Background/Purpose:   The interpretation of the literature on diffuse alveolar hemorrhage (DAH) secondary to ANCA-associated vasculitis (AAV) has been complicated by the small size of reported cohorts, variable and unclear definitions of DAH used, and insufficient description of respiratory failure and its predictors. In addition, patients with severe DAH requiring mechanical ventilation were excluded from the only randomized controlled trial that comprised a sizable number of patients with DAH.  In a well-defined cohort of patients with DAH secondary to AAV, we sought to identify predictors of respiratory failure and to describe the therapeutic efficacy of plasma exchange (PLEX) and rituximab versus cyclophosphamide in patients with or without respiratory failure.

Methods:   Single center historical cohort study of all consecutive patients presenting with DAH secondary to AAV evaluated over a 15 year period. Respiratory failure was defined as the need for mechanical respiratory support. The main outcome was complete remission at 6 months (BVAS/WG=0 and off corticosteroids).  Logistic regression models were developed to examine the predictive role of the baseline clinical characteristics for the development of respiratory failure, and also for the role of PLEX and induction of remission therapy on the main outcome.  To account for potential unequal distribution of important covariates between groups resulting from the lack of random assignment in this observational study (sicker patients received PLEX), a propensity score analysis was performed to assess the probability of receiving PLEX.

Results:  Seventy-three patients with DAH were identified of whom 34 patients developed respiratory failure. Degree of hypoxemia upon initial presentation, higher neutrophil percentage in the bronchoalveolar lavage fluid cell count, and higher C-reactive protein levels were independently associated with the development of respiratory failure. Complete remission at 6 months was achieved in 23 of 32 (72%) patients treated with PLEX versus 32 of 41 (78%) patients treated with no PLEX, p=0.54. After adjustment for important covariates, PLEX therapy was not found to be associated with achieving complete remission at 6 months (OR 0.49, 95% CI 0.12-1.95, p=0.32). Complete remission at 6 months was achieved in 33 of 37 (89%) patients treated with rituximab versus 21 of 31 (68%) patients treated with cyclophosphamide, p=0.02. After adjustment for important covariates, rituximab therapy was independently associated with increased odds of achieving complete remission at 6 months (OR 6.45, 95% CI 1.78-29, p=0.003). There was no difference in long-term survival between patients treated with rituximab compared to those treated with cyclophosphamide.

Conclusion:   The most important predictor of respiratory failure in these patients seems to be the degree of hypoxemia upon presentation. A clear benefit of PLEX added to standard remission induction therapy could not be demonstrated.  Patients with DAH secondary to AAV, including those requiring mechanical ventilation, achieved complete remission by 6 months at a higher rate with rituximab than with cyclophosphamide.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/diffuse-alveolar-hemorrhage-secondary-to-anca-associated-vasculitis-predictors-of-respiratory-failure-and-clinical-outcomes/