Dina Halegoua-De Marzio, MD, is an attending gastroenterologist and transplant hepatologist in practice at Thomas Jefferson University Hospital in Philadelphia, PA. Halegoua-DeMarzio will oversee the Case Conference blog, working with the fellows at Thomas Jefferson for the most interesting and engaging cases of the month.

Disclosure: Halegoua-DeMarzio reports she was an consultant for Intercept Pharma.

BLOG: 36-year-old male with drug reaction with eosinophilia and systemic symptoms


This constellation of symptoms now known as DRESS syndrome was originally recognized as early 1959, once thought to be a “pseudolymphoma”. The term “DRESS” was finally coined in a 1996 report. In a literature search over a span of 12 years, 172 cases were reported, involving 44 drugs. Over half of the reported drugs were associated with only a single reported case. The likely offending drug was determined with high probability in almost 80% of the cases. Based on the summative understanding of these cases, it is generally thought that medications taken for more than 3 months or less than 2 weeks are unlikely to be the culprit agents.

The symptoms of DRESS typically begin 2 to 8 weeks after continued usage of the culprit medication. Fever, lymphadenopathy, malaise, and the characteristic rash are the most frequent symptoms at presentation. Lymphadenopathy is reported in as many as 30% to 60% of patients. While fever and lymphadenopathy are significantly associated in most cases, so are hypereosinophilia (or atypical lymphocytosis) and liver involvement. The hallmark rash is a morbilliform rash that evolves to a uniform erythema. Typically, half of the body surface area is involved and about a quarter of patients will progress to develop an exfoliative dermatitis. Facial edema that can include significant pain of mucus membranes can occur in 50% of cases.

One organ is involved in approximately 90% of patients, with two or more organs involved in 50% to 60% of patients. The liver is most commonly involved, in 60% to 80% of cases with hepatomegaly and jaundice. Liver biopsy, when done, demonstrates necrosis of hepatocytes with granulomatous infiltrates containing eosinophils. Given that the liver is the most common organ involved, severe hepatitis accounts for the majority of deaths associated with DRESS. Kidney involvement affects 10% to 30% of patients typically as acute interstitial nephritis, usually seen in cases related to allopurinol. Older age and preexisting renal dysfunction are predisposing factors. Lung involvement is seen in 5% to 25%. Cough, fever, shortness of breath and hypoxia are the usual symptoms. Other organ systems may be involved, but less commonly.

Symptoms can persist or even worsen despite discontinuation of the culprit drug. Some of the drugs frequently associated include allopurinol, carbamazepine, lamotrigine, phenytoin, sulfasalazine, vancomycin, minocycline, dapsone and sulfamethoxazole (as in this patient). RegiSCAR is a European registry of severe cutaneous adverse reactions (SCAR), such as Stevens-Johnson syndrome, toxic epidermal necrolysis, and DRESS. RegiSCAR developed a scoring system to grade cases of DRESS as no, possible, probable, or definite in their diagnosis.


DRESS is a drug-induced hypersensitivity reaction. While rare, it is important to recognize, because it is a potentially life-threatening condition. The pathogenesis is still not completely understood, but some hypotheses have been proposed. A drug-specific immune response and concurrent viral reactivation are thought to be key factors specific to this syndrome. In one literature review, 80% of patients tested positive for Human Herpes Virus 6 (HHV-6). In another study, specifically a measurable increase in HHV-6 over time was detected in approximately 60% of patients with DRESS, indicating active viral replication in those patients. This hypothesis suggests that not only does the disease temporally relate to HHV-6 DNA detection in serum, but also that viral replication directly correlates to the phenotype and severity of DRESS. A second hypothesis expands upon the first by looking at other viral relationships and found evidence of not just HHV-6, but also EBV and HHV-7 reactivation in 29 out of 40 patients with DRESS. Again, this study reiterated the concept of a drug triggering an immune reaction that, in turn, triggers viral reactivations by a not yet fully understood mechanism. While viral PCR measurements have no effect on treatment as of yet, they may serve as prognostic markers for disease course or complication.

The mainstay of treatment is identification and immediate cessation of the offending medication. Complete recovery is reported in most patients without severe organ involvement and treated with mainly supportive care. Patients with exfoliative dermatitis will require fluid and electrolyte repletion as well as adequate nutritional support. Per AASLD guidelines, in the treatment of drug-induced liver injury, corticosteroids are not necessarily indicated unless a drug hypersensitivity reaction, such as DRESS, or an autoimmune reaction is suspected. Systemic corticosteroids are of unproven benefit for most forms of drug hepatotoxicity, so there is currently no consensus guideline on their use. There is, however, more data to support the use of steroids in lung and kidney injury. Severe hepatocellular injury may progress to acute liver failure and in a few instances the only effective therapy may be liver transplantation. The mortality rate for DRESS is 5% to 10%, regardless of corticosteroid usage. In cases that recover with supportive care, the expected mean time to recovery is 6 to 9 weeks.

In Summary

  • —  Hypereosinophilia (OR atypical lymphocytosis), liver involvement, fever, rash and lymphadenopathy are significantly associated
  • —  2 to 8 weeks between drug exposure and onset
  • —  RegiSCAR’s scoring system has been designed to grade DRESS cases as no, possible, probable, or definite
  • —  Though there is limited data to support it, common practice is…
    • Check EBV, HHV-6, HHV-7 to check for correlation
    • In severe organ involvement, consider adding steroids


Bocquet H, et al. Semin Cutan Med Surg. 1996;15:250-257.

Cacoub P, et al. Am J Med. 2011;124:588.

Chen YC, et al. J Am Acad Dermatol. 2013;68:459.

Kardaun SH, et al. Br J Dermatol. 2007;156:609.

Picard D, et al. Sci Transl Med. 2010;2:46ra62.

Tohyama M, et al. Br J Dermatol. 2007;157:934.

Walsh SA, Creamer D. Clin Exp Dermatol. 36:6–11.1365-2230.