Abstract
Background
Early stratification of disease progression remains one of the major challenges towards the post-coronavirus disease 2019 (COVID-19) era. The clinical relevance of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) nucleic acid load is debated due to the heterogeneity in patients’ underlying health conditions. We determined the prognostic value of nasopharyngeal viral load dynamic conversion for COVID-19.
Methods
The cycling threshold (Ct) values of 28,937 nasopharyngeal SARS-CoV-2 RT-PCRs were retrospectively collected from 3,364 COVID-19 patients during hospitalization and coordinated to the onset of disease progression. The ROC curve was utilized to determine the predictive performance of the rate of Ct value alteration between two consecutive RT-PCR runs within 48 h (ΔCt%) for disease transformation across patients with different COVID-19 severity and immune backgrounds, and further validated with 1,860 SARS-CoV-2 RT-PCR results from an independent validation cohort of 262 patients. For the 67 patients with severe COVID-19, Kaplan-Meier analysis was performed to evaluate the difference in survival between patients stratified by the magnitude of Ct value alteration between the late and early stages of hospitalization.
Results
The kinetics of viral nucleic acid conversion diversified across COVID-19 patients with different clinical characteristics and disease severities. The ΔCt% is a clinical characteristic- and host immune status-independent indicator for COVID-19 progression prediction (AUC = 0.79, 95 % CI = 0.76 to 0.81), which outperformed the canonical blood test markers, including c-reactive protein (AUC = 0.57, 95 % CI = 0.53 to 0.61), serum amyloid A (AUC = 0.61, 95 % CI = 0.54 to 0.68), lactate dehydrogenase (AUC = 0.61, 95 % CI = 0.56 to 0.67), d-dimer (AUC = 0.56, 95 % CI = 0.46 to 0.66), and lymphocyte count (AUC = 0.62, 95 % CI = 0.58 to 0.66). Patients with persistent high SARS-CoV-2 viral load (an increase of mean Ct value < 50 %) during the first 3 days of hospitalization demonstrated a significantly unfavorable survival (HR = 0.16, 95 % CI = 0.04 to 0.65, P = 2.41 × 10–3).
Conclusions
Viral nucleic acid dynamics of SARS-CoV-2 eliminates the inter-patient variance of basic health conditions and therefore, can serve as a prognostic marker for COVID-19.
Keywords: Coronavirus disease 2019 (COVID-19), Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), Viral nucleic acid load, Disease severity
Abbreviations: SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; COVID-19, coronavirus disease 2019; RT-PCR, reverse transcriptional polymerase chain reaction; NAAT, nucleic acid amplification testing; WHO, World Health Organization; ΔCt%, the rate of Ct value alteration between two consecutive RT-PCR runs within 48h; HSH, Huashan Hospital, Fudan University; SSPH, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine; CRI, chronic renal insufficiency; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease
Introduction
Shifting virulence of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and improving herd immunity barrier is shedding light on the post-coronavirus disease 2019 (COVID-19) era. At this critical stage, early stratification of patients with high or low disease progression risk will significantly improve the efficiency of medical resource utilization and provide alerting evidence for appropriate clinical intervention [1], [2]. Therefore, laboratory markers with prognostic value are advocated for disease risk management throughout the clinical practice, though previous investigations have demonstrated that the performance of canonical “host” biomarkers including infection-induced proteins and fibrinolysis products is insufficient for the prognosis of COVID-19 [3], [4], [5], [6], [7].
Viral nucleic acid amplification testing (NAAT) such as real-time polymerase chain reaction (PCR) with respiratory specimens detects the genetic material of targeted virus with desired sensitivity and specificity, therefore became a dominant tool for the infection screening [8]. On the other side, association between SARS-CoV-2 viral load and COVID-19 severity and mortality has been reported [9], [10], [11], but whether the viral nucleic acid quantification is appropriate for identification of the high risk group during clinical practice remains highly debated due to the greatly variable test results among different sampling methods and anatomical location [12], [13]. In this cross-sectional study, we retrospectively reviewed the nasopharyngeal SARS-CoV-2 nucleic acid load transformation throughout the full in-hospital course of 1,852 COVID-19 patients and 1,514 individuals with asymptomatic SARS-CoV-2 infection during the omicron BA2.2 outbreak in Shanghai, China, in early 2022, in order to elucidate the clinical relevance of viral nucleic acid kinetics for COVID-19 prognosis. The established threshold of ΔCt% (the rate of Ct value alteration between two consecutive RT-PCR runs within 48 h) for disease progression prediction was further validated in an independent cohort of 262 COVID-19 patients from another hospital. Our results demonstrated that monitoring the viral nucleic acid load by RT-PCR with high dynamics and sensitivity is a host immune status-independent strategy for the prediction of disease progression, and distinguishes the COVID-19 patients with poor prognosis.
Materials and methods
2.1. SARS-CoV-2 infected patients cohort and nasopharyngeal sampling
A total of 1,852 patients with COVID-19 and 1,514 patients with asymptomatic SARS-CoV-2 infection were enrolled in Huashan Hospital, Fudan University (HSH) during March 25th 2022 to May 15th 2022, the BA2.2 omicron epidemic in Shanghai, China. Moreover, an independent validation cohort of 163 COVID-19 patients and 99 patients with asymptomatic infection were enrolled from Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (SSPH) during May 6th 2022 to May 24th 2022. According to the margin of error formula, the sample size of the cohorts was able to provide enough statistical power with α of 0.05 and acceptable margin of error of 10 %. All these patients were diagnosed and classified with disease severity according to the World Health Organization (WHO) living guidance for clinical management of COVID-19 (23 November 2021) [14]. The patients were hospitalized within 3 days after the first NAAT “positive” confirmation during proactive community screening for SARS-CoV-2 infection. Individuals with severe COVID-19 were assigned to the intensive care unit (ICU). Patients with asymptomatic, mild, or moderate disease were assigned to specialized wards to receive the optimized medical care for their primary disease and / or COVID-19. The nasopharyngeal samples were collected within every 48 h by dedicated sampling nurses. To avoid sampling bias, the patients’ clinical information including disease severity and comorbidities was concealed from the sampling nurses, who received pre-job training of standardized nasopharynx swab sampling procedure including: 1) stabilize the patient’s head with one hand and insert the swab with the other hand from the patient’s anterior nostril. 2) insert the swab with rotation along the direction of nose bottom and into the nasal cavity until there is a sense of resistance. The depth of the swab insertion is approximately 10 cm, as indicated by the mark on the swab. 3) keep the swab inside for approximately 15 s, then rotate it clockwise twice and counterclockwise twice, each time for approximately 5 s. 4) smoothly withdraw the swab along the nose tip and open the sample preservation tube (guanidine salts containing) vertically. Put the swab into the tube immediately, break the tail of swab, and tighten the tube’s cover.
The nucleic acid samples were transported to the PCR laboratory in 4 °C cold chain and tested within 6 h after sampling. Written informed consent was received from all participants. In compliance with the Helsinki Declaration of 1975 as revised in 2013, this study was approved by both the Institutional Review Boards of Huashan Hospital of Fudan University and Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine.
2.1. Performance specification and quality control of RT-PCR system for SARS-CoV-2 nucleic acid quantification
The RNA standard was purified from clinical samples with high SARS-CoV-2 load by QIAamp Viral RNA kti (Qiagen). The vacuum-concentrated RNA sample was quantified and adjusted to a concentration of 2 × 1011 copies/ml by QX200 droplet digital PCR platform (Bio-Rad Labs) using the one-step ddPCR supermix with primers and probe for N gene from China Centers for Disease Control and Prevention [15]. Tenfold serial dilutions of the RNA sample were performed to prepare standards with concentration ranging from 20 copies/ml to 2 × 1011 copies/ml.
Quantitative measurement performance including linearity, limit of quantification (LoQ), and inter- and intra-batch precision was evaluated on the two distinct SARS-CoV-2 RNA RT-PCR kits (Shuoshi Biotechnology, used in the HSH laboratory; Liferiver Biotechnology, used in the SSPH laboratory), which detects the ORF1ab and N gene of SARS-CoV-2 by dual fluorescent probes. In short, linearity of the RT-PCR system was validated in accordance with the Clinical and Laboratory Standards Institute (CLSI) EP6-A guideline, with triplication of the standard at each concentration. The LoQ of the system was evaluated according to the EP17-A2 guideline using 20 replicates of standards with concentrations of 200 copies/ml and 20 copies/ml. The inter- and intra-batch precision of the system was assessed following the CLSI EP05-A3 guideline, with 20 replicates of high- (2 × 107 copies/ml) or low- (2 × 103 copies/ml) copy standards. The performance evaluation assays and all clinical sample assays were performed with the identical experimental environment and reagent lot number following the manufacturer’s instruction to ensure the comparability of all RT-PCR results.
For each SARS-CoV-2 RT-PCR run, nucleic acid extraction and PCR mix preparation were performed by automated nucleic acid robot stations to minimize the manual pipetting errors. Two artificial plasmid-derived low-copy quality controls (2 × 103 copies/ml and 103 copies/ml, Shanghai Center for Clinical Laboratory) were measured following the full procedure of nucleic acid extraction, amplification and data collection together with the clinical samples. The Ct value of ORF1ab and N gene were subjected to the Westgard 12s/22S/10X multi-rules for quality control. Moreover, 2-side nasopharyngeal sampling and subsequently PCR test were conducted for 10 randomly selected patients everyday. Paired t-test of the Ct values of were calculated to ensure the repeatability of sampling.
2.1. Data collection and cleaning
To assess the relative alteration of SARS-CoV-2 viral load from the continuous monitoring during hospitalization, we obtained the cycling threshold (Ct) values of each RT-PCR for nasopharyngeal SARS-CoV-2 detection for both the HSH and SSPH cohorts.
To evaluate the patients’ underlying immune competence in the HSH cohort, we obtained the serum cytokine quantification, white blood cell (WBC) counting, and T/B/NK cell sorting results at the first day of hospitalization from 15 asymptomatic, 71 mild, 82 moderate, and 20 severe COVID-19 patients. All the data were normalized with Z-score for further analysis. Serum markers for COVID-19 severity classification including c-reactive protein (CRP, Genius Medical), serum amyloid A (SAA, Genius Medical), lactate dehydrogenase (LDH, Roche Diagnostics), and D-dimer (DDI, Sysmex) were obtained from each biochemistry and coagulation test. The arterial-venous plasma concentration of pH, pCO2, pO2, HCO3 – was obtained from the blood gas tests.
The medical records were reviewed on all patients to collect clinical parameters including gender, age, SARS-CoV-2 vaccination, underlying disease, together with timeline for symptoms conversion throughout the hospitalization and date of death for the deceased patients. Symptoms were classified into 5 categories with increasing severity according to the WHO COVID-19 living guidance [14] and Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia (National Health Commission of the PRC, Trial ver 9) [16]: 1) asymptomatic or pre-symptomatic infection; 2) fever, fatigue, headache, cough, sore throat, congestion, runny nose, loss of taste or smell; 3) shortness of breath, difficulty breathing or abnormal chest imaging; 4) SpO2 < 93 % on room air, (PaO2/FiO2) < 300 mmHg, respiratory rate > 30 breaths/min or lung infiltrates > 50 %; 5) acute respiratory distress syndrome (ARDS), respiratory failure, shock, or multiple organ dysfunction. Date of disease progression was tagged on the timeline of disease progression for each patient if one developed any higher-categorized symptom(s) than the current condition.
2.1. Immuno-subtyping of COVID-19 patients
Serum cytokines including interleukin (IL)-1, IL-2, IL-6, IL-8, IL-10, and tumor necrosis factor (TNF)-α were quantified by Immulite 1000 immunoassay analyzer (Siemens Healthcare). Absolute cell counts and relative percentage of CD3+ T cells, CD8+ cytotoxic T cells, CD4+ helper T cells, CD16+CD56+ NK cells, and CD19+ B cells of each patient were determined by BriCyte E6 flow cytometer (Mindray Medical) with Multitest 6-color TBNK kit (BD Biosciences). Counts of WBCs, neutrophils, lymphocytes, monocytes, eosinophils, and basophils (percentage) were extracted from the complete blood count assay by the XN-3000 blood analyzer (Sysmex). All these representative immune-related markers were hierarchically clustered by Pearson’s correlation distance, leading to two unsupervised immune-subtypes in the 188 COVID-19 patients (R package ConsensusClusterPlus version 1.56.0). The distinction of these two immune-subtypes was further illustrated by dimension reduction with t-distributed stochastic neighbor embedding (t-SNE) model (R package Rtsne version 0.15).
2.1. Study design and statistical analysis
A flow diagram of this study is illustrated in Fig. 1 A. Comparisons of continuous values were performed using 2-sample t-test or Wilcoxon rank sum test, if appropriate. The effect of each characteristic on time of viral nucleic acid conversion was estimated by Cox proportional hazards regression. The rate of Ct value change (ΔCt%) was calculated as difference between Ct values of two consecutive nasopharyngeal viral RT-PCRs divided by the Ct value of the earlier one, (Ctn-Ctn-1)/Ctn-1.
Fig. 1.
Flow diagram for study design and time interval for nucleic acid negative conversion in COVID-19 patients with different clinical characteristics. A) Flowchart illustrating design strategies and COVID-19 patient cohorts in this study. B) SARS-CoV-2 RT-PCR Ct value (ORF1ab) at the first day of hospitalization distinguished asymptomatic infected individuals from symptomatic COVID-19 patients but did not discriminate COVID-19 cases with different severity. C) Boxplot demonstrating the time interval of nucleic acid negative conversion across different clinical characteristics. D) Univariate and multivariate Cox regression models demonstrated advanced age and comorbidities as independent risk factors for prolonged nucleic acid negative conversion, and SARS-CoV-2 vaccination accelerated viral elimination. E) Time interval of SARS-CoV-2 nucleic acid negative conversation in COVID-19 patients with different disease severity. two-sample t-test, n.s. > 0.05; ** P < 0.01; **** P < 0.0001. Abbreviations: CRI: chronic renal insufficiency; CAD: coronary artery disease; COPD: chronic obstructive pulmonary disease.
The extreme values of serum CRP, SAA, LDH, IL-6, DDI, and peripheral blood lymphocyte counting during hospital admission and disease progression were retained to evaluate their performance for COVID-19 progression prediction. Minimal Euclidean distance fitting to the (0, 1) point of ROC curve was employed to determine the optimal ΔCt% cut-off value for disease progression prediction. Sensitivity and specificity were reported with exact (Clopper-Pearson) 95 % CI. Areas under the ROC curves were compared by DeLong’s test [17]. Overall survival rates were estimated using Kaplan-Meier analysis to assess differences in survival between patient groups stratified by the magnitude of Ct value alteration between late and early stages of hospitalization, or the serum or lymphocyte markers for COVID-19. All the statistical analysis and data visualization was performed in R v4.0.2. A two-sided P < 0.05 was considered statistically significant.
Results
2.1. Patient characteristics
The 3,364 SARS-CoV-2 infected individuals from HSH were 49.41 % female (n = 1,662), had a median age of 56 years (interquartile range = 40 to 69). Besides the 1,481 (44.02 %) asymptomatic individuals, 39.41 % (n = 1,326), 14.24 % (n = 479), and 2.32 % (n = 78) exhibited mild, moderate, or severe COVID-19, respectively. Among all these individuals, 39.45 % (n = 1,327) were with primary diseases and 53.83 % (n = 1,811) were immunized with SARS-CoV-2 vaccine. The 262 individuals in the SSPH cohort were 48.85 % female (n = 128), with a median age of 72 years (interquartile range = 61 to 81.75). The cohort consisted of 105 (40.08 %) asymptomatic individuals, and 117 (44.66 %) mild, 28 (10.69 %) moderate, 12 (4.58 %) severe cases, with a primary disease prevalence of 24.81 % (n = 65) and vaccination coverage of 81.68 % (n = 197) (Table 1 ). The top 10 common comorbidities in the HSH cohort were hypertension (20.24 %, n = 681), diabetes (9.93 %, n = 334), chronic renal insufficiency (CRI) (Stage IV) (5.98 %, n = 201), cardiac insufficiency (4.93 %, n = 166), coronary artery disease (CAD) (4.22 %, n = 142), tumor (4.19 %, n = 141), cerebral infarction (3.24 %, n = 109), chronic bronchitis (1.04 %, n = 35), mental illness (1.01 %, n = 34), chronic obstructive pulmonary disease (COPD) (0.86 %, n = 29) (Table 2 ).
Table 1.
Clinical Characteristics of Cohort Individuals with SARS-CoV-2 Infection. *IQR: interquartile range.
| N = | Gender |
Age (Years) |
Primary Disease | Vaccination |
|||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Male | Female | Median | IQR* | No | Vaccine | Vaccine + Booster | Vaccine + 2Boosters | Data Not Available | |||
| Huashan Hospital (HSH) Cohort | |||||||||||
| Asymptomatic | 1481 | 784(52.94 %) | 697(47.06 %) | 48 | 34–61 | 279(18.84 %) | 431(20.10 %) | 34(2.30 %) | 513(34.64 %) | 496(33.49 %) | 7(0.47 %) |
| Mild | 1326 | 612(46.15 %) | 714(53.84 %) | 59 | 41–70 | 616(45.46 %) | 612(46.15 %) | 39(2.94 %) | 353(26.62 %) | 319(24.06 %) | 3(0.23 %) |
| Moderate | 479 | 252(52.60 %) | 227(47.39 %) | 71 | 62–82 | 355(74.11 %) | 344(71.81 %) | 13(2.71 %) | 67(13.99 %) | 54(11.27 %) | 1(0.21 %) |
| Severe | 78 | 54(69.23 %) | 24(30.77 %) | 80 | 69–89 | 77(98.72 %) | 67(85.90 %) | 0 | 4(5.13 %) | 5(6.41 %) | 2(0.26 %) |
| HSH Total | 3364 | 1702(50.60 %) | 1662(49.41 %) | 56 | 40–69 | 1327(39.45 %) | 1454(43.22 %) | 86(2.56 %) | 937(27.85 %) | 874(25.98 %) | 13(0.39 %) |
| Shanghai Sixth People’s Hospital (SSPH) cohort | |||||||||||
| Asymptomatic | 105 | 55(52.38 %) | 50(47.62 %) | 62 | 45–69 | 22(20.95 %) | 7(6.67 %) | 46(43.81 %) | 39(37.14 %) | 13(12.3 %) | 0 |
| Mild | 117 | 58(49.57 %) | 59(50.43 %) | 75 | 70–87 | 24(20.51 %) | 22(18.80 %) | 77(65.81 %) | 14(11.97 %) | 4(3.42 %) | 0 |
| Moderate | 28 | 16(57.14 %) | 12(42.86 %) | 75.5 | 65–81 | 13(46.43 %) | 12(42.86 %) | 13(46.43 %) | 2(7.14 %) | 1(3.57 %) | 0 |
| Severe | 12 | 5(41.67 %) | 7(58.33 %) | 83 | 74–88.25 | 6(50.00 %) | 7(58.33 %) | 5(41.67 %) | 0(0.00 %) | 0(0.00 %) | 0 |
| SSPH Total | 262 | 134(51.15 %) | 128(48.85 %) | 72 | 61–81.75 | 65(24.81 %) | 48(18.32 %) | 141(53.82 %) | 55(20.99 %) | 18(6.87 %) | 0 |
Table 2.
Common Comorbidities of SARS-CoV-2 Infection in the HSH cohort.
| Top 10 Common Comorbidities | Overall N = | Asymptomatic | Mild | Moderate | Severe |
|---|---|---|---|---|---|
| Huashan Hospital (HSH) Cohort | |||||
| Hypertension | 681 | 140(20.56 %) | 307(45.08 %) | 192(28.19 %) | 42(6.17 %) |
| Diabetes | 334 | 68(20.36 %) | 128(38.32 %) | 112(33.53 %) | 26(7.78 %) |
| Chronic renal insufficiency (Stage IV) | 201 | 32(15.92 %) | 73(36.32 %) | 83(41.29 %) | 13(6.67 %) |
| Cardiac insufficiency | 166 | 27(16.27 %) | 52(31.33 %) | 59(35.54 %) | 28(16.87 %) |
| Coronary artery disease (CAD) | 142 | 22(15.49 %) | 69(48.59 %) | 42(29.58 %) | 9(6.34 %) |
| Tumor | 141 | 19(13.48 %) | 71(50.35 %) | 44(31.21 %) | 7(4.96 %) |
| Cerebral infarction | 109 | 16(14.68 %) | 36(33.03 %) | 40(36.70 %) | 17(15.60 %) |
| Chronic bronchitis | 35 | 4(11.43 %) | 14(40 %) | 13(37.14 %) | 4(11.43 %) |
| Mental illness | 34 | 5(14.71 %) | 7(20.59 %) | 10(29.41 %) | 12(35.29 %) |
| Chronic obstructive pulmonary disease (COPD) | 29 | 4(13.79 %) | 11(37.93 % | 7(24.14 %) | 7(24.14 %) |
| Shanghai Sixth People’s Hospital (SSPH) cohort | |||||
| Diabetes | 9 | 5(4.27 %) | 4(3.81 %) | 0(0.00 %) | 0(0.00 %) |
| Hypertension | 8 | 4(3.42 %) | 2(1.90 %) | 2(7.14 %) | 0(0.00 %) |
| Cardiac insufficiency | 8 | 2(1.71 %) | 2(1.90 %) | 4(14.29 %) | 0(0.00 %) |
| Chronic renal insufficiency (Stage IV) | 6 | 3(2.56 %) | 1(0.95 %) | 2(7.14 %) | 0(0.00 %) |
| CAD | 5 | 0(0.00 %) | 4(3.81 %) | 0(0.00 %) | 1(8.33 %) |
| Chronic liver disease | 5 | 3(2.56 %) | 1(0.95 %) | 0(0.00 %) | 1(8.33 %) |
| COPD | 5 | 1(0.85 %) | 1(0.95 %) | 2(7.14 %) | 1(8.33 %) |
| Cerebral infarction | 5 | 0(0.00 %) | 3(2.86 %) | 1(3.57 %) | 1(8.33 %) |
| Tumor | 5 | 2(1.71 %) | 2(1.90 %) | 1(3.57 %) | 0(0.00 %) |
| Thyroid disorders | 3 | 2(1.71 %) | 1(0.95 %) | 0(0.00 %) | 0(0.00 %) |
2.1. Analytical performance of SARS-CoV-2 RT-PCR tests
By a serial dilution of RNA standards, we confirmed a Ct linearity of the Shuoshi SARS-CoV-2 RNA RT-PCR kit ranging from 8.93 to 41.45 (ORF1ab) and from 7.49 to 41.61 (N); LoQ of the system was at the concentration of 200 copies/ml (CtORF1ab = 41.45 and CtN = 41.61). The Ct linearity of the Liferiver kit was from 8.78 to 41.14 (ORF1ab) and from 7.89 to 40.07 (N), with the LoQ of 200 copies/ml (CtORF1ab = 41.14 and CtN = 40.07). Given the cut-off Ct value for SARS-COV-2 positive detection of 40 (Shuoshi) and 43 (Liferiver) according to the manufacturers’ instruction, we validated that the RT-PCR kits provides comparable results within the Ct value range 8.93 to 40 (ORF1ab) and 7.49 to 40 (N) for the Shuoshi kit, and 8.78 to 41.14 (ORF1ab) and 7.89 to 40.07 (N) for the Liferiver kit. Both the intra- and inter-assay precision of the RT-PCR system were within the coefficient of variation (CV) of 20 % (Ct value-converted copy number). For all the 28,937 SARS-CoV-2 RT-PCR results using the Shuoshi kit and 1,860 results using the Liferiver kit, we observed a significant correlation between the Ct value of ORF1ab and N genes (Shuoshi kit, n = 28,937, Spearman Rho = 0.98, P < 10-5; Zhijiang Kit, n = 1860, Spearman Rho = 0.99, P < 10−5).
2.1. Time interval for SARS-CoV-2 RT-PCR negative conversion
Ct values on the first day of hospitalization distinguished the asymptomatic SARS-CoV-2 infected individuals from the COVID-19 patients, but were insufficient to discriminate the COVID-19 patients with different severity (Fig. 1B and Supplementary Fig. 1A), which is consistent with previous reports [18], [19]. Therefore, we assessed the dynamic of viral nucleic acid elimination across different clinical characteristics of the patient.
According to the validated LoQ performance of RT-PCR, we defined the 2 consecutive SARS-CoV-2 RT-PCR (from nasopharyngeal swab) performed over 24 h apart with Ct values of ORF1ab and N genes greater than or equal to 40 as the criteria for nucleic acid negative conversion.
The mean time for viral RT-PCR negative conversion of the overall population was 10.07 days (95 % CI = 9.77 to 10.36). No significant difference was observed between the two gender groups. Elderly patients and underlying disease demonstrated a prolonged period of virus elimination. In contrast, the SARS-CoV-2 vaccinated group exhibited a significant protective effect to reduce the duration of viral presentation (Fig. 1D). By further detailed survey of the association between nucleic acid negative conversion and patients’ comorbidity, we identified hypertension, diabetes, CRI stage IV, cardiac insufficiency, tumor, and cerebral infarction as risk factors for a prolonged viral elimination period (Supplementary Fig. 1B-D).
The time interval of viral negative conversion increased with disease severity [asymptomatic (mean = 8.89 days) vs mild (mean = 11.07 days, P = 2.66 × 10-12), mild vs moderate (mean = 12.44 days, P = 4.09 × 10-2)]. However, there was no significant difference between the moderate and severe groups (mean = 12.69 days, P = 0.93) (Fig. 1D).
2.1. Dynamic alteration of SARS-CoV-2 viral load predicts the progression of COVID-19
During the period of nucleic acid conversion, a sustained reduction in Ct values (ORF1ab) was detected within 5 days prior to disease progression. It was also noted that a significant alteration in Ct values appeared on the day of progression (P = 6.94 × 10-3) (Fig. 2 A). We also observed that over 95 % of the patients demonstrated a Ct value decrease within three days before the commencement of disease progression (2.96 days) (Fig. 2B). Therefore, we assessed the performance of Ct value alteration for the progression of disease within three days in the 2,622 patients with full medical record of COVID-19 disease transformation from the HSH cohort.
Fig. 2.
Continuous SARS-CoV-2 nucleic acid quantification predicts COVID-19 progression. A) Temporal dynamics in Ct value coordinated by the commencement of disease progression in COVID-19 patients. B) Time based distribution of patients who demonstrated decreased Ct value before disease progression. C-D) Performance of ΔCt% monitoring for COVID-19 progression prediction within three days after Ct value alteration in the HSH (C) and SSPH (D) cohort. D-E) ROC curve of ΔCt% monitoring for COVID-19 progression across distinct disease severity in the HSH (D) and SSPH (E) cohort. Two-sample t-test, * P < 0.05; ** P < 0.01; *** P < 0.001. Abbreviations: ΔCt%: the rate of Ct value change between two consecutive SARS-CoV-2 RT-PCR; Sen.: sensitivity; Spe.: Specificity; AUC: area under the curve.
For 1,689 patients with a pre-progression stable period of disease in the HSH cohort [20], the optimal threshold of ROC curve suggested a 9.95 % reduction in Ct values during continuous monitoring can predict COVID-19 progression with a sensitivity of 0.74 (95 % CI = 0.68 to 0.78) and specificity of 0.75 (95 % CI = 0.66 to 0.79) (area under the curve, AUC = 0.79, 95 % CI = 0.76 to 0.81, Fig. 2C). Among all the 255 patients who developed de-novo disease progression during hospitalization, we observed 44 % cases (n = 113) at day 0 (onset of progression), 14.90 % (n = 38) at day −1 (one day before progression), 13.33 % (n = 34) at day −2, and 12.15 % (n = 31) at day −3, with an optimal cut-off value of ΔCt%ORF1ab. The similar performance of the ΔCt% was also observed in the SSPH validation cohort (AUC = 0.79, 95 % CI = 0.75 to 0.83) (Fig. 2D).
Clinical parameters including gender, age, vaccination, and primary disease didn’t alter the predictive performance (AUC) of Ct value shift for disease progression (Supplementary Fig. 2A-D).
Detailed survey of the ROC curve across the different severity of disease further revealed a similar AUC in progression prediction among the asymptomatic (0.81, 95 % CI = 0.73 to 0.89), mild (0.83, 95 % CI = 0.79 to 0.87, asymptomatic vs mild P = 0.34), moderate (0.76, 95 % CI = 0.70 to 0.82, mild vs moderate P = 0.058), and severe groups (0.68, 95 % CI = 0.54 to 0.83, moderate vs severe P = 0.72) in the HSH cohort (Fig. 2E); and asymptomatic (0.82, 95 % CI = 0.75 to 0.90), mild (0.82, 95 % CI = 0.77 to 0.87, asymptomatic vs mild P = 0.89), moderate (0.87, 95 % CI = 0.78 to 0.97, mild vs moderate P = 0.34), and severe groups (0.71, 95 % CI = 0.48 to 0.94, moderate vs severe P = 0.22) in the SSPH cohort (Fig. 2F). The ΔCt% threshold across different predictive sensitivity for patients with distinct clinical characteristics are listed in Supplementary Table 1. Comparison of the ROC curves between patients with or without specific underlying disease revealed a consistent performance of ΔCt% for disease progression (Supplementary Table 2) across the diverse comorbidity background.
To further validate whether the ΔCt% outperformed the canonical COVID-19 severity-related blood markers for disease transformation prediction, the ROC curves for disease progression was constructed with the extreme values of CRP, SAA, LDH, DDI, IL-6, and peripheral blood lymphocyte count before disease progression. The DeLong’s test demonstrated a significant superior performance of ΔCt% over the serum, blood cell or blood gas test results for disease progression prediction (Table 3 ).
Table 3.
Prognostic Performance of Serum, Blood Cell and Blood Gas Assay for COVID-19 Progression.
| Assay | Optimal Threshold (Unit) | Sensitivity (95 % CI) | Specificity (95 % CI) | AUC (95 % CI) | Sample Size | DeLong's P* |
|---|---|---|---|---|---|---|
| CRP | 20.83 (mg/l) | 45.89 % (38.16 %-52.17 %) | 66.96 % (56.80 %-72.94 %) | 0.57 (0.53,0.61) | 1683 | 3.99 × 10-16 |
| SAA | 66.09 (mg/l) | 60.81 % (47.30 %-71.62 %) | 59.41 % (44.21 %-67.88 %) | 0.61 (0.54,0.68) | 855 | 1.01 × 10-8 |
| LDH | 241.50 (IU/l) | 55.32 % (29.79 %-68.09 %) | 69.68 % (38.06 %-74.95 %) | 0.61 (0.52,0.70) | 458 | 3.66 × 10-4 |
| IL-6 | 8.14 (pg/ml) | 52.89 % (42.14 %-61.16 %) | 63.79 % (51.55 %-71.21 %) | 0.61 (0.56,0.67) | 816 | 3.74 × 10-5 |
| DDI | 2.22 (mg/l) | 54.55 % (36.36 %-68.18 %) | 60.34 % (32.76 %-72.86 %) | 0.56 (0.46,0.66) | 1373 | 6.49 × 10-6 |
| LYMPH#** | 0.97 (×109) | 57.81 % (64.05 %-50.54 %) | 63.90 % (55.91 %-68.38 %) | 0.62 (0.58,0.66) | 1683 | 2.91 × 10-11 |
| pH | 7.38 | 84.62 % (67.69 %-92.31 %) | 29.92 % (11.81 %-38.98 %) | 0.57 (0.49,0.64) | 358 | 2.52 × 10-7 |
| pCO2 | 37.62 (mmHg) | 44.44 % (28.57 %-55.56 %) | 60.16 % (42.57 %-70.51 %) | 0.49 (0.41,0.58) | 358 | 1.93 × 10-10 |
| pO2 | 82.13 (mmHg) | 46.77 % (30.65 %-58.71 %) | 69.65 % (52.92 %-79.38 %) | 0.58 (0.5,0.67) | 358 | 7.35 × 10-6 |
| HCO3– | 26.25 (mmol/l) | 65.57 % (48.08 %-77.05 %) | 40.7 % (23.26 %-52.71 %) | 0.51 (0.43,0.6) | 358 | 3.24 × 10-9 |
*DeLong‘s P-value for comparison of the specific assay and ΔCt%.
**LYMPH#: Peripheral blood lymphocyte count.
According to the ROC curve, we determined the optimal ΔCt% threshold for COVID-19 progression prediction as −20 % for the asymptomatic individuals, −10 % for the mild, moderate, and severe cases, which were further validated in the 262 cases from the SSPH cohort. In short, ΔCt% was able to identify the progressing COVID-19 patients with an AUC of 0.79 (95 % CI = 0.75 to 0.83). A ΔCt% cut-off of −10 % was able to identify the progressing COVID-19 patients with a sensitivity of 72.80 % (95 % CI = 61.11 % to 91.67 %) and specificity of 68.44 % (95 % CI = 56.78 % to 88.85 %) in this validation cohort (Table 4 ).
Table 4.
Prognostic performance for COVID-19 progression of ΔCt% in the SSPH cohort.
| Subgroup | Threshold of ΔCt% | Sensitivity (95 % CI) | Specificity (95 % CI) |
|---|---|---|---|
| Overall | −10 % | 72.80 % (61.11 %-91.67 %) | 68.44 % (56.78 %-88.85 %) |
| Asymptomatic | −20 % | 52.78 % (33.33 %-69.44 %) | 93.69 % (71.36 %-98.06 %) |
| Mild | −10 % | 70.81 % (66.67 %-87.65 %) | 76.54 % (57.76 %-83.23 %) |
| Moderate | −10 % | 76.67 % (49.75 %-95.25 %) | 71.43 % (61.29 %-96.77 %) |
| Severe | −10 % | 63.64 % (47.5 %-97.95 %) | 69.23 % (40.38 %-75 %) |
*CI: confidential interval.
2.1. SARS-CoV-2 viral load dynamics is a host immune-independent indicator for COVID-19 prognosis
COVID-19 symptoms are a result of viral replication and host immune balance [21]. To further evaluate if the dynamics in SARS-CoV-2 viral load can maintain the performance for COVID-19 prognosis prediction across a range of different “host” immune status, we classified 188 COVID-19 patients from the HSH cohort by unsupervised clustering of the immune-related markers including serum cytokine quantification, white blood cell count, and T/B/NK cell profiling. The patients were clustered into two distinct immuno-subtypes (Fig. 3 A). Immune-subtype 1 demonstrated an activated cellular immunity phenotype, with increases in both percentage of lymphocyte (P = 7.56 × 10-36) and T cell count (CD3+, P = 7.45 × 10-13; CD4+, P = 1.40 × 10-11, CD8+, P = 6.97 × 10-3), while immune-subtype 2 manifested a dysfunctional granulopoiesis phenotype with elevated serum IL-6 (P = 2.30 × 10-3) and neutrophil count (P = 2.66 × 10-38) (Fig. 3B).
Fig. 3.
SARS-CoV-2 viral load dynamics is a host immune-independent indicator for COVID-19 prognosis. A) Unsupervised hierarchical clustering of immune-related markers including serum cytokine quantification, white blood cell count, and T/B/NK cell profiling identified COVID-19 patients with two distinct immune backgrounds. B) COVID-19 patients with immuno-subtype 1 demonstrated an activated cellular immunity phenotype while immuno-subtype 2 represented a dysfunctional granulopoiesis phenotype. C) ROC curve of ΔCt% for COVID-19 progression across different immune-subtypes.
Given this distinct immune background [22], we then evaluated the performance of Ct value dynamics for disease prognosis in these two subgroups. For the prediction of disease progression, no significant difference was observed between the AUCs of immuno-subtype 1 (0.74, 95 % CI = 0.61 to 0.87) and subtype 2 (0.77, 95 % CI = 0.67 to 0.88, P = 0.69). With a threshold of −10 %, the ΔCt% demonstrated a similar prognostic performance for COVID-19 progression between the immuno-subgroup 1 (sensitivity: 80 %, specificity: 67 %) and subgroup 2 (sensitivity: 76 %, specificity: 71 %). All these data suggested the ΔCt% as a host immune-independent marker for COVID-19 prognosis.
2.1. Persistent high SARS-CoV-2 viral load pattern indicates poor COVID-19 survival
After validating the performance of time-wise derivative of Ct value for disease transformation prediction, we further observed the temporal alteration of Ct values across different COVID-19 severity, especially in the severe cases. By dividing the time of hospital stay into early hospitalization (day 1 to day 3) and late hospitalization (day 4 to day 10), we observed a significant decrease in the Ct value of ORF1ab and N gene during the late stage of hospitalization among the asymptomatic, mild, moderate, and severe groups with favorable prognosis. However, this significant decrease was not observed in the patients who died from COVID-19 (Fig. 4 A). This phenomenon can also be confirmed by longitudinal tracking of the Ct value during patient hospitalization (Fig. 4B). Among the 67 severe cases with sequential monitoring of SARS-CoV-2 RT-PCR, Kaplan-Meier analysis revealed a significantly improved survival in patients with a (mean CtLateStage – mean CtEarlyStage)/mean CtEarlyStage < 50 % (HR = 0.16, 95 % CI = 0.04 to 0.65, P = 2.41 × 10-3, Fig. 4C, Supplementary Fig. 3). The longitudinal monitoring of Ct value further demonstrated a superior performance over classical infectious or coagulation indicators for high-risk patient stratification (Fig. 4D), which indicated the importance of long-term Ct value monitoring in the risk classification of severe COVID-19 cases.
Fig. 4.
Persistent high SARS-CoV-2 viral load pattern indicates poor COVID-19 survival. A) Significantly decreased SARS-CoV-2 viral load of ORF1ab and N gene between early hospitalization (day 1 to day 3) and late hospitalization (day 4 to day 10) in asymptomatic patients and those with mild, moderate, or severe (survived) disease, but not in those patients who died from COVID-19. B) Temporal dynamic Ct value of ORF1ab and N gene demonstrating a persistent high SARS-CoV-2 viral load in deceased COVID-19 patients in comparison to patients with other disease severity (asymptomatic, mild, moderate, and severe). C) The monitoring of mean Ct value alteration (ORF1ab and N gene) between early and late stage revealed improved survival among 67 patients with severe disease, with > 50 % decrease of SARS-CoV-2 viral load (Kaplan-Meier analysis). D) Ct value alteration monitoring, LDH, and peripheral blood lymphocyte counting were prognostic markers for the survival of COVID-19 patients. Two-sample t-test, n.s. > 0.05; * P < 0.05; **** P < 0.0001.
Discussion
Facing the fifth variant of concern (omicron) of SARS-CoV-2′ s spreading with unprecedented infectiousness [23], [24], [25], medical resource crowding lays a great obstacle on the restoration of social order, especially in the area with limited per capita medical infrastructure and medical service personnel. Therefore, stratifying SARS-COV-2 infected patients with an accurate and predictive classifier is crucial for the precise control of the epidemics. In addition, due to the diverse symptoms of COVID-19, how to manage patients with reliable prognostic markers has become one of the core issues in clinical care.
In the current study, by incorporating the clinical characteristic data, disease transformation records, and temporal conversion of SARS-CoV-2 viral load from over 3,000 COVID-19 patients, we demonstrated that continuous monitoring of SARS-CoV-2 viral load (Ct value) can not only predict the upcoming disease progression, but also serves as a prognostic marker for poor COVID-19 patient survival. The performance of ΔCt% for the prediction of COVID-19 progress was further validated in an independent cohort of 262 patients from another center, confirming the robustness of Ct value monitoring for the stratification of disease severity conversion. Moreover, dynamic viral load monitoring outperformed the classical infectious or coagulation-based blood tests for COVID-19 prognosis. To our best knowledge, this is the first large-scale diagnostic trial reporting the clinical relevance of SARS-CoV-2 viral load monitoring in the prediction of COVID-19 severity conversion.
The balance of viral multiplication and shedding dynamics shaped by virus replicative ability and host defenses ensures that viral load of SARS-CoV-2 in the target organ may be sufficient to cause COVID-19 symptoms [26]. It is reasonable to hypothesize that quantification of the virus component is a direct measurement of disease severity. Previous studies reported viral nucleic acid [9], [27] and plasma nucleocapsid protein [28], [29] as the independent risk factor for ICU admission and mortality among COVID-19 patients. However, these markers have been proven insufficient for clinical decision making due to the natural diversity in patients’ basic health status and immune function. With the viral kinetic data from >3000 COVID-19 patients from two centers and standardized sampling procedure, our study was able to attenuate the bias from pre-analytical errors, and validated the dynamics in viral load as a predictive indicator for infection outcome, which eliminated inter-patient variances of age, vaccination, underlying diseases, and basic immune status that have been defined by classical COVID-19 related serum markers including interleukin profiling, WBC counting and lymphocyte subsetting.
The prompt population screening against COVID-19 pandemic has driven the constant reduction in the cost of SARS-CoV-2 PCR testing since late 2019. It has been well discussed that the PCR test can cost-effectively benefit both the public health [30] and clinical practice [31] when the price of a single test is set to the dollar-level, which is now widely guaranteed by the healthcare system.
Recent studies from two small cohorts reported that high viral load persists in patients with severe disease [32] and associates with poor patient prognosis [33]. One limitation of our study was that we were not able to determine the absolute copy number of SARS-CoV-2 RNA, though the Ct value data in this study was obtained from the RT-PCR systems with a roughly validated linearity, LoQ, and precision. A further quantitative PCR system with a fully validated quantitative performance and traceable standards is required to build a refined predictive model for COVID-19 prognosis across different laboratories [34]. Additionally, this study was performed with the non-invasive nasopharyngeal sampling due to the relative low frequency of severe disease conversion of omicron variant infection, however studies on disease transformation-coordinated viral kinetics should be encouraged in more sampling types including salvia and serum [35], to determine the desired sample type with minimized pre-analytical heterogeneity and the best performance for symptom conversion prediction.
Conclusions
The dynamic alteration of SARS-CoV-2 viral load may perform as a robust prognostic marker for COVID-19 progression that bypasses the heterogeneity of patients’ basic immune status. Establishment of a fully validated system for viral load quantification and traceable standards should be encouraged to facilitate the prognostic transformation of SARS-CoV-2 RT-PCR tests.
CRediT authorship contribution statement
Zhiyuan Wu: Conceptualization, Data curation, Funding acquisition, Writing – original draft. Can Yang: Data curation, Formal analysis. Yutao Shen: Data curation. Qingyun Zhang: Data curation, Investigation. Xuemei Tang: Data curation, Investigation. Di Wang: Investigation. Yu Xu: Investigation. Guojun Cao: Investigation, Methodology. Xiaodong Song: Investigation. Yanchun Ma: Investigation, Methodology. Huajie Fan: Investigation. Hailong Lu: Investigation. Yaju Li: Investigation. Xiangyu Li: Investigation. Yiqin Shen: Investigation. Chen Zhang: Investigation. Min Zhu: Investigation. Xiaoyan Teng: Data curation, Investigation. Yuzhen Du: Data curation, Investigation. Ming Guan: Conceptualization, Funding acquisition, Project administration, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgement
This study was supported by Shanghai “Rising Stars of Medical Talents”–Clinical Laboratory Practitioner Program (grant number 2022-065 to Z.W.), Shanghai Municipal Health Commission (grant number 2022YQ045 to Z.W.; grant number shslczdzk03303 to M.G.), Shanghai Municipal Science and Technology Commission (grant number HS2021SHZX001 to M.G.).
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.cca.2023.117227.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Data availability
No data was used for the research described in the article.
References
- 1.Zarebska-Michaluk D., Flisiak R. Early oral antiviral use in patients hospitalised with COVID-19. Lancet Infect. Dis. 2022;22(12):1650–1651. doi: 10.1016/S1473-3099(22)00522-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Owen A., Diaz J.V., Guyatt G., Lamontagne F., Stegemann M., Vandvik P.O., Agoritsas T. WHO Living Guidelines on antivirals for COVID-19 are evidence-based. Lancet. 2022 doi: 10.1016/S0140-6736(22)02306-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Terpos E., Ntanasis-Stathopoulos I., Elalamy I., Kastritis E., Sergentanis T.N., Politou M., Psaltopoulou T., Gerotziafas G., Dimopoulos M.A. Hematological findings and complications of COVID-19. Am. J. Hematol. 2020;95(7):834–847. doi: 10.1002/ajh.25829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yun H., Sun Z., Wu J., Tang A., Hu M., Xiang Z. Laboratory data analysis of novel coronavirus (COVID-19) screening in 2510 patients. Clin. Chim. Acta. 2020;507:94–97. doi: 10.1016/j.cca.2020.04.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen X., Zhou J., Chen C., Hou B., Ali A., Li F., Hua Z., Wu Y., Yang Q., Chen M., Zhang R., Huang Q., Ding J., Zhang X.E., Men D. Consecutive monitoring of interleukin-6 is needed for COVID-19 patients. Virol. Sin. 2021;36(5):1093–1096. doi: 10.1007/s12250-021-00425-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.De Michieli L., Ola O., Knott J.D., Akula A., Mehta R.A., Hodge D.O., Dworak M., Yang E.H., Gharacholou M., Singh G., Singh R., Gulati R., Jaffe A.S., Sandoval Y. High-sensitivity cardiac troponin t for the detection of myocardial injury and risk stratification in COVID-19. Clin. Chem. 2021;67(8):1080–1089. doi: 10.1093/clinchem/hvab062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Thomas M., Scully M. Clinical features of thrombosis and bleeding in COVID-19. Blood. 2022 doi: 10.1182/blood.2021012247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.T.H. Lee, R. Junhao Lin, R.T.P. Lin, T. Barkham, P. Rao, Y.S. Leo, D. Chien Lye, B. Young, C.-O.R.T. National Centre for Infectious Diseases, Testing for SARS-CoV-2: Can We Stop at 2?, Clin Infect Dis 71(16) (2020) 2246-2248. [DOI] [PMC free article] [PubMed]
- 9.Magleby R., Westblade L.F., Trzebucki A., Simon M.S., Rajan M., Park J., Goyal P., Safford M.M., Satlin M.J. Impact of severe acute respiratory syndrome coronavirus 2 viral load on risk of intubation and mortality among hospitalized patients with coronavirus disease 2019. Clin. Infect. Dis. 2021;73(11):e4197–e4205. doi: 10.1093/cid/ciaa851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Rao S.N., Manissero D., Steele V.R., Pareja J. A systematic review of the clinical utility of cycle threshold values in the context of COVID-19. Infect. Dis. Ther. 2020;9(3):573–586. doi: 10.1007/s40121-020-00324-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Westblade L.F., Brar G., Pinheiro L.C., Paidoussis D., Rajan M., Martin P., Goyal P., Sepulveda J.L., Zhang L., George G., Liu D., Whittier S., Plate M., Small C.B., Rand J.H., Cushing M.M., Walsh T.J., Cooke J., Safford M.M., Loda M., Satlin M.J. SARS-CoV-2 viral load predicts mortality in patients with and without cancer who are hospitalized with COVID-19. Cancer Cell. 2020;38(5):661–671 e2. doi: 10.1016/j.ccell.2020.09.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Evans D., Cowen S., Kammel M., O'Sullivan D.M., Stewart G., Grunert H.P., Moran-Gilad J., Verwilt J., In J., Vandesompele J., Harris K., Hong K.H., Storey N., Hingley-Wilson S., Duhring U., Bae Y.K., Foy C.A., Braybrook J., Zeichhardt H., Huggett J.F. The dangers of using Cq to quantify nucleic acid in biological samples: a lesson from COVID-19. Clin. Chem. 2021;68(1):153–162. doi: 10.1093/clinchem/hvab219. [DOI] [PubMed] [Google Scholar]
- 13.Williams C.M., Pan D., Decker J., Wisniewska A., Fletcher E., Sze S., Assadi S., Haigh R., Abdulwhhab M., Bird P., Holmes C.W., Al-Taie A., Saleem B., Pan J., Garton N.J., Pareek M., Barer M.R. Exhaled SARS-CoV-2 quantified by face-mask sampling in hospitalised patients with COVID-19. J. Infect. 2021;82(6):253–259. doi: 10.1016/j.jinf.2021.03.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.W.H.O., Living guidance for clinical management of COVID-19, 2021. https://www.who.int/publications/i/item/WHO-2019-nCoV-Clinical-2022.2. (Accessed December 1 2022).
- 15.C. CDC, China CDC Primers and probes for detection 2019-nCoV (posted on 24 January 2020), 2020. https://www.who.int/docs/default-source/coronaviruse/whoinhouseassays.pdf. (Accessed June 10 2022).
- 16.N.H.C.o.t. PRC, Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia (Trial Version 9), 2022. http://www.nhc.gov.cn/yzygj/s7653p/202203/b74ade1ba4494583805a3d2e40093d88.shtml. (Accessed May 29 2022).
- 17.Robin X., Turck N., Hainard A., Tiberti N., Lisacek F., Sanchez J.C., Muller M. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinf. 2011;12:77. doi: 10.1186/1471-2105-12-77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wolfel R., Corman V.M., Guggemos W., Seilmaier M., Zange S., Muller M.A., Niemeyer D., Jones T.C., Vollmar P., Rothe C., Hoelscher M., Bleicker T., Brunink S., Schneider J., Ehmann R., Zwirglmaier K., Drosten C., Wendtner C. Virological assessment of hospitalized patients with COVID-2019. Nature. 2020;581(7809):465–469. doi: 10.1038/s41586-020-2196-x. [DOI] [PubMed] [Google Scholar]
- 19.Han M.S., Byun J.H., Cho Y., Rim J.H. RT-PCR for SARS-CoV-2: quantitative versus qualitative. Lancet Infect. Dis. 2021;21(2):165. doi: 10.1016/S1473-3099(20)30424-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.X. Ma, J. Ai, J. Cai, S. Chen, S. Wang, H. Zhang, K. Lin, W. Zhang, H. Wang, Y. Zhang, F. Sun, Y. Li, S. Zhang, L. Shen, S. Chen, G. Lu, J. Xu, X. Chen, W. Zhang, Dynamic Disease Manifestations Among Non-Severe COVID-19 Patients Without Unstable Medical Conditions: A Follow-Up Study — Shanghai Municipality, China, March 22–May 03, 2022, China CDC Wkly (2022).
- 21.Brodin P. Immune determinants of COVID-19 disease presentation and severity. Nat. Med. 2021;27(1):28–33. doi: 10.1038/s41591-020-01202-8. [DOI] [PubMed] [Google Scholar]
- 22.C.-M.-o.B.A. Consortium., A blood atlas of COVID-19 defines hallmarks of disease severity and specificity, Cell 185(5) (2022) 916-938 e58. [DOI] [PMC free article] [PubMed]
- 23.P. Elliott, O. Eales, N. Steyn, D. Tang, B. Bodinier, H. Wang, J. Elliott, M. Whitaker, C. Atchison, P.J. Diggle, A.J. Page, A.J. Trotter, D. Ashby, W. Barclay, G. Taylor, H. Ward, A. Darzi, G.S. Cooke, C.A. Donnelly, M. Chadeau-Hyam, Twin peaks: The Omicron SARS-CoV-2 BA.1 and BA.2 epidemics in England, Science (2022) eabq4411. [DOI] [PMC free article] [PubMed]
- 24.Li Y. Characteristics and Outcomes of Hospitalized Patients in South Africa During the COVID-19 Omicron Wave. J. Am. Med. Assoc. 2022;327(21):2147–2148. doi: 10.1001/jama.2022.5572. [DOI] [PubMed] [Google Scholar]
- 25.Zhang X., Zhang W., Chen S. Shanghai's life-saving efforts against the current omicron wave of the COVID-19 pandemic. Lancet. 2022;399(10340):2011–2012. doi: 10.1016/S0140-6736(22)00838-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.V'Kovski P., Kratzel A., Steiner S., Stalder H., Thiel V. Coronavirus biology and replication: implications for SARS-CoV-2. Nat. Rev. Microbiol. 2021;19(3):155–170. doi: 10.1038/s41579-020-00468-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fajnzylber J., Regan J., Coxen K., Corry H., Wong C., Rosenthal A., Worrall D., Giguel F., Piechocka-Trocha A., Atyeo C., Fischinger S., Chan A., Flaherty K.T., Hall K., Dougan M., Ryan E.T., Gillespie E., Chishti R., Li Y., Jilg N., Hanidziar D., Baron R.M., Baden L., Tsibris A.M., Armstrong K.A., Kuritzkes D.R., Alter G., Walker B.D., Yu X., Li J.Z., R. Massachusetts Consortium for Pathogen SARS-CoV-2 viral load is associated with increased disease severity and mortality. Nat. Commun. 2020;11(1):5493. doi: 10.1038/s41467-020-19057-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang H., Hogan C.A., Verghese M., Solis D., Sibai M., Huang C., Roltgen K., Stevens B.A., Yamamoto F., Sahoo M.K., Zehnder J., Boyd S.D., Pinsky B.A. SARS-CoV-2 Nucleocapsid Plasma Antigen for Diagnosis and Monitoring of COVID-19. Clin. Chem. 2021;68(1):204–213. doi: 10.1093/clinchem/hvab216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Brasen C.L., Christensen H., Olsen D.A., Kahns S., Andersen R.F., Madsen J.B., Lassen A., Kierkegaard H., Jensen A., Sydenham T.V., Madsen J.S., Moller J.K., Brandslund I. Daily monitoring of viral load measured as SARS-CoV-2 antigen and RNA in blood, IL-6, CRP and complement C3d predicts outcome in patients hospitalized with COVID-19. Clin. Chem. Lab. Med. 2021;59(12):1988–1997. doi: 10.1515/cclm-2021-0694. [DOI] [PubMed] [Google Scholar]
- 30.Deckert A., Anders S., de Allegri M., Nguyen H.T., Souares A., McMahon S., Boerner K., Meurer M., Herbst K., Sand M., Koeppel L., Siems T., Brugnara L., Brenner S., Burk R., Lou D., Kirrmaier D., Duan Y., Ovchinnikova S., Marx M., Krausslich H.G., Knop M., Barnighausen T., Denkinger C. Effectiveness and cost-effectiveness of four different strategies for SARS-CoV-2 surveillance in the general population (CoV-Surv Study): a structured summary of a study protocol for a cluster-randomised, two-factorial controlled trial. Trials. 2021;22(1):39. doi: 10.1186/s13063-020-04982-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.A.M. Neilan, E. Losina, A.C. Bangs, C. Flanagan, C. Panella, G.E. Eskibozkurt, A. Mohareb, E.P. Hyle, J.A. Scott, M.C. Weinstein, M.J. Siedner, K.P. Reddy, G. Harling, K.A. Freedberg, F.M. Shebl, P. Kazemian, A.L. Ciaranello, Clinical Impact, Costs, and Cost-effectiveness of Expanded Severe Acute Respiratory Syndrome Coronavirus 2 Testing in Massachusetts, Clin Infect Dis 73(9) (2021) e2908-e2917. [DOI] [PMC free article] [PubMed]
- 32.Zheng S., Fan J., Yu F., Feng B., Lou B., Zou Q., Xie G., Lin S., Wang R., Yang X., Chen W., Wang Q., Zhang D., Liu Y., Gong R., Ma Z., Lu S., Xiao Y., Gu Y., Zhang J., Yao H., Xu K., Lu X., Wei G., Zhou J., Fang Q., Cai H., Qiu Y., Sheng J., Chen Y., Liang T. Viral load dynamics and disease severity in patients infected with SARS-CoV-2 in Zhejiang province, China, January-March 2020: retrospective cohort study. BMJ. 2020;369 doi: 10.1136/bmj.m1443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.N. Neant, G. Lingas, Q. Le Hingrat, J. Ghosn, I. Engelmann, Q. Lepiller, A. Gaymard, V. Ferre, C. Hartard, J.C. Plantier, V. Thibault, J. Marlet, B. Montes, K. Bouiller, F.X. Lescure, J.F. Timsit, E. Faure, J. Poissy, C. Chidiac, F. Raffi, A. Kimmoun, M. Etienne, J.C. Richard, P. Tattevin, D. Garot, V. Le Moing, D. Bachelet, C. Tardivon, X. Duval, Y. Yazdanpanah, F. Mentre, C. Laouenan, B. Visseaux, J. Guedj, C.C.I. French, g. French Cohort Study, Modeling SARS-CoV-2 viral kinetics and association with mortality in hospitalized patients from the French COVID cohort, Proc Natl Acad Sci U S A 118(8) (2021). [DOI] [PMC free article] [PubMed]
- 34.A.A.C.C., AACC Recommendation for Reporting SARS-CoV-2 Cycle Threshold (CT) Values, 2022. https://www.aacc.org/science-and-research/covid-19-resources/statements-on-covid-19-testing/aacc-recommendation-for-reporting-sars-cov-2-cycle-threshold-ct-values. (Accessed June 16 2022).
- 35.Ke R., Martinez P.P., Smith R.L., Gibson L.L., Mirza A., Conte M., Gallagher N., Luo C.H., Jarrett J., Zhou R., Conte A., Liu T., Farjo M., Walden K.K.O., Rendon G., Fields C.J., Wang L., Fredrickson R., Edmonson D.C., Baughman M.E., Chiu K.K., Choi H., Scardina K.R., Bradley S., Gloss S.L., Reinhart C., Yedetore J., Quicksall J., Owens A.N., Broach J., Barton B., Lazar P., Heetderks W.J., Robinson M.L., Mostafa H.H., Manabe Y.C., Pekosz A., McManus D.D., Brooke C.B. Daily longitudinal sampling of SARS-CoV-2 infection reveals substantial heterogeneity in infectiousness. Nat. Microbiol. 2022;7(5):640–652. doi: 10.1038/s41564-022-01105-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No data was used for the research described in the article.