Vitreoretinal Lymphoma: Optimizing Diagnostic Yield and Accuracy - 23/03/22

Doi : 10.1016/j.ajo.2021.09.032

MATTHEW C. SANTOS ^a, ANGELA JIANG ^a,^b, ALBERT S. LI ^a,^c, P. KUMAR RAO ^a, BRADLEY WILSON ^a, GEORGE J. HAROCOPOS ^a,^d,^⁎
^a Washington University School of Medicine, John F. Hardesty Department of Ophthalmology and Visual Sciences (M.C.S., A.J., A.S.L., P.K.R., B.W., G.J.H.), St. Louis, Missouri
^b Oregon Health & Science University, Casey Eye Institute (A.J.), Portland Veteran Affairs Medical Center, Portland, Oregon
^c Zucker School of Medicine at Hofstra/Northwell, Vitreoretinal Consultants of New York, Northwell Health Eye Institute (A.S.L.), Great Neck, New York
^d Washington University School of Medicine, Department of Pathology and Immunology (G.J.H.), St. Louis, Missouri

^⁎Inquiries to George J. Harocopos, MD, 660 S Euclid Ave, MSC 8096-05-07, St. Louis, Missouri, 63110.660 S Euclid Ave

Résumé

PURPOSE

To determine whether the addition of adjunctive tests, including immunohistochemistry (IHC), cytokine analysis, flow cytometry, and IgH gene rearrangement testing, achieves improved diagnostic parameters compared with cytologic smears alone in the detection of vitreoretinal lymphoma (VRL). To determine which of these tests or combination of tests provide the greatest diagnostic utility.

DESIGN

Retrospective review to assess diagnostic value.

METHODS

This single university-affiliated tertiary care center study included data from 237 vitreous biopsies performed between 1999 and 2017 in patients with suspected VRL. From 1999 to 2008-2009, cytologic smears were the sole test performed (84 cases). The protocol initiated in 2008-2009 added the 4 additional diagnostic tests (153 cases). The sensitivity, specificity, positive predictive value, negative predictive value, diagnostic accuracy, and diagnostic yield were calculated. Parameters were calculated for tests individually, for all 5 combined, and all possible 2-, 3-, and 4-test combinations. For cytologic smears, diagnostic parameters were calculated both before and after the addition of adjunctive tests to our protocol and for the entire cohort.

RESULTS

Of the 237 vitreous biopsies, 50 samples (21%) were from patients with confirmed central nervous system lymphoma and/or actively treated central nervous system, systemic, or intraocular lymphoma. Diagnostic yields (95% CI) were 90% (85%–93%) for smears, 82% (72%–89%) for IHC, 91% (85%–96%) for cytokine analysis, 76% (67%–84%) for IgH gene rearrangement, and 50% (40%–60%) for flow cytometry. For smears, the sensitivity pre-protocol was 73% (39%–94%), compared with 87% (69%–96%) post-protocol. IgH gene rearrangement was the only test exhibiting low sensitivity (40%). The combination of smears, IHC, and cytokine analysis exhibited the highest diagnostic parameters, with sensitivity 92%, specificity 98%, and diagnostic yield 100%.

CONCLUSIONS

The combination of cytologic smears, IHC, and cytokine analysis seems to be a reasonable and sufficient protocol for the diagnosis of suspected VRL. IgH gene rearrangement and flow cytometry may be the most expendable tests from our protocol.

Le texte complet de cet article est disponible en PDF.