Non-infectious pulmonary syndromes (NIPS) frequently complicate allogeneic stem cell transplantation (alloSCT). The most common and serious is the bronchiolitis obliterans syndrome (BOS), characterized by irreversible fixed airflow obstruction, impaired quality of life and a high mortality. Treatment for established symptomatic disease is relatively ineffective. Although changes in spirometry parameters at day 80 have been shown to correlate with poorer lung function at one year, no screening test has been proven to clearly predict for the development of NIPS. We sought to determine whether early changes in spirometry parameters predict later development of NIPS. Secondary objectives were to determine pre-transplant predictors of NIPS and the impact of NIPS on relapse, treatment-related mortality (TRM) and survival.
Spirometry and DLCO were performed pre-alloSCT, at day 100 (D100) and one year post-alloSCT. We retrospectively analyzed spirometry, CT and bronchoalveolar lavage results in consecutive patients having alloSCT from 2004-2010 to identify cases of NIPS. Cases of BOS and cryptogenic organizing pneumonia (COP) were defined as per published guidelines (NIH consensus guidelines, 2005). Spirometry trends and baseline variables were compared between patients with and without NIPS to identify early predictors and risk factors for NIPS.
Pulmonary function
257 patients underwent allo-SCT during the study period. 235 survived to D100 and were thus assessable for chronic GVHD (cGVHD) and NIPS. Of these, 23 (9.8%) developed NIPS, 18 with BOS and 5 with COP. Median time of onset was day 367 (IQR 144-544 days). All but one case developed prior to two years post-alloSCT. Change in FEV1.0 (ΔFEV1.0), was the best predictor of later NIPS development, while change in pulmonary diffusion capacity for carbon monoxide was not predictive. Median ΔFEV1.0 from pre-transplant to D100 in patients later developing NIPS was -12% (IQR -25% to -1%) vs -1% (IQR -7% to +6%) in those who did not, p=0.002. From pre-transplant to 1 year, it was -19% (IQR -37% to -6%) vs -3% (-10% to +4%), respectively, p<0.001. Median FEV1.0 at diagnosis was 58% predicted (IQR 46-71%). Once fixed airflow obstruction was established, progressive deterioration in lung function with time was the rule; median change in FEV1.0 during follow-up, post-diagnosis of NIPS, was -2% per annum (IQR -7 to 0%). Median follow-up duration post-diagnosis of NIPS was 1 year, range 0-6 years. For those with BOS and >6 months of follow-up, median change in FEV1.0 per year during follow-up was -8% (IQR -14% to -2%). In contrast, 4 of the 5 patients with COP had substantial improvement in FEV1.0 (11-36%) with normalization of FEV1.0 in 3 of these 4.
Pre-transplant risk factors
Busulfan-based, but not total body irradiation (TBI)-based conditioning increased the risk of NIPS [OR=8.87, 95%CI 3.33-23.63, p<0.001]. No cases of NIPS were seen in 53 patients who received in vivo T-cell depletion with Thymoglobulin as part of their conditioning (p<0.0001).
Survival
At a median follow-up of 44 months, overall survival was 59.1%. There was a trend towards increased all-cause mortality in those patients developing NIPS [HR = 2.02, 95%CI 0.92-4.44, p=0.08] and an increase in TRM [HR = 5.96, 95%CI 2.14-16.62, p=0.001], with eight patients (35%) with NIPS experiencing TRM. No statistically significant difference in disease relapse was seen between those with and without NIPS [HR = 0.24, 95%CI 0.03–1.80], although only one patient with NIPS experienced disease relapse. This may reflect the small numbers of patients with NIPS.
Spirometry is a potentially useful screening test for identification of pre-symptomatic NIPS. We recommend 3-monthly spirometry surveillance for up to two years post-transplant, given the moderately severe obstruction at diagnosis (median FEV1.0 58% predicted) and the incidence of 50% of cases between one and two years post-allo-SCT. Our findings require prospective validation to identify patients in whom earlier intervention may potentially modify the natural history of this disease.
No relevant conflicts of interest to declare.
