Bone Marrow Histology in Hemophagocytic Lymphohistiocytosis.

CONTEXT.­: Bone marrow (BM) samples are obtained through aspiration and trephine biopsy. Hemophagocytic lymphohistiocytosis (HLH) has been largely studied in BM aspirate smears. OBJECTIVE.­: To investigate the histologic features of HLH in trephine biopsy. DESIGN.­: Patients with hemophagocytosis in BM aspirate smears were assigned to HLH (n = 127) and non-HLH (n = 203) groups. We quantified hematoxylin-eosin and CD68 immunohistochemical staining of their trephine biopsies. RESULTS.­: No significant correlation was noted in the hemophagocytosis count between aspirate smears and trephine biopsies. Compared with the non-HLH group, the HLH group had a higher hemophagocytosis count (13 versus 9 per tissue section, P = .046), lower percentage of the adipocytic area (36.7% versus 50.3%, P < .001), and higher percentage of the foamy area (19.1% versus 14.5%, P < .001). The HLH group had more histiocyte infiltrates (total histiocyte density, 9.2% versus 7.3%; P < .001) and more fat-infiltrating histiocytes (histiocyte density of the fat-associated part [HD-FA], 7.6% versus 6.2%; P < .001). We identified the following poor prognostic factors in the HLH group: age 50 years or older (median overall survival [mOS], 95 versus 499 days; P = .04), Epstein-Barr virus-positive T-cell lymphoproliferative diseases (EBV+TLPDs) (mOS, 51 versus 425 days; P < .001), hemophagocytosis count of 6 or higher per tissue section (mOS, 66 versus 435 days; P = .02), and HD-FA of 9% or greater (mOS, 61 versus 359 days; P = .02). Multivariate analysis revealed that age 50 years or older (hazard ratio [HR], 2.38; P < .001), EBV+TLPDs (HR, 2.07; P < .001), and hemophagocytosis count of 6 or higher per tissue section (HR, 2.07; P = .002) were independent prognostic factors for HLH. CONCLUSIONS.­: The HLH group had higher hemophagocytic activity, higher cellularity, a more foamy appearance, more histiocyte infiltrates, and more fat-infiltrating histiocytes. High hemophagocytic activity and marked histiocyte infiltrates in the BM fat were associated with poorer prognosis.

Urine sediment examination: a comparison of automated urinalysis systems and manual microscopy.

BACKGROUND: Several automated instruments examining urine sediment have been introduced. We compared the performance of Sysmex UF-100 and Iris iQ200 with manual microscopy in urine sediment testing. METHODS: Four hundred and thirty-six urine samples were collected. The urine sediments were examined by manual microscopy and these 2 automated urinalysis systems. RESULTS: The within-run CVs for urine samples ranged from 3.4% to 22.3% for the iQ200, 1.6% to 24.2% for the UF-100 and 12.5% to 43.9% for manual microscopy. Between-run CVs on quality-control samples ranged from 6.1% to 32.4% for the iQ200 and 3.5% to 24.7% for the UF-100. The agreement between methods was good for red blood cells and white blood cells counts based on r values of 0.935 to 0.968. However, for epithelial cells, the values measured by different systems were poorly correlated (r=0.888-0.922). The Bland-Altman plot indicated a trend towards the automated cell count being greater than the manual microscopy as the epithelial cell count increased. Casts were difficulty differentiated by 2 automated systems. CONCLUSIONS: These 2 automated urinalysis systems demonstrated good concordance with each other in urine sediment examination. The automated process could be used as a screening procedure but some manual microscopy was still necessary.