CMV) may be associated with severe DRESS and poorer outcomes [31,74,75].

CLINICAL COURSE — The clinical course of DRESS is variable. There are no reliable markers at presentation to predict outcome. However, viral reactivation and, in particular, cytomegalovirus (CMV) reactivation, detected during the course of the disease, is associated with poorer outcomes [52,74].

Acute phase — The cutaneous eruption and visceral involvement generally resolve gradually after drug withdrawal. In up to 90 percent of cases, the cutaneous eruption persists beyond 15 days and then subsides with desquamation. The average time to recovery is approximately seven weeks, but up to 20 percent of patients may have a prolonged course lasting more than 90 days [46,76]. Factors associated with a prolonged course include evidence of viral reactivation, severe liver involvement, and presence of mononucleosis-like, atypical lymphocytosis [76].

Flare-ups — Relapses or flare-ups are common after resolution of the acute disease, occurring in up to 25 percent of cases weeks to months (median of 4.5 months) after resolution [77]. In most cases, flare-ups are cutaneous in nature, but they may manifest as eosinophilia or liver enzyme elevations [77]. Flare-ups tend to be more common in patients treated with systemic corticosteroids and occur more frequently on rapid tapering of corticosteroids [76,78]. Moreover, concurrent reactivation of human herpesvirus (HHV) 6 during flare-ups has been observed [76].

Relapses can also be triggered by medications that may be structurally different from the initial causative drug, most commonly antibiotics given as empirical therapy for suspected infection. The reason for the cross-reactivity to other drugs frequently shown by patients with DRESS is incompletely understood. It may result from an immune hyperactivation triggered by the initial episode of DRESS that leads to polysensitization to multiple drugs [77,79,80]. As such, routine and unnecessary exposure to antibiotics and other medications during the acute phase is discouraged [80].

Long-term outcomes — Various autoimmune sequelae have been noted in DRESS survivors. These include autoimmune thyroiditis, vitiligo, alopecia areata/universalis, autoimmune hemolytic anemia, lupus erythematosus, and type 1 diabetes [81-83]. Fulminant type 1 diabetes has been estimated to occur in 0.5 percent of DRESS patients and may be related to the carriage of human leukocyte antigen (HLA) B62 in Japanese populations [84].

DIAGNOSTIC APPROACH

When to suspect the diagnosis — DRESS should be suspected in a patient who received a new drug treatment in the previous two to eight weeks and presents with an acute cutaneous eruption associated with systemic involvement, such as fever, lymphadenopathy, eosinophilia, or abnormal organ function tests. In such cases, a careful evaluation of any exposures to new drugs and, in particular, to high-risk drugs (table 1) in the previous two to eight weeks is essential. (See 'Assessment of drug causality' below.)

It is important to recognize that DRESS is a dynamic process, and characteristic features are not all present concurrently. It is also possible, though uncommon, for patients with DRESS to have no or minimal cutaneous involvement (<5 percent), absent eosinophilia, and mild, systemic symptoms or organ involvement [4,46]. In such cases, the diagnosis of DRESS requires a high degree of suspicion and clinical judgment.

Assessment of drug causality — Detailed information on drug exposures in the months preceding the illness should be obtained by the patient or caretaker. Two key considerations in the assessment of drug causality for DRESS are:

●Prolonged latency – The latency between drug exposure and onset of disease is prolonged. The reaction typically occurs two to eight weeks after drug exposure. Medications that are taken for less than two weeks or more than three months before the onset of DRESS are unlikely to be the culprit. In some cases, drugs that have been stopped prior to onset of disease can still be suspected, if the drug or drug metabolite is still present in the body due to a long drug half-life or impaired clearance.

●Exposure to high-risk drugs – The majority of DRESS cases are due to a limited number of medications (table 1).

Causality may be further supported by positive patch tests and/or in vitro tests (eg, lymphocyte proliferation assay). (See 'Tests for delayed drug hypersensitivity reactions' below.)

Laboratory investigations — The initial laboratory evaluation in a patient suspected to have DRESS is aimed at confirming the diagnosis, excluding other conditions that mimic DRESS, and evaluating the extent and severity of visceral involvement (table 3). It includes:

●Complete blood cell count with differential and peripheral blood smear – Peripheral eosinophilia >700/microL and the presence of atypical mononuclear cells support the diagnosis of DRESS. Other abnormalities in the complete blood count, such as leukocytosis, neutrophilia, lymphocytosis, and monocytosis, may also be present [4].

●Liver function tests – Serum level of alanine aminotransferase (ALT) twice the upper limit of normal values and/or alkaline phosphatase greater than 1.5 times the upper limit of normal values on at least two occasions indicate liver involvement [4].

●Kidney function tests – Serum creatinine greater than 1.5 times the base value for the patient on at least two occasions and/or proteinuria above 1 g/day, hematuria, and decreased creatinine clearance indicate renal involvement.

●Cardiac enzymes – Cardiac enzymes are performed to exclude cardiac involvement.

●Amylase, lipase – Pancreatic involvement is considered when amylase and lipase are ≥2 times the upper limit of normal values.

●Serology for viral infection:

•HHV-6, HHV-7, Epstein-Barr virus, cytomegalovirus – Testing for human herpesvirus (HHV) is increasingly performed in patients with suspected DRESS, as viral reactivation is an observed phenomenon in these patients. Moreover, the presence of viral infection may be a marker for chronicity as well as a prognostic marker for disease severity. Testing modalities may include quantitative polymerase chain reaction (PCR) or paired serologies [30-34]. (See "Clinical manifestations, diagnosis, and treatment of human herpesvirus 6 infection in adults", section on 'Diagnosis' and "Human herpesvirus 7 infection", section on 'Diagnosis' and "Infectious mononucleosis", section on 'Detection of EBV' and "Overview of diagnostic tests for cytomegalovirus infection".)

•Viral hepatitis – Serology for viral hepatitis (hepatitis A immunoglobulin M [IgM] antibody, hepatitis B surface antigen, hepatitis B core IgM antibody, hepatitis C viral RNA) may be useful in excluding acute viral hepatitis in patients with abnormal liver function test results. (See "Hepatitis A virus infection in adults: Epidemiology, clinical manifestations, and diagnosis", section on 'Diagnosis' and "Hepatitis B virus: Screening and diagnosis in adults" and "Clinical manifestations, diagnosis, and treatment of acute hepatitis C virus infection in adults", section on 'Diagnosis'.)

Tests for delayed drug hypersensitivity reactions

In vivo tests — Patch testing, intradermal testing, and lymphocyte activation assay may be used to confirm drug causality:

●Patch testing – Patch testing positivity in patients with DRESS ranges from 30 to 60 percent [85,86]. Test positivity is dependent on the type of drug. It is more commonly positive in cases due to carbamazepine, beta-blockers, and proton pump inhibitors. It is invariably negative for allopurinol and sulfasalazine [85,86].

●Intradermal testing – Intradermal testing with a delayed reading has been utilized in anecdotal reports. Concentrations are usually 10- to 100-fold lower than those undertaken for immediate immunoglobulin E (IgE)-mediated reactions. Nonetheless, there is a risk of reaction recurrence [87]. Thus, the test should only be performed in exceptional circumstances.

In vitro tests — In vitro tests are only performed in specialized centers and largely remain a research tool. They include:

●Lymphocyte transformation test (LTT) – It measures the proliferation of T cells after stimulation with a drug in vitro and is best performed during the recovery stage of the disease [88,89].

●Drug-induced cytokine assays (interferon [IFN]-gamma, tumor necrosis factor [TNF]-alpha, upregulation of cell activation markers such as CD69).

●Analysis of cytotoxic potential of effector cells (eg, granzyme B, granulysin, CD107).

Positive LTTs have been reported in approximately 50 percent of DRESS cases [88]. The sensitivity and specificity of the test are 73 and 82 percent, respectively [90].

Drug challenge is contraindicated in the evaluation of the causative drug, due to the possibility of recurrence of DRESS. However, challenge has been prudently attempted in specific situations, such as DRESS occurring in the setting of multidrug treatment for HIV infection or tuberculosis [91].

Imaging studies — Ultrasound and computed tomography (CT) may be required to assess for internal organ involvement. Echocardiography and cardiac magnetic resonance imaging (MRI) might be indicated if cardiac involvement is suspected.

Biopsy — A skin biopsy for histopathologic examination should be performed in patients with suspected DRESS. Liver or kidney biopsy may be performed in patients with severe organ involvement. (See 'Histopathologic findings' below.)

Histopathologic findings — The cutaneous histopathologic features of DRESS are heterogeneous and nonspecific [18]. Main histologic reaction patterns associated with DRESS include interface dermatitis with basal vacuolization (75 percent), eczematous (spongiotic) pattern (40 to 75 percent), and features of vascular damage with prominent endothelial cells, perivascular lymphocytic infiltrate, red blood cells extravasation, and vessel wall destruction (50 percent) [92-95]. Coexistence of more than one reaction is common, occurring in more than 60 percent of cases [92].

Dermal infiltration of eosinophils is seen in 20 to 70 percent of cases and neutrophils in up to 40 percent. Atypical lymphocytes can be present in up to 30 percent of cases [92]. The finding of extensive, necrotic keratinocytes is associated with severe disease [35,92,94,95].

Although not routinely performed, biopsy of other organs involved may show nonspecific, inflammatory patterns:

●Lymph nodes – Histologic features of DRESS lymphadenopathy include partial or complete effacement of nodal architecture by a polymorphous infiltrate of immunoblasts, small lymphocytes, eosinophils, and plasma cells. Occasionally, atypical lymphocytes mimicking lymphoma may be seen [96].

●Liver – Liver biopsy may demonstrate an acute hepatitis injury pattern with lobular inflammation, scattered foci of necrotic hepatocytes, and granulomatous infiltrates containing eosinophils. Portal inflammation and cholestasis may also be seen. Confluent hepatocyte necrosis and lobular disarray due to inflammation and regenerative changes are seen in severe cases [97,98].

●Kidney – Renal biopsy may demonstrate tubulointerstitial nephritis with interstitial edema and infiltrates of lymphocytes, histiocytes, eosinophils, and plasma cells [99].

●Heart – Endomyocardial biopsies are rarely performed. Two patterns may be observed: hypersensitivity myocarditis (which shows an eosinophilic and mixed, lymphohistiocytic infiltrate within the myocytes) and acute, necrotizing, eosinophilic myocarditis (which demonstrates additional features of myocyte necrosis) [59].

Criteria for diagnosis confirmation — The most widely used criteria to confirm or exclude the diagnosis of DRESS are those included in the Registry of Severe Cutaneous Adverse Reactions (RegiSCAR) scoring system, based on the main clinical manifestations of DRESS [4]:

●Fever >101.3°F or >38.5°C (core) or >100.4°F or >38°C (axillar)

●Enlarged lymph nodes in at least two different body areas

●Eosinophilia

●Atypical lymphocytes

●Skin involvement (extent, rash suggestive of DRESS, biopsy)

●Organ involvement (eg, at least twofold elevation of liver enzymes on at least two different days)

●Resolution >15 days

A value between -1 and 2 is assigned to each feature (table 4). The cumulative score ranges from -4 to 9 and defines four levels of certainty regarding the diagnosis of DRESS: excluded, possible, probable, and definite. As some of the variables included in the RegiSCAR DRESS score may not be available when the patient is first evaluated, the score is most useful as a retrospective validation of suspected cases.

DIFFERENTIAL DIAGNOSIS — Due to the heterogeneity of clinical presentation, DRESS is often misdiagnosed [100]. The differential diagnosis of DRESS includes the following:

●Other cutaneous adverse drug reactions – These include exanthematous drug eruptions, acute generalized exanthematous pustulosis (AGEP), and Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN):

•Exanthematous drug reactions typically present with morbilliform/maculopapular eruptions and mild, systemic symptoms (low-grade fever, pruritus, mild eosinophilia), but visceral involvement (picture 2A-B) is usually absent. Moreover, the latency time between drug exposure and disease onset is usually shorter (5 to 14 days) than for DRESS. (See "Exanthematous (maculopapular) drug eruption".)

•SJS/TEN typically presents with a cutaneous eruption characterized by coalescing, erythematous macules with atypical target lesions, blisters, erosions, and skin detachment and is associated with severe, mucosal involvement in multiple sites in over 90 percent of cases (picture 3A-C). (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Pathogenesis, clinical manifestations, and diagnosis".)

•AGEP presents with an extensive eruption of hundreds to thousands of pinpoint, nonfollicular pustules (picture 4) a few hours to a few days after the administration of the offending drug. Systemic involvement may occur in 20 percent of cases but is typically mild. Although pustules may be present in DRESS, they are more commonly limited to the head and neck area. (See "Acute generalized exanthematous pustulosis (AGEP)".)

●Lymphomas – Angioimmunoblastic T cell lymphoma typically presents with acute-onset, systemic illness characterized by a rash (50 to 60 percent of cases), generalized lymphadenopathy, hepatosplenomegaly, and B symptoms of fever, night sweats, or weight loss. The cutaneous histology may not demonstrate the presence of neoplastic cells. Diagnosis is confirmed via lymph node biopsy. (See "Clinical manifestations, pathologic features, and diagnosis of angioimmunoblastic T cell lymphoma".)

Sézary syndrome, an aggressive, leukemic variant of cutaneous T cell lymphoma, typically presents with erythroderma and generalized lymphadenopathy and may occasionally mimic DRESS. The diagnosis of Sézary syndrome is based on the evaluation of a skin biopsy for morphology, immunophenotype, and T cell receptor gene clonality of cellular infiltrates. Evaluation of the peripheral blood with flow cytometry and clonality studies may be needed to confirm the diagnosis. (See "Clinical presentation, pathologic features, and diagnosis of Sézary syndrome".)

●Acute cutaneous lupus erythematosus – The acute onset of rash, especially if widespread, and the frequent association with systemic symptoms and laboratory abnormalities may occasionally mimic DRESS (picture 5). The diagnosis of acute cutaneous lupus erythematosus is confirmed by autoantibody testing and histologic examination of skin biopsy that demonstrates interface dermatitis with increased mucin and thickened basement membrane. Direct immunofluorescence may show a continuous band of granular fluorescence at the dermoepidermal junction. (See "Overview of cutaneous lupus erythematosus" and "Clinical manifestations and diagnosis of systemic lupus erythematosus in adults".)

●Hypereosinophilic syndromes – Hypereosinophilic syndromes are characterized by elevated, peripheral eosinophilia (>1500/microL) and involvement of multiple organs, such as the heart, gastrointestinal tract, lungs, brain, and kidneys, without an alternative explanation for the organ damage. Cutaneous manifestations include eczematous eruptions, erythroderma, urticaria, and angioedema. (See "Hypereosinophilic syndromes: Clinical manifestations, pathophysiology, and diagnosis".)

●Viral infections – Examples of viral infections associated with generalized rash that may mimic DRESS include infectious mononucleosis, cytomegalovirus (CMV) infection, measles, HIV infection, viral hepatitis, and dengue.

MANAGEMENT — There are no randomized trials evaluating treatments for DRESS, and evidence-based guidelines are lacking. Because of the high heterogeneity of the clinical manifestations of DRESS, the approach to management is generally based on the severity of skin and organ involvement [78,101,102]. Patients with widespread rash and severe, systemic symptoms should be hospitalized for evaluation and treatment. Organ support and intensive care unit (ICU) care may be needed. Patients with mild symptoms can be managed in an outpatient setting with symptomatic treatment and close clinical and laboratory monitoring for possible organ involvement.

Drug withdrawal, supportive treatment, and monitoring — The identification and withdrawal of the causative medication is the mainstay of treatment for all patients with DRESS (see 'Assessment of drug causality' above). Introducing new medications, including empirical use of antibiotics, should be avoided, if possible. Since the cross-reactivity among aromatic antiseizure medications is well documented, patients with DRESS triggered by carbamazepine or other aromatic antiseizure medications should be treated with nonaromatic agents (eg, valproic acid, topiramate, gabapentin).

Supportive care includes fluid, electrolyte, and nutritional support. Adjunctive measures include gentle skin care with emollients and warm baths/wet dressings.

Regular clinical, laboratory, and imaging monitoring and timely consultation with specialists (eg, hepatologist, nephrologist, pulmonologist) are warranted.

Mild disease (no organ involvement or only mild liver involvement) — Patients with DRESS without clinical, laboratory, or imaging evidence of renal or pulmonary involvement and those with only modest elevation of liver transaminases (<3 times the upper limit of normal) can be treated symptomatically [103,104].

For symptomatic relief of pruritus and skin inflammation, we suggest treatment with topical corticosteroids. High- or super high-potency topical corticosteroids (groups 1 and 2 (table 5)) are applied two to three times a day until disease resolution.

Evidence on the efficacy of topical corticosteroids for the management of DRESS without severe or life-threatening organ involvement is limited. In small case series, patients with DRESS without life-threatening organ involvement who were treated with topical corticosteroids recovered and had a lower incidence of relapse, viral reactivation, and sepsis compared with patients treated with systemic corticosteroids [78,105].

Severe disease with single or multiple organ involvement

Systemic glucocorticoids — For patients with severe disease and involvement of the lungs (dyspnea, abnormal chest radiograph, hypoxemia) or kidneys (creatinine >1.5 times the basal level and proteinuria or hematuria), we suggest oral glucocorticoids as first-line therapy. A moderate to high dose (0.5 to 1 mg/kg per day) of prednisone or prednisone equivalents is given until clinical improvement and normalization of laboratory parameters are achieved and then slowly tapered over the following 8 to 12 weeks, or longer, to avoid relapses [103]. An alternative regimen of intravenous methylprednisolone (250 to 500 mg per day for two to four days) followed by oral prednisolone 1 mg/kg per day tapered over 8 to 12 weeks has also been suggested for severe cases [106,107].

The benefit of systemic glucocorticoids for patients with DRESS and isolated, severe liver involvement is unproven [51]. Patients with signs of acute liver failure (eg, hepatic encephalopathy, coagulopathy) need prompt referral to a liver transplant specialist for further evaluation and care. (See "Drug-induced liver injury", section on 'Management'.)

The efficacy of systemic glucocorticoids for the treatment of DRESS has not been evaluated in randomized trials. However, there is general consensus among experts on their use for patients with severe DRESS, particularly for those with renal and/or pulmonary involvement [45,47,103,106,108].

As systemic glucocorticoids may increase the risk of cytomegalovirus (CMV) reactivation and systemic infections, careful monitoring of infectious complications should be performed [74,78,82,109-111]. (See "Major adverse effects of systemic glucocorticoids".)

Second-line therapies — Second-line therapies for patients with severe DRESS and organ involvement include cyclosporine, intravenous immunoglobulins (IVIGs), and other immunosuppressive agents.

Cyclosporine — Oral cyclosporine at a dose of 3 to 5 mg/kg divided twice daily for seven days, followed by a taper over the next 7 to 14 days, may be given as a second-line therapy for patients with DRESS and severe organ involvement who do not respond to systemic corticosteroids or for patients in whom corticosteroids are contraindicated.

Although evidence remains limited, there are increasing reports on the rapid resolution of symptoms with a short course of oral cyclosporine [112-115]. In a retrospective study that included 5 patients with DRESS treated with cyclosporine and 21 patients treated with systemic corticosteroids, treatment with cyclosporine was associated with a more rapid resolution of fever, normalization of laboratory parameters (leukocytosis, eosinophilia, and liver enzymes), and shorter duration of hospitalization [114].

Intravenous immunoglobulin — There is little evidence to support the use of intravenous immunoglobulins (IVIGs) for DRESS. Although a beneficial effect of IVIG has been reported in a few adult and pediatric patients [116-118], an open study evaluating the efficacy of IVIG in six adult patients with DRESS was prematurely stopped due to an excess of adverse events in five patients, four of whom needed rescue therapy with oral glucocorticoids [119].

Other immunosuppressive agents — There are anecdotal reports of the use of the Janus kinase (JAK) inhibitor tofacitinib at a dose of 10 mg/day for the treatment of refractory, severe DRESS [29,120,121]. The rationale for the use of JAK inhibitors for DRESS is based on the finding of Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling upregulation in DRESS [29]. However, further evidence is needed before tofacitinib can be recommended for severe DRESS.

In a few isolated case reports, cyclophosphamide has been used in patients with severe organ involvement refractory to glucocorticoid therapy [122,123].

Antiviral therapy — The routine use of antiviral medication for DRESS is not recommended because of the spontaneous resolution of viral reactivation in most cases and concern about the toxicity associated with antiviral agents. However, if there is confirmation of viral reactivation and evidence of viral-induced organ damage, or viral reactivation is suspected to be a contributory factor to severe complications (eg, pneumonitis, colitis, encephalitis), antiviral therapy with ganciclovir or valganciclovir is warranted. (See "Ganciclovir and valganciclovir: An overview".)

PROGNOSIS AND FOLLOW-UP — Most patients with DRESS recover completely in weeks to months after drug withdrawal. DRESS survivors should undergo long-term monitoring for the development of autoimmune sequelae. (See 'Long-term outcomes' above.)

The mortality rate among patients with DRESS is estimated to be between 2 and 10 percent [4,44-47]. A mortality rate of 3 percent has been reported in a pediatric population [10]. Older age, severe organ involvement, and multiorgan failure are the main predictors of mortality [45]. Viral reactivation of cytomegalovirus (CMV) during the course of the disease and development of CMV-related complications (eg, hepatitis, colitis) appear to be associated with poorer outcomes [52,74,109]. (See "Epidemiology, clinical manifestations, and treatment of cytomegalovirus infection in immunocompetent adults".)

PREVENTION — For primary prevention of DRESS, unnecessary prescribing of high-risk drugs should be avoided (table 1) [124]. Genetic screening for human leukocyte antigen (HLA) alleles known to be associated with increased risk of severe drug reactions in specific populations may reduce the risk of DRESS (table 2) [124-126]. This was demonstrated in a Taiwanese study that included 2926 patients who had an indication for allopurinol therapy, of whom 571 (20 percent) tested positive for HLA-B*58:01 and received an alternative drug. Among the HLA-B*58:01-negative patients who received allopurinol, no severe cutaneous drug reactions were reported, although seven cases were expected based on the estimated historical incidence nationwide [125].

Patients who recover from DRESS should be educated about the need for strict avoidance of the offending drug and cross-reacting drugs. Moreover, as DRESS survivors may have an increased risk of reaction to structurally unrelated drugs in the months following the acute episode, avoidance of any unnecessary drug treatments is also recommended. Drug allergy labeling needs to be entered in the patient's medical record.

Because of pharmacogenetics susceptibility, avoidance of causative drugs should also be recommended to family members of patients.

SOCIETY GUIDELINES — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Drug allergy and hypersensitivity".)

SUMMARY AND RECOMMENDATIONS

●Definition, etiology, and risk factors – Drug reaction with eosinophilia and systemic symptoms (DRESS) is a rare, potentially life-threatening, drug-induced hypersensitivity reaction characterized by a cutaneous eruption, hematologic abnormalities, lymphadenopathy, and/or internal organ involvement. The majority of DRESS cases are due to antiseizure medications, allopurinol, antibacterial sulfonamides, minocycline, and vancomycin (table 1). Association of DRESS with polymorphisms in the HLA genes has been proven for certain medications in particular at-risk ethnic populations (table 2). (See 'Etiology and risk factors' above.)

●Latency and clinical presentation – In most patients, the reaction begins two to eight weeks after the initiation of the causative drug:

•Cutaneous findings – The skin eruption starts as a maculopapular eruption that may progress to a coalescing erythema (picture 1A-C). Additional findings include purpura, infiltrated plaques, pustules, exfoliative dermatitis, and target-like lesions.

•Systemic symptoms – Systemic symptoms may include fever ≥101.3°F or ≥38.5°C, lymphadenopathy, hematologic abnormalities (eg, leukocytosis, eosinophilia, neutrophilia, atypical lymphocytosis), and signs and symptoms related with visceral involvement (table 3). (See 'Clinical presentation' above.)

•Organ involvement – Involvement of one or multiple organs occurs in most patients with DRESS. Liver injury is most common, occurring in up to 90 percent of cases. Acute interstitial nephritis, interstitial pneumonia, and, less commonly, myocarditis can also occur. (See 'Organ involvement' above.)

●Diagnosis – The diagnosis of DRESS is based on a combination of clinical features (cutaneous findings (picture 1A-C), systemic symptoms), history of exposure to drugs and, particularly, to high-risk drugs (table 1) in the previous two to eight weeks, and laboratory and imaging findings (table 3). Criteria to confirm or exclude the diagnosis of DRESS are those included in the Registry of Severe Cutaneous Adverse Reactions (RegiSCAR) scoring system (table 4).

●Treatment – Identification and withdrawal of the causative drug and supportive treatment are the mainstays of treatment for patients with DRESS. Regular monitoring for organ involvement is warranted (see 'Drug withdrawal, supportive treatment, and monitoring' above):

•For patients with mild disease without organ involvement or only modest elevation of liver transaminases (<3 times the upper limit of normal), we suggest symptomatic treatment of skin inflammation and pruritus with topical corticosteroids (Grade 2C). High or ultra-potent topical corticosteroids (groups 1 and 2 (table 5)) are applied two to three times per day until resolution of the skin eruption. (See 'Mild disease (no organ involvement or only mild liver involvement)' above.)

•For patients with severe disease and involvement of the lungs or kidneys, we suggest oral glucocorticoids as first-line therapy (Grade 2C). A moderate to high dose (0.5 to 1 mg/kg per day) of prednisone or prednisone equivalents is given until clinical improvement and normalization of laboratory parameters are achieved. Systemic glucocorticoids should be tapered slowly over 8 to 12 weeks. Oral cyclosporine is an alternative treatment for patients with severe organ involvement who do not respond to glucocorticoids or for patients in whom corticosteroids are contraindicated. (See 'Severe disease with single or multiple organ involvement' above.)

●Prognosis – Most patients with DRESS recover completely in weeks to months after drug withdrawal. They should also undergo long-term monitoring for the development of autoimmune sequelae. (See 'Prognosis and follow-up' above.)

●Prevention – DRESS survivors should be educated about the need for strict avoidance of the causative drug and cross-reacting drugs. (See 'Prevention' above.)