Critical appraisal
To determine the quality of the included studies, the Newcastle Ottawa Scale was used. For both studies of Chughtai [
15,
16], the risk of bias is low: the selection of patients was considered representative, the cohort selection was done appropriately, surgical records seemed appropriate for selecting patients and outcomes were not present at start of the study. Comparability of cohorts was assessed. The outcomes were derived through record linkage, which imposes medium risk of bias. Follow-up for both the exposed cohort as well as controls was two years and considered appropriate.
The risk of bias for the study of Muller [
17] was considered low since a large number of patients were selected from a cohort of women who underwent SUI surgery with (intervention) or without (controls) mesh. The minimal follow-up duration of 5 years should be adequate and the study corrected for significant confounding factors, including age, ethnicity and pre-existing comorbidities. The study of Cohen Tervaert [
9] has low risk of bias considering the selection of patients with autoimmune disease, but patients were selected from a cohort with known autoimmune symptoms, and therefore, the outcome was present at start of the study. A non-exposed cohort was lacking. There was no description of ascertainment of exposure. Comparability of cohorts based on neither design nor analysis was described. Outcome assessment was done by record linkage. Follow-up was not performed (see Table
1).
Study characteristics
Two of the selected studies were retrospective cohort studies with matched controls and were performed in the USA, by Chughtai et al. [
15,
16]. One study was a national cohort study that has been carried out in the United Kingdom by Muller et al. [
17].
Cohen Tervaert performed the fourth study in the Netherlands, Canada and Belgium. This was a case series [
9]. The summary of included studies is shown in Table
2.
Table 2
Characteristics of the included studies
| N = 12,716 mesh hernia N = 25,432 Colonoscopy | Inclusion: Patients undergoing PP mesh inguinal hernia repair For control: patients undergoing colonoscopy | Cohort: 57.8 (12.8)* Control: 57.8 (12.8)* | At 6 months, 1 year, 2 years and during the entire FU period (average was 6 years) | Cohort: 188 patients (1.5%) had developed SAID at the end of follow-up Control: 413 patients (1.6%) had developed autoimmune disease at the end of follow-up. Result: Adjusted OR 0.91; 95% CI (0.76–1.09). No association was found between hernia mesh repair and the development of SAID at 6-month, 1-year and 2-year FU |
| N = 1507 mesh N = 3014 Colonoscopy N = 1375 Hysterectomy | Inclusion: Women undergoing mesh POP-repair For control: Women undergoing screening colonoscopy (non-surgical cohort) or hysterectomy for benign gynecologic or urogynecologic indications | Cohort: 60.4 (11.5)* Control: 60.4 (11.6)* | At 6 months, 1 year, 2 years and during the entire FU period (average was 6 years) | Mesh vs. Colonoscopy: Mesh cohort: 2.8% of patients developed SAID. Control: 2.8% of patients SAID. Adjusted OR 0.91; 95% CI (0.62–1.34) Mesh vs. vaginal hysterectomy: Mesh cohort: 2.8% of patients developed autoimmune disease. Control: 3.2% of patients developed SAID. Adjusted OR 0.78; 95% CI (0.48–1.26) |
| N = 40 (18 hernia, 4 TVT and 18 TVM) | Inclusion: Patients presenting to several autoimmune clinics, who had previously implanted polypropylene mesh | 49.5 (range 28–75) | Not explicitly described, symptoms of autoimmune disease were recorded at presentation at the clinic, including whether a patient had a PP implant | 18 (45%) of patients were diagnosed with autoimmune disease |
| N = 88 947 mesh surgery N = 3389 non-mesh surgery | Inclusion: Women having first time urinary incontinence surgery with mesh For control: women having first urinary incontinence surgery without mesh | Mesh: 53.1 ± 12* Control: 52.2 ± 12* | Mesh group: 8.7 (6.8–8.7)¥ years Non-mesh group: 9.9 (7.4–9.9)¥ years | Cumulative incidence of autoimmune disease, fibromyalgia or myalgic encephalomyelitis Mesh group: 8.1% at 10 years Non-mesh group: 9% at 10 years |
The studies have been conducted between 2008 and 2019. When combining the three cohort studies eligible for meta-analysis, a total of 104,594 matched participants had PP implants, because of inguinal hernia repair, mesh for POP or SUI. These participants were matched with 33,253 controls.
The control groups in the studies of Chughtai et al. [
15,
16] were extracted from a cohort of colonoscopy patients and a second cohort of patients with a history of vaginal hysterectomy. All subjects in the cohort studies were individually matched by patient characteristics and comorbidities. The women included in the study of Muller et al. [
17] were all women who had SUI surgery either with or without mesh. Patient characteristics, demographic data and comorbidities were comparable.
The study of Cohen Tervaert [
9], being a case series, contained a sample of 714 participants, of whom 40 had a PP mesh implant.
The primary outcome of the studies by Chughtai et al. [
15,
16] was the development of systemic autoimmune disorders (SAID) at the entire follow-up period. The average follow-up period of both studies by Chughtai et al., was 6 years. SAID was defined in one study as an enumeration of various autoimmune disorders (Grave’s disease, Hashimoto’s thyroiditis, pernicious anaemia, autoimmune haemolytic anaemia, autoimmune thrombocytopenic purpura, amyotrophic lateral sclerosis, multiple sclerosis, Guillain–Barré Syndrome, myasthenia gravis, Goodpasture syndrome, vasculitis, celiac disease, pemphigus vulgaris, systemic lupus erythematosus, systemic sclerosis, Sjogren’s syndrome, dermatomyositis, polymyositis, rheumatoid arthritis, ankylosing spondylitis and fibromyalgia) [
16]. Secondary outcomes included development of SAID at 6-month, 1-year and 2-year follow-up time point.
The main outcome measure of the study of Muller et al. [
17] was the first post-operative admission with a record of at least one of 29 autoimmune diseases, fibromyalgia or myalgic encephalomyelitis. Inclusion commenced in 2006 and the study was closed after a minimum follow-up of 5 years for all patients, with a maximum follow-up period of 10 years.
The study of Cohen Tervaert reported on symptoms suggestive of a (systemic) autoimmune disease in the presence of a PP mesh [
9]. Autoimmune disease in the presence of a PP implant in this study was defined as fulfilment of the criteria for the diagnosis of autoimmune/inflammatory syndrome induced by adjuvants (ASIA), Shoenfeld’s criteria [
18]. (Appendix B). Data were collected at presentation to an autoimmune clinic; no further follow-up was described.
Outcomes
Chughtai et al. performed two retrospective cohort studies with matched controls [
15,
16]. In one study subjects were males who had undergone an inguinal hernia repair with mesh, the other study included women with a POP repair with mesh. The source of patients was the New York State Department of Health Statewide Planning and Research Cooperative System (SPARCS) [
19].
The male subjects who were included in the mesh for herniorrhaphy study were matched with a control cohort, consisting of patients undergoing colonoscopy. Controls were excluded if they had a history of mesh-related procedures, a diagnosis of colorectal carcinoma within one month of the colonoscopy or a previous diagnosis of SAID.
In total 12,716 men with a history of (mesh) herniorrhaphy were matched with 25,432 patients who had a colonoscopy. SAID was diagnosed in 188 (1.5%) in the mesh group. In the control group, 413 patients (1.6%) had developed SIAD at the end of follow-up. The adjusted OR was 0.91 (95% CI 0.76–1.09). After matching, the authors concluded that inguinal mesh hernia repair was not associated with the development of SAID [
16].
The women who have been enrolled in the POP with mesh repair study were matched with two cohorts of controls: a surgical and a non-surgical cohort. Controls were either women with a vaginal hysterectomy in their medical history for benign gynecological or urogynecological conditions (surgical cohort) or women who had an indication for a screening colonoscopy (non-surgical cohort).
Two thousand one hundred two women with a mesh-repair for POP were included. These were matched with 37,298 women in the non-surgical control cohort and 7338 women in the surgical control cohort. This resulted in 1507 women with mesh repair matched with 3014 colonoscopy patients and 1375 women with mesh repair matched with 1375 women with vaginal hysterectomy.
Subjects with a (concurrent) history of autoimmune disease, malignancy, mesh-related procedures or prior pelvic floor surgery were excluded. An additional exclusion criterion for the non-surgical cohort was inflammatory bowel disease. In the surgical control cohort women with endometrial hyperplasia with atypia, abnormal vaginal bleeding or benign ovarian pathology were excluded.
In total SAID was diagnosed in 59 women (2.8%) after prolapse with mesh repair, in 1060 women (2.8%) after colonoscopy and in 235 women (3.2%) who had a history of vaginal hysterectomy.
After individual matching by demographics, date of the procedure and comorbidities, no increased risk of developing SAID after a mesh implantation for POP was found. The adjusted OR was 0.91 (95% CI 0.62–1.34) when comparing to the colonoscopy group and 0.78 (95% CI 0.48–1.26) when comparing to the vaginal hysterectomy group.
Muller et al. [
17] performed a national cohort study to compare the incidence of SAID in women having SUI surgery with and without mesh. Patients who had SUI surgery in the English NHS between 2006 and 2013 were included from an administrative database called the Hospital Episode Statistics.
Women were excluded if they had a record of SUI surgery in the previous 3 years or had a history of autoimmune disease, fibromyalgia or myalgic encephalomyelitis within this timeframe.
In total 88,947 women with mesh surgery and 3389 women without mesh surgery for SUI were included. The cumulative incidence of autoimmune disease, fibromyalgia or myalgic encephalomyelitis was 8.1% (95% CI 7.9–8.3%) in the mesh cohort and 9.0% (95% CI 8.0–10.1%) in the control group. The adjusted HR was 0.89 (95% CI 0.79–1.01; p = 0.07).
This study did not demonstrate an increased risk of systemic disease after mesh implantation for SUI.
Finally, we included the study of Cohen Tervaert [
9]. This study described 40 patients with systemic complaints in the presence of a PP mesh implant, who were selected out of a cohort of 714 patients who presented to the authors’ autoimmune clinic. Patients were classified as suffering from autoimmune/inflammatory syndrome induced by adjuvants (ASIA syndrome) when they fulfilled Shoenfeld’s criteria (Appendix B) [
18]
The author described that in 24 out of 40 of the included patients their symptoms started within 1 year after mesh implantation. Ten out of 40 subjects developed ASIA between 1 and 3 years after the implantation of PP, and in 6 patients these symptoms developed later than 3 year after PP implantation. Eighteen out of 40 patients were diagnosed with an International Classification of Diseases (ICD) coded autoimmune disease.