OUP user menu

Usefulness and Reproducibility of Cytomorphologic Evaluations to Differentiate Myeloma From Monoclonal Gammopathies of Unknown Significance

Fuensanta Millá MD, Albert Oriol MD, JoséLuis Aguilar MD, Ana Aventín MD, PhD, Ramon Ayats MD, PhD, Esther Alonso MD, Alícia Domingo MD, Evarist Feliu MD, PhD, Lourdes Florensa MD, PhD, Andrés López MD, PhD, Encarna Pérez-Vila MD, María Rozman MD, PhD, Carmen Sánchez MD, Teresa Vallespí MD, PhD, Soledad Woessner MD, PhD
DOI: http://dx.doi.org/10.1309/34D8-V2KU-23UL-VFBW 127-135 First published online: 1 January 2001

Abstract

We attempted to differentiate monoclonal gammopathies of unknown significance (MGUS) and multiple myeloma (MM) on morphologic grounds and to determine interobserver reproducibility of the differentiation. Cytologists blindly evaluated bone marrow smears from 154 patients with bone marrow plasmacytosis for the proportion of plasma cells with predefined cellular atypias. The single morphologic characteristic that most strongly differentiated MM from MGUS was the presence of nucleoli. The percentage of plasma cells, cytoplasmic contour irregularities, and anisocytosis also predicted a diagnosis of myeloma in multivariate analysis. Six cytologists independently evaluated 68 consecutive cases to determine sensitivity and specificity of these cytomorphologic features. The interobserver coefficient of variation for the plasma cell count was 33%. On consideration of the diagnosis, 36 of 41 MGUS cases and all 24 cases of myeloma were classified correctly. The use of a predesigned score system did not present such a bias, although it did not improve overall efficiency. The plasma cell count is the most predictive characteristic of myeloma from a cytologic viewpoint, but the interobserver variability is high. Interobserver variability is also high in the assessment of morphologic atypia, and atypical traits are not uncommon in plasma cells in MGUS.

Key Words:
  • Myeloma
  • Monoclonal gammopathy of undetermined significance
  • Bone marrow cytomorphologic examination
  • Reproducibility