Comparative Efficacy of Chinese Herbal Injections Combined with Paclitaxel Plus Cisplatin for Non-Small-Cell Lung Cancer: A Multidimensional Bayesian Network Meta-Analysis

Introduction Considering the limitations of pure paclitaxel plus cisplatin chemotherapy in the treatment of non-small-cell lung cancer and the extensive exploration of Chinese herbal injections, this study performed a multidimensional network meta-analysis to systematically evaluate the clinical efficacy and safety of 12 Chinese herbal injections in the treatment of non-small-cell lung cancer. Methods Randomized controlled trials were obtained from several databases according to the eligibility criteria, and the study quality was assessed by the Cochrane risk of bias tool. Data analysis was performed by Stata 13.1 software and WinBUGS 14.0 software. Multidimensional cluster analysis was performed with the “scatterplot3d” package in R 3.6.1 software (PROSPERO ID: CRD42020163503). Results A total of 58 eligible randomized controlled trials involving 4578 patients and 12 Chinese herbal injections were included. Combined with paclitaxel plus cisplatin chemotherapy, Xiaoaiping injection exhibited a better impact on the clinical effective rate than paclitaxel plus cisplatin alone. Shenqifuzheng injection was associated with a preferable response in performance status and reduced leukopenia and gastrointestinal reactions. Kangai injection was dominant in the comprehensive results of the cluster analysis. Conclusions Chinese herbal injections combined with paclitaxel plus cisplatin chemotherapy have a certain adjuvant effect in treating non-small-cell lung cancer, but the results of this study need to be verified by more well-designed, large-sample, multicenter randomized controlled trials.


Introduction
According to the latest global cancer report, both the incidence and mortality of lung cancer ranked first among malignant tumors. Currently, the survival rate of most cancers is steadily increasing, while the 5-year relative survival rate of lung cancer is still stagnant, at 18% [1][2][3][4]. Primary bronchial lung cancer (referred to as lung cancer) is internationally divided into small-cell lung cancer and nonsmall-cell lung cancer (NSCLC) [5]. NSCLC accounts for approximately 85% of the total number of lung cancer cases.
is value has still been increasing in the recent years, and more than 75% of patients are already in an advanced stage of lung cancer at the time of initial diagnosis and miss the opportunity for surgical treatment [6][7][8]. In the recent years, the treatment of NSCLC has improved, and its primary therapeutic regimens involve the combination of two drugs based on platinum drugs, such as vinorelbine plus cisplatin, paclitaxel plus cisplatin (TP), and gemcitabine plus cisplatin [9][10][11]. TP is a commonly used drug in chemotherapy for patients with NSCLC, and its antitumor efficacy has been demonstrated [12][13][14][15][16]. However, TP chemotherapy is often accompanied by the occurrence of adverse events, and how to improve clinical efficacy and reduce adverse reactions has attracted much attention [17,18].
In China, the treatment of traditional Chinese medicine combined with chemotherapy for cancer has been widely used [19,20]. Studies have shown that traditional Chinese medicine has beneficial effects on treating cancer, delaying cancer progression, reducing the complications caused by chemotherapy, and reducing adverse events [21][22][23]. Chinese herbal injections (CHIs), as an integral part of traditional Chinese medicine, play a vital role in treating cancer. Studies have shown that many components in Xiaoaiping injection have significant antitumor effects. Among them, there are dozens of C-21 steroidal saponins, which are the main anticancer components [24]. Modern pharmacological studies have indicated that these components can inhibit tumor cell proliferation, induce apoptosis, reduce the drug resistance of antitumor drugs, and improve body immunity [25][26][27][28][29][30].
Network meta-analysis (NMA) was developed from traditional meta-analysis, extending from the standard twoarm meta-analysis to indirect comparisons of multiple related but different interventions at the same time [31]. While CHIs are widely used in clinical practice, the comparison of the efficacy between multiple varieties cannot be achieved through traditional meta-analysis methods. e greatest advantage of the NMA method is that it can summarize different interventions for the treatment of a certain kind of disease, then carry out quantitative statistical analysis and rank these interventions according to the results of one outcome indicator, and finally, choose the optimal treatment plan. is NMA compares the efficacy of 12 CHIs combined with TP chemotherapy in the treatment of NSCLC by quantitatively synthesizing the evidence. is study aims to provide a reference for the clinical application of CHIs to assist in the treatment of NSCLC.

Method
is NMA was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension statement for reporting of systematic reviews incorporating network meta-analyses of healthcare interventions [32]. A completed PRISMA checklist is included as an additional file (Checklist). A protocol, including the full methods for this review, is available from https://www.crd.york.ac.uk/prospero/ display_record.php?ID�CRD42020163503.

Inclusion and Exclusion
Criteria. Studies were considered eligible for inclusion if the following criteria were met.
(1) Randomized controlled trials (RCTs) of CHIs combined with the TP chemotherapy regimen for treating NSCLC. Studies mentioning "random" were included, regardless of whether blinding methods were used. (2) All patients were diagnosed with stage III or stage IV NSCLC and had an exact pathological or cytological diagnosis. Sex and race were not limited. (3) e control group was treated with the TP chemotherapy regimen, while the observation group received CHI by intravenous drip and the TP chemotherapy regimen. ere was no limitation on the dosages or treatment courses, but other adjuvant therapies (such as Chinese herbal decoction, other Chinese patent medicines, and surgical treatment) should not be used. Symptomatic treatments were given if other complications occurred. (4) e study described efficacy outcomes, such as the clinical effective rate, performance status, leukopenia, and gastrointestinal reactions [33,34]. e primary outcome was the clinical effective rate, according to WHO's criteria for evaluating the efficacy of solid tumors, which could be divided into four levels: complete response (CR), in which visible lesions completely disappeared after >1 month; partial response (PR), in which the tumor area of a single lesion was reduced by ≥ 50% or the sum of the two largest vertical diameter products of multiple lesions was reduced by >50%; stable disease (SD), in which no significant change in condition occurred for, at least, 4 weeks and the tumor size was estimated to increase by <25% or decrease by <50%; and progressive disease (PD), in which new lesions appeared or original lesions were estimated to increase by ≥ 25%. e clinical effective rate was calculated by the following formula: clinical effective rate � (number of CR patients + number of PR patients)/total number of patients × 100%. e secondary outcomes of interest were performance status and adverse drug reactions/adverse drug events (ADRs/ ADEs), including leukopenia and gastrointestinal reactions. e performance status was assessed by the Karnofsky performance status (KPS) score. KPS scores that increased by ≥ 10 points after treatment were considered to improve the performance status; KPS scores that decreased by ≥ 10 points after treatment were considered to reduce the performance status; and KPS scores that increased or decreased by < 10 points were considered stable. e performance status improvement rate � number of patients with improved performance status/total number of patients × 100%. Referring to the "Acute and Subacute Toxicity Standards of Chemotherapy Drugs" formulated by the WHO in 1981, ADRs/ADEs were divided into 5 grades. e incidence of ADRs/ADEs � number of patients with ADRs/total number of patients × 100%.
RCTs were excluded if they met the following criteria: (1) patients with other primary tumors; (2) the intervention measures were non-TP chemotherapy, or the intervention was combined with other Chinese medical treatments (such as Chinese medicine prescriptions, Chinese patent medicines, or massage), as well as Western medical treatments such as radiotherapy and surgery; (3) the administration of CHIs was nonintravenous; (4) nonclinical trials, self-controlled studies, or studies using incorrect randomization methods; (5) duplicate studies (only those with the most recent publication year, larger sample size, and more comprehensive information were retained); (6) studies with incomplete or erroneous data such as the drug name, dosage, and course of treatment; and (7) studies with inconsistent evaluation criteria and those with no related outcome indicators. Unpublished studies and related academic organizations' websites were also researched to supplement RCTs of CHIs combined with chemotherapy for NSCLC. e search strategy, using a combination of mesh terms and free text search terms, included three parts: NSCLC, CHIs, and RCTs. Using PubMed as an example, "Non-Small-Cell Lung Carcinomas [MeSH Terms]," "Non-Small-Cell Lung Carcinoma," "Non-Small-Cell Lung Cancer," "Non-Small-Cell Lung Carcinoma," "Non-Small-Cell Lung Carcinoma," and "Non-Small-Cell Lung Cancer" were applied to identify relevant publications on NSCLC. More details about the strategy of CHIs are provided in Supplementary Materials (available here).

Study Selection and Data Extraction.
Two researchers independently screened the titles and abstracts of potential articles to remove duplicate studies and exclude studies that definitely did not meet the inclusion criteria, and then, they cross checked the results of the included studies. Regarding repeated reports, for example, a report on the results of different periods of the same trial was included as only 1 study. Any divergences were resolved by discussion or the third reviewer. e main components of the extracted data were as follows: (1) basic information of included studies, including the title, first author's name, publication year, and source of the literature; (2) characteristics of patients, including number of patients, age, sex, KPS score before treatment, type of tumor, and clinical stage; (3) interventions, including name, dosage, course of treatment, and treatment cycle of CHIs, as well as chemotherapy drugs; (4) outcome indicators: measured data on the clinical effective rate, performance status, and ADRs/ADEs (leukopenia and gastrointestinal reactions); and (5) quality assessment items, including key factors for the type of study design and risk of bias assessment.

Risk of Bias Assessment.
e Cochrane risk of bias tool (Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0) [35] was used to evaluate the quality of the eligible RCTs by 2 researchers. Any divergences were resolved by discussion or the third reviewer. Based on the characteristics of CHIs, we formulated quality evaluation items for the RCTs as follows: (1) whether random sequence generation and allocation concealment were mentioned (selection bias) and whether follow-up information was mentioned, including the number and reasons for withdrawals and lost follow-ups; (2) whether the blinding of the participants and personnel was mentioned (performance bias); (3) whether the blinding of the outcome assessment was mentioned (detection bias); (4) whether incomplete outcome data were mentioned (attrition bias); (5) description of eligibility criteria; (6) evaluation of ADRs/ADEs; (7) description of the statistical method; (8) funding; and (9) medical ethics.

Statistical Analysis.
First, heterogeneity analysis was performed using the chi-square test, and the amount of heterogeneity was evaluated with the I 2 statistic. It is suitable to perform NMA when slight heterogeneity exists (I 2 ≤ 50%). However, if substantial heterogeneity exists (I 2 > 50%), the causes of heterogeneity should first be considered. en, we decided whether to use a random effects model after excluding clinical heterogeneity factors and, finally, performed a combined analysis. No consistency test was needed in this study because it was based on an indirect comparison comparing different varieties of CHIs combined with chemotherapy. An NMA evidence relationship diagram was performed by Stata 13.1 software (Stata Corp, College Station, TX, USA), which presents indirect comparative relationships between different interventions [36]. e statistical analysis was performed by WinBUGS 14.0 software (MRC Biostatistics Unit, Cambridge, UK). Advanced Bayesian inference with the Markov Chain Monte Carlo (MCMC) method was performed, in which the posterior probability was inferred from the prior probability and, then, estimated and extrapolated under the assumption that MCMC has reached a stable convergence state. When running the WinBUGS program, the number of iterations was set to 200 000, of which the first 10 000 iterations were used for annealing and to eliminate the effect of the initial value. e results of the indirect comparisons are expressed as the median and 95% confidence interval (95% CI) of the odds ratio (OR), which was used as the effect statistic. e relative ordering of effectiveness was obtained according to the ranking result of the median ORs. According to the surface under the cumulative ranking curve (SUCRA) value, which ranged from 0 to 1, the ranking probabilities of different interventions could be estimated; the intervention is absolutely effective when the SUCRA value is 1, while the intervention is absolutely ineffective when it is 0 [37][38][39][40].
en, based on the SUCRA value, clustering analysis was performed on the outcome indicators, aiming to obtain the best intervention among the two clustering indicators [41]. A symmetrical comparison-adjusted funnel plot was used to assess publication bias and showed that there was no obvious publication bias in this study [42,43].

Multidimensional Cluster Analysis.
Multidimensional cluster analysis of the outcome indicators was performed with the "scatterplot3d" package in R 3.6.1 software (Mathsoft, Cambridge, USA). Using the k-means method to cluster these interventions, the number of clusters was adjusted according to the actual problem. e steps of kmeans clustering were as follows: (1) All interventions were randomly divided into k initial classes, and the average of the outcome indicators of these k classes was used as the initial aggregation point. (2) Each intervention was classified into the category of the closest aggregation point; then, the aggregation point of this category was updated to the average of the current outcome indicators. All interventions were Evidence-Based Complementary and Alternative Medicine recategorized and classified.
Step (2) was repeated until all interventions had been assigned, and finally, the ranking of the interventions with three outcome indicators was visualized through a three-dimensional stereogram. Different colors were applied to indicate interventions belonging to different categories.

Literature Selection.
A total of 6593 articles were retrieved in accordance with the predetermined search strategy and data collection method. After deduplication, reading the titles and abstracts, and excluding significantly the irrelevant literature, a total of 1122 RCTs of CHIs combined with the TP chemotherapy regimen in the treatment of NSCLC were collected. After reviewing the full texts of these studies, 58 RCTs , including  12 kinds [91][92][93][94][95][96][97][98][99][100][101] of CHIs (Aidi injection, Huangqiduotang injection, compound Kushen injection, Delisheng injection, Huanchansu injection, Yadanziyouru injection, Kangai injection, Kanglaite injection, Shenfu injection, Shenmai injection, Shenqifuzheng injection, and Xiaoaiping injection) were included in this study. All of the studies included were Chinese literature. e study identification, screening, and inclusion process is illustrated in Figure 1.

Study Characteristics.
A total of 4578 patients were included in this study, with 2310 patients in the observation group and 2268 in the control group. All included RCTs reported the number of patients, as well as their sex, age, TNM stage, estimated survival, and KPS score before treatment. e TP chemotherapy regimen was given to the control group, and the observation group received CHI intravenously. Information on the types, doses, and treatment courses of CHIs in the observation group and information on the doses of chemotherapeutic drugs in the control group is provided in Table 1. Figure 2 shows the network relationship of CHIs combined with TP chemotherapy in the treatment of NSCLC.

Quality Evaluation.
irteen RCTs (22.41%) generated random sequences with the random number table method, and 1 RCT (1.72%) used the lottery method, so their selection bias due to random sequence generation was evaluated as "low risk." e remaining RCTs that only mentioned "random" were evaluated as "unclear." Two RCTs (3.45%) mentioned the implementation of a blinding method, and 1 RCT (1.72%) mentioned the use of a hidden random sequence, so their performance bias and detection bias were evaluated as "low risk." Most RCTs that did not apply random sequence concealment and blindness were evaluated as "unclear." In addition, as all included studies did not have incomplete outcomes or selective reporting, their attrition bias, reporting bias, and other bias were all evaluated as "low risk." Specific information about the risk of bias is shown in Figure 3.
Based on the ranking results of improving the clinical effective rate, the relative ranking of interventions was as follows: Xiaoaiping injection (88.2%) > Kangai injection

Performance Status.
e NMA for performance status included 41 RCTs, 11 interventions, and 10 kinds of CHIs. e following CHIs combined with TP chemotherapy were significantly more effective in improving performance status than using TP chemotherapy alone: Aidi injection (OR � 0.37; 95% CI, 0. 25 Table 2).
Based on the ranking results of improving performance status, the relative ranking of interventions was as follows: Figure 4).
Based on the ranking results of reducing leukopenia, the relative ranking of interventions was as follows: Shenqi- Figure 4).
Based on the ranking results of reducing gastrointestinal reactions, the relative ranking of interventions was as follows: Shenqifuzheng injection (95.63%) > Kangai injection ( (Figure 4). e SUCRA heat map of all the outcome indicators is shown in Figure 5. e warmer colors represent higher SUCRA values, which indicate that this intervention is more likely to be the best intervention.  Evidence-Based Complementary and Alternative Medicine  (Figure 6(a)). In regard to the cluster analysis of 10 interventions that reported both the improvement of performance status and alleviation of leukopenia, Kangai injection and Shenqifuzheng injection combined with TP chemotherapy had advantages in the comprehensive ranking, with TP only yielding the worst result ( Figure 6(b)). When cluster analysis was conducted with 10 interventions that reported the clinical effectiveness rate, improvement of performance status, and alleviation of leukopenia, Kangai injection combined with TP was dominant in the comprehensive ranking of all three outcome indicators, and TP only was evaluated as the     (Figure 6(c)). In regard to the cluster analysis of 11 interventions that reported the clinical effective rate, improvement of performance status, and alleviation of gastrointestinal reactions, Shenqifuzheng injection combined with TP chemotherapy had the potential to be the best intervention according to the comprehensive ranking, while TP only was considered the worst ( Figure 6(d)). e cluster analysis conducted with 10 interventions that reported the clinical effectiveness rate, alleviation of leukopenia, and alleviation of gastrointestinal reactions indicated that Kangai injection and compound Kushen injection combined with TP chemotherapy had the potential to be the best intervention (Figure 6(e)). Kangai injection and compound Kushen injection combined with TP were dominant in the comprehensive ranking of the improvement of performance status, alleviation of leukopenia, and alleviation of gastrointestinal reactions (Figure 6(f )).

Publication Bias.
Publication bias detection was performed using funnel plots for the clinical effective rate and performance status (Figure 7). ere was incomplete symmetry in both funnel plots, and there were certain angles between the correction auxiliary line and the centerline, which indicated that this study has publication bias.

Discussion
is study compared the clinical efficacy and safety of 12 CHIs combined with TP in the treatment of NSCLC with the NMA method. e results are as follows. (1) In terms of the clinical effective rate, based on TP chemotherapy, Aidi injection, compound Kushen injection, Kangai injection, Kanglaite injection, Shenmai injection, and Yadanziyouru injection were significantly more effective than using TP chemotherapy alone. (2) In terms of performance status, based on TP chemotherapy, Aidi injection, compound Kushen injection, Huachansu injection, Yadanziyouru injection, Kangai injection, Kanglaite injection, Shenqifuzheng injection, and Xiaoaiping injection were significantly more effective than using TP chemotherapy alone. (3) In terms of ADRs/ADEs, based on TP chemotherapy, compound Kushen injection, Yadanziyouru injection, Kangai injection, and Shenqifuzheng injection were significantly more effective in reducing leukopenia than TP chemotherapy alone. Based on TP chemotherapy, Kangai injection and Shenqifuzheng injection were significantly more effective in reducing gastrointestinal reactions than TP chemotherapy alone. (4) In the cluster analysis, Kangai injection with TP chemotherapy was dominant in the comprehensive ranking of the clinical effective rate and alleviation of leukopenia. Based on TP chemotherapy, Kangai injection and Shenqifuzheng injection were dominant in the comprehensive  NSCLC is a common malignant tumor that has the characteristics of high mortality and insignificant early symptoms. Due to the low rate of early diagnosis and the rapid development of the disease, most patients have already reached the advanced stage of local cancer or the tumor has already metastasized at the time of diagnosis. Conventional treatments for NSCLC include surgery, radiotherapy, and chemotherapy [8,[102][103][104]. However, while radiation or chemotherapy kills cancer cells, it can also cause serious toxic side effects, complications, and drug resistance. erefore, many doctors and patients with NSCLC are actively exploring other complementary and alternative therapies for cancer treatment [105,106]. Traditional Chinese medicine has been used clinically for the treatment of cancer in China, Japan, and even other Asian countries for thousands of years. In recent years, Chinese medicine as a complementary and alternative medicine has been indicated to have unique advantages in improving the clinical symptoms of cancer patients, improving quality of life, enhancing body immunity, and extending survival. At the same time, the combined use of traditional Chinese medicine and conventional cancer treatment methods such as radiotherapy and chemotherapy also has a therapeutic effect of reducing toxicity and increasing efficacy [107]. According to relevant pharmacological studies of Yadanziyouru injection, which ranked dominantly in performance status, Brucea javanica, a plant in the Simaroubaceae family, has significant antimalarial and antitumor activities. In particular, fatty acid components, such as oleic acid, linoleic acid, palmitic acid, stearic acid, and arachidonic acid, have strong antitumor activity and have great killing effects on tumor cells [108][109][110][111]. e mechanism may be that they can selectively destroy the membranes and mitochondria of tumor cells, inhibit growth and DNA synthesis in tumor cells, block the cell cycle of cancer cells, improve immunity, and reverse the drug resistance to many drugs [112][113][114][115]. Shenqifuzheng injection consists of the traditional Chinese medicines Codonopsis and Astragalus. e two can be used together to promote blood circulation, eliminate blood stasis, and eliminate lumps. Modern pharmacological research shows that Codonopsis can enhance the body's immunity, increase the content of white blood cells, red blood cells, and hemoglobin, adjust gastrointestinal function, and resist fatigue. Astragalus enhances immune function and has a complete immune recovery effect on lymphocyte function in cancer patients [116][117][118].
At present, there are two related NMAs written in Chinese in the databases, both of which have systematically evaluated the clinical efficacy of CHIs combined with TP in the treatment of NSCLC. Compared with those two studies, the advantages of this study are reflected in the following aspects. (1) For the first time, this study used R software to perform a multidimensional clustering analysis on the included outcome indicators, and the results were visualized through a three-dimensional stereogram. Different interventions were compared under multiple outcome indicators.
(2) is study not only updated the search time and included the largest number of studies but also conducted a comprehensive search of 22 CHIs that have been used clinically for tumor treatment at this stage. e search terms used in this study were divided into three parts: NSCLC, CHIs, and RCTs. With a search strategy as the combination of mesh terms and free text, the literature retrieved for this study was comprehensive. (3) Only slight clinical heterogeneity was found among the included studies, as strict inclusion and exclusion criteria were established; the common intervention measure of the included RCTs was TP chemotherapy, and only the WHO evaluation criteria for solid tumor efficacy were adopted as the clinical efficacy evaluation criteria. (4) is study focused on not only important indicators to measure the therapeutic effect of cancer such as the clinical effective rate and improvement in performance status but also the incidence of ADRs/ADEs. Moreover, this study also   performed cluster analysis and comprehensive ranking for the above outcome indicators.

Limitations
Nevertheless, there were still limitations in this study. First, the survival rate is critical in determining and judging the effectiveness of treatment for cancer patients; however, with insufficient information to perform NMA, this study did not evaluate such long-term endpoint outcome indicators. Next, the included studies did not conduct traditional Chinese medicine dialectical analysis. e systematic review should classify the data into different syndromes to provide clearer guidance for clinical practice. In addition, most of the RCTs in this systematic review had low methodological quality, small sample sizes, and incomplete descriptions of randomization and allocation. ese RCTs also lack emphasis on the implementation of blinding methods and the approval of medical ethics, which may affect the reliability of the overall research. erefore, we recommend that RCTs be registered in advance to ensure the transparency of the trial process and improve methodological quality. In addition, RCTs should be implemented in accordance with the latest clinical diagnostic and treatment guidelines. In view of the abovementioned limitations, the results of this study need to be verified by more well-designed, large-sample, multicenter RCTs.

Conclusions
In conclusion, the use of CHIs was associated with improved treatment performance and could be beneficial for patients with NSCLC compared to using TP alone. According to multidimensional cluster analysis, Kangai injection showed the best efficacy with respect to simultaneously increasing clinical efficacy and reducing ADRs/ADEs, followed by Shenqifuzheng injection and compound Kushen injection.
However, more direct comparison studies of two or more CHIs are needed to further confirm the results.

Abbreviations
95% CI: 95% confidence interval ADRs/ADEs: Adverse drug reactions/adverse drug events CHIs: Chinese herbal injections CR: Complete response KPS: Karnofsky performance status NMA: Network meta-analysis NSCLC: Non-small-cell lung cancer OR: Odds ratio PD: Progressive disease PR: Partial response RCTs: Randomized controlled trials SD: Stable disease SUCRA: Surface under the cumulative ranking curve TP: Paclitaxel plus cisplatin.
Data Availability e datasets we generated during the analysis are not publicly available because the analysis process is the core of the results and cannot be made public. Requests to access the datasets of included studies can be met by checking out them in databases.

Ethical Approval
Ethical approval was not necessary in the current study because our study gathered data only from Embase, PubMed, the Cochrane Library, the China National Knowledge Infrastructure Database (CNKI), the Wan-fang Database, the CQVIP Database (VIP), and the China Biology Medicine disc (SinoMed) database, and this procedure   did not address any patient's personal data or harm any patient.

Conflicts of Interest
e authors Bin Sun and Qianqian Zhang were employed by the company Jiangsu Jiuxu Pharmaceutical Co., Ltd. e remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.

Authors' Contributions
Mengwei Ni was involved in conceptualization, methodology, software, formal analysis, investigation, data curation, writing the original draft, visualization, and project administration. Haojia Wang was involved in methodology, formal analysis, investigation, writing the original draft, and visualization. Miaomiao Wang was involved in methodology and writing the review and editing. Wei Zhou was responsible for software and validation. Jiarui Wu was involved in conceptualization, resources, supervision, project administration, and funding acquisition. Bin Sun and Qianqian Zhang were responsible for validation and resources. Xiaotian Fan and Xinkui Liu were responsible for software. Dan Zhang and Xiaoguang Sheng were involved in writing the review and editing. Zhiwei Jing was responsible for resources. Jingyuan Zhang, Ziqi Meng, and Shuyu Liu conducted formal analysis. Siyu Guo and Shanshan Jia conducted investigation. Xiaomeng Zhang performed data Curation.