Previous treatment predicts the efficiency of blood progenitor cell mobilisation: validation of a chemotherapy scoring system

Following mobilising chemotherapy and myeloid growth factors, some heavily pretreated patients do not mobilise adequate numbers of peripheral blood progenitor cells (PBPC). It would be clinically useful to identify such patients in advance. A recent scoring system based on previous therapy may be useful in predicting CD34-positive cell yield. In this study we validated this scoring system on an independent group of 99 patients undergoing 103 harvesting episodes. In 61 patients mobilised with cyclophosphamide 1.5 g/m(2) and G-CSF, those with treatment scores less than 21 yielded significantly more CD34-positive cells than patients with scores greater than 63 (P = 0.0012). Previous treatment with melphalan or carmustine was associated with a significantly lower yield of CD34-positive cells (P = 0.0001). No relationship was seen between the time from previous chemoradiotherapy and harvest outcome. Patients with treatment scores less than 21 required a shorter duration of G-CSF therapy (P = 0.05). Similar findings were seen in 42 further mobilisation cycles undertaken with alternative mobilisation schedules. The present data suggest that a score summarising previous treatment can be used to predict CD34 yields, and could be of clinical use to identify poor PBPC mobilisers in advance.