Diagnostic Role of Flow Cytometry Analysis of Bioptical Samples in the Diagnosis of Lymphoid Tumors

We reviewed flow cytometric immunophenotyping (FCI) data of 470 tissue suspensions suspected of being involved by lymphoma (371 lymph nodes, 263 surgical biopsies and 108 fine needle aspirations,16 spleens, 6 tonsils, and 77 other tissues) and we have compared corresponding histologic diagnosis. The screening panel of FCI (CD45, CD19, CD3, CD4, CD8, and sIgκ/sIgλ ratio) identified three main groups:

1. 239 cases with demonstrable light chain restriction (sIgκ/sIgλ ratio ≤ 0.5 or ≥4);

2. 37 cases with demonstrable T cell proliferation and polyclonal B cell population;

3. 194 cases with polyclonal B cell and normal T cell populations (146 cases) or with not detectable CD45 reactivity and other leucocyte markers (48 cases).

In group 1, 235 cases were then evaluated by means of the following markers: CD5, CD10, CD11a, CD11c, CD20, CD22, CD23, CD30, CD38, CD43, CD79a–b, CD103, CD138, FMC7, and, in group 2, 30 cases were evaluated by means of the following markers: CD1a, CD2, CD5, CD7, CD16, CD26, CD43, CD56, CD57, CD45RA/RO, anti-TCR-ab gd. The complete analysis identified in group 1 four main diagnostic subsets: A (26 cases) characterized by CD5+, CD23+/±, sIg dim (CLL-like); B (36 cases) characterized by CD5+, sIg bright (MCL/CLLv); C (89 cases) characterized by heterogeneous expression of Ig, CD20 and CD43 (lymphoproliferative syndromes CD5−, excluding hairy cell leukemia); D (84 cases) characterized by CD10+, CD20 bright, CD43− (follicular NHL like). In cluster B, FISH analysis for t(11;14) resulted positive in 15/27 cases; in cluster D, FISH analysis for t(14;18) resulted positive in 60/60 cases. Histologic analysis convalidated FCI data in 23 cases of subset A (88%), in 29 cases of subset B (80%) and in 60 cases of subset D (71%). Subset C included 39 cases of LCL and 15 cases of MZL. In group 2, all cases studied expressed an aberrant T phenotype but, with the exception of 1 case of NK proliferation, 1 case of likely Sezary syndrome/Micosis Fungoides (CD4+CD7−CD26−) and 1 case of T lymphoblastic lymphoma (CD3cy+TdT+), in the other 34 cases was not possible identify other peculiar subset. In this group, histologic analysis included 19 cases of NHL T (63%). Considering NHL vs no NHL, sensibility (SE) and specificity (SP) of FCI analysis resulted 88% and 92%, respectively; positive predictive value (PPV) was 96% and negative predictive value (NPV 78%). Considering B-cell NHL only, SE was 91%, SP 95%, PPV 97% and NPV 84%. Even if histology is the basis for the diagnosis of lymphoid tumors, however our study supports that FCI can play an important role mainly in B-NHL diagnosis, combining rapidity of analysis with a multiparametric analysis, also in presence of size limitated samples.