Prolonged severe acute respiratory syndrome coronavirus 2 persistence, attenuated immunologic response, and viral evolution in a solid organ transplant patient

Abstract

Unlike immunocompetent hosts, the duration of viral persistence after infection with severe acute respiratory syndrome coronavirus 2 can be prolonged in immunosuppressed patients. Here, we present a case of viral persistence for over 19 weeks in a patient with a history of solid organ transplant and explore the clinical, virologic, and immunologic course. Our patient still demonstrated viral persistence at 138 days with low polymerase chain reaction cycle threshold values and evidence of continuing viral sequence evolution indicative of ongoing virus replication. These findings have important implications for infection prevention and control recommendations in immunosuppressed patients. Immune response, including neutralizing antibody titers, T cell activity, and cytokine levels, peaked around days 44–72 after diagnosis. Anti-S trimer antibodies were low at all time points, and T cell response was attenuated by day 119. As immune response waned and viral load increased, increased genetic diversity emerged, suggesting a mechanism for the development of viral variants.

1. INTRODUCTION

History of solid organ transplant (SOT) is considered a risk factor for severe illness from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection¹; however, it remains uncertain if clinical outcomes differ significantly from immunocompetent patients.^{2 , 3} Furthermore, the impact of immunosuppression and associated comorbidities on the duration of viral shedding and potential transmissibility remains unclear. In an immunocompetent host, viral shedding detected by polymerase chain reaction (PCR) in nasopharyngeal specimens has been shown to persist up to 46 days from symptom onset,⁴ although the replication-competent virus has only been recovered up to 10 days following symptom onset in mild and moderate cases5, 6, 7, 8 and 20 days in severe cases.⁹

There is emerging evidence for prolonged viral persistence in immunosuppressed patients, including those with a history of SOT,10, 11, 12 hematologic malignancy,13, 14, 15 and autoimmune disease.¹⁶ We present the clinical, virologic, and immunologic course of a patient with a history of orthotopic heart transplantation (OHT) and severe coronavirus disease 2019 (COVID-19) with SARS-CoV-2 persistence for over 19 weeks.

2. CASE PRESENTATION

A 69-year-old woman underwent OHT in December 2019 due to a history of ischemic cardiomyopathy, type 2 diabetes, and chronic kidney disease. She was highly sensitized to HLA antigens and underwent desensitization prior to transplant and also received induction therapy with anti-thymocyte globulin, intravenous immunoglobulin (IVIG), and eculizumab. Her postoperative course was complicated by grade 1B acute cellular rejection in January 2020, which was treated with steroids. In February 2020, she was also diagnosed with deep vein thromboses and pulmonary emboli, treated with apixaban.

The patient was admitted to the hospital in April 2020 after 3 weeks of diarrhea and diagnosed with norovirus by multiplex stool PCR. She was without fever, pulmonary symptoms, or hypoxia; however, routine SARS-CoV-2 PCR screening was positive at hospital day 0 and hydroxychloroquine was administered on days 0–5. On hospital day 14, she developed hypoxia, initially requiring up to 6 L of supplemental oxygen. Computed tomography of the chest demonstrated bilateral ground-glass opacities with peripheral and basilar predominance, consistent with COVID-19 pneumonia. She received tocilizumab 400mg IV on day 16, a second course of hydroxychloroquine on days 16–21, high-dose methylprednisolone on days 20–32, and remdesivir on days 16–21 ( Figure 1). Although she initially improved, she had a mild but persistent oxygen requirement of 1–2 L nasal cannula; however, on hospital day 35 she decompensated and required 100% FiO₂ via non-rebreather face mask and was transferred to the ICU where she was treated for bacterial pneumonia on days 34–41.

FIGURE 1.

Clinical, virologic, and immunologic history. ¹SARS-CoV-2 T cell response assessed using ELISpot. ²Single nucleotide polymorphisms (SNPs). SARS-CoV-2, severe acute respiratory syndrome coronavirus 2 [Color figure can be viewed at wileyonlinelibrary.com]

Over the next several months, she remained hospitalized and required minimal supplemental oxygen via nasal cannula. She remained on an attenuated immunosuppressive course with prednisone, tacrolimus, mycophenolate mofetil, and intermittent IVIG. Her remaining hospital course was complicated by neutropenia requiring filgrastim, cytomegalovirus viremia requiring cytomegalovirus immune globulin, EBV viremia, and failure to thrive. She was eventually discharged to hospice on day 134, readmitted on day 138, and expired on day 140 from presumed bacterial sepsis.

3. LABORATORY INVESTIGATIONS

3.1. Viral detection in patient specimens

Fourteen respiratory tract nasopharyngeal swabs were prospectively collected during the patient’s hospitalization. Additional blood, saliva, and stool samples were collected from days 44, 54, 58, 72, and 119 as part of a longitudinal cohort study. Cycle threshold (Ct) values were assessed using the Roche Cobas SARS-CoV-2 PCR assay, targeting the RdRp and E gene primers or using our in-house assay with TaqPath 1-step RT-qPCR (Thermo Fisher Scientific) and 2019-nCoV CDC EUA Kiet (IDT), utilizing N1, N2, and RP primers. Viral load peaked in nasal swab samples from days 44 and 54 with the Ct values of 11.7 and 11.5, which are translated to 2.2 and 2.4 × 10¹⁰ copies/ml, and remained detectable until her final specimen was obtained on day 138, with a Ct value of 17 (Figure 1). Plasma samples from day 44, 54, and 58 with Ct values of 29.2 (9.6 × 10⁴ copies/ml), 31.8 (1.6 × 10⁴ copies/ml), and 35.2 (2.0 × 10³ copies/ml), rectal swab from day 72 with Ct value of 22.5 (1.5 × 10⁷ copies/ml), and stool from day 119 with Ct value of 28 were also positive for SARS-CoV-2. Viable virus was isolated from a frozen saliva sample from hospital day 54 with a TCID₅₀ of 1176/ml.

3.2. Immunologic studies

Samples obtained from the patient were also analyzed for plasma antibody and T cell response. In accordance with methodology previously published,¹⁷ immunoassays to quantify antibodies against SARS-CoV-2 S trimer were used to measure binding antibody titers, and antibody neutralization assays against SARS-CoV-2 were performed.¹⁷ EC₅₀ for IgM, IgG, and IgA responses against SARS-CoV-2 S-trimer were low at all time points. Neutralizing antibody IC₅₀ was too low to be calculated on day 44, low on days 54 and 58, moderate on day 72, then low again on day 119 (Figure S1).

T cell response was assessed using ELISpot methodology (see Supporting Information for details) and was strong on day 44, weak on day 72, and very weak on day 119 (Figure S2). Reactivity was only demonstrated against the SARS-CoV-2 spike protein but was also noted against the Staphylococcal Enterotoxin B positive control with a similar trend (121 spots on day 44, 17 spots on day 72, and 9 spots on day 119), indicative of a general waning of the cell-mediated response and immune exhaustion. This correlates to clinical lymphocyte panel testing with low absolute CD4, CD8, and B cell counts throughout the hospitalization and with CD8 count highest at day 41 (Table S1). Cytokine levels were analyzed using Milliplex human cytokine/chemokine magnetic bead panels (Millipore Sigma) and the Luminex 200 platform (Luminex) on unstimulated plasma cells. Interleukin 6 (IL-6), interleukin 28, and interferon gamma-induced protein 10 peaked on day 44 and remained low afterward. Interleukin 10, monocyte chemoattractant protein 1, macrophage inflammatory protein 1 beta, and tumor necrosis factor-alpha peaked on day 72. Interleukin 17A and interleukin 1 beta were below the limit of detection at all time points (Table S2).

3.3. Sequencing

Samples from days 44, 54, 58, 72, and 119 were sequenced on an Illumina MiSeq using the ARTIC V3 tiling PCR amplification protocol followed by the Nextera Flex library preparation kit and the 300-cycle MiSeq v2 reagent kit. Reads were trimmed using Trimmomatic and mapped to the Wuhan-Hu-1 reference genome (MN908947.3) using Bowtie2 to identify single nucleotide polymorphisms (SNPs). Consensus genomes were generated using a threshold of 65% alignment for a given base call and minimum coverage of 10X; consensus sequences were then used to generate a maximum-likelihood phylogenetic tree with RAxML after whole-genome alignment with MAFFT ( Figure 2). Isolates from all time points were classified as Pangolin lineage B.1, prevalent in the United States at the time of this study. Nasal swab and saliva samples from days 44, 54, and 58 were nearly identical, differing by only 0–2 SNPs, and had a total of 7–8 mutations compared to the reference strain. By day 119, however, multiple SNPs had accumulated across the genome (n = 11 compared to the reference strain). Notably, changes in the viral genome between days 44–119 included not only the confirmed SNPs above, but also 15 sites where ambiguous base calls indicated the presence of multiple alleles within the sample. This points to the likelihood of active mutation of the virus. Given the consistent classification of isolates within the B.1 lineage and the apparent stepwise evolution between early and late nasopharyngeal swab samples, there was no evidence to suggest reinfection with a different strain. Notably, the plasma sample from day 54 and the rectal swab from day 72 also exhibited an accumulation of SNPs compared to nasal swabs and/or saliva collected on the same day (5–6 SNPs between day 54 plasma and concurrent nasal swab and saliva; 25 SNPs between day 72 rectal swab and saliva), indicative of viral compartmentalization. Furthermore, a D138Y mutation in the S gene, one of the mutations in the Brazilian P1 variant, was detected in the virus from plasma but not from other specimens.

FIGURE 2.

Consensus genomes and maximum-likelihood phylogenetic tree [Color figure can be viewed at wileyonlinelibrary.com]

4. DISCUSSION

While most immunocompetent hosts clear SARS-CoV-2 within 2 weeks, there is heterogeneity in the duration of viral persistence in immunosuppressed patients.11, 12, 13, 14, 15, 16 Here, we present a case of viral persistence for over 19 weeks in a patient with a history of SOT and explore the clinical, virologic, and immunologic course.

Our patient demonstrated viral persistence at 138 days with a Ct value of 17 and evidence of continuing viral evolution indicative of ongoing virus replication. Prolonged RNA positivity was not limited to the respiratory tract, as virus was detected in plasma up to day 58 and stool on day 119. Further studies were precluded as the patient died of SARS-CoV-2 unrelated causes.

These findings have important implications for infection prevention and control recommendations in immunosuppressed patients. When assessing viral persistence in this population, quantitative PCR testing may be more useful than qualitative testing in determining transmission risk, isolation recommendations, and even the role of therapeutic interventions during the late convalescent period of infection.

Immune response, including neutralizing antibody titers, T cell activity, and cytokine levels, peaked around days 44–72, temporally coinciding with a decrease in viral load after day 54. In an immunocompetent host, sufficient IgG response would be expected 2–3 weeks after infection; however, SOT patients have been shown to have a decreased humoral immune response.¹⁸ Concordantly, this patient had insufficient humoral and cell-mediated immune responses, as demonstrated by low anti-S trimer IgM, IgG, and IgA responses at all time points and lack of T cell response by day 119. While binding antibody titers remained low, neutralizing antibody response was relatively strong at day 72, despite viral persistence. On day 119, neutralizing antibody titers later decreased, coinciding with another increase in viral load on days 121 and 138. Cytokine levels were not significantly elevated during prolonged viral shedding, indicating a lack of systemic inflammation. The elevation of IL-10, MCP1, MIP1-beta, and TNF-alpha on day 72 correlates with the initiation of filgrastim on day 64, which has been shown to cause an elevation in several pro-inflammatory cytokines.¹⁹ Overall, this pattern of immunosuppression may be explained early on by the patient’s comorbidities, the use of induction therapy with eculizumab at the time of transplant, as well as treatment for acute cellular rejection (steroids, anti-thymocyte globulin). Furthermore, the patient remained on a modified immunosuppression regimen and received COVID-19 treatment with steroids and IL-blockade, which may have further attenuated her immune response. Although unavailable at the time, this patient would now be eligible for SARS-CoV-2 monoclonal antibody therapy at the time of initial COVID-19 diagnosis, which theoretically could facilitate viral clearance, especially in an immunosuppressed patient with an attenuated humoral immune response.

Taken together, our findings suggest that during early convalescence the immune response peaked, leading to a decreased viral load; however, a subsequent waning of the immune response resulted in viral rebound and the emergence of genetic diversity. This observation of ongoing mutation of the SARS-CoV-2 virus within an immunosuppressed host over a prolonged period of time suggests a mechanism for the development of SARS-CoV-2 variants. The D138Y mutation identified in plasma is one of the targets within the NTD region for anti-SARS-CoV-2 neutralizing antibodies,²⁰ possibly contributing to viral escape. Furthermore, this mutation was identified several months before the emergence of the P1 variant, providing evidence that patterns of mutation acquisition may exist. Lastly, we report discordance between binding antibody EC50 and neutralizing IC50 during late convalescence, while previous reports have shown a direct correlation,¹⁷ although this result may be impacted by the infusion of IVIG the day prior to blood sample collection.

Overall, this case highlights the heterogeneity of SARS-CoV-2 viral persistence and immunologic response in immunosuppressed patients, which can result in viral evolution. Further research and prospective studies are needed to study viral persistence, transmissibility, and immunologic response in this population.

Funding information National Institute of Allergy and Infectious Diseases of the NIH, Grant/Award Number: T32AI114398

SUPPORTING INFORMATION

Additional supporting information may be found in the online version of the article at the publisher’s website.

Supplementary Material