Participant recruitment
All patients with KOA will be recruited from the Tuina department of Yueyang Hospital of Integrated Traditional Chinese and Western medicine, Shanghai University of TCM. Meanwhile, potential patients will also be recruited through posters, the internet and leaflets. In addition, written informed consents will be taken from all patients. A total of 60 patients with KOA chronic pain will be randomly divided into Tuina group and Education group in a 1:1 ratio.
This study protocol has been approved by the Ethics Committee of Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of TCM. The trial was also registered to the Chinese Clinical Trial Registry.
Study design
This is a single-center and parallel randomized controlled trial. A total of 60 participants diagnosed with KOA based on American College of Rheumatology (ACR) criteria (1991 revised version)[33] will be considered as eligible patients. They will be randomly allocated into two equal groups with 30 patients in each group, including a Tuina group and a Education group. The interventions of both groups will last for 30 min and be carried out twice each week. The treatment period will be 12 weeks and the follow-up period will be 48 weeks.
Outcome measurements and MRI scans will be assessed at the baseline, the 6 and 12 weeks after treatment, followed by 48 weeks of clinical follow-up. The changes of clinical variables and cerebral activity of each group will be analyzed after data collection.
Assessor and analyst will be blinded to the group assignment in the study. Outcome assessment and statistical analyses will be performed by independent researchers who are blinded to the patient assignment. The trial flowchart and study design are shown in Figure 1 and Figure 2, respectively.
Inclusion criteria
- meet the diagnostic criteria for KOA set by the ACR in 1991[33]
- are aged between 40 and 60 years, right-handed, left knee pain
- have I-II degree knee-joint radiological change on the Kellgren-Lawrence scale
- have an average knee-pain score on a VAS ≥ 3 (range from 0 to 10) in the past 3 months
- Volunteer to take part in the study and sign the informed consent form
Exclusion criteria
- Are taking analgesics or anesthetics in the past 1 month that may influence brain-imaging outcomes
- Having received any other treatment in the past 1 month
- Are pregnant or lactating women
- Are suffering from mental disease, neurological disease, infectious disease, gastrointestinal disease, cardiovascular and cerebrovascular disease, immunologic disease, respiratory disease or kidney disease, any other chronic pain symptoms, or have a history of brain injury with loss of consciousness
- Are diagnosed as cancer, tuberculosis, rheumatism or rheumatoid arthritis, gout, joint trauma
- Have MRI contraindications such as claustrophobia, cardiac pacemaker, defibrillator, heart stent, intrauterine device
- Have skin lesion around the knee joint
Suspension criteria
- Are not suitable to continue to participate in the study due to adverse events or serious adverse events during the trial
- Are not suitable to continue to participate in the study because of serious deterioration of disease or some complications and special physiological changes
Dropout criteria
- withdrew from the study
- Loss of contact
Sample size
Sample size is required to compare two group mean values. The calculation method of bilateral equality of two samples is adopted[34] for two groups of A = Tuina group and B = Education group. Suppose H0: μA−μB=0, H1:μA−μB≠0. According to the previous relevant research and the pre experiment results[24, 35], we expect the VAS score difference between and after Tuina as μA = 3, and the VAS score difference between and after health care education intervention as μB = 0.7, if assuming that the standard deviation σ of the two groups is equal, that is, σ=2.1. If we set α=0.99 as the type I error, β=0.10 as the type II error, and 1-β=0.90 as the power, the final result is that we need at least 25 samples in each group.
Many MRI studies reported statistical significance with 12 to 21 patients per group[17, 36-38]. In order to have a powerful and repeatable statistical effect, we require 25 patients per group in this trial. Considering a 20% dropout rate and the possible excessive head motion during scanning, it will include 30 KOA participants in each group. In summary, we plan to enroll 60 participants and each group will undergo MRI scans to investigate different central mechanisms between Tuina and health care education treatments.
Randomization
Eligible patients will be randomized in a ratio of 1:1 to the Tuina group and the Education group with 30 patients in each group using simple randomization by the professional responsible for data statistics of the subject. The random number lists will be created will be generated by a random number generator (IBM SPSS Statistics version 21.0 software; IBM Corp., Armonk, NY, USA) and will be sent to the therapist with sequentially numbered, opaque and sealed envelopes by an independent assistant. The therapist will open random-allocation envelopes and allocate the participants accordingly to the Tuina group or the Education group.
Blinding
Participants and therapists will not be blinded to treatment allocation, due to the limitation of the Tuina treatment and the education intervention. The evaluators, data managers, and statisticians will be blinded to the group allocation in the outcome evaluation procedure and data analysis for the sake of reducing the risk of bias.
Interventions
The participants will receive a total of 24 treatments in 12 weeks. The doctors must have at least 10 years of experience with Tuina treatment and be skillful in health care education for the participants. In addition, they are required to have passed a clinical test to make sure that the consistency of the trial is administered.
Tuina group
The selected acupoints for the Tuina treatment include EX-LE02 (heding), EX-LE04 (neixiyan), EX-LE05 (xiyan), GB34 (yanglingquan), SP10 (xuehai), ST34 (liangqiu), ST32 (futu), BL40 (weizhong), BL57 (chengshan), GB31 (fengshi) and Ashi acupoints around the knee joint. An ashi point is a temporary acupoint where there is a sensation of acid, distension and pain, which can appear anywhere in the knee joints[39].
Referring to previous studies[29, 40-42], the Tuina procedures are as follows: First, the patients are in a supine position. The doctor stands on the side of the patient. The Pressing and Kneading Manipulation of Tuina is applied to the anterior acupoints around the affected knee joint for 10 to 15 min, with the thumb to achieve Deqi sensation, which is commonly regarded as an indicator of manipulation efficacy[43, 44]. Second, the patients are in a prone position. Then the physician will perform the same manipulation on the patient’s posterior acupoints around the affected knee joint for 10 to 15 min. It plays the role of relaxing sinews and activating collaterals with the theory of TCM. Every patient will receive 30 min manipulation each treatment, two treatments per week for 12 weeks.
Education group
The education intervention will be given to the participants with two sessions per week. The first health education session is a group meeting using a PowerPoint presentation, which will last 30min. An educational booklet and a video will be given to the patients to do the home-based online e-learning module. The later sessions will be conducted as one-to-one communication focused on personal needs to guide patients to implement the key education in their lives. The program includes two broad components: (1) introduce key concepts of pain biology and (2) present specification of recovery.
Each patient will be given an introduction about the underlying mechanisms, predisposing factors, and prognosis of KOA. Pain will be explained as the conscious part of the response, which can be influenced by many factors as a protective output. Interventionists will formulate specification of daily routines for patients, for example keeping the knees warm, keeping a healthy diet and prevent obesity, maintaining beneficial posture and abandoning bad habitual posture, avoiding overwork, etc.
Concomitant medications and other interventions
Patients with KOA will be instructed not to take any caffeine (tea and coffee, etc.) , any other medications, weight-loss techniques or physical activity, etc. to avoid confound interference for the analgesic effect of intervention of two groups. In cases of severe knee pain, ibuprofen (300 mg per capsule with sustained release) will be allowed as rescue medication and should be recorded on the knee-pain diary. After the end of the study, patients will be taught weight-loss techniques and physical activity. Patients are also asked to record the name, dose, date, frequency and the exact time of the medications used, and to report to the researcher if they take any medications during the study.
Safety evaluation
Adverse events refer to the unexpected responses that occur during or after treatment, which can lead to a hospitalization or even threaten life. The trial should be suspended and immediate process is indispensable whenever we encounter adverse events. We should record the details in the case report forms (CRFs). The efficacy and safety of the intervention will be evaluated after 6, 12 weeks’ intervention and 48 weeks’ follow-up subsequently. In addition, evaluators can be reformed with the clinical symptoms and AEs whenever the patients meet something important.
Outcome measurement
The clinical outcome will be measured by six self-report questionnaires, which can reflect the pain degree, pain sensation, pain emotion and quality of life.
The primary outcome measurements is VAS[45]. VAS is used to assess the current degree of pain. VAS is a 10-point scale selected to quantitatively measure the level of KOA pain during the study, for which 0 means None-of-Pain while 10 represent the unbearable pain.
The secondary outcomes are SF-MPQ[46], WOMAC[47] and SF-36[48]. SF-MPQ is used to assess pain perception experience and focuses on evaluating the sensory and affective components of knee pain. It mainly consists of 15 descriptors (11 sensory and 4 affective) rating on an intensity scale as 0 = none, 1 = mild, 2 = moderate, and 3 = severe. A higher total score indicates more severe pain. The WOMAC is a self-administered questionnaire consisting of 24 items divided into three subscales: pain (5 items), stiffness (2 items) and physical function (17 items). These measurements will be used to subsidiarily evaluate the symptom and quality of life improvement. Furthermore, to investigate the influence of emotional state on the brain activity, HAMD[49] and HAMA[50] will be used.
MRI data acquisition
MRI data will be acquired with a 3.0-T magnetic resonance scanner (SIEMENS MAGNETOM Verio syngo MR B17, Germany) with a 32-channel phase-array head coil at the Medical Imaging Department of Yueyang Hospital of Traditional Chinese and Western Medicine, Shanghai University of TCM. Participants are asked to close their eyes, wear earplugs, keep their heads still throughout the scan, keep relax and stay awake, without thinking of anything particular during the whole scan.
Prior to the blood-oxygen-level-independent (BOLD) resting-state functional images, a high-resolution T1 image for each subject will be acquired with spin echo (SE) sequence, transverse sagittal scan, flip angle (FA) = 9°, pulse repetition time (TR) = 1900ms, echo time (TE) = 2.93ms, field of view (FOV) = 256×256 mm2, slices 160 and slice thickness = 1mm.
Diffusion tensor imaging will be measured with the axial DTI mapping sequences, TR = 10000 ms, TE = 89 ms, matrix = 240 × 240, slice thickness = 2 mm, B-value1 = 0 s/mm2, B-value2 = 1000 s/mm2, 30 directions.
The BOLD resting-state functional images will be obtained with echo-planar imaging (EPI) sequence as: coronal axial scan, 33 slices, 4mm thickness, TE = 30ms, TR = 2000ms, FOV = 220 x 220 mm2, voxel size = 3.4 × 3.4 × 4.0 mm, FA=90°, scanning time 8 min 8 sec, 240 volumes in total.
The MRI outcome include: structure - gray matter density, cortical thickness, subcortical nuclei volumes; diffusion - FA and MD of white matter integrity; fMRI - functional connectivity (FC).
The 3D T1 structure data analysis will be performed by FSL tools (FMRIB Software Library)[51]. SIENAX (part of FSL 5.0)[52] will be used to obtain the volumes of neocortical gray matter (GM), total GM and white matter (WM) etc. The normalized volumes of subcortical regions such as hippocampus and thalamus will be estimated from FMRIB’s integrated registration and segmentation tool (FIRST) (part of FSL 5.0, FMRIB Software Library)[53]. The cortical thickness at each vertex will be obtained using FreeSurfer (http://surfer.nmr.mgh.harvard.edu). DTI data processing pipeline will follow diffusion toolbox in FSL to obtain FA and MD images, and voxewise tract based spatial statistics (TBSS) method will be applied to examine FA/MD changes. For the resting state fMRI data, preprocessing and functional connectivity analysis will be performed by SPM12 software (SPM12, Wellcome Department of Imaging Neuroscience, London, UK; http://www.fil.ion.ucl.ac.uk/spm/) with MATLAB 2013b (Mathworks, Inc., Natick, MA, USA).
Data collection and monitoring
The screeners will collect data on the baseline characteristics when the patients are recruited. Case report form (CRFs) includes observation and scanning time points, outcome measures, adverse events and safety evaluations. Doctors are blinded to evaluate various clinical pain indicators and fill in relevant information timely and accurately according to the requirements of CRFs. Only outcome assessors can access to the CRFs and perform double-data entry. Then two data administrators, who are beyond the research team and blinded to group allocation, will independently receive the completed CRF and enter them into an Excel database (Microsoft, Redmond, WA, USA). They are required to have completed rigorous training for the data monitoring. Then they will entry the real-time data in the Chinese Clinical Trial Registration Center, in which the electronic data management system will be used to track and monitor the test data in real-time in the Department of Science and Technology in Yueyang Hospital.
Statistical analyses
When processing and analyzing clinical data, the intention-to-treat principle will be followed with SPSS 21.0 statistical software (IBM, Armonk, New York, USA) by statisticians who are blinded to the group allocation.
The baseline characteristics will be expressed with descriptive statistics for the two groups, which are reported as the mean±standard deviation. A Kolmogorov-Smirnov test with Lilliefors correction will be used to analyze all quantitative variables to determine whether they follow a normal distribution. Parametric statistics (Tukey test) or nonparametric statistics (Wilcoxon rank-sum test) will be used for the within and between-group analyses in accordance with the results of the homogeneity and normality analyses. When initial homogeneity and normality of data distribution are found, repeated measures analysis of variance (ANOVA) will be implemented, or the Friedman test and Kruskal-Wallis test will be used when initial homogeneity but not normality of data distribution is found, or a linear mixed model will be adjusted for the baseline value if the initial homogeneity is not found. Adverse events in each group will be documented as percentage for safety assessments using the chi-square test or Fisher’s exact test. The statistical significance is defined as P < 0.05, and the 95% confidence interval will be reported.
Quality control
During the processing of the trial, quality control will be conducted with the management of the steering committee. To ensure the consistency of methods, professional trial method and regular monitoring technique should be trained before the researchers participate in the trial. The steering committee and ethics committee should be informed if the study protocol is modified or corrected.