Large-scale screening of SARS-CoV-2 variants in Tokyo, Japan: A 3-year and 9-month longitudinal survey

Abstract

Over nearly four years (March 10, 2021-December 31, 2024), we performed a comprehensive longitudinal analysis of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants among patients in a single hospital in Tokyo, Japan. Using RT-qPCR and Sanger sequencing, complemented by whole-genome sequencing, we tested nasopharyngeal swab samples (n = 4,628) and tracked the emergence and evolution of variants of concern (VOCs). The findings demonstrate the utility of a hospital-based SARS-CoV-2 variant surveillance system for informing clinical decision-making and public health settings, including: i) serving as a reference for selecting appropriate treatments, ii) enabling early detection of VOCs, iii) contributing to the development of hospital infection control guidelines, iv) fostering cooperation with local governments, v) supporting cohort studies, and vi) identifying long-term SARS-CoV-2 infections. This work underscores the importance of real-time variant monitoring for mitigating the effects of pandemics and provides essential epidemiological and clinical data that can guide future outbreak management and policy development.

Keywords: COVID-19; clinical data; epidemiology; sequencing; variant of concern (VOC).

Figure 1.

Inferring SARS-CoV-2 variants using Sanger sequencing. (A) RT-PCR amplification regions targeted in this study. Fragments 1 (covering amino acid residues 1-70) and 2 (covering amino acid residues 332-505) include key mutations that enable the identification of XEC and VOC, respectively. (B) The resulting sequence electropherograms were aligned with the Wuhan-Hu-1 reference sequence (NC_045512.2). (C) SARS-CoV-2 variants were inferred based on mutation patterns using the Stanford SARS-CoV-2 Mutations Analysis tool. (D, E) Prevalence of mutations within the sequenced S region for Variants of Concern (VOC), Variants of Interest (VOI), and the R.1 variant, one of the dominant variants during Japan's 4^th wave, is shown. Mutations present in at least 75% of the sequences associated with each lineage are displayed. For VOC, an asterisk (*) is attached to each notation. Data were obtained from GISAID (accessed on 14 March 2025). The heatmap was generated using the R package outbreakinfo 0.2.0 (). aa: amino acid residues.

Figure 2.

Weekly epidemiological distribution of SARS-CoV-2 variants between March 10, 2021, and December 31, 2024. (A) Data derived from patients diagnosed with COVID-19 at NCGM, Tokyo, Japan (n = 3,423). Sequences with no detectable single mutations were classified as "non_VOC". (B) All domestic samples (≥ 27,000 nucleotides) registered in the GISAID EpiCoV database (n = 642,096). A magnified view of the data from week 46 of 2023 onward is shown in the upper-right corner. (C, D) The number of new SARS-CoV-2 infections and severe cases in Japan and Tokyo based on open data from the Ministry of Health, Labour and Welfare ().

Figure 3.

Relative numbers of newly infected cases in Japan (solid purple), in Tokyo (solid green), severe cases in Japan (dashed black), SARS-CoV-2 variants inferred at NCGM (solid red), and domestic genome sequences registered in the GISAID database (solid orange) are shown. The Y-value at the peak of the 5^th wave (Delta peak, week 33 of 2021) is set to 1.

Figure 4.

Weekly epidemiological relative frequency of SARS-CoV-2 variants between January 1, 2020, and December 31, 2024. (A) All worldwide (n = 16,422,993) and (B) domestic (n = 677,162) samples (≥ 27,000 nucleotides) registered in the GISAID EpiCoV database.

Figure 5.

(A, D) Comparison of Cq values between successfully amplified (n = 30) and not amplified (n = 30) samples during the 2^nd PCR. Among the 60 samples, (B, E) samples collected during the epidemic growth phase of the 8^th wave (weeks 47-48 of 2022) and (C, F) samples collected during the epidemic decline phase (weeks 8-10 of 2023) are shown separately. Cq values were quantified using the SARS-CoV-2 Detection Kit -Multi- (TOYOBO) by amplifying two regions (N1 and N2 primer/probe sets) derived from the N gene. n.d.: not detected.