Birgitta Versluijs

165 Discussion This conclusion seems to contradict our idea of the role of RV in IPS. However, as we are not informed about RV status pre-HCT, and the analysis on outcome was not done for RV and other pathogens separately, we are not convinced that these patients were actually having an infectious pneumonia. It could still mean that persisting RV after HCT, as a risk factor for IPS, triggered immune-mediated lung disease, and that steroids are bene- ficial in the treatment of these patient at the moment of clinical deterioration. Which diagnostic tools contribute to prediction, diagnosis and follow-up of Allo-LS? In the different phases of HC, various diagnostic tools for lung disease are used. We tho- roughly studied the yield of our pre-HCT pulmonary screening program, as described in Chapter 5 . There is a high prevalence of RV in pediatric HCT recipients. In the pre-HCT samples we found RV PCR-positivity in 30-60 % of patients, as described in Chapter 3, 5 and 7. As RV from BAL is such an important predictor for Allo-LS, we implemented pre-HCT BAL screening in all HCT patients, to identify patients at high risk for Allo-LS. We stopped doing parallel routine NPA for RV, as we have shown that PCR-positivity in NPA-only did not increase the risk for Allo-LS ( Chapter 7 ). Only in children who do not need an anesthetic procedure, and therefore cannot have a BAL done, NPA for RV is still performed. In case of BAL RV PCR-positivity, immunosuppression is now given for a longer period after HCT, to prevent Allo-LS. In our center, only Influenza Virus positive patients receive antiviral treatment, all other RV, including RSV, are left untreated. We do not postpone HCT in the presence of RV, as we have seen that RV persists for months in our mostly immunocompromised patient population. So we have not been able to study the impact of recently cleared RV infection (spontaneous or by treatment) on the risk of Allo-LS. The other components of our pre-HCT pulmonary screening program, including BAL for non-viral pathogens, chest HRCT and PFT are not primarily done in the context of Allo-LS. Clinical significant HRCT abnormalities were associated with Allo-LS ( Chapter 5 ), but we have not adjusted our HCT procedures on that finding, so far. It would be inte- resting to study the impact of chest-HRCT again, now we have a larger cohort of patients who have been screened in the same way, and we can apply the HRCT score, as descri- bed in Chapter 4 . The other pre-HCT findings were not related to the development of Allo-LS. These tests are very important though, as they provide information about direct infectious threats (requiring antimicrobial treatment) and baseline pulmonary function, which in our center is an exclusion criterion for the HCT procedure if it is below 50% of normal. Overall, screening outcome had clinical implications for 33% of patients, and each screening modality contributes separately. 9