Comment
Linear IgA bullous dermatosis is a well-documented autoimmune mucocutaneous disorder characterized by linear IgA deposits at the dermoepidermal junction. The development of autoantibodies to antigens within the basement membrane zone leads to both cellular and humoral immune responses that facilitate the subepidermal blistering rash in LABD.2,3 Linear IgA bullous dermatosis affects all ages and races with a bimodal epidemiology. The adult form typically appears after 60 years of age, whereas the childhood form (chronic bullous disease of childhood) appears between 6 months and 6 years of age.3 Medications—particularly vancomycin—are responsible for a substantial portion of cases.1-4 In one review, vancomycin was implicated in almost half (22/52 [42.3%]) of drug-related cases of LABD.4 Other associated medications include captopril, trimethoprim-sulfamethoxazole, phenytoin, and diclo-fenac.3,4 Vancomycin-associated LABD has a substantially shorter time to onset of symptoms, with a mean of 8.6 days compared to 63.8 days for other causative agents.4Resolution of symptoms also occurs more quickly, with remission occurring in 66.7% (16/24) of cases at a mean time of 13 days compared to a 39.2% (11/28) resolution rate with a mean time of 18.9 days following discontinuation of other implicated medications.4 While idiopathic LABD involves the mucous membranes in up to 80% of cases, drug-induced LABD is less commonly associated with mucosal lesions. In an earlier systematic review from 1966 to 2002, 32% (7/22) of reported cases of vancomycin-induced LABD were reported to have mucosal involvement.5,6 In 2012, one group found that most published cases of drug-induced LABD do not use standardized algorithms, such as the Naranjo algorithm, to definitively tie LABD onset to medication use.4 The Naranjo algorithm, devised in 1981, consists of 10 questions that determine the probability of adverse drug reactions.7 In our case, a Naranjo score of 5 suggested a probable adverse drug reaction due to vancomycin use; however, we cannot completely exclude ciprofloxacin in our case in light of a case report of LABD in the setting of IV vancomycin and ciprofloxacin use.8 In our patient, ciprofloxacin had a Naranjo score of 2, which suggested a possible adverse drug reaction. Allopurinol, which does not have any published association with LABD, also had a Naranjo score of 2 in our patient.
The initial treatment of drug-induced LABD is immediate discontinuation of the suspected agent(s) and supportive care.9 Although future avoidance of vancomycin is recommended in patients with a history of LABD, there are reported cases of successful rechallenges.4,10 The early removal of our patient’s cement spacer was discouraged by both the orthopedics and infectious disease consultation services due to potential complications as well as the patient’s gradual improvement during his hospital course.
Dapsone is considered the standard systemic treatment for LABD. Sulfapyridine is an alternative to dapsone, or a combination of these 2 drugs may be used. Corticosteroids can be added to each of these regimens to achieve remission, as in our case.2 Although dapsone was discontinued in the setting of the patient’s AKI, the vancomycin in the dual-eluting spacer was more likely the culprit. A review of 544 postoperative outcomes following the use of an antibiotic-impregnated cement spacer (AICS) during 2-stage arthroplasty displayed an 8- to 10-fold increase in the development of AKIs compared to the rate of AKIs following primary joint arthroplasty.10 While our patient’s AKI was not attributed to dapsone, his prominent peripheral motor neuropathy with resultant bilateral thenar atrophy was a rare complication of dapsone use. While dapsone-associated neuropathy has been reported in daily dosages of as low as 75 mg, it typically is seen in doses of at least 300 mg per day and in larger cumulative dosages.11
Despite having a well-characterized vancomycin-induced LABD in the setting of known vancomycin exposure, our patient’s case was particularly challenging given the continued presence of the vancomycin-impregnated cement spacer (VICS) in the left knee, resulting in vancomycin levels at admission and during subsequent measurements over 2 weeks that were all several-fold higher than the renal clearance predicted.
Vancomycin-associated LABD does not appear to be dose dependent and has been reported at both subtherapeutic1-3 and supratherapeutic levels,5-9 whereas toxicity reactions are more common at supratherapeutic levels.9 The literature on AICS use suggests that drug elution occurs at relatively unpredictable rates based on a variety of factors, including the type of cement used and the initial antibiotic concentration.12,13 Furthermore, the addition of tobramycin to VICSs has been found to increase the rate of vancomycin delivery through a phenomenon known as passive opportunism.14
As AICS devices allow for the delivery of higher concentrations of antibiotics to a localized area, systemic complications are considered rare but have been reported.13 Our report describes a rare case of LABD in the setting of a VICS. One clinical aspect of our case that supports the implication of VICS as the cause of the patient’s LABD is the concentration of bullae overlying the incision site on the left knee. A case of a desquamating rash in a patient with an implanted VICS has been documented in which the early lesions were localized to the surgical leg, as in our case.15 Unlike our case, there was a history of Stevens-Johnson syndrome following previous vancomycin exposure. A case of a gentamicin-impregnated cement spacer causing allergic dermatitis that was most prominent in the surgical leg also has been reported.16 An isomorphic phenomenon (Köbner phenomenon) has been suggested in the setting of vancomycin-induced LABD lesions that intensified at a site of adhesive tape application,17 but the Köbner phenomenon did not appear to be a major factor in our patient. The removal of the patient’s cement spacer was performed to prevent development of a chronic autoimmune response or autoreactivity state against the skin basement membrane zone structural antigen.