The novel coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused a pandemic, threatening global public health. SARS-CoV-2 is the seventh member of the Coronaviridae family known to infect humans and it is responsible for respiratory illness of varying severity (1). Among the clinical symptoms, the most common are fever and cough (2). The mean duration of the virus infection has been estimated to be about 14 days (3). However, the persistence of the virus for more than 30 days has been described (4). Previous studies have found correlation between infection duration and presence of pre-existing diseases (3) or with severe COVID-19 illness (5); conversely, individuals that can remain positive to virus also after symptom resolution were described (6–8). Thus, the factors that impact the persistence of the virus are still uncertain. This report describes the case of a diabetic man affected by Non-Hodgkin Lymphoma with persistent SARS-CoV-2 infection. Whole genome sequence of SARS-CoV-2 was performed on the first oro/nasopharyngeal swab and on a later one collected after more than 1-month, following isolation from VeroE6 cell cultures. The intra-host variability of the viral genome was also assessed.

RT-PCR performed on the two oro/nasopharyngeal swabs confirmed the presence of SARS-CoV-2: IZSPB_141 showed the amplification of E gene with a Ct value of 23.2 and RdRp gene with a ct value of 21.8; IZSPB_123 showed the amplification of E gene with a Ct value of 30.21 and RdRp gene with a Ct value of 26.4. The sample oro/nasopharyngeal swab collected on June 30th 2020 (IZSPB_123) was inoculated in VeroE6 cells, and then incubation was observed at 24 h intervals. The cytopathic effects were observed 72 h after inoculation. Virus replication was confirmed using real-time RT-PCR with RNA extracted from the cell culture medium. To assess the variational spectrum of the oro/nasopharyngeal swab specimen (IZSPB_141) and viral isolates by VeroE6 cell culture (IZSPB_123), whole genome sequencing of the SARS-CoV-2 samples was performed, generating 887.638 and 881.046 number of reads, respectively, of which more than 99% resulted as mapped reads. We obtain the mean coverage of 5.763x for IZSPB_141 sample and 5.807x for IZSP_123 (Supplementary Table 1). Both SARS-CoV-2 sequences belonged to Nextstrain clade 20A, GISAID clade G and Pangolin lineage B.1 (Supplementary Table 2). We detected a total of nine mutations, of which six were shared between the two reconstructed viral sequences (Figure 2). Among the six mutations, two did not result in any amino acid change, while the remaining mutations were non-synonymous variations. Interestingly, the sequence of SARS-CoV-2 isolated from VeroE6 cells harbored three additional mutations, namely C24117T – Spike (S) A852V; C17326T – ORF1b P1287S; C27710T - ORF7a A106V. In particular, mutation (S A852V) was detected in other 84 SARS-CoV-2 sequences available in GISAID database, as per July 28th 2021. Additionally, searching among available sequencing in GISAID database we detected four additional amino acid changes in the same protein position (i.e. A852G; A852S; A852T; A852E). All the nucleotide variants described were predicted by Nextclade tool, thus present in the consensus sequence of the viral genome analyzed and were detected with a frequency >61.9% by LoFreq. When we considered all the variants predicted with a lower frequency than 61.9%, we detected a total of 46 additional variants in IZSPB_141 and 32 additional variants in IZSPB_123 (Supplementary Table 3). Among these, only 8 variants were common in the two sampling; six variants were reported in Figure 2, while two variants were detected with very low frequency between 1.4 and 0.7%.

We reported the case of long persistent SARS-CoV-2 infection in a diabetic patient affected by NHL. The patient initially showed modest typical COVID19 symptoms (i.e., acute respiratory failure) and later became asymptomatic and finally his swab resulted negative after 3 months of infection. Although, the serological assay resulted negative, the clinical condition of the patient was not severe. This finding disagreed with previous studies that reported an increased severity and mortality SARS-CoV-2 infection in adult immunosuppressed patients compared with immunocompetent adults (16, 17). Two infusions of IVIg were administered 2 months after the start of infection. SARS-CoV-2 genome analysis showed that the virus isolated from the oro/nasopharyngeal swab specimen and viral cell line isolates, collected in different times, belong both to pangolin lineage B.1. The two genomes analyzed shared six mutations compared to the reference genome NC_045512.2; however the strain isolated from cell culture harbored three additional mutations. Among these, we noticed a non-synonymous mutation (S A852V) in the gene encoding the Spike protein (S gene), that mediates receptor binding and fusion of the viral and cellular membrane (18), thus determining the ability of the virus to infect the cells and its transmissibility in the host (19). To date, several mutations in the S gene were reported, but it is still unclear whether these variations could influence viral infectivity, transmissibility or immune response (20). The mutation described here (S A852V) was searched in all SARS-CoV-2 genomes available in GISAID database. We detected the same amino acid change in a total of 84 strains. Interestingly, this locus is a polymorphic site, since we found other four additional amino acid changes: hcov S A852G (13 strains); hcov S A852S (186 strains); hcov S A852T (13 strains); hcov S A852E (5 strains). Other two mutations in the same strain, isolated from cell culture were found in genomes already collected in GISAID: specifically, ORF1b P1287S was identified in 83 samples, while ORFa A106V was identified in 21 samples. Nevertheless, the co-segregation of the three variations was found for the first time in our sample. The investigation of the occurrence of these mutations among our two oro/nasopharyngeal samples was not straightforward, since the cultured strain (IZSPB_123) derived from a different oro/nasopharyngeal swab and was sampled about 2 months after the first one (IZSPB_141). Thus, we cannot assess whether these mutation events occurred before or after cell culture, but certainly following the first sequenced virus sample. Viral population diversity was influenced by several factors, e.g., replication fidelity (21, 22). The mutation frequency in SARS-CoV-2 populations is strongly influenced by negative and positive selection pressures (23). Virus replication rates directly impact the accumulation of mutations in the viral genome, enabling the existence of a viral quasispecies (a swarm of different, yet highly related, viral entities) within an infected host (24–26). Thereby, a single swab obtained from an individual or patient can contain different molecules of viral RNA, thus a different level of diversity. This scenario can exert a range of effects on viral fitness (27, 28). Based on this concept, we investigated the presence of mutations with a low frequency in our strains. The aim of this analysis was to determine whether the additional variants identified in sample IZSPB_123 were already present in the previously sequenced genome IZSPB_141 with a frequency <2%. Among the tally of identified mutations, only two were carried in both strains with an allele frequency <1.5%. However, we could not detect in IZSPB_141 the three variations identified in the IZSPB_123 strain. This result suggests two hypotheses: a) the three mutations identified in IZSPB_123 but not in IZSPB_141, may have arisen as a consequence of the in vitro growth of the virus; b) the mutations identified in IZSPB_123 may have arisen de novo after the first diagnosis of SARS-CoV-2 infection. According to the first hypothesis there is evidence to indicate the occurrence of new variations during virus passage in VeroE6 cells (29). The three mutations identified in IZSPB_123 could not be found in any viral genome from Apulia or Basilicata available in the GISAID database, whose sequencing was carried out in our Laboratory following the isolation in VeroE6 cells. Conversely, the latter hypothesis is supported by the finding of the same mutations in other viral samples (none isolated from cells), suggesting that these variations did not are due to amplification errors in sequencing. Thus, these mutations may be the expected result of the constant mutation of SARS-CoV-2, which might be more noticeable in cases of persistent infection. Further studies are required to understand the role of these mutations in COVID-19 ethology and spread.

In conclusion, our findings support the concept that intra-host genetic variations occurs in SARS-CoV-2, especially when the infection is persistent, suggesting that these mutations, in combination with the clinical phenotype of the host, can potentially influence the natural history of the viral infection.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.gisaid.org/, EPI_ISL_525555; EPI_ISL_525554.

Ethical approval was not provided for this study on human participants. After diagnostic routine, strains resulting from the biological material were stored to be processed for further analysis. No personal data or any other information than the type of material and the result of routine analysis were collected from each specimen, inhibiting any correlations of these fully anonymized samples with the respective patients. Thus, according to national regulations and the institutional rules for Good Scientific Practice, the requirement for submission to an ethical committee and for obtaining patients' informed consent was waived. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements. In addition, the document “Deontological rules for processing for statistical or scientific research purposes published according to Article 20, paragraph 4, of Legislative Decree Number 101 of 10 August 2018 - 19 December 2018”, in Article 2 mentions: “These deontological rules do not apply to processing for statistical and scientific purposes related to health protection activities carried out by health professions or health bodies, or with comparable activities in terms of significant personalized impact on the data subject, which shall continue to be governed by the relevant provisions”.

AP, AB, and LC provided substantial contributions to the conception and design of the work and the acquisition, analysis, and interpretation of data and drafting the work. AB, LC, LDe, DS, LP, VR, CM, LDi, AF, and AM performed the acquisition, analysis, and interpretation of data and drafting the work. AB and LC performed the acquisition, analysis, and interpretation of data and critical revising of the work for important intellectual content. All authors have approved the final version to be published and accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

This research was funded by Project COVID-19: epidemiological, clinical, genetic, and social determinants of infection and disease progression (Project Code: COVID-2020-12371675).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.