Cyclosporine as an alternative immunosuppressant for steroid-resistant drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome

Background

Drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome is a type IV drug-induced hypersensitivity characterised by fever, rash, lymphadenopathy, visceral organ involvement and eosinophilia.1 Cutaneous manifestations of DRESS syndrome are by far the most common presenting symptom and can vary from urticarial rash to desquamating or bullous eruptions.1–3

Visceral involvement is also varied and can be multisystem, involving the gastrointestinal system, predominantly the liver and pancreas, kidneys, lungs, heart, thyroid and brain, with approximately half of cases having multiple organ involvement. The most common organ involved is the liver, followed by kidneys and lungs.1 3 Presentation of visceral involvement is varied and often non-specific. Renal involvement commonly presents with proteinuria and interstitial nephritis.2 4 Pulmonary manifestations can present as mild cough and dyspnoea, eosinophilic pneumonia or acute respiratory distress syndrome.5 Visceral involvement is the major cause of morbidity and mortality of DRESS syndrome, and delays in diagnosis and treatment can result in life-threatening outcomes.1 2

While hepatic involvement is the most common gastrointestinal manifestation of DRESS syndrome, other less common gastrointestinal manifestations include oesophagitis, gastritis, colitis, pancreatitis and autoimmune pancreatic involvement resulting in diabetes mellitus.6 Within the gastrointestinal tract, the pancreas was the second most commonly affected organ after the liver, with both early and late pancreatic sequelae associated with DRESS syndrome. Gastrointestinal involvement in DRESS syndrome may also be under-reported due to the non-specific nature of presenting symptoms including diarrhoea, nausea and vomiting.6

There are many inciting medications associated with development of DRESS syndrome. Anticonvulsant medications including carbamazepine and lamotrigine, and antibiotics including minocycline, sulfonamides and allopurinol are among the commonly associated medications with DRESS syndrome.7 Sulfasalazine is a well-known culprit from the reporting pharmacovigilance databases.7

The pathophysiology of DRESS syndrome is still poorly understood and is considered to consist of multiple cellular interactions. The activation of T cells and subsequent production of interleukin-5 (IL-5) is thought to be a primary driver of the eosinophilic predominant hypersensitivity reaction that is observed in DRESS syndrome.1 8 Genetic enzyme deficiencies and genetic associations between human leucocyte antigens and drug hypersensitivity reactions may also predispose to development of DRESS syndrome.8

Resolution of DRESS syndrome typically occurs within weeks of ceasing the offending agent. Systemic immunosuppression is recommended in cases with visceral organ involvement.7 9 Oral corticosteroids are considered the first-line immunosuppressive agent.7 In cases where steroids are contraindicated, or symptoms are steroid resistant or steroid dependent, an alternate immunosuppressive agent should be considered.9

While there is evidence to support the use of other steroid-sparing immunosuppressive agents, including tacrolimus, mycophenolate mofetil and azathioprine in the treatment of classic autoimmune hepatitis, no such literature describes the use of these agents in management of relapsing DRESS syndrome with primary hepatitis involvement.

Case presentation

A woman in her 30s presented with a widespread maculopapular skin eruption involving >50% of her body surface area, fever, oral ulceration and loss of taste 5 weeks after commencing sulfasalazine for a newly diagnosed seronegative arthritis. Based on the clinical picture and initial investigations, a diagnosis of ‘probable’ DRESS syndrome secondary to sulfasalazine was made.2 The patient was commenced on oral prednisolone 50 mg daily and the sulfasalazine ceased, with initial good effect.

Multiple attempts to wean prednisolone resulted in consequent flares in dermatological symptoms, as well as further derangement of liver function tests (LFTs). Prednisolone wean was gradual and protracted, with decreasing prednisolone dose by 5 mg at intervals of at least 4 weeks provided LFTs remained stable. Over the subsequent 9 months, the patient had a chronic relapsing and remitting course of DRESS syndrome that, while responsive to high-dose oral corticosteroid, flared significantly with further cutaneous symptoms and LFT derangement with any attempt to wean below 20 mg daily.

Investigations

Initial laboratory investigations revealed an eosinophil count at the upper limit of normal (0.5×10⁹/L). Remainder of the full blood examination and kidney function tests were within normal limits. There was mild hepatitis with alanine aminotransferase (ALT) 158 U/L, aspartate aminotransferase (AST) 55 U/L, alkaline phosphatase (ALP) 83 U/L, gamma-glutamyl transferase (GGT) 117 U/L and a normal total bilirubin (10 µmol/L). A skin biopsy demonstrated subacute spongiotic reaction pattern with dermal eosinophil infiltration, compatible with a drug reaction. Notably, her LFTs were normal at the time of commencing sulfasalazine.

Upon weaning prednisolone below 20 mg daily, the skin eruption flared significantly, and there was an acute deterioration in LFTs: total bilirubin 170 µmol/L, ALT 1614 U/L, AST 522 U/L, GGT 38 U/L and ALP 90 U/L. There was also now overt eosinophilia (peak level 1.0×10⁹/L). Liver screen was non-revealing for a cause, with negative viral hepatitis serology (including hepatitis A, B and C, Ebstein-Barr virus and cytomegalovirus) and a negative autoimmune hepatitis serology panel.

An abdominal ultrasound showed a normal-appearing liver with no features suggestive of steatosis, cirrhosis or portal hypertension. A liver biopsy, already potentially modified by corticosteroids, demonstrated a moderate severity portal and lobular chronic hepatitis with limited interface hepatitis, lymphocytic infiltrate with rare eosinophils, without fibrosis or plasma cells to suggest classic histological findings of autoimmune hepatitis (figures 1–3). Given the clinical history and presentation, the histopathological findings were most in keeping with hepatic sequelae of DRESS syndrome.

Differential diagnosis

The combination of cutaneous manifestations and drug-induced liver injury is seen in cases of DRESS syndrome, Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN).10 11 While the association between cutaneous and hepatic involvement is most common in DRESS, SJS and TEN can result in devastating and life-limiting outcomes. SJS and TEN are therefore important differential diagnoses that must be excluded. SJS and TEN classically present with more severe dermatological manifestations with less severe hepatic involvement.11

Treatment

On the basis of a small number of published case reports, the patient was commenced on cyclosporine 5 mg/kg daily, leading to normalisation of LFTs over a 2-month period (ALT 525 to 26 U/L, AST 168 to 13 U/L, GGT 50 to 36 U/L) and normalisation of bilirubin (28 to 16 µmol/L).

Given the emergence of steroid-induced side-effects (weight gain, thin skin, easy bruising), budesonide 9 mg orally was subsequently initiated, enabling more rapid weaning and cessation of prednisolone. Once established on full dose of budesonide, prednisolone was successfully weaned from 50 mg, with a dose reduction of 5 mg every fortnight until ceased. There was no observed worsening of LFTs or flare of dermatological symptoms during the prednisolone wean.

Outcome and follow-up

Budesonide was gradually weaned by 3 mg every 2 months until eventually ceased after a 6-month period.

Cyclosporine levels were monitored initially at weekly intervals, aiming levels between 750 and 1000 µg/L. Once therapeutic levels were established, monitoring intervals were extended to monthly for 3 months, and continue at 3-month intervals. Additionally, full blood count, renal function and electrolytes were monitored at the same intervals to monitor for potential complications of renal impairment and myelotoxicity from cyclosporine therapy.

At the time of last review, the patient has remained in a 6-month remission from both a hepatic and dermatological perspective on cyclosporine monotherapy. The patient had not developed any complications from cyclosporine therapy, in particular, no evidence of hirsutism, gingival hyperplasia, nephrotoxicity or neurotoxicity.