Application Of BOS 0-Potential Criteria As a Predictor For Bronchiolitis Obliterans Syndrome In Allogeneic Stem Cell Transplant Patients

Bronchiolitis Obliterans Syndrome (BOS) is a recognized manifestation of chronic GVHD (cGVHD) of the lungs after allogeneic HSCT. Survival in patients with BOS is poor, and this has been attributed to delayed recognition of BOS, at a point when patients have developed severe and irreversible changes to their lungs. BOS is also a significant problem within the lung transplant population, and pulmonologists have adopted a criterion called BOS 0-potential (BOS 0p) based early spirometric changes, that has identified patients at risk for BOS as early as 335 days prior to development of BOS. This is the first study that evaluates the potential of BOS 0p to predict BOS after allogeneic HSCT at an earlier time point. In addition, we determined whether other transplant-related covariates could improve our capability to predict both BOS 0p and BOS. Risk stratification could allow for earlier intervention and better outcomes in an otherwise fatal disease.

A retrospective chart review was performed on consecutive patients 18 and older who received an allogeneic stem cell transplant at the University of Michigan BMT Program between 1/1/2009 and 7/1/2011. A total of 426 patients were identified, and BOS was determined using clinical criteria established by the NIH cGVHD working group. BOS 0p was determined based on criteria used for lung transplantation patients, including: 1. Decline in FEV1>10% of predicted from pre-transplant baseline or 2. Decline in FEF25-75>25% of predicted from pre-transplant baseline. Decline in lung function needed to persist on at least 2 consecutive PFTs, and PFT values could not be consistent with criteria for BOS. The date of BOS 0p was the date of the first PFT demonstrating this decline. Statistical analysis was performed using STATA 12.1. Univariate analysis was performed using Fisher’s Exact test or Wilcoxon Rank Sum test. Risk factors for BOS 0p and BOS were assessed using a competing-risk regression model, where death and relapse represented competing risks.

Among 426 patients, 23 (5.4%) patients met criteria for BOS at a median of 437 days after transplant (95% CI 274,821d). BOS 0p criteria were met in 65 patients (15.2%). BOS 0p occurred at a median of 267 days after transplant, (95% CI 110,475d), and 206 days prior to the development of BOS (95% CI 13,521 days). BOS 0p predicted BOS with a sensitivity of 87%, and a specificity of 89%. The resulting PPV was 31%, and NPV was 99%. On multivariate analysis, the most significant risk factor for development of BOS0p was a history of pulmonary disease pre-transplant, i.e. COPD, asthma (HR 3.5, p<0.001). On univariate analysis of all patients, risk factors for development of BOS included development of BOS0p (p<0.001), history of pulmonary disease pre-transplant (p=0.001), a more severe onset NIH cGVHD score (p=0.01) and an increased eosinophil count on onset of cGVHD (p=0.04). On multivariate analysis, BOS 0p was the strongest predictor of BOS (p<0.001), followed by severity of cGVHD on onset (p=0.01), and history of pulmonary disease pre-transplant (p=0.03). Multivariate analysis of the subgroup of patients with BOS0p revealed that the main risk factor associated with development of BOS was a history of pulmonary disease pre-transplant (p=0.05).

BOS 0p, is an effective spirometric method to identify patients at risk for the development of BOS. BOS 0p can be detected at an earlier time point, providing an opportunity to use earlier, and potentially more effective treatment strategies. Furthermore, patients with a history of pulmonary disease prior to transplant and BOS0p could be the target of prophylactic or close surveillance approaches that could prevent the devastating and often irreversible consequences of BOS.

No relevant conflicts of interest to declare.