The effects of Eculizumab on the pathology of malignant atrophic papulosis

Abstract

Background: Degos disease is a frequently fatal and incurable occlusive vasculopathy most commonly affecting the skin, gastrointestinal tract and brain. Vascular C5b-9 deposition and a type I interferon (IFN) rich microenvironment are held to be pathogenetically important in the evolution of the vascular changes. We recently discovered the use of eculizumab as a salvage drug in the treatment of near fatal Malignant atrophic papulosis (MAP). The effects of eculizumab on the pathology of MAP are explored.

Methods: Archival skin and gastrointestinal biopsy material was procured over a 2.5-year period before and after eculizumab therapy in our index case. Routine light microscopy and immunohistochemical assessment for C3d, C4d, C5b-9, MxA and caspase 3 were examined. Direct immunofluorescent studies were also conducted on select biopsy material.

Results: The patient had received eculizumab as a emergent life saving measure and following rapid improvement he continued with biweekly infusions for 4 years. Although improved he continues to have signs and symptoms of persistent abdominal disease. Pre-Eculizumab biopsies showed an active thrombotic microangiopathy associated with a high type I interferon signature and extensive vascular deposits of C5b-9 in skin and gastrointestinal biopsies. Endothelial cell apoptosis as revealed by Caspase 3 expression was noted. Inflammation comprising lymphocytes and macrophages along with mesenchymal mucin was observed as well. Post-eculizumab biopsies did not show active luminal thrombosis but only chronic sequelae of prior episodes of vascular injury. There was no discernible caspase 3 expression. After 12 months of therapy, C5b-9 was no longer detectable in tissue. The high type I IFN signature and inflammation along with mucin deposition was not altered by the drug. In addition, there was little effect of the drug on the occlusive fibrointimal arteriopathy which appears to be one characterized by extensive myofibroblastic expansion of the intima potentially as revealed by staining for smooth muscle actin without immunoreactivity for desmin and myogenin.

Conclusions: Complement activation and enhanced endothelial cell apoptosis play an important role in the thrombotic complications of MAP. However, the larger vessel proliferative intimal changes appear to be independent of complement activation and may be on the basis of other upstream mechanisms. Monitoring C5b-9 deposition in tissue is likely not of great value in assessing treatment response to eculizumab given the persistence of C5b-9 in tissue for several months despite clinically effective C5 blocking therapy. A more integrated approach addressing upstream and downstream pathways in addition to those attributable to complement activation are critical for the successful treatment of MAP. Eculizumab may be used as salvage therapy in critically ill patients with thrombotic microangiopathy.

Figure 1

A, B, C, and D: Skin histology pre-Eculizumab microvascular changes along with the pattern and extent of C5b-9 deposition and MXA expression in tissue. An erythematous lesion biopsied in July 2009 shows a thrombotic microangiopathy along with dermal chronic microvascular changes characterized by thickened basement membrane zones along with superficial vascular ectasia and foci of vascular drop out. The chronic microvascular changes are present in both the A (200x) and B (1000x) biopsies while intraluminal fibrin deposition is noted in B. (hematoxylin and eosin) In Figure 1A, there is a sparse perivascular lymphocytic infiltrate. The vessels appear hyalinized attributable to vascular basement membrane zone reduplication. In Figure 1B, the vessel is largely devoid of endothelium and contains a loosely adherent thrombus. The vessel is also noticeably thickened. Figure 1C exhibits the pattern of C5b-9 deposition pre-eculizumab. There is prominent deposition of C5b-9 by immunofluroescence in the vessels in the pre-eculizumab biopsy (400x). Figure 1D demonstrates MxA in skin biopsy pre-Eculizumab. There is extensive staining for MxA in the epidermis, endothelium and inflammatory cells in the pre-eculizumab biopsy indicative of a type I IFN rich microenvironment (diaminobenzidene, 1000x).

Figure 2

Skin microvascular changes and immunohistochemical assessment for C5b-9 and MXA staining post-eculizumab. A. Skin histology post-eculizumab. Subsequent to receiving Eculizumab, the patient underwent 6 sequential biopsies 2 weeks to 25 months after commencing the drug. In none of the post-eculizumab biopsies is there evidence of active endothelial cell injury and vascular thrombosis. A common finding in all of the biopsies is one reflective of antecedent episodes of microvascular injury characterized by subepidemal fibrosis with vascular drop out, vascular basement membrane zone duplication and vascular ectasia (diaminobenzidene, 200x). B. Inflammatory cell infiltrate in skin biopsy post-eculizumab. A variable inflammatory cell infiltrate comprising lymphocytes and histiocytes primarily arranged around blood vessels is present in all of the biopsies. The biopsy procured in March 2011, 16 months after the commencement of eculizumab, shows a striking lymphocytic and histiocytic infiltrate with abundant mucin deposition. Despite the exuberant inflammation, active microvascular injury is not seen. In particular discernible vascular thrombosis is not observed (hematoxylin and eosin, 400x). C. MxA in skin biopsy post-eculizumab. In all post-Eculizumab biopsies, there continues to be prominent expression of MxA in the epidermis, inflammatory cells and endothelium (diaminobenzidene, 1000x). D. C5b-9 in skin biopsy post-eculizumab. Nine months after the commencement of the drug, C5b-9 was no longer apparent (diaminobenzidene, 400x).

Figure 3

A and B: Submucosal vessels of intestinal resection specimens pre- and post-eculizumab, respectively. There were a total of 3 intestinal resection specimens available for assessment. One was obtained prior to the commencement of the drug (July 2009) and two were obtained 4 months (February 2010) and 21 months after starting the drug (July 2011). There are aspects of the pathology that was remarkably similar between the pre-eculizumab specimen and the one obtained 2 years after starting therapy. A striking fibro-obliterative arteriopathy affects small and medium-sized vessels of the submucosa as well as mesentery in both the pre- (A) and post-treatment (B) resection specimens (3A, hematoxylin and eosin, 200x, 3B hematoxylin and eosin, 200x). In B, the intimal proliferative changes are very striking comprising hypercellularity of the intima along with an increase in intimal matrix represented by collagen and hyaluronic acid.

Figure 4

A and B: C5b-9 in submucosal vessels of intestinal resection specimens pre- and post-eculizumab, respectively. Immunohistochemical assessment reveals marked deposition of C5b-9 in vessels of the submucosa in the pre-eculizumab specimen (A, diaminobenzidene, 400x) while the intestinal resection specimen procured 21 months after commencing therapy (B, diaminobenzidene 400x) is devoid of vascular staining apart from nonspecific staining within the elastic tissue of blood vessels.

Figure 5

The CD14 stain shows a striking influx of CD14+ monocytes throughout the gastrointestinal tract with margination of monocytes to lie in apposition to the adventitia. Note the infiltration through the medial wall and extension into the intima (diaminobenzidene, 200x).

Figure 6

While the desmin extensively highlighted the medial wall smooth muscle cells, the hyperplastic intima is essentially negative indicative that the intimal cells are not of true medial wall smooth muscle origin (diaminobenzidene, 200x).

Figure 7

In this section of artery, there is very extensive staining for smooth muscle actin, an immunophenotypic profile that is common in myofibroblastic cells of stem cell hematopoietic and monocyte origin (diaminobenzidene, 200x).