Incidence of delayed asymptomatic COVID-19 recurrences in a 6-month longitudinal study

To the Editor,

It has recently been suggested that prior SARS-CoV-2 infection is associated with protection against symptomatic reinfection [1,2]. The role of protective immunity after COVID-19 has been assessed in population-based and cohort studies, where symptomatic recurrences with positive SARS-CoV-2 RT-PCR results were investigated [2,3], usually lacking genomic sequencing to confirm reinfection. However, limited data are available to date about the frequency of long-term asymptomatic reinfections and/or recurrences. Because of their confirmed transmission risk [4], asymptomatic infections also have significant epidemiologic implications in terms of public health control. To answer this question, longitudinal studies with sequential sampling following SARS-CoV-2 infection would be required, ideally including sequencing of viral genomes to discern between reinfection and disease recurrence. Recently, the Centers for Disease Control (CDC) have proposed an investigation protocol for identifying cases with a high index of suspicion for reinfection [5], that prioritizes new detection of SARS-CoV-2 RNA ≥90 days since first infection, whether or not symptoms are present, availability of paired respiratory specimens with a RT-PCR cycle threshold (Ct) value <33, and genomic sequencing to confirm reinfection. An acknowledged limitation of the protocol consists in the exclusion of asymptomatic or mildly symptomatic individuals who never seek testing for SARS-CoV-2. We conducted a prospective study in a cohort of patients hospitalized for microbiologically-confirmed COVID-19 in the first wave, who were longitudinally followed-up during a 6-month period with sequential nasopharyngeal and blood sampling. We evaluated the incidence of late reinfections and recurrences, both symptomatic and asymptomatic, and validated the CDC predictive criteria to identify late reinfections occurring in our cohort by genomic sequencing of the suspected cases. Blood and nasopharyngeal samples were obtained during hospital stay and at 1, 2 and 6 months after patients’ discharge for measuring antibody levels and SARS-CoV-2 RNA. IgG antibody plasma levels against the SARS-CoV-2 internal nucleocapsid protein (N-IgG) and the spike protein (S-IgG) (Anti-SARS-CoV-2 IgG ELISA, Euroimmun, Lubeck, Germany) were tested, and SARS-CoV-2 RNA was detected by RT-PCR (AllplexTM 2019-nCoV Assay, Seegene, Seoul, Korea) which targeted the E, RdRP, and N genes. Genome sequencing of SARS-CoV-2 was performed on nasopharyngeal samples following ARTIC amplicon sequencing protocol for MinIon version V3- Phylogenetic analysis was done using webserver Nextstrain (https://nextstrain.org/), with the SARS-CoV-2 database Nextclade (https://clades.nextstrain.org/).

146 patients admitted for COVID-19 were followed-up. Median age was 64 years, 88 (60.3%) were male, and 72.6% had coexisting comorbid diseases. SARS-CoV-2 shedding lasted a median (Q1-Q3) of 13 (2.2–33.8) days, median (Q1-Q3) time from illness onset to seropositivity was 12 (8–15) days, and peak S-IgG was 5.9 (0.3–7.1) absorbance/cut-off (S/CO) and peak N-IgG 4.1 (0.3–4.9) S/CO. At 1 month after discharge, 40/146 (27%) subjects tested positive for SARS-CoV-2 RNA; 15/127 (11.8%) at 2 months, and 5/134 (3.7%) at 6 months. We analyzed the 5 patients with positive RT-PCR occurring more than 90 days since first COVID-19 diagnosis (Table 1 ). Median (range) time from diagnosis to new detection of SARS-CoV-2 RNA was 183 (167–204) days. Cases included 3 men, with ages ranging from 44 to 73 years, and 3 of them had subjacent comorbidity. Two patients were readmitted to hospital at re-positivity, and 3 patients remained asymptomatic. Only one patient had a Ct<33, and in the other four patients the Cts ranged from 33 to 38. Genomic sequencing was performed in 4 individuals with available paired samples. In the three patients with Ct≥33, all of them asymptomatic, the same clade 20B was detected. In two of them, the clade showed the same hallmark single nucleotide variants. In the third patient, the follow-up sample showed two new mutations, a K374R substitution in the N gene and an A222V substitution in the S gene, probably reflecting adaptive viral changes associated to persistent infection. Genomic sequencing of the symptomatic patient with a Ct of 18 showed phylogenetically distinct genomic sequences; the first sample was member of the clade 20A, and the most recent sample was member of the clade 20B. The 3 patients with asymptomatic recurrence and the symptomatic patient with no sequencing data showed detectable antibody levels at the time of SARS-CoV-2 RNA re-positivity, ranging from 3.01 to 6.01 S/CO for S-IgG and 2.6 to 2.46 S/CO for N-IgG. The patient with symptomatic reinfection had no detectable antibody levels at the time of re-positivity.

Table 1.

Characteristics of patients with late SARS-CoV-2 recurrence or reinfection.

Patient	#1	#2	#3	#4	#5
Sex	Female	Male	Female	Male	Male
Age	44	54	64	73	52
Comorbidity	No	No	Breast cancer in remission	Mental retardation, epilepsy	Refractory hypertension, heart disease, CKD, obesity
Highest WHO severity score at 1st infection	4	4	3	4	3
Therapy for 1st infection	HCQ, AZIT, LPV/r	HCQ, AZIT, LPV/r, steroid bolus	HCQ, AZIT, LPV/r, interferon	HCQ, AZIT	None
Hospital stay at 1st infection, days	10	13	10	11	5
Time to recurrence, days	167	183	181	204	240
Symptoms at recurrence	No	No	No	Yes*	Yes
No. of negative RT-PCR tests before recurrence	6	2	1	0	1
Cycle threshold at recurrence	33	34	38	36	18
Peak S-IgG, S/CO	4.14	13.8	7.24	75.10&	NA
Peak N-IgG, S/CO	9.51	4.89	4.53	NA	NA
S-IgG at recurrence, S/CO	3.04	5.62	6.01	NA	Undetectable
N-IgG at recurrence, S/CO	2.61	1.92	2.46	NA	Undetectable
Genomic sequencing	Same clade 20A	Same clade 20A	Same clade 20A	NA	Different clades, 20A and 20B

*The patient was readmitted to hospital.

HCQ, hydroxychloroquine; AZIT, azithromycin; LPV/r, lopinavir/ritonavir; S/CO, absorbance/cut-off; CKD, chronic kidney disease; NA, not available.

^&Antibodies were detected with Liaison® SARS-CoV-2 S1/2 IgG Diasorin (Saluggia, Italia); positive cut-off ≥15.0 AU/ml.

Our results show that late asymptomatic RT-PCR re-positivity does occur after COVID-19, even 6 months later, and does not necessarily represent new infection, despite the prolonged time interval elapsed and the negativity of subsequent RT-PCR tests since the first diagnosis. Although asymptomatic and symptomatic SARS-CoV-2 re-positivity had been reported, median time to recurrence was usually lower, around 1–2 months [6,7]. We found that the CDC criteria showed to satisfactorily predict reinfection, since none of patients not meeting the proposed criteria showed to be reinfected after genomic sequencing testing, while reinfection was actually confirmed in the suspected case according to criteria. Unfortunately, paired samples were not available for sequencing the viral genomes of one of the patients, who had a symptomatic re-positivity with a Ct value of 36. This patient would not have been classified as a case of suspected reinfection by CDC criteria. Interestingly, confirmed recurrences were accompanied by coexisting detectable antibody levels, as it also occurred with the symptomatic suspected recurrence, while antibodies were not present in the patient with reinfection. These findings reinforce the protective role of antibodies against reinfection. Peak antibody levels after the first SARS-CoV-2 infection in patients with recurrence did not differ from average values observed in the cohort, and S-IgG levels at the time of recurrence were within the range of peak levels. Whether this could have contributed to the absence of symptoms in 3 of the 4 patients is unknown. Immune dysfunction has been implicated among the factors potentially contributing to reactivation of latent persistent virus after COVID-19 [2]. Despite the adequate antibody levels, additional immune deficits, such an insufficient cellular immune response to SARS-CoV-2, might have had a role in the delayed RT-PCR re-positivity.

Our study provides long-term data about the natural history of COVID-19. Asymptomatic recurrences are detected up to six months after COVID-19. The CDC criteria are helpful to distinguish between disease recurrence and reinfection.

Funding sources

This work was supported by the RD16/0025/0038 project as a part of the Plan Nacional Research + Development + Innovation (R + D + I) and cofinanced by Instituto de Salud Carlos III - Subdirección General de Evaluación y Fondo Europeo de Desarrollo Regional; Instituto de Salud Carlos III (Fondo de Investigaciones Sanitarias [grant number PI16/01,740; PI18/01,861; CM 19/00,160; CM20/00,066; COV20–00,005]). The funding agencies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Declaration of Competing Interest

All authors declare no conflict of interest.