A 64-year-old man with coronary atherosclerotic heart disease, hypertension, and type-2 diabetes, with a 40-year history of smoking and alcohol consumption, was confirmed as a COVID-19 case on February 12, 2020. He was identified in a family cluster of COVID-19 virus infection (Figure 1A), in which 4 of 7 family members were confirmed to be COVID-19 cases and whose COVID-19 virus viral sequences were highly homologous. Overall, 3 family members recovered and were discharged by March, but the subject of this case report, who had moderately severe COVID-19, was isolated at the hospital due to persistent COVID-19 virus positivity until August of 2020, with an exception of two 2-week periods when he tested negative (4) (Figure 1A, Supplementary Table S1).
The patient had chills, fever (38.6 °C), sore throat, and loss of appetite from February 1 to his admission on February 12. Clinical examination revealed decreased white blood cell (WBC) and lymphocyte counts (Supplementary Table S2). Chest computed tomography (CT) showed patchy ground-glass opacities in the upper and the lower lobes under the pleura of both lungs and in the middle lobe of the right lung (Supplementary Figure S2). Throughout his hospitalization, no abnormalities were observed in his liver or kidney function or in routine blood examinations (Supplementary Table S3). Supportive evidence that he was immunocompetent was that all absolute cluster of differentiation 4 (CD4) counts were above 350/μL, the CD4/cluster of differentiation 8 (CD8) ratio was above 1, and he was HIV negative; however his CD8 cell counts and natural killer (NK) cell counts were low at times (Supplementary Table S3).
The patient’s neutralizing antibody titer was 1∶2,048 measured 13 days after illness onset; the titer peaked at 1∶8,192 on day 24 and subsequently declined, staying at 1∶384 for over 9 months (Figure 1B; Supplementary Table S4). The highest viral loads, as assessed by PCR cycle thresholds (Ct) in sputum (Ct=17.4) and nasopharyngeal specimens (Ct=23.4), occurred 4 to 5 months (days 114 and 132) after illness onset (Figure 1C, Supplementary Tables S5–S6). Overall, 3 sputum samples (days 107, 112, and 131) were positive for sub-genomic RNA (sgRNA), with respective Ct values of 32.25, 38.15, and 38.30 (Figure 1C). Detailed methods are in Supplementary Materials.
Phylogenetic analyses showed that all viruses belonged to lineage B (5) and were 99.95% to 99.98% homologous with reference strain NC 045512 — evidence that is incompatible with reinfection (Supplementary Figure S1A) (6). Five single nucleotide variants (SNVs) were observed when the case report subject was diagnosed, and sequences among two other family members were identical (the viral load of the third family member was too low to be sequenced). Mutations accumulated across the COVID-19 virus genome, increasing to 14 mutation sites on Day 151. In total, 3 amino acid substitutions in the S protein were observed in serial samples, including H655Y, Y200C, and D614G substitutions. The mutations effectively changed the lineage from B to B.1.1 (Supplementary Figure S1B–C).
Previous case reports have described immunocompromised patients shedding COVID-19 virus RNA for 105 days and 153 days; the longest, previously-reported duration of COVID-19 virus shedding in an immunocompetent COVID-19 patient was 61 days (1-3). To our knowledge, the patient we describe has the longest duration of viral shedding (169 days) with intra-host variants (151 days). The intra-host mutation rate was comparable to that seen with inter-host variants (7-8). However, unlike previous studies, which describe shorter infection periods, the mutations identified in this case appeared across the entire genome rather than in select hotspots, such as S and ORF8 genes (1-2).
Although previous studies suggested that COVID-19 patients positive for COVID-19 virus RNA following a period of being negative have little or no infectiousness (9-10), our evidence suggests infectiousness may last up to 151 days after symptom onset. Although lack of laboratory facilities precluded virus isolation and culture, the observed accumulated mutations and positive sgRNA 3–4 months after infection suggests continuous, on-going viral replication and therefore potential for transmission (1). We believe that more attention to the infectiousness of patients testing positive after a period of testing negative is warranted.
The patient did not have severe clinical symptoms, indicating that prolonged viral shedding can occur in moderately ill cases (11-12). Because frequent nucleic acid testing is normally only done in people with COVID-19-like symptoms, we may be under-detecting occurrence of long-term virus shedding (4).
We anticipate additional follow-up of this individual. Although a single case report may have limited generalizability, our observation that he shed virus despite his neutralizing antibody titers being much higher than titers we have seen with other patients at our facility (13) suggests that immune responses other than humoral responses may have important roles in virus clearance. Perhaps cellular immune responses and innate immune functions are important for eventual clearance of persistent infections. The low CD8 T-cell and NK cell counts may have prolonged the time required for virus elimination or may indicate an exhausted immune response; however, it is not clear which of these two possibilities is at play. Dynamic interactions of killer T-cells, COVID-19 virus infection, and the individual’s immunological function need to be evaluated holistically to understand risk factors for prolonged and infectious COVID-19 virus shedding. It may also be important to evaluate the role of the genetic background of the patient or of virus-host interactions and their contributions to prolonged viral shedding (14).
Acknowledgements: The patient; Zhida Cheng from Yidu Cloud (Beijing) Technology Co., Ltd.
Ethics: This study has obtained consent from participant and is approved by the Medical Ethical Committee of Beijing YouAn Hospital, Capital Medical University (approval number [2020]036).
Conflicts of Interest: No potential conflicts of interest was reported by the authors.