Serum anti-enteric neuronal antibodies in patients with achalasia and their association with clinical profiles

Abstract

BACKGROUND

Immune-mediated mechanisms are implicated in the pathogenesis of achalasia of cardia (AC), which is characterized by inflammation-driven degeneration of myenteric neurons. Circulating autoantibodies targeting the enteric nervous system may serve as potential biomarkers reflecting this underlying neurodegeneration. However, the clinical significance of anti-enteric neuronal antibodies (AENAs) in disease progression and treatment outcomes remains unexplored.

AIM

To evaluate the prevalence of AENA in AC and its correlation with clinical characteristics and treatment outcomes.

METHODS

This prospective cohort study included 97 patients with AC diagnosed according to the Chicago Classification version 4.0 and 98 age- and sex-matched healthy subjects (HS). Indirect immunofluorescence was used to detect AENA and antinuclear antibodies. Multivariable linear regression analyses were performed to explore associations between AENA intensity and clinical parameters in patients with AC.

RESULTS

The AENA-positive rate was significantly higher in patients with AC than HS (60.8% vs 25.5%, P < 0.001). AENA-positive patients had higher endoscopic CARS (contents, anatomy, resistance, and stasis) than AENA-negative patients (5.0 vs 4.0, P = 0.045), whereas Eckardt scores and manometric parameters did not differ between groups. Multivariable regression identified strong AENA positivity as independently associated with higher CARS (β = 0.95, 95% confidence interval: 0.37-1.53, P = 0.002), demonstrating a dose-response relationship (P for trend = 0.005). Patients with strong AENA intensity also exhibited greater barium column width and higher integrated relaxation pressure than those with moderate intensity. After peroral endoscopic myotomy and overlap weighting, AENA-positive patients had worse Eckardt scores at 1- and 6-month follow-up than AENA-negative patients (P = 0.031 and P = 0.005, respectively).

CONCLUSION

Serum AENA is prevalent in AC and correlates with disease severity in a dose-dependent manner and with early treatment response, potentially identifying patients with more severe disease characteristics and poorer outcomes.

Key Words: Achalasia; Anti-enteric neuronal antibody; Enteric nervous system; Disease severity; Treatment outcome

Core Tip: Serum anti-enteric neuronal antibodies (AENA) are significantly more prevalent in patients with achalasia than in healthy subjects. This study identifies a novel dose-dependent association between AENA intensity and endoscopic disease severity, as assessed by the CARS (contents, anatomy, resistance, and stasis) score. Strong AENA positivity correlates with greater esophageal dilation and higher integrated relaxation pressure. Furthermore, AENA positivity is associated with less favorable symptomatic improvement after peroral endoscopic myotomy. These findings suggest that AENA status may serve as a valuable biomarker for identifying a more severe disease phenotype with poorer early treatment outcomes.

INTRODUCTION

Achalasia of cardia (AC) is the most common primary esophageal motility disorder and is characterized by impaired relaxation of the lower esophageal sphincter (LES) and absent or ineffective peristalsis of the esophageal body[1]. Patients with AC typically present with dysphagia to both solids and liquids, regurgitation of undigested food and saliva, chest pain, and weight loss[2]. Although the pathogenesis of AC is not fully understood, it is primarily attributed to degeneration of inhibitory nitrergic neurons within the myenteric plexus[3]. Current therapeutic approaches, including pneumatic dilation, peroral endoscopic myotomy (POEM), and surgical myotomy, alleviate symptoms but can not address the underlying pathophysiology[4]. Furthermore, these interventions carry the potential to disrupt the anti-reflux barrier, which may worsen symptoms and endoscopic manifestations of gastroesophageal reflux disease[5].

From a pathological perspective, AC is a neurodegenerative condition affecting the enteric nervous system. One prevailing hypothesis suggests that immune-mediated ganglionitis, potentially triggered by viral infections in genetically susceptible individuals, leads to progressive loss of myenteric neurons[3,6,7]. This notion is supported by the increased prevalence of autoimmune comorbidities among patients with AC, including thyroid disorders[8] and connective tissue diseases[9]. Moreover, shared genetic backgrounds and pathophysiological mechanisms across various autoimmune conditions may contribute to the higher incidence of comorbid immune-mediated diseases in this population[10]. Subsequent studies have identified upregulation of genes associated with inflammatory activation in patients with AC, including increased tumor necrosis factor (TNF)-related apoptosis-inducing ligand in monocytes (odds ratio = 2.70)[11]. Furthermore, both histological findings and the presence of circulating autoantibodies suggest an immune-mediated attack on esophageal myenteric neurons, involving antibody responses against as-yet unidentified antigens[12].

In patients with AC, inflammatory degenerative neuropathy of the LES is characterized by inflammatory infiltrates, fibrosis, loss of ganglion cells, and reduced numbers of interstitial cells of Cajal[3]. These neuropathic changes suggest a potential immune-mediated process, prompting further investigation into autoimmune responses against the enteric nervous system. A previous study showed that serum from patients with AC reduced inhibitory neurons, such as nitric oxide synthase- and vasoactive intestinal peptide-positive neurons, and increased excitatory cholinergic neurons, including choline acetyltransferase-positive neurons, suggesting a potential role for circulating autoantibodies in disease pathogenesis[13]. Subsequent studies[14,15] further demonstrated that the prevalence of anti-enteric neuronal antibodies (AENA) in patients with AC was significantly higher than in healthy subjects (HS), providing compelling evidence for an autoimmune component in the pathophysiology of AC.

Although AENA may be involved in autoimmune-mediated damage in AC, evidence linking AENA profiles to disease progression and therapeutic response remains limited. Therefore, this study aimed to investigate the prevalence of serum AENA in a Chinese cohort of patients with AC and to evaluate whether AENA status correlates with disease phenotype, severity, and treatment outcomes.

MATERIALS AND METHODS

Study design and subjects

This prospective cohort study was conducted at the Gastroenterology Clinic of Peking Union Medical College Hospital (PUMCH), a high-volume tertiary teaching center, between February 2019 and March 2025. All participants provided written informed consent before enrollment. The study protocol was approved by the PUMCH Ethics Committee (approval No. I-25PJ0734).

Patients diagnosed with AC were consecutively enrolled. The inclusion criteria were as follows: (1) Age between 18 years and 75 years; (2) Fulfillment of the diagnostic criteria for AC according to the Chicago Classification version 4.0[1]; and (3) Completion of symptomatic assessment, esophagogastroduodenoscopy, high-resolution manometry (HRM), and barium esophagram at PUMCH. Individuals who were pregnant or lactating were excluded from the study. Key exclusion criteria related to gastrointestinal comorbidities included gastroesophageal reflux disease, eosinophilic esophagitis, Barrett esophagus, esophageal diverticula or hiatal hernia identified on barium esophagram, and other organic gastrointestinal disorders potentially affecting esophageal function.

HS without gastrointestinal symptoms were recruited from the health examination center during the same period. All HS had normal physical examination and laboratory test results. They were matched to patients with AC by age (± 2 years) and sex. Serum samples from HS were collected as part of routine health assessments, and both antinuclear antibodies (ANA) and AENA testing were performed for all enrolled subjects.

Baseline clinical assessment

All patients with AC underwent comprehensive baseline evaluations, including: (1) Clinical data: Demographic information [sex, age, and body mass index (BMI), disease duration, previous interventions, and autoimmune comorbidities were recorded]; (2) Symptom assessment: Symptom severity was quantified using the Eckardt score[16]; (3) Endoscopic evaluation: Esophagogastroduodenoscopy was performed, and disease severity was graded using the novel endoscopic scoring for achalasia [CARS (contents, anatomy, resistance, and stasis)][17]; (4) HRM: HRM was conducted according to the standard protocol, and key parameters were analyzed following the criteria of the Chicago Classification version 4.0[1], including AC subtype, integrated relaxation pressure (IRP), basal LES pressure (BLESP), and distal contractile integral (DCI); and (5) Radiologic assessment: A barium esophagogram was used to measure the maximum barium column width and assess esophageal tortuosity.

The Eckardt score was used to assess symptom severity in patients with AC. This scoring system combines the frequency scores for dysphagia, retrosternal pain, and regurgitation, with each symptom graded on a scale from 0 to 3 (0 = absent; 1 = occasional; 2 = daily; 3 = every meal), as well as a weight loss component scored from 0 to 3 (0 = no weight loss; 1 = < 5.0 kg; 2 = 5.0-10.0 kg; 3 = > 10.0 kg). The total Eckardt score ranges from 0 (lowest severity) to 12 (highest severity).

Endoscopic evaluation was performed using the novel CARS[17], which has been validated to quantify disease severity[18] and post-POEM response[19]. The CARS assigns 0 to 2 points to each of the domains of contents, anatomy, and resistance at the LES, and 1 point to each component of stasis. Inter-rater reliability (IRR) for the total CARS was assessed in a random subset of baseline examinations using a two-way random-effects intraclass correlation coefficient (ICC).

HRM was performed using an intra-esophageal pressure catheter (GIM6000; Medical Measurement Systems, Enschede, the Netherlands), equipped with four pressure sensors located 5 cm, 10 cm, 15 cm, and 20 cm proximal to the upper border of the LES. The protocol included a 5-minute baseline LES pressure recording, followed by 10 times 5-mL water swallows in the supine position, with 30-second intervals between swallows. Manometric tracings were interpreted according to the Chicago Classification version 4.0[1].

The degree of esophageal distortion on barium esophagogram was classified into four grades: None, mild, moderate, and severe. Mild distortion was defined as slight dilation or localized tortuosity with mild distal narrowing. Moderate distortion was characterized by a “C”-shaped or partial “S”-shaped configuration, whereas severe distortion included an L-type or sigmoid-type esophagus. The grading was determined by senior gastroenterologists and radiologists with more than 10 years of experience.

Treatment and follow-up protocol

A subset of eligible patients subsequently underwent POEM at our institution using the standardized technique described by Inoue et al[20]. Postoperative follow-up assessments were conducted at 1 month and 6 months and included evaluation of Eckardt scores. At the 6-month follow-up, repeat esophagogastroduodenoscopy (with grading of erosive esophagitis according to the Los Angeles classification[21]), HRM, and barium esophagogram were performed.

Detection of serum AENA

Serum AENA was detected according to the protocol described by Wood et al[22]. Intestinal whole-mount preparations were generated from male albino Hartley guinea pigs for AENA detection, following previously established procedures[23]. Animals were euthanized by stunning and exsanguination. The small intestine was immediately excised, flushed with ice-cold Krebs solution (composition, in mM: 120.9 sodium chloride, 5.9 potassium chloride, 1.2 magnesium chloride, 1.2 sodium dihydrogen phosphate, 14.4 sodium bicarbonate, 2.5 calcium chloride, and 11.5 glucose), and divided into segments. The segments were then flushed again with ice-cold Krebs solution, opened along the mesenteric border, maximally stretched, and pinned onto Sylgard resin.

Tissue fixation was performed on the stretched preparations using formalin-picric acid solution (2% formalin containing 0.2% picric acid) for 24 hours. Whole-mount preparations of the submucosal plexus were obtained by microdissection to remove the muscularis externa together with the myenteric plexus.

For indirect immunofluorescence detection, whole-mount preparations were used to assess serum AENA. Tissues were washed three times in phosphate-buffered saline (PBS) containing 0.3% Triton X-100 for 20 minutes and then blocked with 10% normal goat serum (Solarbio, Beijing, China) in PBS containing 0.3% Triton X-100 at room temperature for 2 hours. For neuronal staining, the preparations were incubated overnight at 4 °C with: (1) Mouse anti-Hu antiserum (1:200) as a positive control[24]; (2) PBS as a negative control; and (3) Sera (1:50) from patients with AC or HS as experimental samples. Human whole blood samples were collected into serum tubes (4 mL with coagulant/gel, Greiner Bio-One, Kremsmünster, Austria) after an overnight fast.

Briefly, the samples were centrifuged at 300 × g for 5 minutes and stored at -80 °C until analysis. After three 10-minute washes in PBS, the preparations were incubated with fluorescein isothiocyanate-conjugated secondary antibodies (1:200; Abmart, Shanghai, China) for 1 hour in the dark, followed by five 10-minute washes in PBS. The samples were then cover-slipped and examined under a fluorescence microscope (Axio Imager Z2; Carl Zeiss AG, Oberkochen, Germany) to ensure labeling quality.

All images were captured and evaluated by two independent senior investigators in a double-blind manner. Two additional senior experts adjudicated discrepancies. If disagreement persisted, the assay was repeated, and the final classification was determined based on the repeat results. Fluorescence intensity was defined as follows: “0” = negative (no neuronal staining above background); “1” = moderately positive (staining comparable to anti-Hu controls); and “2” = strongly positive (staining exceeding anti-Hu controls)[23]. Figure 1 illustrates the immunofluorescence staining patterns of AENA. The AENA-positive rate was calculated as the percentage of subjects in each group who were moderately or strongly positive. The AENA strong-positive rate was defined as the percentage of strongly positive subjects within each group.

Figure 1 Immunofluorescence staining patterns of anti-enteric neuronal antibodies. A: Indirect immunofluorescence images of anti-enteric neuronal antibody (AENA)-positive and negative controls; B: Representative indirect immunofluorescence images demonstrating AENA negative (-), moderate-positive, and strong-positive staining intensities. Immunostaining was predominantly localized to the cytoplasm. Large yellow arrows indicate submucosal ganglia, and small yellow arrows indicate myenteric neurons (scale bar = 20 μm). AENA: Anti-enteric neuronal antibody.

Detection of serum ANA

ANA was detected using a standardized indirect immunofluorescence assay (Euroimmun, Lübeck, Germany) in accordance with established clinical protocols for systemic autoimmune diseases[25]. Briefly, sera were diluted to 1:100, incubated on substrate slides for 30 minutes, washed with PBS for 5 minutes, and then incubated with fluorescein isothiocyanate-conjugated anti-human immunoglobulin G for 30 minutes in the dark. After a final 5-minute wash, the slides were mounted with glycerol and independently evaluated by two experienced technicians using a fluorescence microscope (Olympus BX51). ANA positivity was defined as a titer ≥ 1:80, in accordance with established clinical criteria.

Statistical analysis

All statistical analyses were performed using R software (version 4.5.1). Continuous variables with a normal distribution are presented as the mean ± SD, whereas non-normally distributed variables are presented as the median and interquartile range (IQR). For group comparisons, normally distributed data were analyzed using the Student t test, whereas non-normally distributed data were analyzed using the Wilcoxon rank-sum test or Mann-Whitney U test, as appropriate. Categorical variables were compared using the χ² test or Fisher’s exact test, as appropriate. Multivariable linear regression analyses were performed to examine the independent associations between AENA intensity and clinical parameters, adjusting for relevant covariates. Propensity score-based overlap weighting was applied to estimate the average treatment effect in the overlap population and to balance baseline characteristics between AENA-positive and AENA-negative patients. The covariates included in the weighting model were age, sex, disease duration, baseline Eckardt score, and CARS score. Statistical significance was defined as a two-sided P value < 0.05.

RESULTS

Baseline characteristics of the study cohort

A total of 97 patients with AC and 98 HS were enrolled in this study. Age (AC: 44.7 ± 14.3 years vs HS: 41.8 ± 6.6 years, P = 0.135) and sex distribution (AC: 50.5% female vs HS: 56.1% female, P = 0.521) were comparable between the two groups. In the AC cohort, the median disease duration was 4.0 (IQR: 2.0 to 6.0) years, and the median BMI was 20.8 (IQR: 18.8 to 23.4) kg/m². Autoimmune comorbidities were present in 4 patients (4.1%), including Hashimoto’s thyroiditis, immunoglobulin A vasculitis, and rheumatoid arthritis.

According to the Chicago classification version 4.0, type I and type II AC each accounted for 48.5% of cases (47/97). Three patients had inconclusive subtyping due to unsuccessful LES pressure recordings during HRM, which were attributed to excessive esophageal resistance.

Prevalence of serum AENA and ANA

The overall serum AENA positivity rate was significantly higher in patients with AC than in those with HS (60.8% vs 25.5%, χ² = 24.79, P < 0.001; Figure 2). Strong AENA positivity was also more frequent in patients with AC (22.7% vs 4.1%, χ² = 14.59, P < 0.001; Figure 2). However, ANA positivity did not differ significantly between patients with AC and HS (28.9% vs 20.4%, χ² = 1.88, P = 0.171). Notably, even among ANA-negative subjects, AENA positivity remained significantly higher in patients with AC than in HS (59.4% vs 23.0%, χ² = 20.13, P < 0.001; Figure 2).

Figure 2 Distribution and overlap of serum anti-enteric neuronal antibodies and antinuclear antibodies in patients with achalasia. The stacked bar chart displays the distribution of anti-enteric neuronal antibodies (AENA) status stratified by antinuclear antibodies (ANA) classification. Among ANA-negative patients, 59.4% were AENA-positive and 40.6% were AENA-negative, compared with 60.8% and 39.2%, respectively, in the overall achalasia cohort. AENA: Anti-enteric neuronal antibody; ANA: Antinuclear antibodies; AC: Achalasia of cardia; HS: Healthy subjects.

Association between AENA status and clinical phenotype

As shown in Table 1, no significant differences were observed in age (46.7 ± 14.2 years vs 41.6 ± 13.9 years, P = 0.082) or disease duration [4.0 (2.5-5.0) years vs 3.6 (2.0-9.0) years, P = 0.792] between the AENA-positive and AENA-negative groups. The total Eckardt scores (AENA-negative patients: 7.0 vs AENA-positive patients: 6.0, P = 0.905; Table 1) and individual symptom subscores also showed no significant between-group differences (P > 0.05; Supplementary Table 1).

Table 1 Correlation of achalasia of cardia clinical profiles and anti-enteric neuronal antibodies status, mean ± SD/n (%)/median interquartile range.

Parameter	AENA negative (n = 38)	AENA positive (n = 59)	P value
Demography
Female	17 (44.7)	32 (54.2)	0.480
Age, year	41.6 ± 13.9	46.7 ± 14.2	0.082
Weight, kg	57.0 (50.0 to 65.0)	58.00 (53.0 to 66.0)	0.453
BMI, kg/m2	20.3 (18.3 to 21.4)	21.63 (19.2 to 24.2)	0.051
Disease duration, year	3.6 (2.0 to 9.0)	4.0 (2.5 to 5.0)	0.792
Previous intervention	4 (10.5)	11 (18.6)	0.559
Eckardt scores	7.0 (5.0 to 8.0)	6.0 (5.0 to 8.0)	0.905
CARS scores	4.0 (3.0 to 5.0)	5.0 (3.0 to 5.0)	0.045a
HRM parameters
IRP, mmHg	20.2 ± 10.5	17.2 ± 9.0	0.214
BLESP, mmHg	29.8 (22.1 to 40.0)	30.1 (22.4 to 37.5)	0.714
DCI, mmHg/second/cm	47.0 (2.0 to 160.0)	59.5 (2.0 to 176.0)	0.888
AC subtype			> 0.99
Type I	19 (50.0)	28 (47.5)
Type II	18 (47.4)	29 (49.2)
Barium esophagogram
Barium width, mm	41.5 (36.5 to 49.0)	41.5 (35.0 to 47.0)	0.652
Esophageal distortion			0.503
None	18 (56.3)	23 (42.6)
Mild	6 (18.8)	21 (38.9)
Moderate	6 (18.8)	6 (11.1)
Severe	2 (6.3)	4 (7.4)

^aP < 0.05.

Comparisons were performed using Student’s t-test, Mann-Whitney U test, or χ² test, as appropriate. AENA: Anti-enteric neuronal antibodies; BMI: Body mass index; CARS: Contents, anatomy, resistance, and stasis; HRM: High-resolution manometry; IRP: Integrated relaxation pressure; BLESP: Basal lower esophageal sphincter pressure; AC: Achalasia of cardia; DCI: Distal contractile integral.

The IRR for the total CARS score demonstrated good agreement, with an ICC of 0.86 [95% confidence interval (CI): 0.69-0.94]. Notably, the endoscopic CARS score was significantly higher in AENA-positive patients with AC than in AENA-negative patients (median, 5.0 vs 4.0; P = 0.045; Table 1). Among the four components of the CARS system, the “anatomy” subscore, which indicates esophageal dilation, was significantly higher in the AENA-positive group (P < 0.001; Table 2). No statistically significant differences were observed in HRM parameters (IRP, DCI, BLESP, and AC subtypes) or barium esophagogram findings between the groups (all P > 0.05; Table 1).

Table 2 Comparison of individual contents, anatomy, resistance, and stasis components across different intensities of anti-enteric neuronal antibodies, n (%).

CARS component1	AENA negative (n = 38)	AENA positive (n = 58)	P value	AENA moderate-positive (n = 36)	AENA strong-positive (n = 22)	P value
Content			0.750			0.233
None	1 (2.6)	5 (8.6)		5 (13.9)	0 (0.0)
Retained secretions or liquid	22 (57.9)	30 (51.7)		18 (50.0)	12 (54.5)
Retained solids	15 (39.5)	23 (39.7)		13 (36.1)	10 (45.5)
Anatomy			< 0.001b			< 0.001b
Normal caliber	3 (7.9)	4 (6.9)		4 (11.1)	0 (0.0)
Dilated lumen	32 (84.2)	30 (51.7)		23 (63.9)	7 (31.8)
Severely dilated lumen	3 (7.9)	24 (41.4)		9 (25.0)	15 (68.2)
Resistance			0.431			0.017a
None	5 (13.2)	3 (5.2)		2 (5.6)	1 (4.5)
Mild	20 (52.6)	33 (56.9)		25 (69.4)	8 (36.4)
Significant	13 (34.2)	22 (37.9)		9 (25.0)	13 (59.1)
Stasis			0.626			0.059
No evidence for stasis	27 (71.1)	39 (67.2)		21 (58.3)	18 (81.8)
Chronic stasis changes	11 (28.9)	17 (29.3)		13 (36.1)	4 (18.2)
Candida esophagitis	0 (0.0)	2 (3.4)		2 (5.6)	0 (0.0)

^aP < 0.05.

^bP < 0.01.

¹The sample size for anti-enteric neuronal antibodies (AENA)-positive in assessing CARS (contents, anatomy, resistance, and stasis) was 58 (not 59), due to one patient in the moderate-positive AENA group refusing endoscopic assessment.

Comparisons were performed using the Mann-Whitney U test. CARS: Contents, anatomy, resistance, and stasis; AENA: Anti-enteric neuronal antibodies.

Disease severity stratified by AENA intensity

When AENA-positive patients with AC were stratified by fluorescence intensity, they were categorized into a moderate-positive subgroup (n = 37) and a strong-positive subgroup (n = 22). No significant differences were observed in demographic characteristics or symptom severity between the two subgroups (Table 3). However, patients in the strong-positive subgroup exhibited significantly greater disease severity than those in the moderate-positive subgroup. This was reflected by higher total CARS scores (median, 5.0 vs 4.0; P = 0.034; Table 3), particularly in the subscores for anatomical abnormalities (severely dilated lumen: 68.2% vs 25.0%, P < 0.001; Table 2) and LES resistance (significant resistance: 59.1% vs 25.0%, P = 0.017; Table 2). In addition, the strong-positive subgroup demonstrated greater barium column width (45.0 mm vs 38.3 mm, P = 0.029; Table 3) and a higher median IRP [22.1 (13.5-29.3) mmHg vs 14.5 (10.2-19.8) mmHg, P = 0.018; Table 3]. Other HRM parameters and the degree of esophageal distortion remained comparable between the subgroups (all P > 0.05; Table 3).

Table 3 Association between clinical characteristics of achalasia and anti-enteric neuronal antibodies intensity, mean ± SD/n (%)/median interquartile range.

Parameter	AENA moderate-positive (n = 37)	AENA strong-positive (n = 22)	P value
Demographic
Female	21 (56.8)	11 (50.0)	0.815
Age, year	46.6 ± 14.8	46.8 ± 13.5	0.959
Weight, kg	60.0 ± 9.9	59.1 ± 11.6	0.770
BMI, kg/m2	21.9 ± 3.0	21.0 ± 3.3	0.285
Disease duration, year	3.5 (2.0 to 5.0)	4.7 (3.0 to 6.0)	0.100
Previous intervention	7 (18.9)	2 (9.1)	0.461
Eckardt scores	6.8 ± 1.9	6.4 ± 1.6	0.395
CARS scores	4.0 (3.0 to 5.0)	5.0 (4.0 to 5.0)	0.034a
HRM parameters
IRP, mmHg	14.5 (10.2 to 19.8)	22.1 (13.5 to 29.3)	0.018a
BLESP, mmHg	29.3 (23.2 to 33.6)	32.2 (21.7 to 45.8)	0.154
DCI, mmHg/second/cm	56.0 (2.0 to 108.0)	125.0 (0.0 to 361.0)	0.223
AC subtype			0.654
Type I	19 (52.8)	9 (42.9)
Type II	17 (47.2)	12 (57.1)
Barium esophagogram
Barium width, mm	38.3 (35.0 to 45.0)	45.0 (40.0 to 51.0)	0.029a
Esophageal distortion			0.503
None	13 (39.4)	10 (47.6)
Mild	14 (42.4)	7 (33.3)
Moderate	5 (15.2)	1 (4.8)
Severe	1 (3.0)	3 (14.3)

^aP < 0.05.

Comparisons were performed using Student’s t-test, Mann-Whitney U test, or χ² test, as appropriate. AENA: Anti-enteric neuronal antibodies; BMI: Body mass index; CARS: Contents, anatomy, resistance, and stasis; HRM: High-resolution manometry; IRP: Integrated relaxation pressure; BLESP: Basal lower esophageal sphincter pressure; AC: Achalasia of cardia; DCI: Distal contractile integral.

AENA status and treatment outcomes after POEM

Among the 97 patients with AC, 41 (18 AENA-negative and 23 AENA-positive) underwent POEM and completed the subsequent 6-month follow-up assessments (Table 4). To mitigate potential confounding by indication and balance baseline characteristics between the two groups, overlap weighting was applied to this POEM subgroup (Supplementary Table 2). After overlap weighting, AENA-positive patients had worse Eckardt scores than AENA-negative patients at both the 1-month follow-up (median, 1.0 vs 0.0; P = 0.031) and the 6-month follow-up (median, 1.0 vs 0.0; P = 0.005) (Table 4, Figure 3). The magnitude of improvement in Eckardt score, CARS score, barium column width, and manometric parameters from baseline to the 6-month follow-up was also analyzed, and no statistically significant differences were found between AENA-positive and AENA-negative patients (all P > 0.05; Table 4).

Figure 3 Longitudinal changes in Eckardt scores before and after peroral endoscopic myotomy stratified by anti-enteric neuronal antibody status. Individual patient trajectories are depicted by gray lines. Scatter plots represent anti-enteric neuronal antibodies (AENA)-negative (blue) and AENA-positive (red) groups. Horizontal and vertical bars indicate median values with interquartile ranges. Eckardt scores were significantly higher in the AENA-positive group at 1-month and 6-month follow-up (P < 0.05). ^aP < 0.05. NS: Not significant; AENA: Anti-enteric neuronal antibody.

Table 4 Comparison of treatment outcomes between anti-enteric neuronal antibodies-positive and -negative patients with achalasia after overlap weighting, mean ± SD/median interquartile range.

Outcomes	AENA negative (n = 18)	AENA positive (n = 23)	P value
1-month follow-up
Eckardt scores	0.00 (0.00 to 1.00)	1.00 (0.58 to 2.00)	0.031a
Δ Eckardt scores1	6.00 (5.00 to 7.31)	4.68 (4.00 to 6.48)	0.173
6-month follow-up
Eckardt scores	0.00 (0.00 to 1.00)	1.00 (1.00 to 2.00)	0.005b
Δ Eckardt scores1	5.51 (4.00 to 9.00)	5.00 (4.00 to 6.20)	0.233
CARS scores	1.00 (1.00 to 1.43)	1.00 (1.00 to 2.05)	0.915
Δ CARS scores1	2.89 (1.00 to 4.00)	2.98 (1.00 to 4.00)	0.939
Reflux esophagitis			0.497
None	68.9	64.1
Los Angeles-A	10.0	28.1
Los Angeles-B	10.9	2.8
Los Angeles-C	10.1	5.0
IRP, mmHg	1.99 (0.69 to 4.38)	2.36 (0.67 to 5.00)	0.840
Δ IRP, mmHg1	9.67 (-2.05 to 13.77)	13.40 (7.55 to 17.69)	0.359
Barium width, mm	25.10 (11.00 to 30.00)	22.84 (15.39 to 36.55)	0.674
Δ Barium width, mm1	16.82 (9.18 to 24.17)	19.29 (0.16 to 26.08)	0.574

^aP < 0.05.

^bP < 0.01.

¹Changes compared to these at baseline.

Statistical comparisons were performed using weighted Wilcoxon rank-sum tests for continuous variables and weighted χ² tests for categorical variables to account for overlap weights. AENA: Anti-enteric neuronal antibodies; CARS: Contents, anatomy, resistance, and stasis; IRP: Integrated relaxation pressure.

AENA intensity as an independent predictor of endoscopic severity

Figure 4 and Table 5 showed the association between AENA intensity and endoscopic CARS scores. Multivariable linear regression analysis adjusted for age, sex, disease duration, and AC subtype identified strong AENA positivity as an independent factor associated with a 0.95-unit increase in the CARS score (β = 0.95, 95%CI: 0.37-1.53; P = 0.002; Table 5) compared with the AENA-negative group. In contrast, no significant difference was observed in the moderate-positive group (β = 0.08, 95%CI: -0.65 to 0.80; P = 0.838; Table 5). This association remained significant when AENA intensity was analyzed as an ordinal variable (P for trend = 0.005 via ordinal regression), with each one-category increase in AENA intensity corresponding to a 0.43-unit increase in the CARS score (β = 0.43, 95%CI: 0.14-0.73; Figure 5 and Table 6). Sensitivity analysis excluding the 4 patients with documented autoimmune comorbidities showed that the association between AENA positivity and clinical outcomes remained consistent (Supplementary Table 3).

Figure 4 Multivariable linear regression analysis of anti-enteric neuronal antibody intensity and endoscopic severity using the novel endoscopic scoring for achalasia contents, anatomy, resistance, and stasis score. Adjusted predicted mean CARS (contents, anatomy, resistance, and stasis) scores stratified by anti-enteric neuronal antibody intensity; error bars represent 95% confidence intervals. AENA: Anti-enteric neuronal antibody; CARS: Contents, anatomy, resistance, and stasis.

Figure 5 Dose-response association between anti-enteric neuronal antibody intensity and contents, anatomy, resistance, and stasis scores. Predicted marginal means of CARS (contents, anatomy, resistance, and stasis) scores across anti-enteric neuronal antibody intensity levels. Anti-enteric neuronal antibody intensity was analyzed as an ordinal variable (negative = 0; moderate-positive = 1; strong-positive = 2). The solid line indicates the linear trend, and the shaded area represents the 95% confidence interval. AENA: Anti-enteric neuronal antibody; CARS: Contents, anatomy, resistance, and stasis.

Table 5 Multivariable linear regression model for contents, anatomy, resistance, and stasis scores.

Variable	β (95%CI)	P value
AENA intensity (negative)
Moderate-positive	0.08 (-0.65 to 0.80)	0.838
Strong-positive	0.95 (0.37-1.53)	0.002b
Age (per year)	0.00 (-0.02 to 0.02)	0.851
Gender (male)	0.15 (-0.41 to 0.71)	0.596
Disease duration (per year)	0.00 (-0.04 to 0.04)	0.943
AC subtype	-0.06 (-0.64 to 0.52)	0.841

^bP < 0.01.

The model was adjusted for age, sex, disease duration, and achalasia subtype. AENA: Anti-enteric neuronal antibody; CI: Confidence interval; AC: Achalasia of cardia.

Table 6 Dose-response association between anti‐enteric neuronal antibody intensity levels and contents, anatomy, resistance, and stasis scores.

Variable	β (95%CI)	P value
AENA intensity (per category increase)	0.43 (0.14-0.73)	0.005b
Age (per year)	-0.00 (-0.02 to 0.02)	0.990
Gender (male vs female)	0.12 (-0.44 to 0.68)	0.670
Disease duration (per year)	0.00 (-0.03 to 0.04)	0.802
Achalasia subtype	-0.04 (-0.62 to 0.54)	0.893

^bP < 0.01.

The model was adjusted for age, sex, disease duration, and achalasia subtype. Anti-enteric neuronal antibody intensity was analyzed as an ordinal variable (negative = 0; moderate-positive = 1; strong-positive = 2). AENA: Anti-enteric neuronal antibody; CI: Confidence interval.

DISCUSSION

This prospective study, conducted in a large Chinese cohort, demonstrated that serum AENA was significantly more prevalent in patients with AC than in HS. Notably, we identified a novel clinical correlation: The intensity of AENA immunoreactivity was independently associated with endoscopic severity in a dose-dependent manner and with higher LES pressure, greater esophageal dilation, and less favorable short-term symptomatic improvement after POEM. These findings suggest that AENA intensity may serve as a serological marker associated with advanced disease features.

In our cohort, 60.8% of patients with AC were AENA-positive, a prevalence significantly higher than that in HS. Notably, strong AENA positivity was observed in 22.7% of patients with AC, compared with only 4.1% in HS. Since 2003, accumulating studies have reported a high prevalence of antibodies against myenteric neurons in patients with AC using various methodologies (AC: 24.4%-100% vs HS: 0%-12.3%)[14,15,26,27]. Subsequent studies further identified a high positive rate of specific AENA subtypes in AC[15,27]. Serum antineuronal antibodies frequently detected in AC included antibodies against ganglionic acetylcholine receptors[28], glutamic acid decarboxylase-65[29], and PNMA2 (Ma2/Ta) antigen[29]. Our data further strengthen the growing body of evidence from a relatively large cohort, supporting an autoimmune component in AC pathogenesis and providing new insights into the clinical implications of AENA.

A notable finding of our study was the dose-dependent relationship between AENA intensity and endoscopic severity. Multivariable linear regression confirmed that strong AENA positivity was an independent determinant of higher CARS scores. Moreover, the significant dose-response relationship (β = 0.43; P = 0.005) demonstrated a stepwise increase in endoscopic severity with increasing AENA intensity.

However, some earlier, smaller studies failed to identify associations between AENA[26] or autoimmune comorbidities[10] and clinical features. This discrepancy might highlight the importance of adequate sample size and quantification of antibody intensity.

AENA-positive patients showed persistently higher Eckardt scores at both 1 month and 6 months after POEM, despite comparable improvements in manometric parameters between groups. This discrepancy suggests that AENA positivity may be associated with symptom persistence through immune-associated mechanisms beyond simple reduction of LES tone, potentially involving tissue fibrosis, neuromuscular dysfunction, or persistent structural remodeling after the procedure. This interpretation is consistent with our observation that endoscopic severity, as assessed by the CARS score, remained significantly associated with AENA status. The CARS score reflects long-term structural and mucosal changes and has been shown to better reflect treatment response after POEM[19].

Previous studies have demonstrated a dissociation among symptom relief, IRP reduction, and esophagogastric junction distensibility. The Eckardt score is strongly influenced by subjective symptom perception, whereas IRP reflects a single-time-point physiologic measurement. Rohof et al[30] reported a poor correlation between dysphagia relief and IRP in surgically treated patients with AC. Therefore, the association between AENA and endoscopic features may be more clinically informative than correlations with symptom scores or manometric parameters alone.

The poor correlation between IRP and treatment outcomes may be explained by the fact that IRP recorded by HRM mainly reflects active LES muscle tone during swallow-induced relaxation, whereas persistent immune activation and AENA-mediated neuronal injury may promote fibrotic remodeling of the esophagogastric junction[31]. Such remodeling can impair esophagogastric junction distensibility, which has recently been shown to correlate better with post-intervention symptoms than manometric pressure measurements[32,33]. This finding highlights a subgroup of patients characterized by a less favorable symptomatic response, suggesting that factors beyond standard LES disruption may influence postprocedural outcomes.

These observations are consistent with the theory that the severity of the autoimmune attack on the myenteric plexus is a key determinant of the clinical presentation of AC, as first proposed by Kahrilas and Boeckxstaens[34]. Recent research by Liu et al[35] reported that type I AC is characterized by a significantly wider esophageal diameter than type II, which is associated with greater fibrosis, more severe inflammation-driven ganglion loss, and a reduced number of interstitial cells of Cajal in the muscularis propria of the LES[35]. Previous histopathological observations also indicated a correlation between the AC disease course and ganglion degeneration, with the absence of myenteric infiltrate correlating with a shorter symptom duration. In addition, histopathologic features of type III AC show relative preservation of neuronal populations, with only reduced inhibitory signaling and enhanced excitatory signaling observed[36].

A major challenge in understanding AC pathophysiology is distinguishing whether AENA drives neurodegeneration or arises secondarily from neuronal damage. The core pathophysiological abnormality in achalasia is the predominant loss of inhibitory neural control of the esophagus[3], induced by immune activation, as evidenced by histological inflammatory infiltrates extensively documented throughout the disease course[10,37]. Recently, Priego-Ranero et al[12] highlighted this local immune response by identifying ectopic expression of GAD65 and PNMA2 in AC tissues, accompanied by MMP9 activation and fibrosis, which correlated with high positivity for specific autoantibodies directed against these proteins.

These local immune phenomena are increasingly recognized to be driven by serum-mediated mechanisms. In 2006, Bruley des Varannes et al[13] demonstrated that serum from patients with AC induced key pathological features of the disease in an ex vivo healthy human fundus model, characterized by significant reductions in nitric oxide synthase- and vasoactive intestinal peptide-positive neurons, along with impaired nitric oxide-mediated relaxation. Subsequently, our previous in vitro study showed that, compared with weakly positive or negative AENA serum, strongly positive AENA serum promoted neuronal apoptosis by increasing cleaved caspase 3 expression, decreasing Bcl-2 expression, and significantly increasing apoptosis rates[23]. These findings suggested a potential mechanism underlying the neuronal loss characteristic of AC.

More recently, multi-omics data provided strong causal evidence supporting a circulating immune-mediated component in AC pathogenesis. A groundbreaking genome-wide association study involving 4602 patients and 10766 ethnically matched controls identified HLA-DQB1 variation that governs antigen presentation as the predominant risk factor for AC[38]. Moreover, Mendelian randomization analyses demonstrated that specific circulating mediators, such as TNF-related apoptosis-inducing ligand, drive the pathogenesis of AC by shaping a proinflammatory microenvironment[11]. Consistent with previous studies[14,15,26,27], our findings in this high-volume Chinese cohort demonstrated a high prevalence of AENA in patients with AC and its association with clinical severity in a dose-dependent manner. Within this genetically predisposed causal framework, our finding provides further evidence supporting the hypothesis that autoantibodies contribute to AC severity through immune-mediated structural changes in the esophageal wall. However, while causality remains unproven, our dose-dependent clinical associations suggest AENA reflects and may exacerbate disease severity.

Several limitations of this study should be acknowledged. First, although this study employed a prospective design, the assessment of AENA and baseline disease severity remained observational. Therefore, although our findings support a dose-dependent association, they do not establish pathogenic causality. Nevertheless, AENA may remain a clinically relevant serological marker reflecting disease severity and structural burden in AC. Future studies using human enteric neuronal cultures may help elucidate the underlying mechanisms of AENA. Additionally, they may cautiously explore the potential therapeutic implications of immunomodulatory approaches in AENA-positive patients. Second, tissue-based indirect immunofluorescence detects a broad anti-enteric neuronal immune response rather than specific antigenic targets. Moreover, potential interspecies differences may lead to partial cross-reactivity or failure to detect antibodies directed against human-restricted antigens. Despite these limitations, this model remains well established and widely used in the field for detecting AENA. Therefore, our findings provide evidence of anti-enteric neuronal autoimmunity in AC. Further validation using human antigen-specific assays is warranted to confirm the presence of specific pathogenic autoantibodies. Finally, the single-center design, lack of type III achalasia, limited long-term follow-up data for patients undergoing POEM, and a predominantly Chinese study population may constrain the external validity and generalizability of our findings. Because enteric neurons are relatively preserved in type III disease, subtype-specific differences in AENA patterns likely exist. Future multicenter studies including a broader spectrum of achalasia subtypes, particularly type III, are needed to address potential ethnic or genetic differences in immune-mediated susceptibility.

CONCLUSION

This study demonstrates that serum AENA is prevalent in AC and is independently associated with endoscopic disease severity in a dose-dependent manner. Furthermore, AENA positivity identifies a subgroup of patients with higher LES pressure, greater esophageal dilation, and less favorable short-term symptomatic improvement after POEM. These findings strengthen the role of autoimmunity in the pathophysiology of AC and support the potential clinical utility of AENA assessment as a serological biomarker.

ACKNOWLEDGEMENTS

We gratefully acknowledge the Clinical Biobank (ISO 20387) of Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, for providing biological samples and technical support.