Multiple drug sensitization syndrome: A distinct phenotype associated with unrecognized Mycoplasma pneumonia infection.

Introduction

Many adverse drug reactions (ADRs), including Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN), are thought to be caused by delayed-type, cell-mediated immune reactions to a single drug or its related compounds. The term multiple drug hypersensitivity syndrome (MDHS) has been used to describe patients who show delayed-type, cell-mediated immune reactions to 2 or more chemically distinct drugs. Despite the seemingly low prevalence of MDHS, these cases pose the fear of progressing to more severe and widespread forms of ADRs unless the culprit drugs are withdrawn. Thus, it is important to identify patient factors that could increase the risk of multiple drug hypersensitivity (MDH).

Although MDHS was initially defined based on a detailed case history alone or skin tests, a recent study clearly shows that MDHS can be most efficiently proven by an in vitro lymphocyte transformation test (LTT).3, 4 Interestingly, sporadic case reports described that MDH or drug reactions were observed associated with Mycoplasma pneumoniae (MP) infection.5, 6, 7 Indeed, there is mounting evidence suggesting that such infections create a favorable milieu for the initiation and progression of ADRs by abrogating regulatory T-cell (Treg) function.8, 9

We recently experienced a patient with MP infection who subsequently developed a morbilliform eruption 10 days after starting therapy. We also present a comprehensive review of all cases of MDHS previously described in the English- and Japanese-language literature.

Case series

Case 1

A 49-year-old woman with no medical history of drug eruptions presented with a 4-day history of erythematous rashes. Ten days before skin eruptions, therapy with loxoprofen sodium and garenoxacin was started for mild upper respiratory symptoms with low-grade fever. On day 4 of her medication use, the eruption began with pruriginous morbilliform exanthema on her axilla and then spread to the entire body. The patient was admitted to our hospital under suspicion of drug eruptions, and these drugs were withdrawn. Physical examination found confluent, morbilliform, or purplish erythema on her cheeks, trunk, and proximal extremities (Fig 1, A and B) associated with painful lingual erosions. Laboratory data were as follows: white blood cell count of 2620/μL with 39% of neutrophils and 50% of lymphocytes; platelet, 11.1 × 104/μL; a titer of particle agglutination (PA) test for MP, 1:2560 (normal, <40), suggestive of an acute infection with MP. Computed tomography scan of the chest was compatible with pulmonary involvement of MP infection (Fig 1, C). A skin biopsy found a mild perivascular infiltrate and scattered necrotic keratinocytes (Fig 1, D). A diagnosis was made of a morbilliform drug eruption associated with MP infection. The patient was treated with 40 mg/d of prednisolone because of the concern for progression to SJS. Prolonged pulmonary involvement was initially treated with clarithromycin, followed with sitafloxacin (Fig 2). Symptoms resolved completely within 14 days. Positive LTT reactions to loxoprofen sodium were repeatedly detected on the 21st day and 1.5 years after onset, whereas those to clarithromycin and sitafloxacin were only detected less than 100 days after onset.

Fig 1

Clinical and histopathologic findings. A, Purplish erythema diffusely distributed to the lower extremities. B, Closer view of the purplish erythema. C, Computed tomography scan of the chest shows bilateral pulmonary homogeneous ground-glass opacity. D, Scant basal vacuolar change with focal necrosis of keratinocyte, dermal edema, and a mild superficial perivascular lymphocyte infiltrate. (D, Hematoxylin-eosin stain.)

Fig 2

Clinical course of cases listed in Table I in relation to their drug intake. PSL, Prednisolone.

Table I

Case characteristics

Case no.	Age, y/sex	Diagnosis/antecedent MP infection	Initial cutaneous manifestations (L or G)	Extracutaneous manifestation	Medication	Onset of skin reaction after culprit drug intake	LTT (SI) (positive >1.8)		Therapy for the disease	Disease outcome	Study
							<100 d	>100 d
1	49/F present case	Maculopapular rash/yes	Stomatitis (L),Maculopapular rash (G),Purplish erythema (L)	Fever, cough, pneumonia	LoxoprofenClarithromycinSitafloxacinGarenoxacin	10 days	8.5	2.4	Oral prednisolone 40 mg/d	Complete remission	Current case
						NA∗	2.6	1.1
						NA∗	2.5	1.8
						10 d	1.7	NT
2	25/F	Maculopapular rash/yes	Urticaria (G),Maculopapular rash (G)	Fever, cough, arthralgia	LoxoprofenAcetaminophenAmoxicillinLevofloxacin	6 d	2.8	NT	Oral prednisolone 30 mg/d	Complete remission	Takeo et al10
						5 d	12.2	NT
						5 d	1.9	NT
						4 d	3	NT
3	47/F	TEN/yes	Maculopapular rash (G),Atypical target lesion (G)	Fever, cough, dyspnea, pneumonia	Sulbactam/ampicillinSulfamethoxazole/TrimethoprimCeftriaxoneCiprofloxacin	20 d	NT	4.9	Oral prednisolone 50 mg/d, plasma exchange	Complete remission	Gomi et al11
						9 d	0.8	3.8
						12 d	1.1	NT
						12 d	0.8	NT
4	37/M	SJS/serologically positive	Oral and genital erosion (L),Acral erythema and bulla (L)	None	DiclofenacAllopurinolColestimideBezafibrateAmoxicillin	15 d	3.5	NT	Steroid pulse therapy (1g × 3 d), oral prednisolone 50 mg/d	Complete remission	Kubota et al5
						11 d	2.4	1.9
						11 d	11.4	2.4
						11 d	1.9	1.9
						NA∗	NT	1.5

G, Generalized; L, localized; NA, not applicable; NT, not tested.

Drugs sensitized after onset of cutaneous lesions.

Cases from review of the literature

In addition to these cases, a comprehensive MEDLINE and Japan Medical Abstract Society search covering the period from 2007 to 2016 was performed to identify published case reports displaying MDHS using the term multiple hypersensitivity and skin rash with the exclusion of DiHS/DRESS (drug-induced hypersensitivity syndrome/drug reaction with eosinophilia and systemic symptoms). We identified 3 additional cases of MDHS in the literature5, 10, 11: several cases without available information on antecedent or underlying viral and MP infections were excluded. Detailed information and clinical course of the cases are shown in Table I and Fig 2. The most common symptoms were maculopapular rashes (n = 3; 75%), SJS/TEN (n = 2; 50%), fever (n = 4; 100%), cough (n = 3; 75 %), and dyspnea (n = 3; 75%). The site of involvement was the extremities in nearly all cases (n = 4) and trunk in 3 cases, whereas mucous membrane was involved in 3 cases.

In case 4, the authors concluded that the SJS could be caused by allopurinol based on the positive LTT rather than an infectious process: both PA and complement fixation titers were negative at onset, but repeated measurements of PA found seroconversion (1:40) on day 31, which subsequently increased to 1:160 on day 86, suggesting that MP infection would have occurred at onset of drug eruption or near the day of the initial presentation.

Most importantly, all cases, except case 4, had antecedent MP infection in common, and rashes developed 4 to 20 days after drug ingestion. In all cases, MP infection was confirmed by a significant increase in PA and complement fixation titers for MP. Three of the 4 cases had at least 1 mucosal surface involved, with 1 case having 2 mucosal surfaces involved (case 4). The skin manifestations ranged from urticarial rashes to SJS/TEN: of note, 2 of the 4 cases progressed to severe ADRs, such as SJS/TEN, suggesting the significantly greater propensity to progress to severe ADRs. LTT was repeatedly performed on different occasions in cases 3 and 4. Most of positive LTT reactions persisted within 100 days after onset. All cases received systemic corticosteroids at 30 to 50 mg/d, followed by the gradual dose reduction over 21 to 35 days in cases 1, 2, and 4 and 30 weeks in case 3. All cases made a full recovery without any short- and long-term sequelae.

Discussion

Skin manifestations are reported to occur in 8% to 33% of all patients infected with MP. Because MP infection may itself present with fever, rash, and extracutaneous manifestations and, therefore, may confound the role of drug hypersensitivity in the development of skin manifestations, it is quite difficult to distinguish between MP-induced exanthema and drug-induced rash, solely on clinical grounds. Although still somewhat controversial, MP infection would serve to predispose individuals to develop drug hypersensitivity reactions by lowering the activation threshold of drug-specific effector T cells, probably owing to dysfunction of Tregs. Our observation that all previously reported MDHS cases with available information on viral and MP infection were associated with MP infection may give an important clue as to the unique features of MP-induced drug rash. Further studies of a large series of cases will determine whether MDHS could be one of unique features of drug eruptions associated with MP infection.

There are some possible mechanisms acting independently or together that may commit an effector T cell population to respond to multiple medications. One mechanism is the defect of Tregs induced by MP infection. MDHS, however, cannot be solely explained by such a functional defect of Tregs, because, in the LTT reaction, depletion of Tregs from peripheral blood mononuclear cells is found to enhance the reaction but not induce MDH.

Another possibility is that MDH could result from activation of cross-reactive effector T cells that can respond to multiple medications. MP-specific T cells could cross-react with multiple medications presented by the peptide-HLA complex on the surface of an antigen-presenting cell, as demonstrated in Epstein-Barr virus infection: nevertheless, the intensity of positive LTT reactions became decreased with time in case 1, suggesting that this type of sensitization may not be necessarily lifelong.

In all cases, the introduction of culprit drugs preceded the development of rashes by 4 to 20 days, making it more likely that these drugs played a central role. However, clarithromycin and sitafloxacin were introduced 2 to 3 days after the development of maculopapular rashes in case 1, which casts doubt on their contribution to the development of these rashes. Our observation that positive LTT reactions to all of the drugs used were not necessarily detected greater than 100 days after clinical resolution in case 1 suggested that MDH could be transiently observed immediately after and during the development of immune responses associated with MP infection. MP infection could serve to enhance the activation of drug-specific T cells by abrogating Treg function or cause de novo sensitization to the drugs used for treatment for MP infection, like clarithromycin, in case 1. Another possible explanation for MDH is that MP infection could result in a decrease in the threshold for activation of drug-specific effector T cells with a low affinity to the drug that are otherwise unresponsive to the drugs in the setting of no antecedent infection. Because any treatment including antibiotics cannot eliminate the organism from respiratory secretions and severe skin symptoms do not respond well to the treatment, antibiotic therapy that may cause MDHS in some patients is not recommended for MP patients with eruptions. Thus, patients presenting with a history of cough, dyspnea, and fever together with MDH should be screened for MP infection.

This report raises awareness that ADRs associated with MP infection can present with MDHS and suggests that this phenotype may often progress to severe forms of ADRs if the withdrawal of the culprit drugs is delayed. We recommend that any patient with drug eruptions exhibiting MDHS, regardless of the presence of respiratory symptoms, be suspected of underlying or preceding MP infection even if asymptomatic. To minimize the risk of unnecessary immunologic sensitization to multiple medications in the setting of MP infection, LTT should be used to predict potential reactivity to multiple medications in MP patients.