Are routine iron stains on bone marrow trephine biopsy specimens necessary?

AIMS To determine the role of Perls' staining in bone marrow trephine biopsy sections. METHODS The haemosiderin content of 155 Perls' stained, formic acid decalcified trephine biopsy sections was assessed and compared with Perls' stained aspirate samples in 105 cases and haematoxylin and eosin (H&E) stained biopsy sections in all cases. RESULTS An evaluable aspirate film with positive iron or at least seven negative particles was available for 105 biopsies. Only 71 of 95 cases with detectable aspirate iron had haemosiderin detectable on a Perls' stained section. None of 10 samples with a negative aspirate had a positive trephine biopsy. Haemosiderin was positive in 101 of the 155 Perls' stained sections, and was detectable on the H&E stained section in 71 of these cases. In five of 54 cases with negative Perls' staining, a small amount of haemosiderin was thought to be present on H&E staining. CONCLUSIONS Aspirate smears reflect bone marrow iron stores more reliably than formic acid decalcified trephine biopsy sections. The presence of iron in Perls' stained aspirates in 44% of cases with negative Perls' stained sections indicates that iron is often lost from sections during decalcification. However, 61% of cases with unassessable aspirate samples had a positive trephine biopsy Perls' stain, contributing useful clinical information about iron status. Preparation of Perls' stained sections only in cases in which aspirate samples are inadequate for iron assessment and no obvious haemosiderin is present in an H&E stained section could produce savings in staff time and reagent costs.