Linear Magnetic Compression Gastroileostomy Bipartition: Feasibility and Early Outcomes

Abstract

BACKGROUND:

Minimally invasive magnetic compression technique may simplify the completion of an anastomosis and reduce major surgical complications by obviating permanent staple and suture fixation and enterotomy closures.

STUDY DESIGN:

A multisite study of magnetic gastroileostomy bipartition (MagGI procedure) as a revisional option for suboptimal sleeve gastrectomy (SG) was conducted in adults (BMI [kg/m²] greater than or equal to 30.0–less than or equal to 50.0). After 1 magnet was delivered endoscopically to the ileum and another to the gastric antrum, the magnets were laparoscopically approximated to initiate gradual compression anastomosis. The 90-day primary feasibility endpoint was magnet placement (≥90% of cases), magnet passage without reintervention, and patent anastomosis creation. Secondary endpoints were weight reduction and improved metabolic indicators.

RESULTS:

Between November 6, 2023, and July 10, 2024, 20 enrolled patients (mean age 46.1 years, BMI 37.1 ± 0.8, weight 101.0 ± 3.0 kg) underwent SG-revisional MagGI. Primary feasibility endpoint was met: 100.0% magnet placement, median passage 31.0 days; patent anastomosis confirmed (20 of 20). 80.0% (47 of 59) of AEs were low grade (Clavien-Dindo grade I and II). Two intestinal tears (grade III) due to bowel forceps were repaired and resolved; 4 grade-IV SAEs, none related to the magnet device. No anastomotic leak or bleeding; no mortality. Respective 6-month (n = 17) and 12-month (n = 7) weight loss: total 16.9%, 23.4%; excess 57.5%, 66.0%; BMI reduction was 6.2 and 9.1 kg/m². Satisfaction questionnaire (6 months): 100.0% would recommend the procedure.

CONCLUSIONS:

Preliminary findings suggested that SG-revisional MagGI was technically feasible and achieved a promising excess weight loss of 66.0%. Longer-term follow-up in larger cohorts is needed to establish safety and efficacy.

Patients underwent a first-in-human magnetic gastroileostomy study for weight regain after suboptimal sleeve gastrectomy. The procedure was feasible and safe; no stapling or suturing was required, and no retained foreign material was noted. A total of 100.0% formed patent anastomoses without leak or bleeding, resulting in significant, substantial excess weight loss: 6 months (57.5%, n = 17); 1 year (66.0%, n = 7).

Obesity is a highly prevalent disease that continues to threaten health and longevity globally. Patients with obesity (BMI greater than or equal to 30 kg/m²), type 2 diabetes (T2D), and other comorbid diseases are safely and more effectively treated during the long term by metabolic and bariatric surgery (MBS) than by conventional medical therapies.^1-3 Contemporary MBS operations have become increasingly lower in risk due to advances in minimally invasive technique^4,5; yet, acceptance of MBS remains limited to <1.0% of qualified patients, often due to patients’ fear of undergoing surgery.^6,7 Recently, injection-based weight-loss medications, glucagon-like peptide-1 (GLP-1) receptor agonists, have further diverted patients from electing MBS^8-10 although this therapy is often accompanied by severe gastrointestinal and other adverse effects; 53.6% discontinued treatment in less than 1 year, and 72.2% stopped within 2 years (n = 1,548).¹¹

The contemporary magnetic compression anastomosis (MCA) technique has been under preclinical and clinical study as a means of further augmenting minimally invasive MBS by providing a potentially safer technique for anastomosis formation.¹² Linear magnetic anastomosis system (LMAS) technology facilitates endolaparoscopic creation of patent, durable anastomoses without stapling and suturing, enterotomy, or potentially inflammatory implanted stabilizers (eg anastomosis “locking systems,” rings, staples, clips, sutures, glue).^13-16 Twelve publications to date of the LMAS in MBS and general surgery contexts (duodenoileal, gastroileal, gastrojejunal, jejunoileal, cholecystoduodenal) reflect a cautious progression of MCA studies from preclinical testing, through first-in-human (FIH) feasibility investigation, to clinical series reporting early and ongoing outcomes.^17-28

In 20.0% of patients who have undergone the typically very effective MBS sleeve gastrectomy (SG) procedure, revisional surgery is needed because of insufficient weight loss, weight regain, or complications.²⁹ In an FIH stage-1 study, side-to-side LMAS gastroileostomy (MagGI) bipartition used as a revisional procedure after suboptimal SG demonstrated feasibility and promising preliminary efficacy in a small cohort at 6-month follow-up. The MagGI procedure provides dual sources of alimentary flow, 1 through the native duodenal pathway, aimed at retaining nutritionally protective metabolic activity, and a second channel, the bipartition, that rapidly transits undigested chyme to the terminal ileum, stimulating the incretin effect that promotes appetite control and weight loss.²⁷ Here, outcomes of stage-1 MagGI patients who have completed 1-year follow-up are described in conjunction with those of stage-2 patients at 6 months. Satisfaction questionnaire responses of the total cohort are also reported.

METHODS

Study design and conduct

The evaluation was designed as a prospective, open-label, observational study of the investigational Magnet System in gastroileostomy (the “MagGI Study,” Clinicaltrials.gov NCT#06073457). Guidelines regulating clinical development of investigational devices were incorporated into the study protocol. After an initial 90-day feasibility and proof-of-concept evaluation (stage 1) in the Canadian pilot center, the ongoing study was paused for the scheduled safety evaluation by an independent Data Safety Monitoring Board. Its approval initiated stage 2, ongoing safety, and preliminary efficacy assessment. The Ethics Committee of the pilot center and 2 additional study sites in Europe and the Middle East approved the protocol for 30 or fewer participants. Going forward, the Data Safety Monitoring Board reviewed all adverse events (AEs) and serious AEs (SAEs) to ensure patient safety. Sites adhered to the Helsinki Declaration consistent with ISO14155 regulations, 21 CFR Parts 11, 50, 54, 56, and 812 Good Clinical Practices.

Patients and inclusion

Patients who would be appropriate candidates for revision of SG due to weight regain or inadequate weight loss, and T2D recurrence were introduced to the investigational technique and study aims. After consultation with the surgeon and team, patients were able to make an informed decision about their participation in the FIH MagGI Study. Patients who approved the intervention plan provided written informed consent and were enrolled on the FIH MagGI Study.

The primary criteria for patient inclusion were to be 18 to 65 years old at time of consent with a BMI of greater than or equal to 30.0 to less than or equal to 50.0 kg/m² and weight regain after earlier SG (ie 12 months or more previously) or T2D (HbA1_C ≥6.5%) or indicated for a gastroileal GI) diversion. For a period of 1 year, patients were asked to forego additional MBS or reconstructive surgery that would affect body weight. Female patients were asked to use contraception and not become pregnant for the duration of the study. Patients were required to adhere to all protocol requirements for a minimum of 1 year and to attend all follow-up appointments.

The absolute study exclusions were type 1 diabetes; injectable insulin use; uncontrolled T2D; dyslipidemia or sleep apnea, hypertension; earlier intestinal, duodenal, or colonic surgery other than earlier MBS; a plan to perform hiatal hernia repair or cholecystectomy with the investigational study GI anastomosis procedure; earlier trauma, surgery, prostheses, disease, scarring, abnormal anatomy, or genetic expressions that contraindicated the study procedure; helicobacter pylori positive or active ulcer disease; Barrett’s disease; large hiatal hernia; refractory GERD, inflammatory bowel or colonic diverticulitis; any anomaly precluding laparoscopy or orogastric access by catheters and gastroscope; an implantable pacemaker or defibrillator; untreated or poorly controlled psychiatric illness or substance abuse; pregnancy, breastfeeding, or lack of effective contraception; surgery 1 month before or after the study procedure; stroke and transient ischemic attack within 6 months before study consent; gastrointestinal tract anomalies at or distal to magnet placement positions; chronic anticoagulation therapy (except aspirin). Also excluded were patients with active infection requiring antibiotic therapy; those participating in another investigation; known allergies to the device components or contrast media; tobacco or nicotine product cessation or cessation 3 months or less before informed consent; terminal disease; and a positive COVID-19 test before the MagGI procedure.

Anastomosis system for gastroileostomy

The MagGI System is composed of 2 poly(lactic-co-glycolic) acid-flanged neodymium-iron-boron magnets (60- or 70-mm long × 10.2-mm wide × 4.4-mm thick), the Metabolic Delivery System (“delivery system,” 198 cm long), and the Metabolic Laparoscopic Positioning Device (“positioning device”; GT Metabolic Solutions, Inc., San Jose, CA). The 80-mm long × 10.2-mm wide × 4.4-mm thick magnets comprised 3 titanium alloy elements: a central housing with neodymium-iron-boron magnetic core encased in a biocompatible parylene-C coating bounded by 2 adjacent nonmagnetic end components; all are encircled by a biofragmentable poly(lactic-co-glycolic) acid flange (Fig. 1 inset). For ease of detachment of the magnets from the gradually forming anastomosis, the flange eventually fragments, causing the outer nonmagnetic pieces to separate from the central magnetic component, facilitating expulsion of the magnet segments from the body. The shorter 60-mm long magnets were introduced to facilitate the patient’s ability to swallow the magnets, thereby potentially averting the use of 1 or 2 endoscopic procedures.

Figure 1.

(Inset) Left, biofragmentable tripartite magnet; right, magnet fragmented into 3 parts at detachment from the site of magnetic gastroileal (MagGI) anastomosis formation; (A) distal MagGI magnet is placed orogastrically post-pylorus by flexible gastroscope, subsequently transported by laparoscopic magnetic positioning device to the target ileal position 250 to 350 cm proximal to the cecum. (B) After the second, proximal MagGI magnet is delivered by gastroscope to the stomach, the distal magnet is elevated by the positioning device over the transverse colon and brought into apposition with the proximal magnet in the stomach; the mesenteric defect is closed. (C) After the magnets gradually compress and fuse, forming the GI anastomotic bipartition, the biofragmentable flange of the magnet pair detaches from the GI site and passes naturally by peristalsis. (D) Alimentary content now flows through both the native pyloric channel and the patent GI bipartition into the ileal lumen.

The delivery system is a flexible catheter for orogastric transport of the magnet. The laparoscopic positioning devices (with optional magnetic strengths: 12, 18, 21, 24, and 27 Newtons) are used to advance the distal magnet to the target location in the ileum.

Surgical technique

The placement procedure is the same for the 60-, 70-, and 80-mm magnets. The detailed MagGI technique has been previously described.²⁷ In brief, during an inpatient procedure, under propofol sedation, the first in a pair of 70- or 80-mm MagGI System magnets (the distal magnet) was placed orogastrically post-pyloric, in the duodenum using the flexible gastroscope (Olympus America, Center Valley, PA; Fig. 1A). For the 60-mm magnet, after receiving a local anesthetic (ie xylocaine spray), the patient swallowed the magnet with water. Under general anesthesia with continuous hemodynamic and electrocardiographic monitoring, titanium clips were placed laparoscopically as markers in the ileal mesentery site approximately 250 to 350 cm from the ileocecal valve, according to the surgeon’s preference. The magnet was laparoscopically “surfed” through the jejunal lumen with the magnetic positioning device to the ileal marker. The second (proximal) magnet was delivered by gastroscope to the post-SG distal gastric antrum. The positioning device was used to elevate the distal magnet over the transverse colon into side-to-side apposition with the proximal magnet; the 2 magnets aligned, initiating gradual compression of the intervening tissue (Fig. 1B). The positioning device and gastroscope were withdrawn, and the mesenteric defect was closed on the left side with nonabsorbable sutures. During 3 to 4 weeks, the compressed tissues necrose and slough, forming the GI bipartition. The newly fused biofragmentable magnet pair fragments and detaches from the GI site and passes by peristalsis from the body (Fig. 1C). Alimentary content flows through both the pylorus and the GI bipartition into the ileal lumen (Figs. 1D, 2).

Figure 2.

(A) Intraoperative view of the distal linear magnetic gastroileal (MagGI) system magnet placed in the ileum paired side-to-side with the proximal magnet placed in the gastric antrum. (B) Endoscopic view of a patent side-to-side magnetically formed gastroileal anastomosis.

Postoperative care

On postoperative day 1, abdominal x-ray and fluoroscopy using gastrografin were used to confirm magnet pair position. Within 11 to 17 days, patients were reevaluated and required to attend follow-up visits at days 30, 60, 90, 180, 270, and 360. Patients were asked to monitor their stool and report to the surgical office the date(s) of natural magnet passage. If at any time patients experienced a medical issue, they were encouraged to return to the office for assessment.

Endpoints and outcome metrics

Feasibility and performance of the MagGI System at 90 days were confirmed by (1) correct surgical placement and apposition of the MagGI System magnets, (2) expulsion of both magnets with no device-related AEs or surgical reintervention, and (3) a patent MagGI anastomosis visualized fluoroscopically and endoscopically. If MagGI device performance was verified in ≥80% of patients, the primary feasibility endpoint was considered “met.” The primary safety endpoint was ongoing incidence of device- and procedure-related AEs and SAEs based on the Clavien-Dindo (CD) classification of surgical complications (grade I to V).³⁰

Secondary endpoints assessed at 90, 180, and 360 days as interim outcomes in this ongoing study were evolution of weight and BMI; waist circumference; changes in metabolic indicators HbA1_C (%) and blood glucose (mg/dL); and, at 180 days, a custom MagGI procedure satisfaction questionnaire. Weight was analyzed as absolute weight (kg); total weight loss (%TWL: [initial weight − follow-up weight]/[initial weight] × 100); excess weight loss (%EWL: [initial weight − follow-up weight]/[initial weight – ideal body weight] × 100); and the proportion of patients with >5.0% TWL. A recently introduced weight metric to assess revisional MBS procedures was used to capture TWL from the baseline of the primary operation to the follow-up time point of the secondary operation: cumulative percentage TWL ([initial baseline weight of primary procedure – final weight at follow-up time point of secondary procedure] / [initial baseline weight] × 100).³¹

Statistical analysis

Analyses were performed using SPSS statistical software (version 31.0; IBM, Chicago, IL). Primary endpoints were categorical variables, presented as frequencies and percentages. Group differences along categorical variables were assessed using chi-square or Fisher’s exact test. Continuous variables were reported using means and SEM, ranges, 95% CIs, or medians and interquartile ranges. Between-group differences were assessed using independent-samples t-tests. Quantitative measures of change from baseline were evaluated using paired-samples t-tests; tests were 2-tailed and alpha was set at a p value of <0.05.

RESULTS

Patient characteristics

From November 6, 2023, to July 10, 2024, 20 patients (85.0% women; 85.0% White patients) with a mean age of 46.3 ± 1.8 years (range 28.0 to 65.0) underwent MagGI revision of SG. There were no significant differences between groups (stage 1 vs stage 2) with regard to preoperative characteristics (Table 1). Baseline mean absolute weight was 101.0 ± 3.0 kg (74.6 to 127.0); mean BMI was 37.1 ± 0.8 kg/m² (29.1 to 42.3). One patient (5.0%) had T2D before revisional MagGI. All SGs were originally performed 12 months or more before MagGI.

Table 1.

Baseline Characteristics and Perioperative Outcomes of Magnetic Gastroileostomy Revision of Sleeve Gastrectomy

Variable	Combined group, (N = 20)	Stage 1 group, (N = 7)	Stage 2 group (N = 13)	p Value
Preoperative
Earlier sleeve gastrectomy ≥360 d, n (%)	20 (100.0)	7 (100.0)	13 (100.0)	0.999
Females, n (%)	17 (85.0)	5 (71.4)	12 (92.3)	0.270
Ethnicity, White, n (%)	17 (85.0)	6 (85.7)	11 (84.6)	0.999
Age, y, mean ± SEM (range)	46.3 ± 1.8 (28.0–65.0)	44.3 ± 2.4 (37.0–54.0)	47.3 ± 2.4 (28.0–65.0)	0.427
Height, cm, mean ± SEM (range)	164.8 ± 1.7 (154.0–183.0)	166.7 ± 3.4 (157.0–183.0)	163.8 ± 2.0 (154.0–175.0)	0.432
Weight, kg, mean ± SEM (range)	101.0 ± 3.0 (74.6–127.0)	107.9 ± 4.3 (95.3–127.0)	97.3 ± 3.6 (74.6–127.0)	0.088
BMI, kg/m2, mean ± SEM (range)	37.1 ± 0.8 (29.1–42.3)	38.8 ± 0.9 (35.0–42.3)	36.2 ± 1.1 (29.1–41.5)	0.126
Ideal weight, kg, mean ± SEM (range)	68.0 ± 1.5 (59.3–83.7)	69.7 ± 2.9 (61.6–83.7)	67.2 ± 1.6 (59.3–76.6)	0.430
Excess weight, kg, mean ± SEM (range)	33.0 ± 2.2 (10.6–50.4)	38.3 ± 2.3 (27.2–43.3)	30.1 ± 3.0 (10.6–39.3)	0.083
Waist circumference, cm, mean ± SEM (range)	117.9 ± 3.3 (81.0–142.0)	123.6 ± 3.7 (111.0–138.0)	114.8 ± 4.6 (81.0–142.0)	0.214
Weight regain, kg, mean ± SEM (range)	26.4 ± 4.0 (0.9–81.7)	29.6 ± 10.8 (0.9–81.7)	24.7 ± 2.7 (9.2–43.0)	0.575
Type 2 diabetes mellitus, n (%)	1 (5.0%)	0 (0.0%)	1 (7.7%)	0.999
HbA1C, %, mean ± SEM (range)	5.3 ± 0.1 (4.5–6.4)	5.1 ± 0.2 (4.5–5.9)	5.5 ± 0.1 (4.7–6.4)	0.169
Glucose, mg/dL, mean ± SEM (range)	88.5 ± 1.7 (79.0–104.5)	90.6 ± 3.6 (81.1–104.5)	87.3 ± 1.8 (79.0–99.1)	0.373
Perioperative
Operative time, min, mean ± SEM (range)	67.3 ± 4.7 (26.0–114.0)	64.4 ± 6.4 (46.0–93.0)	68.8 ± 6.6 (26.0–114.0)	0.668
Hospital stay, d, mean ± SEM (range)	1.3 ± 0.3 (0.0–7.0)	1.1 ± 0.1 (1.0–2.0)	1.4 ± 0.5 (0.0–7.0)	0.719
Magnet expulsion, d, mean ± SEM (range)	34.5 ± 4.7 (14.0–67.0)	35.8 ± 6.4 (14.0–59.0)	33.0 ± 7.6 (14.0–67.0)	0.778
Magnet expulsion, d, median (interquartile range)	31.0 (22.0–47.5)	31.0 (25.0–58.0)	31.0 (17.8–43.8)	0.999

Perioperative course

Each patient underwent side-to-side MagGI. In 3 stage-1 patients, navigating the 80-mm device through the pylorus to the distal ileum was challenging; therefore, these magnets were placed in the ileal lumen via laparoscopic enterotomy rather than by endoscopically. There were no significant differences between groups in perioperative variables. Mean total MagGI operative time was 67.3 ± 4.7 minutes (26.0 to 114.0). Mean hospital stay was 1.3 ± 0.3 days (0.0 to 7.0); median magnet expulsion occurred at 31.0 days (interquartile range 22.0 to 47.5; Table 1); and patent anastomoses were confirmed radiologically at day 30 in 100.0% of patients.

Feasibility

Feasibility criteria were met in 20 of 20 (100.0%) patients. The MagGI System was correctly positioned and aligned to effect gastroileostomy in all 20 patients. Magnet expulsion occurred without incident in 19 of 20 (95.0%) patients.

Complications

Of the total cohort of 20 patients, 9 (45.0%) experienced at least 1 SAE. Eighty percent (47 of 59) of AEs were mild, CD grade I and II. There were 8 CD grade-III and 4 grade-IV SAEs (20.3%; Table 2). There was 1 case of a 70-mm magnet pair that formed a patent anastomosis, but with tissue partially covering its periphery. That pair was not expelled and underwent endoscopic retrieval at day 168; there were no adverse clinical effects.

Table 2.

Adverse Events by Clavien-Dindo Classification after Magnetic Gastroileostomy Revision of Sleeve Gastrectomy through 360-d Follow-Up

Clavien-Dindo Classification	All patients (n = 20), n (% of total adverse events)
Grade I: deviation from the normal postoperative course without the need for pharmacological treatment or surgical, endoscopic, and radiological intervention. Antiemetics, antipyretics, analgesics, diuretics, and electrolytes, and physiotherapy allowed	22 (37.3)
Grade II: requiring pharmacological treatment with drugs other than those allowed for grade I complications. Blood transfusion and total parenteral nutrition were included	25 (42.4)
Grade III: requiring surgical, endoscopic, or radiological intervention	8 (13.5)
Grade IV: life-threatening complication (including certain CNS complications) requiring intermediate care/ICU management	4 (6.8)
Grade V: death of a patient	0 (0)
Total adverse events	59 (100)

One patient with an SAE presented 8 days postprocedure with abdominal pain; the CT scan appeared negative and the patient was discharged. The same patient presented with similar symptoms 2 days later. Liquid and pneumoperitoneum were seen on CT; surgical exploration revealed peritonitis (though not significant) with a likely small iatrogenic bowel perforation due to the use of standard bowel forceps during exploration. After peritoneal cleaning, an inadvertent tear of the left gastroenterostomy was repaired and the patient later expelled the magnets naturally with no further issues. This perforation occurred during the learning curve of our first 10 patients.

Also during the period of our learning curve, 1 patient experienced an iatrogenic bowel perforation related to the use of standard bowel forceps during the “grasp and slide” magnet placement maneuver (70-mm magnets). In future, different, improved bowel forceps are to be instituted in this procedure to avoid perforation.) On postprocedure day 1, the patient developed symptoms; CT showed pneumoperitoneum and peritonitis with perforation in the small afferent loop. Laparoscopic peritoneal lavage, drainage, and repair resolved the event. One patient (60-mm magnets) sustained an episode of uncontrollable diarrhea, vomiting, and alimentary intolerance which were risks for malnutrition; the patient elected to reverse the gastroileostomy with laparoscopic stapling without sequelae.

The 9 other events resulted in hospitalization. They were gallbladder sludge resulting in cholecystectomy, hypoproteinemia, umbilical hernia, breast carcinoma, anemia, gastroenteritis, food poisoning, peritonitis (unknown origin), and Salmonella infection. These SAEs were not due to the magnet device itself. There was no mortality in the study.

Nutrition

In stage 1 (80-mm magnets), mild nutritional deficiencies in 40.0% of the group were managed conservatively and those that remained within the normal range at 6 months. One patient experienced hypoalbuminemia and 2 episodes of hypoproteinemia on day 168, 1 requiring hospitalization.

Efficacy

Weight

MagGI weight outcomes at 90-, 180-, and 360-day follow-up timepoints are presented in Table 3. Mean absolute weight was reduced from 101.0 ± 3.0 kg at baseline to 83.3 ± 2.9 kg at 180 days and 82.6 ± 4.8 kg at 360 days, representing significant mean changes of 17.2 kg (p < 0.001, n = 17) and 25.4 kg (p < 0.001, n = 7), respectively. Additionally, a significant mean BMI reduction of 9.1 kg/m² (p < 0.001) was observed at 360 days, with a maximum reduction of 14.4 kg/m². In direct correlation, a mean reduction of 19.1 cm in waist circumference was also noted at 360 days (p < 0.001, n = 7). These reductions equated to mean TWLs of 16.9% and 23.4%, and EWLs of 57.5% and 66.0%, at 180 and 360 days, respectively. Ninety-four percent (16 of 17) of patients achieved >5.0% TWL at 180 days, whereas 100.0% (7 of 7) did so at study end (Table 3).

Table 3.

Weight evolution of Magnetic Gastroileostomy Revision of Sleeve Gastrectomy through 360-d Follow-Up for Total Group

Weight parameter	Baseline (N = 20)	90-d follow-up (N = 17)			180-d follow-up (N = 17)			360-d follow-up (N = 7)
	Mean  ± SEM	Mean  ± SEM	Mean change  ± SEM (95% CI)	Within group p value*	Mean  ± SEM	Mean change  ± SEM (95% CI)	Within group, p value*	Mean  ± SEM	Mean change  ± SEM (95% CI)	Within group, p value*
Absolute weight, kg	101.0 ± 3.0	88.5 ± 3.1	11.9 ± 1.3 (9.2–14.7)	<0.001	83.3 ± 2.9	17.2 ± 2.1 (12.7–21.7)	<0.001	82.6 ± 4.8	25.4 ± 3.9 (15.8–34.9)	<0.001
BMI, kg/m2	37.1 ± 0.8	32.5 ± 1.0	4.4 ± 0.5 (3.4–5.3)	<0.001	30.6 ± 1.0	6.2 ± 0.7 (4.7–7.8)	<0.001	29.7 ± 1.3	9.1 ± 1.3 (5.9–12.3)	<0.001
Waist circumference	117.9 ± 3.3	104.8 ± 3.2	10.9 ± 2.0 (6.6–15.3)	<0.001	98.1 ± 2.6	17.7 ± 2.6 (12.1–23.2)	<0.001	104.4 ± 2.8	19.1 ± 4.7 (7.5–30.7)	0.007
Summary weight loss statistics (%)	__	Proportion ≥5.0% TWL	Mean %TWL  ± SEM	Mean %EWL  ± SEM	Proportion ≥5.0% TWL	Mean %TWL  ± SEM	Mean %EWL  ± SEM	Proportion ≥5.0% TWL	Mean %TWL  ± SEM	Mean %EWL  ± SEM
	__	94.1 (16/17)	11.9 ± 1.3	41.7 ± 6.2	94.1 (16/17)	16.9 ± 1.9	57.5 ± 8.2	100.0 (7/7)	23.4 ± 3.2	66.0 ± 8.6

EWL = excess weight loss; TWL, total weight loss.

In Figure 3, evolution of mean absolute weight for stage 1 and stage 2 groups through 270 days is depicted. Figure 4 presents individual patient absolute weight trajectories (n = 7) from baseline of the primary SG through 360-day follow-up with revisional MagGI. Cumulative mean TWL (ie based on absolute weight loss from baseline preprimary SG through 360-day post-MagGI follow-up) was 42.2 ± 2.7%: 5 of 7 (71.4%) patients had lower absolute weights at 1-year after MagGI than those recorded at their weight-loss nadir after primary SG. Figure 5 shows TWL and EWL for the total cohort following revisional MagGI through 360 days of follow-up. Figure 6 shows no significant differences between groups in percentage weight loss out to 270 days of follow-up.

Figure 3.

Evolution of mean absolute weight for stage 1 and stage 2 groups through 270 d (burgundy line = 80-mm magnets; blue line = 60/70-mm magnet).

Figure 4.

Individual patient absolute weight trajectories (n = 7) from baseline of primary sleeve gastrectomy (SG) through 360-d follow-up with revisional magnetic gastroileostomy (MagGI).

Figure 5.

Total weight loss (TWL %) and excess weight loss (EWL, %) for the total cohort following revisional MagGI through 360-d follow-up.

Figure 6.

Total weight loss (TWL) out to 270-d follow-up shows no significant differences between groups (burgundy bar = 80-mm magnets; blue bar = 60/70-mm magnet).

Metabolic indicators

There was 1 confirmed T2D case with a baseline HbA1_C of 6.4%, which fell to 6.2% at 30-day post-MagGI; however, this patient’s data were not available from there on out. Overall cohort mean HbA1_C significantly decreased from 5.3 ± 0.1% at baseline to 4.9 ± 0.1% (p = 0.025, n = 14) at 180 days, and to 4.5 ± 0.2% (p = 0.018, n = 6) at 360 days. At baseline, mean glucose was 88.5 ± 3.5 mg/dL; however, there was no significant change at 180- and 360-day follow-up.

Satisfaction questionnaire

A nonvalidated questionnaire created for the current study to ascertain patient satisfaction levels at 180 days post-MagGI found that 88.0% of patients were satisfied: 13 of 17 (77.0%) reported being “very satisfied” with the GT magnet system for treatment of their obesity, with another 2 of 17 (12.0%) reporting being “somewhat satisfied.” One patient (6.0%) reported being “somewhat dissatisfied.” Of the 17 of 20 patients who completed the questionnaire, all 17 reported that they would “recommend the magnet system for obesity treatment to their family and friends” (Fig. 7).

Figure 7.

Results of a satisfaction questionnaire administered at 6 mo post-sleeve gastrectomy (SG) revisional magnetic gastroileostomy (MagGI; n = 17).

DISCUSSION

The foundation of the current multicenter MagGI series was a preclinical study of MCA gastrojejunostomy using 3 trial magnet sizes in a porcine model.²² The preclinical finding that 80-mm magnets created fully patent, well-vascularized GI anastomoses with minimal inflammation relative to sutured enterotomy sites informed our selection of 80-mm magnets for the subsequent stage-1 FIH pilot study of patients with suboptimal SG who underwent revisional MagGI. In this group (n = 7), feasibility criteria were met per protocol, including magnet placement, expulsion, and radiologically confirmed patent bipartition anastomosis creation.²⁷ In response to 3 early instances of challenging maneuvering of the 80-mm magnets through the pylorus, stage-2 patients (n = 13) received shorter 60- and 70-mm linear magnets to facilitate passage. As a result, at 6-month follow-up, 12 of 13 stage-2 distal magnets moved feasibly through the pylorus without reintervention. In addition, as a precaution against potential nutritional deficits, stage-2 distal magnets were positioned 300 to 350 cm (rather than 250 cm) from the cecum to create a longer common channel. Also, the 60-mm magnet is now swallowable, making the procedure gastroscopy free before laparoscopy.

Magnets were placed successfully in the full cohort (n = 20); 19 of 20 achieved primary technical endpoint success, demonstrating strong MagGI System feasibility performance. The mean operative time of 67.3 ± 4.7 minutes and hospital stay of 1.3 ± 0.3 days were efficient. Median magnet expulsion duration of 31.0 days (Note: abdominal x-rays are performed once a week, and therefore, the expulsion times are overestimated) is representative of completion of patent MCA formation in 100.0% of the cohort. AEs in 80.0% of the cohort were mild, SAEs were resolved effectively, and there was no mortality in the study. No SAEs were related to the LMAS device. During the course of the study, a trend was observed toward an association between 80-mm magnets and a higher number of AEs (60.0% of all AEs) was observed; however, there was no statistically significant difference in AE volume between 80-, 70-, and 60-mm magnets.

Revisional MBS is generally associated with a twice the incidence of severe complications compared with primary procedures.^32,33 However, based on the robust safety profile seen with gradual staple or suture-free MCA in 7 clinical magnetic duodenoileostomy studies (greater than 100 total patients),^18-21,23-25 SG-revisional MagGI is expected to be minimally susceptible to major anastomotic complications: In the ongoing study in the current patient group at 6- to 12-month interim follow-up, there were no anastomotic leaks, bleeding, infection, or obstruction. Although studies of the nonmagnetic procedure, very similar to MagGI, the single-anastomosis sleeve ileal (SASI), are scant for the indication of SG revision, anastomotic safety issues have been observed. In a report of SG-revisional SASI (n = 28), 1 patient 7 months postprocedure experienced severe vomiting associated with anastomotic stenosis.³⁴ A separate study in 44 post-SG SASI patients included 4 patients with intraluminal bleeding at the gastroileal anastomosis (CD grade II), all of which were managed conservatively. The authors suggested that the anastomotic challenges arose from the use of mechanical stapling to form the gastroileal anastomosis because varying tissue thicknesses make appropriate staple load choice difficult to determine.³⁵ The authors strongly suggest that PPIs be maintained in patients to avoid ulcerations.

In terms of potential postoperative nutritional issues, stage-1 (80-mm magnets) mild deficiencies seen in 40.0% of the group were managed conservatively and normalized at 6 months. The sole stage-1 MagGI patient who had experienced hypoalbuminemia and episodes of hypoproteinemia on day 168, later (day 245) sustained a grade-II hypoproteinemia leading to hospitalization; this was treated and resolved without sequelae. In comparison, studies of SASI have reported hypoalbuminemia and hypoproteinemia in high percentages (8.3% to 9.7%, respectively), resulting in takedown of the anastomosis.^36-38 Cognizant of these potential issues, in the 13 stage-2 MagGI patients, use of shorter 60- and 70-mm magnets and creation of a longer common channel (at least 300 cm) were associated with the absence of hypoalbuminemia and hypoproteinemia.

The current total SG-revisional MagGI cohort mean TWL and EWL at 6 months (n = 17) of 16.9 ± 1.9% and 57.5 ± 8.2%, respectively, continued to increase through the 1-year interim endpoint to 23.4 ± 3.2% and 66.0 ± 8.6% (n = 7). Post-SG MagGI improvements in TWL and EWL, respectively, are similar to the systematic review and meta-analytic outcomes of 3 SA SG-revisional MBS operations, including SASI (18.0% to 20.0% and 45.0% to 50.0%; 24.0% to 36.0%, and 60.0% to 65.0%^39-41), SA duodenoileostomy with SG (32.0% EWL; 69.6% EWL⁴²), and 1-anastomosis gastric bypass (NR; 56.9% to 73.4% EWL⁴³). The current mean study CTWL of 42.4 ± 2.9% at the 1-year endpoint (n = 7) was roughly equivalent to the weight nadir of the earlier primary SG, which was intended as a stand-alone procedure. These outcomes suggest that, similar to existing SG revision options, MagGI is a procedure with a promising efficacy profile capable of halting post-SG weight regain and reigniting substantial MBS weight loss.

Among the 17 patients who responded to the satisfaction questionnaire at 180 days, 88.0% expressed confidence in MagGI as a procedure for achieving weight loss, and 17 of 17 patients stated they would recommend it to family and friends. The high satisfaction rating was somewhat surprising, given that some patients experienced grade III to IV complications. This possibly reflects patient selection bias toward highly motivated individuals, the successful resolution of all complications without permanent sequelae, and visible weight-loss outcomes that may have exceeded patients’ earlier SG results. MagGI holds potential to offer patients with obesity who are wary of surgery a markedly effective treatment option that is not a high-risk surgical procedure.

Limitations

The FIH study expansion from the initial 7 patients was intentionally restricted to 20 patients to minimize risk. Pooling stages 1 and 2 may limit the conclusions able to be drawn about the study. The cohort size and observational study design, however, preclude predictive conclusions, and therefore, the findings of this report are preliminary. The study lacks a control group or comparison group to assess the device and procedure directly against procedures with sutured or stapled anastomosis formation technique. Additional research in larger cohorts with level-1 evidence designs is required to definitively demonstrate safety and efficacy.

CONCLUSIONS

There is an increasing emphasis on developing radically minimally invasive, low-risk, effective metabolic bariatric therapies tailored for individual patients. Linear magnets used in MagGI were feasibly placed and formed patent anastomoses in 100.0% of patients. Practicing the magnet “grasp and slide” maneuver using the improved bowel forceps is recommended to reduce the risk of an intestinal tear or perforation, particularly during the surgical learning curve. At 1-year interim follow-up, nearly all of the nadir weight loss attained by the primary SG had been recaptured in TWL of 23.4 ± 3.2%, EWL of 66.0 ± 8.6%, and 42.4 ± 2.9% CTWL. Longer-term research is needed to validate these early findings.

Author Contributions

Conceptualization: Gagner

Funding acquisition: Gagner

Investigation: Gagner, Almutlaq, Ribeiro, Vally-Adam, Couto, Abril, Rodriguez

Methodology: Gagner

Project administration: Gagner, Gnanhoue

Supervision: Gagner

Writing – review & editing: Gagner

Resources: Almutlaq, Ribeiro

Data curation: Gnanhoue

Validation: Gnanhoue

Writing – original draft: Buchwald