|
|
Back to 2018 Program and Abstracts
DOES TREATMENT OF THE HIATUS INFLUENCE THE OUTCOMES OF MAGNETIC SPHINCTER AUGMENTATION FOR CHRONIC GERD?
Matias Mihura*1, Sandra Blitz3, Candice L. Wilshire1, Anee S. Jackson1, Alexander S. Farivar 1, Ralph W. Aye1, Christy M. Dunst2, Brian E. Louie1
1Thoracic Surgery, Swedish Cancer Institute, Seattle, WA; 2MIS Surgery, The Oregon Clinic, Portland, OR; 3Sandra Blitz Statistics, Toronto, ON, Canada
BACKGROUND: Hiatal dissection, restoration of esophageal intra-abdominal length and crural closure are key components of a successful antireflux operation when combined with fundoplication. During magnetic sphincter augmentation (MSA), the hiatus is minimally dissected prior to placement thus relying on native anatomic synergy. The necessity of addressing these components prior to MSA implantation has been questioned in an attempt to improve on objective measures of reflux control. We aimed to compare outcomes of MSA between groups with differing hiatal dissection and closure.
METHODS: We retrospectively reviewed 249 patients who underwent MSA from 2009-2016. Of those, 208 had at least 1-year follow-up and 190 postoperative pH testing. Operative records and/or videos were reviewed to categorize patients into 1 of 4 groups based on hiatal treatment: minimal dissection (MD) - no dissection of the hiatus; crural closure (CC) - minimal posterior dissection with suture closure; formal crural repair (FC) - hiatal dissection, return of length and closure; and extensive dissection without closure (ED) - dissection hiatus, return of length but without closure. The primary outcome was normalization of postoperative DeMeester (DMS) score (≤14.72). Univariable and multivariable logistic regression was used to assess which preoperative predictors achieved normalization. Multivariable models included significant univariable factors, the assigned hiatal treatment group and known predictor.
RESULTS: Of the 190 patients, MD was used in 79 (42%); FC in 39 (21%); CC in 38 (20%) and ED in 34 (18%). Normalization of the DMS occurred in 95 (50%) patients, with MD achieving normalization in 43/79 (54%); FC in 22/39 (56%); CC in 17/38 (45%) and ED 13/34 (38%). Patient characteristics are shown in Table 1.
Univariable analysis identified neither treatment group nor Hill grade were significant; however, hiatal hernia, preoperative DMS, defective LES and gender were significant negative predictors of outcome.
In multivariable analyses, FC was more likely to normalize DMS than MD (OR=2.76, 95%CI=1.04-7.39, p=0.04) or ED (OR=4.65, 95%CI=1.47-14.73, p=0.009). Furthermore, no hiatal hernia was more likely to normalize DMS than a hernia up to 2 cm (OR=0.33, 95%CI=0.15-0.76, p=0.009) or a hernia >2cm (OR=0.14, 95%CI=0.06-0.35, p<0.0001). FC was most likely to achieve normalization regardless of hernia size. (Figure 1) There was a trend that a defective LES was less likely to achieve a normalized DMS (OR=0.43, 95%CI=0.18-1.02, p=0.06).
CONCLUSIONS: Formal hiatal dissection, restoration of intra-abdominal esophageal length and crural closure prior to implantation of a MSA device provides the greatest likelihood of achieving a normal postoperative DeMeester score. The presence of a hiatal hernia or defective LES preoperatively appears to negatively impact MSA outcomes.
Table 1: Preoperative Patient Characteristics
| Variable | Category | Minimal Dissection (MD) | Crural Closure (CC) | Formal Crural Repair (FC) | ED w/o Closure (ED) | | N | | 79 | 38 | 39 | 34 | | Age (yrs) | Median | 53.0 (42.0-60.8) | 50.5 (38.0-62.0) | 53.5 (43.1-60.8) | 49.7 (35.6-56.7) | | Gender | Male | 40 (50.6%) | 21 (55.3%) | 19 (48.7%) | 21 (61.8%) | | | Female | 39 (49.4%) | 17 (44.7%) | 20 (51.3%) | 13 (38.2%) | | BMI | Median | 26.1 (24.3-29.0) | 27.0 (23.4-29.6) | 27.2 (25.5-30.1) | 27.9 (25.7-29.6) | | Esophagitis | None | 50 (63.3%) | 21 (55.3%) | 17 (43.6%) | 24 (70.6%) | | | A | 18 (22.8%) | 14 (36.8%) | 11 (28.2%) | 7 (20.6%) | | | B+ | 11 (13.9%) | 3 (7.9%) | 11 (28.2%) | 3 (8.8%) | | Hill Grade | 1 | 15 (19.0%) | 4 (10.5%) | 4 (10.3%) | 11 (32.4%) | | | 2 | 19 (24.1%) | 7 (18.4%) | 11 (28.2%) | 4 (11.8%) | | | 3 | 23 (29.1%) | 15 (39.5%) | 9 (23.1%) | 5 (14.7%) | | | 4 | 11 (13.9%) | 9 (23.7%) | 9 (23.1%) | 6 (17.6%) | | | Missing | 11 (13.9%) | 3 (7.9%) | 6 (15.4%) | 8 (23.5%) | | Hiatal Hernia Size | None | 44 (55.7%) | 10 (26.3%) | 6 (15.4%) | 15 (44.1%) | | | ≤ 2 cm | 21 (26.6%) | 14 (36.8%) | 8 (20.5%) | 8 (23.5%) | | | 2 - 3 cm | 11 (13.9%) | 10 (26.3%) | 14 (35.9%) | 6 (17.6%) | | | > 3cm | 3 (3.8%) | 4 (10.5%) | 11 (28.2%) | 5 (14.7%) | | DeMeester Score | Median | 33.3 (22.5-50.9) | 46.1 (33.2-61.7) | 32.4 (27.9-55.2) | 39.4 (30.0-50.3) | | % time pH > 4.5 | Median | 9.7 (6.8-15.1) | 14.2 (9.7-19.7) | 10.3 (7.6-15.3) | 11.4 (8.6-13.7) | | Defective Sphincter | No | 16 (20.3%) | 6 (15.8%) | 4 (10.3%) | 9 (26.5%) | | | Yes | 54 (68.4%) | 30 (78.9%) | 34 (87.2%) | 22 (64.7%) | | | Missing | 9 (11.4%) | 2 (5.3%) | 1 (2.6%) | 3 (8.8%) | | GERD-HRQL | Median | 24.0 (18.0-31.0) | 20.0 (14.0-27.0) | 28.0 (22.0-33.0) | 27.0 (24.0-32.0) |
 Figure 1: Outcomes by Treatment Group and Hernia Size
Back to 2018 Program and Abstracts
|
