Efficacy and mechanisms of traditional Chinese medicine for COVID-19: a systematic review

Randomized controlled trials (RCTs), cohort studies (CSs), and case–control studies (CCSs).

Participants were patients with a clear diagnosis of COVID-19, aged ≥ 18 years, regardless of gender and clinical types. We grouped patients according to their clinical types. The classification criteria of clinical types of COVID-19 patients referred to Diagnosis and Treatment Protocol for COVID-19 Patients (Tentative 8th Edition) [8]. The clinical symptoms of mild type patients were mild, and there was no evidence of pneumonia on imaging. Moderate type patients presented fever and respiratory symptoms, and chest radiology suggested pneumonia. Adult COVID-19 patients meeting any of the following were regarded as severe type. (1) Tachypnea, respiratory rate ≥ 30 breaths/min; (2) At rest, oxygen saturation ≤ 93% during air suction; (3) Partial pressure of oxygen (PaO2)/fraction of inspired oxygen (FiO2) ≤ 300 mmHg (1 mmHg = 0.133 kPa); PaO2/FiO2 should be corrected by the following formula in high altitude areas (altitude more than 1000 m): PaO2/FiO2 × [760/atmospheric pressure (mmHg)]. (4) Clinical symptoms were gradually worsening, and lung imaging indicated lesion significantly progressed > 50% within 24 ~ 48 h. COVID-19 patients meeting one of the following conditions were classified into critical type. (1) Respiratory failure demanding mechanical ventilation; (2) Shock; (3) Combined with other organ failure and demanded intensive monitoring and treatment.

In this review, included patients were subsumed into four groups according to the clinical types of COVID-19: mild or moderate types into mild group, severe or critical types into severe group, convalescent type into convalescent group, and unknown type or containing two or more of the above three groups into mixed group.

Patients in the intervention/TCM group were treated with TCM or a combination of TCM and conventional Western medicine (CWM). We did not restrict the dosage form of used TCM. Patients in the control/CWM group were treated with CWM (e.g., antiviral treatment, nutritional support, anti-infection treatment).

Primary outcomes

(1) Proportion of patients progressing to severe cases.

(2) Mortality rate of severe or critical patients.

Because the outcomes of patients were closely related to clinical types, we used the proportion of patients progressing to severe cases as the primary outcome of mild or moderate patients, and the mortality rate as the primary outcome of severe or critical patients.

Secondary outcomes

(1) Total effective rate. We defined the total effective rate as the proportion of the total number of patients whose clinical symptoms improve ≥ 30%.

(2) Clinical cure rate. We defined the clinical cure of COVID-19 patients as achieving all of the following conditions: no fever > 3 days; significant reduction of respiratory symptoms; chest CT images improved markedly; two consecutive negative 2019-nCoV nucleic acids tests (not on the same day).

(3) Lung CT improvement rate. A reduction of 30% or more in the area of the lesion on the lung CT image was considered to be an improvement in lung CT.

(5) Disappearance rate and disappearance time of main symptoms (fever, cough, fatigue);

(6) Discharge rate and length of hospital stay.

(7) The rate and conversion time of negative 2019-nCoV nucleic acids tests for two consecutive times (not on the same day).

(8) Incidence of adverse events.

(9) Related inflammatory or immune indicators including white blood cell count (WBC), lymphocyte count (LYM), lymphocyte percentage (LYM%), C-reactive protein (CRP), and interleukin 6 (IL-6).

The meta-analysis of 7 RCTs [11, 12, 16, 18, 19, 22, 23] demonstrated that TCM could observably lessen the proportion of patients progressing to severe cases [RR = 0.45, 95% CI (0.29, 0.68), I² = 0%, P = 0.0002] (Fig. 3). In addition, 6 RSs [40, 42, 44, 47‐49] evaluated this proportion and reached the same conclusion [RR = 0.26, 95% CI (0.15, 0.46), I² = 0%, P < 0.00001] (Fig. 4). Both RCTs and RSs confirmed that TCM could decrease the proportion of patients progressing to severe cases by more than 55%.

A high-quality RCT [26] we included indicated that the mortality rate of severe or critical patients in TCM group was visibly lower compared with CWM group [38.6% (22/57) vs 75.9% (41/54), RR = 0.51, 95% CI (0.35, 0.73), P = 0.0002]. Five RSs [52, 54, 55, 57, 58] reached the same conclusion [RR = 0.47, 95% CI (0.31, 0.70), I² = 54%, P = 0.0002] (Fig. 5). In summary, TCM could decrease the mortality rate of severe or critical patients by more than 49%.

Efficacy assessment of secondary outcomes was summarized in (Additional file 5).

Meta-analysis indicated that TCM could enhance the total effective rate by 18%, clinical cure rate by 26%, and lung CT improvement rate by 19%. Applying TCM could reduce TCM symptom scores by more than 2.75 points. The discrepancy in the disappearance rate of fever was not statistically significant, but the disappearance time of fever could be shortened by more than 1.05 days in the TCM group compared with the CWM group. TCM could enhance the disappearance rate of cough by more than 33% and disappearance rate of fatigue by more than 28%, but the difference in the disappearance time of cough and fatigue was not statistically significant. TCM could enhance the discharge rate by 33% and shorten the length of hospital stay by 3.07 days, especially in severe or critical cases. In addition, TCM could increase the rate of negative 2019-nCoV nucleic acids tests by 37%, and shorten the conversion time of negative 2019-nCoV nucleic acids tests by more than 1.58 days. The results of RSs showed that the incidence of adverse events in TCM group was 82% of that in CWM group, but the difference did not reach statistical significance in RCTs. However, it was not observed that TCM increased the incidence of serious adverse events in COVID-19 patients, on the contrary, it has been shown that applying TCM could reduce the incidence of serious adverse events in severe or critical cases [98.1% (53/54) vs 78.9% (45/57), P = 0.002] [26]. Applying TCM could increase WBC by 0.25 × 10⁹/L and LYM by 0.23 × 10⁹/L. In addition, TCM could decrease CRP by 7.65 mg/L and IL-6 by 4.81 ng/L. The difference in LYM% between TCM and CWM group did not reach statistical significance.

Studies suggested that 2019-nCoV infection was initiated through the combination of virus with host cell surface receptor ACE2 (angiotensin converting enzyme 2), fusion of virus with cell membrane, and release of virus genome into cells. Among them, Spro (viral spike protein) mediated the activity of receptor binding and membrane fusion [95, 96]. 3CLpro (3C-like protease) in coronaviruses played a vital role in advancing the polyprotein translated from viral RNA [97]. RdRp (RNA-dependent RNA polymerase) was essential for coronavirus replication and transcription and might be a primary target for antiviral drugs [98]. Viral Plpro (papain-like cysteine protease) was considered an important target of antiviral drugs because it was required for SARS-CoV-2 replication and could promote the dysregulation of signaling cascades in infected cells [99]. Nsp14 (nonstructural protein 14) was a functional enzyme related to replication fidelity and involved in mRNA capping, which played an essential role in virus replication [100].

It has been suggested that a proportion of severe COVID-19 patients might suffer from cytokine storm syndrome and might die because of high inflammation driven by the virus [101]. IL-6 (interleukin-6) was an important cytokine with a variety of physiological functions, including regulating immune cell proliferation and differentiation [102]. IL1B (interleukin 1 beta) was considered an essential pro-inflammatory cytokine related to the origination and development of acute respiratory distress syndrome (ARDS) [103]. TNF (tumor necrosis factor) was regarded as a primary inflammatory cytokine that could drive cytokine production, cell survival, or cell death [104]. CCL2 [Chemokine (C–C motif) ligand 2] exhibited a chemotactic activity for monocytes and was one of the key chemokines regulating monocyte/macrophage migration and infiltration [105]. Severe COVID-19 manifested as ARDS with elevated pro-inflammatory cytokines, involving TNF-α, IL-6, IL1B, and CCL2. Therefore, the treatment related to anti-cytokine or anti-cytokine-signaling would be conducive to the prognosis of COVID-19 [106]. Akt1 was one of 3 related serine/threonine-protein kinases implicated in pulmonary fibrosis and lung injury and also played an essential role in immune cell modulation [72].

It has been confirmed that four flavonoids including formononetin, apigenin, luteolin, and kaempferol had in vitro activities against enterovirus 71 infection due to reducing viral replication and protein synthesis [107]. Studies suggested that naringenin might be a promising treatment strategy against COVID-19 due to its antiviral and anti-inflammatory effects [108]. Bicuculline showed in vitro anti-inflammatory activity and attenuated inflammation by decreasing the generation of pro-inflammatory cytokines, like IL1B and TNF-α, and promoting the production of the anti-inflammatory cytokine interleukin-10 [109, 110]. Quercetin manifested antiviral, anti-inflammatory, and immune-enhancing effects in vitro and some animal models [111]. Astragaloside IV could improve immunologic function of RAW264.7 cells through stimulating the NF-κB/MAPK signaling pathway [112]. Isoquercitrin could inhibit herpes simplex virus-1 replication to exhibit an antiviral effect [113]. Rutin exerted the antiviral effect mainly by inhibiting or modifying various viral proteins such as viral neuraminidase and DNA/RNA polymerase [114]. Beta-Carotene exerted anti-inflammatory activity through suppressing the production and expression of inflammatory mediators in lipopolysaccharide-stimulated RAW264.7 cells and macrophages [115]. Rosmarinic acid and salvigenin had anti-inflammatory activity, and the mechanisms included inhibiting the maturation and release of IL-1β [116]. Indirubin exhibited potent anti-inflammatory activity and could significantly downregulate the generation of IL-6, IL1B, and TNF-α [117]. It has been shown that baicalein exerted anti-H5N1 effects through inhibiting the replication of the influenza H5N1 virus and interfering with the H5N1-induced production of IL-6 and TNF-α in macrophages [118]. Studies suggested that calycosin could diminish the levels of TNF-α, IL-6, and IL1B in mice with acute pancreatitis [119]. Artemetin exhibited evident anti-inflammatory activity in many experimental models in rats [120]. Wogonin exerted antiviral effects against herpes simplex virus (HSV) infection by inhibiting viral replication, mRNA transcription, and protein synthesis [121]. 7-Methoxy-2-methyl isoflavone might exert therapeutic effect on COVID-19 through inhibiting inflammatory storms and modulating immune function [90]. Emodin had anti-inflammatory, antiviral, and antibacterial effects, and the anti-inflammatory had been confirmed in various inflammatory models, including asthma, arthritis, and pancreatitis [122].

.