Laparoscopic Hill Repair: 25-Year Follow-Up

Surgical Endoscopy and Other Interventional Techniques https://doi.org/10.1007/s00464-018-6150-z

Laparoscopic Hill repair: 25-year follow-up

Yeseul Park1 · Ralph W. Aye1 · Jeffrey R. Watkins1 · Alex S. Farivar1 · Brian E. Louie1

Received: 29 July 2017 / Accepted: 21 March 2018 © Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract Background The open Hill repair for gastroesophageal reflux disease and hiatal hernia is remarkably durable, with a median 10-year reoperation rate of only 3% and satisfaction of 93%. No long-term data exist for the laparoscopic Hill repair (LHR). Methods Patients who underwent primary LHR at Swedish Medical Center for reflux and/or hiatal hernia at least 5 years earlier (1992–2010) were identified from an IRB-approved database. There were 727 patients who met inclusion criteria, including 648 undergoing repair for reflux and 79 for paraesophageal hernia. Two questionnaires were administered via mail to evaluate long-term quality of life using validated GERD-HRQL, Swallowing score, and global satisfaction score. Outcomes were defined by GERD-HRQL score, Swallowing score, resumption of proton pump inhibitor (PPI) therapy, need for reoperation, and global satisfaction with overall results. Results Two hundred forty-two patients completed and returned the survey (226 lost to follow-up, 90 deceased, 3 denied undergoing LHR, 166 non-responders), of which 52% were male. The average age at the time of surgery was 49.5 years. Median follow-up was 18.5 years (range 6.2–24.7). The average GERD-HRQL score (7.1) and the average Swallowing score (39.9) both indicated excellent symptomatic outcomes. 30% of patients are using daily PPIs. 24 patients (9.9%) required reoperation for failure during the follow-up period, 21 in the reflux group and 3 in the paraesophageal hernia group. Overall, 85% reported good to excellent results, and 76% would recommend the operation. Conclusion LHR shows excellent long-term durability and quality of life similar to the open Hill repair, with 85% good to excellent results at a median follow-up of 19 years and a reoperation rate under 10%. It is surmised that Hill suture fixation of the gastroesophageal junction to the preaortic fascia may confer unique structural integrity compared to other repairs.

Keywords Hill · Laparoscopy · Long-term · Antireflux · Paraesophageal hernia · Fundoplication

The Hill repair for gastroesophageal reflux disease (GERD) laparoscopic antireflux surgery are limited, and there are no re-establishes the gastroesophageal flap valve and anchors published long-term studies of the laparoscopic Hill repair the gastroesophageal junction (GEJ) to the preaortic fascia, (LHR). At a time of increasing concern for the long-term thus differentiating it from other antireflux operations, which use of proton pump inhibitors to treat reflux disease, it is are based on fundoplication. The open Hill repair for GERD incumbent on surgeons to know and report the long-term and hiatal hernia has been shown to be remarkably durable, results of antireflux surgery. We sought to assess outcomes with a 91% satisfaction rate and a reoperation rate of only of LHRs done at least 5 years previously. 2% at up to 25-year follow-up [1]. Moreover, Hill sutures, when added to laparoscopic Nissen fundoplication for paraesophageal hernia, dramatically decrease long-term anatomic Materials and methods recurrence rates and may have independent structural advan- tages [10]. However, publications on long-term outcomes of A retrospective review of 822 patients undergoing LHR at Swedish Medical Center from 1992 to 2010 was performed. 95 patients were excluded: 14 underwent LHR as * Ralph W. Aye a second antireflux operation, 36 required conversion to [email protected] an open repair, and 45 had less than 5-year follow-up. The 1 Division of Thoracic Surgery, Swedish Cancer Institute remaining 727 patients underwent a primary LHR, with and Medical Center, 1101 Madison, Suite #900, Seattle, 648 (89.1%) having the repair for GERD and 79 (10.9%) for WA 98104, USA

Vol.:(0123456789)1 3 Surgical Endoscopy paraesophageal hernia (PEH). The Swedish Medical Center of subgroups and various studies utilized the T test, with the Institutional Review Board approved this project. significance level set at ap < 0.05. The Hill repair was performed laparoscopically in all cases with a standardized technique which included complete mobilization of the hiatus, trimming of the hernia Results sac except along the lesser curvature, non-division of the short gastric vessels, closure of the hiatus, and placement Of the 727 patients who underwent a primary LHR, 90 were of four non-absorbable repair sutures each of which passed deceased and 3 denied (in error) ever having LHR. Of the sequentially through the anterior sling fibers of the lower remaining 634, 226 could not be contacted and were accord- esophageal sphincter, then the posterior sling fibers, then the ingly categorized as lost to follow-up, leaving 408 available preaortic fascia just superior to the celiac axis. The sutures for survey. With 166 declining to participate, the remaining were partially tied over a 45–48 Fr bougie and then adjusted 242 completed surveys for an overall follow-up of 38% and utilizing single-port water perfused manometrics. a survey response rate of 59% of those available (Fig. 1). All patients received the same structured survey via mail. These 242 responders were the basis of further analysis. Patient consent was obtained before administering ques- Two hundred thirteen (88%) of 242 participants under- tionnaires. A staff member (YP) contacted and interviewed went LHR for reflux and 29 (12%) for PEH. On aver- each of the non-responders using a scripted format over tel- age, PEH patients were older (55.2 years) than GERD ephone. If requested by the patient, the interviewer sent the (47.4 years) at the time of surgery (p < 0.01). Additionally, survey via email. the PEH patients had a significantly shorter median follow- Two questionnaires were implemented to appraise long- up of 12.3 years relative to the 18.6-year median of the reflux term quality of life after LHR. The first investigated the patients (p = 0.03, Table 2). During the follow-up period, patient’s symptoms of heartburn, dysphagia, odynophagia, 16% underwent endoscopic dilation. 24 patients (9.9%; 21 and bloating as well as antireflux medication interfering GERD, 3 PEH) underwent reoperative repair for failure, giv- with daily life using a six-point validated Gastroesophageal ing a reoperation rate of 9.9% for the GERD subgroup and Reflux Disease-Health Related Quality of Life (GERD- 10.3% for the PEH subgroup. A significantly higher propor- HRQL) scale (range 0–50, 0 = best) [2]. It included a sat- tion of women than men underwent reoperation (p = 0.01, isfaction score with regard to current condition (e.g., satis- Table 1). Recurrences requiring reoperation were considered fied, neutral, or dissatisfied), type and frequency of reflux failures (see separate analysis). remedies taken more than once a week (antacids, H 2 block- The median follow-up for the remaining 218 was ers, and/or proton pump inhibitors [PPIs]), and any further 18.5 years (range 6.2–24.7), with 113 (52%) males and a esophageal procedures undertaken following LHR (dilations mean age of 49.5 years at surgery. Participants obtained and reoperations). Two five-point Likert scales evaluated an average 7.1 GERD-HRQL score (range 0–50, 0 = best). global assessment and the likelihood of recommending LHR More than 24% of patients attained the best possible score to family members with a similar condition. of zero, e.g., complete absence of heartburn, dysphagia, The second survey utilized a Swallowing score (range odynophagia, bloating, or medical therapy. The average 0–45, 45 = best) described by Dakkak and Bennett [3]. Swallowing score was 39.9 (range 0–45). 61% reported the Patients ranked difficulty with swallowing nine specified maximum score of 45. More than half of the participants food consistencies. Foods not eaten out of preference were conveyed little to no avoidance of food due to dysphagia, treated as problematic, unless they specifically mentioned necessity to eat slower relative to others, or regurgitation. no issues with foods of identical consistency. Three sup- One hundred eighteen patients (46%) reported utilizing plementary questions addressed avoidance of food due to antireflux medication more than once a week (21 antacids, dysphagia, the need to eat slower than others, and regurgi- 21 H2 blockers, 76 PPIs, 13 multiple), but only 66 (30%) tation using five-point Likert scales, ranging from “absent” are on daily PPIs (Fig. 2). As a whole, 85% reported good to “occurs daily.” to excellent outcomes (Fig. 3), and 76% would recommend Surveys were administered independent of current use or the operation to a family member with a similar condition non-use of medication, e.g., those who had resumed medi- (Fig. 4). On a three-point scale, the majority of respondents cation were instructed to answer according to their current (67%) felt satisfied with their current condition, 17% felt condition on medication. We compared our data with short- neutral, and 16% felt dissatisfied. term results of LHR and long-term results of open Hill repair Responders who underwent LHR for GERD versus PEH and to published reports of other laparoscopic antireflux indicated similar quality of life results (Table 2). operations with long-term follow-up. Comparing the primary group to those who underwent Mean and median calculations implemented standard reoperation revealed no differences with regard to most qual- statistical analysis. All categorical variable juxtapositions ity of life responses, including mean GERD-HRQL scores

1 3 Surgical Endoscopy

Fig. 1 Study design. The break- down of identified patients into exclusions and inclusions

Table 1 Demographics and Primary (n = 218) Reoperatives (n = 24) p Value Total (n = 242) quality of life Male (%) 113 (52) 6 (25) 0.012* 119 (49) Mean age, years (range) 49.9 ± 10.9 45.3 ± 8.7 0.075 48.3 (22–78) Median follow-up, years (range) 18.5 ± 5.5 19.2 ± 5.8 0.640 18.5 (6.2–24.7) GERD-HRQL score 7.1 ± 9.0 8 ± 8.4 0.603 7.2 ± 8.9 Swallowing score 39.9 ± 10.0 35.1 ± 12.8 0.018* 39.4 ± 9.4 Reflux medication use (%) 101 (46) 9 (38) 0.412 110 (45) Daily PPIs (%) 66 (30) 7 (29) 0.900 73 (30) Dilation procedure (%) 35 (16) 6 (25) 0.269 41 (17) Avoids food daily (%) 14 (7) 5 (21) 0.014* 19 (8) Eats slower daily (%) 41 (19) 8 (33) 0.011* 49 (20) Regurgitates daily (%) 2 (1) 2 (8) 0.192 4 (2)

For GERD-HRQL, lower = better. For Swallowing, higher = better *Significant

Table 2 LHR for GERD vs. PEH GERD (n = 213) PEH (n = 29) p Value

Mean age (years) 47.4 ± 11.9 55.2 ± 11.1 0.001* Median follow-up (years) 18.6 ± 5.4 12.3 ± 4.8 0.035* Reoperation rate (%) 21 (10) 3 (10) 0.935 GERD-HRQL score 7.2 ± 9.0 6.9 ± 8.4 0.852 Dysphagia score 39.6 ± 10.3 37.9 ± 10.9 0.369 Daily PPIs (%) 67 (31) 6 (21) 0.238

For GERD-HRQL, lower = better. For Swallowing, higher = better *Significant

(p = 0.67; Table 1). However, reoperative patients obtained Fig. 2 Antireflux medication type and frequency resumed by primary a lower mean Swallowing score (35.1) than those who had LHR patients. Respondents concurrently utilizing more than one rem- a primary repair (p = 0.02). Consistently, they had more edy were accounted for multiple times

1 3 Surgical Endoscopy

More than 2500 LHRs have been performed globally since its introduction in 1992 [4]. Short-term clinical results were reported as good to excellent in 92% at 8 months following surgery [5]. The disclosed daily PPI usage ranges from 4 to 15% at 1-year follow-up [5–7]. A 15-month follow- up analysis that applied the same quality of life assessments exhibited an average 8.1 GERD-HRQL score and 38.1 Swal- lowing score [7]. The Hill repair is unique among antireflux procedures in securing the GEJ itself to the preaortic fascia intra- abdominally. The resistance to axial forces by GEJ fixation potentially reduces herniation [4], the most common cause of failure following Nissen fundoplication. In addition, this firm fixation allows the Hill repair to be utilized in the case of short esophagus, typically without the need for Collis Fig. 3 Global satisfaction with overall LHR outcomes of primary gastroplasty [8]. Furthermore, in contrast to LNF, LHR does patients not significantly raise lower esophageal sphincter pressure, being based instead on reconstruction of the gastroesophageal valve, thus making it applicable for patients with poor esophageal motility [9]. Short-term results with the LHR are equivalent to the laparoscopic Nissen fundoplication (LNF) in clinical and objective parameters measuring clinical suc- cess and repair failure at 1-year follow-up [6], but they have not been compared long-term. The structural components of Hill repair may have par- ticular durability. In a large multi-institution retrospective study of over 1100 patients undergoing open Hill repair with a 10-year mean follow-up, 91% rated their outcomes as good to excellent, with a reoperation rate of 1.9 and a 23% resumption of anti-secretory medication [1]. Furthermore, the addition of Hill sutures to laparoscopic Nissen repair for paraesophageal hernia achieved a substantial reduction in 60-month anatomic recurrences from 45 to 5% [10], perhaps due to fixation of the GEJ to the preaortic fascia, which may Fig. 4 Patient’s likelihood of recommending LHR to family members unload axial stress from the fundoplication. The strength suffering from comparable condition of the Hill repair may come at a small price, as the dilation rate of 16% is a bit high and likely reflects the “non-floppy” smaller diameter calibration of the repair over a 45–48 Fr frequent dysphagia symptom incidences of avoiding food bougie. (p = 0.01) and eating slower (p = 0.01). The higher failure rate in this study compared to open Hill repair is not readily explained, but likely in part reflects the early learning curve transition from open to laparoscopic repair. It may also be that open repair results in more robust Discussion scarring and adhesions, which may be an advantage in fix- ing the repair in position, as reoperations following laparo- This is one of the few studies evaluating long-term outcomes scopic repair are often notable for limited adhesions. It is of laparoscopic antireflux surgery beyond 10 years, and the also possible that the quality and depth of suture placement only one specifically evaluating LHR. It confirms that the with open repair may be sturdier, though evidence for this excellent early clinical outcomes obtained with the LHR are is lacking. sustained long-term, with 85% good to excellent subjective The results of this study compare favorably to other outcomes, a reoperation rate for failure of only 10%, and long-term studies of laparoscopic antireflux surgery. The only 30% resumption of daily PPIs in the remaining 90% at only other publication with a median follow-up beyond a median follow-up of 18.5 years. 10 years is by Robinson et al., a 20-year follow-up of LNF

1 3 Surgical Endoscopy

[11]. This study showed similar good long-term results. The References reoperative rate of 18% was higher than the 10% reported in this study, but resumption of PPIs at 43% was probably 1. Aye RW, Rehse D, Blitz M, Kraemer SJ, Hill LD (2011) The equivalent, as it was not stratified by frequency of use and Hill antireflux repair at 5 institutions over 25 years. Am J Surg 201:597–602 included patients who had undergone reoperation for failure. 2. Velanovich V (2007) The development of the GERD-HRQL Moreover, the 30% resumption of PPI therapy in this current symptom severity instrument. Dis Esophagus 20:130–134 study excludes the 10% who underwent reoperation. Patients 3. Dakkak M, Bennett J (1992) A new dysphagia score with objec- indicated similar long-term satisfaction rates following both tive validation. J Clin Gastroenterol 14:99–100 4. Aye RW, Gupta A (2015) The Hill antireflux operations repair procedures, with 75% of LNF and 71% of LHR respondents and its variants. In: Swanstrom LL, Dunst CM (eds) Antireflux expressing maximal satisfaction with the operation. Several surgery. Springer, New York, NY, pp 117–126 publications with 10-year follow-up after laparoscopic fun- 5. Aye RW, Hill LD, Kraemer SJ, Snopkowski P (1994) Early results doplication report reoperative rates of 2–4%, with daily PPI with the laparoscopic Hill repair. Am J Surg 165(5):542–546 6. Aye RW, Swanstrom LL, Kapur S, Buduhan G, Dunst CM, Knight consumption ranging from 3 to 26% [12–14]. A, Malmgren JA, Louie BE (2012) A randomized multiinstitu- Limitations of the study include its retrospective nature, tion comparison of the laparoscopic Nissen and Hill repairs. Ann the purely subjective assessment of outcomes, and the Thorac Surg 94(3):951–958 number of patients lost to follow-up. The overall follow- 7. Schneider AM, Aye RW, Wilshire CL, Farivar AS, Louie BE (2017) Tri-comparison of laparoscopic nissen, hill, and nissen-hill up of 33% is not ideal, though it is not surprising with a hybrid repairs for uncomplicated gastroesophageal reflux disease. median follow-up of over 18 years, and the survey response J Gastrointest Surg 21:434 rate of 59% of those available for follow-up was higher 8. Jobe BA, Aye RW, Deveney CW, Domreis JS, Hill LD (2002) than expected. Likewise, there was no attempt to correlate Laparoscopic management of giant type III hiatal hernia and short esophagus. Objective follow-up at three years. J Gastrointest Surg patients’ subjective responses to objective outcomes and it 6(2):181–188 is known that symptomatic appraisal of GERD correlates 9. Aye RW, Mazza DE, Hill LD (1997) Laparoscopic Hill repair in imperfectly with objective pathologic reflux in antireflux patients with abnormal motility. Am J Surg 173(5):379–382 surgery patients [15]; nonetheless, quality of life is of high 10. Levy G, Aye RW, Farivar AS, Louie BE (2017) A combined nissen plus hill hybrid repair for paraesophageal hernia improves importance to the patient, and thus a valid metric for long- clinical outcomes and reduces long-term recurrences compared term benefit to the patient [16]. with laparoscopic nissen alone. J Gastrointest Surg 21(1):121–125 This study encompasses the largest reported follow-up of 11. Robinson B, Dunst CM, Cassera MA, Reavis KM, Sharata A, laparoscopic antireflux surgery beyond 10 years. It confirms Swanstrom LL (2015) 20 years later: laparoscopic fundoplication durability. Surg Endosc 29(9):2520–2524 that laparoscopic antireflux surgery is durable and that the 12. Salminen PT, Hiekkanen HI, Rantala AP, Ovaska JT (2007) Com- majority of patients achieve long-term symptomatic benefit parison of long-term outcome of laparoscopic and conventional and freedom from the use of daily PPIs. These results pro- Nissen fundoplication. Ann Surg 246(2):201–206 vide a counterpoint to the oft-heard impression that antire- 13. Dallemagne B, Weerts J, Markiewicz S, Dewandre JM, Wahlen C, Monami B, Jehaes C (2006) Clinical results of laparoscopic flux surgery progressively declines and does not provide fundoplication at 10 years after surgery. Surg Endosc 20:159–165 lasting benefit, and give the surgeon reasonable confidence 14. Kornmo TS, Ruud TE (2008) Long-term results of laparoscopic in recommending antireflux surgery as an alternative to Nissen fundoplication due to gastroesophageal reflux disease. long-term PPI use. A ten year follow-up in a low volume center. Scand J Surg 97:227–230 Acknowledgements 15. Khajanchee YS, O’Rourke RW, Lockhart BA, Patterson EJ, This work was partially supported by funding for Hansen PD, Swanstrom LL (2002) Postoperative symptoms and research personnel from the Ryan Hill Research Foundation and the failure following antireflux surgery. Arch Surg 137(9):1008–1014 Foundation for Surgical Fellowships. 16. Borgaonkar MR, Irvine EJ (2000) Quality-of-life measurement in gastrointestinal and liver disorders. Gut 47:444–454 Compliance with Ethical Standards

Disclosures The authors Yeseul Park, Ralph W Aye, Jeffrey R Wat- kins, Alex S Farivar, and Brian E Louie have no conflicts of interest or financial ties to disclose.

1 3