Invasive acupuncture for gastroparesis after thoracic or abdominal surgery: a systematic review and meta-analysis

Objectives This meta-analysis aimed to systematically evaluate the efficacy of acupuncture in treating postsurgical gastroparesis syndrome (PGS) after thoracic or abdominal surgery. Design Systematic review and meta-analysis. Data sources Twelve databases (PubMed, Embase, Cochrane Library Cochrane Central Register of Controlled Trials (CENTRAL), Medline (Ovid) (from 1946), Web of Science, EBSCO, Scopus, Open Grey, China National Knowledge Infrastructure (CNKI), Wanfang Database, Chinese Scientific Journals Database (VIP) and China Biology Medicine disc (CBM)) and three registration websites (WHO International Clinical Trials Registry Platform (ICTRP), ClinicalTrials.gov, and Chinese Clinical Trial Registry (ChiCTR)) were searched from the inception to September 2022, and citations of the included literature were screened. Eligibility criteria All randomised controlled trials addressing invasive acupuncture for PGS. Data extraction and synthesis Key information on the included studies was extracted by two reviewers independently. Risk ratio (RR) with 95% CI was used for categorical data, and mean difference with 95% CI for continuous data. The quality of evidence was assessed using Grading of Recommendations Assessment, Development and Evaluation. Outcomes were conducted with trial sequential analysis (TSA). Results Fifteen studies with 759 patients met the inclusion criteria. Subgroup analyses revealed that compared with the drug group, the drug and acupuncture group had a greater positive effect on the total effective rate (TER) (nine trials, n=427; RR=1.20; 95% CI 1.08 to 1.32; P-heterogeneity=0.20, I2=28%, p=0.0004) and the recovery rate (RCR) (six trials, n = 294; RR = 1.61; 95% CI 1.30 to 1.98; P-heterogeneity=0.29, I2=19%, p<0.0001) of PGS after abdominal surgery. However, acupuncture showed no significant advantages in terms of the TER after thoracic surgery (one trial, p=0.13) or thoracic/abdominal surgery-related PGS (two trials, n = 115; RR=1.18; 95% CI 0.89 to 1.57; P-heterogeneity=0.08, I2=67%, p=0.24) and the RCR after thoracic/abdominal surgery (two trials, n=115; RR=1.40; 95% CI 0.97 to 2.01; P-heterogeneity=0.96, I2=0%, p=0.07). The quality of evidence for TER and RCR was moderate certainty. Only one study reported an acupuncture-related adverse event, in the form of mild local subcutaneous haemorrhage and pain that recovered spontaneously. TSA indicated that outcomes reached a necessary effect size except for clinical symptom score. Conclusion Based on subgroup analysis, compared with the drug treatment, acupuncture combined drug has significant advantages in the treatment of PGS associated with abdominal surgery, but not with thoracic surgery. PROSPERO registration number CRD42022299189.


GENERAL COMMENTS
In the current study, the authors investigate the efficacy and safety of acupuncture in the treatment of postsurgical gastroparesis syndrome (PGS) after thoracic or abdominal surgery in the form of meta-analysis by including 12 databases and 3 trials. They including 15 studies with 759 patients and found that acupuncture had a greater positive effect on the TER and RR. They declared acupuncture has a significant advantage in the treatment of PGS associated with abdominal surgery, while it hasn't in PGS associated with thoracic surgery.
I have some comments that should be addressed by the authors. 1.The authors should provide the full names for the abbreviations the first time they appeared. 2.The authors should check their manuscript carefully and make consistence of the style. 3. Clinical importance of this meta-analysis should be emphasized. What information have the authors provided for clinicians and researchers? 4. Is there any meta-analysis focused on this topic, the author should discuss the priority of this one. More discussions were needed. 5. Trial sequential analysis should be performed to see if more investigations were needed. (Fu W, et al. Oncotarget. 2017;8: 9806-9822.) written, with clear objectives and a well-constructed analysis. However, I have a few comments / concerns: -The authors have defined "total effective rate" and "recovery rate" as their primary study outcomes. It is unclear what these relate to, and whether the definitions are homogeneous across the 12 and 8 trials, respectively, which assess these outcomes. Please comment and discuss.
-The authors have concluded that acupuncture may be safe to perform due to improvements in clinical symptoms and lack of demonstrated serious adverse events. I would suggest that a more cautious and conservative statement is used (or this is removed from the manuscript conclusion entirely) given that adverse outcomes associated with acupuncture were reported in only two trials. Furthermore, heterogeneity among trials which used the clinical symptom score was very large (I2=100%), despite the authors' subgroup analysis based on the site of surgery. 3. Line 145: They noted "Subgroup studies were conducted", but I guess it would be "subgroup analyses". 4. In Statistical Analyses section, they should mention they used the Egger test for the publication bias test. 5. Although the Egger test was not significant, the funnel plot in S- Figure 4 was substantially asymmetric. It might indicate publication biases, and it should be at least mentioned as a limitation of this study in Discussion section. 6. The English should be checked by native English speakers.

VERSION 1 -AUTHOR RESPONSE
Reviewer: 1 Dr. Jing He, Guangzhou Women and Children's Medical Center Comments to the Author: In the current study, the authors investigate the efficacy and safety of acupuncture in the treatment of postsurgical gastroparesis syndrome (PGS) after thoracic or abdominal surgery in the form of meta-analysis by including 12 databases and 3 trials. They including 15 studies with 759 patients and found that acupuncture had a greater positive effect on the TER and RR. They declared acupuncture has a significant advantage in the treatment of PGS associated with abdominal surgery, while it hasn't in PGS associated with thoracic surgery.
I have some comments that should be addressed by the authors. Comment 1. The authors should provide the full names for the abbreviations the first time they appeared.
Response 1. Thank you for your comments, and we have carefully revised our manuscript to make sure that full names for the abbreviations are provided at the first time they appeared.
Comment 2. The authors should check their manuscript carefully and make consistence of the style. Response 2. We apologize for this confusion generated by the previous version of the manuscript. We have checked it carefully and kept the style consistent and invited native English speakers for language corrections to make it readable. We really hope that the readability of our manuscript has been substantially improved. Note from Editor-While a trial sequential analysis would be a useful addition this is not a mandatory revision under our formatting guidelines.
Reviewer: 2 Dr. Sameer Bhat, The University of Auckland Comments to the Author: Fan and colleagues performed a systematic review and meta-analysis to assess the efficacy and safety of invasive acupuncture for treating postsurgical gastroparesis syndrome after thoracic or abdominal surgery. Overall, the study is well researched and written, with clear objectives and a wellconstructed analysis. However, I have a few comments / concerns: Comment 1. The authors have defined "total effective rate" and "recovery rate" as their primary study outcomes. It is unclear what these relate to, and whether the definitions are homogeneous across the 12 and 8 trials, respectively, which assess these outcomes. Please comment and discuss. Response 1. Thank you. We have explained this in detail in our manuscript. At present, we have added this part in the Outcome assessment of Methods section [Line 155]. In the trials included in this meta-analysis, the evaluation criteria for the total effective rate (TER) and the recovery rate (RCR) of postsurgical gastroparesis syndrome (PGS) were centered on these proposed by the China Association of Chinese Medicine (CACM), which are the combination of objective indicators (gastric juice volume, gastrointestinal radiography) and subjective indicators (whether nausea, vomiting, and bloating disappear). Recovered: Objective indicators: There was no gastric juice volume, and gastrointestinal radiography indicated good gastric motility. Subjective indicators: Symptoms such as nausea, vomiting, and bloating disappeared after eating without gastrointestinal decompression. Effective: Objective indicators: The gastric juice volume decreased significantly (＜150ml/d), and gastrointestinal radiography indicated gastric hypomotility slightly. Subjective indicators: Symptoms such as nausea, vomiting, and bloating appeared without gastrointestinal decompression, and inability to eat. TER refers to the proportion of recovered and effective people in all participants. RCR refers to the proportion of recovered people in all participants. Similar situations have been found in published meta-analysis literatures in English [1,2].
The authors have concluded that acupuncture may be safe to perform due to improvements in clinical symptoms and lack of demonstrated serious adverse events. I would suggest that a more cautious and conservative statement is used (or this is removed from the manuscript conclusion entirely) given that adverse outcomes associated with acupuncture were reported in only two trials. Response 2. We strongly agree with you that there were too few literatures on "adverse events related to acupuncture" to reach the conclusion that "acupuncture may be safe", and more trials are needed to study it. Therefore, we removed this conclusion from the Abstract section and the Discussion section. And we described it more cautious and mentioned it as a limitation of our study in the Discussion section [Line 419, 486].
Comment 3. Furthermore, heterogeneity among trials which used the clinical symptom score was very large (I2=100%), despite the authors' subgroup analysis based on the site of surgery. Response 3. We initially considered that the large heterogeneity of trials reported the clinical symptom score was mainly due to the inconsistent scoring criteria used in them. We attempted a meta-analysis of the three trials which adopted completely consistent scoring criteria. Unfortunately, there was still significant heterogeneity. This may be because the CSS was subjective indicator and there were possibilities of bias. At present, there are too few relevant trials to analyze the reasons of heterogeneity. This requires more high-quality trails for further study and we have discussed it in the Discussion section and mentioned it as a limitation in the meta-analysis [Line 408, 479]. Thank you again for raising important questions to help us to improve the quality of the manuscript. We have tried our best to revise and explain these comments / concerns. If there are any more, we would be very happy to revise and explain them. Comment 6. The English should be checked by native English speakers. Response 6. We apologize for the poor language of our manuscript. We worked on the manuscript for a long time and repeated addition and removal of sentences obviously led to poor readability. We have invited native English speakers for language corrections to make our manuscript readable. These revises will not influence the content and framework of the manuscript. And here we did not list the changes but marked in red-colored text in the revised manuscript. We really hope that the language level has been improved substantially.

GENERAL COMMENTS
Thank you for considering my comments. All of my concerns were adequately addressed.