Copyright 2013 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Bullous pemphigoid (BP) has been previously described to develop after vaccination in 26 patients. Immunoblotting or enzyme-linked immunosorbent assays (ELISAs), which were performed for 7 of these patients, have always shown circulating autoantibodies against BP180 and/or BP230 antigens. A case of anti–laminin-332 mucous membrane pemphigoid (MMP) that developed shortly after a diphtheria tetanus vaccination is described, with a review of the literature on postvaccination BP.
A 29-year-old man developed an acute eruption of oral and cutaneous blisters and erosions 2 days after receiving a diphtheria tetanus vaccination. The histopathological, immunohistochemical, immunofluorescent, ELISA, and immunoblotting assay results were compatible with anti–laminin-332 MMP. The serum autoantibodies reacted with the α3 and β3 subunits of laminin-332. The disease was controlled by administering a combination of glucocorticosteroids and dapsone.
Conclusions and Relevance
The development of acute MMP shortly after a diphtheria tetanus vaccination may have been serendipitous, a result of a nonspecific bystander activation of the immune system, or due to structural mimicry between domains of the toxoid molecule and a subunit of laminin-332.
Vaccine administration is an effective tool to prevent infectious diseases, but its powerful stimulus on the immune system has generated the fear of exacerbating preexisting autoimmune diseases or inducing autoimmune disorders in otherwise healthy individuals.1 Bullous pemphigoid (BP) is the most frequent autoimmune blistering skin disease. It affects predominantly elderly individuals but has also been reported in children. The etiologic characteristics of BP are unknown, but there have been 26 BP cases suspected of having been induced by vaccinations (Table).2- 21 In all of the postvaccination BP cases in which immunoblotting and/or enzyme-linked immunosorbent assay (ELISA) studies were performed, the target antigens were found to be BP180 and/or BP230.3,8,12,16,17 We report a unique case of mucous membrane pemphigoid (MMP) that developed acutely in a 29-year-old man 2 days after he received a diphtheria tetanus (DT) vaccination. The results of laboratory studies were compatible with anti–laminin-332 MMP, which to our knowledge has not been previously described following vaccination.
An otherwise healthy 29-year-old man was admitted to our department as a result of a sudden mucocutaneous eruption of blisters that occurred 2 days after receiving a DT vaccine for a traumatic skin cut. There were no available medical records of previous DT vaccinations, although the pediatric population in Israel is vaccinated regularly several times during childhood. The patient did not have a history of skin diseases or allergies to medications. The superficial cut was 1.5 cm long. It was treated with povidone-iodine solution and closed with adhesive surgical tape strips. There was no wound infection, and the patient did not receive any systemic antibiotics. On physical examination, there were tense blisters along with flaccid blisters and erosions on the face, inguinal area, and lower abdomen (Figure 1). There were also widespread erosions on the oral mucosa (Figure 1). The Nikolsky sign was absent, and the rest of the physical examination had normal results. Histopathologic analysis of a skin blister revealed a subepidermal blister with complete epidermal-dermal separation and a sparse mononuclear cell infiltrate in the papillary dermis (Figure 2A and B). Immunohistochemical staining demonstrated type IV collagen on the dermal floor of the split (Figure 2C). Direct immunofluorescence microscopy showed linear deposits of IgG and C3 along the basement membrane zone. An indirect immunofluorescence test on the patient's serum, which was performed on monkey esophagus, revealed linear deposits of IgG along the basement membrane zone with a titer of 1:160. Indirect immunofluorescence microscopy performed on 1M sodium chloride–split human skin detected linear deposits of IgG along the dermal side of the split with a titer of at least 1:40 (Figure 2D). The results of ELISA (EUROIMMUN Medizinische Labordiagnostika AG kit) did not demonstrate anti–BP180-NC16A and anti–BP230-C-terminal domain antibodies. In addition, ELISA did not detect antibodies against the C-terminal and N-terminal domains of BP180, the entire ectodomain of BP180, and the N-terminal domain of BP230 (M. Hertl, MD, written communication, April 2, 2012). Immunoblotting analysis of the patient’s serum, using human dermal extract as the antigenic source, demonstrated IgG reactivity to an approximately 120-kDa band. The dilutions for all immunoblotting studies were 1:80. Immunoblotting using recombinant human collagen VII as the antigenic target did not detect anti–collagen VII IgG antibodies (K.B. Yancey, MD, written communication, June 29, 2012). Immunoblotting analysis using a purified human laminin-332 (Abcam) as the antigenic target revealed binding of IgG antibodies to 130-kDa and 145-kDa bands (Figure 3). These 2 bands corresponded in their molecular weights to processed α3 and unprocessed β3 laminin-332 subunits, respectively. The α3 subunit of laminin-332 can be processed twice. The first processing step consists of a cleavage of the carboxyl-terminal globular domains G4 and G5 and generation of a fragment of 165 kDa. An additional cleavage within the amino terminal domain IIIa of the α3 subunit generates a fragment of 145 kDa.
A, Nose; B, groin; C, oral mucosa.
A, Specimen showing a subepidermal blister with complete epidermal-dermal separation (hematoxylin-eosin [H&E], original magnification ×100). B, Scant mononuclear infiltrate in the dermal floor of the blister (H&E, original magnification ×200). C, Immunohistochemical staining demonstrating collagen IV at the dermal floor (original magnification ×100). D, Human 1M sodium chloride–split skin indirect immunofluorescence showing linear deposits of IgG at the dermal floor (original magnification ×400).
Immunoblotting showing 130-kDa and 145-kDa bands (arrows), corresponding to β3 and α3 subunits, respectively, of human laminin-332 in serum samples from a patient (A) and a healthy control (B). M indicates marker.
As for the β3 subunit, it is not processed, and it has previously been reported to have a molecular weight of 140 kDa.22 The 2 bands were dissected from the gel, trypsinized, and analyzed by means of tandem liquid chromatography mass spectrometry analysis (LC-MS/MS) performed on the LTQ Orbitrap device (Thermo). Identification was performed with Discoverer software, version 1.3, against the human part of the Swiss-Prot database, against a specific database of human laminin, and against decoy databases using the Mascot search engine. Results of the LC-MS/MS analysis confirmed that the 130-kDa and 145-kDa bands corresponded to the β3 and α3 subunits of human laminin-332, respectively.
The complete blood cell count, routine chemical analysis, urinalysis, and tests for immunoglobulins, antinuclear antibodies, complements C3 and C4, and rheumatoid factor had results within the normal limits. A workup for a possible underlying malignancy, which included chest radiography, abdominal ultrasonography, and tests of serum levels of prostate-specific antigens, all had results within the normal ranges.
Because of severe mucosal involvement, the patient was treated initially for 3 days with intravenous methylprednisolone pulse therapy, 500 mg/d, followed by oral prednisone, 100 mg/d. The dosage of prednisone was then slowly tapered down. A slight mucosal exacerbation occurred when prednisone was tapered to 15 mg/d, with mild ulcerative gingivitis and mild conjunctivitis. Therefore, dapsone, 100 mg/d, was added and complete remission obtained. Further tapering down to 5-mg prednisone led to the appearance of desquamative gingivitis, and the patient is currently receiving prednisone, 10 mg/d, and dapsone, 100 mg/d, and in complete remission.
The Table reviews and summarizes the 26 cases of BP that developed after vaccination that have been reported in the literature. Including our case of MMP, the reported cases comprise 19 male and 8 female patients aged 2 months to 90 years. The most common types of preceding vaccinations were anti-influenza vaccines (9 cases [33%]) and the diphtheria tetanus pertussis (DTP) vaccine (11 cases [41%]). There were 2 cases after a tetanus booster only, and our case occurred after a DT vaccine. In 10 of 11 cases, the DTP vaccine used was the Tetracoq vaccine [Aventis Pasteur], which combines DTP vaccine with a poliomyelitis vaccine and is administered as part of the first regular vaccination in infants. Other vaccines that were combined with the DTP vaccine included hepatitis B (5 cases), Haemophilus influenzae b (4 cases), influenza (1 case), pneumococcus (2 cases), and meningococcus C (1 case). In addition, there were 2 cases in which BP occurred after a BCG vaccine only, 1 case in which it developed after a herpes zoster vaccine only, and 1 case after an anti–swine flu vaccine only. The appearance of the BP lesions occurred between 5 hours and 5 weeks after vaccination. Immunoblotting and/or ELISA studies were performed for 7 patients with BP, demonstrating anti-BP180 antibodies with or without anti-BP230 antibodies in 6 patients and anti-BP230 antibodies without anti-BP180 antibodies in only 1 patient. Anti-BP180 and anti-BP230 antibodies were not detected in our patient, but additional testing revealed the presence of anti–laminin α3 and anti–laminin β3 antibodies. This was congruent with the results of immunohistochemical staining, which demonstrated collagen IV bound to the floor of the blister, and a salt–split human skin indirect immunofluorescence analysis, which demonstrated IgG antibodies at the dermal base. Epidermolysis bullosa acquisita was ruled out by the presence of collagen IV at the dermal floor of the blister, and a negative result of the immunoblotting analysis for collagen VII antibodies. Anti–p-200 pemphigoid was unlikely because of the lack of a 200-kDa band in the immunoblotting analysis using human dermal extract. Also, in contrast to our case, anti–p-200 pemphigoid is characterized by superficial inflammatory infiltrates, usually dominated by neutrophils,23 and has a milder course and a more prompt response to therapy.24 The immunoblotting using human skin extract demonstrated a 120-kDa band, which was initially thought to correspond to the linear IgA bullous dermatosis (LAD-1) antigen, but ELISA performed for the entire ectodomain of BP180 had negative results.
Anti–laminin-332 MMP is considered to be a severe refractory form of pemphigoid presenting often as a cicatricial form but also as a noncicatricial form of MMP.25 The oral involvement was severe in our patient, but there was also marked cutaneous involvement and the lesions did not scar. Anti–laminin-332 MMP is associated with an increased risk for a malignant tumor.26,27 A workup for malignancy in our patient had negative results.
Laminin-332 is a heterotrimeric glycoprotein situated as the basement membrane, consisting of α3, β3, and γ2 subunits. Most of the patients with anti–laminin-332 MMP were reported to have autoantibodies to the α3 and γ2 subunits, and less frequently to the β3 subunit.28 Our patient demonstrated antibodies to both α3 and β3 subunits. The combination of prednisone and dapsone has been reported to be effective in controlling anti–laminin-332 MMP,29,30 as was also the outcome in our patient.
Vaccines have triggered autoimmune phenomena such as the appearance of DNA antibodies, localized disorders that are self-limiting (eg, reactive arthritis), and systemic diseases, some of which are transient (eg, immune thrombocytopenic purpura) and others that induce lifelong disability (eg, systemic lupus erythematosus).31 The possible role of vaccines in autoimmune disease induction is difficult to prove in humans because of the need to perform large, complex, and expensive epidemiological studies. Garcia-Doval et al32 did not find an increased incidence of BP in people 65 years or older following an influenza vaccine, but in 2 of the reported cases of postvaccination BP (1 infant and 1 adult) the patients developed a relapse of BP after having received an additional vaccination.4,18 In another case, no recurrence of BP was observed following readministration of the Tetracoq vaccine, 1 year after the initial development of BP.3 Bullous pemphigoid developing after vaccination is likely to be a rare event among vaccinated individuals, or the relationship might be serendipitous.
In the present case, the appearance of MMP was acute and developed only 2 days after a DT booster vaccination. It has been suggested that postvaccine autoimmunity may be induced by antigen-specific molecular mimicry or by a nonspecific bystander activation.33 It is possible that our patient had preexisting anti–laminin-332 circulating antibodies and the DT vaccination activated the immune system to increase their serum levels and induce an overt disease, ie, “a bystander activation.” Unfortunately, there was no patient serum sample obtained before the onset of the disease available to test this possibility. Alternatively, a putative molecular mimicry based on structural similarity between laminin-332 and the DT toxoids might have been involved to produce cross-reactive autoantibodies. In a quest for a possible structural mimicry between DT toxoids and proteins comprising the basement membrane zone in the human skin, we have searched the DELTA-BLAST algorithm34 in the blastp suite (http://blast.ncbi.nlm.nih.gov/). Our search revealed an alignment score of 50.1 (e-value, 4 × 10−7) between the Clostridium neurotoxin N-terminal receptor binding domain of Clostridium tetani and the G4 domain of the human laminin α3 subunit, both of which belong to the LamG superfamily. The LamG superfamily is a heterogeneous group of proteins that share highly conserved laminin G–like domains that mainly serve as Ca2+-mediated receptors that participate in a variety of biological functions including cell adhesion, cell signaling, and differentiation.35 A study using bacterial-expressed recombinant proteins has suggested that the serum of most cicatricial pemphigoid patients recognizes the G domains of the α3 subunit of laminin-332.36 This may argue for a biological mimicry that might have been operative in our case, but more experimental evidence is needed.
Corresponding Author: Reuven Bergman, MD, Department of Dermatology, Rambam Health Care Campus, Haifa 31096, Israel (firstname.lastname@example.org).
Accepted for Publication: February 19, 2013.
Published Online: May 8, 2013. doi:10.1001/jamadermatol.2013.741.
Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Sezin and Bergman.
Acquisition of data: All authors.
Analysis and interpretation of data: Sezin, Egozi, and Bergman.
Drafting of the manuscript: Sezin and Bergman.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Sezin.
Obtained funding: Bergman.
Administrative, technical, and material support: All authors.
Study supervision: Bergman.
Conflict of Interest Disclosures: None reported.
Additional Contributions: The authors thank the Smoler Proteomics Center at the Technion Israel Institute of Technology, Haifa, Israel, for performing the LC-MS/MS analysis.
Sezin T, Egozi E, Hillou W, Avitan-Hersh E, Bergman R. Anti–Laminin-332 Mucous Membrane Pemphigoid Developing After a Diphtheria Tetanus Vaccination. JAMA Dermatol. 2013;149(7):858-862. doi:10.1001/jamadermatol.2013.741