[Demonstration of trivalent iron in decalcified bone marrow specimens. Comparative study of standard Perls stain on sections and the Perls pre-reaction on fragments].

UNLABELLED: In the literature the usefulness of the Perls reaction for Iron (Fe+++) stores estimation on bone marrow is largely accepted. However, due to decalcification, there is a considerable disagreement about the accuracy of this staining on osteomedullary tissue. In fact loss of Fe+++ has been claimed by some Authors, and denied by others. In order to check this controversial point in this paper Perls reaction, as a methodological variant, was directly anticipated on fragments (pre-Perls). In this way numerous osteomedullary sternal specimens from autopsies (non hematological cases) were observed in comparison with normal Perls reaction on sections (standard-Perls). In standard-Perls series (A) decalcification was achieved by HCl (alone or in combination with EDTA) or by formic acid. On pre-Perls treated fragments decalcification was an intrinsic effect of the hydrochloride component during Prussian Blue formulation. Nine different procedures of fixation were tested respectively for standard-Perls and pre-Perls. Pre-Perls was performed on pre-fixed (B series) as well as post-fixed material (C series). For the microscopical evaluation was applied the classical grading system proposed by Lundin for histology. Maximum grade was only observed in B and C, was the same in both these series and varied from 2 to 4 depending on autopsic case. Results strongly indicate the superiority of pre-Perls versus standard-Perls, with a differential grading ranging from 2 to 4 grades, in seven procedures of B (two procedures were equivalent in A and B) and in all the procedures of C. The comparison between B and C series showed inferiority of B in four procedures. For control (D series) standard-Perls was systematically performed on sections from pre-Perls: no differences between pre-Perls and pre-Perls + standard-Perls were observed indicating a complete permeation of fragments by Perls reagents. In general terms morphology in pre-Perls B was better than in pre-Perls C and substantially identical to standard procedures of A. Other data are analytically reported.

CONCLUSIONS: Pre-Perls staining (with special reference to five procedures of B) seems to be recommendable for decalcified bone marrow biopsies, in conjunction with standard processing, when an high sensitivity in detection of Fe+++ deposits is required or in absence of aspirate. Pathologists could subsequently decide if this methodological variant is suitable or not for an effective alternative choice in their routine according to morphological and immunohistochemical adequacy.