Efficacy and safety of TCM combined with chemotherapy for SCLC: a systematic review and meta-analysis

Abstract

Background

Chemotherapy is the standard treatment for small cell lung cancer (SCLC), but chemotherapy resistance and adverse reactions remain major problems. Although Traditional Chinese Medicine (TCM) is wildly applied for patients with SCLC in China, the evidence of TCM in the treatment for SCLC is limited.

Purpose

To evaluate the efficacy and safety of TCM combined with chemotherapy for patients with SCLC.

Method

We conducted a systematic search of PubMed, EMBASE, the Chinese National Knowledge Infrastructure, the VIP Information Database, and the Wanfang Database for randomized-controlled trials (RCTs) that are relevant. The included studies were reviewed by two investigators, with relevant data extracted independently. The effect estimate of interest was the relative risk (RR) or mean difference with 95% confidence intervals (95% CIs).

Results

22 RCTs involving 1887 patients were included in this study. Compared with patients treated with chemotherapy© alone, those with Chinese herbal medicine and chemotherapy (TCM-C) had better therapeutic effects (RR = 1.295, 95% CI 1.205–1.391, P < 0.001), KPS scores (RR = 1.310, 95% CI 1.210–1.418, P < 0.001), 1-year survival rate (RR = 1.282, 95% CI 1.129–1.456, P < 0.001), 3-year survival rate (RR = 2.109, 95% CI 1.514–2.939, P < 0.001), and 5-year survival rate (RR = 2.373, 95% CI 1.227–4.587, P = 0.01). The incidence of gastrointestinal reaction (RR of = 0.786, 95% CI 0.709–0.870, P < 0.000) and bone marrow depression (RR = 0.837, 95% CI 0.726–0.965, P = 0.014) in TCM-C group were lower than that in the C group.

Conclusion

The systematic review indicated that TCM combined with chemotherapy may improve therapeutic effect, quality of life, and prolong survival time. More large-scale and higher quality RCTs are warranted to support our findings.

PROSPERO registration number

CRD42016038016.

Electronic supplementary material

The online version of this article (10.1007/s00432-020-03353-0) contains supplementary material, which is available to authorized users.

Background

Lung cancer is still a global health issue today, especially in China. There are more than 2.09 million newly diagnosed lung cancers around world in 2018 (Chen et al. 2015a). More than 0.7 million new lung cancer patients were Chinese (Hong et al. 2015). Meantime, it is the leading cause of cancer mortality for both men and women in China (Chen et al. 2015b). SCLC, an aggressive disease accounting for 13–15% of all cases of lung cancer (Shi et al. 2019), is known for early metastatic spread to regional lymph nodes and distant sites and rapid growth (Kalemkerian and Schneider 2017). Considering its sensitivity, chemotherapy has become the standard treatment for SCLC (Waqar and Morgensztern 2017). However, the majority of patients, who usually exhibit sensitive responsiveness to chemotherapy, subsequently relapse with resistant diseases very soon. Meantime, the side effects of chemotherapy seriously affect the treatment and patients’ quality of life. According to the reports, approximately 80% of patients with limited-stage small cell lung cancer (LS-SCLC) and almost all patients with extensive small cell lung cancer (ES-SCLC) relapse or progress within 1 year after treatment, and about 95% of them eventually die from disease progression (Dazzi et al. 2013).

In Asia, TCM is widely used to treat various diseases (Lv et al. 2018), including for China’s SCLC patients during or after chemotherapy (Konkimalla and Efferth 2008), with some clinical trials showing that the combination of TCM and chemotherapy can enhance the therapeutic effect. Also, TCM is considered as an important supplementary therapy with beneficial effects for SCLC in reducing chemotherapy-related side effects and improve the quality of life (Wang 2009). In the past 20 years, more clinical trials concerning the efficacy and safety of TCM for SCLC were conducted (Wang 2009, 2014; Fu and Zhang 2013; Li and Zheng 2001; Xu and Wei 2010; Shao 1999; Zhang and Zhang 2013; Wu and Ma 2015; Dang et al. 2002; Ran et al. 2010, 2011, 2014; Wo 2013; Dong et al. 2014; Sun and Shao 2008; Wang et al. 2013, 2007; Zhang et al. 2013; Zhao et al. 2004; Zhang and Wang 2018; Sun 2017; Qiao and Zhao 2013), but are less convincing confined to their small sample sizes. Therefore, we conducted this systematic review to evaluate the efficacy and safety of TCM combined with chemotherapy for patients with SCLC.

Methods

This study was conducted based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (Moher et al. 2015).

Inclusion and exclusion criteria

Studies were included if they met the following PICO(S) [participants, intervention, comparators, outcomes (study designs)] criteria:

Participants

Adult patients diagnosed with small cell lung cancer regardless of age, gender, and chemotherapy regimen. The diagnosis must have been confirmed by pathological or cytology diagnosis.

Interventions

Based on the Pharmacopoeia of the People’s Republic of China 2010, decoctions, tablets, pills, powders, granules, capsules, oral liquids, and injections are all formulations of TCM (Committee CP 2010), which, typically consisting of two or more herbs to achieve specific effects under specific conditions, are all determined by physicians according to TCM diagnostic and therapeutic theory.

Comparator

Participants in the TCM-C group should receive the conventional chemotherapy and TCM for at least 28 days during or after chemotherapy. Participants in the C group should be treated by the conventional chemotherapy alone. There is no limit to the number of TCM herbs and formulas.

Outcome measures

The main outcome measure is therapeutic effect according to standard for therapeutic effect evaluation of solid tumor by Response Evaluation Criteria in Solid Tumors (RECIST) (Eisenhauer et al. 2009). Survival time, quality of life evaluated with Karnofsky score, and adverse events are second outcome measures.

Studies

Only randomized-controlled trials (RCTs) were included.

Studies were excluded for: (1) adult patients diagnosed with non-small cell lung cancer; (2) patients did not receive TCM; (3) patients did not receive chemotherapy; (4) patients received tumor resection or targeted therapy; (5) duplicate publications (only the largest publication kept); (6) case reports, letters, reviews, conference abstracts, animal experiments, and expert opinions.

Search strategy

Relevant publications evaluating the efficacy and safety of TCM for SCLC were searched in the Chinese National Knowledge Infrastructure (CNKI), the VIP Information Database, Wanfang Database, PubMed, EMBASE, ClinicalTrails.gov, and the Cochrane Library from their inception up to September 15, 2019.

Study selection and data extraction

Studies were independently selected by two reviewers according to the above inclusion and exclusion criteria. Two reviewers independently extracted the following detailed information of each trial: the first author’s name and publication year; the characteristics of patients; intervention information: composition, dosage and treatment time of TCM and chemotherapy; primary and secondary outcome measures. Missing information is obtained by contacting the author. Disagreements about data extraction were resolved through discussion with third parties.

Risk of bias and quality assessment

Four authors will independently assess the methodological quality of included studies. The methodological quality of the included RCTs will be assessed according to the guidance of the Cochrane Handbook for Systematic Review of Interventions, Version 5.1.0 (Higgins and Green 2011), which includes the following seven criteria: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessments (detection bias), incomplete outcome data (attrition bias), selective outcome reporting (reporting bias), and other sources of bias. The GRADE system (GRADEpro GDT 2015) was used to evaluate quality of evidence and certainty of conclusions (GRADEpro G. GRADEpro GDT 2015).

Data synthesis and analysis

The data were processed and analyzed by Stata 13.0 (Stata Corp, College Station, TX). Random-effects models were performed to calculate pooled effects. Fixed-effect models were performed if statistical heterogeneity was absent (heterogeneity test, P ≥ 0.10), dichotomous data were presented as pooled Risk Ratio (RR) with 95% confidence intervals (95% CIs), while continuous data were presented as Mean Difference (MD) with 95% Cis. Assessment of heterogeneity was performed using Cochran’s Q test and Higgins’s I²; I² > 50% and a P value < 0.10 suggested significant heterogeneity (Higgins et al. 2003). Sensitivity analysis was performed by sequentially omitting each study to examine the robustness of the results. Potential publication bias was evaluated using Begg’s funnel plot and Egger’s test (Egger et al. 1997).

Results

Overall

The study selection was showed in Fig. 1. There were 5235 potential articles identified through literature search from 7 electronic databases. 1349 articles remained after removing duplicates. 3801 articles were excluded after we reviewed the title and abstract. 15 studies were excluded for following reasons: non-SCLC patient (n = 7), conference abstracts (n = 1), not an RCT studies (n = 1); different intervention (n = 2), different outcome (n = 2), and same data source (n = 1). Finally, 22 eligible studies were included in our study.

Fig. 1.

Flow diagram

Table 1 shows the characteristic information of the included studies. Included studies were all single-centered and conducted in China. And these studies are all published in Chinese from 1999 to 2018. 1887 subjects were analyzed. Sample sizes in included studies ranged from 56 to 166. The therapeutic effect was reported in 17 studies. The survival time was reported in seven studies, KPS was reported in nine studies, and adverse events were reported in 14 studies. Peripheral blood counts of white cells were reported in seven studies.

Table 1.

Characteristics of studies included

Study	T/C (n)	Stage	Interventions	Control	Baseline similarity
Wang et al. (2009)	36/38	–	TCM injection, 250 ml, ivgtt, qd, 42 days Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*2 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*2 courses Or: Etoposide, 100 mg/m2, ivgtt, day 1–3, 21 days*2 courses Carboplatin, AUC = 5, ivgtt, day 1, 21 days*2 courses	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*2 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*2 courses Or: Etoposide, 100 mg/m2, ivgtt, day 1–3, 21 days*2 courses Carboplatin, AUC = 5, ivgtt, day 1, 21 days*2 courses	Comparable (p > 0.05)
Wang et al. (2013)	28/28	–	TCM injection, 250 ml, ivgtt, qd, 64 days Docetaxel, 75 mg/m2, ivgtt, day 1, 21 days*4 courses Cisplatin, mg/m2, ivgtt, day 1, 21 days*4 courses	Docetaxel, 75 mg/m2, ivgtt, day 1, 21 days*4 courses Cisplatin, mg/m2, ivgtt, day 1, 21 days*4 courses	Comparable (p > 0.05)
Ran et al. (2011)	44/44	Limited disease	TCM formula, 130 ml, po, TID, 2 years (1st–3rd month: 4 weeks/month; 4th–12th month: 14 days/month; 13th–24th month: 7 days/month) Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Comparable (p > 0.05)
Wang et al. (2014)	32/30	–	TCM injection, 20 ml, ivgtt, qd, 84 days Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Comparable (p > 0.05)
Ran et al. (2010)	35/35	Limited disease	TCM formula, 130 ml, po, TID, 2 years (1st–3rd month: 4 weeks/month; 4th–12th month: 14 days/month; 13th–24th month: 7 days/month) Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses Radiotherapy 50–56 Gy/time, 25–28 time/5–6 week	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses Radiotherapy 50–56 Gy/time, 25–28 time/5–6 week	Comparable (p > 0.05)
Dong et al. (2014)	40/40	–	TCM formula, 200 ml, po, BID, 84 days Etoposide, 80–120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60–80 mg/m2, ivgtt, day 1, 21 days*4 courses	Etoposide, 80–120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60–80 mg/m2, ivgtt, day 1, 21 days*4 courses	Comparable (p > 0.05)
Qiao et al. (2013)	40/40	TCM-C: limited disease: 16, extensive disease: 24; C: limited disease: 16, extensive disease: 24;	TCM formula, 200 ml, po, BID, 4 months Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days*4 courses Cisplatin, 20 mg/m2, ivgtt, day 1–5, 21 days*4 courses	Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days*4 courses Cisplatin, 20 mg/m2, ivgtt, day 1–5, 21 days*4 courses	Comparable (p > 0.05)
Zhang et al. (2018)	52/52	TCM-C: limited disease: 40, extensive disease: 12; C: limited disease: 41, extensive disease: 11;	TCM formula, 200 ml, po, BID, 84 days Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days*4 courses Cisplatin, 20 mg/m2, ivgtt, day 1–5, 21 days*4 courses	Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days*4 courses Cisplatin, 20 mg/m2, ivgtt, day 1–5, 21 days*4 courses	Comparable (p > 0.05)
Sun et al. (2017)	41/41	TCM-C: limited disease: 11, extensive disease: 30; C: limited disease: 19, extensive disease: 29;	TCM injection, 250 ml, ivgtt, qd, 28 days*2 courses 1st cycle (21 days): Carboplatin, 300 mg/m2, ivgtt, day 1, Etoposide, 100 mg/m2, ivgtt, day 2–6 2nd cycle (21 days): Cyclophosphamide 1000 mg/m2, ivgtt, day1,8 Doxorubicin 40 mg/m2, ivgtt, day 1 Cisplatin, 50 mg/m2, ivgtt, day 3–5	1st cycle (21 days): Carboplatin, 300 mg/m2, ivgtt, day 1, Etoposide, 100 mg/m2, ivgtt, day 2–6 2nd cycle (21 days): Cyclophosphamide 1000 mg/m2, ivgtt, day 1, 8 Doxorubicin 40 mg/m2, ivgtt, day1 Cisplatin, 50 mg/m2, ivgtt, day 3–5	Comparable (p > 0.05)
Zhang et al. (2013)	46/34	–	TCM formula, 300 ml, po, BID, 112 days Etoposide, 100 mg/m2, ivgtt, day 1–4, 28 days*4 courses Cisplatin, 35 mg/m2, ivgtt, day 1–4, 28 days*4 courses	Etoposide, 100 mg/m2, ivgtt, day 1–5, 28 days*4 courses Cisplatin, 35 mg/m2, ivgtt, day 1–5, 28 days*4 courses	Comparable (p > 0.05)
Ran et al. (2014)	60/60	Extensive disease	TCM formula, 130 ml, po, TID, 56 days Etoposide, 100 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Etoposide, 100 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day1, 21 days*4 courses	Comparable (p > 0.05)
Zhang et al. (2013)	36/42		TCM formula, 300 ml, po, BID, 2 years (1st–3rd month: 4 weeks/month; 4th–12th month: 14 days/month; 13th–24th month: 7 days/month) Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Comparable (p > 0.05)
Wu et al. (2015)	44/44	–	TCM formula, 300 ml, po, BID, 84 Days Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days* 6–8 courses Cisplatin, 75 mg/m2, ivgtt, day 1, 21 days* 6–8 courses	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days* 6–8 courses Cisplatin, 75 mg/m2, ivgtt, day 1, 21 days* 6–8 courses	Comparable (p > 0.05)
Wang et al. (2007)	26/38	TCM-C: limited disease: 20, extensive disease: 6; C: limited disease: 27, extensive disease: 11;	TCM patent medicine, 10 ml, po, TID, week 1–2, 3 weeks* 3–5 courses Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days* 3–5 courses Cisplatin 20 mg/m2, ivgtt, day 1–5, 21 days* 3–5 courses	Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days* 3–5 courses Cisplatin 20 mg/m2, ivgtt, day 1–5, 21 days* 3–5 courses	Comparable (p > 0.05)
Sun et al. (2008)	84/82	–	TCM pill, 2 g, TID, 9 weeks Cyclophosphamid, 800 mg/m2, ivgtt, day1 and day 8, 21 days* 3 courses Vincristine 2 mg/m2, ivgtt, day 1 and day 8, 21 days* 3 courses Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days* 3 courses Cisplatin 30 mg/m2, ivgtt, day 1–5, 21 days* 3 courses	Cyclophosphamid, 800 mg/m2, ivgtt, day 1 and day 8, 21 days* 3 courses Vincristine 2 mg/m2, ivgtt, day 1 and day 8, 21 days* 3 courses Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days* 3 courses Cisplatin 30 mg/m2, ivgtt, day 1–5, 21 days* 3 courses	Comparable (p > 0.05)
Xu et al. (2010)	42/46	–	TCM injection 250 ml, ivgtt, qd, 48 days Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days* 2 courses Cisplatin 20 mg/m2, ivgtt, day 1–5, 21 days* 2 courses	Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days* 2 courses Cisplatin 20 mg/m2, ivgtt, day 1–5, 21 days* 2 courses	Comparable (p > 0.05)
Wo et al. (2013)	37/37	Extensive disease	TCM formula, 250 ml, po, BID, 84 Days Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day 1, 21 days*4 courses	Etoposide, 120 mg/m2, ivgtt, day 1–3, 21 days*4 courses Cisplatin, 60 mg/m2, ivgtt, day1, 21 days*4 courses	Comparable (p > 0.05)
Shao et al. (1999)	92/48	–	TCM pill, 2.5 g, TID, 8 weeks Cyclophosphamid, 600 mg/m2, ivgtt, day 1 day 8, day 15, 21 days*2 courses Vincristine 2 mg/m2, ivgtt, qd, 21 days*2 courses Etoposide, 100 mg/m2, ivgtt, day1-5, 21 days*2 courses Cisplatin 20 mg/m2, ivgtt, day 1–7, 21 days*2 courses	Cyclophosphamid, 600 mg/m2, ivgtt, day 1 day 8, day 15, 21 days*2 courses Vincristine 2 mg/m2, ivgtt, qd, 21 days*2 courses Etoposide, 100 mg/m2, ivgtt, day 1–5, 21 days*2 courses Cisplatin 20 mg/m2, ivgtt, day 1–7, 21 days*2 courses	Comparable (p > 0.05)
Zhao et al. (2004)	64/62	–	TCM pill, 2.5 g, TID, 8 weeks Etoposide, 100 mg/m2, ivgtt, day 1–3, 21 days*3 courses Cisplatin, 25 mg/m2, ivgtt, day 1–3, 21 days*3 courses	Etoposide, 100 mg/m2, ivgtt, day 1–3, 21 days*3 courses Cisplatin, 25 mg/m2, ivgtt, day 1–3, 21 days*3 courses	Comparable (p > 0.05)
Fu et al. (2013)	16/14	Limited disease: 21, Extensive disease: 9	TCM powder, 5 g, BID, 28 days*2–4 courses Carboplatin, 300 mg/m2, ivgtt, day 1, 28 days*2–4 courses Etoposide, 100 mg/m2, ivgtt, day 1–3, 28 days*2–4 courses	Carboplatin, 300 mg/m2, ivgtt, day 1, 28 days*2–4 courses Etoposide, 100 mg/m2, ivgtt, day 1–3, 28 days*2–4 courses	Comparable (p > 0.05)
Dang et al. (2002)	20/19	Limited disease: 30, Extensive disease: 9	TCM formula, 250 ml, po, BID, 28 Days Vinorelbine 40 mg/m2, ivgtt, day 1 and day 8, 21 days*2 courses Cisplatin, 50 mg/m2, ivgtt, day 1–3, 21 days*2 courses	Vinorelbine 40 mg/m2, ivgtt, day 1 and day 8, 21 days*2 courses Cisplatin, 50 mg/m2, ivgtt, day 1–3, 21 days*2courses	Comparable (p > 0.05)
Li et al. (2001)	49/49	–	TCM formula, 250 ml, po, BID, 56 Days Carboplatin, 300 mg/m2, ivgtt, day 1, 21 days*4 courses Doxorubicin 40–50 mg/m2, ivgtt, day 1, 21 days*4 courses Etoposide, 100 mg/m2, ivgtt, day 3–5, 21 days*4 courses	Carboplatin, 300 mg/m2, ivgtt, day 1, 21 days*4 courses Doxorubicin 40–50 mg/m2, ivgtt, day 1, 21 days*4 courses Etoposide, 100 mg/m2, ivgtt, day 3–5, 21 days*4 courses	Comparable (p > 0.05)

Risk of bias and quality assessment

According to the Cochrane Handbook for Systematic Review of Interventions (Higgins and Green 2011) (Fig. 2), the methodological quality of three studies (3/22, 14%) presented a detailed description of how patients were randomized. No studies adequately reported the blinding of the investigator, patients, and assessor. The method of allocation concealment was reported in three studies. Incomplete outcome and selected reporting were reported in all studies.

Fig. 2.

Assessment of risk of bias

The results of the GRADE evaluation of studies which evaluated therapeutic effect are presented in Table 2. All the reasons for downgrading are labeled.

Table 2.

The results of the GRADE evaluation

Studies	Quality assessment						Quality
	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations
Dang (2002)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Dong (2014)	Randomised trials	Seriousb	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⊕⃝ MODERATE
Fu (2013)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Li (2001)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Qiao (2013)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Ran (2014)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Shao (1999)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Sun ( 2008)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Sun (2017)	Randomised trials	Seriousb	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⊕⃝ MODERATE
Wang (2009)	Randomised trials	Seriousb	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⊕⃝ MODERATE
Wang (2013)	randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	none	⊕⊕⃝⃝ LOW
Wang (2014)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	none	⊕⊕⃝⃝ LOW
Wu (2015)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Xu (2010)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Zhang (2018)	Randomised trials	Very seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Zhang (2013)	Randomised trials	Very seriousb	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝⃝ LOW
Zhao (2004)	Randomised trials	Seriousa	No serious inconsistency	No serious indirectness	No serious imprecision	None	⊕⊕⃝ MODERATE

^aThe study did not descrip the details of how patients were randomized and not report the blinding of the investigator, patients, and assessor

^bThe study did not report the blinding of the investigator, patients, and assessor

Efficacy and safety

All outcomes of meta-analysis are showed in Table 3.

Table 3.

Meta-analysis on outcomes

Outcome	Pooled RR (95% CI)	Z	P	Heterogeneity
				I2(%)	Ph
Therapeutic effect	1.295 (1.205–1.391)	7.06	0.000	0.0%	0.753
KPS	1.310 (1.210–1.418)	6.69	0.000	35.5%	0.134
1-Year survival rate	1.282 (1.129–1.456)	3.82	0.000	32.0%	0.184
3-Year survival rate	2.109 (1.514–2.939)	4.41	0.000	0.0%	0.798
5-Year survival rate	2.373 (1.227–4.587)	2.57	0.010	0.0%	0.862
Gastrointestinal reaction	0.786 (0.709–0.870)	4.64	0.000	0.0%	0.633
Bone marrow depression	0.837 (0.726–0.965)	2.45	0.014	10.0%	0.352

RR relative risk, CI confidence interval

Efficacy

Therapeutic effect

The therapeutic effect according to Response Evaluation Criteria in Solid Tumors (RECIST) between two groups was evaluated in 17 (Wang 2009, 2014; Fu and Zhang 2013; Li and Zheng 2001; Xu and Wei 2010; Shao 1999; Wu and Ma 2015; Dang et al. 2002; Ran et al. 2014; Dong et al. 2014; Sun and Shao 2008; Wang et al. 2013; Zhang et al. 2013; Zhao et al. 2004; Zhang and Wang 2018; Sun 2017; Qiao and Zhao 2013) studies. Figure 3 indicates that the therapeutic effect was enhanced in TCM-C group when compared with C group (RR = 1.30, 95% CI 1.21–1.40, P < 0.001) with low heterogeneity (I² = 0%, P = 0.753).

Fig. 3.

Meta-analysis on therapeutic effect

Quality of life (KPS scores)

There were nine (Wang 2009, 2014; Fu and Zhang 2013; Xu and Wei 2010; Shao 1999; Sun and Shao 2008; Wang et al. 2013; Zhao et al. 2004; Sun 2017) studies compared patients' quality of life by KPS scores before and after the treatments. Figure 4 shows that the KPS scores were enhanced in TCM-C group (RR = 1.31, 95% CI 1.21–1.42, P < 0.001) with low heterogeneity (I² = 35.5%, P = 0.142).

Fig. 4.

Meta-analysis on KPS scores

1-Year survival rate

Seven (Shao 1999; Zhang and Zhang 2013; Ran et al. 2010, 2011; Wo 2013; Sun and Shao 2008; Zhao et al. 2004) studies reported 1-year survival rate. Figure 5 shows that the 1-year survival rate in TCM-C group was higher than that in C group (RR = 1.28, 95% CI 1.13–1.46, P < 0.001) with low heterogeneity (I² = 32%, P = 0.184).

Fig. 5.

Meta-analysis on 1-year survival rate

3-Year survival rate

Seven (Shao 1999; Zhang and Zhang 2013; Ran et al. 2010, 2011; Wo 2013; Sun and Shao 2008; Zhao et al. 2004) studies reported 3-year survival rate. Figure 6 shows that the 3-year survival rate in TCM-C group was higher than that in C group (RR = 2.11, 95% CI: 1.51 ~ 2.94, P < 0.001) with low heterogeneity (I² = 0%, P = 0.798).

Fig. 6.

Meta-analysis on 3-year survival rate

5-Year survival rate

Five (Ran et al. 2010, 2011; Wo 2013; Sun and Shao 2008; Zhao et al. 2004) studies reported 5-year survival rate. Figure 7 indicates that the 5-year survival rate in TCM-C group was higher than in C group (RR = 2.37, 95% CI 1.23–4.59, P = 0.01) with low heterogeneity (I² = 0%, P = 0.862).

Fig. 7.

Meta-analysis on 5-year survival rate

Safety

Gastrointestinal reactions

Gastrointestinal reactions were recorded in 14 (Wang 2009; Fu and Zhang 2013; Li and Zheng 2001; Xu and Wei 2010; Zhang and Zhang 2013; Wu and Ma 2015; Dang et al. 2002; Ran et al. 2010, 2011, 2014; Wo 2013; Sun and Shao 2008; Wang et al. 2007; Zhang et al. 2013; Zhao et al. 2004) studies. Figure 8 shows that the incidence of gastrointestinal reactions in TCM-C group was lower than that in C group (RR = 0.79, 95% CI 0.71–0.8, P < 0.001) with low heterogeneity (I² = 0%, P = 0.633).

Fig. 8.

Meta-analysis on gastrointestinal reactions

Bone marrow depression

Bone marrow depression was recorded in six (Fu and Zhang 2013; Zhang and Zhang 2013; Ran et al. 2010, 2011; Wo 2013; Wang et al. 2007; Zhang et al. 2013) studies. Figure 9 shows that the incidence of bone marrow depressions in TCM-C group was lower than that in C group (RR = 0.84, 95% CI 0.73–0.96, P = 0.014) with low heterogeneity (I² = 10%, P = 0.352).

Fig. 9.

Meta-analysis on bone marrow depressions

Subgroup analysis

Outcomes of subgroup analysis are showed in Table 4.

Table 4.

All outcomes of subgroup analysis

Subgroups	Number of studies	Pooled RR (95% CI)	Z	P	Heterogeneity
					I2 (%)	Ph
Therapeutic effect
Formula	9 (Fu and Zhang 2013; Li and Zheng 2001; Zhang and Zhang 2013; Wu and Ma 2015; Dang et al. 2002; Ran et al. 2014; Dong et al. 2014; Zhang et al. 2013; Qiao and Zhao 2013)	1.39 (1.23, 1.58)	5.27	0.000	0.0%	0.651
Chinese paten drug	3 (Shao 1999; Sun and Shao 2008; Zhao et al. 2004)	1.22 (1.12, 1.34)	4.37	0.000	0.0%	0.907
Injection	5 (Wang 2009, 2014; Xu and Wei 2010; Wang et al. 2013; Sun 2017)	1.24 (1.04, 1.47)	2.45	0.014	0.0%	0.433
KPS
Formula	1 (Fu and Zhang 2013)	1.17 (0.72, 1.89)	0.63	0.535	–	–
Chinese paten drug	3 (Shao 1999; Sun and Shao 2008; Zhao et al. 2004)	1.26 (1.15, 1.39)	4.78	0.000	74.4%	0.02
Injection	5 (Wang 2009, 2014; Xu and Wei 2010; Wang et al. 2013; Sun 2017)	1.39 (1.21, 1.59)	4.70	0.000	0.0%	0.975
Gastrointestinal reaction
Formula	9 (Fu and Zhang 2013; Li and Zheng 2001; Zhang and Zhang 2013; Dang et al. 2002; Ran et al. 2010, 2011, 2014; Wo 2013; Zhang et al. 2013)	0.85 (0.72, 1.00)	2.00	0.045	0.0%	0.640
Chinese paten drug	3 (Sun and Shao 2008; Wang et al. 2007; Zhao et al. 2004)	0.74 (0.64, 0.85)	4.08	0.000	28.4%	0.248
Injection	2 (Wang 2009; Xu and Wei 2010)	0.76 (0.57, 1.02)	1.80	0.072	65.8%	0.087

RR relative risk, CI confidence interval

We performed subgroup analysis for the administration of TCM. Subgroup analysis (Figs. 10, 11, 12) indicated that the therapeutic effect was enhanced in Formula subgroup (RR = 1.39, 95% CI 1.23–1.58, P = 0.000, I² = 0.0%), Chinese paten drug subgroup (RR = 1.22, 95% CI 1.12–1.34, P = 0.000, I² = 0.0%), and Injection subgroup (RR = 1.24, 95% CI 1.04–1.47, P = 0.014, I² = 0.0%); KPS scores were enhanced both in Chinese paten drug subgroup (RR = 1.26, 95% CI 1.15–1.39, P = 0.000, I² = 74.4%) and injection subgroup (RR = 1.39, 95% CI 1.21–1.59, P = 0.000, I² = 0.0%); the incidence of gastrointestinal reactions was decreased in Formula subgroup (RR = 0.85, 95% CI 0.71–1.00, P = 0.045, I² = 0.0%) and Chinese paten drug subgroup (RR = 0.74, 95% CI 0.64–0.85, P = 0.000, I² = 28.4%). The number of studies of Formula subgroup in KPS scores and the number of studies of injection subgroup are too few to be performed subgroup analysis.

Fig. 10.

Subgroup analysis of therapeutic effect

Fig. 11.

Subgroup analysis of KPS

Fig. 12.

Subgroup analysis of gastrointestinal reactions

Publication bias

It showed some evidence of publication bias as indicated by the Egger test (P = 0.026) and funnel plots for the primary outcome (Figs. 13, 14).

Fig. 13.

Egger test of primary outcome

Fig. 14.

Funnel plot of primary outcome

Sensitivity analysis

Sensitivity analysis was performed by sequentially omitting each study to examine the robustness of the primary outcome. It is suggested that the pooled RR values of therapeutic effect, quality of life (KPS scores), 1-year survival rate, 3-year survival rate, 5-year survival rate, gastrointestinal reactions, and bone marrow depression are stable (Figs. 15, 16, 17, 18, 19, 20, 21).

Fig. 15.

Sensitivity analysis of therapeutic effect

Fig. 16.

Sensitivity analysis of KPS

Fig. 17.

Sensitivity analysis of 1-year survival rate

Fig. 18.

Sensitivity analysis of 3-year survival rate

Fig. 19.

Sensitivity analysis of 5-year survival rate

Fig. 20.

Sensitivity analysis of gastrointestinal reactions

Fig. 21.

Sensitivity analysis of bone marrow depression

Discussion

As we know, the treatment for SCLC has not changed recently. Compared with cyclophosphamide plus doxorubicin and vincristine (developed in the 1970s), combination of platinum, etoposide, and topotecan which were used as first-line and second-line drugs, respectively, have shown no survival benefits (Waqar and Morgensztern 2017). To achieve greater survival benefits, TCM as adjuvant to chemotherapy is widely used for SCLC in China. This meta-analysis of RCT studies showed that TCM therapy as adjuvant to chemotherapy might be beneficial for SCLC in enhancing therapeutic effect, quality of life, and prolonging survival time. Meantime, patients who received TCM may have less chemotherapy-related side effects.

To our knowledge, our study is the first systematic review and meta-analysis to evaluate the efficacy and safety of TCM as adjuvant to chemotherapy for SCLC. Although we observed a favorable result in improving efficacy, some aspects need to be considered when discussing our findings. First, most of the TCM interventions mentioned in this study are not the same. According to the theory of TCM, a personalized treatment plan should be used based on the patient's physical condition. Therefore, we are referring to the general concept of TCM, not a single TCM formula or a single herb. For therapeutic effect, we also did a subgroup analysis dependent on different kinds of TCM. The results suggested that compared with chemotherapy alone, formula, Chinese paten drug and injection all have better therapeutic effects for SCLC patients.

Main chemotherapy-related side effects include bone marrow suppression and gastrointestinal reaction. According to our results, the incidence of bone marrow suppression and gastrointestinal reaction were both less in patients receiving TCM during or after chemotherapy. TCM which was considered as a mean of adjuvant therapy does not increase the adverse effects; it, however, reduced incidence of chemotherapy-related side effects. Because the evaluation criteria adopted by each study were different, we only analyzed the difference in the incidence of chemotherapy-related side effects. We did not further analyze the difference in degrees of side effects.

There are some limitations in this study. First, the search strategy was performed and cross-checked independently by two researchers in two databases, but our analysis may still miss a few eligible studies. Moreover, we only searched studies published in Chinese and English. Some studies published in Korean, Japanese, or other languages might be missed in our analysis. Second, the quality of randomized-controlled trials included is not high. Although the quality of the included studies is not very high, the results of our study are relatively credible and reliable based on the results of sensitivity analysis. And we believe that there will be more and more high-quality research in the future, and we will update our research regularly. According to the funnel plot and Egger test, there is publication bias of this systematic review. This may because a positive result is more likely to be published during the research publication process. Third, SCLC is divided into two stages, extensive stage and limited stage, but this study did not conduct a subgroup analysis about different stages. Although disease stage was mentioned in ten studies (Fu and Zhang 2013; Dang et al. 2002; Ran et al. 2010, 2014; Wo 2013; Wang et al. 2007; Zhang and Wang 2018; Sun 2017; Qiao and Zhao 2013; Higgins and Green 2011), the statistical results according to different disease stages were only conducted in four studies (Ran et al. 2010, 2014; Wo 2013; Higgins and Green 2011). Therapeutic effect was only conducted in one study (Higgins and Green 2011) (extensive disease; RR = 1.36, 95% CI 1.04–1.79). Therefore, we cannot further analyze which staged patients can benefit more from the traditional Chinese medicine.

Conclusion

According to this systematic review and meta-analysis, we can draw a conclusion that TCM combined with chemotherapy may improve therapeutic effect, quality of life, and prolong survival time. More RCTs with more participants and high quality are warranted to support our findings.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

SCLC

Small cell lung cancer

RCTs

Randomised-controlled trials

RR

Relative risk

MD

Mean difference

95% Cis

95% Confidence intervals

C

Group chemotherapy

TCM-C

Group chemotherapy plus TCM

LS-SCLC

Limited-stage small cell lung cancer

ES-SCLC

Extensive small cell lung cancer

TCM

Traditional Chinese medicine

RECIST

Response Evaluation Criteria in Solid Tumors

CNKI

Chinese National Knowledge Infrastructure

Author contributions

BH put forward the idea of research. Search strategy was developed and conducted by SC, JX, and YB. XZ and SH independently screen the titles and abstracts of all included studies. Data extraction was performed by ZZ, RL, XZ, and HZ. QG, RQ, and JJ conducted a meta-analysis. Manuscript was written by SC.

Funding

This study is supported by the National Natural Science Foundation (81774294) and the National Natural Science Foundation (81673961).

Compliance with ethical standards

Conflict of interest

The authors declare that they have no competing interests.

Ethical approval

Not applicable.

Consent for publication

Not applicable.