Acupuncture and moxibustion for malignant tumor patients with psychological symptoms of insomnia, anxiety and depression: a systematic review and Meta-analysis

Abstract

OBJECTIVE: To evaluate the efficacy and safety of acupuncture and moxibustion therapy (AMT) for cancer-related psychological symptoms (CRPS) of insomnia, depression and anxiety.

METHODS: Seven databases were searched for randomized controlled trials (RCT) comparing AMT to routine care or conventional drug for alleviating CRPS of insomnia, depression, and anxiety before April 2020. Two independent reviewers performed the data extraction and assessed the risk of bias.

RESULTS: A total of 30 RCTs involving 2483 cancer patients were enrolled. The pooled analysis indicated that the treatment group was significantly better than the control group in improving the depression effective rate [RR= 1.29, 95% CI(1.12, 1.49), P =0.0004], the quality of life (QOL) [MD =1.11, 95% CI(0.80, 1.42), P < 0.000 01], and reducing Self-rating Anxiety Scale (SAS) [MD =﹣7.75, 95% CI(﹣10.44, ﹣5.05), P < 0.000 01]. But there was no statistically significant difference between two groups in improving the insomnia effective rate [RR= 1.18, 95% CI(0.93, 1.51), P =0.18]. The subgroup analysis showed the effectiveness of different intervention on CRPS. Compared with routine care, AMT helps relieve CRPS better evaluated by Pittsburgh Sleep Quality Index (PSQI), Hamilton Depression Scale (HAMD), and Self-rating Depression Scale (SDS), and depression effective rate. Compared with conventional drug, AMT performs better evaluated by SDS, depression effective rate and QOL. Moreover, the conventional drug showed higher treatment efficacy on improving insomnia effective rate compared with AMT. Compared to conventional drug, AMT plus conventional drug resulted in a significant reduction on CRPS such as PSQI, HAMD, SDS, and SAS, and also had a meaningful improvement on insomnia effective rate, depression effective rate and QOL. Fewer published reports were found on the adverse events of AMT than the conventional drug.

CONCLUSION: The results suggested that AMT might be effective in improving CPRI; however, a definite conclusion could not be drawn because the quality of trials are low. Further large-scale and high-quality RCTs to verify the efficacy and safety of AMT on CRPS are still warranted.

1. INTRODUCTION

As one of the leading causes of morbidity and mortality worldwide, cancer is steadily increasing in rate, resulting in a huge global burden over the past 30 years. The global cancer burden is expected to be 28.4 million cases in 2040, a 47% rise from 2020.¹ Meanwhile cancer survivors grow rapidly with ongoing therapy. More than 3% of the adult in Western countries have survived for 5 years as cancer cases. By 2022, 64% of 18 million cancer cases in the United States are estimated to survive 5 years or more; 40% survive 10 years or more and 15% survive 20 years or more after diagnosis.² 4.3 million new cancer cases and 2.9 million new cancer deaths estimated occurred in China in 2018,³ with 36.9% of cancer patients surviving at least 5 years after diagnosis.⁴ Surviving cancer patients generally suffer from various symptoms and expect long-term survival as a chronic disease.⁵ On account of cancer invasion or its treatment, cancer treatments increase the risk for numerous late effects, thus making cancer survivors psychologically and medically vulnerable.⁶

Cancer-related psychological symptoms (CRPS) includes highly prevalent symptoms like depression, anxiety and insomnia with the overall prevalence at 30%-40%.^{7,⇓,⇓,⇓- 11} Untreated depression and ongoing insomnia are associated with decreased immune responses and lower survival rates.¹² CRPS links to cancer incidence and mortality risk, which presented a negative correlation with the quality of life (QOL) and positive correlation with physical symptoms.^{13,⇓- 15} Failing to identify and treat CRPS promptly may lead to serious consequences on QOL in cancer patients,¹⁶ which can be alleviated with appropriate interventions.

There are at risk for antidepressant, anxiolytic and hypnotic use up in five years after the diagnosis of cancer patients.¹⁷ However, pharmacological treatments of CRPS are not curative and in the meanwhile they often result in additional side effects (dizziness, residual daytime sedation, fatigue, daytime sleepiness, and headache).^18,19 With the study of drug-drug interaction between antidepressants and cancer treatment, direct consequences have shown that patients on tamoxifen and antidepressant have doubled the risk of relapse to cancer in two years.¹² Financial distress is also an important and common factor contributing to the suffering of advanced cancer patients and their caregivers.²⁰ Therefore cancer patients and providers alike are interested in evidence-based nonpharmacologic alternatives like acupuncture and moxibustion to relieve these symptoms.

Acupuncture is performed by taking fine needles to insert in the acupoints and stimulate needling response, either manually or electrically, to treat side effects of conventional cancer treatment. Moxibustion involves burning the moxa on or above the skin at acupoints with stick-on moxa, sandwiched moxibustion or indirect moxibustion.²¹ Acupuncture and moxibustion therapy (AMT), originated from the system of Traditional Chinese Medicine, has been in use at least 2500 years in China and other Asian countries.²¹ AMT has a good effect on CRPS and is easy to operate, safe, and economical. National Cancer Institute (NCI) have recommended “Acupuncture for Cancer Symptom Management” to target common symptom clusters such as pain, sleep disturbance, fatigue, chemotherapy side effects, psychological distress, and impairment in quality of life.^{22,⇓,⇓,⇓,⇓,⇓,⇓- 29} As effective interventions, acupuncture and moxibustion have been confirmed to improve financial distress and relieve the effect of CRPS on the quality of life of advanced cancer patients.²²

Although three published reviews have reported the effects of acupuncture and moxibustion on CRPS,^30,⇓-32 there is no Meta-analysis of AMT on CRPS. Meanwhile the included studies in the three published reviews are in low quality without enough recent studies. With the rapid development in recent years, more related studies with high quality have been published. Therefore, this study aims to screen recent studies on that issue for a systematic review and Meta-analysis, thus examining the clinical efficacy and safety of AMT on CRPS of insomnia, anxiety and depression systematically.

2. METHODS

2.1. Protocol register

The protocol for this systematic review was registered in the Prospective Register of Systematic Reviews (PROSPERO): CRD42020197084.

2.2. Search strategy

The following databases were searched from their inception until April 15, 2020: PubMed, EMBASE, Cochrane Library, China National Knowledge Infrastructure Database (CNKI), China Science and Technology Journal Database (VIP), Chinese Biomedical Literature Database (CBM) and Wanfang Database. The complete manuscripts of all relevant studies published in English and Chinese were retrieved. The search terms included (acupuncture OR single needle OR electric acupuncture OR blood-letting puncture OR head acupuncture OR scalp acupuncture OR ear needle OR auricular acupuncture OR dermal needle OR acupoint catgut embedding OR fire needle OR meridian OR acupotomy OR moxibustion method OR moxibustion OR direct moxibustion OR indirect moxibustion OR heat-sensitive moxibustion OR needle warming moxibustion OR ginger-separated moxibustion OR salt-separated moxibustion) AND (cancer OR tumour OR neoplasms OR malignancy OR carcinoma) AND (sleep disorder OR sleep disturbance OR insomnia OR depression OR depressive disorder OR anxiety).

2.3. Inclusion and exclusion criteria

2.3.1 Types of studies

To collect evidence of high quality, only randomized controlled trials (RCTs) were included in this systematic review. Only full articles were included.

2.3.2 Types of participants

The participants who were diagnosed with the psychological symptoms like insomnia, anxiety and/or depression due to interior-stirring by cancer would be included. Additionally, patients with mental disorders would be excluded. Serious organic disease and complications or pregnancy would not be considered.

2.3.3 Types of intervention

All of the cancer patients with psychological symptoms were required to undergo basic cancer treatment in both the treatment and control groups.

Treatment interventions. In the treatment group, the intervention would be AMT combined with/without conventional drug or routine care. These included studies could perform the forms of acupuncture (e.g., manual acupuncture, electro-acupuncture, intradermal needling and scalp-acupuncture) and/or moxibustion (e.g., direct or indirect moxibustion, heat-sensitive moxibustion, warm needling, ginger-separated moxibustion or salt-separated moxibustion). The other forms of AMT, such as acupressure, transcutaneous electrical nerve stimulation, point injection, laser irradiation, far infrared moxibustion, auricular acupressure, auricular acupuncture and cupping would be excluded.

The conventional drug would be the regular and frequent use on the treatment of insomnia, depression, and anxiety of CRPS such as paroxetine, sertraline, diosmin tablets, estazolam or diazepam. The routine care included nursing intervention, explanation of disease knowledge, life guidance and psychological health education.

Control interventions. In the control group, the patients would receive conventional drug or routine care that prevent the psychological symptoms like insomnia, anxiety and/or depression. Placebo controls or sham acupuncture for the purpose of blinding are encouraged.

The following treatment comparisons would be studied:

Acupuncture and/or moxibustion compared with routine care;

Acupuncture and/or moxibustion compared with conventional drug;

Acupuncture and/or moxibustion plus conventional drug compared with conventional drug.

2.3.4 Types of outcomes

2.3.4.1 Primary outcomes

(a) the effective rate of the psychological symptoms like insomnia, anxiety or depression.^30,⇓-32

The effective rate of insomnia was used to assess the sleep effectiveness. Recovery: sleep time returned normal or more than 6 h a night, with deep sleep and wake up energetically; Remarkable effect: sleep quality was obviously improved with sleep time and depth sleep increased; Effective: sleep quality was improved but some symptoms still existed; Invalid effect: no significant improvement or reverse aggravation of insomnia after treatment. The effective rate of insomnia = recovery rate + marked effective rate.

The HAMD reduced rate was used to assess the therapeutic effectiveness as the effective rate of depression. HAMD reduced rate (%) = (score before treatment-score after treatment) /score before treatment × 100%, cure rate (reduced rate>75%), effective rate (reduced rate 50%-75%), improved rate (reduced rate 25%-49%), invalid rate (reduced rate<25%). The effective rate of depression = recovery rate + marked effective rate.

(b) Scales or indices for the quality evaluation of the psychological symptoms like insomnia, anxiety or depression, the Pittsburgh Sleep Quality Index (PSQI), Hamilton Depression Scale (HAMD), Self-rating Depression Scale (SDS), Self-rating Anxiety Scale (SAS).

2.3.4.2 Secondary outcomes

The secondary outcomes of this review mainly included the quality of life (QOL) and adverse events.

2.4. Selection of studies and data extraction

Two authors (MA Fangfang and LI Bingxue) screened all the records independently and abstracted all related data from the enrolled papers, including first author’s name, country, year of publication, age, gender, sample size, diagnosis standard, details of interventions for treatment groups and control groups, cointerventions, outcome measures, treatment duration, and adverse events. Disagreements were resolved by discussion or consulting a senior researcher (WANG Xiaomin).

2.5. Quality assessment

The methodological quality was assessed according to the Cochrane Handbook version 5.3. Two authors (MA Fangfang and LI Bingxue) independently extracted the data from the included studies. The risk of bias quality was assessed based on 7 domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias. There were three potential bias judgments according to the Cochrane criteria after assessing all the domains: (a) low risk; (b) high risk; and (c) unclear risk. Disagreements were resolved by discussion between all of the authors.

2.6. Statistical analysis

For dichotomous data, we present the treatment effect as relative risk with 95% confidence intervals (CIs). For continuous data, we used the mean difference (MD) with 95% CIs to measure the treatment effect. In the case of outcome variables with different scales, we used the standardized mean difference (SMD) with 95% CIs. The χ² test for heterogeneity and the I² test were used to evaluate the heterogeneity of the included studies. When P≥ 0.10, I²≤ 50%, there was no statistical heterogeneity between the results included in the study and the fixed effect model was used for analysis. If P < 0.10, I²>50%, it was considered as statistically different. Then we made an analysis on the heterogeneity sources to determine whether the sensitivity analysis and subgroup analysis were needed, whether random effects models can be used for analysis or not. Descriptive analysis was used if the clinical heterogeneity was too large or the literature was too few to perform Meta-analysis. Methodological quality evaluation was performed by the bias risk map drawn by Revman 5.3 software (Cochrane Library, London, UK). When there were more than 10 articles on a certain outcome indicator, an inverted funnel plot was used to assess the published bias of the included study. All the analyses were done using Revman 5.3.

3. RESULTS

3.1. Literature Search Results

A total of 2474 studies were retrieved at the initial search. After removing 799 duplicates, 1675 studies were identified for further analysis. Through screening the titles and abstracts, 1615 studies were excluded because they were literature reviews, expert opinions, commentaries, case reports, case series, nonclinical studies, animal research or irrelevant intervention. Of the remaining 60 studies, by reading the full text, 30 studies met our inclusion criteria, as shown in Figure 1.

Figure 1. Flow diagram of studies included in the review.

CNKI: China National Knowledge Infrastructure Database; VIP: China Science and Technology Journal Database; CBMdisc: China Biology Medicine disc; EMbase: Excerpt Medica Database; RCT: randomized controlled trial.

3.2. Characteristics of included studies

The basic characteristics of the included studies are listed in Table 1.^{33,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓- 62} By integrating the 30 included studies, a total of 2483 cases were reported with 1235 cases in the treatment group and 1248 cases in the control group. The publication period was from 2008 to 2020. One study was published in English and the other 29 studies were published in Chinese. Six studies were for female breast cancer patients. All of the included patients were concentrated in China with a wide variation in the age of subjects (18-80 years). All studies included at least one of the main outcome indicators, which corresponding to the diagnosis standards of the psychological symptoms like insomnia, anxiety and/or depression due to interior-stirring by cancer.

Table 1.

Basic characteristics of the included studies

Study		Sample size (T/C)		Age year (T/C)		Gender (Male /Female)							Disease diagnosis					Control						Intervention			Main treated acupoints and treatment frequency			Course		Outcome measure
Chen YH et al 201333		30/30		T: 38-83 C: 24-82		T: 18/12 C: 20/10							Gastric cancer +anxiety and depression					Routine care						Moxibustion+C			MT: Zusanli (ST36), Xuehai (SP10), Sanyinjiao (SP6) and Zhongwan (CV12). 1 time per day for 30 min			2 w		SDS, SAS
Chen J 201834		18/18		T: 66.43±5.43 C: 65.13±5.98		T: 10/8 C: 9/9							Tumor+depression					Paroxetine 20-60 mg/d+anti-tumor therapy						Acupuncture +moxibustion+c			AT: Neiguan (PC6), Shenmen (TF4), Yifeng (SJ17), Baihui (GV20), Sishencong (EX-HN1), Hegu (LI4) and Taichong (LR3). 1 time per day for 30 min MT: Guanyun (CV4) and Zusanli (ST36). 1 time per day for 30 min			6 w		HAMD, depression efficiency
Dan ZJ 201335		27/28		T: 52.222±7.153 C: 51.296±6.866		T: 14/13 C: 16/12							Tumor+insomnia					Estazolam 1mg/d						Acupuncture +boxed moxibustion			AT: Baihui (GV 20), Shenting (GV24), Yintang (EX-HN3), Shenmen (TF4), Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day for 30 min MT: Shenque (CV8) and Guanyun (CV4). 1 time per day for 30 min			7 d		Sleep efficiency, PSQI
Deng XY et al 201836		30/30		T: 48.93±4.03 C: 50.13±4.09		T: 12/18 C: 16/14							Tumor+depression					Sertraline 10 mg/d+ anti-tumor therapy						Acupuncture+Electric acupuncture			AT: Yinlingquan (SP9), Fenglong (ST40), Hegu (LI4), Neiguan (PC6), Taichong (LR3), Sanyinjiao (SP6), Yintang (EX-HN3) and Baihui (GV 20). 2 times per week for 30 min EA: Yinlingquan (SP9) and Fenglong (ST40). 2 times per week for 30 min			4 w		HAMD, QLQ-C30
Deng XY et al 201937		30/30		T: 53±9 C: 49±11		T: 11/19 C: 10/20							Tumor+depression					Sertraline 50 mg/d						Acupuncture+C			AT: Hegu (LI4), Taichong (LR3), Neiguan (PC6), Shenmen (TF4), Zusanli (ST36), Sanyinjiao (SP6), Yintang (EX-HN3) and Baihui (GV 20). 2 times per weekfor 30 min			4 w		HAMD, QLQ-C30
Feng Y et al 201138		40/40		T: 63.80±5.47 C: 63.60±4.26		T: 26/14 C: 27/13							Malignant neoplasms+insomnia and depression					Fluoxetine 20 mg/d						Acupuncture+C			AT: Fenglong (ST40), Yinlingquan (SP9), Xuehai (SP10), Sanyinjiao (SP6), Yintang (EX-HN3), Baihui (GV 20), Sishencong (EX-HN1), Neiguan (PC6) and Shenmen (TF4). 1 time per day for 20-30 min			30 d		SDS, HAMD, depression efficiency, PSQI
Feng XT et al 201939		70/70		35-78 (61.51±6.24)		61/79							Liver cancer+anxiety and depression					Paroxetine 20 mg/d						Acupuncture +chinese herb+c			AT: the acupoints of liver meridian and heart meridian, Shenmen (TF4), Danzhong (CV 17), Taichong (LR3), Qimen (LR14) and Xingjian (LR2). 5 times per weekfor 30 min			8 w		SDS, SAS
Liao LL 201940		35/34		T: 63.63±16.30 C: 63.18±15.04		T: 14/21 C: 13/21							Tumor+insomnia					Routine care						Intradermal needling +moxibustion			IN: Yintang (EX-HN3) and Shenmen (TF4). 2~3 times per week; 3~4 times per day (acupressing 1~2 min per point) MT: Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day (5 min per point)			2 w		PSQI
Liu YL 201741		30/30		T: 51.63±9.14 C: 54.63±9.89		T: 10/20 C: 12/18							Tumor+depression					Routine care						Thermal moxibustion			MT: Baihui (GV 20), Shenmen (TF4), Neiguan (PC6), Taichong (LR3) and Ganshu (BL18). 1 time per day (20 min per point)			21 d		HAMD, depression efficiency
Liu XF et al 201942			40/40		T: 46.37±14.53 C: 49.32±13.22	Female						Breast cancer+anxiety and depression					Diosmin tablets 2.7 g/d						Needle warming moxibustion				Jianyu (LI15), Waiguan (SJ5), Yinlingquan (SP9), Quchi (LI11), Shuifen (IN9) and Zusanli (ST36). 1 time per 2 days for 30 min			28 d		SDS, SAS, QLQ-C30
Liu YP et al 201943			40/40		T: 63±13 C: 63±12	T: 24/16 C: 20/20						Tumor+depression					Escitalopram 10 -20 mg/d						Acupuncture+C				AT: Taichong (LR3), Hegu (LI4), Baihui (GV 20) and Yintang (EX-HN3). 5 times per week for 30 min			12 w		HAMD
Mo ZY et al 201744			58/60		T: 48.22±10.28 C: 48.60±11.02	Female						Breast cancer+anxiety and depression					Routine care						Ginger moxi-bustionwithbox				MT: Hegu (LI4), Zusanli (ST36), Taichong (LR3) and Yongquan (KI1). 2 times per day for 30 min			7 d		SDS, SAS
Pei Y et al 201045			31/36		T: 51.76±10.21 C: 48.34±8.79	Female						Breast cancer+anxiety and depression					Citalopram 20 mg/d						Acupuncture				AT: Feishu (BL13), Xinshu (BL15), Geshu (BL17), Ganshu (BL18), Pishu (BL20) and Shenshu (BL23). 5 times per week for 30 min			6 w		SDS, SAS, HAMD
Peng XH et al 201746			93/97		T: 59.36±12.54 C: 60.95±11.11	T: 68/36 C: 68/36						Tumor+insomnia					Estazolam 1 mg/d						Acupuncture +moxibustion				AT: Baihui (GV 20), Shenting (GV24), Yintang (EX-HN3), Shenmen (TF4), Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day for 30 min MT: Shenque (CV8) and Guanyun (CV4). 1 time per day for 30 min			7 d		Sleep efficiency
Shen LP et al 201647			50/50		T: 54. 93±14.45 C: 58. 09±12.11	T: 28/22 C: 31/19						Lung cancer+insomnia					Zolpidem 10 mg/d+ Paroxetine 20 mg/d (if needed)						Acupuncture+Electric acupuncture+C				AT: Zhaohai (KI6), Shenmai (BL62), Shenmen (TF4), Yintang (EX-HN3), Sishencong (EX-HN1), Anmian(Extra 8), Hegu (LI4), Taichong (LR3), Zusanli (ST36), Qihai (CV6) and Quchi (LI11). 1 time per day for 30 min EA: Shenmen (TF4), Anmian(Extra 8), Sishencong (EX-HN1), Hegu (LI4), Taichong (LR3), Zusanli (ST36), Qihai (CV6) and Quchi (LI11). 1 time per day for 30 min			4 w		SDS, SAS, PSQI
Shi Y et al 201448			40/40		NR	NR						Tumor+insomnia					Diazepam 5 mg/d						Scalp-acupuncture				AT: Baihui (GV 20), Sishencong (EX-HN1). 1 time per day for 60 min SA: Ezhongxian (MS2~3). 1 time per day for 60 min			7 d		PSQI
Song JR et al 201549			60/60		NR	NR						Tumor+insomnia					Estazolam 1 mg/d						Acupuncture boxed moxibustion				AT: Baihui (GV20), Shenting (GV24), Yintang (EX-HN3), Shenmen (TF4), Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day for 30 min MT: Shenque (CV8) and Guanyun (CV4). 1 time per day for 30 min			7 d		Sleep efficiency, PSQI
Liu XF et al 201942			40/40		T: 46.37±14.53 C: 49.32±13.22			Female			Breast cancer+anxiety and depression					Diosmin tablets 2.7 g/d					Needle warming moxibustion					Jianyu (LI15), Waiguan (SJ5), Yinlingquan (SP9), Quchi (LI11), Shuifen (IN9) and Zusanli (ST36). 1 time per 2 days for 30 min			28 d		SDS, SAS, QLQ-C30
Liu YP et al 201943			40/40		T: 63±13 C: 63±12			T: 24/16 C: 20/20			Tumor+depression					Escitalopram 10-20 mg/d					Acupuncture+C					AT: Taichong (LR3), Hegu (LI4), Baihui (GV 20) and Yintang (EX-HN3). 5 times per week for 30 min			12 w		HAMD
Mo ZY et al 201744	58/60				T: 48.22±10.28 C: 48.60±11.02		Female		Breast cancer+ anxiety and depression						Routine care							Ginger moxi-bustionwithbox			MT: Hegu (LI4), Zusanli (ST36), Taichong (LR3) and Yongquan (KI1). 2 times per day for 30 min			7 d			SDS, SAS
Pei Y et al 201045	31/36				T: 51.76±10.21 C: 48.34±8.79		Female		Breast cancer+anxiety and depression						Citalopram 20 mg/d							Acupuncture			AT: Feishu (BL13), Xinshu (BL15), Geshu (BL17), Ganshu (BL18), Pishu (BL20) and Shenshu (BL23). 5 times per week for 30 min			6 w			SDS, SAS, HAMD
Peng XH et al 201746	93/97				T: 59.36±12.54 C: 60.95±11.11		T: 68/36 C: 68/36		tumor+insomnia						Estazolam 1 mg/d							Acupuncture +moxibustion			AT: Baihui (GV 20), Shenting (GV24), Yintang (EX-HN3), Shenmen (TF4), Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day for 30 min MT: Shenque (CV8) and Guanyun (CV4). 1 time per day for 30 min			7 d			Sleep efficiency
Shen LP et al 201647	50/50				T: 54. 93±14.45 C: 58. 09±12.11		T: 28/22 C: 31/19		Lung cancer+insomnia						Zolpidem 10 mg/d+ Paroxetine 20 mg/d (if needed)							Acupuncture+Electric acupuncture+C			AT: Zhaohai (KI6), Shenmai (BL62), Shenmen (TF4), Yintang (EX-HN3), Sishencong (EX-HN1), Anmian(Extra 8), Hegu (LI4), Taichong (LR3), Zusanli (ST36), Qihai (CV6) and Quchi (LI11). 1 time per day for 30 min EA: Shenmen (TF4), Anmian(Extra 8), Sishencong (EX-HN1), Hegu (LI4), Taichong (LR3), Zusanli (ST36), Qihai (CV6) and Quchi (LI11). 1 time per day for 30min			4 w			SDS, SAS, PSQI
Shi Y et al 201448	40/40				NR		NR		Tumor+insomnia						Diazepam 5 mg/d							Scalp-acupuncture			AT: Baihui (GV 20), Sishencong (EX-HN1). 1 time per day for 60 min SA: Ezhongxian (MS2-3). 1 time per day for 60 min			7 d			PSQI
Song JR et al 201549	60/60				NR		NR		Tumor+insomnia						Estazolam 1 mg/d							Acupuncture boxed moxibustion			AT: Baihui (GV20), Shenting (GV24), Yintang (EX-HN3), Shenmen (TF4), Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day for 30 min MT: Shenque (CV8) and Guanyun (CV4). 1 time per day for 30 min			7 d			Sleep efficiency, PSQI
Wang Y 201550	40/40				NR		Female		Breast cancer+anxiety and depression						Routine care							Five-element acupuncture+ wheat grain moxibustion			The acupoints according to five-element acupuncture therapy. 1 time per week			6 w			SDS, SAS
Wang BN et al 201951	40/40				T: 52.47±9.62 C: 50.16±7.34		Female		Breast cancer+insomnia						Routine care							Intradermal needling			IN: Shenmen (TF4), Sanyinjiao (SP6), Anmian (Extra 8), Zusanli (ST36) and Taibai (SP3). 5 times per week; 3 times per day (acupressing 1 min per point)			4 w			PSQI
Wang WG et al 202052	40/40				T: 39.52±8.35 C: 39.36±9.27		T: 24/16 C: 22/18		Brain tumor+ Depression								Routine care			Acupuncture					AT: Neiguan (PC6), Renzhong (GV26), Sanyinjiao (SP6), Chize (LU5), Jiquan (HT1), Fengchi (GB20), Weizhong (BL40), Tianzhu (BL10) and Wangu (BL12). 2 time per day for 30 min			4 w			SDS, SAS
Wei S 201953			33/33		49.21±6.32		Female							Gynecological malignant tumor+insomnia					Alprazolam 0.4 mg/d	Thunder-fire moxibustion+C					MT: Shenmen (TF4), Neiguan (PC6), Sanyinjiao (SP6)，Xinshu (BL15), Pishu (BL20), Zusanli (ST36), Danshu (BL19) and Qiuxu (GB40). 1 time per day for 30 min			2 w			PSQI, Sleep efficiency, QLQ-C30
Xia Q et al 201754			23/23		NR		18/28							Tumor+depression					fluoxetine 20 mg/d	Scalp-acupuncture					SA: Dingzhongxian (MS5), Ezhongxian (MS1), Epangxian(MS2~ MS4), Nieqianxian (MS10), Niehouxian (MS11). 6 times per week for 30 min			4 w			HAMD, SDS, depression efficiency
Xia WM et al 201955			37/37		T: 47.6±8.2 C: 46.2±7.5		T: 10/27 C: 13/24							Thyroid cancer+ insomnia, anxiety and depression					Routine care	Moxibustion					MT: Baihui (GV 20). 1 time per day for 20 min			7 d			SDS, SAS, PSQI
Zeng XL 201556			29/28		T: 60.28±10.22 C: 61.71±11.58		T: 12/17 C: 13/15							Tumor+depression					Sertraline 50 mg/d	Acupuncture +moxibustion					AT: Neiguan (PC6), Shenmen (TF4), Laogong (PC8), The Stabilizing Four-Shen points and the Four-Shen pointsof Jin's Three-needle Therapy. 1 time per day for 60 min. MT: 1 time per 2 days for 20 min.			4 w			HAMD, SDS, QLQ-C30
Zhang YT 201757			35/34		T: 64.06±9.155 C: 64.46±10.947		T: 18/17 C: 14/20							Tumor+insomnia					Routine care+Estazolam 1 mg/d (if needed)	Thermal moxibustion+C					MT: Shenque (CV8), Xinshu (BL15), Pishu (BL20), Shenmen (TF4) and Sanyinjiao (SP6). 1 time per day (20 min per point)			7 d			PSQI
Zhang CH et al 201758			32/32		T: 56.9±5.7 C: 56.8±6.1		T: 13/19 C: 14/18							Tumor+depression					Amitriptyline 50-150 mg/d	Acupuncture +chinese herb+c					AT: Zusanli (ST36), Sanyinjiao (SP6), Feishu (BL13) and Yinlingquan (SP9). 3 times per week for 30 min			6 w			HAMD, depression efficiency
Zhang YL et al 201759			31/31		18-65 (53.55±11.60)		38/24							Colorectal cancer+anxiety and depression					Loperamide 4 mg/d	Needle warming moxibustion					Guanyun (CV4), Shenque (CV8), Tianshu (ST25) and Zusanli (ST36). 1 time per day (10 min per point)			7 d			SDS, SAS
Zhang B 201760			30/30		T: 59.8±4.3 C: 58.4±5.2		T: 13/17 C: 15/15							Tumor+depression					Chinese herb	Five-element acupuncture+C					The acupoints according to five-element acupuncture therapy. 1 time per week			6 w			HAMD
Zhao YL 201561			93/97		T: 59.36±12.541 C: 60.95±11.112		T: 60/33 C: 66/31							Tumor+insomnia					Estazolam 1 mg/d	Acupuncture+moxibustion					AT: Baihui (GV 20), Shenting (GV24), Yintang (EX-HN3), Shenmen (TF4), Zusanli (ST36) and Sanyinjiao (SP6). 1 time per day for 30 min MT: Shenque (CV8) and Guanyun (CV4). 1 time per day for 30 min			7 d			Sleep efficiency, PSQI
Zhu JJ et al 201762			50/50		T: 63.38±8.34 C: 62.78±7.85		T: 37/13 C: 38/12							Liver cancer+depression					Escitalopram 10 mg/d	Cake-separated moxibustion +Chinese herb					MT: Fengfu (GV16), Geshu (BL17) and Danshu (BL19). 1 time per day for 30 min			6 w			HAMD, depression efficiency

Notes: T: the treatment group; C: the control group; NR: not reported; AT: acupuncture; MT: moxibustion; EA: Electric acupuncture; IN: intradermal needling; SA: scalp-acupuncture; PSQI: the Pittsburgh Sleep Quality Index; HAMD: the Hamilton Depression Scale; SDS: the Self-rating Depression Scale; SAS: the Self-rating Anxiety Scale; QLQ-C30: Quality of Life measurement Scale.

We divided the studies into three subgroups according to their different treatment comparisons. The first subgroup was AMT versus routine care in 8 studies;^{33,40,41,44,50,⇓-52,55} the second subgroup was AMT versus conventional drug (CD) in 11 studies;^{35,42,45,46,48,49,54,56,59,61,62} the third subgroup was AMT plus CD versus CD in 11 studies.^{34,36,⇓,⇓ -39,43,47,53,57,58,60} Total treatment sessions ranged from 7 d to 12 weeks, and intervention time ranged from 30 to 60 min.

3.3. Risk of bias

According to the Cochrane Handbook for Systematic Reviews of Interventions, we assessed the risk of bias of the included literature. The details of the risk of bias (ROB) assessment in 30 studies are provided in Figure 2. 19 RCTs described the appropriate random sequence generation method in detail while the other 11 studies did not. 19 studies using allocation concealment with sealed envelopes, nine studies did not report, while 2 studies broke randomization and thus were assigned with high risk. Because it is impossible to blind participants and personnel from acupuncture and moxibustion to conventional treatment, 13 studies were assessed to be a high ROB. Only 1 study was at low ROB for reporting the blinding of participants and personnel, while the other 16 studies were at unclear ROB. All of the included studies reported the complete outcome data, and we considered them to be at low ROB. 24 studies had reported the details of adverse reactions or published protocols with a low ROB, while the other 6 studies did not report. 20 studies reported the patients’ baseline characteristics, while the other 10 studies were assessed to be at unclear ROB for lack of reporting details. The overall quality of the trials was assessed as having moderate to high risk of bias.

Figure 2. Risk of bias graph.

A: risk of bias item among included studies; B: risk of bias item in each included study.

3.4. Effectiveness assessment

3.4.1 Evaluation of efficacy in insomnia

3.4.1.1. Primary outcome measure: the effective rate of insomnia

Seven studies^{38,49,50,52,56,60,64} showed the equivalent effects of AMT on the cancer related insomnia effective rate, 391 cases in the treatment group and 399 cases in the control group, with high heterogeneity between the studies (P < 0.000 01, I² = 83%). The Meta-analysis was detailed in Figure 3A. The Meta-analysis showed no significant difference between the 2 groups [n = 790, RR = 1.18, 95% CI (0.93, 1.51), P = 0.18].

Figure 3. Results of the Meta-analysis of the efficacy of insomnia.

A: Meta-analysis forest map of the effective rate of insomnia; B: Meta-analysis forest map of PSQI. AMT: acupuncture and moxibustion therapy; CD: conventional drugs; PSQI: Sleep Quality Index.

The subgroup analysis was explained the high heterogeneity by different intervention of treatment. The Meta-analysis showed that 4 studies^38,49,52,64 showed significant difference on the effective rate of insomnia between AMT and conventional drugs [n = 555, RR = 0.89, 95% CI (0.81, 0.98), P = 0.02] with low heterogeneity (P = 0.32, I² = 14%), 3 studies^50,56,60 showed significant difference on the effective rate of insomnia between AMT + CD and conventional drugs [n = 235, RR = 2.18, 95% CI (1.34, 3.56), P = 0.002] with middle heterogeneity (P = 0.09, I² = 59%).

3.4.1.2. Primary outcome measure: the Pittsburgh Sleep Quality Index (PSQI)

Eleven studies^{38,41,43,50-52,54,56,58,60,64}reported changes in PSQI, 490 cases in the treatment group and 493 in the control group, with high heterogeneity between the studies (P < 0.000 01, I² = 95%), so random effects model was used. The Meta-analysis was detailed in Figure 3B. The results of Meta-analysis showed that the difference in the reductions of the PSQI between the treatment group and the control group was statistically significant [n = 984, MD = −1.95, 95% CI (−3.01, −0.90), P = 0.0003].

A subgroup analysis was conducted to explore whether heterogeneity could be partially explained by different intervention of treatment. The Meta-analysis showed that 3 studies^43,54,58 showed significant difference on PSQI between AMT and routine care [n = 223, MD = −3.77, 95% CI (−5.43, −2.12), P < 0.000 01] with high heterogeneity (P = 0.002, I² = 84%), 4 studies^38,51,52,64 showed no significant difference on PSQI between AMT and conventional drugs [n = 445, MD = 0.10, 95%CI (−0.75, 0.94), P = 0.82] with high heterogeneity (P = 0.005, I² = 77%), 4 studies^41,50,56,60 showed significant difference on PSQI between AMT + CD and conventional drugs [n = 315, MD = −2.77, 95% CI (−3.67, −1.87), P < 0.000 01] with high heterogeneity (P = 0.001, I² = 81%).

3.4.2 Evaluation of efficacy in depression

3.4.2.1 Primary outcome measure: the effective rate of depression

Seven studies^{41,44,46,57,59,61,65} showed the equivalent effects of AMT on the cancer related depression effective rate, 244 cases in the treatment group and 243 cases in the control group, with moderate heterogeneity between the studies (P = 0.02, I² = 59%). The Meta-analysis was detailed in Figure 4A. The Meta-analysis showed significant difference between the 2 groups [n = 487, RR = 1.29, 95% CI (1.12, 1.49), P = 0.0004].

Figure 4. Results of the Meta-analysis of the efficacy of depression.

A: Meta-analysis forest map of the effective rate of depression; B: Meta-analysis forest map of HAMD. AMT: acupuncture and moxibustion therapy; CD: conventional drugs; HAMD: Hamilton Depression Scale.

The subgroup Meta-analysis showed that 1 studies⁴⁴showed significant difference on the effective rate of depression between AMT and routine care [n = 60, RR = 3.00, 95% CI (1.71, 5.25), P = 0.0001], 3 studies^57,59,65 showed significant difference on the effective rate of depression between AMT and conventional drugs [n = 203, RR = 1.18, 95% CI (1.05, 1.32), P = 0.005] with low heterogeneity (P = 0.50, I² = 0%), 3 studies^41,46,61 showed significant difference on the effective rate of depression between AMT + CD and conventional drugs [n = 224, RR = 1.28, 95% CI (1.09, 1.50), P = 0.003] with low heterogeneity (P = 0.27, I² = 24%).

3.4.2.2 Primary outcome measure: HAMD

Twelve studies^{37,39-41,44,65,48,57,59,61,63,65} reported changes in HAMD, 383 cases in the treatment group and 387 in the control group, with high heterogeneity between the studies (P < 0.000 01, I² = 94%), so standard random effects model was used. The Meta-analysis was detailed in Figure 4B. The results of Meta-analysis showed that the difference in the reductions of the HAMD between 2 groups was statistically significant [n = 770, SMD = −1.10, 95% CI (−1.77, −0.43), P = 0.001].

The Meta-analysis of the subgroups showed that 1 studies⁴⁴showed significant difference on HAMD between AMT and routine care [n = 60, SMD = −0.54, 95% CI (−1.06, −0.03), P = 0.04], 5 studies^{39,48,57,59,65} showed no significant difference on HAMD between AMT and conventional drugs [n = 330, SMD = −0.62, 95% CI (−1.87, 0.62), P = 0.33] with high heterogeneity (P < 0.000 01, I² = 96%), 6 studies^{37,40,41,46,61,63} showed significant difference on HAMD between AMT+CD and conventional drugs [n = 380, SMD = −1.59, 95% CI (−2.47, −071), P = 0.0004] with high heterogeneity (P < 0.000 01, I² = 93%).

3.4.2.3 Primary outcome measure: SDS

13 studies^{36,41,42,45,47,48,50,53,55,57-69,62} reported changes in SDS, 519 cases in the treatment group and 525 in the control group, with high heterogeneity between the studies (P < 0.000 01, I² = 92%), so random effects model was used. The Meta-analysis was detailed in Table 2 and supplementary Figure 1A. The results of Meta-analysis showed that the difference in the reductions of the SDS between 2 groups was statistically significant [n = 1044, MD = −9.40, 95% CI (−12.21, −6.59), P < 0.000 01].

Table 2.

Summary of the Meta-analysis

Risk factor		Trials (n)	Participants (n)	Mean difference (95% CI)	Heterogeneity	P value
SDS		13	1044	−9.40 (−12.21,−6.59)	<0.000 01	<0.000 01
	AMT and routine care	5	412	−13.28 (−17.92,−8.64)	<0.000 01	<0.000 01
	AMT and conventional drugs	5	312	−6.66 (−12.66,−0.65)	<0.000 01	0.03
	AMT+CD and conventional drugs	3	320	−7.44 (−9.26, −5.63)	0.91	<0.000 01
SAS		10	861	−7.75 (−10.44, −5.05)	<0.000 01	<0.000 01
	AMT and routine care	5	412	−8.61 (−12.38, −4.84)	<0.000 01	<0.000 01
	AMT and conventional drugs	3	209	−7.12 (−14.61, −0.37)	<0.000 01	0.06
	AMT+CD and conventional drugs	2	240	−6.60 (−11.32, −1.88)	0.07	0.006
QOL		5	323	1.11 (0.80, 1.42)	0.15	<0.000 01
	AMT and conventional drugs	2	137	1.36 (0.99, 1.74)	0.31	<0.000 01
	AMT+CD and conventional drugs	3	186	0.93 (0.59, 1.28)	0.28	<0.000 01

Notes: CI: confidence interval; SDS: self-rating depression scale; SAS: self-rating anxiety scale; QOL: quality of life; AMT: acupuncture and moxibustion therapy; CD: conventional drugs.

The Meta-analysis of the subgroups showed that 5 studies^{36,47,53,55,58} showed significant difference on SDS between AMT and routine care [n = 412, MD = −13.28, 95%CI (−17.92, −8.64), P < 0.000 01] with high heterogeneity (P < 0.000 01, I² = 93%), 5 studies^{45,48,57,59,62} showed difference on SDS between AMT and conventional drugs [n = 312, MD = −6.66, 95%CI (−12.66, −0.65), P = 0.03] with high heterogeneity (P < 0.000 01, I² = 95%), 3 studies^41,42,50 showed significant difference on SDS between AMT + CD and conventional drugs [n = 320, MD = −7.44, 95% CI (−9.26, −5.63), P < 0.000 01] with low heterogeneity (P = 0.91, I² = 0%).

3.4.3 Evaluation of efficacy in anxiety

3.4.3.1 Primary outcome measure: SAS

Ten studies^{36,42,45,47,48,50,53,55,58,62} reported changes in SAS, 427 cases in the treatment group and 434 cases in the control group, with high heterogeneity between the studies (P < 0.000 01, I² = 88%), so random effects model was used. The Meta-analysis was detailed in Table 2 and supplementary Figure 1B. The results of Meta-analysis showed that the difference in the reductions of the SAS between 2 groups was statistically significant [n = 861, MD = −7.75, 95% CI (−10.44, −5.05), P < 0.000 01].

The Meta-analysis of the subgroups showed that 5 studies^{36,47,53,55,58} showed significant difference on SAS between AMT and routine care [n = 412, MD = −8.61, 95% CI (−12.38, −4.84), P < 0.000 01] with high heterogeneity (P < 0.000 01, I² = 88%), 3 studies^45,48,62 showed no difference on SAS between AMT and conventional drugs [n = 209, MD = −7.12, 95%CI (−14.61, −0.37), P = 0.06] with high heterogeneity (P < 0.000 01, I² = 95%), 2 studies^42,50 showed significant difference on SAS between AMT+CD and conventional drugs [n = 240, MD = −6.60, 95% CI (−11.32, −1.88), P = 0.006] with high heterogeneity (P = 0.07, I² = 71%).

3.4.4 Quality of life (QOL) and adverse events

3.4.4.1 Secondary outcomes: QOL

Five studies^{36,37,42,53,56} reported changes in quality of life, 162 cases in the treatment group and 161 cases in the control group, with moderate heterogeneity between the studies (P = 0.15, I² = 40%), so random effects model was used. The Meta-analysis was detailed in Table 2 and supplementary Figure 2. The results of Meta-analysis showed that the difference in the growth of the QOL between 2 groups was statistically significant [n = 323, MD = 1.11, 95% CI (0.80, 1.42), P < 0.000 01].

The Meta-analysis of the subgroups showed that 2 studies^42,56 showed significant difference on QOL between AMT and conventional drugs [n = 137, MD = 1.36, 95%CI (0.99, 1.74), P < 0.000 01] with high heterogeneity (P = 0.31, I² = 1%), 3 studies^36,37,53 showed significant difference on QOL between AMT+CD and conventional drugs [n = 186, MD = 0.93, 95% CI (0.59, 1.28), P < 0.000 01] with high heterogeneity (P = 0.28, I² = 22%).

3.4.4.2 Secondary outcomes: adverse reactions

Ten studies^{38,44,46,48,51,53,54,57,59,62} investigated adverse events, whereas other 20 studies did not report. One studie⁴⁶ reported a total of 16 mild adverse events in 80 cases including mild skin bruising (n = 8), dizziness (n = 5), and headache (n = 3) which were improved after symptomatic treatment. One studie⁵¹ investigated that adverse events of acupuncture included vertigo, reduced ability of behave while adverse events of taking sleeping pills “Estazolam” included hangover, nausea or dizziness, and insomnia rebound. Two studies^38,48 reported a total of 5 adverse events including pain (n = 1), hematoma (n = 2), palpitation (n = 1) and dizziness (n = 1) in the treatment group; 16 patients’ adverse reactions including fatigue (n = 1), dizziness (n = 3), nausea (n = 4), sweating (n = 2), and dry mouth (n = 5) in the control group. Two studies^48,51 reported the incidence of adverse reactions in the treatment group was lower than that in the control group with significant difference. Five studies^{44,53,57,59,62} reported that no significant adverse reactions were found in either group.

3.5. Publication bias

Based on the primary outcome measure including PQSI, HAMD, SDS and SAS in RevMan 5.3 software, we used a funnel plot to have a qualitative analysis of publication bias. The distribution of graphical cues was not symmetric, and a few points were distributed beyond the funnel, indicating that there might be publication bias in our study that influenced the results of our analysis, as shown in supplementary Figure 3.

3.6. Sensitivity analyses

We found that the results of heterogeneity comparing the PSQI, the effective rate of insomnia, HAMD, SDS, SAS scale on AMT VS conventional drugs or routine care were not significantly reduced by omitting the study sequentially.

However, the results of heterogeneity comparing the effective rate of depression significantly reduced from [RR = 1.29, 95% CI (1.12, 1.49), P = 0.0004, I² = 59%] to [RR = 1.21, 95% CI (1.11, 1.32), P<0.0001, I² = 0%] after excluding the Liu’s study.⁴¹ Therefore, the Liu’s study⁴¹ was regarded as the source of heterogeneity, it offered inferior evidence for the effect of AMT on cancer-related depression.

4. DISCUSSION

4.1. Acupuncture and moxibustion Intervention Mechanism

As a new and promising field of research, acupuncture and moxibustion therapy is strongly needed by more and more cancer patients and survivors for various cancer-associated symptoms.^{63,⇓- 65} The most common integrative medicine therapies mentioned on the 45 National Cancer Institute (NCI)-designated comprehensive cancer center were exercise (97.8%) and acupuncture and meditation (88.9%).⁶⁶ A number of cancer centers in the U.S., including Dana-Farber Cancer Institute (DFCI) in Boston, Memorial Sloan Kettering Cancer Center in New York, and M.D. Anderson Cancer Center in Houston are integrating acupuncture into cancer care.

The mechanism of AMT on cancer-related psychological problems has been studied as a new and promising field of research. The co-occurrence of multiple psycho-neurological symptoms, including pain, sleep disturbance, fatigue, depression, anxiety, and cognitive disturbance among adult cancer survivors, may be associated with hypothalamic-pituitary-adrenal (HPA), sympathetic nervous system (SNS) dysregulation,⁶⁷ the kynurenine metabolic pathway activation,⁶⁸ the adenosinergic system,⁶⁹ and neuroendocrine-immune mechanisms.^70,71 Laboratory studies have provided evidence for acupuncture’s effect on sleep, depression, anxiety and psychological distress in cancer survivals by stimulating the HPA axis,⁷² SNS system,⁷³ circadian rhythms, cytokines and neurotransmitters(such as IL-1, IL-6, neuropeptide Y, serotonin, melatonin, dopamine, norepinephrine, gamma-aminobutyric acid, β-endorphins),^{74,⇓- 76} which plays a significant role in all important targets in the psychiatric treatment of depression and anxiety.^77,78 At present, more studies are needed on the definite mechanism of acupuncture and moxibustion in patients with CRPS.

4.2. Summary of main results

The aim of this study is to evaluate the effectiveness and safety of AMT in cancer patients with CRPS. After literature screening, we pooled the data from 30 studies involving 2483 patients for further analysis. Our pooled analysis indicates that compared with routine care or combined with conventional drugs, AMT seems to improve cancer-related psychological symptoms to a certain extent, especially in insomnia, depression and anxiety; besides, it is likely to significantly reduce PSQI, HAMD, SDS, SAS and increase insomnia effective rate, depression effective rate and QOL with mild adverse reactions. In addition, there is no statistically difference in the therapeutic effects between the AMT and conventional drug groups. Therefore, compared with conventional drugs, AMT has the same effects as antidepressants and anti-anxiety medication to some extent in improving cancer-related depression and anxiety.

Another systematic review also suggested that acupuncture was slightly more effective than benzodiazepines of anti-insomnia drug, with effective rate for acupuncture and benzodiazepines being 91% and 75% respectively, and equally as effective as zolpidem for improving sleep.^79,80 Adverse reactions of AMT have been observed as being relatively mild, mainly including bruises, soreness, nausea, and dizziness which can be effectively avoided by strengthening the aseptic operation specification, and improving the professional ability of doctors. Therefore, acupuncture and moxibustion treatment for CRPS is considered as being safe and effective, which is worthy of clinical application.

4.3. Quality of evidence

Based on the selected criteria, 30 studies involving a total of 2483 cancer survivors have been included in this study to illustrate the effect of acupuncture and moxibustion on the improvement of CRPS. The quality of evidence differs from low to high based on the Cochrane Bias Evaluation Tools and Jadad scale. Among the 30 studies included, 14 studies are of high quality, 3 studies are of moderate quality, and 13 studies are of low quality. Therefore, the quality of the included studies is generally moderate. The inappropriate random method, allocation concealment, and a lack of blinding in most studies have exaggerated the results of the outcome measures. In this study, 76.67% of the literature and 10% of the literature have correctly reported randomized grouping and allocation of hidden information, respectively. This could lead to overestimate.

4.4. Discussion of heterogeneity

There was heterogeneity in the effectiveness of AMT on the total effective rate of psychological symptoms. To explore its source, we adopted subgroup analysis and found that clinical heterogeneity could not be explained for sorts of western medicine and many factors, its methodological heterogeneity has been mainly affected by the quality of the included literature and the subjective bias of literature quality evaluation. Similarly, by changing the research effect model and adopting the sensitivity analysis method, we’ve found that the heterogeneity is significantly reduced after omitting the study of Liu.⁴¹ To explore its reasons, it has been found that they all have problems such as low quality, the flaw in test design, or small sample sizes. This indicates that the results of this Meta-analysis, to a certain extent, have been affected by the risk of bias.

It seems that AMT is an effective and relatively safe complementary approach for the treatment of cancer-related psychological symptoms. However, substantial heterogeneity was a limitation of the analysis of AMT for cancer-related insomnia, depression and anxiety. The variety of intervention and control groups may have contributed to heterogeneity. But subgroup analysis (according to the type of intervention, type of the acupuncture and moxibustion) could not totally explain the heterogeneity.

4.5. Limitations and advantages

There were several limitations of this systematic review and Meta-analysis:

Firstly, the evaluation criteria are inconsistent for cancer-related insomnia, depression and anxiety. There are 11 studies using PSQI, and 7 studies using the effective rate of insomnia with cancer-related insomnia. There are 12 studies using the HAMD, 13 studies using the SDS, 7 studies using the effective rate of depression with cancer-related depression. Therefore, it might be one of the sources of heterogeneity. However, due to the limitation of the number of included literature and the lack of integrated data, this review doesn’t distinguish these criteria. Therefore, it is suggested that clinical studies should be strengthened in future studies, and clinical effectiveness should be evaluated by the internationally unified effectiveness evaluation standards. Thus, the results would have more authenticity and reliability.

Secondly, the methodological quality of most included studies is relatively low and results in a latent risk of bias, which to some extent, weakens the credibility and reliability of the evidence of AMT for cancer-related psychological symptoms in this systematic review and Meta-analysis. For example, with few control groups being placebo controls or false AMT, it is difficult to eliminate the placebo effect. Although the word “random” appeared in all the included literature, only 19 studies described the random method correctly. In most of the included studies, allocation concealment and blind method are not clear, which may lead to potential implementation bias and selectivity bias. Besides, most of the included RCTs were conducted in China, and only one article was published in English. And the results of the funnel plot indicates potential publication bias.

Compared with published systematic review and Meta-analysis, literature retrieval in this study is more comprehensive with systematic literature retrieval strategy and a larger number of literature is included, ensuring more reliable and scientific evidence in this study. Regardless of the potential bias and limitations of this review, all of the included studies show that AMT intervention seems to have significant effects in improving cancer-related psychological symptoms. As an effective adjunct treatment, AMT intervention has turned to a superior complementary and alternative medicine with a treatment modality of potential efficacy and few adverse effects, reducing the economical burden on health care resources and the community.

In conclusion, in this systematic review and Meta-analysis of RCTs, the efficacy and safety of acupuncture and moxibustion in the treatment of cancer-related psychological symptoms are confirmed. Acupuncture and moxibustion have the potential to become an effective therapy than routine care, and superior to conventional drugs for it was of the same efficacy on cancer-related insomnia, anxiety and depression but caused less adverse reactions.

Nevertheless, limited by the quality of existing published studies, extreme heterogeneity in the analyses remains unexplained with few high-quality studies included. Therefore, further properly designed RCTs with well-designed, rigorous, large sample, and multicenter prospective are needed before the AMT therapy could be widely recommended in the treatment of cancer-related symptoms, so as to reduce the influence of publication bias, improve the credibility of research and guide clinical practice better. The long-term efficacy of AMT intervention on cancer-related psychological symptoms should also be evaluated.