A surgical therapy for Alzheimer's disease with lymphaticovenular anastomosis

Abstract

Background

Deep cervical lymphaticovenular anastomosis (dcLVA) surgery is able to control aging-associated Alzheimer's disease in patients. However, the efficacy rate remains unknown.

Objective

This study is designed to test the surgery efficacy in the treatment of mild-to-moderate AD patients.

Methods

This is a single-center retrospective study of dcLVA treatment of mild-to-moderate AD for 3 months. A total of 41 patients received the surgery, in which lymph vessels and lymph nodes in the district III of cervical area were identified using indocyanine fluorescence dye. The afferent lymphatics of the obstructed lymph nodes were connected to the jugular vein to fix the lymphatic blockage under a fluorescent microscope. The efficacy rate was examined at 3-month post-surgery by clinical scores and biomarkers.

Results

Lymph flow obstruction was observed on both sides of cervical area in the AD patients. The obstruction was successfully resolved through the surgery, and AD progression was attenuated or even reversed in the patients according to improvement in the scales of MMSE, ADL, NPI, CDR-SB, and CGI-EI. The average effectiveness rate was 50% by the CDR-SB score improvement. The efficacy was higher with shorter disease duration but not influenced by age and APOE4 genotype. Aβ_42/40 ratio and p-tau181 were improved in more than 67% patients. There were 2 cases of mild adverse reactions that were controlled immediately by regular treatments.

Conclusions

The data demonstrate that dcLVA surgery is an effective and safe therapy for AD in mild-to-moderate patients with 50% efficacy rate as measured by improvement of the CDR-SB score.

Introduction

Alzheimer's disease (AD) is a neurodegenerative disease, a major cause of dementia, accounting for 60% to 80% of all types of dementia. ¹ The clinical manifestations of AD include cognitive dysfunction, decline in daily living abilities, and mental behavioral abnormalities. The main pathological characteristics of AD include inflammatory plaques formed by extracellular amyloid-β (Aβ) deposits outside neurons and neurofibrillary tangles formed by abnormal hyperphosphorylation of tau protein within neurons. ² Existing clinical care for AD are mainly based on symptomatic treatments with limited strategies that reverse or stop progression of the condition. Lecanemab (trade name Leqembi) is a humanized monoclonal antibody approved in January 2023 by FDA of US for clearance of Aβ protofibrils and Aβ plaques. ³ This medicine can improve cognitive impairments in early-stage AD patients, but shows no effect on moderate and severe AD patients. Additionally, its high cost and side effects (brain edema, brain hemorrhage, brain atrophy, etc.) limit its clinical application.

Deep cervical lymphatic-venous anastomosis (dcLVA) is a new surgical therapy for AD patients,^4,5 which is based on the finding of central lymphatic system function in clearance of large molecular metabolic wastes (such as Aβ and tau proteins).^6,7 In the system, the meningeal lymphatic system was first reported in 2015,^8,9 which completely changed the traditional anatomical concept of “no lymphatic system in the brain”. ¹⁰ Dysfunction of the brain lymphatic system provides a plausible mechanism for the amyloid protein Aβ deposition, tau protein deposition, and inflammatory responses in the brain of AD patients.^6,7,11

The brain lymphatic system contains several network systems: the glymphatic system, ¹² meningeal lymphatic vessels,^8,9 nasopharyngeal lymph plexus, ¹³ and cervical lymphatic system. ¹⁴ The cervical lymphatic system has two different networks, subcutaneous and deep cervical networks. The deep cervical network has been reported for removal of brain metabolic products in several studies,^13,15,16 although the function of subcutaneous network has also been reported recently. ¹⁷ The metabolic wastes in the interstitial fluid of the brain tissue enter the glymphatic system first and then sequentially goes through the meningeal lymphatic vessels, nasopharyngeal lymph plexus, and deep cervical lymphatic network to move asway from the brain.^6,7 Several excellent studies have demonstrated functional degeneration of each component of the brain lymphatic system under aging, thereby accelerating the onset and progression of AD by accumulation of the metabolic wastes. ¹⁸ More importantly, the dysfunction precedes Aβ and tau deposition, ¹⁹ and becomes a common target in the exploration of surgical, pharmacological, and physical therapies of AD in mouse models. ²⁰ Here, we report the efficacy of dcLVA therapy in 41 mild and moderate AD patients at the Zhengzhou Central Hospital Affiliated to Zhengzhou University with 3-month follow-up.

Methods

Study design

This is retrospective single center study of efficacy rate of dcLVA in the treatment of AD, in which bilateral dcLVA was performed in the patients and the intervention efficacy was determined at three-month post-surgery. The study was conducted between August 23, 2024 and April 15, 2025 at the Zhengzhou Central Hospital Affiliated to Zhengzhou University, Zhengzhou, China. All participants were comprehensively assessed by specialists in the Cognitive Impairment Center to ensure precise diagnosis. A multidisciplinary team meticulously evaluated the risks of surgery and anesthesia to ensure the safety of the patients.

Evaluation approaches

Clinical dementia rating sum of boxes (CDR-sb)

Reflects cognitive status and daily living function, total score ranges from 0–18.0 points, 0.5–4.0 points indicate possible mild cognitive impairment, 4.5–9.0 points indicate mild dementia, 9.5–15.5 points indicate moderate dementia, 16.0–18.0 points indicate severe dementia.

Mini-Mental state examination (MMSE)

Includes orientation, memory, attention, recall ability, language skills in five dimensions, total score 30 points (0–30 points), higher scores indicate better patient cognitive function.

Activities of daily living scale (ADL)

Evaluates daily living abilities, total score ≤16 points indicate completely normal, >> 16 points indicate varying degrees of functional decline, maximum 56 points, higher scores indicate worse functionality.

Neuropsychiatric inventory (NPI)

Mainly evaluates psychiatric symptoms in dementia patients, scoring range 0–144, higher scores indicate more severe mental symptoms.

Clinician's interview-based impression of change plus caregiver input (CIBIC-plus) ²¹

Developed from the Clinical Global Impression of Change in 1991, assessing the cognition, behavior, and function of patients through face-to-face talks between the attending physician and the patient or family members. This scale uses an 8-point scoring method (0～7 points), based on the comparison of the patient's condition with the initial situation: 0: Not evaluated; 1: Significant improvement; 2: Improvement; 3: Slight improvement; 4: No change; 5: Slight deterioration; 6: Deterioration; 7: Severe deterioration.

Clinical global impression scale (CGI)^21,22

Originally designed by WHO and revised by the American NIMH in 1976, used to evaluate overall clinical efficacy, initially serving as the primary clinical efficacy indicator for various mental illnesses. Over time, it has gradually been applied to AD and other dementia studies. This scale includes three parts: Severity Index (SI), Overall Efficacy Assessment (GI), and Efficacy Index (EI), where EI is the efficacy index (0–4.00), EI >> 1 indicates effective treatment, and higher indices suggest better efficacy. Since the original form's descriptions for evaluating treatment effects are somewhat vague, we reviewed extensive domestic and international literature to provide specific quantified evaluation standards (see Table 1).

Table 1.

Evaluation criteria for therapeutic efficacy by the Clinical Global Impression-Efficacy Index (CGI-EI) scale (revised edition).

Treatment effect		Evaluation criteria
4 score	Significantly effective	The symptoms disappeared completely or almost.
3 score	Effective	Improvement in one of the following criteria: 1. MMSE increase ≥2 pointsS1−S3; 2. MMSE was stable, and the total score of NPI decreased by ≥30% or the key single score were significantly decreased (≥4)S3,S4; 3. ADL decreased ≥5S3,S5,S6; 4. CDR-SB decreased ≥1.5 pointsS7.
2 score	Slightly effective	Improvement as assessed by the scales (MMSE, ADL, NPI, and CDR-SB), but the assessor and caregiver rated the patient as having improved (meeting the criteria for slight improvement on the CIBIC-plus scale)S8,S9.
1 score	No change	1. MMSE was stable (change ±1 point)S3,S10; 2. No improvement according to the assessment physician and caregiver (no change in CIBIC-plus)S11.
	Aggravation	Having one of the following criteria: 1. MMSE decrease ≥2 pointsS12; 2. Assessment of patient deterioration as perceived by physicians and caregivers (meeting CIBIC-plus criteria for deterioration)S13.

The CGI-EI scale was modified in current study to precisely evaluate the improvement of AD patients. The therapeutic effect of dcLVA was quantified in the patients by scores with redefined evaluation criteria. The citation references in Table 1 are in the supplemental material.

Inclusion criteria

(1) Meet the diagnostic criteria for AD (2018 NIA-AA diagnostic criteria), ²³ mild-to-moderate patients CDR-SB < 16; (2) Have complete neuropsychological assessments, physical examinations, cranial magnetic resonance imaging (MRI) medical records; (3) Research subjects have stable and reliable caregivers, or at least frequent contact with caregivers (at least 4 days per week, at least 2 h per day); (4) Individuals or their families voluntarily participate in this study, provide biological samples, and sign informed consent forms.

Exclusion criteria

(1) Existence of severe mental illness (diagnosed or requiring medication-controlled schizophrenia, bipolar affective disorder, mental retardation, etc.); (2) Existence of other causes leading to dementia, including vascular dementia, central nervous system infections (such as viral encephalitis, tuberculous meningitis, syphilis, etc.), concurrent with other neurodegenerative diseases (such as Parkinson's disease, multiple system atrophy, Lewy body dementia, frontotemporal dementia, Creutzfeldt-Jakob disease, Huntington's disease, etc.), traumatic dementia, other physical and chemical factors (such as drug poisoning, alcoholism, carbon monoxide poisoning), endocrine imbalance (such as thyroid disease, parathyroid disease), and other factors (vitamin deficiency, epilepsy, etc.) leading to dementia; (3) Patients already diagnosed with malignant tumors, organ dysfunction. Elimination Criteria: drop off the follow-up. This study complies with the Helsinki Declaration and corresponding laws and regulations and was approved by the Ethics Committee of Zhengzhou Central Hospital (Ethics Number: ZXYY202472). Before the surgery, all patients or their families signed the informed consent form for this clinical research projects, agreeing to truthfully provide information and allowing clinical data to be used in scientific publications; (4) Severe AD patients diagnosed with the parameters below.

Clinical data collection

The collected information includes: (1) General information, medical history, physical examination, treatment course, etc.; (2) Preoperative baseline (within 1 week before surgery) and postoperative 3-month ± 7-day neuropsychological assessment results (MMSE, Montreal Cognitive Assessment, NPI, ADL, CDR-SB, Digital Span Test, Geriatric Depression Scale, Hachinski Ischemic Score, Auditory Verbal Learning Test, Pittsburgh Sleep Quality Index, Hamilton Anxiety Rating Scale, Hamilton Depression Rating Scale, Trail Making Test), which were measured by the same trained cognitive impairment specialist physician; (3) Postoperative 3-month CGI-EI evaluation; (4) Preoperative baseline (within 1 week before surgery) and postoperative 3-month ± 7-day peripheral blood or cerebrospinal fluid (CSF) biomarkers (Aβ_1–42, Aβ₁₋₄₀, Aβ_42/40, p-tau181, α-synuclein); (5) Apolipoprotein E (APOE) genotype; (6) Preoperative imaging (cranial MRI, magnetic resonance angiography, susceptibility-weighted imaging, hippocampal thin-layer MRI, ¹⁸F-fluorodeoxyglucose (FDG)-positron emission tomography (PET), some completed Aβ-PET); (7) Postoperative adverse reactions.

Surgical procedure

• 1. Preoperative preparation. The geriatric department selected the candidate patients by evaluating medical history, symptoms, and family history. The cognitive function was tested using several standard score systems including MMSE, CDR, and the Montreal Cognitive Assessment. Imaging examinations included cranial MRI, PET-computerized tomography (CT), hippocampus thin-layer scan, chest CT, cardiac ultrasound. CSF was collected by lumbar puncture and the biomarkers including Aβ protein, tau protein, and p-tau181 were determined in the CSF as well as plasma.

• 2. Surgical risk assessment. Coagulation function was evaluated to avoid intraoperative and postoperative bleeding risks. Heart and lung functions were determined to ensure the patient tolerance to general anesthesia and allergy to medicines was recorded. The surgical treatment plan was acknowledged, and the consent forms were signed by the patients or the family members.

• 3. Surgical operation. The patient laid supine, after satisfactory general anesthesia, 1 ml indocyanine green injection is given subcutaneously at both sides below the mastoid behind the ears and at the angle of the mandible, fluorescent detection is used for surface marking, iodophor disinfection of the neck surgical area, routine placement of sterile towels and drapes. During surgery, the head is slightly tilted back and to the left, a longitudinal incision about 5.0 cm long was made along the posterior edge of the sternocleidomastoid muscle on the right side, skin and subcutaneous tissues were cut successively, hemostasis is achieved by cauterizing bleeding points with electrodes, fascia tissue is bluntly dissected, the sternocleidomastoid muscle is identified, and the sternocleidomastoid muscle is retracted towards the inner side of the neck, taking care to protect the external jugular vein and cutaneous nerves. In microscopic view, the zone III lymph nodes were located in the deep cervical, and the patency of lymphatic vessels were observed with the fluorescence in afferent lymphatic vessels under the fluorescence microscope (KINEVO 900, Carl Zeiss Meditec AG, Germany). The blocked lymph nodes were identified by fluorescence accumulation from the afferent lymphatic vessels. The matching veins were identified nearby the afferent lymphatic vessels. The proximal ends of afferent lymphatic vessels and distal end of the vein were anastomosed using 10-0 non-absorbable microsurgical sutures. After completion of the anastomosis, good patency was confirmed by fluorescence appearance in the distal vein vessel. The same procedure was conducted to complete anastomosis on the other side. Fascia tissue was sutured with 5-0 absorbable sutures, skin was sutured intermittently with collagen protein sutures, and the wound was dressed with sterile bandages, completing the surgery.

Styles of lymphatic-venous anastomosis: Small veins have lower pressure compared to major veins, and finer veins are preferred in the anastomose to match the pressure of lymphatic vessels. There are currently no unified standards for lymphatic-venous anastomosis methods regarding the position, number, and type of anastomosis. Simultaneous anastomosis of multiple lymphatic vessels with the same vein cavity (Octopus style) and 1:1 end-to-end anastomosis (Paired style) are all used. Our group primarily adopts the “Octopus style” for its minimal invasion nature, which connects multiple lymphatic vessels to a single vein simultaneously, resembling the tentacles of an octopus. Its clinical significance lies in: (1) improving lymphatic drainage efficiency by significantly increasing lymphatic fluid outflow through multipoint connections; (2) maximizing the use of venous resources by handling multiple lymphatic vessels through a single vein when venous resources are limited, optimizing resource allocation; (3) reducing tissue damage by decreasing the number of repeated operations and time compared to traditional individual anastomoses, lowering the risk of tissue damage. However, this surgical style requires a high level of skills in the surgeon's microsurgical operation. Quality of the anastomosis is examined to avoid venous obstruction or leakage at the anastomosis site.

• 4. Anesthesia. The surgery was completed under general anesthesia with tracheal intubation. Anesthesia induction was performed with combination of 3 drugs, such as sufentanil (0.2–0.5 μg/kg), etomidate (0.15–0.2 mg/kg), and rocuronium bromide (0.6 mg/kg). Elderly patients had less dosages of induction medicines as appropriate. Maintenance of anesthesia was administrated by combination of sevoflurane inhalation (1%-2%), and vein injection of propofol (2–4 mg/kg/h). Intraoperative monitoring includes invasive blood pressure, electrocardiogram, SPO₂, PETCO₂/T and other vital signs, maintaining blood pressure fluctuation within a range of 5–10%. Depth of anesthesia was adjusted accordingly.

Statistical analysis

Data of 41 patients were statistically analyzed using SPSS 27.0 software, and graphs were drawn using Graphpad Prism (version 8.0). Data of counts are expressed as percentage of change in each group using chi-square tests; baseline and 3-month data were compared for difference using two-tailed Student's t-tests and Welch's Correction test for adjustment; differences among three groups were analyzed using One-way ANOVA and Kruskal-Wallis correction tests, and LSD was used for pairwise comparisons. p < 0.05 indicates statistical significance.

Results

General information

A total of 41 AD patients were included in this study with 18 males, 23 females, aged 45–84 years at average of 63.98 ± 8.38 years, education duration 10.02 ± 4.31 years, disease duration 3.73 ± 1.95 years, 17 (41.46%) had hypertension, 7 (17.07%) had diabetes, 10 (24.39%) had hyperlipidemia, 8 (19.51%) had coronary heart disease, perioperative blood pressure and blood sugar were stably controlled, and symptoms of coronary heart disease patients were stable. APOE gene testing was completed in 39 cases, with 14 carrying the APOE E4 gene, including 2 homozygous (E4/E4 type), 11 heterozygous of E3/E4 type, and 1 heterozygous of E2/E4 type. According to the CDR-SB score results, AD severity was graded, with 7 mild cases and 34 moderate cases. The AD patients had a higher rate of comorbid hypertension and coronary heart disease, with a higher incidence of APOE E3/E3 genotype (Supplemental Table 1).

dcLVA surgery

The surgery was performed with an incision (Figure 1A) at one side of cervical area to expose the deep lymphatic vessels and lymph nodes. The lymphatic vessels and lymph nodes were identified with indocyanine fluorescence dye under the microscope (Figure 1B). The obstructed lymph node was identified by fluorescence in the afferent lymphatic vessels, in combination with accumulation of fluorescent dye in the lymph node and no appearance of fluorescence in the efferent lymphatic vessels. The afferent lymphatic vessels were anastomosed to the nearby vein vessel as described in the method section (Figure 1C). The success of anastomosis was verified by flow of fluorescence-containing lymph fluid into the vein vessel after anastomosis to bypass the blocked lymph node (Figure 1D). Anatomy basis of the surgical operation is indicated by a diagram (Figure 1E).

Figure 1.

dcLVA surgery. (A) Incision of dcLVA surgery. The length and position of incision are shown in the cervical area of an Alzheimer's disease patient. (B) Obstructed lymph node. Indocyanine fluorescence was used to identify obstructed lymph vessel/node. (C) Anastomosis of lymphatic and vein vessels. Multiple afferent lymphatic vessels of the obstructed lymph node were anastomosed simultaneously to single vein vessel. (D) Verification of successful anastomosis. The success was indicated by flow of fluorescence-containing lymph fluid into the vein vessel after anastomosis. (E) Diagram for anatomy basis of the surgical operation.

Efficacy of dcLVA therapy

To assess the efficacy of dcLVA surgery, the neurological psychological evaluation was conducted at the baseline and 3-months post-operation. Eighteen patients completed the evaluation offline in the hospital and 23 patients completed the evaluation at home by phone. In the basis of offline data, 83.33% patients were improved by the upward scores of MMSE (15/18 patients, p < 0.001), 83.33% patients were improved as indicated by the downward scores of NPI (15/18 patients, p < 0.01), 66.67% patients were improved as indicated by the downward ADL scores (12/18 patients, p >> 0.05), and 50% patients were improved as shown by the upward scores of CDR-SB (9/18 patients, p >> 0.05) (Figure 2A-E and Supplemental Table 2). As a gold standard for AD treatment goal, CDR-SB score usually became worse along the progression course of AD. In current study, this downward trend was blocked even reversed in 9 out of 18 patients in the 3-month study (Figure 2D). The 9 responders by the improved CDR-SB score suggests a 50% efficacy of the surgery in the AD patients. The responders and non-responders were compared in the basal conditions including biomedical indices, metabolic parameters and inflammatory markers (Supplemental Table 3). The responders had a shorter disease duration over the non-responders (2.22 ± 0.83 versus 4.33 ± 1.94 years, p = 0.008). The data suggest that dcLVA surgery improved the dementia status in AD patients by enhancing the cognitive function and neuropsychiatric capacity.

Figure 2.

Improvement of clinical scale scores by dcLVA in Alzheimer's disease patients. (A) The MMSE score at 3-month was higher than that of the baseline (p < 0.001); (B) The NPI score at 3-month was significantly lower than that of the baseline (p < 0.001); (C) The ADL score at 3-month was not significantly changed over the baseline (p > 0.05); (D) The CDR-SB score at 3-month was not significantly changed than that of baseline (p > 0.05). (E) Efficacy rate in the basis of percentage of improved patients in each type evaluation. **p < 0.01, ***p < 0.001 by paired student's t-test (n = 18).

Efficacy rate

In terms of CDR-SB improvement, the efficacy rate of dcLVA surgery was 50% in the AD patients. However, in reference to CGI-EI improvement, the efficacy rate was even higher at 75.61%. Based on the evaluation criteria of CGI-EI scale, we defined the impact of dcLVA surgery at four levels (Table 1). The results from the defined evaluation were used to calculate the efficacy rate. Overall efficacy rate was 75.61% in the AD patients, with 60.98% being effective and 14.63% slightly effective in the total 41 patients (Table 2). To evaluate the efficacy more precisely, we quantified the efficacy using the efficacy index in CGI-EI, which included surgery-related adverse reactions in the efficacy analysis. The results showed that the effectiveness rate of dcLVA surgery was the same at 75.61% in the AD patients (Supplemental Table 4). This indicates that surgery-related adverse reactions had a minimal impact on overall efficacy, suggesting high safety of the surgery.

Table 2.

Efficacy rate of dcLVA surgery in AD patients by the CGI-EI score.

Treatment effect	Mild-to-moderate (N = 41)	Effective rate [n/N (%)]
Effective n (%)	25 (60.98)	31/41 (75.61)
Slightly effective n (%)	6 (14.63)
No change/aggravation n (%)	10 (24.39)

The efficacy rate was calculated by the improved score of CGI (Clinical Global Impression) scale. The effective, slightly effective and non-effective (No change/aggravation) data were obtained by comparing the scores at the baseline and 3-month. The chi-square test was used for analysis of significant difference between the two time points. The efficacy rate was calculated in 3 categories, total patients (n = 41). Overall effective rate was 75.61% (effective + slightly effective) in all 41 patients. p < 0.05 was considered significant difference.

Factors influencing the surgery efficacy

Several factors including disease duration, age, and APOE4 genotypes were examined for their impact in the dcLVA surgery efficacy. The efficacy was not significantly influenced by disease duration below or above 5 years (Figure 3A). The efficacy rate was not influenced by patient's ages in all age groups (Figure 3B). About impact of genotypes in the efficacy rate, APOE4 genotypes (heterozygous or homozygous) exhibited no significant impact in the surgery efficacy (Figure 3C). The data suggest that the surgery efficacy was not influenced by the factors above. However, within 5 years of disease duration, a shorter disease duration was associated with a higher efficacy of the surgery treatment (Supplemental Tables 5–7).

Figure 3.

Impact of multiple factors in the efficacy rate of dcLVA. (A) Disease duration impact in efficacy rate (<5 years, n = 27; >> 5 years, n = 14); (B) Patient age impact in the efficacy rate (<65 years, n = 23; 65∼75 years, n = 14; >> 75 years, n = 4); (C) Impact of APOE4 genotypes in the efficacy rate (APOE4 positive, n = 14; APOE4 negative, n = 25). Two-way ANOVA test was used for statistical analysis of impact of different factors except genotype in the dcLVA efficacy. Student's t-test was used in the analysis of genotype effect. *p < 0.05.

dcLVA improved ad biomarkers in plasma and CSF

The biomarkers of AD, such as Aβ_42/40 ratio and p-tau181 level, were examined in the plasma and CSF. The data were compared at the baseline and end of 3-month post-surgery for each patient (Figure 4A-D). The Aβ_42/40 ratio was significantly improved by the surgery with an elevation in both plasma and CSF, respectively (Figure 4A, B). The p-tau181 value was significantly improved by the surgery with a decrease in the plasma and CSF, respectively (Figure 4C, D). The data suggest that the AD biomarkers, Aβ_42/40 ratio and p-tau181, were significantly improved in the patients by dcLVA surgery in both plasma and CSF.

Figure 4.

Aβ_42/40 ratio and p-tau181 in the plasma and CSF. (A) Plasma Aβ_42/40 ratio in each patient. (B) CSF Aβ_42/40 ratio in each patient. (C) Plasma p-tau181 in each patient. (D) CSF p-tau181 in each patient. *p < 0.05, ** p < 0.01, *** p < 0.001 by paired student's t-test (Plasma, n = 12; CSF, n = 5).

The efficacy rate of dcLVA therapy was examined with improvement of the biomarkers by percentage of patients with the biomarker improvements. By plasma Aβ_42/40 ratio, the efficacy rate was 92% in the AD patients (n = 13); By CSF Aβ_42/40 ratio, efficacy was 83% (n = 6) (Figure 5A, B). By the plasma p-tau181 reduction, the efficacy rate was 85% (n = 13), and by CSF p-tau181 reduction, the efficacy rate was 67% (n = 6), respectively (Figure 5C, D). These data suggest that the efficacy rate of dcLVA was between 67%-92% in AD patients by the favorite changes in biomarkers.

Figure 5.

Efficacy rate of dcLVA surgery in AD patients by improvement of biomarkers. The efficacy rate was calculated by the percentage of patients with improved biomarkers of Aβ_42/40 ratio or p-tau181. The analysis was conducted in both plasma and CSF. (A) Efficacy rate by patient percentage with improved value of plasma Aβ_42/40 ratio. (B) Efficacy rate by patient percentage with improved CSF Aβ_42/40 ratio. (C) Efficacy rate by patient percentage with improved plasma p-tau181. (D) Efficacy rate by patient percentage with improved CSF p-tau181. In plasma, n = 13; in CSF, n = 6.

Safety of dcLVA

The average operation time was 2.62 ± 0.56 h (1.67–4.92 h), with little blood loss (all less than 10 ml). Within the 3-month period after the surgery, there were 2 cases of adverse reactions: 1 patient developed delirium within 24 h after the surgery, and recovered within 24 h. 1 patient developed urinary retention after the surgery, and recovered after treatment with medication. All these adverse reactions were mild, and under control without any serious outcome.

Discussion

Our results suggest that dcLVA surgery is an effective and safe therapy of AD patients by improving the clinical symptoms and biomarkers in the mild and moderate AD patients. LVA surgery is traditionally used to treat patients with peripheral lymphedema, and its application to AD therapy ⁴ was based on the role of meningeal lymphatics in the pathogenesis of AD found in the mouse models. ¹⁶ Currently, there are two reports on dcLVA therapy of AD patients, one case report by Xie's group with 6 month follow-up, ⁴ and one small study of 6 patients by the Ninth People's Hospital of Shanghai Jiao Tong University with 5 week follow-up. ⁵ Both studies indicate that the surgery is effective in the treatment of mild AD patients, but the efficacy rate was not available in those studies due to the limitation of sample size or short term. This issue has been addressed in current study of 41 cases for 3 months with an efficacy rate around 50% in terms of CDR-SB score improvement in the AD patients. The efficacy rate is supported by the improved levels of biomarkers in the plasma and CSF.

In the AD patients, we observed that deep cervical lymphatic vessels and lymph nodes became blocked leading to lymph fluid retention in current study. The retention was resolved by dcLVA as indicated by appearance of the fluorescent dye in the vein blood upon completion of anastomosis. In the 3-months post-surgery, were no significant side effects of the surgery.

The study provides a reference for selection of dcLVA candidates. The inclusion and exclusion criteria were set up for selection of the AD patients in this clinical trial. The efficacy rate may be further improved by identification of precise location of lymphatic dysfunction in the whole central lymphatic system. The central lymphatic system includes four parts, with the deep cervical lymphatic system being the bottom portion. Dysfunctions of any upstream portions may reduce the efficacy rate of dcLVA surgery. Clinical imaging technologies have been explored to assess the functions of each section of the brain lymphatic system,^24,25 the standard protocols remain to be established for clinical application of the technologies to dcLVA surgery. In this condition, there is no international standard yet for the application of dcLVA surgery to AD patients.

Our results suggest that the therapy efficacy is unrelated to anesthesia. In general, anesthesia impairs cognitive function of elderly in individual-specific manner and AD patients have a high risk of cognitive impairment by anesthesia medicines as reviewed. ²⁶ In terms of brain lymphatic system, anesthesia has been reported to improve ²⁷ as well as inhibit ²⁸ the glymphatic function in the mouse models. Sleep is reported to improve clearance of the brain wastes as demonstrated in the study of sleep impact in the brain metabolites.^29,30 However, the sleep effect from one time anesthesia may not last for 3 months. In current study, anesthesia was performed by combination of two medicines, sevoflurane and propofol. Both medicines have mild activities in the impairment of cognitive function of patients. ²⁶ These facts do not support a role of anesthesia in the therapeutic effects of dcLVA in AD patients in current study.

Current study suggests that dcLVA effects was related to reduction of Aβ and tau protein accumulation. A recent report suggests that the deep cervical lymph nodes exhibited Aβ protein accumulation in AD patients, ³¹ supporting that the deep cervical lymphatic system represents a drainage channel for brain Aβ proteins. In the current study, dcLVA surgery improved the biomarkers, such as an increase in Aβ_42/40 ratio and reduction in p-tau181 levels in the plasma and CSF. The dcLVA surgery is similar to the classical LVA in the treatment of peripheral lymphedema by increasing lymph fluid drainage. Therefore, we consider AD as a type of brain lymphedema.

In summary, current study suggests that dcLVA is an effective and safe therapy to AD patients in the control of disease progression with an efficacy rate around 50% by CDR-SB score improvement. The efficacy of dcLVA surgery is reduced by disease duration above 2.5 years in AD patients. However, this study is limited by several factors. As a single center and retrospective study, the observations are limited to comparison of the conditions at the baseline and 3-month post-surgery. The patient continued taking general medicines in control of blood pressure and other conditions after the surgery for the best treatment effects. Effect of the general medicines could not be excluded as there was no non-surgical control group. The conclusion is also limited by lack of Aβ quantification in the brain through PET-CT, which was not done due to unavailability of the test in our hospital during the study. P-tau217 was not examined due to lack of reagents during the study. If the sample size is larger and the study duration is longer (>>3 months), the neurological psychological evaluation data would be more convincing. In general, this study provides a foundation for a multicenter randomized controlled trial in the future.