Acupuncture Therapy plus Hyaluronic Acid Injection for Knee Osteoarthritis: A Meta-Analysis of Randomized Controlled Trials

Objective This meta-analysis aimed to investigate the effectiveness of acupuncture therapy plus hyaluronic acid injection versus hyaluronic acid injection alone for patients with knee osteoarthritis. Methods Relevant randomized controlled trials that compared the combined effect of acupuncture therapy and hyaluronic acid injection with hyaluronic acid injection alone for knee osteoarthritis patients were included. 10 studies were included in this meta-analysis, and the relative risk (RR) and weight mean difference (MD) with 95% CI for the Lysholm knee score (LKSS), visual analogue scale (VAS), and effective rate (ER) were evaluated by using RevMan 5.3 software. Besides, the bias assessment of the included studies was evaluated using the Cochrane risk of bias tool, and the GRADE (Grading of Recommendations, Assessment Development, and Evaluation) system was applied to assess the overall quality of the evidence. Results A total of 10 studies involving 998 participants were included in this study. Compared to hyaluronic acid injection alone, the combined therapy significantly reduced pain on the visual analogue scale (VAS) and improved the ER and knee function on the Lysholm knee score (LKSS). Of these, the pooled LKSS (MD = 8.09, 95% CI = [7.02, 9.16], p < 0.00001, 7 studies) and ER (RR = 1.23, 95% CI 1.15 to 1.31, p < 0.00001, 8 studies) suggested that combination therapy yielded a significantly higher ER and improved the LKSS scores to a greater degree than hyaluronic acid injection alone in patients with KOA. The VAS (MD = −1.39, 95% CI = [−1.99, −0.79], p < 0.00001, 7 studies) showed that the combined therapy significantly reduced pain than hyaluronic acid injection alone. The quality of evidence for the main outcomes was from very low to low according to the GRADE system. Conclusion Current evidence suggests that acupuncture therapy combined with hyaluronic acid injection is more effective in alleviating pain, improving the ER and knee function compared with hyaluronic acid injection alone. However, considering the low quality, small size, and high risk of the studies identified in this meta-analysis, more higher methodological quality, rigorously designed randomized controlled trials with large sample sizes are needed to confirm the results.


Introduction
Knee osteoarthritis is a highly common chronic degenerative disease in middle-aged and elderly people, with more female patients than males, which not only seriously affects joint function and the quality of life, but also becomes a serious public health problem worldwide [1][2][3]. With accelerating aging process of the social population, the morbidity of degenerative osteoarthritis in the world has increased significantly [4]. KOA is generally thought to be caused by a Among these conservative therapies, acupuncture, known as an ancient, complementary, and alternative therapeutic technique [11], has been empirically practiced and improved over thousands of years in Asian countries and verified to be the most effective and popular therapies in treating the pain symptoms and functional disability of patients with KOA based on integral concepts and syndrome differentiation of the TCM system. e effectiveness of acupuncture for treating many diseases has been verified in a series of high-quality clinical trials [12][13][14]. According to the ACR guidelines, acupuncture is conditionally recommended for KOA for its pain relief, affordability, and safety [15][16][17][18]. In addition, hyaluronic acid products, commonly used as alternative intraarticular injection and recommended by the Food and Drug Administration for treatment of KOA in 1997 [19], have the function of viscoinduction properties and increasing intraarticular lubrication, inhibiting inflammatory mediators and promoting repair of cartilage to delay the progression of KOA [20,21]. Altman et al. [22] has reported that HA injections are generally well tolerated, provide a longer duration of symptomatic relief, and improve knee function for patients with knee OA.
However, acupuncture therapy combined with HA injection is also a common method for KOA treatment, but whether the combined application of these two methods is better than HA alone is still lacking of systematic evaluation. erefore, we conducted the present meta-analysis of randomized controlled trials (RCTs) to assess the efficacy of acupuncture therapy plus HA injection compared with that of HA alone in patients with knee OA.

Bias Assessment of the Included Studies.
Risk of bias was assessed independently by two reviewers using the Cochrane risk of bias tool for the following seven criteria: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other bias. e judging criteria were as follows: "Yes" for low risk, "Unclear" for unclear, and "No" for high risk. A third reviewer appraised the discrepancy and made the final decision regarding the ratings.

Statistical Analyses.
e weighted mean difference (MD) with 95% CI was applied for continuous data, and the pooled relative risk (RR) and 95% confidence interval (CI) was reported for the clinical efficacy rate (ER). e potential heterogeneity among the included studies was assessed by the chi-square test and the inconsistency index statistic (I 2 ), and the fixed-effects model was applied to analyze the data without obvious heterogeneity (p > 0.1 and I 2 < 50%); otherwise, for p > 0.1 and I 2 > 50%, it was considered that the trials included were heterogeneous and a random-effects model was used, and we would try to explore the potential sources of it by sensitivity analysis. In view of different acupuncture therapy techniques, subgroup analyses were further performed. Additionally, the potential publication bias was assessed by applying the funnel plot if the number of studies for an outcome was adequate (n ≥ 10).
2.6. Quality of Evidence. We used the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) working group method to appraise the quality of evidence, and the GRADEpro software (version 3.6 for Windows, GRADE Working Group) was applied.

Literature Selection.
We identified 1071 studies based on our retrieval strategies, of which 992 studies were removed by screening the titles and abstracts since they were not qualified according to the predefined inclusion criteria. A further 69 studies were excluded according to the exclusion criteria through reviewing the full texts. Finally, ten studies were included in the meta-analysis. Figure 1 describes the detailed selection process for relevant studies.

Characteristics of Included
Literature. In the 10 eligible randomized controlled trials involving 998 patients (470 men and 528 women), 501 received acupuncture therapy plus hyaluronic acid injection in the intervention group and 497 received hyaluronic acid injection in the control group. Among them, the age range of participants adopted into the meta-analysis ranged from 24 to 86. All of the included studies showed no significant difference with the baseline, and all of that were conducted in China and published in Chinese. ese included trials were all published between 2012 and 2018.

Methodological Quality of Included Studies.
e risk of bias assessment for the included studies is shown in Figures 2 and 3. In the generation of randomization sequence, patients were randomized by using the random number table in 6 studies [23,26,[28][29][30]32] and by applying the SAS software in one study [31] and the remaining 3 studies [24,25,27] only mentioned "random" or "randomization" without describing the explicit randomization technique. Noticeably, no studies described allocation concealment and method of blinding in a detailed way. Given the characteristics of the design between the intervention and control groups, it was not possible to blind participants and personnel, thus all studies were judged to be at high risk of bias. Additionally, unclear risk of bias was observed across all studies for detection bias and low risk of bias for attrition and reporting bias. [23-27, 29, 32]. e meta-analysis indicated superior effects of combination therapy on LKSS improvement (MD � 8.09, 95% CI � [7.02, 9.16], p < 0.00001). However, a heterogeneity test (p for heterogeneity � 0.001, I 2 � 72%) indicated that there was moderate statistical heterogeneity between studies; therefore, a random-effects model was used and we conducted subgroup analyses based on distinct acupuncture techniques: warm acupuncture (WA), acupotomy (AT), and acupotomy plus warm acupuncture (WA + AT). Subgroup analyses showed remarkably increases of the Lysholm scores in the WA combination therapy with severe heterogeneity (MD � 8.05, 95% CI � [6.65, 9.44], p < 0.00001, p for heterogeneity � 0.0001, p � 86%); AT combined with HA injection similarly enhances Lysholm scores relative to the control group treated with HA injection alone (MD � 7.90, 95% CI � [5.84, 9.96], p < 0.00001) with no heterogeneity (p for heterogeneity � 0.51, I 2 � 0%) and AT plus WA (MD � 8.62, 95% CI � [5.82, 11.42], p < 0.00001) (Figure 4).

Random sequence generation (selection bias)
Allocation concealment (selection bias)

Quality of Evidence.
e GRADEpro was utilized to evaluate the quality of evidence for the meta-analysis. We assessed the Lysholm knee score (LKSS), the visual analogue scale (VAS), and the clinical effective rate (ER). Details are shown in the GRADE evidence profile, and the summary of findings is shown in Table 2.
e results show that the quality of evidence was from low for the assessment of LKSS because 3 studies mentioned "random" or "randomization" without describing the detailed randomization technique. Besides, all of the studies did not specifically describe the method of allocation concealment, there was a high risk of performance bias across the studies due to the difficulty of blinding to participants and personnel, and the number of adverse events of included studies was not sufficiently reported, thus the quality of evidence of publication bias was downgraded. For VAS, the quality of evidence was very low because of the selection and performance bias, the significant heterogeneity, and the insufficient adverse events. For ER, the quality of evidence was low due to the selection and performance bias and the insufficient adverse events.

Preferred Acupuncture Points.
We collected the data related to the acupuncture points selected from the enrolled studies and found that six acupuncture points were most frequently used to treat KOA, namely, Neixiyan (Ex-LE4), Xuehai (SP10), Yanglingquan (GB34), Zusanli (ST36), Weizhong (BL40), and Shenshu (BL23). e overall acupuncture points of these included studies are shown in Appendix 2.

Adverse Events.
Of the 10 studies, 2 studies [23,31] reported that no serious adverse events were observed during the treatment, one study [30] reported that there were injection site pain (2 cases in the experimental group and 1 case in the control group) and knee discomfort (1 case

Principal Findings.
To our knowledge, this is the first meta-analysis regarding published evidence to investigate the effectiveness of the combination of acupuncture therapy plus hyaluronic acid injection versus hyaluronic acid injection alone for treating KOA patients. Acupuncture therapy and hyaluronic acid injection as routine therapies are commonly used to treat patients with knee osteoarthritis based on traditional Chinese medicine or modern Western medicine. However, no strong evidence provided by previous researches has demonstrated whether the combination therapy has a more positive impact on KOA than hyaluronic acid injection alone. erefore, in the present study, a metaanalysis of relevant RCTs was designed to provide a reliable quantitative evaluation of existing evidence on the effectiveness of acupuncture therapy and hyaluronic acid injection for the treatment of KOA.
To comprehensively assess the value of acupuncture therapy plus hyaluronic acid injection, data on Lysholm knee score (LKSS), visual analogue scale (VAS), and clinical effective rate (ER) were chosen to evaluate for improvements in multiple dimensions among patients with KOA. Based on our meta-analysis, the pooled data of LKSS and VAS showed that the combined application of acupuncture therapy and HA injection improved knee function and alleviated pain to a significantly greater degree (MD � 8.09, 95% CI � [7.02, 9.16], p < 0.00001; MD � − 1.39, 95% CI � [− 1.99, − 0.79], p < 0.00001, respectively), than HA alone. e estimated effect sizes of ER suggested that the effective rate was significantly better in the combination therapy group than in the HA injection alone group (RR � 1.23, 95% CI 1.15 to 1.31, p < 0.00001).
Overall, the effectiveness of acupuncture therapy plus HA injection for patients with KOA was confirmed by this meta-analysis. Regarding acupuncture points, we summarised the most commonly adopted six acupuncture points based on the included studies and thereby provide recommendations for clinical and research settings.

Limitations of ese Studies.
e current meta-analysis has a number of limitations that must be acknowledged, as shown in the following: (1) No studies enrolled in this meta-analysis had mentioned allocation concealment, blinding of participants and personnel, and blinding of outcome assessments.

AT
Heterogeneity: tau 2 = 0.02; chi 2 = 2.46, df = 1 (P = 0.12); I 2 = 59% Test for overall effect: Z = 8.64 (P < 0.00001)  Evidence-Based Complementary and Alternative Medicine endowed, widely researched and practiced, and all published in Chinese language, but KOA is a worldwide disease. Besides, the risk of publication bias with the funnel plot cannot be evaluated because of limited number of trials. erefore, the results might have language and reporting bias.

WA+AT
(4) e duration of therapy in the majority of the included studies lasted 4-5 weeks, and only four studies had mentioned the follow-up period ranging from 3 months to 1 year. However, KOA is a chronic disease which should include adequate duration of therapy and follow-up periods in the observations of researches. (5) Our review may be affected by the high heterogeneity, and AEs were not sufficiently reported.

Conclusion
is meta-analysis of 10 randomized controlled trials provides evidence confirming that acupuncture therapy plus hyaluronic acid injection can improve the ER and LKSS and reduce VAS in patients with KOA compared with Total events 145 125 Heterogeneity: chi 2 = 0.31, df = 2 (P = 0.86); I 2 = 0% Test for overall effect: Z = 3.40 (P = 0.0007)

MA
Total events 67 49 Heterogeneity: chi 2 = 0.24, df = 1 (P = 0.62); I 2 = 0% Test for overall effect: Z = 3.51 (P = 0.0004) 79   High risk of performance bias; one of the studies describe the allocation concealment. 2 Adverse events were not sufficiently reported. 3 Significant heterogeneity. 8 Evidence-Based Complementary and Alternative Medicine hyaluronic acid injection alone. However, the findings should be interpreted cautiously because of the poor methodological quality and heterogeneity of the included studies. Consequently, further rigorously designed and higher quality trials with a larger sample size are necessary for overcoming the limitations of the current study and enhancing the strength of evidence.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Yang Zheng and Xiangdong Duan contributed equally to the design and conception of the study. e manuscript was drafted by Yang Zheng and Haibo Hu and reviewed by Renfu Quan. e search strategy was improved by all authors and performed by Yang Zheng and Mengran Wang, who also independently screened the potential studies. Conglin Ren and Haipeng Xu extracted data from the included studies, and Shangfeng Qi assessed the risk of bias and finished data synthesis. Renfu Quan arbitrated any discrepancies and ensured that no errors occur during the study. Besides, all authors have read and approved the final version of the manuscript.