2.1 Study design
This was a prospective cohort study. The patients were enrolled from the Department of Neurology or Geriatrics of 5 hospitals in Xi'an, China. The experimental protocol was reviewed by the Ethics Committee of the First Affiliated Hospital of Xi'an Jiaotong University (approval number: XJTU1AF2015LSL-066). All subjects signed an informed consent form. The study was registered in Clinical Trials (registration number: NCT02711683) and was funded by the clinical research project of the First Affiliated Hospital of Xi'an Jiaotong University.
2.2 Participants And Eligibility Criteria
Each patient was examined by a doctor in the clinic and was given a diagnosis of AD according to their clinical history, laboratory examination and brain MRI.
All patient met the following inclusion criteria: 1) age of 50–85 years old (including those who were 50 and 85 years old), either sex; 2) met the diagnostic criteria for suspected AD by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer‘s Disease and Related Disorders Association (NINCDS-ADRDA) (1984)[11]; 3) mild to moderate AD patients, that is, patients with 11 points ≤ MMSE total score ≤ 26 points (or patients with an elementary school education level: 11 points ≤ MMSE total score ≤ 22 points); [12, 13] 4) total Hachinski ischemic scale (HIS) [14] score ≤ 4 points; 5) memory loss for at least 12 months, with a tendency of progressive deterioration; 6) brain magnetic resonance imaging (MRI) scan suggest a significant possibility of AD (medial temporal lobe atrophy (MTA) visual rating scale[15] grade 2 or higher, Fazekas scale[16] ≤ 2); 7) no obvious physical signs during nervous system examination; 8) no prior history of treatment with donepezil or with treatment history of donepezil and a relatively stable disease status; 9) stable and reliable caregivers, with the ability to contact the caregivers frequently (at least 4 days a week, and at least 2 hours a day); the caregivers helped the patients participate throughout trial; 10) elementary school or higher education level and the ability to complete the cognitive ability measurement and other tests specified in the protocol; and 11) signed an informed consent form.
Exclusion criteria were 1) previous nervous system diseases (including stroke, optic neuromyelitis, Parkinson's disease, epilepsy, etc.); 2) mental illness, according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR) criteria[17], including schizophrenia and other mental illness, bipolar disorder, and severe depression or paralysis; 3) unstable or severe heart, lung, liver, kidney, or hematopoietic diseases; 4) uncorrectable visual and auditory disorders that affected completing neuropsychological tests and scale assessments; and 5) simultaneous use of other cholinesterase inhibitors or Memantine.
2.3 Interventions
According to the patient's condition, family income, compliance and other factors, the doctor consults with the patients’ caregivers to formulate the treatment plan. Patients with mild to moderate AD were divided into the donepezil alone group and donepezil combined NBP group according to different treatment regimens: donepezil alone group - donepezil 5 mg, once daily; donepezil combined NBP group: donepezil 5 mg daily plus NBP 200 mg, three times a day for 48weeks.
2.4 Trial End Point And Evaluation Indexes
The primary end point index was the Alzheimer's disease assessment scale-cognitive subscale (ADAS-Cog/12) score [18]. Secondary end point indexes were the clinician's interview-based impression of change plus caregiver input (CIBIC-Plus) scale [19], Alzheimer's disease cooperative study ADL scale (ADCS-ADL) [20], and neuropsychiatric inventory (NPI) scale [21]. The ADAS-Cog can assess 6 cognitive fields, including memory, language, orientation, logic, social cognition, and attention. The lower the score is, the lighter the cognitive damage. The CIBIC-Plus reflects the improvement of an individual’s clinical symptoms by interviewing the individual and his/her caregiver. The score ranges from 1 to 7 points, where 1 represents maximum improvement, 4 represents no change, and 7 represents maximum deterioration. The clinician's impression of the severity of the disease based on the interview with the subject at baseline served as the reference for CIBIC-Plus scoring in follow-ups. Clinicians evaluating the CIBIC-Plus were not aware of the scores of other scales. The ADCS-ADL is a clinician assessment standardized questionnaire composed of 23 items that assesses the actual performance of specific actions and behaviors of the individual observed by the caregiver in the past 4 weeks. The score ranges from 0 to 78 points, and the lower the score is, the more severe the disorder. The NPI includes 10 items regarding behavior and 2 items regarding the autonomic nervous system. The total score ranges from 0 to 144, and the higher the score is, the more severe the damage.
Safety assessments included physical examinations, vital signs, and adverse event (AE) reports. Each patient attended 5 visits, including baseline, 12 weeks, 24 weeks, 36 weeks, and 48 weeks. At each follow-up, the above scales were evaluated, and AEs were recorded.
To make ensure the reliable of all assessment, a blind method was used. Neuropsychological assessors didn’t know what treatment each patient received. Each center at least had two neuropsychological assessors, one responsible for the assessment of the MMSE and ADAS-cog, another responsible for the evaluation of ADL, NPI, CIBIC-plus. The scale assessors and doctors received uniform training in scale testing and disease diagnosis. The reliability of the cognitive tests and diagnoses between assessors was greater than 0.90. All trainees passed a scale consistency test before participating in the trial.
2.5 Statistical Analysis
The difference between the scale scores at 48 weeks (ADAS-cog, NPI, ADCS-ADL, and CIBIC-plus) and the scale scores at baseline was used as an indicator of changes in cognitive function, mental behavior, daily living ability, and overall efficacy. All data were used to build a database in EpiData and were analyzed using SPSS 18.0 statistical software. Variables related to the study participants that conformed to an approximate normal distribution are expressed as the mean ± standard deviation (x ± s), variables that had a severely skewed distribution are expressed in quartiles, and categorical variables are expressed as value (%). The x2 test, t test or rank sum test was used for comparisons between groups according to the different types of data, and multivariate logistic regression was used for the multivariate analysis. P < 0.05 (double sided) represented statistical significance for all tests.