Abstract
Background
Cancer patients are at an elevated risk for SARS-CoV-2 infection and require vigilant monitoring. The efficacy and safety of Azvudine in treating COVID-19 among this vulnerable group remain under-researched.
Methods
We conducted a multicenter, retrospective cohort study of nine hospitals involving cancer patients with COVID-19 hospitalized from December 2022 to January 2023. To minimize immortal time bias, patients who died or were discharged within 24 h of drug administration were excluded. A 2:1 propensity score matching (PSM) was performed to balance baseline characteristics, explicitly adjusting for tumor metastasis, tumor classification, and COVID-19 vaccination status. The primary outcome was 30-day all-cause death, and the secondary outcome was composite disease progression. Robustness was verified using Probit regression and double robust estimation with Augmented Inverse Probability Weighting (AIPW).
Results
A total of 1,829 patients were included in the final matched cohort (1,200 in the control group and 629 in the Azvudine group). Azvudine treatment was significantly associated with a reduced risk of all-cause death (adjusted hazard ratio [HR]: 0.65, 95% CI: 0.452–0.939, P = 0.022). Furthermore, unlike unadjusted analyses, the multivariate model demonstrated a significant reduction in composite disease progression (adjusted HR: 0.70, 95% CI: 0.518–0.936, P = 0.016). These findings were consistent across all sensitivity analyses. Regarding safety, the Azvudine group exhibited a higher incidence of hypokalemia (P < 0.001) and Grade ≥3b platelet count decreased (P = 0.002). Significant fluctuations in lymphocyte counts (increased rates of both decrease and increase) were also observed.
Conclusion
Azvudine significantly reduced the risk of all-cause mortality and disease progression in hospitalized cancer patients with COVID-19. Although the drug was associated with increased risks of specific laboratory abnormalities (hypokalemia and thrombocytopenia), the substantial survival benefit outweighs these manageable risks. Routine clinical monitoring of electrolytes and blood counts is recommended during treatment.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12885-026-15800-1.
Keywords: Cancer patients, COVID-19, Azvudine, All-cause death, Composite disease progression
Introduction
In December 2019, the initial outbreak of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in Wuhan, China, precipitated the worldwide pandemic of Coronavirus disease 2019 (COVID-19). To date, the virus has infected over 750 million individuals and is responsible for approximately 5 million fatalities globally [1]. The pathogenesis of SARS-CoV-2 is intricately associated with the immune system, including overproduction of pro-inflammatory factors (such as IL-6, IFN-γ, and TNF-α) and anti-inflammatory factors (such as IL-4 and IL-10) [2, 3]. In severe patients, this delicate equilibrium is disrupted, leading to an uncontrolled release of cytokines known as a ‘cytokine storm’. This hyperinflammatory state accelerates the progression to systemic inflammatory conditions and can culminate in life-threatening multi-organ failure [4, 5].
The ongoing pandemic poses a significant risk to immunocompromised individuals, particularly cancer patients undergoing treatments like chemotherapy and immune checkpoint inhibitors, which can further weaken their immune defenses [6]. Although various COVID-19 vaccines offer some degree of protection, their real-world efficacy in cancer patients has been suboptimal. This has led to the consideration of oral antiviral medications as a strategy for early viral clearance to mitigate the risk of severe illness in this vulnerable group.
Azvudine, the first pioneering oral antiviral drug for COVID-19 in Henan, China, received conditional approval in July 2022 for the treatment of moderate COVID-19 patients in adults. Clinical trials have demonstrated Azvudine’s ability to effectively suppress SARS-CoV-2 replication and accelerate the conversion to negative nucleic acid status in patients with mild to moderate symptoms [7, 8]. Additionally, it appears to bolster immune responses by safeguarding the thymus and enhancing lymphocyte profiles [8]. Real-world studies indicate that Azvudine can reduce the duration to negative nucleic acid conversion, all-cause death, and composite disease progression [9, 10]. Nevertheless, there is scant data on its efficacy and safety of Azvudine in cancer patients with COVID-19.
This study aims to assess the clinical efficacy and safety of Azvudine in treating hospitalized cancer patients with COVID-19. By gathering data from nine hospitals across Henan Province, we have formed a multicenter, large-scale, real-world, retrospective cohort. The study compares the outcomes of cancer patients treated with Azvudine to those in a control group, thereby providing robust evidence to support its clinical use.
Methods
Study design and population
We conducted a multicenter, retrospective cohort study involving nine tertiary hospitals in Henan Province, China, from December 5, 2022, to January 31, 2023. The study design followed a similar framework to our previous investigation of Azvudine in elderly COVID-19 patients [11], but focused specifically on oncological populations. The protocol was approved by the Institutional Review Board of the First Affiliated Hospital of Zhengzhou University (Approval No. 2023-KY-0865-001). The study adhered to the STROBE guidelines and the ethical standards set forth in the 1975 Declaration of Helsinki.
Participant selection
Adult patients (≥ 18 years) with histologically confirmed malignant tumors and RT-PCR–confirmed SARS-CoV-2 infection were eligible. Exclusion criteria included: (1) receipt of antiviral agents other than Azvudine; (2) contraindications to Azvudine (e.g., severe hepatic or renal impairment, defined as eGFR < 30 mL/min/1.73 m² or Child-Pugh class C; (3) concomitant use of medications with narrow therapeutic windows primarily metabolized by CYP3A4 (e.g., alfentanil, quinidine); (4) pregnancy or lactation; (5) patients who were discharged or died within 24 h of drug administration. Treatment protocols followed the “COVID-19 Diagnosis and Treatment Plan (trial versions 9 or 10)” issued by the National Health Commission of the People’s Republic of China [12, 13].
Data sources
We collected electronic medical records of patients from nine hospitals in Henan Province. These records encompassed a comprehensive range of information, including demographic characteristics, hospital admission date, ICU admission, registered death, diagnoses, prescription and drug dispensing records, procedures, imaging data, and laboratory test results. The data collection spanned from the initiation of Azvudine administration to five half-lives post the final dose. In instances where multiple abnormalities were detected, the most severe results were chosen for analysis.
Treatment and outcomes
Cancer patients with COVID-19 eligible for the study were divided into two groups: the control group, which consisted of patients who received standard treatment without any antiviral agents, and the Azvudine group, where patients were administered 5 mg of Azvudine once daily for a maximum of 14 days. We observed patients from the index date until the date of outcome events or 30 days. The primary outcome of the study is all-cause death within 30 days of diagnosis. All-cause mortality refers to death from any causes that occurs in the study population during the study’s defined follow-up period. It records the patient’s survival status from enrollment to a predefined observation endpoint (discharge date, or day 30). The secondary outcome is composite disease progression, which is defined as progression to severe/critical illness or death. Disease severity was classified according to the “COVID-19 Diagnosis and Treatment Plan (trial version 10)”.
Safety
Safety was evaluated by the incidence of adverse events (AEs) overall, and specifically those AEs of grade 3 or higher. AEs were classified according to the Common Terminology Criteria for Adverse Events, Version 5.0 (CTCAE 5.0) [14]. The primary AEs monitored in this study were abnormal laboratory results. Monitoring commenced following drug administration and continued until five half-lives post the final dose. In instances where the severity of an AE varied, the highest grade observed was recorded.
Statistical analysis
All analyses were conducted using R version 4.0.3. A two-sided P value < 0.05 was considered statistically significant. To minimize selection bias, we performed 2:1 propensity score matching (PSM) without replacement, using a caliper width of 0.2 standard deviations of the logit of the propensity score. The propensity score was estimated via logistic regression incorporating all baseline variables listed in Table 1, including demographic characteristics (gender, age, BMI), the severity of COVID-19 at admission, vaccinated doses, use of systemic steroids, use of antibiotics, comorbidities, tumor classification, tumor metastasis and laboratory parameters. The balance between groups was assessed using standardized mean differences (SMD < 0.1 considered acceptable). The characteristics of baseline covariates were assessed using independent sample t-tests or Mann–Whitney U tests for continuous variables, and chi-square tests for categorical variables, in both unmatched and PSM scenarios. Missing baseline data were addressed through multiple imputation.
Table 1.
Baseline characteristics of cancer patients with COVID-19 before and after propensity score matching
| Characteristics | Before matching | After 2:1 matching | ||||
|---|---|---|---|---|---|---|
| Control (n=6366) | Azvudine (n=648) | P value | Control (n=1200) | Azvudine (n=629) | P value | |
| Age, mean (SD), year | 62.03 (12.50) | 65.86 (13.33) | <0.001 | 65.32 (13.06) | 65.56 (13.24) | 0.714 |
| Gender, n (%) | 0.219 | 1 | ||||
| Male | 3746 (58.8) | 398 (61.4) | 731 (60.9) | 383 (60.9) | ||
| Female | 2620 (41.2) | 250 (38.6) | 469 (39.1) | 246 (39.1) | ||
| BMI, mean (SD), kg/m2 | 23.29 (3.80) | 23.50 (3.79) | 0.192 | 23.57 (4.02) | 23.49 (3.78) | 0.66 |
| Vaccination doses, n (%) | <0.001 | 0.958 | ||||
| None | 1793 (28.2) | 251 (38.7) | 464 (38.7) | 239 (38.0) | ||
| One dose | 246 (3.9) | 29 (4.5) | 48 (4.0) | 28 (4.5) | ||
| Two doses | 1048 (16.5) | 92 (14.2) | 160 (13.3) | 88 (14.0) | ||
| Three doses | 3239 (50.9) | 271 (41.8) | 522 (43.5) | 269 (42.8) | ||
| Four doses | 39 (0.6) | 4 (0.6) | 5 (0.4) | 4 (0.6) | ||
| Five doses | 1 (0.0) | 1 (0.2) | 1 (0.1) | 1 (0.2) | ||
| Severity at admission, n (%) | <0.001 | 0.6 | ||||
| Mild | 812 (12.8) | 59 (9.1) | 128 (10.7) | 59 (9.4) | ||
| Moderate | 5003 (78.6) | 448 (69.1) | 847 (70.6) | 444 (70.6) | ||
| Severe | 551 (8.7) | 141 (21.8) | 225 (18.8) | 126 (20.0) | ||
| Concomitant systemic steroid, n (%) | <0.001 | 0.125 | ||||
| No | 4903 (77.0) | 397 (61.3) | 800 (66.7) | 396 (63.0) | ||
| Yes | 1463 (23.0) | 251 (38.7) | 400 (33.3) | 233 (37.0) | ||
| Antibiotics, n (%) | <0.001 | 0.773 | ||||
| No | 4894 (76.9) | 325 (50.2) | 630 (52.5) | 325 (51.7) | ||
| Yes | 1472 (23.1) | 323 (49.8) | 570 (47.5) | 304 (48.3) | ||
| Comorbidities, n (%) | ||||||
| Diabetes | 844 (13.3) | 105 (16.2) | 0.043 | 190 (15.8) | 102 (16.2) | 0.885 |
| Hypertension | 1584 (24.9) | 218 (33.6) | <0.001 | 392 (32.7) | 206 (32.8) | 1 |
| Liver diseases | 913 (14.3) | 88 (13.6) | 0.639 | 186 (15.5) | 88 (14.0) | 0.429 |
| Cardio-cerebral diseases | 1245 (19.6) | 144 (22.2) | 0.116 | 282 (23.5) | 142 (22.6) | 0.699 |
| Kidney diseases | 438 (6.9) | 154 (23.8) | <0.001 | 232 (19.3) | 136 (21.6) | 0.272 |
| Chronic respiratory diseases | 1547 (24.3) | 210 (32.4) | <0.001 | 368 (30.7) | 201 (32.0) | 0.608 |
| Autoimmune diseases | 102 (1.6) | 15 (2.3) | 0.235 | 28 (2.3) | 14 (2.2) | 1 |
| Tumor classification, n (%) | <0.001 | 0.925 | ||||
| Digestive System Tumors | 1462 (23.0) | 135 (20.8) | 252 (21.0) | 131 (20.8) | ||
| Genitourinary System Tumors | 631 (9.9) | 80 (12.3) | 154 (12.8) | 76 (12.1) | ||
| Respiratory System Tumors | 1471 (23.1) | 145 (22.4) | 257 (21.4) | 141 (22.4) | ||
| Head and Neck Tumors | 254 (4.0) | 20 (3.1) | 33 (2.8) | 19 (3.0) | ||
| Hematologic and Lymphatic System Tumors | 495 (7.8) | 104 (16.0) | 173 (14.4) | 99 (15.7) | ||
| Multiple Sites or Other Tumors | 2053 (32.2) | 164 (25.3) | 331 (27.6) | 163 (25.9) | ||
| Tumor metastasis, n (%) | <0.001 | 0.946 | ||||
| No | 4747 (74.6) | 540 (83.3) | 991 (82.6) | 521 (82.8) | ||
| Yes | 1619 (25.4) | 108 (16.7) | 209 (17.4) | 108 (17.2) | ||
| Laboratory parameters, mean (SD) | ||||||
| Neutrophil, ×109/L | 4.77 (4.41) | 5.67 (8.96) | <0.001 | 5.40 (5.51) | 5.35 (4.72) | 0.868 |
| Lymphocyte, ×109/L | 1.40 (3.89) | 1.43 (3.61) | 0.847 | 1.34 (2.86) | 1.40 (3.58) | 0.687 |
| Glucose, mmol/L | 6.10 (2.77) | 6.59 (2.75) | <0.001 | 6.56 (3.36) | 6.54 (2.72) | 0.923 |
| Alanine aminotransferase, IU/L | 28.25 (52.71) | 31.86 (49.94) | 0.095 | 34.12 (86.07) | 31.97 (50.53) | 0.563 |
| Aspartate aminotransferase, IU/L | 36.90 (110.69) | 39.26 (105.84) | 0.605 | 41.71 (124.60) | 39.27 (107.39) | 0.677 |
| Alkaline phosphatase, IU/L | 108.63 (123.27) | 101.68 (93.00) | 0.163 | 102.88 (101.25) | 102.03 (94.30) | 0.862 |
| Gamma-glutamyl transpeptidase, IU/L | 66.46 (129.36) | 67.93 (126.34) | 0.782 | 71.13 (132.24) | 68.30 (127.91) | 0.66 |
| Albumin, g/L | 38.41 (19.63) | 36.69 (13.23) | 0.029 | 36.93 (12.55) | 36.79 (13.31) | 0.826 |
| Total bilirubin, μmol/L | 15.15 (35.39) | 12.32 (16.77) | 0.045 | 13.47 (18.99) | 12.34 (17.00) | 0.207 |
| Creatine, μmol/L | 91.66 (193.12) | 83.19 (96.13) | 0.27 | 83.05 (93.24) | 83.52 (97.49) | 0.919 |
| Glomerular filtration rate, ml/min | 137.17 (190.99) | 133.40 (185.31) | 0.631 | 135.42 (193.55) | 131.38 (178.67) | 0.664 |
| Cholesterol, mmol/L | 4.25 (2.17) | 4.16 (2.01) | 0.305 | 4.19 (1.64) | 4.16 (2.03) | 0.808 |
| Triglyceride, mmol/L | 1.57 (2.98) | 1.52 (1.99) | 0.664 | 1.49 (1.83) | 1.52 (2.01) | 0.711 |
| High-density lipoprotein, mmol/L | 1.09 (1.11) | 1.13 (1.85) | 0.428 | 1.10 (1.69) | 1.13 (1.88) | 0.71 |
| Low-density lipoprotein, mmol/L | 2.47 (1.40) | 2.57 (2.40) | 0.136 | 2.53 (2.40) | 2.57 (2.43) | 0.746 |
| Prothrombin time, s | 13.02 (5.15) | 16.21 (9.04) | <0.001 | 15.19 (7.94) | 15.77 (8.63) | 0.149 |
| Activated partial thromboplastin time, s | 29.73 (7.19) | 27.78 (9.13) | <0.001 | 28.38 (8.12) | 28.11 (8.97) | 0.517 |
| C-reactive protein, mg/L | 44.89 (59.96) | 56.39 (68.80) | <0.001 | 57.66 (70.96) | 56.12 (69.35) | 0.658 |
| Procalcitonin, ng/ml | 1.42 (8.21) | 1.26 (6.39) | 0.63 | 1.26 (5.72) | 1.29 (6.48) | 0.934 |
Time-to-event analyses were conducted using Kaplan-Meier curves and log-rank tests. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using Cox proportional hazards models, adjusted for all covariates retained after matching. The proportional hazards (PH) assumption was verified using Schoenfeld residuals. A time-dependent covariate-based non-proportional hazards Cox regression model was formulated when the PH assumption was not met.
Prespecified subgroup analyses stratified by baseline covariates were performed to validate the consistency of results across population subsets. Missing data were handled using Multiple Imputation by Chained Equations (MICE), generating 5 imputed datasets.Sensitivity analyses included alternative matching using a probit model. To further validate the robustness of the causal inference and minimize bias from model misspecification, we performed a double robust estimation using the Augmented Inverse Probability Weighting (AIPW) method. This approach combines a model for the treatment assignment (propensity score) with a model for the outcome, ensuring unbiased estimation of the treatment effect even if one of the models is misspecified [cite reference if needed]. The results were visualized using a forest plot to compare estimates across different statistical methodologies.
Results
Baseline characteristics
In the study, a total of 32,864 hospitalized patients diagnosed with COVID-19 were collected from nine hospitals in Henan province. Of these, 7,014 cancer patients met the inclusion and exclusion criteria and were allocated into two groups: 6366 in the control group and 648 in the Azvudine group. Subsequently, we used 2:1 PSM to control for confounding variables, resulting in 1200 patients in the control group and 629 patients in the Azvudine group (Fig. 1).
Fig. 1.
The flowchart of study design
Table 1 presents the baseline characteristics of patients included before and after PSM. Prior to matching, compared with the control group, patients in the Azvudine group were older, more severe on admission, and had more comorbidities of diabetes, hypertension, and kidney disease. Most patients were treated with concomitant systemic steroids. Significant differences were observed in several laboratory parameters, including Neut, GLU, ALB, TBIL, HDL, PT, APTT, and CRP. After matching, the characteristics between the two groups were statistically balanced (all P > 0.05; Fig. S1).
All-cause death and composite disease progression
All-cause death was the primary outcome. During the follow-up period, 47 deaths occurred in the Azvudine group compared with 104 in the control group. The Kaplan-Meier method and log-rank test revealed a significantly lower risk of all-cause death in the Azvudine group compared with the control group (log-rank P = 0.01) (Fig. 2A). After multivariate adjustment, Azvudine was associated with a 35% lower risk of all-cause death compared to controls (HR: 0.65, 95% CI: 0.452–0.939, P = 0.022) (Fig. 3).
Fig. 2.
Kaplan–Meier curves of control and Azvudine groups. A Cumulative hazard of all‐cause death and B composite disease progression
Fig. 3.
Multivariate Cox proportional hazards regression analysis of all-cause death and composite disease progression in control and Azvudine groups. Adjusted for all baseline covariates in Table 1. HR: Hazard Ratio; 95%CI: 95% confidence interval
The secondary outcome was composite disease progression. After the 30-day follow-up period, a total of 146 composite disease progression events occurred in the control group and 80 in the Azvudine group. According to the Kaplan-Meier analysis, the cumulative incidence of disease progression was lower in the Azvudine group, showing a trend toward significance (Log-rank P = 0.064) (Fig. 2B). However, in the multivariate adjusted Cox proportional hazards model, Azvudine treatment was significantly associated with a reduced risk of composite disease progression (HR: 0.70, 95% CI: 0.518–0.936, P = 0.016) (Fig. 3).
Subgroup analysis
We performed prespecified subgroup analyses stratified by key clinical covariates to assess heterogeneity in treatment effects on all-cause mortality and composite disease progression outcomes in cancer COVID-19 patients treated with Azvudine, using formal tests with interaction P values. (Fig. 4).
Fig. 4.
Subgroup analyses for the effectiveness of Azvudine in reducing the risk of (A) all-cause death and (B) composite disease progression in cancer patients with COVID-19
The results showed that patients older than 60 years had lower all-cause death than those younger than or equal to 60 years (P value of interaction = 0.031). There were no significant subgroup differences in all-cause death associated with age, hospital admission severity, vaccination dose, antibiotic use, steroid hormone use, diabetes, liver disease, heart and brain disease, kidney disease, chronic respiratory disease, autoimmune disease, tumor classification, and tumor metastasis (all P values for the interaction were > 0. 05). Furthermore, there were no significant differences in the composite disease progression results between the Azvudine and control groups. Overall, these results suggest that Azvudine may provide enhanced protection when used in elderly patients older than 60 years, which is consistent with a previous study [11].
Sensitivity analyses
To verify the robustness of our findings, we performed a sensitivity analysis using a Probit regression model for propensity score estimation (Table S1). The Kaplan-Meier analysis in the sensitivity cohort demonstrated that the Azvudine group had a significantly lower cumulative hazard of all-cause death compared to the control group (Log-rank P = 0.0024, Figure S2A). regarding composite disease progression, the Kaplan-Meier curves showed a strong trend toward risk reduction in the Azvudine group (Log-rank P = 0.051, Figure S2B).
Crucially, the multivariate Cox proportional hazards regression models confirmed the consistency of the results. Azvudine treatment remained significantly associated with a reduced risk of all-cause death (adjusted HR: 0.62, 95% CI: 0.436–0.894, P = 0.01). Furthermore, contrary to the unadjusted analysis, a significant reduction in the risk of composite disease progression was observed after multivariate adjustment (adjusted HR: 0.74, 95% CI: 0.550–0.990, P = 0.043) (Figure S3). These findings align with the primary analysis, reinforcing the validity of our conclusions.
Moreover, to further substantiate these findings and address potential confounding, we employed doubly robust estimation. The doubly robust analysis maintained the HR for all-cause death at 0.67 (P = 0.033) and composite disease progression at 0.70 (P = 0.019) (Figure S4). Collectively, these analyses robustly confirm the association between Azvudine and reduced all-cause mortality, while indicating that the effect on the composite progression endpoint is less consistent across statistical models.
Safety analysis
The safety profile of Azvudine was evaluated by analyzing the incidence of adverse events in cancer patients with COVID-19. The Azvudine group exhibited a distinct safety profile characterized by a higher incidence of specific laboratory abnormalities compared with the control group (Table 2). Regarding hematological parameters, the analysis revealed a complex impact on lymphocyte dynamics. Azvudine treatment was associated with a significantly higher incidence of lymphocyte count decreased (All grades: P < 0.001; Grade ≥ 3b: P < 0.001). Conversely, the incidence of lymphocyte count increased was also significantly higher in the Azvudine group (All grades: P = 0.004; Grade ≥ 3b: P = 0.038), suggesting a potential drug-induced immune mobilization or turnover. Furthermore, while the overall incidence of platelet count decrease was similar between the two groups (P = 0.2), the risk of severe thrombocytopenia (Grade ≥ 3b platelet count decreased) was significantly elevated in the Azvudine group (9.3% vs. 5.3%, P = 0.002).
Table 2.
Incidence of adverse event of the study population receiving normal or Azvudine treatment
| Adverse events (n, %) | Available dataa | All grades | Grade ≥ 3b | |||||
|---|---|---|---|---|---|---|---|---|
| Control | Azvudine | Control | Azvudine | P value | Control | Azvudine | P value | |
| Lymphocyte count decreased | 5609 | 593 | 1548 (27.6%) | 220 (36.9%) | < 0.001 | 659 (12%) | 115 (19%) | < 0.001 |
| Lymphocyte count increased | 5609 | 593 | 101 (1.8%) | 21 (3.5%) | 0.004 | 10 (0.2%) | 4 (0.7%) | 0.038 |
| Neutrophil count increased | 1712 | 283 | 209 (12.2%) | 45 (15.9%) | 0.084 | 77 (4.5%) | 19 (6.7%) | 0.11 |
| PLT count decreased | 2158 | 408 | 282 (13.1%) | 64 (15.7%) | 0.2 | 115 (5.3%) | 38 (9.3%) | 0.002 |
| Anemia | 1779 | 288 | 641 (35.7%) | 119 (42%) | 0.084 | 173 (9.7%) | 34 (12%) | 0.3 |
| Hypophosphatemia | 1116 | 149 | 260 (23%) | 28 (19%) | 0.2 | 0 | 0 | |
| Hypokalemia | 3641 | 456 | 892 (24.4%) | 193 (43%) | < 0.001 | 190 (5.2%) | 51 (11%) | < 0.001 |
| Hyperkalemia | 3641 | 456 | 173 (4.7%) | 29 (6.4%) | 0.3 | 5 (0.1%) | 1 (0.2%) | 0.5 |
| ALT increased | 1779 | 353 | 256 (14.7%) | 74 (21%) | 0.002 | 28 (1.6%) | 9 (2.5%) | 0.2 |
| AST increased | 1829 | 366 | 299 (16.2%) | 79 (21.7%) | 0.015 | 30 (1.6%) | 12 (3.3%) | 0.037 |
| ALP increased | 1767 | 348 | 135 (7.6%) | 34 (9.8%) | 0.2 | 2 (0.1%) | 0 (0%) | > 0.9 |
| GGT increased | 1469 | 254 | 294 (19.9%) | 49 (18.9%) | 0.8 | 48 (3.3%) | 4 (1.6%) | 0.15 |
| Hyperuricemia | 1564 | 307 | 57 (3.6%) | 11 (3.6%) | > 0.9 | 0 | 0 | |
| CREA increased | 1785 | 367 | 92 (5.2%) | 20 (5.5%) | 0.8 | 19 (1.1%) | 5 (1.4%) | 0.6 |
| Hypoglycemia | 418 | 115 | 71 (17%) | 23 (20%) | 0.5 | 0 | 0 | |
| Hypercholesterolemia | 293 | 53 | 14 (4.8%) | 4 (7.5%) | 0.5 | 0 | 0 | |
| Hypertriglyceridemia | 203 | 34 | 34 (7%) | 8 (23.8%) | 0.3 | 4 (2.0%) | 2 (5.9%) | 0.2 |
Abbreviations: PLT platelets, Hb hemoglobin, ALT alanine aminotransferase, AST aspartate aminotransferase, ALP alkaline phosphatase, UA uric acid, CREA creatinine, Glu glucose, CH Cholesterol, TG Triglyceride
aNumber of people who completed the follow-up of data collection for this indicator
bSeverity grades were defined according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 5.0
Significant differences were also observed in electrolyte and liver function indicators. The Azvudine group showed a markedly higher susceptibility to hypokalemia, with significant increases in both all-grade (43% vs. 24.4%, P < 0.001) and Grade ≥ 3b events (11% vs. 5.2%, P < 0.001). In terms of liver function, Azvudine treatment was associated with hepatotoxicity signals. The incidence of all-grade ALT elevation was higher in the Azvudine group (21% vs. 14.7%, P = 0.002), although Grade ≥ 3b ALT elevation did not differ significantly (P = 0.2). However, AST levels were more profoundly affected, with significant increases observed in both all-grade (21.7% vs. 16.2%, P = 0.015) and Grade ≥ 3b elevations (P = 0.038) in the Azvudine group.
Dynamic changes of leukocyte subsets during treatment
We further investigated the longitudinal dynamics of leukocyte subsets in both groups over the 15-day drug administration period to evaluate hematological safety. At baseline, the immunological profiles exhibited significant heterogeneity; specifically, lymphocyte counts in the Azvudine group were significantly lower compared to the control group, which showed levels generally above the normal range (Fig. 5A). However, following treatment initiation, the dynamic trajectories of lymphocyte count in the Azvudine group showed a recovery trend and converged with those of the control group. No statistically significant inter-group differences were observed at post-treatment time points, indicating that Azvudine did not impede immune reconstitution. Regarding other subsets, the baseline levels of total leukocytes, monocytes, basophils, and eosinophils were generally below the normal range in both cohorts (Fig. 5B–E). Importantly, throughout the observation period, the temporal trends of these cells, as well as neutrophil counts (Fig. 5F), remained comparable between the Azvudine and control groups, with no significant drug-induced suppression or abnormalities observed. These longitudinal data suggest that Azvudine maintains a favorable hematological safety profile in cancer patients with COVID-19.
Fig. 5.
Dynamic changes of leukocyte subsets during treatment. After treatment, the dynamic levels of (A) lymphocytes (B) leukocytes (C) monocytes (D) basophils (E) eosinophils (F) neutrophils in the Azvudine and control groups
Discussion
Following the adjustment of national COVID-19 policies in December 2022, China witnessed a significant increase in SARS-CoV-2 infections. Cancer patients, particularly vulnerable due to an immunocompromised state from malignancies and treatments like chemotherapy, radiotherapy, immunotherapy, or surgery, face a markedly increased risk of contracting SARS-CoV-2 infection and developing severe disease compared with the general population [2, 6]. Therefore, antiviral therapy is critical for hospitalized cancer patients with COVID-19, warranting heightened attention from clinicians. Nimalrevir/ritonavir and Azvudine are the recommended antivirals in the “COVID-19 Diagnostic and Therapeutic Regimen (Trial 9th or 10th Edition)”. However, due to the high price and limited availability of nimalrevir/ritonavir, an increasing number of patients are being treated with Azvudine. Currently, there are limited studies on the association of Azvudine with COVID-19 in cancer patients. Preliminary clinical trials have shown that Azvudine may inhibit SARS-CoV-2 replication, enhance immune function, and shorten the time to nucleic acid conversion in mild and moderate patients [15]. Kaican Zong et al. have also noted that a reduced risk of in-hospital mortality in overall COVID-19 patients treated with Azvudine [16]. Given the limitations of existing studies, such as small cohorts and single center, there is an imperative need for multicenter, large-scale, real-world studies to investigate the efficacy and safety of Azvudine in cancer patients with COVID-19, thereby informing in clinical practice and decision-making.
Our multicenter, real-world study provides robust evidence regarding the efficacy and safety of Azvudine in hospitalized cancer patients with COVID-19. By applying rigorous statistical adjustments for oncological characteristics (tumor metastasis and classification) and excluding early-death outliers, we demonstrated that Azvudine treatment was significantly associated with a 35% reduction in the risk of all-cause mortality and a 30% reduction in composite disease progression. These findings were consistently supported by sensitivity analyses, including the Probit model and double robust estimation (AIPW), confirming that early antiviral intervention can effectively interrupt the disease trajectory in this vulnerable population.
SARS-CoV-2 is a self-limiting disease that, unlike viruses such as hepatitis B, hepatitis C, and HIV, is fatal not only because of the virus itself but also because of the excessive inflammatory response and a range of inflammatory factors it causes [17]. Previous studies have established the efficacy of oral antivirals, such as Nirmatrelvir/Ritonavir and Molnupiravir, in preventing severe COVID-19 [18, 19]. However, data specifically focusing on cancer patients—who often face contraindications due to drug-drug interactions with chemotherapeutic agents—remain limited. Our results extend the current knowledge by identifying Azvudine as a viable and effective alternative. Mechanistically, Azvudine not only inhibits SARS-CoV-2 replication but has also been suggested to preserve thymus function and promote immune recovery [7, 8]. Additionally, in vitro and in vivo experiments have confirmed that Azvudine has significant tumor growth inhibitory activity [20]. This dual mechanism is particularly beneficial for cancer patients, who are often in a state of immunocompromise. The significant reduction in disease progression observed in our study suggests that Azvudine effectively lowers the viral load, thereby preventing the cascade of inflammation that leads to critical illness and death. Azvudine reduces viral load, restores and promotes thymic function, improves lymphocyte profiles, and mitigates inflammation and organ damage.
The surveillance of adverse drug reactions is crucial for mitigating medication risks and safeguarding public health. There is little data on the safety of Azvudine in the treatment of COVID-19, and one previous study has linked Azvudine with side effects such as headache, dizziness, elevated aminotransferases, nausea, and elevated D-dimer levels [21]. Our analysis in this specific cancer cohort revealed a distinct toxicity profile that warrants clinical attention. Unlike studies in the general population which reported a benign safety profile, we observed a higher incidence of specific laboratory abnormalities in the Azvudine group [22]. First, electrolyte disturbance: A significant increase in hypokalemia (including Grade ≥ 3 events) was identified. This may be attributed to the gastrointestinal side effects of the drug overlapping with cancer-related malnutrition or cachexia. Second, hematological impact: We observed a higher risk of Grade ≥ 3 thrombocytopenia. Given that cancer patients often have limited bone marrow reserve due to prior chemotherapy, they appear more sensitive to the potential myelosuppressive effects of nucleoside analogs. Third, lymphocyte dynamics: Interestingly, we observed a “dual effect” on lymphocytes—significant increases in both lymphocyte reduction and elevation. This fluctuation likely reflects a drug-induced acceleration of immune cell turnover or mobilization rather than simple suppression. Fourth, hepatotoxicity: While mild elevations in liver enzymes were common, and Grade ≥3b AST elevation was statistically higher, severe liver failure was rare. These findings indicate that while Azvudine provides a survival advantage, it is not without risks. We strongly recommend proactive monitoring of serum potassium, platelet counts, and liver function during treatment, with prompt supportive care (e.g., potassium supplementation) to manage these manageable adverse events.
This study has several limitations. First, as a retrospective observational study, residual confounding cannot be entirely eliminated. Specifically, detailed granular data on ongoing anti-cancer regimens (e.g., specific chemotherapy lines or immunotherapy agents) were incomplete due to the emergency context of data collection. However, we incorporated tumor metastasis and tumor classification into our propensity score model as robust proxies for advanced disease severity and active treatment needs, which significantly strengthened the validity of our comparisons. Second, the study was conducted during the Omicron wave, and the results may need validation for future variants. Third, the safety analysis relied on retrospective electronic records, which might underestimate mild, undocumented adverse events.
In conclusion, Azvudine significantly improved the survival outcomes and reduced disease progression in cancer patients with COVID-19. Although the drug was associated with increased risks of hypokalemia, thrombocytopenia, and AST elevation in this fragile population, the survival benefit outweighs these manageable risks. Clinicians should prioritize Azvudine use in high-risk cancer patients while ensuring strict monitoring of hematological and biochemical indicators to maximize patient safety.
Supplementary Information
Acknowledgements
We thank all participating hospitals, including the First Affiliated Hospital of Zhengzhou University, Henan Provincial Chest Hospital, Henan Infectious Disease Hospital, Luoyang Central Hospital, Nanyang Central Hospital, the Fifth People’s Hospital of Anyang, Shangqiu Municipal Hospital, Guangshan County People’s Hospital, and Fengqiu County People’s Hospital. We thank all the generous volunteer subjects who participated in the study. We sincerely thank all the doctors, nurses and medical technicians who bravely fight the COVID-19 pandemic.
AI technology
We have not used AI Technology to develop any portion of the manuscript.
Abbreviations
- BMI
Body mass index
- Neut
Neutrophils
- Lymph
Lymphocytes
- Glu
Glucose
- ALT
Alanine aminotransferase
- AST
Aspartate aminotransferase
- ALP
Alkaline phosphatase
- GGT
Gamma-glutamyl transpeptidase
- ALB
Albumin
- TBIL
Total bilirubin
- CREA
Creatinine
- e-GFR
Glomerular filtration rate
- TG
Triglycerides
- CH
Cholesterol
- HDL
High-density lipoprotein
- LDL
Low-density lipoprotein
- PT
Prothrombin time
- APTT
Activated partial thromboplastin time
- CRP
C-reactive protein
- PCT
Procalcitonin
Authors’ contributions
Lei Li and Wenjie Liu contributed equally to this work. Zujiang Yu and Zhigang Ren conceived and designed the study; Zhigang Ren, Wenjie Liu, Yongjian Zhou, Ling Wang, Juan Wang, Guotao Li, Guowu Qian, Shixi Zhang, Hong Luo, Silin Li and Donghua Zhang managed the patients; Lei Li, Wenjie Liu, Mengzhao Yang, Ling Wang and Guowu Qian collected the data; Lei Li and Mengzhao Yang analyzed the data; Lei Li wrote the manuscript; All authors reviewed and approved the manuscript.
Funding
This work was supported by the National Key Research and Development Program of China (2023YFC3043514 to Z.Y. and 2022YFC2303100 to Z.R.), Central Plains Talent Program-Central Plains Youth Top Talents to Z.R., Young and Middle-aged Academic Leaders of Henan Provincial Health Commission (HNSWJW-2022013 to Z.R.), Scientific and Technological Innovation Talents of Universities in Henan Province (24HASTIT063 to Z.R.), the Scientific Research and Innovation Team of The First Affiliated Hospital of Zhengzhou University (ZYCXTD2023002 to Z.Y. and QNCXTD2023002 to Z.R.).
Data availability
The data that support the findings of this study are from nine hospitals but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the corresponding author.
Declarations
Ethics approval and consent to participate
The study followed the Helsinki Declaration and Rules of Good Clinical Practice. This study was approved by the Institutional Review Board of the First Affiliated Hospital of Zhengzhou University (Approval No. 2023-KY-0865-001). The need for informed consent to participate was waived by the Institutional Review Board of the First Affiliated Hospital of Zhengzhou University due to the retrospective nature of the study and the use of anonymized clinical data.
Consent for publication
Not applicable. This retrospective cohort study employed anonymized data and did not require individual informed consent for publication.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Lei Li and Wenjie Liu contributed equally to this work.
Contributor Information
Zujiang Yu, Email: johnyuem@zzu.edu.cn.
Zhigang Ren, Email: fccrenzg@zzu.edu.cn.
References
- 1.WHO COVID-19 dashboard. https://data.who.int/dashboards/covid19/cases?n=c.
- 2.Wekking D, Senevirathne TH, Pearce JL, Aiello M, Scartozzi M, Lambertini M, De Silva P, Solinas C. The impact of COVID-19 on cancer patients. Cytokine Growth Factor Rev. 2024;75:110–8. [DOI] [PubMed] [Google Scholar]
- 3.Hu B, Huang S, Yin L. The cytokine storm and COVID-19. J Med Virol. 2021;93(1):250–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fajgenbaum DC, June CH. Cytokine Storm. N Engl J Med. 2020;383(23):2255–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Nile SH, Nile A, Qiu J, Li L, Jia X, Kai G. COVID-19: pathogenesis, cytokine storm and therapeutic potential of interferons. Cytokine Growth Factor Rev. 2020;53:66–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sobhani N, Mondani G, Roviello G, Catalano M, Sirico M, D’Angelo A, Scaggiante B, Generali D. Cancer management during the COVID-19 world pandemic. Cancer Immunol Immunother. 2023;72(11):3427–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ren Z, Luo H, Yu Z, Song J, Liang L, Wang L, Wang H, Cui G, Liu Y, Wang J, et al. A Randomized, Open-Label, Controlled Clinical Trial of Azvudine Tablets in the Treatment of Mild and Common COVID-19, a Pilot Study. Adv Sci (Weinh). 2020;7(19):e2001435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhang JL, Li YH, Wang LL, Liu HQ, Lu SY, Liu Y, Li K, Liu B, Li SY, Shao FM, et al. Azvudine is a thymus-homing anti-SARS-CoV-2 drug effective in treating COVID-19 patients. Signal Transduct Target Ther. 2021;6(1):414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Li F, Zheng K, Qi X, Cui K, Yang J, Hao Z. A retrospective analysis of Azvudine in patients with COVID-19 and pre-existing cancer. J Cancer. 2024;15(8):2442–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yang H, Wang Z, Jiang C, Zhang Y, Zhang Y, Xu M, Zhang Y, Wang Y, Liu X, An Z, et al. Oral azvudine for mild-to-moderate COVID-19 in high risk, nonhospitalized adults: Results of a real-world study. J Med Virol. 2023;95(7):e28947. [DOI] [PubMed] [Google Scholar]
- 11.Sun R, Wang H, Sun J, Yang M, Zhang S, Hu X, Yu B, Song Z, Han N, Luo H, et al. Effectiveness and Safety of Oral Azvudine for Elderly Hospitalized Patients With COVID-19: A Multicenter, Retrospective, Real-World Study. Adv Sci (Weinh). 2025;12(13):e2404450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.China,NHCotPsRo. COVID-19 diagnosis and treatment plan (trial version 9). 2022. https://www.gov.cn/zhengce/zhengceku/2022-03/15/5679257/files/49854a49c7004f4ea9e622f3f2c568d8.pdf.
- 13.China,NHCotPsRo. COVID-19 diagnosis and treatment plan (trial version 10). 2023. https://www.gov.cn/zhengce/zhengceku/2023-01/06/5735343/files/5844ce04246b431dbd322d8ba10afb48.pdf.
- 14.CommonTerminology Criteria for Adverse Events (CTCAE). Version 5.0. https://dctd.cancer.gov/research/ctep-trials/for-sites/adverse-events/ctcae-v5-5x7.pdf.
- 15.Zhu KW. Efficacy and safety evaluation of Azvudine in the prospective treatment of COVID-19 based on four phase III clinical trials. Front Pharmacol. 2023;14:1228548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zong K, Zhou H, Li W, Jiang E, Liu Y, Li S. Azvudine reduces the in-hospital mortality of COVID-19 patients: a retrospective cohort study. Acta Pharm Sin B. 2023;13(11):4655–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lai CC, Shih TP, Ko WC, Tang HJ, Hsueh PR. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): The epidemic and the challenges. Int J Antimicrob Agents. 2020;55(3):105924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zheng B, Tazare J, Nab L, Green AC, Curtis HJ, Mahalingasivam V, Herrett EL, Costello RE, Eggo RM, Speed V, et al. Comparative effectiveness of nirmatrelvir/ritonavir versus sotrovimab and molnupiravir for preventing severe COVID-19 outcomes in non-hospitalised high-risk patients during Omicron waves: observational cohort study using the OpenSAFELY platform. Lancet Reg Health Eur. 2023;34:100741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jayk Bernal A, Gomes da Silva MM, Musungaie DB, Kovalchuk E, Gonzalez A, Delos Reyes V, Martín-Quirós A, Caraco Y, Williams-Diaz A, Brown ML, et al. Molnupiravir for Oral Treatment of Covid-19 in Nonhospitalized Patients. N Engl J Med. 2022;386(6):509–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang Q, Liu X, Wang Q, Zhang Y, Jiang J, Guo X, Fan Q, Zheng L, Yu X, Wang N, et al. FNC, a novel nucleoside analogue inhibits cell proliferation and tumor growth in a variety of human cancer cells. Biochem Pharmacol. 2011;81(7):848–55. [DOI] [PubMed] [Google Scholar]
- 21.da Silva RM, Gebe Abreu Cabral P, de Souza SB, Arruda RF, Cabral SPF, de Assis A, Martins YPM, Tavares CAA, Viana Junior AB, Chang J, et al. Serial viral load analysis by DDPCR to evaluate FNC efficacy and safety in the treatment of mild cases of COVID-19. Front Med (Lausanne). 2023;10:1143485. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ren Z, Yang M, Su G, Qian G, Yuan Y, Yu J, Li S, Wang C, Lu M, Luo H, et al. Real-world effectiveness and safety of azvudine in hospitalized patients with SARS-CoV-2 infection: A multicenter, retrospective cohort study. J Infect. 2024;89(6):106355. [DOI] [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
The data that support the findings of this study are from nine hospitals but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the corresponding author.