Moxibustion for alleviating side effects of chemotherapy or radiotherapy in people with cancer

Hong Wei Zhang; Zhi Xiu Lin; Fan Cheung; William Chi‐Shing Cho; Jin‐Ling Tang

doi:10.1002/14651858.CD010559.pub2

. 2018 Nov 13;2018(11):CD010559. doi: 10.1002/14651858.CD010559.pub2

Moxibustion for alleviating side effects of chemotherapy or radiotherapy in people with cancer

Hong Wei Zhang ^1,^✉, Zhi Xiu Lin ¹, Fan Cheung ¹, William Chi‐Shing Cho ², Jin‐Ling Tang ³

Editor: Cochrane Gynaecological, Neuro‐oncology and Orphan Cancer Group

PMCID: PMC6517257 PMID: 30480754

Abstract

Background

Moxibustion, a common treatment in traditional Chinese medicine, involves burning herbal preparations containing Artemisia vulgaris on or above the skin at acupuncture points. Its intended effect is to enhance body function, and it could reduce the side effects of chemotherapy or radiotherapy and improve quality of life (QoL) in people with cancer.

Objectives

To assess the effects of moxibustion for alleviating side effects associated with chemotherapy, radiotherapy or both in people with cancer.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library, MEDLINE via Ovid, Embase via Ovid and AMED (Allied and Complementary Medicine Database) from their inception to February 2018. We also searched databases in China including the Chinese BioMedical Literature Database (CBM), Chinese Medical Current Contents (CMCC), TCMonline, Chinese Dissertation Database (CDDB), China Medical Academic Conference (CMAC) and Index to Chinese Periodical Literature from inception to August 2017. Registries for clinical trials and other resources were also searched.

Selection criteria

We included randomised controlled trials (RCTs) comparing moxibustion treatment, including moxa cone and moxa stick, versus sham, no treatment or conventional treatment.

Data collection and analysis

Two review authors (HWZ and FC) independently extracted data on study design, participants, treatment and control intervention, and outcome measures, and they also assessed risk of bias in the included studies. We performed meta‐analyses, expressing dichotomous outcomes as risk ratios (RR) and continuous outcomes as mean differences (MD), with 95% confidence intervals (CI).

Main results

We included 29 RCTs involving 2569 participants. Five RCTs compared moxibustion versus no treatment, 15 compared moxibustion plus conventional treatment versus conventional treatment, one compared moxibustion versus sham moxibustion, and eight compared moxibustion versus conventional medicine. The overall risk of bias was high in 18 studies and unclear in 11 studies. Studies measured outcomes in various ways, and we could rarely pool data.

Moxibustion versus no treatment: low‐certainty evidence from single small studies suggested that moxibustion was associated with higher white blood cell counts (MD 1.77 × 10⁹/L; 95% CI 0.76 to 2.78; 80 participants, low‐certainty evidence) and higher serum haemoglobin concentrations (MD 1.33 g/L; 95% CI 0.59 to 2.07; 66 participants, low‐certainty evidence) in people with cancer, during or after chemotherapy/radiotherapy, compared with no treatment. There was no evidence of an effect on leukopenia (RR 0.50, 95% CI 0.10 to 2.56; 72 participants, low‐certainty evidence) between study groups. The effects on immune function (CD3, CD4, and CD8 counts) were inconsistent.

Moxibustion versus sham moxibustion: low‐certainty evidence from one study (50 participants) suggested that moxibustion improved QoL (measured as the score on the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire‐C30 (EORTC QLQ‐C30)) compared with sham treatment (MD 14.88 points; 95% CI 4.83 to 24.93). Low‐certainty evidence from this study also showed reductions in symptom scores for nausea and vomiting (MD −38.57 points, 95% CI −48.67 to −28.47) and diarrhoea (MD −13.81, 95% CI −27.52 to −0.10), and higher mean white blood cell count (MD 1.72 × 10⁹/L, 95% CI 0.97 to 2.47), serum haemoglobin (MD 2.06 g/L, 95% CI 1.26 to 2.86) and platelets (MD 210.79 × 10⁹/L, 95% CI 167.02 to 254.56) when compared with sham moxibustion.

Moxibustion versus conventional medicines: low‐certainty evidence from one study (90 participants) suggested that moxibustion improved WBC count eight days after treatment ended compared with conventional medicines (MD 0.40 × 10⁹/L; 95% CI 0.15 to 0.65). Low‐certainty evidence from two studies (235 participants) suggested moxibustion improved serum haemoglobin concentrations compared with conventional medicines (MD 10.28 g/L; 95% CI 4.51 to 16.05).

Moxibustion plus conventional treatment versus conventional treatment alone: low‐certainty evidence showed that moxibustion plus conventional treatment was associated with lower incidence and severity of leukopenia (WHO grade 3 to 4) (RR 0.14, 95% CI 0.01 to 2.64; 1 study, 56 participants), higher QoL scores on the EORTC QLQ‐C30 (MD 8.85 points, 95% CI 4.25 to 13.46; 3 studies, 134 participants, I² = 26%), lower symptom scores for nausea and vomiting (RR 0.43, 95% CI 0.25 to 0.74; 7 studies, 801participants; I² = 19%), higher white blood cell counts (data not pooled due to heterogeneity), higher serum haemoglobin (MD 3.97 g/L, 95% CI 1.40 to 6.53; 2 studies, 142 participants, I² = 0%). There was no difference in platelet counts between the two groups (MD 13.48 × 10⁹/L; 95% CI −16.00 to 42.95; 2 studies, 142 participants; I² = 34%).

Most included studies did not report related adverse events, such as burning or allergic reactions.

Authors' conclusions

Limited, low‐certainty evidence suggests that moxibustion treatment may help to reduce the haematological and gastrointestinal toxicities of chemotherapy or radiotherapy, improving QoL in people with cancer; however, the evidence is not conclusive, and we cannot rule out benefits or risks with this treatment. High‐quality studies that report adverse effects are needed.

Plain language summary

Moxibustion for alleviating side effects of chemotherapy or radiotherapy in people with cancer

The issue  Moxibustion is used in traditional Chinese medicine to enhance quality of life and relieve the side effects of conventional treatments for a variety of diseases. As its application involves the burning of a herbal preparation, it can also cause some undesirable side effects itself, such as allergic reactions, burns and infection.

The aim of the review  We conducted this systematic review to understand whether moxibustion can reduce common side effects of chemotherapy and radiotherapy and improve well‐being in people with cancer.

Selection criteria  We reviewed 29 studies involving 2569 people with different types of cancer, receiving chemotherapy, radiotherapy or both.

What are the main findings?  We found some small single studies showing various beneficial effects of moxibustion on increasing blood cells and promoting immunological function, decreasing gastrointestinal symptoms caused by toxicity of chemotherapy or radiotherapy (such as nausea and vomiting), and improving quality of life. However, the poor reporting and high risk of bias in study methods reduced the certainty of the evidence.

What is the certainty of the evidence?  The evidence was of low or very low‐certainty.

What are the conclusions?  There is presently no good evidence to support or oppose the use of moxibustion in people receiving treatment for cancer. High‐quality studies are needed, which should include reporting of adverse effects.

Summary of findings

Background

Description of the condition

Cancer rates have steadily increased over the past few decades, placing a huge burden on health systems worldwide. The global economic burden of cancer has more than doubled over the past 30 years. Cancer is the leading causes of morbidity and mortality worldwide, with approximately 14 million new cases reported in 2012, and it is expected that annual cancer cases will rise to 22 million within the next two decades. The five most common sites of cancer diagnosed in men in 2012 were the lung, prostate, colon/rectum, stomach and liver, and in women they were the breast, colon/rectum, lung, cervix and stomach (WHO 2014).

Conventional cancer treatments include surgery, radiotherapy, chemotherapy and psychosocial support (WHO 2002). Advances in chemotherapy and radiotherapy in recent years have greatly improved treatment results (WHO 2014). However, cytotoxic drugs and ionising radiation also cause many distressing side effects. Some of these are serious enough to prompt discontinuation of treatment (Redmond 1996; Robbins 2002; WHO 2014). The side effects most commonly associated with chemotherapy and radiotherapy include fatigue, pain, and nausea and vomiting (Henry 2008; Stasi 2003). Other side effects include bone marrow suppression leading to anaemia; hair follicle cell damage leading to alopecia; gastrointestinal damage leading to diarrhoea and oral ulceration and skin reactions to radiation (Robbins 2002; WHO 2014). Although new drug development programmes have been undertaken to reduce the side effects of cancer therapy, satisfactory treatment still is not readily available to a large proportion of patients receiving chemotherapy and/or radiotherapy. New, effective treatments that can reduce chemotherapy‐ and radiotherapy‐associated adverse effects are needed, especially non‐pharmacological strategies with minimum harm (Cho 2010; Ellebaek 2008; Herrstedt 2007; Jordan 2007; Lotfi‐Jam 2008; Redmond 1996).

Description of the intervention

Moxibustion is a common treatment in traditional Chinese medicine and has been used in China and other Asian countries for millenia (Cho 2009). Moxibustion involves burning herbal preparations containing Artemisia vulgaris (mugwort) on or above the skin at acupuncture points. Moxibustion techniques commonly used in clinical practice to treat side effects of conventional cancer treatment involve either direct moxibustion with a traditional moxa stick (stick‐on moxa) (Yu 2003), or indirect moxibustion, achieved by placing insulating materials such as salt, monkshood cake, sliced ginger or garlic between the skin and a burning moxa cone (Chen 2000; Zhao 2007). The leaves of A vulgaris or mugwort, in Chinese called ai ye, are the main material used for moxibustion. Other Chinese herbs may be sometimes used in combination with mugwort. Mugwort is considered to be warm, acidic and bitter. It has the ability to warm the body's meridians, thereby promoting better circulation. According to Chinese medicine theory, the meridians are the channels inside the human body that circulate vital energy (in Chinese called qi and blood). Besides promoting the flow of vital energy through meridians, moxibustion, which stimulates some specific acupoints located along the meridians upon burning, is considered to have some specific treatment effects, such as strengthening the body's vital energy or facilitating digestion. Although practiced widely in East Asia, it is also associated with some adverse effects, such as allergic reactions, burns and infections (Chan 2014).

How the intervention might work

A systematic review demonstrated that acupuncture point stimulation, as performed through electro acupuncture and acupressure, may reduce chemotherapy‐induced nausea or vomiting (Ezzo 2006). Moxibustion is widely used in China and in other East Asian countries to reduce cancer pain and fever in people with cancer and to lessen the adverse effects of radiotherapy and chemotherapy (Lee 2010; Zhang 2008). Many clinical studies of moxibustion for people with cancer receiving chemotherapy or radiotherapy have indicated that it could alleviate some of the adverse effects of treatment, such as fatigue, nausea and vomiting, diarrhoea, alopecia and pain, as well as improving quality of life (Chen 2000; Chen 2008; Gao 2010; Jiang 2002; Kim 2010; Kuai 2008; Qiu 2008; Song 2003; Shen 2008; Zhang 2008; Zhao 2007).

The clinical effects of moxibustion may be attributable to the actions of enhancing immunity, relieving bone marrow suppression and producing an anti‐oxidative effect (Chen 2000; Cui 2007; Huang 1999; Jiang 2002; Pei 2007; Xu 2003a; Yu 2002a; Yu 2003; Zhao 2007). The infrared radiation peak (around 7.5 μm) of traditional indirect moxibustion with monkshood cake, ginger slices and garlic slices as the medium matches that of infrared radiation on human skin at some acupoints such as LI 4 (hegu), indicating involvement of a sympathetic vibration of infrared radiation from indirect moxibustion and the acupoints. These mechanisms of action (including thermal action, infrared radiation and sympathetic vibration) and their pharmacological effects may contribute to the therapeutic efficacy of moxibustion (Shen 2006). In addition, actions exerted on the acupoints by moxibustion may elicit systemic effects through transmission along meridians.

Why it is important to do this review

Given its potential effect, low cost and simplicity of application, moxibustion may be a valuable adjuvant treatment option for many people with cancer. However, practitioners should also consider the possible side effects related to moxibustion. A recently published systematic review on moxibustion for cancer care found limited evidence supporting the effectiveness of moxibustion for reducing cancer‐related nausea and vomiting (Lee 2010). Review authors evaluated moxibustion as the sole treatment for cancer, or as an adjunct to chemotherapy or radiotherapy; however, they did not clearly specify outcome measurements. Our systematic review focuses primarily on the effects of moxibustion for alleviating the side effects of chemotherapy, radiotherapy or both in people with cancer. We used a transparent and clearly defined systematic method to comprehensively evaluate the evidence. Findings from this systematic review should help to inform medical practitioners, patients and researchers about the effectiveness and safety of moxibustion for people with cancer receiving radiotherapy, chemotherapy or both.

Objectives

To assess the effects of moxibustion for alleviating side effects associated with chemotherapy, radiotherapy or both in people with cancer.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs). For randomised cross‐over trials, we included only phase 1 data because treatment carryover effects were likely.

Types of participants

We included participants of any age with any kind of malignant disease receiving chemotherapy, radiotherapy or both.

Types of interventions

The intervention was any type of moxibustion treatment, defined as burning moxa on or above any acupoint or at some specified region of the body. Commonly used techniques include direct and indirect moxibustion with a moxa cone or moxa stick. Direct moxibustion with a moxa cone consists of placing a small cone‐shaped moxa directly on the skin and burning it; in indirect moxibustion, a medium (salt, garlic, ginger, monkshood cake or any other herbs) separates the skin and the burning cone. In moxibustion with moxa stick, a practitioner lights one end of the moxa stick, which is roughly similar to a cigar in shape and size, and holds it for several minutes or even one hour close to the area being treated until the area turns red. We also included moxibustion treatments that involve burning materials made of moxa and/or other medicinal herbs, with or without the aid of an instrument, because these approaches are considered traditional moxibustion treatments. We excluded moxa needle therapy, which consists of inserting a needle into an acupoint and wrapping the end of the needle in an ignited moxa, because this treatment method also involves acupuncture. The acupuncture treatment combined with moxibustion makes it impossible to evaluate whether the treatment effect is due only to moxibustion.

The intervention in the control group may include a sham, no treatment or other conventional treatments that are currently accepted and widely used for patients receiving chemotherapy, radiotherapy or both, and may include treatments for raising white or red blood cell counts and haemoglobin levels, or for enhancing immunity. We did not accept other herbal or complementary medicines as a control intervention when there was no validated evidence about their effectiveness.

Basic oncological treatment (chemotherapy, radiotherapy) or supportive care should be identical in the intervention and control groups. We excluded the studies with Chinese medicines as the co‐administered treatment between groups because they may vary individually.

Types of outcome measures

Primary outcomes

Incidence and severity of chemotherapy‐ or radiotherapy‐related toxicities, as reported according to internationally accepted criteria for common toxicities (e.g. World Health Organization (WHO) (Miller 1981), Eastern Cooperative Oncology Group (ECOG), or National Institutes of Health (NIH) criteria for adverse effects)

Secondary outcomes

Quality of life (QoL) as measured by a validated instrument (e.g. the 36‐Item Short Form Health Survey (SF‐36), the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire‐C30 (EORTC QLQ‐C30), or the World Health Organization QoL (WHOQOL)
Patient‐reported physical and psychological indices of symptom distress using a validated scale (e.g. visual analogues scale (VAS))
Other objective outcome measures aimed at assessing side effects of chemotherapy or radiotherapy (e.g. blood cell counts, measures of immunological function)
Modification or cessation of cancer treatments as the result of side effects or adverse effects, which may be measured as continuous or dichotomous data
Adverse events in the treatment and control groups (including serious and moderate ones), which may or may not be related to moxibustion treatment. We compared the possible occurrence of adverse events between the moxibustion group and the control group

The above outcome measurements were collected immediately after treatment and at the end of follow‐up.

Search methods for identification of studies

We searched for articles in all languages, applying no date restrictions.

Electronic searches

We searched the following databases:

the Cochrane Central Register of Controlled Trials (CENTRAL; 2018, Issue 2), in the Cochrane Library;
MEDLINE via Ovid (1946 to January week 4 2018);
Embase via Ovid (1980 to 2018 week 6);
AMED (Allied and Complementary Medicine Database) (1985 to January 2018).

We also searched trials registries and Chinese databases, including Chinese BioMedical Literature Database (CBM), Chinese Medical Current Contents (CMCC), TCMonline, Chinese Dissertation Database (CDDB), China Medical Academic Conference (CMAC) and Index to Chinese Periodical Literature from their inception time to August 2017.

The search strategies in CENTRAL, MEDLINE and Embase are in Appendix 1,Appendix 2 and Appendix 3.

We identified all relevant articles on PubMed and used the 'related articles' feature to carry out further searches for newly published articles.

Searching other resources

We searched the following registries for ongoing trials: metaregister (www.controlled‐trials.com/mrct), Physicians Data Query (/www.ncbi.nlm.nih.gov), www.clinicaltrials.gov and www.cancer.gov/clinicaltrials.

To identify ongoing studies and grey literature, we searched USA CenterWatch Clinical Trials Listing Service (www.CenterWatch.com) and OpenSIGLE (System for Information on Grey Literature in Europe).

We checked the references of all included studies and relevant reviews to find further relevant articles.

Data collection and analysis

Selection of studies

We used the search strategy described above to obtain titles and abstracts of studies that may be relevant to the review. We entered all references from electronic databases into NoteExpress and removed duplicates. Two review authors (HWZ and FC) independently reviewed these titles and abstracts, discarding studies that were not eligible for the review and retaining those with potentially relevant data or information. We retrieved full texts of potentially eligible articles for further assessment, labelling each as 'include', 'exclude' or 'unclear' on full‐text review. We resolved disagreements by discussion and consensus. When the article fell into the unclear category due to unclear information or missing data, we contacted the trial authors for clarification, recording all communications.

Data extraction and management

Two review authors (HWZ and FC) independently carried out data extraction, using a pre‐tested data extraction form. When we found more than one publication of a study, we grouped reports together, using the publication with the most recent and complete data to extract outcomes. When earlier reports were the only ones to publish relevant outcomes, we used these data, noting any discrepancies between published versions. A third review author (ZXL) resolved disagreements between the two review authors in consultation with them.

For included trials, HWZ abstracted the following data as recommended in Chapter 7 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

General information: published or unpublished, author, country, publication language, publication year, journal citation.
Trial design.
Participants: inclusion and exclusion criteria, total number enrolled and number in each comparison group, baseline characteristics, setting.
Interventions: administration route, timing of intervention, comparison intervention and any co‐intervention, expertise of practitioner.
Risk of bias in trials (see Assessment of risk of bias in included studies).
Follow‐up: length of follow‐up, reason for and number of dropouts and withdrawals, method of analysis.
Outcomes reported: the incidence and severity of chemotherapy‐ or radiotherapy‐related toxicities, QoL, patient‐reported physical and psychological indices of symptom distress based on a validated scale, any other objective outcome measures aimed at assessing side effects of chemotherapy or radiotherapy, modification or cessation of cancer treatments as the result of side effects or adverse effects, and incidence and types of adverse events resulting from moxibustion.
For each outcome: outcome definition (with diagnostic criteria if relevant).
Unit of measurement (if relevant).
For scales: upper and lower limits, and whether high or low score is good.
Results: number of participants allocated to each intervention group.
For each outcome of interest: sample size and missing participants.

Data on outcomes were extracted as follows:

For dichotomous outcomes (e.g. adverse events), we extracted the number of participants in each treatment arm who experience the outcome of interest and the number of participants assessed at endpoint to estimate a risk ratio (RR).
For continuous outcomes (e.g. QoL), we extracted the final value and the standard deviation of the outcome of interest and the number of participants assessed at endpoint in each treatment arm at the end of follow‐up to estimate the mean difference (MD) (if trials measured outcomes on the same scale) or standardised mean differences (SMD) (if trials measured outcomes on different scales) between treatment arms and standard error.

Assessment of risk of bias in included studies

Two review authors (HWZ and FC) independently assessed the risk of bias in the included studies, resolving any discrepancies by discussion and reaching conclusions by consensus. If disagreements persisted, a third review author (ZXL) helped to make the final decision.

To detect potential selection bias, performance bias, detection bias, attrition bias and reporting bias, we addressed the following six domains in the assessment of risk of bias.

Selection bias.
- Random sequence generation.
- Allocation concealment.
Performance bias.
- Blinding of participants and personnel (participants and treatment providers) on subjective and objective outcomes.
Detection bias.
- Blinding of outcome assessment.
Attrition bias.
- Incomplete outcome data: we recorded the proportion of participants whose outcomes were not reported at the end of the study, coding a satisfactory level of loss to follow‐up for each outcome, such as:
  - Low risk of bias, if fewer than 20% of participants were lost to follow‐up, and reasons were similar in both treatment arms.
  - High risk of bias, if more than 20% of participants were lost to follow‐up, or reasons for loss to follow‐up differed between treatment arms.
  - Unclear risk of bias, if authors did not report loss to follow‐up.
Reporting bias.
- Selective reporting of outcomes.
Other possible sources of bias.
- Baseline characteristics.

We categorised the risk of bias for each outcome, within and across included studies, into three levels: low, unclear and high risk of bias. On the basis of this assessment, we used the GRADE system to further evaluate the certainty of evidence for each individual outcome (Higgins 2009). This involved consideration not only of risk of bias (methodological quality) but also of directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias. The empirical evidence for each individual outcome was graded into four levels: high, moderate, low or very low certainty in accordance with the GRADE approach.

Measures of treatment effect

We used the following measures of the effect of treatment.

For dichotomous outcomes, we used the RR with 95% confidence interval (CI). To help determine the applicability of the results to individual participants, we planned to calculate the number needed to treat for a beneficial outcome (NNTB) across a range of assumed control risks if needed.
For continuous outcomes, we used the mean difference between treatment arms (with its 95% CI).

Unit of analysis issues

We analysed outcomes based on randomised participants. In the case of multiple intervention groups within a study, we performed pair‐wise comparisons relevant to the study objective. If necessary, we combined relevant groups to make a single comparison or split them to make multiple comparisons.

Dealing with missing data

Conducting available case analysis, we considered the potential impact of missing data in the 'Risk of bias' table and in interpretation of the results. We did not impute missing outcome data for any of the outcomes.

Assessment of heterogeneity

We assessed heterogeneity between studies through visual inspection of forest plots, by estimation of the percentage of heterogeneity between trials that cannot be ascribed to sampling variation (Higgins 2003), by a formal statistical test of the significance of the heterogeneity (Deeks 2001), and, if possible, by subgroup analyses. When heterogeneity was present, we first reviewed study components such as participants, interventions and outcomes to decide whether the heterogeneity was substantial. If that were the case, we investigated and reported on possible reasons.

Assessment of reporting biases

Due to the widespread comparisons in the included studies, we did not undertake funnel plot analysis as planned.

Data synthesis

When clinically similar studies were available, we pooled their results in meta‐analyses.

For any dichotomous outcomes, we calculated the RR for each trial and then as a pooled effect estimate.
For continuous outcomes, we pooled the MDs between treatment arms at the end of follow‐up if all trials measured the outcome on the same scale.

We used random‐effects models with inverse variance weighting for all meta‐analyses (DerSimonian 1986).

Subgroup analysis and investigation of heterogeneity

We undertook post hoc subgroup analysis based on the different conventional medicines in the control group. We did not conduct the planned subgroup analyses based on type of cancer, indirect or direct moxibustion, age of participants and duration of moxibustion treatment due to the widespread comparisons and limited number of included studies.

Sensitivity analysis

We planned a sensitivity analysis to explore the influence of adequate sequence generation and blinding as well as the possible influence of including data from the first period in a cross‐over study, but we were not able to carry this out due to the paucity of relevant included studies.

Summary of findings table

Based on the methods described in Chapter 11 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), we prepared a 'Summary of findings' table to present the review findings. We presented results for the following outcomes.

Incidence and severity of chemotherapy‐ or radiotherapy‐related toxicities.
QoL (EORTC QLQ‐C30).
Patient‐reported physical and psychological indices of symptom distress.
- Nausea/vomiting (EORTC QLQ‐C30, WHO grade 3 to 4).
- Diarrhoea (EORTC QLQ‐C30).
Objective outcome measures aimed at assessing side effects of chemotherapy or radiotherapy.
- Leukopenia (WHO grade 3 to 4).
- WBC count (× 10⁹/L).
- Haemoglobin (g/L).
- Platelets (× 10⁹/L).

We used the GRADE system to rate the certainty of the evidence (Schünemann 2011), downgrading for inconsistency, design limitations (risk of bias), imprecision, indirectness and other factors, such as publication bias, where appropriate. Where the evidence was based on single studies, or where there was no evidence on a specific outcome, we included the pre‐specified outcome in the 'Summary of findings' tables and graded or explained accordingly. Two review authors (HWZ and FC) performed the grading, resolving differences by discussion and, if necessary, by involving a third review author (ZXL).

Results

Description of studies

See Characteristics of included studies; Characteristics of excluded studies; and Characteristics of studies awaiting classification.

Results of the search

Our searches yielded 1224 records. After removing duplicates, we screened the titles or abstracts of 823 records. We read the full texts of 202 records and finally included 29 RCTs in the review (Chen 2000; Chen 2015; Cheng 2005; Cheng 2016; Enkhtuya 2010; Fan 2001; Gao 2013; Hao 2014; Li 2011; Li 2012; Li 2014a; Li 2015; Li 2016; Liang 2002; Mo 2016; Ruan 2014; Tian 2015; Wang 2014; Wu 2013; Xu 2014a; Xu 2014b; Yang 2014; Yin 2013; Yu 2004; Yuan 2014; Zhang 2013 Zhang 2016a; Zhang 2016b; Zhu 2017). Eight of the full texts contained insufficient or ambiguous information (Cui 2010; Lan 2013; Li 2014b; Liang 2012; Qiu 2015; Zhang 2014b; Zhang 2014c; Zhang 2014d), justifying their inclusion in the Characteristics of studies awaiting classification section, pending responses from the investigators (Figure 1). After careful comparison, we considered two records to pertain to the same study as Zhang 2013, while eight pertained to Yu 2004.

Included studies

We included 29 studies involving 2569 participants. The sample sizes ranged from 24 to 332 participants. All studies were reported in Chinese; 28 were conducted in Chinese hospitals, whereas Enkhtuya 2010 took place in a Mongolian hospital. The participants had a variety of cancers, including nasopharyngeal carcinoma, gastric cancer, respiratory system cancer, primary non‐small cell lung cancer, breast cancer and cervical cancer. All but two studies, Fan 2001 and Liang 2002, reported participant gender, and overall, 1279 (49.8%) of participants in these studies were men.

In 2 studies, participants received simultaneous chemotherapy and radiotherapy plus moxibustion (Chen 2000; Cheng 2005); in 19 studies, participants received simultaneous moxibustion plus chemotherapy (Chen 2015; Cheng 2016; Enkhtuya 2010; Hao 2014; Li 2011; Li 2014a; Li 2015; Liang 2002; Ruan 2014; Wang 2014; Wu 2013; Xu 2014a; Xu 2014b; Yang 2014; Yin 2013; Yuan 2014; Zhang 2013; Zhang 2016b; Zhu 2017); in 1 study, participants received moxibustion before and after chemotherapy (Fan 2001); in 6 studies, participants received moxibustion after chemotherapy and/or radiotherapy (Gao 2013; Li 2012; Li 2016; Mo 2016; Tian 2015; Zhang 2016a); and in 1 study, participants received simultaneous moxibustion with radiotherapy (Yu 2004).

Fifteen studies compared moxibustion plus conventional treatment versus conventional treatment alone (Chen 2000; Chen 2015; Enkhtuya 2010; Gao 2013; Hao 2014; Li 2015; Li 2016; Ruan 2014; Xu 2014a; Yang 2014; Yin 2013; Yuan 2014; Zhang 2013; Zhang 2016b; Zhu 2017); eight compared moxibustion versus conventional medicines (Cheng 2005; Cheng 2016; Fan 2001; Li 2012; Li 2014a; Mo 2016; Tian 2015; Wang 2014); five compared moxibustion versus no treatment (Li 2011; Liang 2002; Wu 2013; Yu 2004; Zhang 2016a); and one compared moxibustion versus sham moxibustion (Xu 2014b).

In four studies (Enkhtuya 2010; Li 2011; Zhang 2013; Zhang 2016b), practitioners placed at least three continuous moxa cones directly on acupoints RN4 (guanyuan) or bilateral BL17 (geshu) and BL19 (danshu), and the treatment duration ranged from 5 to 10 days. Three studies used a direct grain‐sized moxa cone placed on the acupoints ST36 (zusanli), DU14 (dazhui), BL13 (feishu) or RN4 (guanyuan) with continuous 5, 9 or 18 cones, and the treatment duration ranged from 12 to 42 days (Gao 2013; Xu 2014a; Zhang 2016a). In two studies, a specially made direct moxa box on was used on acupoints RN13 (shangwan), RN12 (zhongwan), RN10 (xiawan), ST25 (tianshu), PC6 (neiguan), and ST36 (zusanli) (Cheng 2016; Li 2015). In seven studies, an indirect moxa cone was placed on salt on acupoint RN8 (shenque), or on ginger placed on the acupoints DU14 (dazhui), BL17 (geshu), BL20 (pishu), BL21 (weishu), ST36 (zusanli) or RN12 (zhongwan), with treatment duration ranging from 3 to 65 days (Chen 2000; Chen 2015; Cheng 2005; Li 2011; Li 2014a; Xu 2014b; Yuan 2014). Nine studies used a moxa stick (Fan 2001; Hao 2014; Liang 2002; Li 2016; Mo 2016; Tian 2015; Yang 2014; Yin 2013; Yu 2004), generally for about 10 to 30 minutes on acupoints ST36 (zusanli), SP6 (sanyinjiao) and RN8 (shenque) for 5 to 50 days. One study used moxa stick on acupoints RN8 (shenque), on a paste of grounded herbs (chaihu (Bupleuri Radix), chuanxiong (Chuanxiong Rhizoma), dangshen (Codonopsis Radix), maidong (Ophiopogonis Radix), wuweizi (Schisandrae Chinensis Fructus), danggui (Angelicae Sinensis Radix), huangqi (Astragali Radix) and shexiang (Moschus) for about 2 hours per treatment and 3 times per week, with duration of 126 days (Wu 2013). One study used indirect moxa box on ginger, which was placed on the bilateral acupoints ST36 (zusanli) and KI 1 (yongquan) (Zhu 2017). Only three studies used complementary acupoints based on syndrome differentiation, according to Chinese medicine theory (Hao 2014; Liang 2002; Yang 2014).

The conventional medicines used in the control group included leucogen, batilol, berbamine or recombinant human granulocyte colony‐stimulating factor (G‐CSF) injection. The conventional treatments administered in both groups were generally supportive and symptomatic ones, including granisetron for the prevention of vomiting.

In Xu 2014b, a moxa cone was placed on a slice of ginger in the treatment arm, and for the sham moxibustion, a piece of board in thickness of 0.6 mm was placed on a slice of ginger between the skin and moxa cone to insulate the heat of burned moxa cone.

Excluded studies

The main reasons for excluding studies were lack of random allocation (Chen 2010), moxibustion combined with other therapy (Chen 2006), Chinese medicine as the control intervention (Liu 2002), no targeted outcome (Chen 1991), varied conventional treatment depending on the symptoms in the control group or both groups (Liang 2014; Zhang 2014a; Zhong 2014), no chemotherapy or radiotherapy (Wang 2016), and duplicate or fake reports (Xu 2002a).

Most relevant studies took place in China, and many authors used self‐developed scales or national criteria to assess the treatment outcome (Zheng 2002).

Risk of bias in included studies

See Figure 2; Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Only nine studies reported using a random number table or computer programme (Hao 2014; Li 2012; Li 2016; Mo 2016; Ruan 2014; Wang 2014; Xu 2014a; Xu 2014b; Zhang 2016a). Other studies mentioned only the random allocation without any further information. In three studies, there was a high imbalance in the number of cases between groups (Fan 2001; Li 2016; Yu 2004), which we considered conferred high risk of bias. One study reported using an envelop during random allocation. Only Chen 2000 reported adequate procedures for allocation concealment.

Blinding

No included study reported any procedure for undertaking blinding of participants or doctors, even the study using sham moxibustion. We considered all of the studies except Xu 2014b to be at high risk of performance bias. Xu 2014b used sham moxibustion, but there was no description of blinding measures. The risk of performance bias was unclear in this study. For detection bias, no study reported information on blinding. Because we thought the lack of blinding had less influence on the objective compared to subjective outcomes, we considered objective outcomes to be at unclear risk of bias and subjective ones to be at high risk.

Incomplete outcome data

All but three of the included studies reported complete data (Xu 2014b; Yu 2004; Zhang 2016a). In Xu 2014b, data were missing for 2/27 participants in both groups. In Yu 2004, data were missing for 2/38 and 9/30 participants, with no explanation for the reason. In Zhang 2016a, there were no data for 2/35 participants in the treatment group and 3/35 in the control group.

Selective reporting

There were no protocols available for included studies, but the review outcomes described in the Methods were generally reported, so we considered the risk of reporting bias to be unclear.

Other potential sources of bias

Eight studies presented baseline data, which were comparable between groups (Enkhtuya 2010; Hao 2014; Ruan 2014; Xu 2014a; Xu 2014b; Yang 2014; Zhang 2016a; Zhu 2017). The other studies reported only that some baseline data between groups were comparable without any detailed information, or they provided no information about comparability.

We assessed the overall risk of bias as unclear in 11 studies (Cheng 2005; Cheng 2016; Gao 2013; Li 2011; Li 2012; Li 2014a; Liang 2002; Tian 2015; Wang 2014; Wu 2013; Zhang 2016a) and high in 18 others (Chen 2000; Chen 2015; Enkhtuya 2010; Fan 2001; Hao 2014; Li 2015; Li 2016; Mo 2016; Ruan 2014; Xu 2014a; Xu 2014b; Yang 2014; Yin 2013;Yu 2004; Yuan 2014; Zhang 2013; Zhang 2016b; Zhu 2017).

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4

Summary of findings for the main comparison. Moxibustion versus no treatment for side effects of chemotherapy or radiotherapy in cancer patients.

Moxibustion versus no treatment for side effects of chemotherapy or radiotherapy in cancer patients
Patient or population: patients receiving chemotherapy or radiotherapy for cancer treatment Settings: hospital Intervention: moxibustion Comparison: no treatment
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No of participants  (studies)	Certainty of the evidence  (GRADE)
	Assumed risk	Corresponding risk
	No treatment	Moxibustion treatment
The incidence and severity of toxicities: leukopenia (WHO grade 3 to 4)	111 per 1000	56 per 1000  (11 to 284)	RR 0.50   (0.10 to 2.56)	72  (1 study)	⊕⊕⊝⊝  Low^a
QoL	No evidence
Patient‐reported symptom: nausea/vomiting	No evidence
Patient‐reported symptom: diarrhoea	No evidence
Objective outcome measure: WBC count (× 10‎⁹/L)	Mean WBC counts (× 10‎⁹/L) in the control group was 3.60	Mean WBC counts (×10‎⁹/L) in the intervention group was 5.37 (4.36 to 6.38)	MD 1.77  (0.76 to 2.78)	80  (1 study)	⊕⊕⊝⊝  Low^a
Objective outcome measure: haemoglobin (g/L)	Mean haemoglobin (g/L) in the control group was 10.24	Mean haemoglobin (g/L) in the intervention groups was 11.57 (11.44 to 11.7)	MD 1.33  (1.20 to 1.46)	66  (1 study)	⊕⊕⊝⊝  Low^a
Objective outcome measure: platelets (× 10⁹/L)	No evidence
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; RR: risk ratio; WBC: white blood cells; WHO: World Health Organization.
GRADE Working Group grades of evidence  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Moxibustion versus sham treatment for side effects of chemotherapy or radiotherapy in cancer patients
Patient or population: patients receiving chemotherapy or radiotherapy for cancer treatment Settings: hospital Intervention: moxibustion treatment Comparison: sham
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No of participants  (studies)	Certainty of the evidence  (GRADE)
	Assumed risk	Corresponding risk
	Sham	Moxibustion treatment
The incidence and severity of chemotherapy‐ or radiotherapy‐related toxicities	No evidence
QoL (EORTC QLQ‐C30)	Mean QoL (EORTC QLQ‐C30) in the control group was  62.5	Mean QoL (EORTC QLQ‐C30) in the intervention group was 77.38 (67.33 to 87.43)	MD 14.88  (4.83 to 24.93)	50  (1 study)	⊕⊕⊝⊝  Low^a
Patient‐reported symptom: nausea/vomiting (EORTC QLQ‐C30)	Mean nausea/vomiting score (EORTC QLQ‐C30) in the control groups was  46.67	Mean nausea/vomiting score (EORTC QLQ‐C30) in the intervention group was 8.10 (−0.2 to 18.2)	MD −38.57  (−48.67 to −28.47)	50  (1 study)	⊕⊕⊝⊝  Low^a
Patient‐reported symptom: diarrhoea (EORTC QLQ‐C30)	Mean diarrhoea score (EORTC QLQ‐C30) in the control group was  30	Mean diarrhoea score (EORTC QLQ‐C30) in the intervention group was 16.19 (2.48 to 29.9)	MD −13.81  (−27.52 to −0.1)	50  (1 study)	⊕⊕⊝⊝  Low^a
Objective outcome measure: WBC count (× 10⁹/L)	Mean WBC count (× 10⁹/L) in the control group was  4.1	Mean WBC count (× 10⁹/L) in the intervention group was 5.82 (5.07 to 6.57)	MD 1.72  (0.97 to 2.47)	50  (1 study)	⊕⊕⊝⊝  Low^a
Objective outcome measure: haemoglobin (g/L)	Mean haemoglobin (g/L) in the control group was  9.67	Mean haemoglobin (g/L) in the intervention group was 11.73 (10.93 to 12.53)	MD 2.06  (1.26 to 2.86)	50  (1 study)	⊕⊕⊝⊝  Low^a
Objective outcome measure: platelets (× 10⁹/L)	Mean platelet count (× 10⁹/L) in the control group was 172.9	Mean platelet count (× 10⁹/L) in the intervention group was 383.69 (339.92 to 427.46)	MD 210.79  (167.02 to 254.56)	50  (1 study)	⊕⊕⊝⊝  Low^a
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; MD: mean difference; EORTC QLQ‐C30: European Organization for Research and Treatment of Cancer Quality of Life Questionnaire; QoL: quality of life.
GRADE Working Group grades of evidence  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Moxibustion + conventional treatment versus conventional medicine alone for side effects of chemotherapy or radiotherapy in cancer patients
Patient or population: patients receiving chemotherapy or radiotherapy for cancer treatment Settings: hospital Intervention: moxibustion plus conventional treatment  Comparison: conventional treatment^a
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect  (95% CI)	No of participants  (studies)	Certainty of the evidence  (GRADE)
	Assumed risk	Corresponding risk
	Conventional treatment	Moxibustion plus conventional treatment
The incidence and severity of toxicities: leukopenia (WHO grade 3 to 4)	107 per 1000	15 per 1000  (1 to 283)	RR 0.14   (0.01 to 2.64)	56  (1 study)	⊕⊕⊝⊝  Low^b
QoL (EORTC QLQ‐c30)	The mean QoL (EORTC QLQ‐c30) in the control groups was  65	The mean QoL (EORTC QLQ‐c30) in the intervention groups was 73.85 (69.25 to 78.46)	MD 8.85  (4.25 to 13.46)	134  (3 studies)	⊕⊕⊝⊝  Low^b
Patient‐reported symptom: nausea/vomiting (WHO grade 3 to 4)	152 per 1000	65 per 1000  (38 to 112)	RR 0.43   (0.25 to 0.74)	801  (7 studies)	⊕⊕⊝⊝  Low^b
Patient‐reported symptom: diarrhoea	33 per 1000	6 per 1000 (0 to 128)	RR 0.19 (0.01 to 3.88)	61 (1 study)	⊕⊕⊝⊝  Low^b
Objective outcome measure: WBC count (× 10⁹/L)	2 studies (N = 200) both reported that moxibustion was associated with a slightly higher mean white blood cell count compared with control (MD 0.5 × 10⁹/L; 95% CI 0.12 to 0.88; MD 1.5 × 10⁹/L; 95% CI 1.14 to 1.86). One (N = 62) found no evidence of a difference compared with control (MD 0.41 × 10⁹/L; 95% CI −0.22 to 1.04).		Not pooled due to high heterogeneity	262  (3 studies)	⊕⊕⊝⊝  Low^a
Objective outcome measure: haemoglobin (g/L)	The mean haemoglobin (g/L) in the control groups was  108	The mean haemoglobin (g/L) in the intervention groups was  111.97 (109.4 to 114.53)	MD 3.97  (1.4 to 6.53)	142  (2 studies)	⊕⊕⊝⊝  Low^b
Objective outcome measure: platelet (× 10⁹/L)	The mean platelet (× 10⁹/L) in the control group was 170	The mean platelet (× 10⁹/L) in the intervention group was 183.48 (154 to 212.95)	MD 13.48 (−16.00 to 42.95)	142  (2 studies)	⊕⊕⊝⊝  Low^b
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; EORTC QLQ‐C30: European Organization for Research and Treatment of Cancer Quality of Life Questionnaire; QoL: quality of life; RR: risk ratio; WBC: white blood cells; WHO: World Health Organization.
GRADE Working Group grades of evidence  High quality: further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: we are very uncertain about the estimate.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Leukopenia (WHO grade 3 to 4)	1	72	Risk Ratio (IV, Random, 95% CI)	0.5 [0.10, 2.56]
2 WBC count (× 10⁹/L)	1	80	Mean Difference (IV, Random, 95% CI)	1.77 [0.76, 2.78]
3 Haemoglobin (g/L)	1	66	Mean Difference (IV, Random, 95% CI)	1.33 [0.59, 2.07]
4 Platelets (× 10⁹/L)	1	65	Mean Difference (IV, Fixed, 95% CI)	30.80 [8.03, 53.57]
5 CD3 (g/L)	1	57	Mean Difference (IV, Random, 95% CI)	5.30 [1.46, 9.14]
6 CD4 (g/L)	2	113	Mean Difference (IV, Random, 95% CI)	5.42 [3.01, 7.82]
7 CD8 (g/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
8 IgA (g/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
9 IgM (g/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
10 IgG (g/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Karnofsky score	1	50	Mean Difference (IV, Random, 95% CI)	10.86 [5.10, 16.62]
2 QoL (EORTC QLQ‐C30) (version 3.0)	1	50	Mean Difference (IV, Random, 95% CI)	14.88 [4.83, 24.93]
3 Nausea/vomiting (EORTC QLQ‐C30) (version 3.0)	1	50	Mean Difference (IV, Random, 95% CI)	‐38.57 [‐48.67, ‐28.47]
4 Diarrhoea (EORTC QLQ‐C30) (version 3.0)	1	50	Mean Difference (IV, Random, 95% CI)	‐13.81 [‐27.52, ‐0.10]
5 WBC count (× 10⁹/L)	1	50	Mean Difference (IV, Random, 95% CI)	1.72 [0.97, 2.47]
6 Haemoglobin (g/L)	1	50	Mean Difference (IV, Random, 95% CI)	2.06 [1.26, 2.86]
7 Platelets (× 10⁹/L)	1	50	Mean Difference (IV, Random, 95% CI)	210.79 [167.02, 254.56]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Haematologic (adults) (WHO grade 3 to 4)	1	72	Risk Ratio (IV, Random, 95% CI)	0.57 [0.15, 2.20]
2 WBC count (× 10⁹/L) at the end of treatment	8		Mean Difference (IV, Random, 95% CI)	Subtotals only
2.1 Moxibustion vs batilol + leucogen	3	191	Mean Difference (IV, Random, 95% CI)	0.84 [0.12, 1.57]
2.2 Moxibustion vs leucogen + vitamin C + vitamin E	1	84	Mean Difference (IV, Random, 95% CI)	‐0.09 [‐0.84, 0.66]
2.3 Moxibustion vs G‐CSF	1	90	Mean Difference (IV, Random, 95% CI)	‐3.06 [‐3.30, ‐2.82]
2.4 Moxibustion vs ondansetron + batilol	1	163	Mean Difference (IV, Random, 95% CI)	3.01 [2.25, 3.77]
2.5 Moxibustion vs batilol + leucogen + G‐CSF(optional))	1	72	Mean Difference (IV, Random, 95% CI)	0.99 [0.76, 1.22]
2.6 Moxibustion vs leucogen + berbamine	1	102	Mean Difference (IV, Random, 95% CI)	1.80 [0.95, 2.65]
3 WBC count (× 10⁹/L) after follow‐up (8 days)	1	90	Mean Difference (IV, Random, 95% CI)	0.40 [0.15, 0.65]
4 Haemoglobin (g/L)	2	235	Mean Difference (IV, Random, 95% CI)	10.28 [4.51, 16.05]
5 Platelets (× 10⁹/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
6 Karnofsky score	1	82	Mean Difference (IV, Random, 95% CI)	6.70 [2.37, 11.03]
7 CD3 (g/L)	2	247	Mean Difference (IV, Random, 95% CI)	0.69 [‐0.64, 2.02]
8 CD4 (g/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
9 CD8 (g/L)	2		Mean Difference (IV, Random, 95% CI)	Subtotals only
10 IgA (g/L)	1	41	Mean Difference (IV, Random, 95% CI)	2.84 [2.30, 3.38]
11 IgG (g/L)	1	41	Mean Difference (IV, Random, 95% CI)	7.31 [6.05, 8.57]
12 IgM (g/L)	1	41	Mean Difference (IV, Random, 95% CI)	2.06 [1.66, 2.46]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Leukopenia (WHO grade 3 to 4)	1	56	Risk Ratio (IV, Random, 95% CI)	0.14 [0.01, 2.64]
2 WBC count (× 10⁹/L)	3		Mean Difference (IV, Random, 95% CI)	Subtotals only
3 Nausea/vomiting (WHO grade 3 to 4)	7	801	Risk Ratio (IV, Random, 95% CI)	0.43 [0.25, 0.74]
4 Vomiting (CTCAE v3.0 grade 3 to 5)	1	169	Risk Ratio (IV, Random, 95% CI)	0.07 [0.00, 1.14]
5 Diarrhoea (WHO grade 3 to 4)	1	61	Risk Ratio (IV, Random, 95% CI)	0.19 [0.01, 3.88]
6 Haemoglobin (g/L)	2	142	Mean Difference (IV, Random, 95% CI)	3.97 [1.40, 6.53]
7 Platelets (×10⁹/L)	2	142	Mean Difference (IV, Random, 95% CI)	13.48 [‐16.00, 42.95]
8 Karnofsky score	4	252	Mean Difference (IV, Random, 95% CI)	7.21 [5.74, 8.68]
9 QoL (EORTC QLQ‐c30)	3	134	Mean Difference (IV, Random, 95% CI)	8.85 [4.25, 13.46]
10 QoL (FACT‐G)	1	332	Mean Difference (IV, Random, 95% CI)	11.51 [10.64, 12.38]
11 QoL (FACT‐L)	1	60	Mean Difference (IV, Fixed, 95% CI)	10.04 [7.63, 12.45]
12 Physical well‐being (FACT‐L)	1	72	Mean Difference (IV, Random, 95% CI)	‐4.33 [‐6.25, ‐2.41]
13 IgA (g/L)	1	61	Mean Difference (IV, Random, 95% CI)	0.55 [0.21, 0.89]
14 IgG (g/L)	1	61	Mean Difference (IV, Random, 95% CI)	2.11 [1.19, 3.03]
15 IgM (g/L)	1	61	Mean Difference (IV, Random, 95% CI)	0.40 [0.19, 0.61]

Methods	Design: parallel RCT Randomisation method: not reported Blinding: no Power calculation: no Dropouts/withdrawals: no
Participants	Patients with nasopharyngeal carcinoma (stage TNM III and IVa), Karnofsky > 60 Number (treatment/control): 56 (28/28) Mean age (range): 43.7 (18‐71) Gender (M/F): 30/26 Country: Guangdong province, China Setting: hospital
Interventions	Indirect moxa cone on salt + conventional treatment vs conventional treatment Treatment group Indirect moxa cone on salt 10 continuous indirect moxa cones on salt on the acupoint RN8 (shenque), once per day Treatment duration: 30 days Chemotherapy with doxorubicin and cisplatin for 1 course Radiotherapy for 7 to 8 weeks Conventional treatment Control group Chemotherapy with doxorubicin and cisplatin for 1 course Radiotherapy for 7 to 8 weeks Conventional treatment Conventional treatment Symptomatic treatment
Outcomes	WHO Recommendations for Grading of Acute and Subacute Toxicity (1981) at the end of treatment Haemoglobin WBC Platelets Nausea/vomiting Oral mucosa Hair loss
Notes	—
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described
Allocation concealment (selection bias)	Low risk	Envelope was used for random allocation
Incomplete outcome data (attrition bias)   All outcomes	Low risk	There was no loss to follow‐up. All participants were included in the analysis.
Selective reporting (reporting bias)	Unclear risk	Few outcome measures were reported
Other bias	Unclear risk	It was mentioned that the groups were comparable, but no baseline characteristics data were presented.
Blinding of participants and personnel (performance bias)   All outcomes	High risk	No blinding
Blinding of outcome assessment (detection bias)   Subjective outcomes	High risk	No blinding
Blinding of outcome assessment (detection bias)   Objective outcomes	Low risk	No blinding; however, objective outcomes detected by machine were not influenced substantially

Study	Reason for exclusion
Chen 1991	No relevant outcome measure was available in the study report
Chen 2006	Moxibustion plus acupuncture treatment was used as the treatment intervention.
Chen 2010	There was no random allocation for intervention assignment.
Chen 2012	No relevant outcome measure was available in the study report.
Ding 2008	Although random allocation was mentioned in the abstract, sequence of admission to hospital determined group assignment.
Fan 2011	Although random allocation was mentioned in the abstract, visiting order determined group assignment.
Ge 2011	No relevant outcome measure was available in the study report.
Guo 2011	No relevant outcome measure was available in the study report.
Huang 2015	No relevant outcome measure was available in the study report.
Jiang 2002	Acupressure was applied after moxibustion in the treatment group.
Jin 2003	Acupuncure was used together with moxibustion in the treatment group.
Li 2007	No relevant outcome measure was available in the study report.
Liang 2014	The conventional medicine in 2 groups varied depending on the symptoms.
Liu 2001	The treatment duration was different between the comparison groups.
Liu 2002	The combination treatment of chemotherapy and Chinese herbal medicine was used as the control intervention.
Liu 2006	The treatment group assignment was determined by the patient preference.
Liu 2013	The treatment information about chemotherapy was not provided.
Long 2012	It was a cross‐over study, but no phase I study data were available.
Ou 1992	No information about the outcome data was provided.
Qiu 2008	Acupressure was administered with moxibustion in the treatment group.
Shao 2012	No relevant outcome measure was available in the study report.
Shen 2002	No relevant outcome measure was available in the study report.
Shen 2010	It has the same results on some outcomes as Fan 2001. It was possibly a fake article.
Shen 2011	No relevant outcome measure was available in the study report.
Song 2003	No relevant outcome measure was available in the study report.
Tang 2011	Warm needle moxibustion was used as the treatment intervention.
Wang 2016	The patients with cancer pain might not receive chemotherapy or radiotherapy.
Xiang 2011	Acupressure was administered with moxibustion in the treatment group.
Xu 2008	No random allocation
Yao 1998	The sequential balanced coefficient method was used for intervention assignment.
Zhang 2014a	The dose of G‐CSF in the control group varied.
Zhong 2011	Acupressure was administered with moxibustion in the treatment group.
Zhong 2014	The conventional medicine in the control group varied according to the side effects of chemotherapy.

Methods	Cross‐over RCT
Participants	Patients with lung cancer, Karnofsky > 60
Interventions	Moxa stick + conventional treatment vs conventional treatment
Outcomes	WHO Recommendations for Grading of Acute and Subacute Toxicity (Vomiting) (Miller 1981)
Notes	Information on the number of participants in the treatment and control group in phase I and II trial study was not reported. No response has been received after sending enquiry mail at the time of drafting the review.

Methods	RCT
Participants	Cancer patients receiving chemotherapy
Interventions	Indirect moxibustion on ginger + conventional treatment vs conventional treatment
Outcomes	Nause and vomiting assessed by a problematic grading criteria
Notes	Description on the outcome assessment time was unclear. No response has been received to enquiry mail at the time of drafting the review.

Methods	RCT
Participants	The digestive people with cancer receiving chemotherapy
Interventions	Moxa stick treatment on herbal paste + chemotherapy vs chemotherapy
Outcomes	Grading of toxic effect, WBC, blood platelet
Notes	The treatment duration was not reported. No response has been received after sending enquiry mail at the time of drafting the review.

Methods	RCT
Participants	People with cervical cancer receiving radiotherapy
Interventions	Moxa stick treatment + conventional medicine vs conventional medicine
Outcomes	Radiation Therapy Oncology Group (lower GI)
Notes	The moxibustion treatment lasted during the whole period of radiotherapy; however, no information about the treatment duration was reported. No response has been received to enquiry mail at the time of drafting the review.

Methods	RCT
Participants	Cancer patients receiving radiotherapy
Interventions	Indirect moxa cone on ginger + conventional medicine vs conventional medicine
Outcomes	WBC count, Hb, platelets
Notes	The moxibustion treatment duration was not provided. No response has been received to enquiry mail at the time of drafting the review.

Methods	RCT
Participants	Cancer patients receiving chemotherapy with expected survival time of more than 3 months and Karnofsky > 60
Interventions	Indirect moxa cone on ginger vs conventional medicine
Outcomes	WBC count, Hb, platelets, Karnofsky score
Notes	The information about treatment duration was unclear. No response has been received to enquiry mail at the time of drafting the review.

PERMALINK

Moxibustion for alleviating side effects of chemotherapy or radiotherapy in people with cancer

Hong Wei Zhang

Zhi Xiu Lin

Fan Cheung

William Chi‐Shing Cho

Jin‐Ling Tang

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Summary of findings table

Results

Description of studies

Results of the search

1.

Included studies

Excluded studies

Risk of bias in included studies

2.

3.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

Summary of findings for the main comparison. Moxibustion versus no treatment for side effects of chemotherapy or radiotherapy in cancer patients.

Summary of findings 2. Moxibustion versus sham treatment for side effects of chemotherapy or radiotherapy in cancer patients.

Summary of findings 3. Moxibustion versus conventional medicines for side effects of chemotherapy or radiotherapy in cancer patients.

Summary of findings 4. Moxibustion + conventional treatment versus conventional medicine alone for side effects of chemotherapy or radiotherapy in cancer patients.

Moxibustion versus no treatment

Incidence and severity of chemotherapy‐ or radiotherapy‐related toxicities: leukopenia

1.1. Analysis.

Other objective outcome measures aimed at assessing side effects of chemotherapy or radiotherapy

WBC counts

1.2. Analysis.

Haemoglobin concentration

1.3. Analysis.

Lymphocyte counts

1.5. Analysis.

1.6. Analysis.

1.7. Analysis.

Platelets

1.4. Analysis.