Histological findings in BSLE are relatively consistent across cases and resemble lesions seen in DH. Bullae demonstrate separation of the epidermis from the basement membrane zone (BMZ). The epidermal roof is typically intact and the blister cavity contains fibrin and abundant neutrophils. There is also a neutrophil-dominated inflammatory infiltrate of the upper dermis [16] and dermal edema. In some cases the infiltrate is most pronounced in the dermal papillae as seen in DH, and in other cases it is distributed evenly in a band beneath the BMZ as seen in LAD [22]. Some monocytes and eosinophils are normally present within the infiltrate, and in some cases there is histological evidence of necrotizing vasculitis demonstrated by leukocytoclasis, extravasation of erythrocytes and vessel necrosis [21]. Basal keratinocyte vacuolisation, BMZ thickening and epidermal atrophy characteristic of primary SLE lesions are generally absent in the lesions of BSLE.

Immunoglobulin deposits in the upper dermis and the BMZ are a consistent finding in BSLE. Peri-lesional and unaffected skin demonstrate deposits of IgG, IgA, IgM and complement at the BMZ under direct immunofluorescence (DIF). A review of immunohistochemistry in 30 cases of BSLE reported IgG is present in 93 % of cases and IgM and IgA are present in approximately 70 % of cases [23]. Complement is present in 77 % of cases and has been reported to be more commonly observed in lesional skin than in clinically uninvolved skin [16]. Two major patterns of immunoglobulin deposition have been described with a granular pattern present in approximately 60 % of cases and a linear pattern present in approximately 40 % [23]. In some cases, thready or fibrillar deposits have been reported or a mixed pattern compromising a band of homogenous deposits punctuated by scattered granular deposits [16]. Interestingly, irrespective of the pattern of immunoglobulin deposition, the clinical and histological features of BSLE are consistent.

The results of indirect immunofluorescence (IIF) in BSLE patients correlate with findings on DIF. In those patients with granular immunoglobulin deposition on DIF, IIF is negative. In those patients with a linear pattern of immunoreactants on DIF, IIF on normal skin is usually negative but positive on sodium chloride-split skin, most showing dermal binding [24].

Immunoelectron microscopy (IEM) studies demonstrate immunoglobulin deposition as a continuous band of granular reaction products in the upper dermis beneath the lamina densa. Occasionally, deposition is observed on the lamina densa and in the deeper dermis, the region where anchoring fibrils are seen [16]. The epidermo-dermal cleavage plane is usually the area of immunoglobulin deposition [20].

Immunoblotting has demonstrated reaction between antibodies in the sera of BSLE patients and 290- and 145-kDa autoantigens extracted from normal human dermis [25, 26]. These autoantigens have been identified as components of type VII collagen [27], the target antigen in EBA. Fusion proteins of smaller components of the collagen VII NC-1 domain were created and tested by both immunoblotting and ELISA by one of us (DM), and two epitope regions were identified within the type III fibronectin repeat region [12]. These epitope regions were the same ones that were recognised by the EBA sera in the study, at that time not explaining the clinical differences in phenotype of the two conditions. The epitopes were independently verified by Jones using 8mer peptides from within these two epitope regions, confirming cross-reaction of BSLE sera and EBA sera with two peptide regions within each of the two previously defined epitope regions [28]. Four epitope regions within the NC1 domain were also found by Woodley’s group in addition to reactivity against the collagenous domain adjacent to the NC2 domain [13].

Camisa and Sharma [29] first proposed diagnostic criteria for BSLE in 1983 which were later revised after the administration of salt-split skin immunofluorescence [30]. The criteria essentially necessitated (1) a diagnosis of SLE by ARA criteria, (2) a vesiculobullous eruption, (3) subepidermal histopathology consistent with DH and leukocytoclastic vasculitis, (4) negative or positive IIF for BMZ autoantibodies and (5) positive DIF at the BMZ.

Gammon and Briggaman then divided patients into two types depending on the presence of antibodies to type VII collagen. Those patients with antibodies to collagen VII identifiable by IIF or IEM, usually with linear or mixed patterns of immunoglobulin deposition, were designated type I BSLE or BSLE-I [16]. Patients who did not demonstrate antibodies to type VII collagen were designated type II BSLE or BSLE-II. Clinically it is not possible to distinguish between the two types of BSLE; rather, subtyping is performed through immunohistochemistry.

A third type of BSLE was proposed by Yell et al. in 1995, who also suggested revised criteria for diagnosis [2]. They highlighted that there existed BSLE patients with classical clinical and histological features whose sera bound to epidermal rather than dermal epitopes and that these patients should be designated type III BSLE.

As our understanding of the pathogenesis underling BSLE evolves, these criteria are likely to change accordingly. However, at present, a diagnosis of BSLE is still made according to the revised criteria put forward by Camisa’s group with further subtyping performed according to the results of immunofluorescence, immunoblotting and ELISA.

BSLE is a rare bullous dermatosis in patients with SLE. It is characterised by clinical and histological features resembling BP or DH and a heterogeneous immunological profile, usually characterised by autoimmunity to components of type VII collagen, much like EBA. As our understanding of the pathology of this interesting dermatological condition has evolved, so too have the criteria and profiling of BSLE. The distinct clinical, histological and immunological features of BSLE represent a unique bullous disease phenotype.