# Diagnostic Performance of Second-Generation Fujifilm SILVAMP TB LAM in Nigerian Adults with Presumptive Tuberculosis

## TL;DR

A new urine test for tuberculosis (TB) called FujiLAM v.2 shows similar accuracy to its older version in diagnosing TB in Nigerian adults.

## Contribution

The study evaluates and compares the diagnostic accuracy of the second-generation FujiLAM v.2 assay with its first-generation counterpart in a Nigerian population.

## Key findings

- FujiLAM v.2 had a sensitivity of 58.3% and specificity of 97.3%.
- The new assay showed strong agreement with the first-generation test (97.1% agreement, κ = 0.80).
- No significant differences were found between FujiLAM v.2 and v.1 across TB or HIV subgroups.

## Abstract

Background: Tuberculosis (TB) remains a leading cause of death worldwide, with major diagnostic challenges in low-resource settings. Urine-based lipoarabinomannan (LAM) assays provide a non-invasive option, particularly for people living with HIV who may struggle to produce sputum. Fujifilm has developed a second-generation SILVAMP TB LAM assay (FujiLAM v.2) to improve diagnostic performance. This study aimed to evaluate the diagnostic accuracy of FujiLAM v.2 among Nigerian adults with presumptive TB and directly compare it with the first-generation assay (FujiLAM v.1) using the same set of urine samples. Methods: We conducted a retrospective analysis among Nigerian adults with presumptive TB (n = 178). Stored urine samples collected in Abuja were retested with FujiLAM v.2 and compared with results previously obtained using FujiLAM v.1 on the same specimens. Xpert MTB/RIF and culture served as the reference standard. Sensitivity, specificity, agreement with the reference standard (Cohen’s Kappa), and differences (McNemar’s test) were assessed. Results: FujiLAM v.2 demonstrated sensitivity of 58.3% and specificity of 97.3%, comparable to FujiLAM v.1 (58.3% and 98.0%, respectively). No significant differences were found between test versions across TB or HIV subgroups (p > 0.05). Overall agreement between assays was 97.1% (κ = 0.80). Conclusions: FujiLAM v.2 showed diagnostic performance consistent with the first-generation assay, with similar sensitivity and specificity to the earlier version. These findings support its potential as a rapid, non-sputum-based diagnostic tool to complement TB testing in resource-limited settings. Further studies are needed to assess its implementation within TB diagnostic algorithms, including studies with multiple production lots.