Latent Nuclear Antigen Flow-FISH Improves the Detection of Circulating Viroblasts during KSHV/HHV-8-Associated Multicentric Castleman Disease Flares

Kaposi sarcoma-associated herpes virus (KSHV)/Human Herpesvirus 8 (HHV-8)-associated multicentric Castleman disease (MCD) is an inflammatory lymphoproliferative disorder mainly affecting immunocompromised hosts. Gold-standard diagnosis of KSHV/HHV-8 MCD currently relies on lymph node biopsy showing plasma-cell or mixed type Castleman disease associated to the pathognomonic presence of KSHV/HHV-8-infected cells located in the mantle zone. Hemophagocytic lympho-histiocytosis can complicate KSHV/HHV-8 MCD and lead to multiple organ failure and coagulopathy, delaying invasive sampling procedures. Our group previously reported the detection of circulating KSHV/HHV-8-infected cells (called KSHV/HHV-8-infected viroblasts, or KIVs) using standard flow cytometry in a small series of patients with KSHV/HHV-8-MCD flare. These cells had the same phenotype as those described in lymph node, were IgM, lambda and CD38 positive, and CD20 and CD24 negative. While specificity was high (100%) in this preliminary study, flow cytometry lacked sensitivity. The Flow-FISH technique, combining fluorescence in situ hybridization and flow cytometry, offers the advantage of viral transcripts direct and could improve the detection and characterization of KIVs during KSHV/HHV-8 MCD flares. Flow-FISH targeting Latent Nuclear Antigen (LNA) transcript of KSHV/HHV-8 was performed on peripheral blood mononuclear cells (PBMC) obtained from a large cohort of patients with KSHV/HHV-8 MCD flares and compared to standard flow cytometry. Fifty patients with KSHV/HHV-8 MCD flare were included in the study. All had histological confirmation of KSHV/HHV-8 MCD. Forty (80%) were male with a median age of 54. Thirty-two (64%) had a history of KSHV/HHV-8 MCD before the present flare, and 27 (54%) were living with HIV. Fifteen (30%) had a history of KS and none had history of PEL. Median C-reactive protein (CRP) levels and KSHV/HHV-8 whole blood viral load were 92 mg/L and 5.9 log copies/mL, respectively. Standard multiparametric flow cytometry was performed on PBMC in all patients. This technique was able to detect KIV, previously described as IgM+CD38highCD24-lambda+ cells, in 31 (62%) patients with KSHV/HHV-8 MCD flare (flow-cytometry positive group, or FC+). The percentage of IgM+CD38highCD24-lambda+ cells varied from 0.01% to 9.23% (median at 0.29%) among the CD3-CD14- population. Further extracellular characterization showed variable expression of CD19, CD20, CD27 and CD86 antigens. Flow-FISH was performed in 13 individuals in the FC+ group and showed the presence of LNA transcripts in all patients with LNA+ cells varying from 0.20% to 9% (median at 0.98%) of the CD3-CD14- population. All infected cells had a KIV phenotype and no other infected population was detected. Flow-FISH was performed on 11 samples from the 19 patients with KSHV/HHV-8-MCD flare but without detectable IgM+CD38highCD24-lambda+ population (FC- group). We were able to detect a significant LNA+ population in six additional cases (6/11, 54%, FC-FF+) with LNA+ cells varying from 0.01% to 0.1% (median at 0.04%) of the CD3-CD14- population. Once again, all infected cells had a KIV phenotype and no other infected population was detected. Two patients from the FC-FF- group received corticosteroids and/or etoposide before sampling. Overall, among the 42 patients for whom both standard flow cytometry and Flow-FISH could be performed, 37 (88%) had a detectable KIV population, supporting a role of Flow-FISH in enhancing sensitivity to detect these cells. We then compared the 31 patients from the FC+ group to the 19 patients from the FC- group. CRP level was significantly higher in the FC+ group (median at 108 mg/L vs 60 mg/L, p=0.03), as well as KSHV/HHV-8 DNA viral load (median at 6.3 copies/mL vs 5.2 copies/mL, p=0.012). The platelet count was significantly lower in the FC+ group (79 G/L vs 200 G/L, p=0.004). Combination of flow cytometry and Flow-FISH had a sensitivity of 88% to diagnose KSHV/HHV-8 MCD flare. Combination of flow cytometry and Flow-FISH emerges as a rapid and sensitive tool when suspecting KSHV/HHV-8 MCD flare, that might help the clinician to start an appropriate and urgent treatment without waiting for the result of lymph node biopsy that will further confirm the diagnosis.