Investigation of lymphoid lesions of the head and neck using combined fine needle aspiration cytology and flow cytometry: Accuracy and pitfalls

Objective We reviewed the diagnostic utility of combined fine needle aspiration cytology (FNAC) and flow cytometry (FC) in the diagnosis of lymphoid lesions of the head and neck. Method In total, 1402 patients with combined FNAC-FC reports were correlated with follow-up information. Rapid on-site evaluation (ROSE) of cytological specimens was performed in 52% of cases. Results In total, 211 lymphoid malignancies were identified, including 198 non-Hodgkin lymphoma (NHL) and 13 Hodgkin lymphoma (HL). Accuracy measures for NHL were: sensitivity 95.5%; specificity 99.9%; PPV 99.5%; NPV 99.2%; accuracy 99.3%. Only seven of 13 cases of HL were detected by FNAC-FC. False negative cases included HL (six cases), diffuse large B-cell lymphoma (four), T-cell lymphoma (two), follicular lymphoma (one), marginal zone cell lymphoma (one) and B-cell NHL, not otherwise specified (one). Two false positive results were identified: one immunoblastic hyperplasia reported as suspicious for HL and one case reported as suggestive of NHL that was found to be reactive hyperplasia. Cases collected with ROSE had a significantly lower rate (P < 0.0001) of insufficient cells for FC analysis (7.0%) than cases where ROSE was not performed (16.4%). Sensitivity (P < 0.0001) and NPV (P = 0.0023) were significantly higher for ROSE-collected specimens. None of the false-negative NHL cases had ROSE performed. Conclusions FNAC-FC is a highly sensitive and specific test for NHL. Diagnostic errors mostly involved HL, large cell lymphomas and T-cell lymphomas. ROSE results in a significantly higher adequacy rate for FC and higher sensitivity for NHL.