Impact of Change in Sizing Protocol on Outcome of Magnetic Sphincter Augmentation

Abstract

Objective:

To evaluate and compare magnetic sphincter augmentation (MSA) device sizing protocols on postoperative outcomes and dysphagia.

Background:

Among predictors of dysphagia after MSA, device size is the only factor that may be modified. Many centers have adopted protocols to increase device size. However, there are limited data on the impact of MSA device upsizing protocols on surgical outcomes.

Methods:

Patients who underwent MSA were implanted with 2 or 3 beads above the sizing device’s pop-off point (POP). Clinical and objective outcomes >1 year after surgery were compared between patients implanted with POP+2-versus-POP+3 sizing protocols. Multiple subgroups were analyzed for the benefit of upsizing. Preoperative and postoperative characteristics were compared between the size patients received, regardless of protocol.

Results:

A total of 388 patients were implanted under POP+2 and 216 under POP+3. At a mean of 14.2 (7.9) months, pH normalization was 73.6% and 34.1% required dilation, 15.9% developed persistent dysphagia, and 4.0% required removal. The sizing protocol had no impact on persistent dysphagia (P=0.908), pH normalization (P=0.822), or need for dilation (P=0.210) or removal (P=0.191). Subgroup analysis found that upsizing reduced dysphagia in patients with <80% peristalsis (10.3% vs 31%, P=0.048) or distal contractile integral >5000 (0% vs 30.4%, P=0.034). Regardless of sizing protocol, as device size increased there was a stepwise increase in the percent male sex (P<0.0001), body mass index >30 (P<0.0001), and preoperative hiatal hernia >3 cm (P<0.0001), Los Angeles grade C/D esophagitis (P<0.0001), and DeMeester score (P<0.0001). Increased size was associated with decreased pH normalization (P<0.0001) and need for dilation (P=0.043) or removal (P=0.014).

Conclusions:

Upsizing from POP+2 to POP+3 does not reduce dysphagia or affect other MSA outcomes; however, patients with poor peristalsis or hypercontractile esophagus do benefit. Regardless of sizing protocol, preoperative clinical characteristics varied among device sizes, suggesting size is not a modifiable factor, but a surrogate for esophageal circumference.

Magnetic sphincter augmentation (MSA) was introduced in 2008 with the goal of providing surgeons a technically standardized outpatient alternative to fundoplication with comparable outcomes and an improved side effect profile.¹ Several comparative studies of MSA and fundoplication have demonstrated that this goal was largely achieved. MSA is subject to less technical variability and reported symptomatic and objective reflux control rates are consistent across several centers and comparable to fundoplication.^2,3 However, dysphagia remains the most common complaint after MSA, and device sizing remains a source of variability.

Device size is determined with the aid of a manufacturer-provided sizing tool, which wraps around the esophagus and measures esophageal circumference in a number of beads ranging from 13 to 17. Studies have demonstrated that the use of a smaller device is associated with higher rates of dysphagia.^4,5 As a result, most centers have adopted a practice of upsizing, that is, selecting a device larger than the sizing tool’s recommendation in an effort to reduce the rate of dysphagia. Variability among published sizing protocols is rooted in 3 domains: measurement technique (eg, the extent of dissection, location, and angle of the sizing tool), who gets upsized (eg, all patients, patients with preoperative risk factors for dysphagia, patients with smaller esophageal caliber), and by how much they get upsized. Despite this variability, some have espoused the superiority of upsizing as a foregone conclusion. However, there are no data comparing the efficacy of different sizing protocols.

Reported risk factors for dysphagia after MSA have included smaller device sizes, preoperative dysphagia, normal hiatal anatomy, and poor peristaltic contractions. Among these, the only factor over which surgeons have any agency is device size. However, data on the impact of changing device sizing protocols on MSA outcomes are limited. Therefore, we designed this study to assess the impact of different MSA sizing protocols on persistent dysphagia and reflux control outcomes and to identify the populations of patients who benefit from upsizing the most.

METHODS

Study Population

This was a retrospective review of prospectively collected data of patients who underwent MSA between April 2013 and August 2021 at Allegheny Health Network hospitals (Pittsburgh, PA). Patients aged 18 to 89 years with no history of prior foregut surgery who had at least 1-year follow-up were included in the study. Severe esophageal dysmotility, esophageal stricture, or a known allergy to titanium, nickel, or ferrous material were considered contraindications for MSA. This study was evaluated and approved by the Institutional Review Board of the Allegheny Health Network (IRB 2020-064-WPH).

Preoperative Assessment

All patients underwent a detailed clinical evaluation and completed the Gastroesophageal Reflux Disease-Health Related Quality of Life (GERD-HRQL) questionnaires before MSA implantation. The GERD-HRQL consists of 16 symptom-specific questions scored for severity from 0 to 5. A score of 3 or greater on the heartburn, regurgitation, dysphagia, or gas-bloat–specific items was considered clinically significant. The questionnaire also assesses overall patient satisfaction and proton pump inhibitor (PPI) use.

All patients underwent a routine objective foregut evaluation before MSA, including the following tests:

• Esophagogastroduodenoscopy: Esophagitis graded according to Los Angeles (LA) classification, the presence of Barrett esophagus, and size and type of hiatal hernia (HH) were evaluated. Severe esophagitis was defined as LA grade C or D esophagitis. HH size was measured from the gastroesophageal junction to the crural impression. The presence of a HH was further subdivided into small HH (≤3 cm), large HH (>3 cm), or paraesophageal hernia based on preoperative endoscopic findings.

• Videoesophagram: Anatomy and gross pharyngeal and esophageal motility were assessed to identify any potential masses, mucosal lesions, diverticula, strictures, or HH.

• High-resolution impedance manometry (HRIM): A 4.2 mm solid-state catheter with 36 pressure transducers was used for data collection. Sensors were calibrated and then a baseline measurement was taken before the measurement of 10 serial liquid swallows 20 seconds apart. Standard manometric variables including lower esophageal sphincter (LES) characteristics, esophageal body metrics, and bolus clearance were assessed by using Manoview software (Medtronic).

• Esophageal pH monitoring: Patients underwent 48-hour Bravo (Medtronic) pH monitoring, or 24-hour pH impedance (Diversatek). PPIs were discontinued for 10 days before testing. A DeMeester score >14.7 was considered abnormal distal esophageal acid exposure.

Surgical Procedure

The MSA procedure was completed laparoscopically and consisted of complete reduction and excision of the HH sac if indicated, complete circumferential mediastinal dissection of the esophagus, and restoration of at least 3 cm intra-abdominal esophageal length in all patients. Posterior crural closure was performed with interrupted sutures without mesh. This procedure was carried out for all patients regardless of HH size. A small tunnel was created between the posterior vagal nerve trunk and esophagus at the level of the gastroesophageal junction to place the LINX device.

Protocol for Sizing LINX Device

The LINX Reflux Management System Esophageal Sizing Tool (Ethicon, Johnson & Johnson) has a soft circular curved tip actuated by coaxial tubes via a handset. The handle contains numerical indicators that correspond to the size range of the LINX MSA device (Ethicon, Johnson & Johnson). When the tip is extended, it loops back on itself forming a ring with the magnetic cap. After dissection and creating space around the esophagus excluding the posterior vagus, the sizing tool is placed around the esophagus and tightened until the magnetic cap completing the ring opens, which defines the pop-off point (POP), expressed as a number of MSA beads. Distal esophageal circumference was measured twice with the laparoscopic sizing tool before implanting the LINX device. The LINX device is manufactured in sizes ranging from 13 to 17 according to the bead count. Each bead is 5.8 mm in diameter and 4.3 mm wide. In our practice, to err on the side of larger-sized devices, patients whose measured esophageal circumference was in between two sizes were sized based on the larger of the 2 POP sizes. Early in our practice, we selected the device size by adding 2 beads (POP+2) above the measured POP; however, since March 2018, we have increased it to 3 beads (POP+3) from the POP. This temporal point was used to define the POP+2 and POP+3 sizing protocol groups for analysis.

Postoperative Management and Outcome Assessment

All patients were evaluated and educated by a registered dietician specialized in esophageal surgery nutritional therapy before surgery and during routine postoperative clinic visits. Patients were started on a solid diet on the day of surgery and were instructed to eat small portions of solid food every hour while awake for the first 8 weeks after surgery in an effort to prevent the restriction of the device from opening and to facilitate device actuation.

Patients who developed postoperative dysphagia despite dietary recommendations were considered for endoscopic evaluation and sequential balloon dilation under fluoroscopic guidance to ensure the opening and closure of the MSA device. Patients were prescribed prednisone 20 mg BID for 7 days after dilation to mitigate fibrous encapsulation of the device. We previously reported that early in our practice 22 patients underwent dilation in the first 8 weeks after surgery with only 21% resolution of dysphagia.⁶ We subsequently modified our protocol to wait until 8 weeks after surgery before considering dilation.

Subjective postoperative outcomes were evaluated at routine visits 2 weeks, 6 weeks, 6 months, and annually after the operation. Patients were assessed for resolution of their reflux symptoms, the use of antisecretory medication, and procedure-related complications. Postoperative Gastroesophageal Reflux Disease-Health Related Quality of Life (GERD-HRQL) questionnaires were evaluated. Persistent dysphagia was defined as a score >3 on the GERD-HRQL “difficulty swallowing” item at least 1 year after MSA. Objective testing, including upper endoscopy, pH monitoring, and HRIM, were performed at 1-year follow-up.

Statistical Analysis

Demographic and preoperative clinical characteristics, device size, sizing protocol, GERD-HRQL total score, GERD-HRQL symptom-specific item scores (heartburn, regurgitation, gas bloating, and dysphagia), freedom from PPI, patient satisfaction, DeMeester score, HRIM characteristics, the need for dilation, and device removal rates were used to for analysis. First, a comparative logistic regression analysis between POP+2 and POP+3 sizing protocol groups was performed. Then multiple subgroup analyses were performed to compare the impact of the sizing protocol on the outcomes of persistent dysphagia and pH normalization. A total of 34 subgroups were assessed for both outcomes. Each subgroup consisted of patients stratified by the presence or absence of one of the following variables: age >45 years; female sex; body mass index (BMI) >30; preoperative dysphagia; HH absent, <3 cm, or ≥3 cm; DeMeester score >14.7; DeMeester score >50; no esophagitis; esophagitis LA grade A/B or LA grade C/D; LES resting pressure >45; integrated relaxation pressure (IRP) <15; distal contractile integral (DCI) <750, 750 to 5000, or >5000; contractile amplitude <43; percent peristalsis <80%; or bolus clearance <10%, <20%, or <30%. Finally, a comparative analysis between actual device sizes received, regardless of sizing protocol, was performed to assess for preoperative differences between device size groups. Values for continuous variables are expressed as either mean (SD) or median with an interquartile range when appropriate. Values for categorical variables are presented as frequency and percentage. Statistical analysis was performed by means of the nonparametric Mann-Whitney U test, Wilcoxon signed-rank test, and Fisher exact test when appropriate. A P value <0.05 was considered significant. Statistical analysis was performed using SAS software (SAS Institute Inc.).

RESULTS

A total of 604 patients underwent laparoscopic MSA during the study period. The population was 62.7% female with a median [interquartile range (IQR)] age of 57 (46–64) and a BMI of 29.0 (26–32). On endoscopic evaluation 10.8% had LA grade C or D esophagitis and 24.2% had a >3 cm HH. DeMeester score was 29.0 (17–43).

At a mean (SD) follow-up of 14.2 (7.9) months, the median (IQR) GERD-HRQL total score improved from 31.0 (18–48) to 5.0 (2–12) (P<0.0001), 85.6% of the patients were satisfied with the outcomes of MSA, 90.0% had discontinued PPI use, and 73.6% had normalization of their distal esophageal acid exposure. Dilation for dysphagia was performed in 34.1% of patients with 15.9% developing persistent dysphagia. However, compared to the preoperative dysphagia rate of 29.2%, this was a significant decrease (P<0.0001). Device removal was required by 4.0% of patients for dysphagia and 1.3% of patients for poor reflux control. Patients with smaller devices (13/14 beads) were more likely to need dilation for dysphagia (39.2% vs 27.6%, P=0.003), develop persistent dysphagia (17.4% vs 8.8%, P=0.002), and require device removal for dysphagia (7.8% vs 1.6%, P=0.0006). Persistent dysphagia was lower in patients with >3 cm HH (13.5% vs 23.8%, P=0.012).

Impact of Sizing Protocol on Outcomes

There were 388 (64.2%) patients implanted using the POP+2 sizing protocol and 216 (35.8%) patients using POP+3. Postoperative GERD-HRQL and objective testing outcomes were compared between sizing protocol groups as shown in Table 1. Upsizing from POP+2 to POP+3 had no impact on symptomatic or objective reflux control, patient satisfaction, or freedom from PPIs (P>0.05). Postoperative manometry demonstrated that POP+3 had a shorter median (IQR) LES total length [2.8 (2–3) vs 3.1 (3–4), P=0.003] and a higher IRP [14.9 (11–19) vs 13.0 (9–17), P=0.012]. All other manometric characteristics were comparable (P>0.05). The median (IQR) dysphagia score in the POP+3 group was 0.0 (0–2), which was lower than the POP+2 group at 1.0 (0–2) but this difference did not reach statistical significance (P=0.156). Similarly, the percent of patients with a zero dysphagia score was higher with POP+3, but not significantly (51.9% vs 44.8%, P=0.107). The need for dilation for dysphagia, the development of persistent dysphagia, and the rate of removal for dysphagia were all comparable between POP+2 and POP+3 groups (P>0.05).

TABLE 1.

Comparison of Postoperative Quality of Life and Objective Outcomes Between Sizing Protocols

Characteristic	POP+2 (N=388)	POP+3 (N=216)	P
GERD-HRQL components, median (IQR)
Total score	5.0 (2–12)	4.0 (2–12)	0.336
Heartburn item	0.0 (0–4)	0.0 (0–3)	0.194
Regurgitation item	0.0 (0–4)	0.0 (0–4)	0.900
Gas bloating item	2.0 (0–3)	1.0 (0–3)	0.542
Patient satisfaction, N (%)	297 (85.6)	177 (85.5)	1.000
50% reduction GERD-HRQL total score, N (%)	261 (78.4)	146 (78.1)	1.000
Freedom from PPI, N (%)	319 (91.4)	168 (87.5)	0.177
DeMeester score, median (IQR)	6.1 (2–16)	4.3 (2–16)	0.387
pH normalization, N (%)	204 (73.9)	116 (73.0)	0.822
Device removal for reflux symptoms, N (%)	5 (1.3)	3 (1.4)	1.000
Dysphagia-related outcomes
Persistent dysphagia, N (%)	61 (15.7)	35 (16.2)	0.908
GERD-HRQL dysphagia item, median (IQR)	1.0 (0–2)	0 (0–2)	0.156
Need for dilation, N (%)	125 (32.2)	81 (37.5)	0.210
Device removal for dysphagia, N (%)	12 (3.1)	12 (5.6)	0.191
LES characteristics, median (IQR)
LES total length (cm)	3.1 (3–4)	2.8 (2–3)	0.003
LES intra-abdominal length (cm)	1.8 (1–3)	1.7 (1–2)	0.347
LES resting pressure (mm Hg)	28.4 (20–35)	27.3 (23–37)	0.410
Integrated relaxation pressure (mm Hg)	13.0 (9–17)	14.9 (11–19)	0.012
Esophageal body characteristics, median (IQR)
Distal contractile integral (mm Hg-s-cm)	1910 (1109–3517)	1852 (1073–3225)	0.122
Contractile amplitude (mm Hg)	86.7 (63–128)	83.8 (67–148)	0.555
Percent peristaltic	100.0 (80–100)	90.0 (80–100)	0.646
Incomplete bolus clearance (%), median (IQR)	10.0 (0–50)	10.0 (0–40)	0.610

POP+2 indicates 2 beads above the sizing device pop-off point; POP+3, 3 beads above the sizing device pop-off point.

Impact of Sizing Protocol in Multiple Subgroup Analyses

Sizing protocol resulted in significant or borderline significant differences in persistent dysphagia rates within subgroups based on 3 variables: LES resting pressure, DCI, and percent peristalsis, as shown in Table 2. Among the 36 (5.8%) patients with a DCI >5000, upsizing to POP+3 resulted in significantly less persistent dysphagia (0% vs 30.4%, P=0.034). Among the 57 (9.4%) patients with a preoperative LES resting pressure >45, the rate of dysphagia was higher in POP+2 (32.1%) compared to POP+3 (6.3%), but this difference did not reach statistical significance (P=0.067). Upsizing also reduced the rate of dysphagia among the 71 (11.8%) patients with a percent peristalsis <80 (10.3% vs 31%, P=0.048). Within this <80% peristalsis subgroup, 53 underwent postoperative pH testing. POP+3 resulted in much lower pH normalization at 59.1% compared to 83.9% with POP+2, but this difference did not reach statistical significance (P=0.058). The remaining subgroups had comparable persistent dysphagia and pH normalization rates between sizing protocol groups (P>0.1).

TABLE 2.

Subgroup Comparison of Sizing Protocol for Impact on Dysphagia

Variables	Subgroup	POP+2 dysphagia, n (%)	POP+3 dysphagia, n (%)	P
LES resting pressure	>45	9 (32.1)	1 (6.3)	0.067
	≤45	50 (14.3)	34 (17.2)	0.389
DCI	<750	49 (16.0)	34 (19.4)	0.380
	750–5000	42 (14.8)	34 (21.0)	0.116
	>5000	7 (30.4)	0 (0)	0.034
Percent peristalsis	<80	13 (31)	3 (10.3)	0.048
	≥80	45 (13.7)	32 (17.4)	0.303

POP+2 indicates 2 beads above the sizing device pop-off point; POP+3, 3 beads above the sizing device pop-off point.

Preoperative and Postoperative Factors and Device Size, Regardless of Sizing Protocol

Regardless of sizing protocol, during the study period, there were 129 (21.4%) patients who were implanted with size 13 beads, 215 (35.6%) with size 14 beads, 156 (25.8%) with 15 beads, 72 (11.9%) with 16 beads, and 29 (4.8%) with 17 beads. The assessment of preoperative factors between device sizes showed that as device size increased there was a stepwise increase in the prevalence of male sex (P<0.0001), BMI >30 (P<0.0001), >3 cm HH (P<0.0001), and LA Grade C or D esophagitis (P<0.0001) (Fig. 1). There was a similar trend with preoperative DeMeester score (P<0.0001) (Table 3). As device size increased from 13 to 17, age went from 54.0 (42–61) to 57.0 (47–64) to 58.0 (50–65) to 57.0 (48–64) to 56.0 (44–66) (P=0.037). Preoperative GERD-HRQL total score (P=0.904) and the prevalence of preoperative dysphagia (P=0.701) were comparable between device size groups. On preoperative manometry, IRP decreased in a stepwise manner as device size increased (P<0.0001). LES intra-abdominal length (P=0.033), LES resting pressure (P=0.0007), DCI (P=0.0012), and contractile amplitude (P=0.017) were also significantly different, but there was no direct or indirect stepwise relationship with increasing device size (Table 3).

FIGURE 1.

Bar graphs showing that as device size increased from 13 to 17 beads, there was a significant stepwise increase in the frequency of male sex (P<0.0001), body mass index (BMI) >30 (P<0.0001), preoperative hiatal hernia >3 cm (P<0.0001), and Los Angeles grade C or D esophagitis (P<0.0001).

TABLE 3.

Comparison of Preoperative pH Monitoring and Manometric Characteristics by Device Size

	Device size
Variables, median (IQR)	13 (N=129)	14 (N=215)	15 (N=156)	16 (N=72)	17 (N=29)	P
DeMeester score	22.3 (12–33)	26.0 (16–39)	36.8 (24–49)	33.0 (20–53)	40.1 (30–62)	<0.0001
LES total length	2.9 (2–4)	2.9 (2–4)	2.8 (2–4)	2.9 (2–4)	2.7 (2–4)	0.822
LES intra-abdominal length	0.9 (0–2)	1.2 (0–2)	1.3 (0–2)	0.0 (0–2)	0.5 (0–2)	0.033
LES resting pressure	24.4 (15–33)	22.8 (16–33)	30.3 (11–30)	16.1 (10–29)	15.4 (12–24)	0.0007
Integrated relaxation pressure	8.1 (4–12)	7.6 (4–11)	5.2 (2–9)	5.0 (2–9)	4.6 (0–6.2)	<0.0001
Distal contractile integral	1699 (1.1–2.9k)	1804 (1.1–3.0k)	1593 (0.9–2.5k)	1269 (0.7–2.2k)	997 (0.7–1.5k)	0.0012
Contractile amplitude	87.8 (68–109)	90.0 (64–123)	78.9 (60–115)	79 (56–107)	64.7 (49–91)	0.017
Percent peristaltic	100.0 (80–100)	100 (90–100)	100 (90–100)	100 (90–100)	100 (85–100)	0.784
Incomplete bolus clearance	0.0 (0–20)	0.0 (0–10)	0.0 (0–20)	0.0 (0–20)	10.0 (0–40)	0.078

Postoperatively, the rate of pH normalization decreased in a stepwise manner with increasing device size, with the exception of size 16, which was an increase from 65% at size 15 to 72%, before dropping to 38% with size 17 (P<0.0001) (Fig. 2). A similar trend was observed in the need for dilation (P=0.043). The rate of persistent dysphagia was highest in size 13 at 20.8% and decreased in a stepwise manner with increasing device size to 8% at size 16, but increased slightly to 10% at size 17. However, this trend did not reach significance (P=0.174). A similar trend was observed with the need for the removal of dysphagia, but the relationship was significant (P=0.014). There was no difference in GERD-HRQL improvement (0.322), freedom from PPIs (P=0.147), satisfaction (P=0.094), or device removal for reflux (P=0.942). Postoperative manometry characteristics are stratified by device size in Table 4. There was a stepwise decrease in IRP with increased device size (P<0.0001), similar to the trend observed in preoperative IRP (P<0.0001) (Fig. 3). However, postoperative IRP was significantly higher than preoperative IRP for every device size (P<0.0001), with the exception of size 17, which was borderline significant (P=0.075).

FIGURE 2.

Bar graphs comparing postoperative outcomes between device sizes. There was a significant difference in pH normalization (P<0.0001), need for dilation (P=0.043), and need for device removal for dysphagia (P=0.014), but not for persistent dysphagia (P=0.174).

TABLE 4.

Comparison of Postoperative Manometric Characteristics by Device Size

	Device size
Variables, median (IQR)	13 (N=58)	14 (N=90)	15 (N=66)	16 (N=28)	17 (N=13)	P
LES total length (cm)	2.9 (2–4)	3.0 (3–4)	2.9 (2–4)	3.2 (3–4)	2.3 (2–3)	0.150
LES intra-abdominal length (cm)	1.6 (1–3)	1.8 (1–2)	2.1 (1–3)	2.0 (2–3)	0.0 (0–2)	0.071
LES resting pressure (mm Hg)	30.6 (25–37)	29.1 (21–37)	26.2 (20–33)	24.8 (19–33)	19.8 (18–32)	0.064
Integrated relaxation pressure (mm Hg)	16.7 (13–20)	14.5 (11–18)	12.4 (10–16)	10.5 (8–14)	9.4 (6–13)	<0.0001
Distal contractile integral (mm Hg-s-cm)	1853 (1.2–3.5k)	2644 (1.1–4.0k)	1925 (1.1–2.9k)	1649 (1.2–2.9k)	915.3 (0.7–1.6k)	0.038
Contractile amplitude (mm Hg)	88.7 (67–139)	95.6 (66–148)	83.5 (66–107)	82.9 (67–106)	60.3 (45–88)	0.053
Percent peristaltic	90.0 (80–100)	100.0 (80–100)	90.0 (80–100)	90.5 (98–100)	90.0 (90–100)	0.872
Incomplete bolus clearance (%)	20.0 (10–50)	10.0 (0–40)	10 (0–40)	9.5 (0–43)	10.0 (0–10)	0.100

FIGURE 3.

Box whisker plots showing a significant stepwise decrease in (A) preoperative integrated relaxation pressure (IRP) (P<0.0001) and (B) postoperative IRP (P<0.0001) as device size increased from 13 to 17 beads.

DISCUSSION

The history of antireflux surgery has been a quest to achieve reflux control without the development of postoperative symptoms. In the 1950s, Rudolf Nissen introduced fundoplication with immediate success in terms of reflux control. Studies with 20-year follow-up have demonstrated excellent and durable outcomes after open Nissen fundoplication. However, many patients experienced troublesome postoperative dysphagia and bloating symptoms, which deflated antireflux surgery’s appeal. In the 1960s, Dor and Toupet sought to ameliorate this situation with their eponymous partial fundoplications. In the 1970s, DeMeester demonstrated that dysphagia and bloating could be reduced without compromising reflux control with a slight modification to the 360-degree wrap, commonly known as the “short and floppy Nissen.” The 1980s saw the rise and fall of the Angelchik device, the first attempt at an antireflux implant, which resulted in such high rates of device migration, erosion, postoperative dysphagia, and removal that the procedure was almost entirely abandoned by the 1990s. The advent of laparoscopic surgery reduced the morbidity of all types of fundoplication but did nothing to improve dysphagia or bloating. As a result, a treatment gap developed as some patients with medically refractory GERD were unwilling to abide by the risks of postoperative complications. In 2008, MSA was introduced as a standardized outpatient antireflux surgery with the potential to improve the ARS side effect profile and address this treatment gap. Now, 15 years later, MSA is widely accepted as an effective alternative to fundoplication with a somewhat improved side effect profile, yet dysphagia remains a problem. Outcome studies have identified poor peristalsis, esophageal hypocontractility, preoperative dysphagia, and smaller device size as predictors of dysphagia after MSA. However, among these predictors, small device size is the only potentially modifiable factor. As a result, many centers have adopted protocols aimed at increasing device size. However, this is the first study to assess the impact of device upsizing on dysphagia and reflux control. We found that in the overall population, upsizing from POP+2 to POP+3 has no impact on dysphagia or reflux control. However, within the subgroups of patients with poor peristalsis, hypercontractile esophagus, or hypertensive lower esophageal sphincter, the risk of dysphagia can be reduced by upsizing.

The relationship between device size and postoperative dysphagia is well established. Our center previously demonstrated an inverse relationship between dysphagia and device size, which we have redemonstrated in this present study.⁶ These findings are consistent with the findings reported by other centers. One study of 68 patients undergoing MSA found that 58.5% of patients with device size 12–14 had dysphagia compared to 30% of those with size 15–17, P=0.026.⁵ However, they do not detail how their devices were sized. A group from Austria studied postoperative dysphagia in 268 patients who underwent MSA and found that device size 13 was an independent predictor of postoperative dysphagia. They report a tailored sizing protocol, using POP+3 for the smallest esophagi, a judgment call for medium size esophagi, and POP+2 for larger esophagi.⁷ The natural conclusion from these results is that smaller devices have a higher resistance at the EGJ. This was demonstrated in a previous study that found higher postoperative IRP, a measure of EGJ outflow resistance, in patients with smaller devices, consistent with our present findings.⁸ The result of this higher outflow resistance is that the esophageal body must work harder with smaller sizes to achieve bolus clearance without dysphagia. This concept is akin to the proposed mechanism for dysphagia after an overly tight fundoplication.

Logically upsizing to a looser device should provide less EGJ outflow resistance, enabling easier bolus clearance and yielding lower rates of dysphagia. However, the most unexpected finding of this study was that upsizing from POP+2 to POP+3 resulted in a higher IRP and did not change bolus clearance or dysphagia rates. Therefore, dysphagia after MSA cannot simply be due to a tight device. The most likely explanation for this finding is that another factor contributes to increased outflow resistance after MSA, the fibrous capsule. The MSA was the first implantable titanium device in the GI tract. However, decades of orthopedic and orthodontic literature have demonstrated that titanium implants induce a foreign body response, which attempts to isolate the implant from the host by walling it off with a thick fibrous capsule. A similar fibrous capsule forms around the MSA. The impact of this fibrous capsule on esophageal physiology has been evaluated in studies looking at MSA device removal, as the capsule is not commonly removed with the device. A study from our center of 40 patients who underwent removal for heartburn or dysphagia found that IRP after device removal did not change but remained increased compared to preoperative IRP. In addition, patients who underwent removal for dysphagia maintained adequate subjective and objective reflux control.⁹ Other studies have similarly found that the fibrous capsule can contribute to postoperative dysphagia and postremoval reflux control without revisional antireflux surgery.¹⁰ These findings suggest that the fibrous capsule provides substantial EGJ resistance, even in the absence of the MSA device. Orthopedic studies have demonstrated that when the interface between host tissue and an implant is loosened, the fibrous capsule can become substantially thicker.^11,12 This finding suggests that MSA upsizing may result in a thicker capsule by similarly loosening the interface between the esophageal wall and the titanium beads. The resistance provided by this thicker capsule likely negates any effects of upsizing on dysphagia or reflux control.

The choice of device size can have a major impact on surgical outcomes at the extremes. Studies have demonstrated that a device tight enough to compress the esophagus puts the patient at elevated risk of dysphagia and erosion.¹³ Conversely, an overly large device may slide down onto the cardia and fail to be in an appropriate position to augment the LES. However, the assumption that the choice of device size within these boundaries can influence MSA outcomes is challenged by the findings of the present study. Regardless of whether 2 or 3 beads are added, device size is based on the measured esophageal circumference, a nonmodifiable anatomic characteristic. The MSA sizing tool may use the unconventional unit of measurement “number of beads,” but fundamentally device size is a surrogate for esophageal circumference. This notion is consistent with our findings that reflux symptom control, acid exposure, and dysphagia-related outcomes were unaffected by upsizing from POP+2 to POP+3, yet varied with device size regardless of upsizing. In addition, it provides an explanation for the indirect relationship between preoperative IRP and device size, factors that should be independent if device size were purely an intraoperative modifiable factor. Mathematically, circumference is directly proportional to radius, and based on Poiseuille’s law of resistance ( $R\propto \frac{\eta L}{r^4})$ , indirectly proportional to resistance.¹⁴ Therefore, as esophageal circumference decreases, EGJ outflow resistance increases. This mathematical principle is also consistent with previous studies associating small-caliber esophagus with dysphagia in patients with normal findings on barium esophagram, esophagogastroduodenoscopy, and manometry.^15,16 However, data on the relationship between anatomic esophageal circumference and esophageal physiology are limited. Further studies are warranted to assess the relationship between anatomic and physiological esophageal measures and their impact on postoperative dysphagia.

The strongest evidence that device size reflects innate anatomic variance and is not a modifiable surgical technique is the finding that demographic and preoperative clinical characteristics were different between patients who received different device sizes. This unexpected finding has not been previously reported in the literature. Specifically, male sex, BMI >30, preoperative HH size, DeMeester score, and severe esophagitis all increased with each increasing device size. These preoperative factors are interrelated and important in the pathophysiology of GERD. Previous studies have demonstrated that male sex and increased BMI are associated with an increased risk for HH and erosive esophagitis.¹⁷ In addition, large HHs are a well-established cause of distal esophageal acid exposure and erosive esophagitis. Therefore, device size is a marker for the severity of preoperative GERD. This finding suggests that sizing is not a modifiable operative technique, but rather a reflection of inherent anatomic and physiological variance in patients. A large device size indicates a large esophageal circumference, which may require a larger hiatal opening, and likely contributes to the increased risk of HH. Previous studies have demonstrated that as preoperative HH size increased the risk of hernia recurrence, and therefore poor reflux control, increased in a stepwise manner.¹⁸ This is consistent with our finding that device size was directly associated with HH size and postoperative abnormal esophageal acid exposure.

In the overall population, the absence of a large HH was associated with a higher rate of persistent dysphagia and was more common as device size decreased. These findings are consistent with a study assessing predictors of dysphagia after MSA in a cohort of 380 patients, which found that both small device size and the absence of large HH were associated with an increased risk of dysphagia after MSA on univariate analysis. However, on multivariate analysis, the absence of a large HH was associated with almost 3 times more dysphagia, while device size was not an independent predictor of dysphagia.⁶ These findings suggest that the relationship between device size and dysphagia may reflect the relationship between HH and dysphagia. The presence of a large HH causes an elevation in the resistance at the EGJ. Repair of the hernia greatly reduces this resistance. However, MSA increases resistance. In a patient with an intact hiatus, there is only an increase in resistance, which may be perceived as dysphagia. By contrast, patients with a large HH, which was tolerated as it grew in size slowly over time, may be more accustomed to a higher postoperative resistance or even experience a net decrease in resistance, resulting in lower rates of perceived dysphagia after MSA.

Upsizing was ineffective in the total population, but patients with poor peristalsis, hypertensive native sphincter, or esophageal body hypercontractility benefitted from upsizing. Interestingly, these upsizing sensitive subgroups are similar to previously identified predictors of dysphagia. A study of persistent postoperative dysphagia in 380 patients who underwent MSA found that <80% peristalsis (odds ratio: 2.5, P=0.031) was an independent predictor of persistent dysphagia. In addition, they found that there was a stepwise increase in the rate of dysphagia from 16.2% to 42.9% as DCI increased from >4000 to >7000.⁶ The findings of the present study suggest that these risk factors can be mitigated with a looser MSA device.

Although the findings of this study suggest that upsizing will not aid in outcome optimization, appropriate risk stratification and patient selection can improve outcomes. In our practice, risk factors, such as preoperative dysphagia, are not treated as contraindications to surgery. Despite being the most common complaint after MSA, the prevalence of dysphagia actually decreases after MSA. This decrease was demonstrated in the present study and is consistent with previous studies, which have found that although preoperative dysphagia is a risk factor for persistent postoperative dysphagia (odds ratio: 2.19, P=0.037), the prevalence of dysphagia decreased from 35% preoperatively to 15.5% after MSA (P<0.001).¹⁸ As a result, it is our practice to use preoperative factors associated with both favorable outcomes and complications to provide patients with a tailored risk-benefit assessment. The decision to proceed with surgery is ultimately determined through an informed shared-decision–making model.

We acknowledge certain limitations in this study including the nonrandomized sizing protocol and inherent variability in sizing technique. Accurate measurement using the sizing tool depends on port placement, patient body habitus, degree of dissection, and angulation of the sizing tool. Standardized protocols within our practice such as performing a complete dissection in all patients and measuring esophageal size twice to confirm accuracy were designed to mitigate variability; however, these factors are not completely unavoidable and are potential sources of variability within the current MSA sizing technique. In addition, because there are only 5 sizes to choose from, there is an element of judgment that goes into sizing patients whose measured esophageal circumference lies between 2 sizes. Our practice has been to err on the side of a larger device; however, this is another source of variability with the current sizing tool.

There is concern that an increase by only 1 bead may not be sufficient to affect postoperative dysphagia. However, this interpretation of our results is somewhat lacking. First, the probability that our results were due to chance was only 9.2% in this large cohort of 604 patients. In addition, early MSA studies did not add any additional beads at all from the measured circumference and demonstrated dysphagia rates of 43% to 68% during the immediate postoperative period and 11% at 1 year.^19,20 More recent studies using protocols with 2 to 3 additional beads reported similar rates. One study, which used 2 to 3 beads to upsize, reported immediate dysphagia in 63.2% of patients and persistent dysphagia at 1 year in 15.5%.⁴ Another study with a similar sizing protocol compared patients with severe and nonsevere GERD and found dysphagia rates at 1 year of 25.4% and 14.1%, respectively.²¹ Another study, which tailored POP+2 and POP+3 to esophagus size reported that 37% of 268 patients had at least some degree of dysphagia. Therefore, as clinical practice has evolved to include more upsizing protocols, reported rates of dysphagia rates have remained relatively constant. These findings suggest that upsizing does not have a substantial impact on dysphagia rates in the overall population and that the results of the present study were not due to an insufficient increase in size. Based on these findings there is not sufficient evidence to recommend further upsizing beyond POP+3. Upsizing too much may increase the risk of device migration and the need for reoperation.

CONCLUSIONS

This was the first study to compare 2 different sizing protocols for the MSA device. We demonstrated that in the overall population upsizing from 2 beads (POP+2) to 3 beads (POP+3) above the measured esophageal circumference does not impact the rate of postoperative dysphagia. However, in patients with poor peristalsis, hypercontractile esophagus, and hypertensive lower esophageal sphincter, there is a benefit to upsizing the MSA device. Regardless of sizing protocol, patients who were implanted with different-sized devices had significantly different demographic and preoperative clinical characteristics, with increased HH size, erosive esophagitis, and abnormal distal esophageal acid exposure. This finding suggests that the relationship between device size and postoperative dysphagia may be related to nonmodifiable factors such as esophagus size. Therefore, further upsizing is not warranted. To optimize outcomes and achieve dysphagia-free reflux control, surgeons should emphasize patient selection and risk stratification based on nonmodifiable predictors of dysphagia, such as adequacy of the esophageal body.