Treatment efficacy analysis of traditional Chinese medicine for novel coronavirus pneumonia (COVID-19): an empirical study from

Theoretic base of TCM comprehension on NCP

TCM comprehension on pathological evolution of NCP and medication paradigm were analyzed based on syndrome differentiation. As NCP belongs to the category of epidemic disease, the pathogen is generally attributed to dampness toxin according to the main symptom characteristics of this disease. TCM believes that NCP locates in lung, and is closely related to spleen and stomach, and its pathological changes involve in heart, liver and kidney in the later stages. Dampness pathogen can change into cold-dampness (寒湿) pathogen following the Yin body constitution (阴性体质), and also become the dampness-heat (湿热) pathogen following the Yang body constitution (阳性体质). Clinical observation shows that dampness toxin can directly invade into middle energizer (中焦) in partly NCP patients, and leads to the dysfunction of Qi movement. If the treatment method is appropriate and sufficient healthy Qi (正气) gradually recovers, the pathogen will be driven out, consequently the patient will enter into the recovery period. At the same time clinical manifestations appear some symptoms of Qi and Yin deficiency (气阴两虚证). Thus, the pathological evolution of NCP in TCM can be summarized as dampness toxin invading defense exterior (卫表) in early stage, and then enters the lungs and influences spleen function, eventually involves heart, liver and kidney, which causes Yang Qi collapse (阳气外脱) by excess pathogen and Yin and Yang separates from each other (阴阳离决). If the treatment is timely and suitable, sufficient healthy Qi can eliminate the pathogen, syndromes with deficiency of Qi and Yin in lung and spleen will be manifested (Fig. 1).

Although the pathogen between NCP and SARS have certain similarities in virus origination, the TCM pathogenesis of NCP and SARS appear to be different (Table 1).

Table 1 Characteristics of TCM Pathogenesis in NCP and SARS

Except using CMs treatment according to TCM different syndromes, the application of achievements from modern pharmacological research on CMs, targeting the pathological changes of NCP at different periods will also improve therapeutic effects. For example, Ephedra sinica Stapf, Schizonepeta tenuifolia Briq., Perillae Folium and Lonicerae Japonicae Flos have antipyretic and analgesic actions, and Ma Xing Shi Gan Decoction (麻杏石甘汤), Arctium lappa L. and Poria cocos (Schw.) Wolf. are able to regulate immune function and suppress inflammatory cytokine storm, suggesting that TCM plays a comprehensive beneficial regulating role in the treatment of NCP through multi-level and multi-pathways.

Characteristics of patients and TCM utilization

Demographic characteristics of patients were shown in Table 2. The average age of all the 54 patients was 55.07 years old. Male patients (52.03 years old) were younger than female patients (60.25 years old) in average (t = 2.128, P = 0.028). The average length of hospital stay was 8.96 days. Patients over 45 years old stayed 9.79 days in hospital in average, which was longer than 7.64 days of patients under 45 years old (t = 2.232, P = 0.034).

Table 2 Demographic characteristics and TCM features of patients

The common symptoms were correlative with TCM clinical features (Table 3). Fever was negatively related to cough and shortness of breath, and was in positive correlation to abnormal digestion, Red tongue with white coating (舌红苔白) and deep pulse (沉脉). Cough was in negative correlation to shortness of breath and abnormal digestion. Shortness of breath was negatively related to deep pulse.

Table 3 Relevant of Symptoms and TCM clinical features

NCP patients were prescribed with 87 kinds of CMs in total, and the most frequently used CMs were listed according to their classification (Table 4).

Table 4 The most frequently used CMs for NCP patients

Clinical effect of TCMs treatment

Multiple linear regression was employed with the length of hospital stay as dependent variable, and patients’ age, gender, symptoms and CMs they administrated as independent variables (Table 5). The method of backward elimination was chosen to get the optimal fitting degree of regression model (Table 6).

Table 5 Assignment of independent variables in multiple linear regression

Table 6 Regression coefficients in multiple linear regression

The regression equation of length of hospital stay was:

$${text{Y}},{ = }, – ,4.999, + ,0.106;{text{X}}_{1} , + ,4.538;{text{X}}_{2} , – ,3.474,{text{X}}_{3} , – ,2.665,{text{X}}_{4} , + ,1.925,{text{X}}_{5} , – ,1.510,{text{X}}_{6} , + ,7.864,{text{X}}_{7} , + ,9.771,{text{X}}_{8} ,,{text{with}};{text{R}}^{2} , = ,0.569,{text{and}},{text{adjusted}};{text{R}}^{2} , = ,0.353.$$

According to the standard coefficients, the factor affecting the length of hospital stay for the most was administrating category of ISRD, followed by administrating other CMs, male, and cough.

Patients received blood test both when admitted to hospital and discharge from hospital (Table 7).

Table 7 Values of blood test when admission and discharge of hospital

The indicators of blood test included white cell count (WBC), percentage of neutrophils (Neu %), percentage of lymphocytes (Lym %), percentage of mononucleosis (Mon %), percentage of eosinophils % (EOS %), percentage of alkaline granulocytes (Bas %), number of neutrophils (Neu#), number of lymphocytes (Lym#), number of single-core cells (Mon#), number of eosinophils (Eos#), number of alkaline granulocytes (Bas#), red cell count (RBC), hemoglobin concentration (HGB), hematocrit (HCT), average red blood cell volume (MCV), average red blood cell hemoglobin concentration (MCH), platelet number (PLT), average platelet volume (MPV), platelet distribution width (PDW) and platelet pressure (PCT).

(Table 7 Values of blood test when admission and discharge of hospital)

By comparing the parameters of blood test between admission and discharge of hospital, the values of WBC, EOS %, Neu#, Lym#, Eos#, Bas#, PLT and PCT increased, while values of RBC, HGB, HCT, and MPV decreased. The result indicated that TCM treatment significantly ameliorated the immune ability against SARS-CoV-2 in patients.

Network pharmacology of QEPDD and NHC-recommended formulas

The five disease stages consist of nine syndromes and nine formulas. The sovereign medicinal of mild disease stage contained Pogostemon Cablin (Blanco), Atractylodes Lancea (Thunb.)Dc., Scutellariae Radix, Chaihu Radix Bupleuri, Forsythiae Fructus; the moderate stage contained gypsum, Atractylodes Lancea (Thunb.)Dc., Polygoni Cuspidati Rhizoma Et Radix, Pogostemon Cablin (Blanco), Verbenae Herb; the severe stage contained Ephedra Herba, gypsum, Lepidii Semen Descurainiae Semen, buffalo horn; the critical stage contained Panax Ginseng C. A. Mey., Aconiti Lateralis Radix Praeparata; and the recovery stage contained Hedysarum Multijugum Maxim., Ophiopogon japonicus (Linn. f.) Ker–Gawl, Panacis Quinquefolii Radix.

There are 21 CMs in the QFPDD, and 16 CMs in recommended formulas of five disease stages. After deleting duplication, 32 different CMs were selected in total. As the Ophiopogon japonicus (Linn. f.) Ker–Gawl., gypsum, and buffalo horn were not found in TCMSP database, finally 29 CMs were picked in this study. There are 201 compounds in recommended formulas of five disease stages and 288 compounds in the QFPDD after screening. Comparing with the compound in sovereign medicinal, we discovered kaempferol, beta-sitosterol, Stigmasterol, quercetin, luteolin, Genkwanin, diop, isorhamnetin participated in three or more disease stages. At the same time, these compounds were not unique to a single Chinese herbal medicine, but many common CMs. Comparing the compounds in QFPDD with those in recommended formulas of the first four disease stages, seven of the eight compounds (except isorhamnetin) were found to be representative in QFPDD and in recommended formulas of three or four disease stages.

The recommended formulas included 164 types of putative targets in mild disease stage, 147 types in moderate stage, 150 types in severe stage, 88 types in critical stage and 112 types in recovery stage. Totally, there were 204 types of different targets in recommended formulas, and 240 targets in QFPDD. After comparing the putative targets of five disease stage, it was found that 169 of the 204 targets were common targets, among which 58 were involved in the treatment of all five disease stages. By comparing QFPDD with recommended formulas of the first four stages, it was found that only 9 of 248 targets have nothing to do with QFPDD, and 60 targets were common targets.

According to the compounds (C)-putative targets (T) network (C-T network) related to the five stages, the degree value of quercetin, luteolin, kaempferol, acacetin and genkwanin were all greater than 30 and exist in a variety of CMs, which acted on different disease stages (Fig. 2). And the degree value of a total of 25 targets was greater than 30, among which the first five were CYP1B1, ESR1, ESR2, AR, and ABCG2, respectively.

Then a C-T network of QRPDD was constructed similarly. By considering the large number of targets, only 60 common targets in QFPDD and recommended formulas of the first four disease stages were selected for visualization (Fig. 3). For the compounds, kaempferol, quercetin, luteolin, galangin, luteolin, isorhamnetin and the degree value were all greater than 30. For the targets, the first five were CYP1B1, ABCG2, CA7, CA4 and ESR2, respectively.

The hexagon on the left represents the compounds, and the different colours from top to bottom represent the different compounds in mild, moderate, severe, critical and recovery stages of the disease. The right circle represents the putative targets, the yellow represents the 58 common targets of the five disease stages, and the orange represents 197 non-common targets. The edges show the relationship between the compounds and putative targets obtained by SEA prediction. Size of the point is related to its degree value, and the larger the value, the larger the point.

The blue hexagon represents the compound, and the red circle points represent the 60 targets shared by QFPDD and the recommended compound of the first four disease stages. The rest is consistent with Fig. 2.

As shown in Fig. 4, according to the enrichment results of pathways corresponding to the five disease stages, figure B, C and D show a large number of enrichment pathways, while E and F are more concentrated. Nitrogen metabolism is presented in the first four diseases stages, with a high enrichment rate as well as a low p value, and the expression of arachidonic acid metabolism in the critical disease stages was the most obvious, indicating that formulas were more inclined to inhibit pathogens in the critical diseases stage. Ovarian steroidogenesis exists mainly in severe and critical disease stages. For QFPDD, nitrogen metabolism, linoic acid metabolism and steroid hormone biosynthesis have high enrichment rate (Figure A), which is roughly consistent with the results of mild and moderate stages.

Figure A represents QFPDD and figure B-F represent recommended formulas of mild, moderate, severe, critical and recovery stages of the disease, respectively. The x axis means enrichment ratio and the y axis means the name of KEGG pathways.