Joint Hypermobility, Autonomic Dysfunction, Gastrointestinal Dysfunction, and Autoimmune Markers: Clinical Associations and Response to Intravenous Immunoglobulin Therapy

Abstract

INTRODUCTION:

We examined autoimmunity markers (AIM) and autonomic dysfunction in patients with chronic neurogastroenterological symptoms and their relationship to joint hypermobility/hypermobility spectrum disorder (JH/HSD).

METHODS:

AIM positivity was defined as a diagnosis of known autoimmune/autoinflammatory disorder with at least 1 positive seromarker of autoimmunity or at least 2 positive seromarkers by themselves. Three cohorts were studied: (i) retrospective (n = 300), (ii) prospective validation cohort (n = 133), and (iii) treatment cohort (n = 40), administered open-label intravenous immunoglobulin (IVIG).

RESULTS:

AIM positivity was found in 40% and 29% of the retrospective and prospective cohorts, the majority of whom (71% and 69%, respectively) had autoinflammatory disorder. Significantly more patients with AIM had elevations of C-reactive protein (31% vs 15%, P < 0.001) along with an increased proportion of cardiovascular autonomic dysfunction (48% vs 29%; P < 0.001), small fiber neuropathy (20% vs 9%; P = 0.002), and HLADQ8 positivity (24% vs 13%, P = 0.01). Patients with JH/HSD were more likely to have AIM (43% vs 15%, P = 0.001) along with more severe autonomic and gastrointestinal (GI) symptom scores. IVIG treatment was associated with robust improvement in pain, GI, and autonomic symptoms, but adverse events were experienced by 62% of patients.

DISCUSSION:

Autoimmune markers and autonomic dysfunction are common in patients with unexplained GI symptoms, especially in those with JH/HSD. Many patients seem to respond to IVIG treatment, but this needs to be confirmed by controlled trials. These results highlight the need for vigilance for autoimmune and autonomic factors and JH/HSD in patients with neurogastroenterological disorders. Clinicaltrials.gov, NCT04859829.

INTRODUCTION

An autoimmune etiology or association is often present in disorders affecting the peripheral motor, sensory, and autonomic nervous systems. Patients presenting primarily with these syndromes are commonly noted to have gastrointestinal (GI) symptoms (1,2). GI and autonomic dysfunction have also been seen in greater frequency in patients with joint hypermobility/hypermobility spectrum disorder (JH/HSD) (3,4). Despite these known associations, patients whose primary complaints are GI in nature are seldom investigated in detail for comorbid autonomic dysfunction, JH, or an underlying autoimmune disorder. Uncovering an immunological basis for the symptoms in these patients is important because of the lack of satisfactory treatment options.

Our principal aim was therefore to examine markers of autoimmunity and their association with dysautonomia and JH/HSD in a large patient population with refractory GI symptoms without an obvious etiology. To further test the role of putative autoimmunity in the pathogenesis of these symptoms, we also treated a separate cohort of these patients with empirical immunomodulation using intravenous immunoglobulin (IVIG) in a prospective open-label trial.

METHODS

Patients and treatments

Definition of “autoimmune marker positivity.”

This was defined as a known diagnosis of an autoimmune/autoinflammatory disorder (AIDX) with at least 1 circulating seromarker associated with autoimmunity, or in patients not meeting criteria for such a disorder, having 2 or more such seromarkers. Patients with a history of celiac disease were not classified as autoimmune marker (AIM) positive only on the basis of positive celiac serology (i.e., they had to have at least 1 additional autoantibody).

Retrospective cohort.

We first conducted a chart review on patients presenting to a single experienced gastroenterologist at a tertiary referral center during the period between March 2016 and October 2018. The primary inclusion criterion was the presence of unexplained GI symptoms (including pain) reflecting more than 1 region of the gut. Patients with single organ symptomatology (e.g., dysphagia due to achalasia or reflux symptoms only), inflammatory bowel disease, chronic pancreatitis, or other likely etiological factors (such as uncontrolled diabetes or thyroid/endocrine disorder, malignancy, or Parkinson disease) were not included. As per our practice, these patients routinely underwent multiple tests including serology and whole GI transit studies. In addition, tilt-table testing was performed for those who had symptoms of dysautonomia such as postural cardiovascular symptoms, temperature intolerance, and/or sweating abnormalities. These patients also underwent skin biopsies to look for objective signs of neuropathy. Details of these tests are provided in the Supplementary Digital Content (see Supplementary Material, http://links.lww.com/AJG/D318).

Prospective (validation) cohort.

This consisted of new patients presenting during the period between September 2019 and January 2021 and with the same inclusion/exclusion criteria. Patients were usually seen in a multidisciplinary clinic also staffed by a psychiatrist/pain management specialist (G.T.), a clinical psychologist (N.G.), and a rheumatologist (Z.M.). Patients were invited to participate in a registry with the aims of validating the findings of the retrospective cohort and to provide more granular clinical information using standardized patient-reported outcome (PRO) questionnaires (covering multiple GI symptom domains, autonomic symptoms, JH, pain, headache/migraine, and Rome IV (5) classification). Details of these questionnaires are provided in the Supplementary Digital Content (see Supplementary Material, http://links.lww.com/AJG/D318).

Pilot trial of IVIG.

This was a separate study initiated in March 2017 (before the prospective study and continuing in parallel with it thereafter). Patients were offered IVIG if they were suspected to have an underlying immune disorder based on a history of a known AIDX and/or at least 1 autoantibody (the presence of “AIM” was not an entry criteria since this definition had not been arrived at before this study was initiated). IVIG was administered at a dose of 2 gm/kg body weight administered monthly; patients with at least 3 monthly courses were included in the final analysis. As a primary endpoint, we used a 15-point overall treatment effectiveness (OTE) questionnaire reported on a scale from −7 (worst) to 7 (best) points (6); secondary endpoints included changes in specific clinical symptoms measured by various PRO instruments.

Statistical methods.

Statistical significance was calculated by using the 2-sample t test or Pearson χ² test. Multiple comparisons were assessed by using the false discovery rate method. For the treatment cohort, the mean OTE score was compared with zero by using the one-sample t test. Statistical significance for the secondary measures was tested if the primary outcome was significant. The Hochberg method was then used to account for multiple comparisons among the secondary scales. Computations were performed using R software version 4.2.2.

Human subjects approval

The entire study was approved by the Institutional Review Board of the Johns Hopkins University School of Medicine, and the IVIG trial was registered with ClinicalTrials.gov (NCT04859829).

RESULTS

Retrospective cohort

There were 300 patients (89% female; mean age 40.5 ± 16.7 years) in this cohort. Presenting complaints included abdominal pain (89%), bloating/distention (88%), constipation (79%), nausea (78%), early satiety/fullness (69%), vomiting (50%), dysphagia (36%), and diarrhea (34%). Postural symptoms (dizziness, lightheadedness, syncope) were reported by 77% of patients and abnormal sweating by 53%. A history of migraine/recurrent headaches was present in 74%. One hundred twenty patients (40%) had AIM of which 85 (71%) had a history of a known AIDX (see Supplementary Table S1, http://links.lww.com/AJG/D319). However, GI and other symptoms were not attributed to the AIDX by the referring physician. Autoantibody markers and their frequency are provided in Supplementary Digital Content (see Supplementary Table S2, http://links.lww.com/AJG/D319).

Prospective cohort

There was a total of 133 patients (81% female; mean age 42.1 ± 16.9) in this cohort. These patients fulfilled diagnostic criteria for multiple Rome IV categories (see Supplementary Table S3, http://links.lww.com/AJG/D319). Alternatively, some of these patients could also be given diagnoses based on the results of gut transit studies, performed in 102 patients: 54 (53%) of these met standardized criteria for gastroparesis (7), 12 (12%) and 65 (64%) met criteria for delayed small bowel and colonic transit, respectively (8). In addition, 31 (30%) patients also had esophageal delay.

Of the 133 patients, 39 (29%) had AIM of whom 27 (69%) had a history of a known AIDX (see Supplementary Table S4, http://links.lww.com/AJG/D319). The most prevalent diagnoses were similar to the retrospective group—Hashimoto thyroiditis, Sjogren/Sicca syndrome, systemic lupus erythematosus/lupus-like syndrome, undifferentiated connective tissue disorder, and mixed connective tissue disorder. Supplementary Digital Content (see Supplementary Table S5, http://links.lww.com/AJG/D319) lists the autoantibodies and their frequency in these patients; the most prevalent antibodies were the same as that in the retrospective cohort (anti-ganglioside, anti-thyroid peroxidase, antinuclear antibodies, and anti-smooth muscle antibody [ASMA]).

Phenotypic differences between patients with and without AIM

We next examined phenotypic differences in patients with and without AIM in the prospective cohort based on the answers to the PRO questions (Table 1). JH/HSD was reported in 74% of patients in the AIM-positive group vs only 40% in the AIM-negative group (P = 0.001). Dysautonomia scores were much higher in AIM-positive patients (47 vs 35; P = 0.002) as was the severity of pelvic floor symptoms (pelvic floor distress inventory scores of 100 vs 60; P = 0.001).

Table 1.

Comparison of symptoms and demographics between patients with and without autoimmune markers in the prospective cohort^a

To further validate the definition of AIM, we used an alternative categorization method that distributed patients into 2 different groups based only on a known AIDX (see Supplementary Table S7, http://links.lww.com/AJG/D319). This classification, unlike the one using AIM, was not able to yield phenotypic differences.

Clinically distinct features of patients with JH/HSD

Several significant clinical differences were found between patients with and without JH/HSD (Table 2). They were more likely to be female (94% vs 68%, P < 0.001), and the prevalence of AIM was nearly 3-fold higher (43% vs 15%; P = 0.001). Upper GI symptoms were also significantly worse in this group, as were constipation-related symptoms and quality of life as well as pelvic floor symptoms. Pain scores particularly as measured by PROMIS intensity and interference measures were worse in patients with JH/HSD, along with Headache Screening Questionnaire migraine but not TTH scores.

Table 2.

Comparison of symptoms and demographics between patients with and without joint hypermobility/hypermobility spectrum disorder in the prospective cohort^a

The severity of dysautonomia in the JH/HSD group was worse, with nearly twice as many patients meeting the Composite Autonomic Symptom Score (COMPASS-31) derived definition and a mean score that was 16 points higher. To put this in perspective, we compared these scores with those reported elsewhere on healthy volunteers, patient with diabetic polyneuropathy, small fiber neuropathy, and scleroderma (9); patients with JH/HSD had the highest scores among all the groups (Figure 1).

Figure 1.

COMPASS-31 scores (means ± Standard Error of Mean), a measure of autonomic dysfunction, of patients with JH/HSD and without^a, as compared with healthy volunteers and patients with other disorders^b. ^aPatients in this study. ^bOther data from a previously published report and provided for comparison (8). *COMPASS scores in all disease categories were significantly higher than healthy controls (P < 0.001, one-way Analysis of Variance). **COMPASS scores in patients with JH/HSD were significantly higher than all other disease categories (P < 0.001). COMPASS scores in all other disease categories, including patients without JH/HSD were not significantly different than each other. COMPASS, composite autonomic symptom score; JH/HSD, joint hypermobility/hypermobility spectrum disorder; SSC, systemic scleroderma. Numbers above the error bars indicate the number of patients in the respective groups.

Association between AIM and objective markers of disease (combined cohorts)

Next, we attempted to further validate the AIM definition by examining its association with objective laboratory and other tests. To increase statistical power, we pooled prospective and retrospective cohorts, giving a total of 433 patients (Table 3) with results of blood tests and whole gut scintigraphy (routinely done in all patients) and tilt-table testing and skin biopsies (only done in patients with clinical suspicion of dysautonomia).

Table 3.

Comparison of objective tests and demographics between patients with and without autoimmune markers (AIM)^a

Clear differences were found in the AIM-positive group, including the presence of high C-reactive protein (31% vs 15%; P < 0.001), high Erythrocyte Sedimentation Rate (23% vs 12%; P = 0.008), and Human Leukocyte Antigen DQ8 (24% vs 13%; P = 0.01). The AIM group also had more patients who were tilt-table positive (48% vs 29%; P < 0.001) and had small fiber neuropathy (20% vs 9%; P = 0.002).

Response to IVIG in treatment cohort

This cohort included 40 patients with symptoms refractory to a variety of medications (see Supplementary Table S11, http://links.lww.com/AJG/D319) and treated with IVIG. Eight patients were excluded because they did not complete more than 2 courses (because of lack of insurance coverage in 4 and adverse effects in 2) or did not complete the surveys (n = 2), leaving 32 patients for analysis. Baseline features of these patients are described in Table 4. OTE scores are shown in Figure 2 with a mean of 1.8 (SD 3.2) that was significantly better than zero (95% Confidence Interval 0.6–2.9, P = 0.004), with an effect size (ES) of more than 0.5 SDs (indicating at least a medium effect). Secondary endpoints were also markedly improved after IVIG (Table 4). The largest ES (1.5) was observed for the COMPASS-31, but scores were only available for 12 patients. The Patient Assessment of Upper Gastrointestinal Disorders-Quality of Life had the largest ES (0.9) among scales measured for all subjects which corresponds to a “great or very great deal better” (10); mean scores on Patient Assessment of Constipation-Quality of Life improved by 0.5, also indicating a clinically significant response (11). The sample size was too small to allow for any meaningful analysis of predictors of response. Adverse effects were common and seen in 62% of patients with headaches as the most frequent and serious side effect (Table 5).

Table 4.

Baseline features of patients treated with IVIG

Figure 2.

Distribution of overall treatment effect (OTE) scores after treatment with IVIG scatter dot plot of individual responses with mean and 95% Confidence Interval (CI) indicated in grey. IVIG, intravenous immunoglobulin.

Table 5.

Change in patient-reported outcome scales and subscales after IVIG treatment

DISCUSSION

Based on detection of autoantibodies, an autoimmune pathogenesis has been suspected previously in small series of patients with constipation, irritable bowel syndrome, gastroparesis, and paraneoplastic conditions with GI symptoms (12–15). However, antibodies that have been shown to be pathogenic are very few (12,16–19), although numerous case series of patients with GI dysmotility and/or “Irritable Bowel Syndrome (IBS)” have been described with an inflammatory infiltrate around damaged enteric neurons or muscle, with or without circulating antibodies (20–22). In this study, we found a relatively high (29%–40%) proportion of patients with AIM. A diagnosis of an autoimmune disorder was not particularly challenging in the majority of these patients who had a history of a known AIDX (69%–71%). However, the AIM definition was able to discriminate between clinical phenotypes better than using AIDX by itself, indicating the possibility of as yet unidentified autoimmune disorders in the minority of patients who did not have a known disorder. Finally, the AIM group was enriched for patients with inflammatory markers and autonomic dysfunction with a higher background of HLA DQ8 positivity and therefore may serve as a way of identifying patients with an underlying autoimmune pathogenesis.

In this study, we found common autoantibodies such as antinuclear antibody, anti-thyroid peroxidase in addition to a variety of rarer antibodies. However, 2 of the most prevalent antibodies, anti-ganglioside antibodies and ASMA, have generally not been reported in this context. Anti-ganglioside antibodies have classically been described in Guillain-Barre syndrome and its variants that affect the peripheral nervous system and may represent an immune reaction to Campylobacter jejuni. (23) However, in our study, the presence of these antibodies did not correlate with skin neuropathy (see Supplementary Table S12, http://links.lww.com/AJG/D319). ASMA have generally been associated with autoimmune hepatitis, but also reported in up to 60% of patients with Sjogren, which is commonly associated with autonomic and GI symptoms (24,25) and often without overt liver disease (26). The prevalence of AIM in our patient groups may be significantly underestimated as we used a relatively limited set of commercially available autoantibodies and did not test for several others (e.g., anti-vinculin reported in patients with postinfectious IBS (27), and Fibroblast Growth Factor-3 and Trisulfated Heparin Disaccharide antibodies in small fiber neuropathy (28)). We hope that this study will encourage other investigators to investigate these possibilities and ideally, lead to a quest for more specific, but as yet unknown, immune markers (including both humoral and cellular components).

In our patients with AIM, we found more than a 2-fold greater prevalence of elevated C-reactive protein and Erythrocyte Sedimentation Rate values, consistent with an underlying inflammatory disorder. This group was more likely to express HLA-DQ8 which has been associated with several other autoimmune disorders including rheumatoid arthritis (29), autoimmune thyroiditis (30), T1 diabetes mellitus (31), and Addison disease (32). Although patients with Postural Tachycardia Syndrome commonly have GI symptoms and often suspected to have an autoimmune etiology (33–36), there have been only scattered reports of autonomic testing in patients presenting primarily with GI dysmotility and in whom autonomic dysfunction was not previously suspected (37). We found that nearly a third of all of our patients and nearly half of patients with AIM had objective evidence of autonomic dysfunction. Another novel aspect of our study was the large numbers of patients who underwent skin biopsy for neuropathy; small fiber neuropathy (SFN) was found in 25% of all our patients and in 31% of patients with AIM. The importance of SFN in neurogastroenterology has not been well studied, although a recent retrospective study of patients with SFN found a 54% prevalence of functional GI disorders, predominantly IBS, and functional constipation (38). Our results therefore underline the importance of maintaining a high state of vigilance for a systemic neural triopathy, possibly autoimmune in nature, that may involve the enteric, autonomic, and peripheral nervous systems in patients presenting with neurogastroenterological symptoms which are otherwise dismissed as “functional” or “disorders of gut-brain interaction.” Our results suggest that gastroenterologists should pay particular attention to a history of hypothyroidism and sicca symptoms, and appropriate testing should be performed. Furthermore, objective tests such as tilt-table and skin biopsies can often provide evidence of “organic” disease when GI examinations are negative.

This study also examined the relationship between JH and GI and/or autonomic symptoms in the prospective cohort. JH/HSD is known to be associated with GI dysfunction, dysautonomia, and other comorbidities in a third to as many of three-quarters of patients (39,40). A major contribution of this study has been to reliably quantify the severity of GI and autonomic symptoms in these patients. Upper and lower GI (including pelvic floor dysfunction) symptom scores were much higher patients with JH/HSD. The severity of dysautonomia (as measured by the COMPASS score) was markedly higher in patients with JH/HSD than that reported in other disorders of the peripheral nervous system (Figure 1). Pain (including migraine, but not tension-type headache) was also prominent using a variety of indices. Furthermore, none of our patients met the Rome IV diagnostic criteria for centrally mediated abdominal pain syndrome, widely believed to be the cause of pain in the kinds of patients whom we describe here (41). At the same time, measures of anxiety and depression were modest in both groups, lending support to the belief that these symptoms are not a result of psychological comorbidities.

Another novel finding of this study is the much higher prevalence of AIM in patients with JH/HSD (43% vs 15%, see Table 2); conversely 74% of patients with AIM had JH/HSD as compared with 40% without (Table 1). Such an association has not been described before and adds to the myriad comorbidities noted in patients with JH/HSD. Little is known about the pathogenesis of autonomic or GI symptoms in these patients, although extracellular matrix abnormalities may be important, as shown by mouse models (42,43). However, the prevalence of JH/HSD in the general population is probably much higher than those that become symptomatic (44), suggesting the possibility of a “second-hit” that is acquired after birth. The strong association with AIM suggests that autoimmunity may be one such factor, perhaps triggered by an infection or other environmental factors. It can be speculated that patients with JH are predisposed to autoimmunity because of aberrant transforming growth factor β signaling that may contribute to both connective tissue abnormalities and immune dysfunction, as has been reported in other disorders in the literature (45).

Table 6.

Adverse events reported with intravenous immunoglobulin (n, %)

We also report here the results of a prospective open-label trial of IVIG in patients suspected of having autoimmunity. IVIG resulted in significant improvements overall and in multiple GI, autonomic symptoms including abdominal pain. In this study, the response of patients who met our subsequent definition of AIM (n = 20) was not significantly different than those who did not (n = 12), which is not surprising given the small sample size (results not shown). Therefore, we cannot as yet predict which patients will respond best to IVIG. It should also be noted that IVIG treatment was associated with significant adverse events, with the incidence of headache and aseptic meningitis apparently higher in this population than that reported in general (33).

The strengths of our observations are the inclusion of a prospective cohort that was administered multiple validated questionnaires, the large numbers of patients who were systematically studied using a comprehensive set of objective tests to measure whole gut transit and autonomic neuropathy, and a single gastroenterologist with a uniform clinical approach. Nevertheless, there are several significant limitations of this study that we acknowledge, including the tertiary nature of the practice, the use of selected antibody markers of unknown relevance to the pathogenesis, and importantly, the lack of a randomized controlled trial design for IVIG treatment. Furthermore, although IVIG was used in this study to demonstrate proof of concept for immunomodulation, it is associated with significant costs and neurological adverse effects. Finally, we did not assess changes in the gut microbiome (colonic or small bowel) which is important as GI dysbiosis is now clearly implicated in the pathogenesis of so-called functional bowel disorders and may also be a potential driver of autoimmune/autoinflammatory disease (46,47).

In conclusion, we describe 4 features that are commonly found in patients with unexplained GI symptoms: JH, autonomic dysfunction, GI dysfunction, and autoimmune markers. We have termed these features collectively as “JAG-A,” if all 4 are present; however, patients can also present with various combinations of these features. Patients with JH appear to be particularly prone to display autoimmune markers and dysautonomia (in addition to having more severe GI symptoms and pain). It is therefore important for gastroenterologists to screen for JH, autonomic dysfunction, and pay attention to a history of AIDXs in patients presenting with neurogastroenterological symptoms. These findings need to be validated by additional prospective (and controlled) studies, coupled with further research on the pathobiology of the putative autoimmune condition.

CONFLICTS OF INTEREST

Guarantor of the article: Pankaj Jay Pasricha

Specific author contributions: P.J.P.: study concept and design, analysis and interpretation of data, drafting of the manuscript, obtained funding; M.M.: acquisition of data; L.V.: acquisition of data; G.B.: acquisition of data; J.B.: acquisition and interpretation of data, editing of manuscript; K.H.: acquisition of data, editing of manuscript; M.P.: acquisition and interpretation of data, editing of manuscript; R.B: acquisition of data; Z.H.M.: interpretation of data, editing of manuscript; N.G.: interpretation of data, editing of manuscript; B.G.: interpretation of data; critical revision of the manuscript for important intellectual content; J.H.: analysis and interpretation of data, drafting of the manuscript, statistical analysis; G.T.: analysis and interpretation of data, drafting of the manuscript, obtained funding.

Financial support: Supported in part by a gift from the Amos Family to the Johns Hopkins University School of Medicine

Potential competing interests: None to report.