The authors evaluated NG-TEST Acineto-5, a lateral flow immunoassay, on 433 whole-genome-sequenced bacterial isolates. The panel included single producers of common acquired carbapenemases and isolates carrying multiple enzymes. The study team reported that the assay detected all targeted carbapenemases and every tested variant and combination, which the authors state translated to 100% sensitivity in their dataset. They also reported no false positives among isolates lacking the targeted carbapenemases, yielding 100% specificity in this collection. These performance metrics were derived directly from the described isolate panel and the investigators’ analysis.
Short, direct claim: the assay identified every targeted enzyme in the tested samples. That result, if reproducible, would support rapid phenotypic identification at the point of care or in clinical laboratories. The manuscript provides those detection outcomes; however, the source did not report external validation in routine clinical workflows beyond the sequenced collection.
The collection was dominated by Acinetobacter baumannii (n = 365) and included Acinetobacter pittii (n = 33); thirteen additional Acinetobacter species were present at low counts (1–7 isolates each). The authors enumerated targeted enzyme groups: OXA-23-like (n = 126), OXA-24-like (n = 14), OXA-58-like (n = 8), OXA-143/253/542 (n = 10), VIM-like (n = 9), and NDM-like (n = 25). Co-producers were represented as well: 20 isolates carried two carbapenemases and 2 isolates carried three. The study therefore covered both single-enzyme and multi-enzyme backgrounds, giving a broad view of assay performance across genotypes detected by whole-genome sequencing.
The investigators included isolates that carry intrinsic OXA-type genes. They observed weak positive signals in four of five A. radioresistens isolates that harbor intrinsic blaOXA-23-like genes; the authors interpreted those weak signals as consistent with low-level gene expression rather than strong acquired resistance gene expression.
Data indicated that the same antibody pair within the assay detected OXA-24-like, OXA-143-like, and OXA-542 enzymes, demonstrating cross-reactivity within that subgroup. The manuscript reported no false positives among isolates without targeted carbapenemases, which the authors used to support a specificity estimate of 100% for this sample set. Sensitivity was reported as 100% across all targeted enzymes and combinations in the collection.
The source does not report certain operational details in the abstract: procedural steps, turnaround time, limits of detection, or comparison with alternative rapid molecular or phenotypic assays were not provided there. The main text, the authors state, contains all data produced in the work; readers should consult the full manuscript for granular assay procedures and any supplementary analytical metrics.
The authors emphasized that detection of NDM and VIM may have specific clinical consequences because these metallo-β-lactamases are not inhibited by durlobactam. They also noted an association—reported in the literature and cited in the manuscript—between these enzymes and increased cefiderocol MICs, which can limit therapeutic options for carbapenemase-producing Acinetobacter. Rapid identification of these enzymes, the investigators argued, could therefore inform infection-control measures, epidemiological surveillance, and therapeutic decision-making, given the frequent multidrug-resistant phenotype of carbapenemase-producing isolates.
Short declarative point: knowing whether an isolate produces NDM or VIM can affect anticipated activity of specific agents and stewardship choices.
The authors disclosed that several contributors (A.C., L.F., M.L.Q., and C.M.) are employees of NG Biotech and participated in the development of the evaluated kit; NG-TEST Acineto-5 devices were provided free of charge. The manuscript states that NG Biotech was not involved in experimental evaluation, data analysis, interpretation, or manuscript preparation. The study team reported that data analysis and interpretation were carried out by investigators not involved in kit development and that results were independently reviewed by two researchers from each investigator team.
Ethics statements in the preprint indicate that isolates were de-identified and that relevant ethical guidelines and approvals were followed. The authors also declared that all data produced are contained in the manuscript.
This report is a preprint and has not undergone peer review; the authors explicitly state it should not be used to guide clinical practice. The abstract does not provide operational parameters such as assay workflow times, limits of detection, or head-to-head comparisons with other rapid diagnostics—details that the source did not report in the abstract and that readers must consult the full manuscript to confirm. The study’s dataset is laboratory-based, drawn from whole-genome-sequenced isolates; the authors did not describe prospective clinical implementation or performance in routine diagnostic laboratories within the abstract. Finally, manufacturer involvement in kit provision and author employment is declared; the investigators reported independent analysis and review, but external, peer-reviewed replication will be necessary to confirm these findings.
The authors posted the full manuscript and data with the preprint; subsequent peer review and independent validation studies are the next documented steps needed before clinical adoption.