IgA vasculitis preceded by pulmonary hemorrhage in a patient with ulcerative colitis during treatment with vedolizumab

Ulcerative colitis (UC) is an inflammatory bowel disease (IBD) characterized by relapsing and remitting inflammation affecting colon and rectum. Although erythema nodosum and pyoderma gangrenosum are well-known cutaneous manifestation of UC, vasculitis is a rare symptom related to this disease1. Vedolizumab is an antibody against gut-selective α4β7 integrin and approved for UC and Crohn’s disease. Vedolizumab is rarely associated with cutaneous side effects including vasculitis. We herein report a case of IgAV that developed during vedolizumab therapy for UC.A 71-year-old Japanese woman visited our hospital complaining of purpura on her limbs occurring a few days earlier. She suffered from UC refractory to corticosteroid and mesalazine for more than 11 years, but it was well-controlled by vedolizumab treatment for the last 28 months. She also had recurrent episodes of hemoptysis two months before the appearance of the palpable purpura, which was one month after 16th administration of vedolizumab. She had no history of smoking or respiratory disease. However, she had been coughing with small amount of yellow phlegm for a few months before the onset of hemoptysis. Chest computed tomography (CT) showed occurrence of ground glass opacity and consolidation with thickening of the tracheal walls in the lower left lobe compared to the image before starting vedolizumab treatment (Fig. 1a,b). The sputum cytopathology was class 2, and bacterial culture detected Haemophilus parainfluenzae and Pseudomonas aeruginosa. One-month of oral erythrosine therapy induced only slight improvement of the lung consolidation (Fig. 1c). One month after 17th administration of vedolizumab, palpable purpura of 1 to 3 cm in size appeared on the lower limbs (Fig1. d), followed by hemoptysis a few days later. She did not complain of joint pain, numbness, or dyspnea. Skin biopsy specimen showed leukocytoclastic vasculitis (Fig. 1e,f). Immunohistochemistry of immunoglobulin A (IgA) revealed IgA deposition on the vascular walls of the superficial dermis (Fig. 1g). Laboratory findings were unremarkable, except elevated serum IgA (533 mg/dl). Serum anti-neutrophil cytoplasmic antibodies were negative. A diagnosis of IgA vasculitis (IgAV) was made. The purpura disappeared promptly by 20 mg/day of oral prednisolone. Chest CT scan at 2 weeks after the initiation of prednisolone treatment showed substantial improvement of the lung consolidation (Fig. 1h). Although our patient was suspected to have lung infection causing pulmonary hemorrhage, improvement of consolidation in the lung field was limited even after antibiotic therapy. However, it was completely improved with prednisolone treatment. Although bronchoscopy was not performed in this case, we considered it as diffuse alveolar hemorrhage due to IgAV in consideration of multifocal abrasive shadows and combined shadows of air-space consolidation with increased lung field density on CT scan in addition to the exacerbation of the hemoptysis after the appearance of purpura of IgAV. We thus thought that pulmonary hemorrhage was the initial symptom of IgAV.IgAV can be triggered by infection, medications, and malignancy2. IgAV rarely cause pulmonary hemorrhage3. In our case, pulmonary hemorrhage without renal, gastrointestinal or joint symptoms preceded cutaneous purpura by 2 months and relapsed intermittently after the appearance of the purpura. Pulmonary hemorrhage can be triggered by infection4. Recently, a relationship between IBD, particularly UC, and airway lesions such as bronchiectasis has gained attention5. Because vedolizumab inhibits gut-selective lymphocyte migration, it is unlikely that it directly affected immune system of the skin and lungs. However, it was possible that vedolizumab treatment altered the intestinal microbiota and affected the immune environment in the lungs as a consequence6. Therefore, preceding unrecognized subclinical bronchiectasis combined with lower respiratory tract infection by Haemophilus parainfluenzae and Pseudomonas aeruginosa may have caused abnormal immune responses leading to the development of IgAV.While a recent study focusing on IgAV in patients with IBD indicates that tumor necrosis factor-α blockers may promote the onset of IgAV7, vedolizumab is rarely associated with IgAV. Although Gold et al. reported a case of IgAV as an infusion reaction of vedolizumab in a patient with Crohn’s disease8, our case cannot be explained by the same mechanism. Vedolizumab was not discontinued in our patient on the ground of the following points. Firstly, UC was well-controlled by vedolizumab, and there was no exacerbation of UC even after the onset of IgAV. Secondly, the mechanism of action of vedolizumab is specific to T cells homing to the intestinal tract. Thirdly, compared to TNF inhibitors, vedolizumab has much fewer reports of side effects of vasculitis. Indeed, Desai et al reported a case of IgAV in a patient with UC on vedolizumab treatment that was successfully treated by systemic steroid without discontinuing vedolizumab9. Despite vedolizumab treatment was continued, our patient did not report relapse of purpura and hemoptysis even after the discontinuation of prednisolone. There was no exacerbation of UC in the meanwhile. However, since vedolizumab is a relatively new drug, it is necessary to investigate whether vedolizumab increases the risk of IgAV in patients with IBD.