Downloaded from on September 24, 2015 - Published by

Original paper

Heat-sensitive moxibustion in patients with osteoarthritis of the knee: a three-armed multicentre randomised active control trial Rixin Chen,1 Mingren Chen,1 Tongsheng Su,2 Meiqi Zhou,3 Jianhua Sun,4 Jun Xiong,1 Zhenhai Chi,1 Dingyi Xie,1 Bo Zhang1

For numbered affiliations see end of article. Correspondence to Professor Rixin Chen, The Affiliated Hospital with Jiangxi University of TCM, No 445 Bayi Avenue, Nanchang City 330006, People’s Republic of China; [email protected] RC and MC contributed equally. Accepted 25 April 2015 Published Online First 21 May 2015

To cite: Chen R, Chen M, Su T, et al. Acupunct Med 2015;33:262–269.


ABSTRACT Background In China, heat-sensitive moxibustion (HSM) is used for knee osteoarthritis (KOA) to reduce pain and improve physical activity. However, there is little high-quality evidence of its effectiveness. Objective To evaluate the effectiveness of HSM in the treatment of KOA compared with usual care. Methods We performed a multicentre, randomised controlled trial. In total, 432 patients with KOA were randomly assigned to one of three groups (HSM, conventional moxibustion, or conventional injection with sodium hyaluronate). The primary end point was the guiding principle of clinical research on new drugs in the treatment of KOA (GPCRND-KOA). Measurements were obtained at baseline and after 1 and 6 months (month 7) of study. Result For GPCRND-KOA, there were significant differences among the three groups after treatment at months 1 and 7. Pairwise comparisons showed that HSM was more effective than the conventional drug. There was no difference in any measures between conventional moxibustion and the conventional drug. Compared with conventional moxibustion, HSM resulted in greater improvement in all outcomes. Conclusions This trial provided some evidence of the superiority of HSM in patients with KOA, suggesting that the observed differences might be due to superiority effects of a heat-sensitive point, although the effect of expectation cannot be ruled out. Trial registration number The trial was registered at Controlled Clinical Trials: ChiCTRTRC-09000600.

BACKGROUND Knee osteoarthritis (KOA) is a common and disabling health problem in the elderly and significantly affects the lives of

many.1 2 Management of KOA is primarily aimed at pain relief and functional recovery.3 Although pharmacological treatments are widely used, they are associated with adverse events, such as gastrointestinal irritation and bleeding, perforating ulcers and renal and hepatic toxicity.4 Therefore, patients are recommended to follow non-pharmacological treatments in China, including complementary and alternative medicines, such as acupuncture and moxibustion. Moxibustion has been studied sufficiently to demonstrate that it does have benefits for KOA.5 6 One recent systematic review evaluated moxibustion for osteoarthritis and included six randomised controlled trials (RCTs) for KOA; meta-analysis showed favourable effects of moxibustion on the response rate.7 An experimental study found that indirect moxibustion could relieve pain in an experimental rat model of KOA and suggested the existence of sustained inhibitory modulation by endogenous opioids.8 Heat-sensitive moxibustion (HSM) is a form of treatment for pain and dysfunction associated with musculoskeletal conditions in China that involves administering suspended moxibustion at heat-sensitive acupuncture points.9 10 It has been used for various diseases, such as KOA, lumbar disc herniation and fibromyalgia syndrome. It originated from the observation that suspended moxibustion at certain body locations elicited heat sensitisation.11–13 Patients appear to become thermally sensitised to moxibustion stimulation at certain locations, as indicated by sensations of strong warmth or heat penetrating into the body (heat penetration), warmth spreading

Chen R, et al. Acupunct Med 2015;33:262–269. doi:10.1136/acupmed-2014-010740

Downloaded from on September 24, 2015 - Published by

Original paper around the stimulation site (heat expansion), or warmth conducted in certain directions and reaching some body regions or even internal organs remote from the stimulation sites (heat transmission).13 These heat-sensitised locations are not fixed, but may, during the progression of disease, change within a certain range, centred on acupuncture points. However, conventional moxibustion is applied at fixed acupuncture points. From our empirical evidence we formulated the following hypothesis: moxibustion at heat-sensitive acupuncture points has better efficacy than that at fixed acupuncture points.12 13 Some small sample studies have evaluated the effect of HSM. For example, Kang et al14 treated 40 patients with KOA by suspended moxibustion. The HSM group showed a higher recovery rate than the conventional moxibustion group (n=80) (80.95% vs 21.05%, respectively).14 Xie et al15 observed that HSM treatment reduced joint pain and improved functional disability, compared with the conventional moxibustion (n=120). However, these studies do not provide conclusive evidence of the effect of HSM, as the RCTs were limited by methodological defects, including small sample size, variability of control group, poor clinical practice and missing information. We therefore carried out a rigorous multicentre RCT with a large sample size, with the aim of assessing the effectiveness of HSM for treating KOA compared with conventional moxibustion or conventional drug. METHODS Design

We performed a multicentre (four centres in China), randomised, active controlled trial. Our trial was carried out in four hospitals, including the Affiliated Hospital of Jiangxi University of Traditional Chinese Medicine (TCM) in Nanchang, the first Affiliated Hospital of Anhui University of TCM in Hefei, Jiangsu TCM Hospital in Nanjing and Shanxi TCM Hospital in Xian. Patients were gathered through hospital-based recruitment and newspaper advertisements. Eligible participants were randomly allocated into one of three groups (the HSM group, the conventional moxibustion group, or the conventional drug treatment group) with a 1:1:1 allocation ratio by the central randomisation system (figure 1). This system was provided by the China Academy of Chinese Medical Sciences, using computer telephone integration technology to integrate computer, internet and telecom. The random number list was assigned by interactive voice response and interactive web response.16 The evaluation of participants and the analysis of results were performed by professionals blinded to the group allocation. Sample size

We calculated the sample size according to our previous pilot study. A two-sided 5% significance level and

Chen R, et al. Acupunct Med 2015;33:262–269. doi:10.1136/acupmed-2014-010740

95% power were considered to detect an effective rate difference (70% for HSM and 50% in the other groups). It was calculated that about 120 participants in each group were required. Assuming a dropout rate of 20%, the sample size would be 144 for each group (total n=432). Participants Recruitment

Patients were recruited in China between 30 December 2011 and 30 January 2013. Written consent was obtained from each participant before the start of this trial. We obtained oral and written consent from each participant before collecting information at the first visit. Inclusion criteria

We included participants who met the following criteria: diagnosed with idiopathic KOA according to the guiding principle of clinical research on new drugs (GPCRND-KOA),17 with a score of more than five points; moderate to severe swelling KOA; aged 38– 70 years and of either sex. Additionally, according to the KOA diagnosis standard, knee joints appeared swollen; a floating patella test was negative; patients accepted the treatment protocol in this trial; and acupuncture point heat-sensitisation phenomenon was present in the region bounded by SP9 (Yinlingquan), GB34 (Yanglingquan), ST34 (Liangqiu) and SP10 (Xuehai). The patients were instructed not to take any regular drugs before and during the treatment periods and were provided with the usual care instructions and rescue medication for KOA. Exclusion criteria

Participants were excluded if they were experiencing or had a history of the following: serious lifethreatening disease, such as heart disease or disease of brain blood vessels, liver, kidney and haematopoietic system, or if they were psychotic; diabetes, diabetic polyneuropathy and polyneuropathic disturbances; were pregnant or in the lactation period. The following conditions also led to exclusion: acute knee joint trauma or ulceration of local skin; complications of serious genu varus/valgus and flexion contraction. Study interventions

The treatments, which included moxibustion treatments, were provided by specialised acupuncturists who had at least 5 years’ training and 5 years’ experience using a standardised protocol. All treatment regimens and outcome assessment methods were standardised between the four centres and basic information about this study and the monitoring process was disseminated through workshops before the start of the study. All the patients were asked not to use any additional treatments such as physiotherapy or regular pain-killing drugs. 263

Downloaded from on September 24, 2015 - Published by

Original paper

Figure 1

A flowchart of the study process. ITT, intention to treat.

In the two moxibustion groups, a moxa stick (diameter 22 mm, length 120 mm) was applied ( produced by Jiangxi provincial TCM Hospital, China). The patients usually lay down in a comfortable supine position for treatment, with 24–30°C temperature in the room. Patients were asked to wear loose trousers, to allow the knees to be exposed.

HSM group

For the HSM group, the moxa sticks were lit by the therapist and held over the rectangle bounded by SP9, GB34, ST34 and SP10 to identify any heat sensitisation phenomenon present. The smouldering moxibustion stick was suspended about 3 cm over the skin to search for the acupuncture point heat-sensitisation phenomenon (box 1). When an acupuncture point showed at least one sensation listed in box 1, the therapists marked the point as a heat-sensitive point. We attempted to identify all sensitive points in each patient by repeated testing. 264

The therapists then began treatment, starting with the most heat-sensitive point. Treatment sessions ended when patients felt that the acupuncture point Box 1 Heat-sensitisation phenomenon at acupuncture points Heat penetration Heat penetrating from the skin into subcutaneous tissues. Heat expansion Heat expanding away from the stimulation site to surrounding cutaneous and subcutaneous tissues. Heat transmission Perceiving a stream of heat conducted in certain directions, or perceiving heat in some body regions or in the joint cavity. Non-thermal sensations Instead of thermal sensations, some patients perceiving aching, heaviness, pain, numbness, pressure, or cold in local or distant locations of stimulation.

Chen R, et al. Acupunct Med 2015;33:262–269. doi:10.1136/acupmed-2014-010740

Downloaded from on September 24, 2015 - Published by

Original paper heat-sensitisation phenomenon had disappeared, usually after 30–60 min. Patients received the treatment twice a day in the first week (5 days a week) and once a day from the second week to sixth week (a total of 35 sessions over 30 days). Conventional moxibustion group

A licensed doctor performed moxibustion at three acupuncture points around the affected knee— EX-LE5 (Xiyan, two points) and EX-LE2 (Heding)— for a total of 35 sessions for 30 days throughout the trial. Most procedures were similar to those used in the first group, except that each point was treated for 15 min at a time, according to current practice.18 The total time comprised 45 min for every treatment. In the treatment, patients usually felt local warmth without burning pain and might experience mild hyperaemia in the local region. The sensation of acupuncture point heat-sensitisation phenomenon was not pursued and not avoided in the treatment. Conventional drug group

According to pragmatic design, the control group did not use a placebo procedure, but received sodium hyaluronate intra-articular injection as the conventional drug. Many studies have confirmed the efficacy and safety of intra-articular hyaluronan for KOA.19 The injection was given every 6 days (2 mL) for a total of five times.

Table 1 List of items in the guiding principle of clinical research on new drugs in the treatment of knee osteoarthritis (GPCRND-KOA) Item



Pain or discomfort in night when lying in bed



Pain in activity or some position Pain in non-activity No


1 km 300 m–1 km 0.05, Fisher exact test). Outcome measures Total GPCRND-KOA score

Table 3 presents a comparison of GPCRND-KOA scores. The total score for each group decreased significantly during the 1-month trial and the 6-month follow-up visit ( p60 years, n (%) Sex, n (%) Female Male Duration of knee pain, n (%) 10 Years BMI (kg/m2), mean (SD) BMI (kg/m2), min–max GPCRND-KOA grade, n (%) Severe Moderate Knee circumference (cm), mean (SD) GPCRND-KOA score, mean (SD) Previous treatment (past half year), n (%) Pharmaceutical intervention Physiotherapy Previous acupuncture treatment BMI, body mass index; GPCRND-KOA, guiding



Heat-sensitive moxibustion group (n=144)

Conventional moxibustion group (n=144)

Conventional drug group (n=144)

55 (5.1) 41–70 19 (3.2)

53 (5.2) 40–70 20 (13.8)

56 (5.0) 42–66 18 (12.5)

56 (38.9) 88 (61.1)

53 (36.8) 91 (63.2)

57 (39.6) 87 (60.4)

78 (54.2) 39 (27.1) 27 (18.8) 22.1 (2.1) 15.4–30.2

75 (52.1) 46 (31.9) 23 (15.0) 23.2 (2.2) 16.2–31.1

73 (50.7) 40 (35.1) 31 (21.5) 23.1 (2.3) 13.4–33.1

85 (59.0) 59 (40.1) 41.2 (3.2) 11.2 (3.3)

88 (61.1) 56 (38.9) 42.2 (3.2) 11.3 (3.2)

82 (56.9) 62 (43.1) 41.3 (2.7) 12.1 (2.9)

76 (52.8) 12 (8.3) 18 (12.5) principle of clinical research on new drugs

81 (56.3) 79 (54.9) 16 (11.1) 13 (9.0) 16 (11.1) 16 (11.1) in the treatment of knee osteoarthritis score .

Chen R, et al. Acupunct Med 2015;33:262–269. doi:10.1136/acupmed-2014-010740

Downloaded from on September 24, 2015 - Published by

Original paper Table 3

Comparison of GPCRND-KOA scores* Month 1


Month 7

Mean (SD)

95% CI

Mean (SD)

95% CI

Heat-sensitive moxibustion group (n=144) 2.8 (1.8) 2.6 to 3.0 3.6 (1.6) 3.3 to 3.9 Conventional moxibustion group (n=144) 4.9 (2.8) 4.4 to 5.3 6.4 (1.5) 6.2 to 6.6 Conventional drug group (n=144) 5.6 (2.1) 5.4 to 5.9 7.0 (1.9) 6.6 to 7.3 Comparison of the three groups F Value 3.6 4.2 p Value 0.031 0.020 Heat-sensitive moxibustion group versus conventional moxibustion group q Value 3.8 5.9 p Value 0.035 0.023 Heat-sensitive moxibustion group versus conventional drug group q Value 6.2 8.2 p Value 0.011 0.0096 Conventional moxibustion group versus conventional drug group q value 2.1 2.7 p Value 0.130 0.091 *Comparison of the three groups by analysis of covariance. Pairwise comparison for the two groups by Student–Newman–Keuls (q test). All data are based on an intention to treat. GPCRND-KOA, guiding principle of clinical research on new drugs in the treatment of knee osteoarthritis score.

than those in conventional moxibustion group or conventional drug group at 1 and 7 months; however, there were no significant differences between the conventional moxibustion and conventional drug groups at either time point.

the HSM and conventional moxibustion groups was significant at both time points. Significant differences between the HSM group and the conventional drug group were also evident; however, there was no significant difference between the conventional moxibustion and conventional drug groups at both time points.

Knee circumference

In comparison with baseline, significant reductions in each group were seen at months 1 and 7 ( p

Heat-sensitive moxibustion in patients with osteoarthritis of the knee: a three-armed multicentre randomised active control trial.

In China, heat-sensitive moxibustion (HSM) is used for knee osteoarthritis (KOA) to reduce pain and improve physical activity. However, there is littl...
672KB Sizes 1 Downloads 5 Views