George Makdisi, MD, Francis C. Nichols, III, MD, Stephen D. Cassivi, MD, Dennis A. Wigle, MD, PhD, K. Robert Shen, MD, Mark S. Allen, MD, and Claude Deschamps, MD Division of General Thoracic Surgery, Mayo Clinic College of Medicine, Rochester, Minnesota
Background. Minimally invasive procedures have become common, and more reoperations for failed antireflux procedures are performed laparoscopically. We wanted to study the outcomes of laparoscopic reoperations for reflux. Methods. Medical records of all patients who underwent reoperation without esophageal resection after previous antireflux procedures between January 2000 and October 2012 were reviewed. Results. Seventy-five patients were included in this report: 56 (77%) women and 19 (23%) men. Median age was 58 years. The previous operation was laparoscopic antireflux procedures in 65 (87%) patients. The median interval between the last antireflux procedure and laparoscopic reoperation was 42 months. The median body mass index (BMI) was 28.7. All patients were symptomatic. Intraoperative findings included recurrent hiatal hernia in 47 (63%) patients, incompetent fundoplication in 14 (19%) patients, tight fundoplication in 8 (11%) patients, and tight crura in 2 (3%) patients. Laparoscopic Nissen fundoplication was performed in 57 (76%)
patients, partial posterior fundoplication was performed in 12 (16%) patients, partial anterior fundoplication was performed in 3 (4%) patients, removal of crural stitches was performed in 2 patients, and a combination of partial posterior fundoplication and removal of crural stiches was performed in 1 patient. Complications occurred in 13 (15%) patients. Improvement in symptoms was observed in 70 (93%) patients in early postoperative follow-up and in 59 (78%) patients in long-term follow-up. Functional results were classified as excellent in 59 (78%) patients, good in 6 (7%) patients, fair in 7 (8%) patients, and poor in 3 (4%) patients. Conclusions. Laparoscopic reoperation for failed antireflux operations is a complex procedure, but it is safe and effective in selected patients. Reoperation after a failed antireflux repair results in excellent or good functional status in a majority of patients, but these results may deteriorate over time.
A
failed antireflux procedures and analyzed factors affecting postoperative morbidity and short- and longterm outcomes.
ntireflux operations have a failure rate of 5% to 30% [1–7] after open and laparoscopic antireflux procedures. The failure rate after laparoscopic repair may worsen as these patients are followed for a longer period. Although not all patients with recurrent reflux symptoms or obstructive symptoms require reoperation, it is estimated that 3% to 6% of patients who undergo primary antireflux operations will eventually require reoperation [5, 8]. Reoperation for failed antireflux repair is usually performed through transabdominal or transthoracic techniques. With the development of laparoscopic techniques, more reoperations for failed antireflux procedures are being performed through minimally invasive techniques [1–3, 5]. Early reports after laparoscopic reoperations are encouraging [6, 9, 10], but little is known regarding long-term outcomes. We reviewed our experience with minimally invasive laparoscopic repair for Accepted for publication May 13, 2014. Presented at the Fiftieth Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Jan 25–29, 2014. Address correspondence to Dr Deschamps, Division of General Thoracic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: [email protected].
Ó 2014 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2014;98:1261–6) Ó 2014 by The Society of Thoracic Surgeons
Patients and Methods Between January 2000 and October 2012, 2,402 patients underwent antireflux operations at the Mayo Clinic in Rochester, Minnesota. Reoperation for failed primary antireflux procedures was performed in 395 (16%) symptomatic patients. Medical management of symptoms had failed in these patients. A laparoscopic approach was attempted in 83 (21%) patients, with 8 requiring conversion to an open approach because of extensive adhesions in 7 (8%) patients and intraoperative esophageal perforation in 1 (1%) patient. Those 8 patients were excluded from further analysis. Data about previous procedures, symptoms, preoperative studies, intraoperative findings, technique of reoperation, complications, and follow-up were collected, and descriptive analyses were performed. Operative findings (types of surgical failure) as described by the surgeon performing the reoperation were classified according to Hinder [11]: 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2014.05.036
GENERAL THORACIC
Laparoscopic Repair for Failed Antireflux Procedures
GENERAL THORACIC
1262
MAKDISI ET AL LAPAROSCOPIC REPAIR FOR FAILED ANTIREFLUX PROCEDURES
Type I: Complete or nearly complete disruption of the fundoplication Type II: Malpositioned fundoplication Type III: Slipped fundoplication Type IV: Transhiatal herniation Functional status was considered excellent if the patient was eating a general diet without preoperative symptoms or without requiring medication. Functional status was good if symptoms were minimal and if daily medication or frequent dilation was not required. Functional status was fair if symptoms were improved by daily medication or if frequent dilation was required. Functional status was poor if symptoms were unchanged or worse or if the patient required reoperation [12, 13]. Approval from The Mayo Clinic Institutional Review Board was granted for the study and all patients gave consent for the research. All data were reported and analyzed using SPSS Statistics, version 17 (SPSS Inc, Chicago, IL), as proportions, means (standard deviations), or medians (ranges). The means of continuous variables were compared using appropriate parametric or nonparametric tests. Categorical variables and proportions were compared using Fisher’s exact test, and Pearson’s c2 test was used to identify variables significant for the prediction of complications and reoperation failures. Factors included in the model were age, sex, weight, previous type of operation (open versus laparoscopic), presenting symptom (reflux versus obstructive symptoms), and addition of a Collis wedge gastroplasty. A p value less than 0.05 was considered statistically significant.
Results There were 75 patients: 56 (77%) women and 19 (23%) men formed the basis of the study. Median age at the time of reoperation was 58 years (range, 20–88 years). Median body mass index (BMI) was 28.7 kg/m2 (range, 14–39 kg/m2), and 9 patients (12%) had a BMI greater than 35 kg/m2. Fifty-four patients (72%) had their initial operations elsewhere. The initial approach was laparoscopic in 65 (87%) patients and open in 10 (13%) patients. A single previous repair had been performed in 72 patients (96%) and 3 patients had undergone 2 previous repairs. Previous procedures included laparoscopic Nissen fundoplication in 60 patients, open Nissen fundoplication in 10 (13%) patients, partial posterior fundoplication in 3 patients, and anterior partial fundoplication and gastropexy in 1 patient each. Additional previous procedures in those patients included crural mesh implantation in 2 patients, lengthening gastroplasty in 2 patients, and gastrostomy tube insertion and Rossetti modification in 1 patient each. Twelve (16%) patients had undergone at least 1 previous laparoscopy or laparotomy for a condition other than gastroesophageal reflux or hiatal hernia. Median time between reoperation and previous surgical intervention was 42 months (range, 5 days–17 years). Twenty-two patients (29%) were operated on within 1 year of the most recent operation and 37 (49%) underwent operation within 2 years. All patients were
Ann Thorac Surg 2014;98:1261–6
symptomatic (Table 1). Pain was present in 66 (88%) patients, pyrosis was present in 55 patients (73%), dysphagia was present in 48 (64%) patients, and previous episodes of aspiration were present in 18 (24%) patients. At least 1 esophageal dilation was performed in 9 (12%) patients (range, 1–5 procedures). Weight loss was observed in 8 (11%) patients (median loss, 7.6 kg; range, 2–15 kg). Barium swallow examination was performed in 70 patients (93%): documented expected postoperative findings were present in 6 (8%) patients; slipped or herniated wraps were present in 46 (61%) patients; loose, disrupted, or unwrapped fundoplication was present in 12 (16%) patients, and significant narrowing of the distal esophagus with delayed esophageal emptying was present in 11 (15%) patients. Esophagoscopy was performed in 54 (72%) patients: documented expected postoperative findings were seen in 9 (12%) patients, hiatal hernia or slipped wrap was seen in 38 (51%) patients, esophagitis was seen in 38 (51%) patients, Barrett’s esophagus was seen in 11 (15%) patients, and a Schatzki ring was seen in 2 patients. Esophageal manometry was performed in 44 (59%) patients and was normal in 27 (36%) patients; 17 (35%) patients had significant dysmotility. Computed tomography of the abdomen and chest was performed in 19 (25%) patients and was read as normal in 5 patients. It showed sliding hiatal hernia in 12 (16%) patients and malposition of the fundoplication and complete dehiscence in 1 patient each. A pH study was done in 11 (15%) patients and showed significant reflux in 4 (5%) patients. Operative findings (types of surgical failure) as described by the surgeon performing the reoperation included type I in 14 (19%) patients, type II in 3 (4%) patients, type III in 37 (49 %) patients, and type IV in 5 (7%) patients. In addition, 12 (16%) patients were found to have an obstructing fundoplication, and in 2 (3%) patients, the crural closure was too tight (Table 2). The laparoscopic redo antireflux procedure was a Nissen fundoplication in 57 (76%) patients, a partial posterior fundoplication (Toupet procedure) in 13 (17%) patients, a partial anterior fundoplication (Dor procedure) in 3 (4%) patients, removal of crural stitches in 1 patient, and a combination of partial posterior fundoplication and removal of crural stiches in 1 patient. Additional procedures included wedge gastroplasty for esophageal lengthening in 37 patients (49%). Thirty-five (47%) of the gastroplasties were performed with a Nissen
Table 1. Preoperative Symptoms and Findings Symptoms/Findings Pain Pyrosis Dysphagia Esophagitis Aspiration Obstruction Weight loss Barrett’s esophagus
Patients (%) 66 55 48 38 18 16 8 11
(88%) (73%) (64%) (51%) (24%) (21%) (11%) (15%)
MAKDISI ET AL LAPAROSCOPIC REPAIR FOR FAILED ANTIREFLUX PROCEDURES
Table 2. Findings at Reoperation Operating Room Findings Type I Type II Type III Type IV Obstructing fundoplication Tight crura
Patients (%) 14 3 37 5 12 2
(19) (4) (49) (7) (16) (3)
Type I, complete or nearly complete disruption of the fundoplication; type II, malpositioned fundoplication; type III, slipped Nissen fundoplication; type IV, transhiatal herniation.
fundoplication and 2 were performed with a partial anterior fundoplication (Table 3). There were no operative deaths. Complications occurred in 13 (17%) patients, including prolonged nausea and vomiting in 5 (7%) patients; surgical wound pain in 4 (5%) patients; atrial fibrillation, urinary retention, ileus, and esophageal perforation requiring reoperation in 2 patients each; and ulnar nerve injury, pneumothorax requiring a chest tube insertion, and severe dysphagia requiring reoperation in 1 patient each. Median hospitalization was 3.3 days (range, 1–15 days). Median follow-up was 25 months (range 30 days–12.8 years) and was complete in all patients. Intermittent pain was the most common frequent complaint in 11 (15%) patients followed by dysphagia in 8 (11%) patients and pyrosis in 6 (8%) patients. Three patients (4%) required esophageal dilation postoperatively. Significant weight loss was observed in 5 patients (7%). During follow-up, barium swallow examination was performed in 25 patients (36%), which documented expected postoperative findings in 13 (17%) patients, significant narrowing of the distal esophagus with delayed esophageal emptying in 5 (7%) patients, hiatal hernia in 5 (7%) patients (3 sliding and 2 paraesophageal hernias), and delayed gastric emptying in 2 (3%) patients. Esophagoscopy was performed in 21 (28%) patients, which documented normal and expected postoperative findings in 9 (12%) patients, hiatal hernia in 5 (7%) patients (3 sliding and 2 paraesophageal hernias), Barrett’s esophagus in 4 patients (new onset in 2 patients), and esophagitis in 3 patients. Esophageal manometry was performed in 7 (9%) patients and was normal in 6 (8%) of them. One patient (1%) had significant dysmotility. Computed Table 3. Procedures at Reoperation Procedures Nissen fundoplication With wedge gastroplasty Without wedge gastroplasty Toupet procedure Dor procedure With wedge gastroplasty Without wedge gastroplasty Crural stitch removal
Patients (%) 57 35 18 13 3 2 1 2
(76%) (47%) (29%) (17%) (4%) (3%) (2%) (3%)
1263
tomography of the abdomen and chest was performed in 6 (8%) patients and were read as normal in 5 patients and showed sliding hiatal hernia in 1 patient. Radionuclide gastric emptying studies were performed in 6 (8%) patients and were read as normal in 4 patients and delayed in 2 patients. Twenty-four– hour pH studies were performed in 2 (3%) patients and significant reflux was seen in 1 patient. Improvement of preoperative symptoms was observed in 70 (93%) patients in early postoperative follow-up (3 months). Overall, 65 patients (87%) were improved. Functional results were classified as excellent in 59 patients (78%), good in 6 (7 %) patients, fair in 7 (8%) patients, and poor in 3 (4%) patients. Two (3%) patients underwent reoperation at 1 month and 6.5 years, respectively. One patient had a left thoracotomy, cut gastroplasty, and Nissen fundoplication, and the other patient had laparoscopy, Nissen fundoplication conversion to a posterior fundoplication, and distal esophageal myotomy. Overall analysis revealed that preoperative factors including age, sex, previous surgical approach, symptoms, interval between operations, obesity, and performance of gastroplasty at time of reoperation were not significantly associated with postoperative morbidity, length of stay, and functional results (Table 4).
Comment As experience in laparoscopic repair for gastroesophageal reflux disease and paraesophageal hernia has increased, the number of reoperations has increased as well [5, 7, 10, 11, 13–16]. In our practice, 16% of all antireflux operations are reoperations. This reoperation rate is much higher than the recently reported rate of 3% to 6% [1] and reflects the tertiary nature of our practice. The key to a successful outcome includes both proper patient selection and careful operating technique [12]. An important factor a surgeon should consider in choosing the surgical approach is the likelihood that a safe and technically proper reconstruction of the antireflux mechanism can be performed [17]. It is critical to understand the pattern of failure when planning the optimal surgical approach for reoperation [12, 18, 19]. To that effect, valuable information includes operative details from the initial operation or operations, current anatomy of the esophagogastric junction, including possible esophageal shortening, size of the hiatal hernia, and any associated endoluminal pathologic condition such as stricture, esophagitis, and Barrett’s esophagus [12, 20]. Recurrent reflux may indicate an inadequate primary repair caused by either poor construction or faulty positioning of the fundoplication or subsequent disruption of the fundoplication or the crura, or both. Alternatively, dysphagia symptoms could be related to a tight, twisted, or herniated fundoplication or a tight crura or the presence of a paraesophageal hernia. The basic principles of reoperation should be identical to those of the primary procedure, including (1) complete reduction of the hiatal hernia if present, (2) establishment of an adequate
GENERAL THORACIC
Ann Thorac Surg 2014;98:1261–6
GENERAL THORACIC
1264
MAKDISI ET AL LAPAROSCOPIC REPAIR FOR FAILED ANTIREFLUX PROCEDURES
Ann Thorac Surg 2014;98:1261–6
Table 4. Factors Affecting Morbidity, Hospital Stay, and Long-Term Functional Outcome
Subgroup/Number Age
Background. Minimally invasive procedures have become common, and more reoperations for failed antireflux procedures are performed laparoscopically. We wanted to study the outcomes of laparoscopic reoperations for reflux. Methods. Medical records of all patients who underwent reoperation without esophageal resection after previous antireflux procedures between January 2000 and October 2012 were reviewed. Results. Seventy-five patients were included in this report: 56 (77%) women and 19 (23%) men. Median age was 58 years. The previous operation was laparoscopic antireflux procedures in 65 (87%) patients. The median interval between the last antireflux procedure and laparoscopic reoperation was 42 months. The median body mass index (BMI) was 28.7. All patients were symptomatic. Intraoperative findings included recurrent hiatal hernia in 47 (63%) patients, incompetent fundoplication in 14 (19%) patients, tight fundoplication in 8 (11%) patients, and tight crura in 2 (3%) patients. Laparoscopic Nissen fundoplication was performed in 57 (76%)
patients, partial posterior fundoplication was performed in 12 (16%) patients, partial anterior fundoplication was performed in 3 (4%) patients, removal of crural stitches was performed in 2 patients, and a combination of partial posterior fundoplication and removal of crural stiches was performed in 1 patient. Complications occurred in 13 (15%) patients. Improvement in symptoms was observed in 70 (93%) patients in early postoperative follow-up and in 59 (78%) patients in long-term follow-up. Functional results were classified as excellent in 59 (78%) patients, good in 6 (7%) patients, fair in 7 (8%) patients, and poor in 3 (4%) patients. Conclusions. Laparoscopic reoperation for failed antireflux operations is a complex procedure, but it is safe and effective in selected patients. Reoperation after a failed antireflux repair results in excellent or good functional status in a majority of patients, but these results may deteriorate over time.
A
failed antireflux procedures and analyzed factors affecting postoperative morbidity and short- and longterm outcomes.
ntireflux operations have a failure rate of 5% to 30% [1–7] after open and laparoscopic antireflux procedures. The failure rate after laparoscopic repair may worsen as these patients are followed for a longer period. Although not all patients with recurrent reflux symptoms or obstructive symptoms require reoperation, it is estimated that 3% to 6% of patients who undergo primary antireflux operations will eventually require reoperation [5, 8]. Reoperation for failed antireflux repair is usually performed through transabdominal or transthoracic techniques. With the development of laparoscopic techniques, more reoperations for failed antireflux procedures are being performed through minimally invasive techniques [1–3, 5]. Early reports after laparoscopic reoperations are encouraging [6, 9, 10], but little is known regarding long-term outcomes. We reviewed our experience with minimally invasive laparoscopic repair for Accepted for publication May 13, 2014. Presented at the Fiftieth Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Jan 25–29, 2014. Address correspondence to Dr Deschamps, Division of General Thoracic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: [email protected].
Ó 2014 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2014;98:1261–6) Ó 2014 by The Society of Thoracic Surgeons
Patients and Methods Between January 2000 and October 2012, 2,402 patients underwent antireflux operations at the Mayo Clinic in Rochester, Minnesota. Reoperation for failed primary antireflux procedures was performed in 395 (16%) symptomatic patients. Medical management of symptoms had failed in these patients. A laparoscopic approach was attempted in 83 (21%) patients, with 8 requiring conversion to an open approach because of extensive adhesions in 7 (8%) patients and intraoperative esophageal perforation in 1 (1%) patient. Those 8 patients were excluded from further analysis. Data about previous procedures, symptoms, preoperative studies, intraoperative findings, technique of reoperation, complications, and follow-up were collected, and descriptive analyses were performed. Operative findings (types of surgical failure) as described by the surgeon performing the reoperation were classified according to Hinder [11]: 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2014.05.036
GENERAL THORACIC
Laparoscopic Repair for Failed Antireflux Procedures
GENERAL THORACIC
1262
MAKDISI ET AL LAPAROSCOPIC REPAIR FOR FAILED ANTIREFLUX PROCEDURES
Type I: Complete or nearly complete disruption of the fundoplication Type II: Malpositioned fundoplication Type III: Slipped fundoplication Type IV: Transhiatal herniation Functional status was considered excellent if the patient was eating a general diet without preoperative symptoms or without requiring medication. Functional status was good if symptoms were minimal and if daily medication or frequent dilation was not required. Functional status was fair if symptoms were improved by daily medication or if frequent dilation was required. Functional status was poor if symptoms were unchanged or worse or if the patient required reoperation [12, 13]. Approval from The Mayo Clinic Institutional Review Board was granted for the study and all patients gave consent for the research. All data were reported and analyzed using SPSS Statistics, version 17 (SPSS Inc, Chicago, IL), as proportions, means (standard deviations), or medians (ranges). The means of continuous variables were compared using appropriate parametric or nonparametric tests. Categorical variables and proportions were compared using Fisher’s exact test, and Pearson’s c2 test was used to identify variables significant for the prediction of complications and reoperation failures. Factors included in the model were age, sex, weight, previous type of operation (open versus laparoscopic), presenting symptom (reflux versus obstructive symptoms), and addition of a Collis wedge gastroplasty. A p value less than 0.05 was considered statistically significant.
Results There were 75 patients: 56 (77%) women and 19 (23%) men formed the basis of the study. Median age at the time of reoperation was 58 years (range, 20–88 years). Median body mass index (BMI) was 28.7 kg/m2 (range, 14–39 kg/m2), and 9 patients (12%) had a BMI greater than 35 kg/m2. Fifty-four patients (72%) had their initial operations elsewhere. The initial approach was laparoscopic in 65 (87%) patients and open in 10 (13%) patients. A single previous repair had been performed in 72 patients (96%) and 3 patients had undergone 2 previous repairs. Previous procedures included laparoscopic Nissen fundoplication in 60 patients, open Nissen fundoplication in 10 (13%) patients, partial posterior fundoplication in 3 patients, and anterior partial fundoplication and gastropexy in 1 patient each. Additional previous procedures in those patients included crural mesh implantation in 2 patients, lengthening gastroplasty in 2 patients, and gastrostomy tube insertion and Rossetti modification in 1 patient each. Twelve (16%) patients had undergone at least 1 previous laparoscopy or laparotomy for a condition other than gastroesophageal reflux or hiatal hernia. Median time between reoperation and previous surgical intervention was 42 months (range, 5 days–17 years). Twenty-two patients (29%) were operated on within 1 year of the most recent operation and 37 (49%) underwent operation within 2 years. All patients were
Ann Thorac Surg 2014;98:1261–6
symptomatic (Table 1). Pain was present in 66 (88%) patients, pyrosis was present in 55 patients (73%), dysphagia was present in 48 (64%) patients, and previous episodes of aspiration were present in 18 (24%) patients. At least 1 esophageal dilation was performed in 9 (12%) patients (range, 1–5 procedures). Weight loss was observed in 8 (11%) patients (median loss, 7.6 kg; range, 2–15 kg). Barium swallow examination was performed in 70 patients (93%): documented expected postoperative findings were present in 6 (8%) patients; slipped or herniated wraps were present in 46 (61%) patients; loose, disrupted, or unwrapped fundoplication was present in 12 (16%) patients, and significant narrowing of the distal esophagus with delayed esophageal emptying was present in 11 (15%) patients. Esophagoscopy was performed in 54 (72%) patients: documented expected postoperative findings were seen in 9 (12%) patients, hiatal hernia or slipped wrap was seen in 38 (51%) patients, esophagitis was seen in 38 (51%) patients, Barrett’s esophagus was seen in 11 (15%) patients, and a Schatzki ring was seen in 2 patients. Esophageal manometry was performed in 44 (59%) patients and was normal in 27 (36%) patients; 17 (35%) patients had significant dysmotility. Computed tomography of the abdomen and chest was performed in 19 (25%) patients and was read as normal in 5 patients. It showed sliding hiatal hernia in 12 (16%) patients and malposition of the fundoplication and complete dehiscence in 1 patient each. A pH study was done in 11 (15%) patients and showed significant reflux in 4 (5%) patients. Operative findings (types of surgical failure) as described by the surgeon performing the reoperation included type I in 14 (19%) patients, type II in 3 (4%) patients, type III in 37 (49 %) patients, and type IV in 5 (7%) patients. In addition, 12 (16%) patients were found to have an obstructing fundoplication, and in 2 (3%) patients, the crural closure was too tight (Table 2). The laparoscopic redo antireflux procedure was a Nissen fundoplication in 57 (76%) patients, a partial posterior fundoplication (Toupet procedure) in 13 (17%) patients, a partial anterior fundoplication (Dor procedure) in 3 (4%) patients, removal of crural stitches in 1 patient, and a combination of partial posterior fundoplication and removal of crural stiches in 1 patient. Additional procedures included wedge gastroplasty for esophageal lengthening in 37 patients (49%). Thirty-five (47%) of the gastroplasties were performed with a Nissen
Table 1. Preoperative Symptoms and Findings Symptoms/Findings Pain Pyrosis Dysphagia Esophagitis Aspiration Obstruction Weight loss Barrett’s esophagus
Patients (%) 66 55 48 38 18 16 8 11
(88%) (73%) (64%) (51%) (24%) (21%) (11%) (15%)
MAKDISI ET AL LAPAROSCOPIC REPAIR FOR FAILED ANTIREFLUX PROCEDURES
Table 2. Findings at Reoperation Operating Room Findings Type I Type II Type III Type IV Obstructing fundoplication Tight crura
Patients (%) 14 3 37 5 12 2
(19) (4) (49) (7) (16) (3)
Type I, complete or nearly complete disruption of the fundoplication; type II, malpositioned fundoplication; type III, slipped Nissen fundoplication; type IV, transhiatal herniation.
fundoplication and 2 were performed with a partial anterior fundoplication (Table 3). There were no operative deaths. Complications occurred in 13 (17%) patients, including prolonged nausea and vomiting in 5 (7%) patients; surgical wound pain in 4 (5%) patients; atrial fibrillation, urinary retention, ileus, and esophageal perforation requiring reoperation in 2 patients each; and ulnar nerve injury, pneumothorax requiring a chest tube insertion, and severe dysphagia requiring reoperation in 1 patient each. Median hospitalization was 3.3 days (range, 1–15 days). Median follow-up was 25 months (range 30 days–12.8 years) and was complete in all patients. Intermittent pain was the most common frequent complaint in 11 (15%) patients followed by dysphagia in 8 (11%) patients and pyrosis in 6 (8%) patients. Three patients (4%) required esophageal dilation postoperatively. Significant weight loss was observed in 5 patients (7%). During follow-up, barium swallow examination was performed in 25 patients (36%), which documented expected postoperative findings in 13 (17%) patients, significant narrowing of the distal esophagus with delayed esophageal emptying in 5 (7%) patients, hiatal hernia in 5 (7%) patients (3 sliding and 2 paraesophageal hernias), and delayed gastric emptying in 2 (3%) patients. Esophagoscopy was performed in 21 (28%) patients, which documented normal and expected postoperative findings in 9 (12%) patients, hiatal hernia in 5 (7%) patients (3 sliding and 2 paraesophageal hernias), Barrett’s esophagus in 4 patients (new onset in 2 patients), and esophagitis in 3 patients. Esophageal manometry was performed in 7 (9%) patients and was normal in 6 (8%) of them. One patient (1%) had significant dysmotility. Computed Table 3. Procedures at Reoperation Procedures Nissen fundoplication With wedge gastroplasty Without wedge gastroplasty Toupet procedure Dor procedure With wedge gastroplasty Without wedge gastroplasty Crural stitch removal
Patients (%) 57 35 18 13 3 2 1 2
(76%) (47%) (29%) (17%) (4%) (3%) (2%) (3%)
1263
tomography of the abdomen and chest was performed in 6 (8%) patients and were read as normal in 5 patients and showed sliding hiatal hernia in 1 patient. Radionuclide gastric emptying studies were performed in 6 (8%) patients and were read as normal in 4 patients and delayed in 2 patients. Twenty-four– hour pH studies were performed in 2 (3%) patients and significant reflux was seen in 1 patient. Improvement of preoperative symptoms was observed in 70 (93%) patients in early postoperative follow-up (3 months). Overall, 65 patients (87%) were improved. Functional results were classified as excellent in 59 patients (78%), good in 6 (7 %) patients, fair in 7 (8%) patients, and poor in 3 (4%) patients. Two (3%) patients underwent reoperation at 1 month and 6.5 years, respectively. One patient had a left thoracotomy, cut gastroplasty, and Nissen fundoplication, and the other patient had laparoscopy, Nissen fundoplication conversion to a posterior fundoplication, and distal esophageal myotomy. Overall analysis revealed that preoperative factors including age, sex, previous surgical approach, symptoms, interval between operations, obesity, and performance of gastroplasty at time of reoperation were not significantly associated with postoperative morbidity, length of stay, and functional results (Table 4).
Comment As experience in laparoscopic repair for gastroesophageal reflux disease and paraesophageal hernia has increased, the number of reoperations has increased as well [5, 7, 10, 11, 13–16]. In our practice, 16% of all antireflux operations are reoperations. This reoperation rate is much higher than the recently reported rate of 3% to 6% [1] and reflects the tertiary nature of our practice. The key to a successful outcome includes both proper patient selection and careful operating technique [12]. An important factor a surgeon should consider in choosing the surgical approach is the likelihood that a safe and technically proper reconstruction of the antireflux mechanism can be performed [17]. It is critical to understand the pattern of failure when planning the optimal surgical approach for reoperation [12, 18, 19]. To that effect, valuable information includes operative details from the initial operation or operations, current anatomy of the esophagogastric junction, including possible esophageal shortening, size of the hiatal hernia, and any associated endoluminal pathologic condition such as stricture, esophagitis, and Barrett’s esophagus [12, 20]. Recurrent reflux may indicate an inadequate primary repair caused by either poor construction or faulty positioning of the fundoplication or subsequent disruption of the fundoplication or the crura, or both. Alternatively, dysphagia symptoms could be related to a tight, twisted, or herniated fundoplication or a tight crura or the presence of a paraesophageal hernia. The basic principles of reoperation should be identical to those of the primary procedure, including (1) complete reduction of the hiatal hernia if present, (2) establishment of an adequate
GENERAL THORACIC
Ann Thorac Surg 2014;98:1261–6
GENERAL THORACIC
1264
MAKDISI ET AL LAPAROSCOPIC REPAIR FOR FAILED ANTIREFLUX PROCEDURES
Ann Thorac Surg 2014;98:1261–6
Table 4. Factors Affecting Morbidity, Hospital Stay, and Long-Term Functional Outcome
Subgroup/Number Age
