Multisite Skin Biopsies vs Cerebrospinal Fluid for Prion Seeding Activity in the Diagnosis of Prion Diseases

Abstract

Recent studies have revealed that autopsy skin samples from cadavers with prion diseases (PRDs) exhibited a positive prion seeding activity similar to cerebrospinal fluid (CSF). It is worthwhile to validate the findings with a large number of biopsy skin samples and compare the clinical value of prion seeding activity between skin biopsies and concurrent CSF specimens. To compare the prion seeding activity of skin biopsies and CSF samples and to determine the effectiveness of combination of the skin biopsies from multiple sites and numerous dilutions on the diagnosis for various types of PRDs. In the exploratory cohort, patients were enrolled from September 15, 2021, to December 15, 2023, and were followed up every 3 months until April 2024. The confirmatory cohort enrolled patients from December 16, 2023, to June 31, 2024. The exploratory cohort was conducted at a single center, the neurology department at Xuanwu Hospital. The confirmatory cohort was a multicenter study involving 4 hospitals in China. Participants included those diagnosed with probable sporadic Creutzfeldt-Jakob disease or genetically confirmed PRDs. Patients with uncertain diagnoses or those lost to follow-up were excluded. All patients with PRDs underwent skin sampling at 3 sites (the near-ear area, upper arm, lower back, and inner thigh), and a portion of them had CSF samples taken simultaneously. In the confirmatory cohort, a single skin biopsy site and CSF samples were simultaneously collected from a portion of patients with PRDs. The skin and CSF prion seeding activity was assessed using the real-time quaking-induced conversion (RT-QUIC) assay, with rHaPrP90-231, a Syrian hamster recombinant prion protein, as the substrate. In the exploratory cohort, skin samples were tested at dilutions of 10-2 through 10-4. In the confirmatory cohort, skin samples were tested at a dilution of 10-2. A total of four 15-μL wells of CSF were used in the RT-QUIC assay. Correlations between RT-QUIC results from the skin and CSF and the final diagnosis of enrolled patients. In the exploratory cohort, the study included 101 patients (mean [SD] age, 60.9 [10.2] years; 63 female [62.4%]) with PRD and 23 patients (mean [SD] age, 63.4 [9.1] years; 13 female [56.5%]) without PRD. A total of 94 patients had CSF samples taken simultaneously with the skin biopsy samples. In the confirmatory cohort, a single skin biopsy site and CSF sample were taken simultaneously in 43 patients with PRDs. Using an experimental condition of 10-2 dilution, the RT-QUIC positive rates of skin samples from different sites were comparable with those of the CSF (skin: 18 of 26 [69.2%] to 74 of 93 [79.6%] vs CSF: 71 of 94 [75.5%]). When tested at 3 different dilutions, all skin sample positivity rates increased to over 80.0% (79 of 93 for the near-ear area, 21 of 26 for the upper arm, 77 of 92 for the lower back, and 78 of 92 for the inner thigh). Combining samples from skin sites near the ear, inner thigh, and lower back in pairs yielded positivity rates exceeding 92.1% (93 of 101), significantly higher than CSF alone (71 of 94 [75.5%]; P =.002). When all skin sample sites were combined and tested at 3 dilution concentrations for RT-QUIC, the sensitivity reached 95.0% (96 of 101). In the confirmatory cohort, the RT-QUIC positive rate of a single skin biopsy sample was slightly higher than that of the CSF (34 of 43 [79.1%] vs 31 of 43 [72.1%]; P = .45). Results of this diagnostic study suggest that the sensitivity of an RT-QUIC analysis of a combination of 2 or more skin sites was superior to that of CSF in diagnosing PRDs.