Nasopharyngeal viral load at admission is not an independent predictor of thromboembolic complications in unvaccinated COVID-19 hospitalized patients.

Abstract

COVID-19 patients may develop thrombotic complications, and data regarding an association between nasopharyngeal viral load and thrombosis is scarce. The aim of our study was to evaluate whether SARS-CoV-2 nasopharyngeal viral load upon admission is a useful prognostic marker for the development of thromboembolic events in patients hospitalized for SARS-CoV-2 infection. We performed a retrospective study of all hospitalized patients with a positive PCR test for SARS-CoV2 who had deep vein thrombosis (DVT), pulmonary embolization (PE), or arterial thrombosis diagnosed during their clinical course in a single academic center. The study population was divided according to the cycle threshold (Ct) value upon admission in patients with high viral load (Ct < 25), intermediate/medium viral load (Ct 25–30), and low viral load (Ct > 30). A regression model for propensity was performed matching in a 1:3 ratio those patients who had a thrombotic complication to those who did not. Among 2,000 hospitalized COVID-19 patients, 41 (2.0%) developed thrombotic complications. Of these, 21 (51.2%) were diagnosed with PE, eight (19.5%) were diagnosed with DVT, and 12 (29.2%) were diagnosed with arterial thrombosis. Thrombotic complications occurred as frequently among the nasopharyngeal viral load or severity stratification groups with no statistically significant differences. Univariate logistic regression revealed increased odds for thrombosis only in mechanically ventilated patients OR 3.10 [1.37, 7.03] (p = 0.007). Admission SARS-CoV-2 nasopharyngeal viral loads, as determined by Ct values, were not independently associated with thromboembolic complications among hospitalized patients with COVID-19.

Highlights

• Type of Research: Single-center retrospective case-control study.

• Key Findings: Among 2,000 hospitalized COVID-19 patients, 41 (2.0%) developed thrombotic complications.

• Thrombotic complications occurred as frequently among the three nasopharyngeal viral load groups and stratification of severity groups with no statistically significant differences.

• Take home Message: SARS-CoV-2 nasopharyngeal viral load at admission was not independently associated with thromboembolic complications among COVID-19 hospitalized patients.

Introduction

SARS-CoV-2, which causes the coronavirus disease (COVID-19), emerged in late 2019. As of April 13, 2022, it has infected over 500 million people and caused more than 6.1 million deaths worldwide [1]. It has been documented that nearly a third of the patients infected have an asymptomatic course; however, the other two-thirds of patients develop mild to severe symptoms [2, 3]. Most patients have mild symptoms and can be treated at home; yet, some individuals develop pneumonia which can lead to respiratory failure, requiring hospitalization. Increased age, obesity, diabetes, and cardiovascular diseases are some of the main risk factors for a severe clinical presentation [4].

The main target of SARS-CoV-2 is the respiratory system, but other systems and organs are also affected. Hospitalized COVID-19 patients often present thrombotic complications, particularly, patients admitted to the intensive care unit (ICU) [5–7]. Venous thromboembolism (VTE), such as deep vein thrombosis (DVT) and pulmonary embolism (PE), is reported as the most common thromboembolic event in hospitalized COVID-19 patients. Despite receiving antithrombotic therapy at prophylactic doses, COVID-19 patients still present higher rates of VTE compared to otherwise hospitalized patients, with 10–40% VTE complications in hospitalized COVID-19 patients [8].

The intensity of antithrombotic therapy remains controversial [9]. Some randomized controlled trials (RCTs) have shown that the use of therapeutic heparin reduced mortality or VTE [10, 11]; however, differences in the design limit the interpretation and application of these results, as reflected in the conditional recommendations of multiple guidelines [12–14]. Until these results of more RCTs are available to determine safety and effectiveness of this therapeutic approach, clinicians need to rely on individual factors to identify the patients that may benefit from more aggressive treatment. Since inflammatory and coagulopathy markers are commonly elevated in hospitalized COVID-19 patients, there are not clear cutoffs values to predict VTE in these patients [9].

There is a need to identify a more reliable predictor of VTE in hospitalized COVID-19 patients. Quantitative nasopharyngeal viral load at admission has been proposed as a predictor for disease severity upon admission [15]. However, no further analysis has been done to establish if there is a correlation between nasopharyngeal viral load and VTE risk in particular. If there were, then nasopharyngeal viral loads could be used as a risk stratification factor to identify patients that would benefit from a more aggressive antithrombotic therapy.

The aim of this study was to analyze whether SARS-CoV-2 nasopharyngeal viral load at admission is a viable prognostic marker for the development of VTE in hospitalized COVID-19 patients.

Methods

Study population and setting

This study was performed with prior IRB review and approval no. I004738 and in accordance with institutional, national (Mexican General Health Law in Health-Related Research Article 17) and international (Helsinki Declaration of 1964 and 2012 revision) regulations and standards. This retrospective, case–control study consisted of hospitalized patients with diagnosis of symptomatic infection of COVID-19 between March 2020 and June 2021. All of them presented a positive reverse transcription-polymerase chain reaction (RT-PCR) result for SARS-CoV-2 at Hospital San Jose Tecnologico de Monterrey in Monterrey, Mexico.

All COVID-19 patients who developed DVT, PE, or arterial thrombosis during their hospitalization were included in the study. A regression model for propensity scoring was performed based on gender, age, smoking status, diabetes mellitus, and systemic arterial hypertension. Patients with thrombosis (DVT, PE, or arterial thrombosis) were matched in a 1:3 fashion to those without thrombotic problems using nearest neighbor algorithm with a 0.05 clamp.

Patients who presented thrombotic events at admission or had risk factors that predisposed them to thrombotic events (neoplasia, pregnancy, antiphospholipid syndrome, coagulopathy or thrombophilia, atrial fibrillation, use of contraceptives, or a history of a previous thrombotic event) were excluded from the study.

Nasopharyngeal viral load assessment

Nasopharyngeal viral load was assessed at admission by collecting a nasopharyngeal swab and performing a RT-PCR assay. The RT-PCR identifies three of the main genes of the SARS-CoV-2 genome: Orf1ab, N gene, and S gene. Orf1ab detects only genomic SARS-CoV-2 RNA, whereas N and S may also detect any subgenomic RNA released by lysed cells. This instrument provides a Cycle threshold (Ct) value for each target, representing the number of replication cycles needed to detect the amplification above a pre-set threshold. The relationship between Ct and the nasopharyngeal viral load is inversely proportional, which means the higher the viral load, the lower the Ct value. Based on previously published cut-offs¹⁷, the study population was divided according to the Ct value upon admission in patients with high (Ct < 25), intermediate (Ct 25–30), and low (Ct > 30) nasopharyngeal viral load.

Statistical analyses

Data analysis was performed in SPSS v 25 (IBM, Boston). Normality of distribution was assessed using the Kolmogorov–Smirnov Test. Normally distributed data was tested using t-student test and ANOVA with Tukey as a post-hoc test and results described in mean and standard deviation. Non-normally distributed data was tested using nonparametric tests including Kruskal–Wallis and Mann–Whitney-U, with results expressed as medians and ranges. Ordinal and Nominal data was tested using the Chi-Square test or Fisher's Exact where applicable, and results are expressed as frequencies and percentages.

Univariate logistic regressions were performed using the enter method, with appraisal of the Omnibus test for coefficients and the Hosmer–Lemeshow Test for data fitness. Results are expressed in Odds Ratios (OR). Additionally, multivariate analysis was performed using multivariate logistic regression using the forward stepwise Wald approach, with appraisal of the Omnibus test for coefficients and the Hosmer–Lemeshow Test for data fitness. A p-value of < 0.05 was used to determine statistical significance.

Results

Among 2,000 hospitalized COVID-19 patients, 41 (2.0%) developed thrombotic complications during their hospital stay. Of these, 21 (51.2%) were diagnosed with PE, eight (19.5%) were diagnosed with DVT, and 12 (29.2%) were diagnosed with arterial thrombosis. One hundred twenty-three patients were identified to serve as a control group.

No significant differences were found in baseline characteristics such as age, gender, or presence of comorbidities such as diabetes mellitus, systemic arterial hypertension, obesity, and smoking. None were vaccinated against COVID-19. Overall baseline characteristics are displayed in Table 1.

Table 1.

Baseline characteristics of hospitalized COVID19 patients

Variable	No thrombotic complications N (%)	Thrombotic complications N (%)	P-value
Age (years) [mean, SD]	61.9 (14.6)	61.9 (14.1)	.326
Female	29 (23.6%)	10 (24.4%)	.535
Male	94 (76.4%)	31 (75.6%)
BMI [median (IQR)]	30 (17–53)	29.7 (22.9–44.4)	.274
Overweight	40 (32.8%)	14 (35.9%)	.431
Obesity	62 (50.8%)	19 (48.7%)	.352
Comorbidities
Smoking	8 (6.5%)	4 (9.8%)	.732
Hypertension	73 (59.3%)	23 (56.1%)	.425
Diabetes Mellitus	62 (50.4%)	21 (51.2%)	.536
Clinical Presentation
Days of symptoms prior to admission [median (IQR)]	7 (0–24)	8 (1–30)	.225
Highest oxygen requirement at admission			.472
Oxygen by low-flow nasal cannula	83 (67.5%)	29 (70.7%)
Oxygen by high-flow nasal cannula	29 (23.6%)	8 (19.5%)
Mechanical ventilation	2 (1.6%)	1 (2.4%)
Laboratory values at admission
Leukocytosis: WBC > 11 × 109 cells/L	43 (35.2%)	18 (46.2%)	.151
Lymphopenia: lymphocyte count < 1 × 109 cells/L	70 (57.9%)	29(76.3%)	.054
Anemia: hemoglobin < 12 g/dL	15 (12.5%)	9 (24.1%)	.120
Thrombocytopenia: platelet count < 150 × 109 cells/L	27 (22%)	7 (17.9%)	.659
Inflammatory markers at admission
C-reactive protein (mg/dL) [median (IQR)]	16.3 (0.8–55.8)	19.4 (0.2–38.9)	.930
Procalcitonin (ng/mL) [median (IQR)]	0.2 (0–5.3)	0.28 (0–4.3)	.080
Ferritin (ng/mL) [median (IQR)]	1,688 (40–10,074)	1,696 (22.5–5,585)	.592
D-dimer (ng/mL) [median (IQR)]	259 (53–41,436)	405 (102–43,586)	.001

SD standard deviation, BMI body mass index, WBC white blood count

Thrombotic complications

Thrombotic complications occurred as frequently among the three nasopharyngeal viral load groups and stratification of severity groups with no statistically significant differences. Moreover, mean nasopharyngeal viral load between patients with thrombotic events and patients without thrombotic events was similar, as was their body mass index (BMI) and platelet count, depicted in Table 1 and 2. Patients with thrombotic complications had higher median D-dimer than those without thrombosis (p = 0.034). Additionally, mean oxygen saturation at admission in patients with thrombotic complications was lower (77.1% ± 15.2%) compared to patients without thrombotic complications (85.8% ± 10.6%) (p = 0.001). Further analysis and post-hoc reveals differences are attributable to PE specifically, as ANOVA with Tukey reveals significant differences between patients with PE and patients without thrombotic complications. There were no differences in oxygen saturation at admission between DVT, arterial thrombosis, and the non-thrombosis group.

Table 2.

Frequency of thrombotic complications by viral load stratification

Variable	No thrombotic complications N (%)	Thrombotic complications N (%)	P-value
Viral load by nasopharyngeal swab			.121
High viral load	64 (52.0%)	24 (58.5%)
Medium viral load	46 (37.4%)	9 (22.0%)
Low viral load	13 (10.6%)	8 (19.5%)

Treatments and outcomes of patients with and without thrombotic complications are summarized in Table 3. Thrombosis was equally frequent in patients who received steroids, convalescent plasma, or anticoagulation than in those who did not. Patients on mechanical ventilation had thrombotic complications more frequently than patients without mechanical ventilation in 61.0% vs. 36.0% of the cases, respectively, (p = 0.006). Patients with thrombotic events presented higher in-hospital mortality than patients without thrombotic events (51.2% vs. 26.0%, respectively, p = 0.002). Individual analysis by DVT, PE, or arterial thrombosis revealed significantly increased frequency of mechanical ventilation (p = 0.02) and patient outcomes, such as superinfection, length of stay in the ICU and in-hospital mortality (p = 0.006).

Table 3.

Treatment and outcomes of both studied cohorts

Variable	No Thrombotic complications N (%)	Thrombotic complications N (%)	P-value
Treatment
Steroids	118 (95.9%)	39 (95.1%)	.559
Anticoagulation	119 (96.7%)	41 (100.0%)	.573
Convalescent Plasma	48 (39.0%)	20 (48.8%)	.180
Outcomes
Intensive care unit	3 (2.4%)	30 (73.2%)	.001
Mechanical ventilation	45 (36.6%)	25 (61.0%)	.010
Superinfection	34 (27.6%)	22 (53.7%)	.002
Acute kidney injury	2 (18.2%)	9 (22%)	.001
Bleeding	15 (12.2%)	3 (7.3%)	.566
Thrombosis			.001
None	123 (100.0%)	0 (0.0%)
Pulmonary embolism	0 (0.0%)	21 (51.2%)
Deep vein thrombosis	0 (0.0%)	8 (19.5%)
Arterial thrombosis	0 (0.0%)	12 (29.3%)
In-hospital mortality	32 (26.0%)	21 (51.2%)	.003

Regression analysis

The univariate logistic regression of nasopharyngeal viral load group, severity score, mechanical ventilation, pharmacotherapy including steroids, anticoagulation, and convalescent plasma revealed only increased odds for thrombosis in mechanically ventilated patients, OR 3.10 [1.37,7.03] (p = 0.007). Multivariate analysis revealed an adjusted OR of 3.25 [1.49,7.04] (p = 0.003) for mechanical ventilation. Additional multivariate analysis on in-hospital mortality revealed that mechanical ventilation was statistically significant with an adjusted OR of 38.83 [13.83–109.35] (p = 0.001).

Response Operator Curve analysis featuring viral load (Ct) and thrombosis events revealed a poor performance with an Area Under the Curve of 0.511, p = 0.835. (Fig. 1).

Fig. 1.

Response Operator Curve analysis featuring viral load (Ct) and thrombosis events

Discussion

To our knowledge, this study is the first to analyze the association between SARS-CoV-2 nasopharyngeal viral load and the development of VTE in COVID-19 patients. Despite being a common complication among hospitalized COVID-19 patients, the precise mechanisms of COVID-19 induced thrombosis are not fully understood. Several models have been proposed, including virus-induced endothelial damage through ACE2 receptor, a SARS-CoV-2 target receptor. In consequence, SARS-CoV-2 viral load were a priori a plausible predictor for VTE.

According to our data, SARS-CoV-2 nasopharyngeal viral load at admission, determined by Ct values, was not independently associated with VTE among hospitalized COVID-19 patients. In our population, VTE was equally frequent in patients with high, intermediate, and low viral loads. Furthermore, the mean viral load among patients with thromboembolic events was similar to the control group. Importantly, there were no significant differences in the baseline characteristics of the groups. Notably, none of the patients in the present study were vaccinated against COVID-19, since vaccination wasn’t available at that moment in Mexico”.

Prior studies have shown a significant association between SARS-CoV-2 nasopharyngeal viral load on admission and disease severity and mortality [16–18]. Pujadas et al. [17] demonstrated a significant difference in survival probability between those with high nasopharyngeal viral load and those with low nasopharyngeal viral load (p = 0.0003) using a univariate survival analysis. Moreover, high nasopharyngeal viral load has been independently associated with risk of intubation [19]. Other studies have found association between high nasopharyngeal viral load and increased risk of respiratory failure [20], ICU admission [21], acute kidney injury [22], liver injury [23], increased length of hospital stay and persistence of sequelae at six months [24]. Therefore, the report of quantitative SARS-CoV-2 nasopharyngeal viral load at admission has been proposed as well as its use as a prognostic factor to help in the stratification of patients.

There is no consensus towards the role of viral load in disease severity. Salto-Alejandre et al. [25] found in a multivariable analysis that the number of virus copies in the nasopharyngeal swabs was not independently associated with an unfavorable clinical result. Similarly, other studies did not find an association between initial nasopharyngeal viral load and clinical outcome, symptoms, oxygen requirements, ICU admission, and overall survival, although these studies are limited by the number of patients analyzed [26, 27]. Considering there is a well-documented correlation between time since symptoms onset and Ct values [28], some authors recognize Ct could be a confounder [25]. In this study, notably, the time since onset of symptoms was similar between patients with VTE and the control group, and we observed no correlation with Ct at admission.

Since the prognostic role of SARS-CoV-2 nasopharyngeal viral load in the disease severity and complications remains ambiguous, the use of other parameters as predictors of thrombosis needs to be explored. Male gender, delayed admission after illness onset, and invasive mechanical ventilation have been reported as predictors of prolonged viral RNA shedding (> 15 days), reflected by Ct value persistently < 40 [29]. Interestingly, in our study, mechanical ventilation was a significant risk factor for thrombosis; however, further studies are needed to demonstrate if there is a correlation between length of SARS-CoV-2 RNA shedding and thrombosis. In addition, an association between Ct value and hematological and biochemical markers has been reported [30, 31]. Future studies of our group are aim to study if there is an association between biochemical markers and nasopharyngeal viral loads.

The establishment of guidelines to prevent thromboembolic complications in hospitalized COVID-19 patients remains a challenge because of its high morbidity and mortality and the lack of a good predictor for risk stratification. In our study, patients admitted to the ICU were found to be at greater risk of developing thromboembolic events. This is in accordance with what is reported in the literature, the rates of VTE in COVID-19 patients admitted to the ICU and receiving prophylaxis are up to fourfold higher than the rates of VTE demonstrated in typical ICU patients [9]. As expected, in this study the patients with VTE had higher median D-dimer than the non-thrombosed patients. The use of D-dimer levels above certain cut-off has been suggested to guide the intensity of the antithrombotic therapy; however, its use should be cautious since D-dimer is commonly elevated in these patients.

Limitations in our study include its retrospective nature and the use of data documented in the medical record and thus could have misclassified patient characteristics or outcomes. Besides, it describes findings on an early population before massive vaccination. Furthermore, nasopharyngeal viral load quantification was only performed at a single time point through nasopharyngeal swabs, and we have no data on the dynamics according to the clinical outcomes. Additionally, patients were on multidrug treatments that included thrombotic prophylaxis that could have impacted the observed results. Studies from larger centers should provide further evidence by comparing thrombosis rates in patients admitted to ICU with and without COVID-19 to determine if it’s a risk modifier in this cohort.

Conclusion

In the present study, admission SARS-CoV-2 nasopharyngeal viral loads, as determined by Ct values, were not independently associated with thromboembolic complications among unvaccinated hospitalized patients with COVID-19. Further studies to delineate the underlying mechanisms and interactions between SARS-CoV-2 and the immune system leading to thrombosis would contribute to the prevention and management of thromboembolic complications in COVID-19 patients.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated during and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

None.