Trial design
Because of the heterogeneous components of Chinese herbs, the onset time and half-time of Chinese herbs are hard to confirm(24). According to previous clinical observation and preliminary trial(11), we decided to set the intervention period or control period for 4 weeks and the wash-out interval for 1 week. Each cycle comprises a 4-week intervention period with MBYD combining conventional western medicine and a 4-week control period with placebo with the same western medicine in random order. The wash-out interval prior to switching each period makes carryover effects minimized. Hence, there are 29 weeks in all involving 3 cycles in this study (Figure1). The conventional western medicine that the patients take still follows his/her original therapeutic schedule. In view that patients with MG cannot stop taking conventional western medicine for a few days, which may aggravate their condition rapidly and lead to serious adverse events, only MBYD or placebo granules are washed out while western medicine are maintained during the wash-out interval. The protocol is conducted in light of the SPIRIT reporting guidelines(25) and obeys the Conventional Protocol Items: CONSORT extension for reporting N-of-1 trials (CENT) 2015 and CONSORT extension for reporting N-of-1 trials for traditional Chinese medicine (CENT for TCM) 2019. (24, 26)
Participants
The First Affiliated Hospital of Guangzhou University of Chinese Medicine (GZUCM) that we conduct the study is one of the centers for MG treatment in China and many MG patients coming to visit every year. That means adequate patients will be recruited at the inpatient and outpatient visits in the hospital.
The patients will be selected if they adapt to the inclusion criteria as follows:
- Patients aged between 18 to 65 and gender is not limited
- Patients diagnosed with MG based on ‘The Chinese guidelines for the diagnosis and treatment of myasthenia gravis (2020)’: patients with typical clinical features of MG (fluctuating myasthenia) excluded from other diseases and meeting any of the following three points, including pharmacological examination, electrophysiological characteristics or serum antibody detection (27)
- Patients attached to the spleen-stomach deficiency syndrome or spleen-kidney deficiency syndrome will be enrolled which is based on ‘The guidelines for clinical diagnosis and treatment of internal medicine of traditional Chinese Medicine: myasthenia gravis (2020)’ (28)
- Patients identified as class II or III according to myasthenia gravis foundation of America (MGFA) clinical classification (29) and are in the stable stage of MG at least 3 months
- Patients with Quantitative Myasthenia Gravis (QMG) score more than 6
- Patients treated with glucocorticoid administration should not take more than 15mg prednisone (or an equivalent dose of other glucocorticoids) per day.
The exclusion criteria are following:
- Female patients who are pregnant or lactating or have a pregnancy plan during the trial
- Patients with other autoimmune diseases (e.g., polymyositis, multiple sclerosis, rheumatoid arthritis) that may impact the assessment and treatment
- Patients with severe heart, kidney, liver, lung, hematological system, infectious diseases or cancer
- Patients with neuropsychiatric disorders that cannot cooperate
- Patients received plasma exchange, glucocorticoid or gamma-globin pulse therapy within 3 months
- Patients received thymectomy within half a year
- Patients with hypersensitivity of any drug in this trial
- Patients participating in other trials
- Patients consumed antibiotics, probiotics or anti-acids during the previous two months
The withdrawal criteria are following:
- Patients with serious adverse effect (SAE) or suspected unexpected serious adverse reaction (SUSAR)
- Patients with myasthenic crisis or worsening of symptoms requiring any treatment other than the trial medication
- Patients who decide to withdraw from the trial or lose follow-up
The finished cycles before the withdrawal will be analyzed as part of the trial. We will still follow up with withdrawn participants until their planned end in order to appraise any adverse effects of the trial medication.
Interventions
In consideration of the hardship in guaranteeing the quality of the decoction, we intend to use granules in this study. The conventional western medicine administration such as pyridostigmine, low-dose prednisone and immunosuppressants is in the light of their prestudy treatment schedule. In order to avoid the effects of other Chinese herbs, any drug containing ingredients from Chinese herbs is not allowed except MBYD. Moreover, patients are not allowed to take antibiotics, probiotics, or anti-acids, which can obviously affect gut microbiota.
The MBYD consists of Astragalus Membranaceus (Huangqi), Radix Fici Simplicissimae (Wuzhimaotao), Rhizoma Atractylodis Macrocephalae (Baizhu), Radix Codonopsitis (Dangshen), Poria Cocos (Fuling), Radix Bupleuri (Chaihu), Cimicifuga Foetida (Shengma), Pericarpium Citri Reticulatae (Chenpi), Angelica Sinensis (Danggui), Semen Coicis (Yiyiren), Rhizoma Dioscoreae (Shanyao), Radix Liquiritiae (Gancao), Radix Scrophulariae (Xuanshen). Patients with MG who have spleen qi-deficiency syndromes and kidney deficiency syndromes are usually diagnosed with spleen-kidney deficiency syndrome according to TCM theory and in that case, Morinda officinalis (Bajitian), Cistanche Salsa (Roucongrong), Semen Cuscutae (Tusizi), Cornus Officinalis (Shanzhuyu), Taxillus sutchuenensis Danser (Sangjisheng), Rosa laevigata Michx (Jinyingzi), which are together used to replenish kidney qi, will be added in MBYD. Both the MBYD granules and the placebo granules were manufactured by Guangdong Yifang Pharmaceutical Co. Ltd.(10g/bag, 2 bags three times a day, last number: J2004021). The placebo granules were identical to the MBYD granules in appearance, texture, color and odor.
Outcomes
Given the multi-component and multi-target characterization of TCM, an all-round assessment of its effective evaluation is significant. In this study, the primary outcome is the QMG score. The secondary outcomes are following:
- TCM syndrome score. Assessing TCM syndrome score is based on ‘Diagnosis standards for common syndromes in traditional Chinese medicine’ (30) issued by the Chinese Medicine Diagnostic Branch of China Association of Traditional Chinese Medicine. The measurement information of each symptom and sign of MG patient are weighted and summed according to the values of different symptoms and signs.
- The scores of the following scales: myasthenia gravis composite (MGC), myasthenia gravis activities of daily living profile (MG-ADL), myasthenia gravis quality of life (MG-QOL)
- The level of CD4+ Foxp3+ Treg cells and cytokines (IL-4, IL-17A, INF-γ, TGF-β) in the peripheral blood
- The alteration of gut microbiota. Before the trials begin and after each period, we will collect the patients’ fresh fecal samples and then immediately store them in the laboratory freezer at -80℃ until analyses. We will analyze the gut microbiota in the stool samples in order to explore the mechanism of MBYD in MG therapy at the microbiological level by the 16S rRNA gene sequencing.
- The reduction of side effects of western medicine for MG treatment
The safety assessments consist of the following items:
- Laboratory tests, including serum alanine aminotransferase, aspartate aminotransferase, urea nitrogen, creatinine and routine blood tests
- Vital signs, physical examination and electrocardiograph
- Adverse events occurring during the study, including MG crisis, drug allergy, hepatorenal insufficiency, and other adverse drug reactions
The outcomes and the safety assessments will be conducted before the trials begin and at the end of each period.
Sample size
We conducted the preliminary trial that 5 patients had finished the first cycle. The mean and standard deviation(SD) of QMG scores (the primary outcome) difference between the MBYD and placebo groups was (3.0±2.5). According to the sample size calculation of paired sample mean comparison, we used the two-sided test and set the type I error α of hypothesis test as 0.05 and the error β of class II as 0.1. The sample size was calculated as 10 by using PASS 15.0 software. Considering the 20% of lost follow-up rate, we set the sample size as 12.
Randomization, blinding and treatment allocation
We mark the intervention (MBYD) as A and the control (placebo) as B. Then, each cycle can be recorded as AB or BA. The clinical research associate (REA) uses SPSS 26.0 to set a random seed and generate a random number. Each generated odd number is regarded as AB and even number is regarded as BA. The first to third digits are allotted to the first participant, the fourth to sixth digits are allotted to the second participant, and so on. For example, if the first six digits of the generated number is 453728, then the first participant will receive 3 cycles as BA, AB and AB and the second participant will receive 3 cycles as AB, BA and BA.
The allocation result will be written and put into an opaque and sealed envelope. The granule storage and distribution will be in charge of the REA. Participants and investigators will both be blinded until the end of the trials.
In our experience, MBYD is usually mild and fewer side effects. In case of SAEs related to MBYD, the supervisor and primary researchers shall decide whether emergency blind breaking is necessary, in order to know what kind of medicine the patient is taking in time. When breaking the blind, researchers should write the date, the reason for opening the blind and sign on the envelope.
Statistical methods
All data collected will be recorded on the CRFs and then inputted into the Electronic Data Capture(EDC) system and proofread by two independent researchers. Each cycle containing the MBYD and the placebo period is considered as a pair in N-of-1 trials. We will combine the results of all complete cycles, analyse them, and then produce a posterior probability of the difference between the MBYD treatment and the placebo. If data are normally distributed, paired 𝑡 test will be used for a single case. And if the data do not obey normal distribution, Paired Wilcoxon signed rank tests will be conducted to analyze the data (31).The data will be analysed by using SPSS 26.0 software, and P < 0.05 is considered statistically significant.