Properly stained slides are usually pink to the naked eye. Bluish discoloration can arise from too thick a smear, prolonged staining time, inadequate washing, or excessively alkaline buffer in the dyes. A slide also may be blue because the blood contains abnormally high amounts of plasma proteins in such diseases as multiple myeloma. Excessively pink appearance can result from opposite problems: short staining times, prolonged washing, dye that is too acidic, or mounting the slides with coverslips before they are adequately dried. Use of unclean slides, inadequately filtered stain, and dust settling on the smear will lead to the appearance of precipitates.

Microscopic examination should begin by scanning the entire slide at low power (×10 or ×20 objective) to determine
the adequacy of staining looking for uniformity of color: nuclei of leukocytes should appear purple and red cells pink, rather than brick-red or yellow. Review at this low magnification also allows detection of overt abnormalities in cell number, type, and aggregation, and permits finding the optimal areas to evaluate all the blood components. A systematic approach should include evaluation of the lateral and feather edges of blood smears prepared by the wedge method, where a disproportionate distribution in these areas of any large abnormal cells, platelet clumps, and microfilarial parasites, if present, may occur. Red cells are best evaluated in the thin areas of the smear, where they are present in a single layer, closely apposed but not overlapping, and exhibiting normal central pallor. Leukocytes may be best examined in the thick areas of the smear, where many cells are present in a single field of view, but abnormal lymphocyte morphology is best appreciated in the thin areas. Review at higher power (×40 to ×50 objective) allows for an assessment of cell size and a closer examination of the nuclear and cytoplasmic features of individual cells. Oil immersion lenses (×50 and ×100) allow for a detailed evaluation of the quality of the nuclear chromatin and presence of nucleoli, and of cytoplasmic components such as granules, vacuoles, and inclusions. This power provides confirmation of any suspicious cells or organisms seen at lower power. Depending on the procedures defined in each laboratory, the manual differential count consists of identifying and classifying 100, 200, 500, or 1,000 white blood cells and reporting each type as a percentage. Note should also be made of any atypical findings in the leukocytes including abnormal nuclear segmentation and granulation of the neutrophils, or the presence of atypical lymphocytes. The presence in a blood smear of nucleated red cells, megakaryocytes, macrophages, and immature leukocytes, which are normally not found in circulating blood, deserves special note and reporting.

The final reporting also includes an assessment of red cell morphology and an estimate of the platelet count. Evaluation of red cell morphology includes reporting of cell size, estimation of the hemoglobin content, shape, inclusions, or structural abnormalities. In the blood of a healthy individual, the red cells appear as uniform discs that range 6 to 8 µm in diameter with a slightly pale central area. Red blood cells (RBCs) smaller than 6 µm in diameter are microcytic, and those that are larger than 9 µm are macrocytic. Abnormal variation in cell size, or anisocytosis, is commonly seen in anemias and is reflected in a wider RDW (RBC distribution width) or coefficient of variation of RBC volume in automated blood counts. The color of the red cells reflects the hemoglobin content, and increased area of central pallor is seen in hypochromic cells, whereas a denser central coloration is observed in hyperchromic cells or in spherocytes. Abnormal (increased) variation in RBC shape, or poikilocytosis, is reported by identifying the presence of specific atypical forms including elliptocytes, spherocytes, target cells, sickle cells, tear drop cells (dacrocytes), schistocytes, acanthocytes, and echinocytes, among others. Reporting of inclusions or granules in the red cells (such as Howell-Jolly bodies, basophilic stippling, or Pappenheimer bodies) or intracellular parasites (e.g., Plasmodium sp. or Babesia sp.) requires evaluation at higher magnification (×40 to ×100).

Areas evaluated for red cell morphology are also used to roughly estimate platelet counts on peripheral blood smears; however, automated counting is far more accurate and reliable. A normal platelet count corresponds to one platelet per every 10 to 30 RBCs, which is approximately 7 to 20 platelets per oil immersion field ×100. A small fraction of platelets may show giant morphology in a normal individual depending on how soon the blood smears were prepared after blood is drawn. However, an increase in the percent of giant platelets that exceeds the normal size range (2-4 µm in diameter) should be noted as it can be associated with neoplastic conditions, immune thrombocytopenia, macrocytic anemias, or Bernard-Soulier syndrome.

The aspirate smear is indispensable for evaluation of the cytologic details of the marrow cells. Specifically, the aspirate smear is necessary to evaluate the presence of dysplasia in the myeloid and erythroid lineages, cell inclusions, and parasites, which cannot be well appreciated on other marrow preparations. A bone marrow nucleated differential count allows assessment of the proportions of different cell lineages, and comparison to expected normal ranges. Furthermore, the manual differential count performed on an adequate aspirate smear is the gold standard for detecting, identifying, and quantifying any abnormal cells (i.e., dysmorphic cells or blasts), and establishing information needed for the classification of hematologic malignancies.

Examination of the aspirate smear begins at low power (×4 or ×10 objective) scanning for particles and for any clusters of abnormal cells such as those of extrinsic metastatic neoplasms, which tend to be more cohesive than hematopoietic cells. At this magnification, it is also possible to estimate marrow cellularity, although that is best judged on the biopsy section. To do so, the ratio of fat to cellular elements in the particles is assessed, together with the cell density around the particles. Presence of lymphoid aggregates and evaluation of megakaryocytes is also assessed at low power. A crush preparation is useful for the assessment of marrow fibrosis, focal disease (i.e., plasma cell myeloma, lymphoma, granulomas, and metastatic carcinoma), the determination of cellularity, and megakaryocyte numbers.

In a normal bone marrow, at medium to high power (×20 to ×40), the predominant marrow component should be segmented granulocytes. An evaluation at this power can give a preliminary assessment of the maturation of the myeloid and erythroid cells and can identify the best areas in which to perform a manual differential count. Bone marrow cells should be counted in cellular areas adjacent to or in the trail of the particles, where the cells are well dispersed, show good cytologic detail, and where lysed cells do not predominate. Areas with excessive air-drying artifact should be avoided, as cells in these areas show suboptimal morphology. Only intact cells should be counted, as naked
nuclei lacking cytoplasmic features cannot be confidently subclassified. The differential count is performed at high (×40) or oil (×50, ×100) magnification to allow optimal identification of cell types. The count includes nucleated myeloid and erythroid cells in various stages of maturation, promonocytes, monocytes, mast cells, lymphocytes, and plasma cells. Cells that should not be included in the count include macrophages, megakaryocytes, osteoblasts, osteoclasts, stromal, or extrinsic cells. The myeloid to erythroid ratio is calculated by expressing a ratio of all the granulocytes and monocytes and their precursors to the erythroblasts in various maturation stages. Touch preparations should be examined with the same parameters as used for aspirate smears; however, these specimens tend to be characterized by suboptimal morphology and are less reliable.

Histologic sections represent thin slices through the specimen, which preserves the overall architecture of the marrow and connective tissue elements. Compared to aspirate smears, histologic sections allow better evaluation of the marrow cellularity and number of megakaryocytes. In the clot section, however, the cellular marrow is somewhat contracted and lacks the full architecture, thus the trephine biopsy section is ideal for assessment of cellularity. Similar to the estimate performed on bone marrow aspirate smears, cellularity is evaluated by comparing the volume of the hematopoietic cells to the adipose cells and stromal elements that make up the total marrow space. Cellularity is assessed with reference to the patient’s age, because bone marrow cellularity varies with age. The subcortical intertrabecular spaces are frequently hypocellular and therefore these areas should be excluded from the cellularity assessment.

Several histologic serial sections are examined to increase the likelihood of identifying focal diseases. A systematic approach helps to ensure that maximal information is obtained from the morphologic review. Evaluation should begin at low power (×4 to ×10) to assess the length, adequacy, and
cellularity of the bone marrow specimen. This power provides an opportunity to examine the general pattern, any presence of focal lesions, abnormal cell clusters, and quality of bone structure. Megakaryocyte numbers can be appreciated at low and medium magnification, which also allow for assessment of the relative ratio of myeloid to erythroid cells. Maturation of hematopoiesis is better evaluated at higher magnification (×20 to ×40), which shows the cytology. Maturation of each cell lineage has characteristic features: including increased nuclear segmentation and cytoplasmic granularity in the myeloid lineage, and chromatin condensation and cytoplasmic eosinophilia in the erythroid lineage.

Normally, immature myeloid cells are found along the bony trabeculae without forming clusters or large aggregates. Erythroid cells are distributed interstitially in small distinct islands. Megakaryocytes are scattered as individual cells in the interstitium, and in normal individuals, one to three are typically seen per high-power field. Detection of lymphomas and metastatic neoplasms is more reliable in histologic sections than on aspirate smears, and their histologic pattern provides important clues for the specific diagnosis, which can then be confirmed by immunohistochemistry.

The bony trabeculae should always be evaluated. Osteosclerosis or thickening can be seen as part of myelofibrosis or metabolic diseases of bone, whereas trabecular thinning is seen in osteopenia. Other characteristic patterns of pronounced osteoid seams, irregular bone resorption, or “mosaic” trabeculae patterns, are seen in such conditions as osteomalacia, osteitis fibrosa, and Paget disease, respectively.

Reticulin staining to evaluate for bone marrow fibrosis is routine in most laboratories, and should be scored based on established scoring systems. Additional stains, including immunohistochemistry, are performed depending on suspicious morphologic findings or specific clinical indications. Interpretation of these studies requires understanding of common artifacts and background, use of positive controls and negative controls, common expression patterns, and comparison with morphology seen on the corresponding area of the H&E section.