Birgitta Versluijs

167 Discussion outcome of children and adults could be explained by factors like pulmonary situation pre-HCT (COPD, smoking), co-morbidities, different risk factors for developing Allo-LS influencing response to therapy (RV, less GVHD) etcetera. There are few publications on treatment and outcome of IPS and BOS in children after HCT. In general the methodology of these studies is insufficient. Most studies are single center, retrospective analyses with small numbers of patients. Response is not always defined uniformly, and intervention is not blinded. With these remarks in mind, Table 2 gives an overview of the literature and can lead to some conclusions. IPS therapy is always corticosteroid based. In the various reports either an increased dose of steroids, or the addition of TNF-alpha blockade with etanercept or infliximab are evaluated. In most IPS studies Initial Response (defined as recovery of respiratory symptoms and/or cessation of supplemental oxygen requirement at day 28), Overall Sur- vival and Non Relapsed Mortality (here defined as death, not due to relapsed disease, excluding progressive lung disease) are stated. With regard to initial response our cohort (as described in Chapter 8 ) performed worse, with only 53% patients responding, versus 71-81% reported by others. This is partly due to the definition of response. We considered the need of salvage therapy as a sign of initial therapy failure, without restrictions to the day of this event. In most of the other papers response was defined as recovery at day 28 after first or second line therapy. When applying these criteria in our cohort of IPS patients Initial Response rate would be 60%, still slightly lower than what was seen in the other groups. Responders to therapy tend to do so within 8-17 days. 7,8,37 Overall survival rates varied between 21% and 50%. Cause of death is not always men- tioned, but IPS contributed to death in 18-50%, and non-relapse mortality (NRM) other than lung disease was high (13-46%), mostly due to infection or GvHD in another organ. Here our results are in line with others. Treatment of BOS in children after HCT is also based on corticosteroids. Different doses, given systemically and by inhalation, in combination with other immune suppressive agents, sometimes with the addition of immune modulating agents like azithromycin or pravastatin are studied. 13,37-40 In contrast with IPS studies, overall survival in BOS is better (57%-78%), with less children dying from progressive lung disease (11%-30%) or NRM (0-11%). The BOS patients in our cohort did relatively poor with respect to overall survival, but remarkably there was normalization of lung function in the majority of survivors. This is in contrast with most studies, especially in adults. This could be due to early diagnosis and the fact that respiratory virus played a role in the development of BOS in a large proportion of patients, influencing reversibility of the immunological phenomena in these cases. Or it might be the anti-fibrotic effect of imatinib, which we gave to all our BOS patients. 9