The Morphologic Diagnosis Of Multiple Myeloma

The morphologic diagnosis of multiple myeloma and its distinction from reactive plasmacytosis relies on both the quantity of plasma cells seen and the qualitative plasma cell abnormalities, as previously mentioned. In general terms, more than 30% of plasma cells in a marrow aspirate smears constitute a major diagnostic criterion for multiple myeloma, although such a percentage may occur in other conditions, such as rheumatic disorders, inflammatory reactions of the bone marrow, and in particular, HIV infection. The demonstration of light-chain isotype restriction has become a major criterion for distinguishing malignant from reactive cases. Nuclear-cytoplasmic asynchrony is more evident in malignant plasma cells. Recently, the international myeloma working group has suggested criteria for the distinction of monoclonal gammopathy of unknown significance (MGUS), smoldering myeloma, and symptomatic myeloma. These criteria do not rely on strict morphologic criteria, but stress the clonal or light-chain isotype restriction of plasma cells.2

Despite attempts to define morphologic and or cyto-chemical features that distinguish neoplastic plasma cells from normal plasma cells, no morphologic criteria appear to be pathognomonic. Spindle-shaped crystal deposits and other cellular inclusions (see Figures 81.1 and 81.2) typical of myeloma can occasionally be seen in normal plasma cells. The diffuse eosinophilic pink staining at the cell periphery, the so-called "flame" cells (see Figure 81.3) seen mostly in IgA myeloma, can occur in other variants, as can Russell bodies which, while initially thought to be organisms, are now known to be immunoglobulin (Ig), have been seen in normal plasma cells. Variations in cell size, in multiple nucleolarity, and even cytochemical staining are not pathognomonic (see Figure 81.4). Most plasma cells stain positively for acid phosphatase, nonspecific esterase, and periodic acid-Schiff, but these are not specific for myeloma. The presence of plasmablasts (see Figure 81.5), nucleolated plasma cells often with less eccentricity and less intense staining of cytoplasm, is classical of myeloma. The number of plasmablasts has also been shown to correlate with prognosis. The presence of a large number of plasmablasts per se is probably the most important morphologic feature that distinguishes myeloma from reactive plasmacytosis, and a population of plasmablasts greater than 2% of plasma cells have been shown by the Mayo Clinic to be an adverse prognostic factor.3

In early myeloma, plasma cells distribute in an interstitial pattern. At a later stage, they form dense aggregates on endosteal surface, followed by nodules and sheets in advanced disease. These patterns correlate with

Figure 81.1 Bone marrow aspirate of a patient with myeloma. Needle-shaped, azurophilic crystalline inclusions in cytoplasm of plasma cells representing Ig inclusion material

Figure 81.2 Bone marrow aspirate of a patient with myeloma. Mott cells are plasma cells with multiple bluish cytoplasmic inclusion bodies

survival, as do other major prognostic factors: plasma cell labeling index and serum ^-microglobulin.4-6

The growth pattern of myeloma on trephine is also predictive of the type of skeletal defects and correlates strongly with magnetic resonance imaging findings. Nodules of plasma cells are associated with osteolytic lesions, whereas interstitial and "sarcomatous" types are associated with osteoporosis.7

The percentage of plasma cells on bone marrow aspirates is used as a diagnostic criterion for myeloma. According to the World Health Organization classification, major diagnostic criteria require 30% plasma-cytosis, while minor criteria require 10-30% for the diagnosis of myeloma. For MGUS, marrow plasmacy-tosis is defined as <10%.

Monoclonality of myeloma can be demonstrated on trephine by kappa or lambda light-chain restriction by immunohistochemical stain. CD138/syndecan-1 is a useful immunohistochemical marker of normal and neoplastic plasma cells on bone marrow trephine.

0 0

Post a comment