How are you taking part in this consultation?

You will not be able to change how you comment later.

You must be signed in to answer questions

    The content on this page is not current guidance and is only for the purposes of the consultation process.

    Interventional procedure overview of laparoscopic insertion of an inactive implant for gastro-oesophageal reflux disease

    Closed for comments This consultation ended on at   Request commenting lead permission

    Evidence summary

    Population and studies description

    This interventional procedures overview is based on a total of around 130 people from 3 studies across 6 papers. The studies include 1 prospective observational study (Bjelovic 2020) and 2 retrospective cohort studies (Feka 2024 and Fringeli 2024c). The remaining 3 articles include 2 retrospective chart reviews associated with Fringeli (2024c), (Fringeli, 2024a and 2024b; further analysis on subgroups within the sample in Fringeli [2024c]) and 1 follow-up article from the Bjelovic (2024) study (Harsanyi 2024). All participants in all studies had the procedure. This is a rapid review of the literature, and a flow chart of the complete selection process is shown in figure 1. This overview presents 3 studies as the key evidence in table 2 and table 3.

    The prospective study was a single-arm, multicentre study of 50 people by Bjelovic (2020). This was based in Hungary, Serbia and Switzerland. Follow-up time was 6 weeks, 3 months, 6 months and 12 months. The aim of the study was to assess the safety and effectiveness of the procedure in those experiencing GORD, with small HH and no IOM. It was also to obtain CE marking. The 4-year follow-up results are reported in Harsanyi (2024). There are 2 unpublished papers for the 3-year and 5-year follow-up results. These manuscripts have not yet been accepted for publication so cannot be included in this overview.

    The first retrospective observational single-arm study of 40 people was published by Feka (2024). This is a multicentre trial from 2 hospitals, one based in the UK and the other in Austria. Follow-up time was 3 months. The aim of this study was to assess safety and effectiveness in those experiencing GORD and who have IOM.

    The second retrospective cohort study and 2 retrospective chart reviews by Fringeli (2024a, 2024b, 2024c) were based in Switzerland. These studies include a significant overlap of people, but with a variation in follow-up length and clinical characteristics.

    The retrospective cohort study by Fringeli (2024c) included 40 people who were followed up for 3 months. The study included those with HH less than 10 cm in size. Some people (77.5%) had IOM but not all. Fringeli (2024c) reports the initial overall analysis of this study. Following this, further analysis was carried out in 2 subgroups of 20 people and 30 people in 2 retrospective chart reviews (Fringeli 2024a, 2024b).

    The Fringeli (2024a) study is a retrospective chart review that included 20 people who had IOM and a 12-month follow-up. The population included in this review were also part of the Fringeli (2024c) study. The aim of the review was to assess effectiveness and safety outcomes in this group of people.

    The second retrospective chart review included 30 people (Fringeli 2024b). The aim of the study was to assess the safety and effectiveness of the procedure in those with large HH (between 4 cm and 10 cm in size). Follow-up time was up to 6 months after the procedure. They almost entirely overlap with those in the Fringeli (2024c) study as well as some in the Fringeli (2024a) study.

    Table 2 presents study details.

    Figure 1 Flow chart of study selection

    Table 2 Study details

    Study no.

    First author, date

    country

    Characteristics of people in the study (as reported by the study)

    Study design

    Inclusion criteria

    Intervention

    Follow-up

    1

    Bjelovic, 2020

    Hungary, Serbia, Switzerland

    N=50

    (n=47 at 1-year follow-up)

    Mean age: 51.5 years (SD 11.8).

    Female: 22 (44%)

    Prospective, single-arm, multicentre study

    (Surgery performed at 4 hospitals)

    • People aged 18 to 75

    • Documented typical GORD symptoms, present for more than 6 months, that respond to PPIs

    • Requires daily PPI

    • Subject has a 24-hour pH monitoring proven GORD performed while off any anti-reflux medication or after discontinuation for at least 7 days before testing

    • Total distal oesophageal pH must be more than 4 for more than 4.5% of the time during a 24-hour monitoring

    Exclusion criteria include:

    • A history of gastroesophageal surgery

    • HH larger than 3 cm

    • Oesophageal dysmotility disorder (IOM)

    • Oesophagitis grade C or D, according to the Los Angeles classification. The Los Angeles classification system is a method for diagnosing and grading reflux oesophagitis. There are 4 grades, A to D, with grade A indicating the most mild and grade D indicating the most severe oesophagitis.

    • BMI over 35 kg/m2

    RefluxStop device

    6 weeks,

    3 months,

    6 months,

    12 months

    2

    Harsanyi, 2024

    (follow-up of Bjelovic 2020)

    N=44 at 4 years

    Baseline characteristics same as described in Bjelovic (2020)

    Prospective, single-arm, multicentre study

    (Surgery performed at 4 hospitals)

    Inclusion criteria same as described in Bjelovic (2020)

    RefluxStop device

    Effectiveness results: 4 years.

    Safety results: years 1 to 4 and at any additional visits

    3

    Feka, 2024

    Austria and UK

    N=40

    Female: 15 (37.5%)

    Mean age: 48.93 years (SD 4.59)

    Retrospective, observational, single-arm study

    • People aged 18 years and older

    • More than 1 year of GORD

    • History of PPI usage

    • Have IOM (as diagnosed according to criteria according to Chicago classification v4.0). The Chicago classification version 4.0 is a system for classifying oesophageal motility disorders

    • GORD diagnosis via 24-hour impedance pH testing

    • BMI less than 35 m2/kg

    Exclusion criteria include:

    • History of oesophageal or gastric malignancies

    • Oesophageal strictures or stenosis

    RefluxStop device

    3 months

    4

    Fringeli, 2024c

    Switzerland

    N=40

    Female: 16 (40%)

    Mean age: 60 years (SD 51 to 71)

    Mean BMI (kg/m2): 26.3 (24.6 to 28.9)

    Retrospective single-arm cohort study

    Included all people who underwent surgery between May 2020 and April 2022. To be eligible for surgery people must:

    • be aged 18 years or older

    • with documented GORD or typical symptoms of GORD

    • with HH smaller than 10 cm

    • no previous gastric or oesophageal surgery

    • no long-segment Barrett's oesophagus

    RefluxStop device

    4 weeks,

    3 months

    5

    Fringeli, 2024a

    Switzerland

    (subgroup of Fringeli 2024c)

    N=20

    Female: 10 (50%)

    Mean age: 57.4 years (SD 12.6).

    Has IOM

    Mean BMI (kg/m2): 26.5 (SD 4.8)

    Retrospective chart review

    This review included the first cohort of people with the criteria below who had RefluxStop in the Fringeli (2024c) study.

    • those with 12-month follow-up data

    • aged 18 years or older

    • with documented GORD

    • concurrent with IOM (identified via video-oesophagram with inefficient or slow emptying of the oesophagus or on manometry with less than 70% contractile waves or an amplitude of less than 30 mmHg)

    Exclusion criteria include:

    • HH larger than 10 cm

    • Long-segment Barrett's oesophagus

    • History of oesophageal or gastric surgery

    RefluxStop device

    4 weeks,

    3 months,

    12 months

    6

    Fringeli, 2024b,

    Switzerland

    (subgroup of Fringeli 2024c)

    N=30

    Female: 14 (46.7%)

    Mean age: 61 years (SD 15)

    Has HH between 4 cm and 10 cm

    Mean BMI (kg/m2): 26.8 (SD 4.3)

    Retrospective chart review

    This review included people with the criteria below, many of whom had RefluxStop in the Fringeli (2024c) study.

    • Those with 6-month follow-up data

    • Aged 18 years and older

    • With documented GORD

    • Require daily PPI

    • Diagnosed with a large HH (defined as an axial hernia length of 4 cm to 10 cm)

    • Diagnosis of HH was made preoperatively by gastroscopy, high-resolution oesophageal manometry or video-oesophagram

    Exclusion criteria include:

    • History of oesophageal or gastric surgery

    • Long-segment Barrett's oesophagus

    • HH larger than 10 cm

    RefluxStop device

    Day 1,

    4 weeks,

    3 months

    6 months

    Table 3 Study outcomes

    First author, date

    Efficacy outcomes

    Safety outcomes

    Bjelovic, 2020

    Primary efficacy outcome:

    GERD-HRQL scores (0 to 50 points)

    Baseline: 28.8 (SD 7.3)

    6 months: 3.4 (SD 6.0; p<0.0001) – improvement of 88%. 45 out of 47 people had more than 50% improvement of GERD-HRQL total score at 6 months.

    12 months: 44 people had more than 50% improvement in GORD symptoms compared with baseline. 3 people had less than 50% improvement.

    Secondary efficacy outcomes:

    Median heartburn subscores:

    Median baseline: 4.0 (range 0 to 5)

    Median 12 months: not provided

    Daily regurgitation (Foregut questionnaire):

    Baseline: 88% (out of 50 people) had daily regurgitation

    12 months: 97.8% (out of 47 people) had no or minimal occasional episodes of regurgitation.

    Statistically significant reduction (p<0.001).

    Changes in dysphagia

    Baseline: 15 (30%) people had dysphagia (n=50)

    6 months: 11 (22%) people had resolved dysphagia. 4 (9%) people continued minimal to moderate dysphagia (p<0.001; n=47)

    12 months: 2 (4%) people reported minimal dysphagia (p<0.001; n=47)

    Odynophagia

    Baseline: 13 (26%) people had odynophagia (n=50)

    6 months: 0 (0%) people (p<0.001; n=47)

    12 months: 1 (2%) person (p<0.001; n=47)

    PPI medication

    Baseline: Everyone took daily PPI medication

    6 months: 0 people took PPI.

    12 months: 1 person took PPIs (device position was too low)

    24-hour pH monitoring

    This was measured as the mean percentage of overall time (within 24 hours) with a pH less than 4.

    Baseline: 16.35%

    6 months: 0.8% (p<0.001). 98% of people had a normal 24-hour pH result.

    This is a 95% improvement of mean value.

    Gas bloating (those with a score above 2 in the related GERD-HRQL survey):

    Baseline: 84%, median: 4.0 (0 to 5)

    12 months: 19% (p<0.0001), Median: not provided

    At 12 months, gas bloating:

    Disappeared: 30 people

    Improved: 7 people

    Unchanged: 2 people

    Worsened: 0 people

    Patient satisfaction relating to GORD:

    Baseline: 1 (2%) person was satisfied. 45 (90%) were dissatisfied

    6 months: 44 (94%) people were satisfied, 2 (4%) were dissatisfied

    12 months: 43 (91%) people were satisfied, 1 (2%) was dissatisfied. 2 (4%) further people were dissatisfied, and the authors suggest that this was unrelated to GORD due to their normal 24-hour pH monitoring results at 4-year follow-up.

    5 out of 6 people who did not have 4-year follow-up data were satisfied at their last visit. Further information was not provided.

    Primary safety outcome:

    No serious device-related adverse events

    No serious adverse events related to RefluxStop device reported during the 6-month and 1-year follow-up.

    6 serious procedure-related adverse events

    Occurred in 4 people:

    • Infection – included both mediastinal abscess and empyema, probably due to the infected mediastinal haematoma

    • Bleeding

    • Release of fundoplication sutures occurred in 1 person at 6 months. They had a second operation which was successful.

    Secondary safety outcome:

    No device-related adverse events

    No non-serious adverse events related to RefluxStop device reported during the 6-month and 1-year follow-up. No device deficiencies and no removal of device.

    Procedure-related adverse events

    Postoperative dysphagia

    No new cases of dysphagia or odynophagia at 6 months or 12 months

    Harsanyi, 2024

    (Follow-up from Bjelovic 2020 paper)

    At 4 years, follow-up data were available in 44 out of 50 people. In addition, they reported outcomes from 3 people at 3 years carried forward.

    Primary efficacy outcome:

    Median GERD-HRQL score (IQR) (0 to 50 points):

    Baseline: 29.5 (24 to 33)

    4 years: 3.0 (0 to 9.2)

    This is a 90% reduction compared with baseline.

    Secondary efficacy outcome:

    Daily regurgitation (Foregut questionnaire):

    Baseline: 43 out of 50 (86%) people had daily regurgitation.

    At 4 years they report the inverse: 42 out of 44 (95.5%) had no or minimal regurgitation. 3 people with missing 4-year follow-up data reported no regurgitation at 3 years. So 45 out of 47 had no or minimal regurgitation. Everyone experienced improvement in regurgitation.

    Changes in dysphagia

    1 report of dysphagia:

    Baseline: 1 person reported severe dysphagia (score 5)

    3 years: Same person reported mild dysphagia at 3-year follow-up (score 2). Their dysphagia score was improved from baseline.

    46 out of 47 people reported no dysphagia.

    Changes in odynophagia

    0 reports of odynophagia

    PPI use:

    Baseline: Everyone used daily PPIs (n=50)

    4 years: 2 out of 44 (4.5%) people used daily PPIs. Both had normal 24-hour pH monitoring results.

    Gas bloating

    Baseline: (data not provided, but reported in Bjelovic 2020)

    4 years:

    • Disappeared: 30 (68%)

    • Improved: 11 (25%)

    • Unchanged: 2 (5%)

    • Worsened: 1 (2%)

    Patient satisfaction relating to GORD:

    Baseline: reported in Bjelovic 2020

    4 years: 2 out of 44 (4.5%) were dissatisfied. Both had normal 24-hour pH monitoring results, suggesting that GORD was unlikely to be the reason for dissatisfaction

    Missing data:

    • 1 person died from COVID-19

    • 2 people missed 4-year follow-up

    • 1 person terminated within the first year of the study after a broken needle was left subcutaneously. This was removed under local anaesthesia. They terminated dissatisfied, with a high GERD-HRQL score, no regular daily PPI at 6 months and no regurgitation. They refused 24-hour pH testing.

    • 2 people terminated within the first year of study. Both were satisfied, did not take PPIs had a low average GERD-HRQL score of 2.5, no regurgitation.

    Primary safety outcomes:

    No serious device related adverse events

    • No adverse events related to RefluxStop device

    • No device deficiencies

    • No migration or erosion

    • No oesophageal dilation required

    2 serious procedure-related adverse events occurred following surgery which were reported in Bjelovic 2020:

    1. Infection with abscess

    2. Haematoma

    Both treated to resolution.

    No other serious adverse events were reported in the follow-up period.

    Postoperative dysphagia

    No new cases of dysphagia or odynophagia at 4 years.

    Secondary safety outcomes:

    No device related adverse event

    1 procedure related adverse event:

    Recurrence of acid reflux symptoms in 1 person who had a 24-hour pH of less than 4. Device was found to be positioned too low. This position was categorised as a 'failure risk' after surgery and at 4 year follow-up.

    Events relating to dysphagia and odynophagia are reported in the efficacy outcomes section.

    Feka, 2024

    (Austria and UK)

    Aim: assess safety and effectiveness in those with GORD and IOM

    Primary efficacy outcome:

    GERD-HRQL total score (they do not clarify which version of the questionnaire they used or range for scores)

    Baseline (SD): 32.83 (5.08)

    3 months (SD): 6.6 (3.71; p<0.001)

    Secondary efficacy outcomes:

    Mean heartburn subscore

    Baseline mean (SD): 17.4 (2.07)

    3 months mean (SD): 3.83 (1.88; p<0.001)

    Mean regurgitation subscore

    Baseline mean (SD): 10.85 (2.85)

    3 months mean (SD): 1.63 (1.68; p<0.001)

    Changes in dysphagia (%)

    Baseline: 4 (10%) people

    3 months: 1 (2.5%) people(p-value not provided)

    Percentage of people using PPIs daily

    Baseline: 31 (77.5%) people

    3 months: 5 (12.5%; p<0.001) people

    35 (87.5%) people no longer required PPI

    Patient satisfaction relating to GORD

    Baseline: 40 (100%) reported dissatisfied or neutral. 0 reported satisfied.

    3 months: 4 (10%) reported dissatisfied or neutral. 36 (90%) reported satisfied.

    Dissatisfaction reduction of 90%. P-value not provided.

    Operative practice

    Median hospital stay was 3 days

    There were 2 (5%) serious adverse events

    1. 1 serious adverse event occurred in 1 person and was classified as Clavien–Dindo 3b in surgical severity. The newly reconstructed His angle and RefluxStop device had reherniated into the thoracic cavity and required revisional surgery at the first postoperative day. During the laparoscopic revision, the device was removed and a Dor fundoplication was performed. They believe it was due to high BMI causing high intra-abdominal pressure.

    2. In 1 person, device migration occurred 1 month after surgery. They experienced epigastric pain for 3 days before the device was retrieved. The study suggests this could have been caused by a small haematoma where the device was implanted.

                Postoperative dysphagia

                No new cases of dysphagia at 3 months.

                Fringeli, 2024c

                Switzerland

                Aim: To assess safety and effectiveness outcomes

                Primary efficacy outcome:

                Feasibility of the procedure, measured by the proportion of people who had the device implanted in the correct position. Everyone had the device implanted in the correct position.

                Secondary efficacy outcomes:

                Median GERD-HRQL scores (0 to 75 points)

                (n=38)

                Baseline (IQR): 35 (28.5 to 49)

                3 months (IQR): 2 (0 to 3; p<0.0001)

                Heartburn subscore

                Statistically significant reduction (p<0.0001). Scores not provided (from figure 3)

                Regurgitation subscore

                Statistically significant reduction (p<0.0001). Scores not provided (from figure 3)

                Changes in dysphagia

                3 months: The 20 people who had previously suffered from dysphagia showed a reduction in severity or complete resolution of symptoms. No p-values provided.

                Patient satisfaction relating to GORD (n=38)

                Baseline: 28 (73.7%) reported either dissatisfaction or neutral

                3 months: 100% reported improvement

                No further figures or p-values provided.

                Operative practice

                Median operating time: 57.5 minutes (IQR, 51.75 to 64.25 minutes).

                Median hospital stay: 4 days (IQR 3 to 5).

                3 months: RefluxStop device was in correct location in everyone. No device-related complications were observed.

                Secondary outcomes:

                Operative practice

                Conversion to laparotomy, n (%): 1 (2.5%)

                Intraoperative complication, n (%): 1 (2.5%)

                6 postoperative complications within 3 months graded to Clavien–Dindo classification system:

                Grade 2: 1 (2.5%)

                Grade 3a: 3 (7.5%)

                Grade 3b: 2 (5%)

                Description of postoperative complications:

                1. 1 person underwent urgent laparoscopic reoperation the same day due to postoperative haemorrhage caused by dissection of the short gastric vessels at the fundus

                2. 1 person reported persistent fatigue due to pericardial effusion

                3. 1 person developed a trocar hernia in the epigastric area. They underwent direct open closure of the trocar hernia defect

                4. 3 people had dysphagia. Events relating to dysphagia are reported in the efficacy outcomes section.

                No device-related complications or re-operations during 4-week and 3-month follow-up.

                Postoperative dysphagia

                The author reports no new onset of dysphagia at 3 months. They also report that after surgery 3 people (7.5%) had severe, persistent dysphagia with frequent vomiting and inability to eat a normal diet. Due to the severity of the symptoms, early dilations were performed 3 to 4 weeks following surgery. These people had IOM prior to surgery and 2 had preoperative dysphagia. 1 was new onset after the follow-up period of 3 months.

                Due to the severity of the symptoms they performed early dilations 3 to 4 weeks after surgery. In 1 person 18 mm to 20 mm balloons were successful. The other 2 people required further endoscopic dilations with the endoflip balloon to 25 mm.

                Fringeli, 2024a,

                Switzerland

                Aim: To assess safety and effectiveness in those with IOM

                There is overlap in people and their respective outcomes in this article and those in the Fringeli (2024c and 2024b article)

                Mean GERD-HRQL score (SD) (0 to 75 points):

                Baseline SD: 40.7 (16.0)

                3 months: 4.8 (8.3; p<0.001)

                12 months: 5.7 (7.7; p<0.001)

                No worsening of GORD symptoms reported at 12-month follow-up in anyone.

                12-month outcomes categorised as:

                Excellent: 13 people (65%)

                Good: 2 people (10%)

                Fair: 3 people (15%)

                Poor: 2 people (10%)

                Of the 2 people with poor outcomes:

                • 1 person had recurrence of GORD within 12 months. They attributed this to weight gain of 10 kg

                • 1 person had persistence of symptoms, particularly heartburn. They could not identify a cause and was considered multifactorial in nature

                Mean heartburn subscores (SD):

                Baseline: 17.4 (8.3)

                12 months: 2.4 (3.7; p<0.001)

                Mean regurgitation subscores (SD):

                Baseline: 17.7 (7.5)

                12 months: 1.6 (3.5; p<0.001)

                Changes in dysphagia

                Changes in scores for the whole cohort (n=20) were not statistically significant.

                12 people experienced dysphagia before surgery.

                Mean baseline scores (SD): 2.7 (1.4)

                Mean 12-month scores (SD): 1.0 (1.0; p<0.005)

                PPI usage:

                16 people were able to completely discontinue use of PPIs

                Patient satisfaction relating to GORD

                Baseline: 25% satisfied

                12 months: 90% satisfied

                Operative practice

                RefluxStop procedure was feasible in all 20 people. In everyone, video-oesophagram confirmed the device was in the right location and adequate reduction in HH.

                • Median operating time: 59.5 minutes (IQR 50.25 to 64 minutes).

                • Median hospital stay: 3.5 days (IQR 3 to 4 days)

                • 1 person (5%) required a second operation during 12-month follow-up, due to persistent left-sided thoracic pain, which was considered unrelated to the device.

                • 1 person (5%) had a procedural complication. An implant that had split into 5 parts and had migrated from its initial position. It is thought this was due to insufficient surgical closure of the passageway where the deployment tool was used to hold the device in place during the procedure. It was not due to pressure-induced tissue damage, so it is considered a procedural complication rather than device malfunction.

                • No device related reoperations or complications at surgery during 12-month follow-up

                • For 1 person, the procedure was converted to open surgery due to adhesions and bleeding when trying to establish access laparoscopically. They had had previous open surgery. Once bleeding was controlled, the intended procedure was performed as described in Methods.

                Postoperative dysphagia

                • Among the 8 people without dysphagia before the procedure, 3 developed dysphagia after 12 months. Change in dysphagia scores for the 8 people were not statistically significant (p<0.075).

                • The 3 people (15%) who had dysphagia experienced persistent severe dysphagia requiring endoscopic dilation after surgery. 2 people had dysphagia before surgery and 1 person developed dysphagia after surgery. All 3 people had successful repeated endoscopic dilations with complete resolution of dysphagia.

                Fringeli, 2024b,

                Switzerland

                Aim: To assess safety and effectiveness in those with a larger HH (between

                4 to 10 cm).

                There is overlap in people and their respective outcomes in this article and those in the Fringeli (2024c and 2024a article)

                Primary outcomes:

                The RefluxStop procedure was feasible in everyone in the chart review.

                Secondary outcomes:

                Mean GERD-HRQL total scores (SD; 0 to 75 points)

                Mean baseline scores (SD): 37.6 (15.5)

                Mean 6-month scores (SD): 3.1 (5.4; p<0.001)

                6-month outcomes (n=29) categorised as:

                Excellent: 26 people (89.7%)

                Fair: 1 person (3.5%)

                Poor: 2 people (6.9%)

                Both people with poor results showed improvement in symptoms.

                Mean heartburn subscores (SD)

                Baseline: 16.5 (8.1)

                6 months: 1.1 (2.3; p<0.001)

                Mean regurgitation subscores (SD)

                Baseline:16.5 (7.2)

                6 months: 0.9 (2.0; p<0.001)

                Mean dysphagia subscores (SD)

                Changes in subscores for the whole cohort were not statistically significant. At baseline, 15 (50%) people had dysphagia. Among them, 14 (93.3%) people had IOM.

                Mean dysphagia subscores of those with dysphagia at baseline (n=15)

                Baseline: 2.7 (1.5)

                6 months: 0.6 (0.9; p<0.001)

                Patient satisfaction relating to GORD

                Baseline: 8 (27%) people were satisfied. 17 (57%) were dissatisfied

                6 months: 28 (97%) people were satisfied, 0 (0%) were dissatisfied

                Operative practice

                • Median operating time: 56 minutes (IQR 52 to 63).

                • Median hospital stay: 4 days (IQR 3 to 5)

                • Adequate reduction in HH and correct location of device in everyone (100%)

                Primary outcomes:

                Device related complications

                No device-related complications during 3- and 6-month follow-up.

                No device-related reoperations for anyone.

                Secondary outcomes:

                Postoperative complications within 90 days

                • 1 person experienced persistent fatigue at 4 weeks due to pericardial effusion.

                • 1 person had persistent dysphagia 12 weeks after operation and required endoscopic dilations

                Postoperative dysphagia

                Of the 15 people who did not have dysphagia at baseline, 10 had IOM (66.7%).

                New onset dysphagia occurred in 5 people. 4 people had no intervention and had changes in diet only.

                Recurrence of hiatus hernia

                3 months: No recurrence of HH

                6 months: 1 person had recurrence of HH (3.3%).

                The person had a large HH measuring 8 cm. It is thought to have been a consequence of severe vomiting from food poisoning. The upper third of the stomach and RefluxStop device had slipped through the hiatus. They received emergency surgery outside the institution to reposition the fundus and RefluxStop device.

                Operative practice

                • 29 out of 30 people: surgery was performed laparoscopically

                • 1 person had a procedure that was converted to open surgery, due to adhesions and bleeding. They had had open surgery previously

                Procedure technique

                All 6 papers detailed the procedure technique and devices used. There were no significant variations in technique or devices used.

                Only 1 device was used for this procedure (RefluxStop, Implantica, Zug, Switzerland).

                The procedure

                Hiatus hernia reduction

                In 5 papers, HH was treated first before implanting the device (Bjelovic 2020, Fringeli 2024a, 2024b, 2024c, Harsanyi 2024). Feka (2024) did not specify treatment of HH in their methods, but this step was implied. HH was reduced and the sac was excised (if present) in the studies by Bjelovic (2020) and Fringeli (2024a, 2024c). The HH was closed, avoiding compression of the oesophagus. Excessive fat pads at the angle between the LOS and the stomach were removed in all studies (Bjelovic 2020, Fringeli 2024a, 2024b, 2024c, Feka 2024, Harsanyi 2024).

                Implantation of device

                In all articles, everyone had the operation using a standardised surgical technique, similar to other laparoscopic anti-reflux procedures (LARs). All studies highlighted the importance of high placement of the device above the upper edge of the LES for successful treatment (Bjelovic 2020, Fringeli 2024a, 2024b, 2024c, Feka 2024, Harsanyi 2024).

                All studies dissected quite extensively around the oesophagus in the mediastinum and as high as possible. Fringeli (2024a) described that they aimed to have an intra-abdominal length of at least 4.5 cm with a small traction on the oesophagus. Similarly, Bjelovic (2020) described that optimal placement of the device is more than 1 time the device size above the upper edge of the LOS. In all studies, the device was introduced laparoscopically using a specially made trocar and deployment tool (Implantica, Zug, Switzerland) (Bjelovic 2020, Fringeli 2024a, 2024b, 2024c, Feka 2024, Harsanyi 2024).

                Efficacy

                GERD-HRQL questionnaire scores

                All 6 papers reported data on GERD-HRQL questionnaire scores. All studies showed reduction in GERD-HRQL scores from baseline to follow-up. A high GERD-HRQL score indicates worse GORD symptoms and a low GERD-HRQL score indicates minimal GORD symptoms. 5 out of 6 papers reported statistical significance in reduction. Only the 4-year follow-up study did not provide a p-value (Harsanyi 2024).

                In the prospective single-arm study of 50 people, mean GERD-HRQL scores were 28.8 (SD 7.3) at baseline, which decreased at 6 months follow-up to 3.4 (SD 6.0; p<0.0001) (Bjelovic 2020). In the follow-up study, the score had remained reduced at 4 years. At year 1, the median had been 29.5 (IQR 24 to 33) which decreased to 3.0 (IQR 0 to 9.2) by year 4. The p-value for the 4-year result was not provided (Harsanyi 2024). The GERD-HRQL questionnaire used in these 2 articles had a score range of 0 to 50.

                In the retrospective observational single-arm study of 40 people, mean GERD-HRQL scores decreased from 32.83 (SD 5.08) before surgery to 6.60 (SD 3.71) at 3 months (p<0.001; Feka 2024). The authors did not clarify the range or the version of the GERD-HRQL questionnaire they used.

                In the retrospective cohort study of 40 people, median GERD-HRQL score at baseline was 35 (IQR 28.5 to 49). At 3 months this decreased to 2 (IQR 0 to 3; p<0.0001). In both retrospective chart reviews there were statistically significant reductions in GERD-HRQL scores. In Fringeli (2024a) with 20 people with IOM, the mean baseline score was 40.7 (SD 16.0). This decreased to 4.8 (SD 8.3) at 3 months and to 5.7 (SD 7.7) at 12 months (p<0.001) for both. In Fringeli (2024b), made up of 30 people with larger HH, the mean baseline score was 37.6 (SD 15.5). At 6 months this reduced to 3.1 (SD 5.4). The GERD-HRQL questionnaire used in these 3 articles had a score range of 0 to 75. As previously discussed, populations in the 2024a and 2024b studies almost entirely overlap with the 2024c study. All results were statistically significant.

                Heartburn – subscore of GERD-HRQL questionnaire

                All 6 papers measured changes in heartburn symptoms using subscores of components of the GERD-HRQL questionnaire, but not all studies reported the heartburn scores separately.

                In the retrospective observational single-arm study of 40 people, mean heartburn scores decreased from 17.4 (SD 2.07) before surgery to 3.83 (SD 1.88) at 3 months (p<0.001; Feka 2024).

                In a retrospective chart review of 20 people (with IOM), the mean baseline score was 17.4 (SD 8.3). At 12 months, this decreased to 2.4 (SD 3.7; p<0.001) (Fringeli 2024a).

                In Fringeli (2024b) the retrospective chart review of 30 people (with larger HH) had a baseline heartburn score of 16.5 (SD 8.1). At 6 months, this had reduced to 1.1 (SD 2.3; p<0.001).

                In the retrospective cohort study with 40 people, heartburn scores were not provided, but they noted that the reduction was statistically significant (p<0.0001; Fringeli 2024c). There was significant overlap between populations in the 2024a, 2024b and 2024c studies.

                In the prospective single-arm study of 50 people, by Bjelovic (2020) and 4-year follow-up study (Harsanyi 2024) heartburn scores were not provided.

                Regurgitation – subscore of GERD-HRQL questionnaire

                All 6 papers reported changes in regurgitation symptoms using subscores of components of the GERD-HRQL questionnaire.

                In the prospective single-arm study of 50 people, 88% of participants experienced daily regurgitation. At 12 months, 98% of people had no or minimal regurgitation episodes (p<0.001; Bjelovic 2020). In the 4-year follow-up study, 42 out of 44 participants (96%) had no or minimal regurgitation (Harsanyi 2024).

                In the retrospective observational single-arm trial of 40 people, mean regurgitation scores decreased from 10.85 (SD 2.85) before surgery to 1.63 (SD 1.68) at 3 months (p<0.001; Feka 2024).

                In a retrospective chart review of 20 people (with IOM), the mean baseline score was 17.7 (SD 7.5). At 12 months, this decreased to 1.6 (SD 3.5; p<0.001; Fringeli 2024a).

                In Fringeli (2024b), the retrospective chart review of 30 people (with larger HH), had a baseline regurgitation score of 16.5 (SD 7.2). At 6 months, this had reduced to 0.9 (SD 2.0; p<0.001).

                In the retrospective cohort study, regurgitation scores were not provided, but they noted that the reduction was statistically significant (p<0.0001; Fringeli 2024c). There was significant overlap between subjects in the 2024a, 2024b and 2024c studies.

                Changes in dysphagia (difficulty swallowing)

                All 6 papers measured changes in dysphagia. This section reports changes in symptoms in those who had preoperative dysphagia. New onset and postoperative dysphagia are described in the safety section below.

                In a prospective single-arm study of 50 people, 15 (30%) people experienced dysphagia. At 6 months, 4 (9%) people reported minimal to moderate dysphagia (p<0.001). At 12 months, 2 (4%) people reported minimal dysphagia (p<0.001) (Bjelovic 2020). In the follow-up study, 1 person reported mild dysphagia (a score of 2 in the GERD-HRQL questionnaire) at 3-year follow-up. This was an improvement from their baseline score of 5 (severe). At 4 years 46 out of 47 people reported no dysphagia (Harsanyi, 2024).

                In the retrospective observational single-arm trial of 40 people, 4 (10%) people had dysphagia at baseline. At 3 months, 1 (2.5%) person had dysphagia. A p-value was not provided (Feka 2024).

                In the retrospective cohort study with 40 people, 20 out of 40 (50%) reported preoperative dysphagia (Fringeli 2024c). At 3 months, there was a reduction in severity or complete resolution of symptoms in those who had dysphagia. A p-value was not provided. 2 of those who reported preoperative dysphagia required treatment (dilations) after the procedure. They also had IOM. Due to the severity of the symptoms, early dilations were performed 3 to 4 weeks after surgery (Fringeli 2024c).

                In the retrospective chart review of 20 people (with IOM), changes in dysphagia scores for the whole cohort were not statistically significant. In the 12 people who had dysphagia before surgery the mean dysphagia score reduced from 2.7 (SD 1.4) at baseline to 1.0 (SD 1.0) at 12 months (p<0.005) (Fringeli 2024a).

                In the retrospective chart review of 30 people (with larger HH), changes in the whole cohort were not statistically significant. 15 out of 30 (50%) people had dysphagia at baseline, with a mean dysphagia score of 2.7 (SD 1.5). At 6 months, this had reduced to 0.6 (SD 0.9; p<0.001).

                There was significant overlap between subjects in the Fringeli 2024a, 2024b and 2024c studies.

                Odynophagia (pain swallowing)

                Changes in odynophagia were reported in 2 papers (1 study) (Bjelovic 2020, Harsanyi 2024).

                At baseline 13 out of 50 (26%) of people had odynophagia. At 6 months, no-one reported odynophagia and at 12 months 1 person (2%) reported odynophagia (p<0.001 for both time periods; Bjelovic 2020). In the 4-year follow-up, 0 out of 50 people reported odynophagia between the years of 1 and 4 (Harsanyi 2024). Of the 50, 3 people did not have 4-year follow-up data so their 3-year data was carried forward (Harsanyi, 2024).

                The Harsanyi (2024) paper reports 2 events of odynophagia that were not reported in the Bjelovic (2020) paper. 1 person complained of odynophagia 1 day after surgery. At 6 months they were verified to have a motility disorder. For the second person with odynophagia, symptoms disappeared within a few months after surgery (Harsanyi 2024).

                No other papers reported odynophagia outcomes.

                Proton pump inhibitor (PPI) usage

                All 6 papers measured changes in usage of proton pump inhibitor medication before and after the procedure.

                In the prospective single-arm study of 50 people, everyone took daily PPI medication preoperation. At 6 months, 0 people took PPI and at 12 months 1 person was taking PPIs. In the person who started taking PPIs at 12 months, the device had been implanted too low and had been at risk of failure surgically (Bjelovic 2020). In the 4-year follow-up study, 2 out of 44 (5%) people were using daily PPIs. Both had normal 24-hour pH monitoring results (Harsanyi 2024).

                In the retrospective observational single-arm trial of 40 people, 31 (78%) people took daily PPIs. At 3 months, this reduced to 5 (13%) people (p<0.001) (Feka 2024).

                In a retrospective chart review of 20 people, at baseline everyone used PPI. At 12 months, 16 people were able to completely discontinue use of PPIs (Fringeli 2024a).

                The 2 other studies by Fringeli (2024b and 2024c) did not report changes in PPI usage.

                24-hour pH monitoring

                1 paper reported changes in 24-hour pH monitoring scores. This was measured as the mean percentage of overall time (within 24 hours) with a pH less than 4.

                In the prospective single-arm study of 50 people, at baseline the mean percentage of time with a pH less than 4 was 16%. At 6 months, this reduced to 1% (p<0.001). 98% of people had a normal 24-hour pH result. This was a 95% improvement of the mean value (Bjelovic 2020).

                In the 4-year follow-up study, 24-hour pH monitoring only occurred in those whose operation was considered at risk of failure. Potential failure operations included those in people who were dissatisfied with their quality of life relating to GORD, taking daily PPIs or had GERD-HRQL questionnaire results with less than 50% improvement. Of those who had an operation that was at risk of failure, 1 person had a pathologic pH test result. The paper considers this an objectified failed operation and is thought to be due to the device being positioned too low (Harsanyi 2024).

                No other papers reported changes in 24-hour pH monitoring.

                Gas bloating

                2 papers reported changes in gas bloating (Bjelovic 2020 and Harsanyi 2024). Both report percentages of people who experienced gas or bloating as identified through the GERD-HRQL questionnaire.

                In the prospective single-arm study of 50 people, 84% of people experienced gas bloating (a median score of 4.0, range 0 to 5). At 12 months, this had reduced to 19% of people (p<0.0001). After 12 months, gas bloating had disappeared in 30 people, improved in 7 people, remained unchanged in 2 people and did not worsen in anyone (Bjelovic 2020).

                In the 4-year follow-up study, gas bloating had disappeared in 30 (68%) people, improved in 11 (25%) people, remained unchanged in 2 (5%) people and worsened in 1 (2%) person (Harsanyi 2024).

                Patient satisfaction relating to GORD – score of additional question in GERD-HRQL questionnaire

                All 6 papers measured patient satisfactionrelating to GORD by adding a question to the GERD-HRQL questionnaire. The question was 'How satisfied are you with your current quality of life related to GERD?'. Possible responses included 'satisfied', 'neutral' or 'dissatisfied'.

                In the prospective single-arm study of 50 people, baseline results showed that 1 (2%) person was satisfied and 45 (90%) were dissatisfied with their quality of life related to GORD. At 6 months, 44 (94%) people were satisfied and 2 (4%) were dissatisfied. At 12 months, 43 (91%) were satisfied, 1 (2%) was dissatisfied (Bjelovic 2020). 2 (4%) more people were dissatisfied for reasons that Bjelovic (2020) have suggested are not related to GORD as they had normal 24-hour pH monitoring results. In the 4-year follow-up study, 2 out of 44 (5%) people were dissatisfied due to reasons unrelated to GORD. 42 out of 44 (96%) people were either satisfied or neutral at 4 years after the procedure (Harsanyi 2024).

                In the retrospective observational single-arm trial of 40 people, all subjects were either dissatisfied or neutral about their quality of life relating to GORD. No-one was satisfied. At 3 months post-procedure, this had decreased to 4 (10%) people who were dissatisfied or neutral. 36 (90%) of people were satisfied. This is a dissatisfaction reduction of 90%. P-values were not provided (Feka 2024).

                In the retrospective cohort study of 40 people, 28 (74%) reported dissatisfaction or neutrality when asked about their quality of life relating to GORD. At 3 months, 100% reported improvement and no-one was dissatisfied. No further information or p-values were provided (Fringeli 2024c).

                In the retrospective chart review of 20 people (with IOM), 25% of people were satisfied preoperation, compared with 90% satisfied at 12-month follow-up (Fringeli 2024a).

                In the retrospective chart review of 30 people (with larger HH) had 8 (27%) satisfied people compared with 28 (97%) satisfaction at 6 months (Fringeli 2024b).

                There was significant overlap between subjects in the 2024a, 2024b and 2024c studies.

                Safety

                Postoperative dysphagia

                This section describes postoperative dysphagia. Most studies describe this as new onset dysphagia which developed following the procedure (Bjelovic 2020, Harsanyi 2024, Feka 2024, Fringeli 2024a, Fringeli 2024b). 1 study was not clear about whether the postoperative dysphagia was new onset or if the participants experienced dysphagia before the operation (Fringeli 2024c).

                In the retrospective cohort study with 40 people, the authors report that there was no new onset of dysphagia recorded after 3 months. They report that 3 people had postoperative dysphagia requiring dilations (see section 'Changes in dysphagia'). All 3 people had IOM and 2 of them had experienced dysphagia preoperatively (Fringeli, 2024c). The authors do not clarify whether the third person had preoperative dysphagia, or if they developed new onset dysphagia postoperatively. Due to the severity of the symptoms, early dilations were performed 3 to 4 weeks after surgery (Fringeli, 2024c).

                In the retrospective chart review of 20 people (with IOM), there were 8 people who did not report preoperative dysphagia. Of those people, 3 developed dysphagia after 12 months following the procedure. Change in dysphagia scores for these 8 people were not statistically significant (Fringeli 2024a).

                In the retrospective chart review of 30 people (with larger HH), new-onset dysphagia occurred in 5 people, 4 of whom had no intervention and only changes in diet (Fringeli 2024b).

                There was significant overlap between participants in the Fringeli 2024a, 2024b and 2024c studies.  

                There were no new cases of dysphagia after surgery in the prospective single-arm study of 50 people at 12 months, or during the 4-year follow-up in the subsequent article (Bjelovic 2020 and Harsanyi 2024). There were no new cases of dysphagia in the retrospective observational single-arm trial of 40 people (Feka 2024).

                Recurrence of HH

                Recurrence of HH occurred in 2 people in 2 different studies.

                In the retrospective observational single-arm study of 40 people, 1 person had recurrence of HH where the newly reconstructed oesophageal and stomach angle and the device had herniated into the thoracic cavity. This was classified as a Clavien–Dindo grade 3b. They had a second operation the day after where the device was removed and they had a Dor fundoplication (Feka 2024).

                In the retrospective chart review of 30 people (with larger HH), 1 person had recurrence of HH. Their pre-existing hernia was 8 cm in size. The reherniation occurred after the person experienced severe vomiting caused by food poisoning, resulting in the upper third of the stomach and device to enter the thoracic cavity. They required emergency surgery to reposition the stomach and device (Fringeli 2024b).

                Device migration

                Device migration occurred in 2 people in 2 different studies (Feka 2024 and Fringeli 2024a).In the retrospective observational single-arm study of 40 people, device migration occurred in 1 person 1 month after the operation. The device entered the stomach cavity and left the body naturally. They experienced epigastric pain for 3 days before the device was removed. The paper suggests that it could have been caused by a small haematoma where the device had been implanted (Feka 2024). In the retrospective chart review of 20 people, 1 person experienced device migration. The device did not migrate into the stomach cavity. The device had split into 5 parts: 3 parts were sitting close to the stomach and 2 parts were close to the spleen and greater omentum. All 5 parts were recovered by diagnostic laparoscopy. It is thought the migration was due to insufficient closure of the passageway where the deployment tool was used to hold the device in place during the operation (Fringeli 2024a).

                Haemorrhage or bleeding (haematoma)

                2 people experienced postoperative bleeding in 2 different studies (Bjelovic 2020 and Fringeli 2024c).

                In the prospective single-arm study of 50 people, 1 person experienced bleeding which was categorised as a severe adverse event. They believed it was probably caused by the short gastric vessels. The person had clot evacuation and drainage (Bjelovic 2020).

                In the retrospective cohort study of 40 people, 1 person experienced a haemorrhage caused by dissection of the short gastric vessels at the top of the stomach (fundus). This required urgent laparoscopic operation on the same day (Fringeli 2024c).

                Infection

                Infection occurred in 1 person in the prospective single-arm study of 50 people (Bjelovic 2020). The infection occurred 3 times in the same person and resulted in both a mediastinal abscess and empyema. The authors report this was probably caused by an infected mediastinal haematoma. The infection did not spread to the pouch with the device. Once the subject had healed from the infection the study claim they had an excellent treatment result (Bjelovic 2020).

                Recurrence of acid reflux symptoms

                1 person in 1 study experienced recurrence of acid reflux symptoms as verified by a pathological 24-hour pH monitoring result. The device was found to be positioned too low and was categorised as a failed operation during follow-up (Harsanyi 2024).

                Release of sutures

                In the prospective single-arm study of 50 people, fundoplication sutures released in 1 person between the 6 and 12 month visit. This was classified as a serious adverse event. Reoperation was performed and they had a successful 12-month follow-up (Bjelovic 2020).

                Persistent left-sided thoracic pain

                1 person in 1 study experienced persistent left-sided thoracic pain at 12 months. It required a second operation. The pain was considered unrelated to the device (Fringeli 2024a).

                Persistent fatigue due to pericardial effusion

                There were 2 reports of a person experiencing persistent fatigue due to pericardial effusion. This was reported in the retrospective cohort study by Fringeli (2024c) and the retrospective chart review by Fringeli (2024b). It is unclear if this is the same event reported twice or 2 separate events.

                Incisional hernia

                1 person experienced a hernia at the insertion site for laparoscopic surgery (trocar hernia) (Fringeli 2024c). 1 person experienced an incisional hernia twice (Bjelovic 2020).

                Death

                1 person died in 1 study due to COVID-19 (Harsanyi 2024).

                Other adverse events

                The prospective single-arm study of 50 people further reported 1 moderate adverse event of pleuritis in 1 person, and 4 mild adverse events including the removal of part of a broken needle that had been left subcutaneously in 1 person, 1 accidental intraoperative hepatic lesion, 1 postoperative delayed gastrointestinal paralysis for 1 day and 1 procedural pneumothorax (Bjelovic 2020).

                Anecdotal and theoretical adverse events

                Expert advice was sought from consultants who have been nominated or ratified by their professional society or royal college. They were asked if they knew of any other adverse events for this procedure that they had heard about (anecdotal), which were not reported in the literature. They were also asked if they thought there were other adverse events that might possibly occur, even if they had never happened (theoretical).

                They listed the following anecdotal and theoretical adverse events:

                • silicone allergies or induced hypersensitivity

                • foreign body reaction

                • erosion risk

                • oesophagus dilation.

                6 professional expert questionnaires for this procedure were submitted. Full details of what the professional experts said about the procedure are in the specialist advice questionnaires for this procedure.

                Validity and generalisability

                • Evidence is mainly from small studies, with each study having 50 people or fewer. Sample size calculations were reported and sufficient numbers of people participated in 1 study (Bjelovic 2020). The remaining 4 papers (2 studies) did not report sample size calculations or said that they did not calculate them (Feka 2024, Fringeli 2024a, 2024b, 2024c).

                • In the 4-year follow-up study, there was missing data for 6 out of 50 people. The sample size calculation (reported in Bjelovic 2020) mentioned a sample size of 45 was necessary. Of those who terminated the trial early:

                  • 1 person died from COVID-19

                  • 2 people had 3-year follow-up data which showed they were satisfied, not low GERD-HRQL scores (average 1.3), no PPI usage and no regurgitation

                  • 2 people left the study within the first year. Both were satisfied, did not take PPIs, had a low GERD-HRQL score (2.5) and no regurgitation

                  • 1 person had a broken needle left under the skin that required removal under local anaesthesia. They were dissatisfied, had a high GERD-HRQL score, no PPI usage and no regurgitation.

                • Of the 3 studies, 1 was based in Hungary, Serbia and Switzerland, another was based in Switzerland, and 1 was based in Austria and the UK.

                • The retrospective cohort study of 40 people had significant overlap in sample with those in the retrospective chart reviews (Fringeli 2024a, 2024b, 2024c). The retrospective chart review of 20 people (Fringeli 2024a) is entirely a subgroup of the retrospective cohort study (Fringeli 2024c). The second retrospective chart review (Fringeli 2024b) is almost entirely a subgroup of the retrospective cohort study (Fringeli 2024c). There is also some overlap of people in the chart reviews (Fringeli 2024a, 2024b).

                • There are some potential sources of bias in these studies. As all studies were single arm and were not comparative studies, principles of randomisation cannot be applied. Due to the nature of the procedure, blinding could also not be applied, leading to selection bias. Sham procedures would not have been feasible. In the retrospective chart review of 30 people, the authors report selection bias towards people with ineffective or weak oesophageal motility (Fringeli 2024b).

                • Follow-up time for 5 out of 6 papers was 12 months or less. 1 paper was a 4-year follow-up (Harsanyi, 2024).

                • There was very little variability in the person inclusion and exclusion criteria across the 6 papers. The main variation was that in the retrospective chart review of 20 people, they selected those with IOM (Fringeli 2024a). In the retrospective chart review of 30 people, they selected those with larger HH (4 to 10 cm in size; Fringeli 2024b).

                • There was little variability in the procedure technique. The same device was used in every procedure.

                • 1 study (2 papers) was funded by the device manufacturer, Implantica (Zug, Switzerland) (Bjelovic 2020, Harsanyi 2024). Feka (2024) reports that 2 of their authors received travel grants and speaker honoraria but do not say from who. In the Fringeli papers, 1 author, Joerg Zehetner, received reimbursement from Implantica for educational activities. There are also overlaps in authors in different studies. Joerg Zehetner, who was an author in the Fringeli (2024a, 2024b, 2024c) papers was also an author on the Feka (2024) paper.

                Any ongoing trials

                • NCT05870163 Post-Market Registry for the Evaluation of RefluxStop in GERD Treatment. This is a single-arm, prospective, open-label, multicentre study. There are 91 enrolled people: 43 in Germany, 24 in Switzerland, 19 in Italy and 5 in Norway. Estimated study completion date is 31 December 2029.

                • RXI005/RENEW RCT (RCT RefluxStop vs. Nissen): to be conducted in Austria (submitted to Ethical Committee and waiting for approval), Switzerland (submitted to Ethical Committee and waiting for approval), Germany (submitted to Ethical Committee and waiting for approval), Italy (under preparation), Spain (under preparation), the UK (under preparation).

                • RXI008/IM REVOLUTION RCT (RCT RefluxStop vs.PPI): under preparation and to be conducted in Austria, Switzerland, Germany, Italy, Spain and the UK.