Birgitta Versluijs

145 Outcome predictors in Allo-LS in children Of 21/22 children who are alive without Allo-LS the PFT at last follow up, at a median of 7.0 (range 0.7-9.4) years after Allo-LS, show a median FEV1% of 79 (SD 21.6), median FVC% was 76 (SD 17.9), median FEV1/FVC was 0.94 (SD 0.10), median RV/TLC was 28 (SD 10.6), median TLCO was 80 (SD 15.5) % of normal. All patients were without pul- monary symptoms. In 13 patients, we had serial data on PFT over time, showing a trend to worsening of lung function after 2 years (see Figure S3). Discussion To our knowledge this is one of the very few reports on prognostic factors for outcome in children with Allo-LS. In 53 HCT recipients developing Allo-LS, the largest pediatric cohort published thus far, we studied predictors for response to first line treatment with MP-pulse therapy. 45% of patients were ‘poor responders’ resulting in very poor Allo-LS free survival (4%). Predictors for poor outcome were previous systemic treatment for GvHD and mechanical ventilation at time of diagnosis. PCR positivity for RV at diagnosis of Allo-LS was an indicator for favorable outcome. Failure to therapy was noted already very early after start of therapy. In BOS patients, inadequate increase of FEV1% after the first MP pulse predisposed for poor long term outcome. Responders to therapy were found to have good survival chances with improvement of clinical symptoms and almost normalization of pulmonary function. Limitation of this cohort analyses are the relatively small numbers of patients, and the long study period in which a few important changes in policy regarding pulmonary complications took place (e.g. standard MP-pulse from 2005, screening for pulmonary disease from 2008). Over time we have become more aware of the risks and symptoms of Allo-LS, possibly leading to earlier diagnosis. In those patients who were RV positive prior to HCT, immune suppressive therapy was tapered slower, this may have resulted in the observed decrease in incidence of Allo-LS after 2008 from 25% to 10% (data not shown). However, we found no significant difference in Therapy Response and Allo-LS free survival between patients transplanted before or after 2008. There are very few studies aiming to identify factors correlating with therapy response and mortality prediction in Allo-LS. In one study on IPS, patients requiring less oxygen, and not needing mechanical ventilation had higher response rates, and there was a trend to better response in those starting treatment within 3 days of supplemental oxygen sup- port. 6 This is in line with our results. The importance of FEV1% in the characterization of severity of BOS is long established. Williams et al. showed a relation between FEV1% < 35% and mortality in adults with BOS after HCT. 15 In our cohort we could not confirm a relation between FEV1% at time of Allo-LS and mortality, but we did not have such low FEV1% values at Allo-LS diagnosis. 8