Coronaviruses primarily cause respiratory infections in humans. Among the seven recognized human coronaviruses (HCoVs), HCoV-229E, HCoV-NL63, HCoV-OC43, and HCoV-HKU1 were the most prevalent endemic HCoVs before the emergence of severe acute respiratory syndrome-CoV-2 [1–3]. Common HCoVs can be detected using multiplex reverse transcription real-time PCR (RT-qPCR) for respiratory viruses. The Allplex respiratory panel assay (Seegene, Seoul, Korea), a widely used RT-qPCR kit in Korea and globally [4, 5], targets HCoV-OC43, HCoV-NL63, and HCoV-229E; however, HCoV-HKU1 is not included. Two HCoV-HKU1 cases were misidentified as HCoV-OC43 using the Allplex assay in Italy [6]. Therefore, we conducted, for the first time, a comprehensive validation study to determine whether the Allplex assay misidentifies HCoV-HKU1 as HCoV-OC43 and to characterize the laboratory manifestations, extent of misidentification, and potential clinical implications.

Nucleic acids from all specimens identified as HCoV-OC43 using the Allplex assay between November 2018 and April 2024 at Hanyang University Guri Hospital, Guri, Korea, were analyzed using pan-coronavirus sequencing targeting the RNA-dependent RNA polymerase gene region [7, 8] and a RT-qPCR assay certified for in vitro diagnostic use under the European CE marking (PowerChek Coronavirus 229E/OC43/NL63/HKU1 real-time PCR kit; KogeneBiotech, Seoul, Korea). Allplex results were analyzed using Seegene Viewer (v3.06–v3.33; Seegene) and interpreted per the manufacturer’s guidelines. The Institutional Review Board of Hanyang University Guri Hospital approved the study protocol (2024-11-018) and waived the need for informed consent given the study’s retrospective nature.

Ninety-five specimens tested positive for HCoV-OC43 using the Allplex assay. Among these, 62 were species-identified via sequencing (HCoV-OC43, N=47; HCoV-HKU1, N=15). PowerChek assay results were consistent with the sequencing results of these 62 specimens. Additionally, 26 HCoV-OC43 cases and four HCoV-HKU1 cases were detected using the PowerChek assay among 30 specimens in which sequencing failed due to low viral load. Neither method yielded valid results for three specimens with Ct values of 38.5–41.1 in the Allplex assay. In HCoV-OC43 cases, Ct values were comparable between the Allplex and PowerChek assays (Fig. 1A). Conversely, HCoV-HKU1 cases demonstrated substantially lower Ct values in the PowerChek assay than in the Allplex assay (Fig. 1B).

Among two cases with low Allplex Ct values (29.02 and 29.23, respectively), in the HCoV-HKU1 case, the amplification efficiency decreased in the exponential phase. The final fluorescence intensity was substantially lower (1,400 relative fluorescence units [RFU]) than that for the HCoV-OC43 case (4,700 RFU) (Fig. 2A). A similar trend was observed in two cases with high Ct values (39.56 and 39.62), where final fluorescence intensities were 550 RFU for HCoV-HKU1 and 1,150 RFU for HCoV-OC43 (Fig. 2B). Across Ct values, final fluorescence intensity was consistently lower for HCoV-HKU1 than for HCoV-OC43 (Fig. 2C).

We experimentally confirmed that the Allplex assay misidentifies HCoV-HKU1 as HCoV-OC43 in 20% of cases initially labeled as HCoV-OC43. Misidentification was characterized by delayed Ct values (>28) despite high viral concentrations and reduced amplification efficiency with lower final RFU in amplification curves. Although these findings may not directly affect clinical interpretation, thorough assessment of amplification curves could help ensuring the reliability for RT-qPCR assays and identifying any need for additional testing. Reduced RT-qPCR amplification efficiency may result from suboptimal primer binding [9], and partial binding of HCoV-OC43 primers to HCoV-HKU1 sequences may account for the observed misidentification. However, the precise mechanism could not be determined owing to the lack of publicly available molecular information on the assay.

Clinically, the four common HCoVs do not markedly differ from one another in their manifestations, treatment strategies, or prognoses [1–3, 10]. The misclassification of HCoV-HKU1 as HCoV-OC43 is not expected to have substantial clinical implications in routine clinical settings. As of 2024, in Korea, HCoV-HKU1 can only be identified using the FilmArray assay (BioMérieux, Paris, France) among routinely used clinical RT-qPCR kits. Therefore, had the nature of the Allplex assay been communicated to clinicians, the detection of HCoV-OC43/HKU1 using the Allplex assay could have provided clinically relevant information by suggesting a potential viral etiology of respiratory symptoms. Additionally, providing information on assay characteristics may be helpful in contexts such as research on diagnostic accuracy or epidemiologic surveillance, where precise viral identification is essential.

This study has certain limitations. First, the single-institution setting limits the generalizability of our findings. Second, prolonged nucleic acid storage during the study period may have influenced test performance. Finally, the mechanism underlying misidentification could not be elucidated, and a corrective solution could not be proposed.

In conclusion, our findings highlight the crucial role of clinical laboratory in thoroughly evaluating the available data when reviewing test results to ensure accurate outcomes.