Herniation of an abdominal antireflux fundoplication into the chest: what does it mean?

Transcription

1 European Journal of Cardio-Thoracic Surgery Advance Access published December 9, 2013 European Journal of Cardio-Thoracic Surgery (2013) 1 6 doi: /ejcts/ezt553 ORIGINAL ARTICLE Herniation of an abdominal antireflux fundoplication into the chest: what does it mean? Yannick Deswysen, Francesco Volonté, Christian Gutschow, Renato Romagnoli, Paolo Strignano, Aous Ouazzani, Luc Verstraete, Charles De Gheldere, Maximillien Thoma, Vincent Uluma, Felicia Ungureanu, Jean-Yves Mabrut and Jean-Marie Collard* THORACIC Unit of Upper Gastro-Intestinal Surgery, Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Brussels, Belgium * Corresponding author. Unit of Upper Gastro-Intestinal Surgery, Cliniques Universitaires Saint-Luc, Avenue Hippocrate 10, B-1200 Brussels, Belgium. Tel: ; fax: ; jean-marie.collard@uclouvain.be ( J.-M. Collard). Received 13 April 2013; received in revised form 29 September 2013; accepted 16 October 2013 Abstract OBJECTIVE: The specific contribution of the herniation of an abdominal antireflux fundoplication into the chest to symptomatic and therefore surgical failure remains unclear. METHODS: The study was conducted in 189 consecutive fundoplication patients, categorized as patients reoperated on for chest herniation of either an abdominal 360 (Group 1; n = 95) or a partial (Group 2; n = 10) fundoplication, and patients having undergone an intrathoracic 360 fundoplication for short oesophagus (Group 3; n = 84; reference group). There were four subgroups in Group 1: 1A: wrap still complete and perioesophageal; 1B: wrap still complete but perigastric; 1C: wrap still perioesophageal but partially disrupted and 1D: wrap perigastric and partially disrupted. RESULTS: The prevalence of defective symptoms (heartburn and regurgitation) was significantly lower (P < ) in Group 3 (0.0%) and Subgroup 1A (3.7%) than in Subgroups 1B (84.4%), 1C (86.7%) and 1D (100%) and Group 2 (100%). The prevalence of obstructive symptoms (dysphagia, chest pain, necrosis and perforation) was significantly higher (P < ) in Subgroup 1A (100%) than in Subgroups 1B (57.8%), 1C (60.0%) and 1D (25.0%). The prevalence of a short oesophagus, an abdominal wall hernia repair and high abdominal pressure episodes in reoperated patients were 13.7, 36.2 and 67.2%, respectively. CONCLUSIONS: Unlike perigastric or partial fundoplication, a 360 perioesophageal abdominal fundoplication, when herniated into the chest, is still effective against reflux. Obstructive symptoms are due to either diaphragmatic strangulation or perigastric migration of the wrap (slipknot effect). Short oesophagus, weakness of the abdominal wall and high abdominal pressure episodes favour the herniation process. Keywords: Reoperation of fundoplication Chest Oesophagus Gastroesophageal reflux disease Hiatal hernia INTRODUCTION Various anatomical anomalies following antireflux fundoplication have been described [1 9], each one explaining the failure to provide the expected symptomatic relief to patients suffering from gastroesophageal reflux disease (GERD). Most commonly described are too long or too tied to wrap, construction or slip down of the wrap around the stomach, breakdown of the fundoplication, absence or breakdown of the crura closure, the latter made too snug, construction of the fundoplication with the greater curve of the stomach rather than with the gastric fundus, excessive fibrosis of the hiatal sling post-cauterization and herniation of the wrap into the chest. These anatomical anomalies mostly coexist at reoperation, rendering the specific contribution of each one of them to the failure process unclear. The present study aims at specifying the role of chest herniation of the fundoplication in the genesis of defective or obstructive oesophageal symptoms occuring either immediatelly after operation or at follow-up, i.e. after a symptom-free period. A second objective is to evaluate to which extent other factors may influence the failure pattern in these patients. Contributing factors are concommittant partial disruption of the fundoplication, concommittant location of the fundoplication around the stomach instead of around the oesophagus, the presence of a short oesophagus [10], a history of sudden increases in abdominal pressure and diffuse weakness of the abdominal wall (which had justified wall-repairing surgery). For this study, those patients who had their initial abdominal fundoplication herniated into the chest were extracted from the senior author s series of patients who underwent redo antireflux surgery following unsatisfactory symptomatic outcomes. The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 2 Y. Deswysen et al. / European Journal of Cardio-Thoracic Surgery MATERIALS AND METHODS Patients The present study is based on 189 consecutive patients operated on between 1986 and 2010 (Fig. 1). They were 72 women and 117 men ranging in age from 23 to 82 years (mean: 52 years; SD 13.24). Among those, 105 patients were reoperated on by the senior author (Jean-Marie Collard) for herniation into the chest of either an abdominal 360 fundoplication (Group 1, n = 95) or an abdominal partial fundoplication (Group 2, n = 10). The remaining patients underwent an intrathoracic 360 fundoplication as first antireflux operation for GERD linked to a short oesophagus (reference group, Group 3, n = 84). Group 1 patients were categorized clinically as those who never experienced any symptomatic relief following antireflux operation (early failure, n = 31) and those who became symptomatic again after a symptom-free period ranging from 1 to 300 months (late failure, n = 64). Group 1 patients were further subdivided into four subgroups according to anatomical findings at reoperation: Subgroups 1A: herniation of an intact 360 perioesophageal fundoplication into the chest (n = 27); 1B: herniation of a 360 fundoplication that also had slipped down around the stomach (n = 45); 1C: herniation of a 360 fundoplication that had partially broken down (n = 15) and 1D: herniation of a 360 fundoplication that had partially broken down and slipped down around the stomach (n = 8). The number of previous antireflux operations in combined Groups 1 and 2 was 1, 2 and 3 in 88, 14 and 3 patients, respectively. The last antireflux operation had been carried out by either laparotomy (n = 46) or laparoscopy [11] (n = 58), and in one case by left thoracotomy, the fundoplication needing to be replaced below the diaphragm. The time elapsed between the previous antireflux operation and the reoperation ranged from 1 day to 305 months (median: 24 months). The local ethics committee approved the study and waived patient written consent. Preoperative work-up Radiological anatomy of the gastroesophageal (GE) junction in all patients was assessed by a preoperative barium swallow study including radiographs taken in both the upright and supine positions [12, 13]. The oesophagus was considered short whenever the junction between upper gastric folds and the unwrinkled oesophageal mucosa was above the diaphragm, while the oesophageal body was straight in the upright position. Ninety-four patients, i.e. 84 of Group 1 and all 10 of Group 2, underwent upper gastrointestinal (GI) endoscopy in search of erosive oesophagitis or Barrett s. One hundred and twenty-five patients underwent 24-h oesophageal ph monitoring either before the reoperation in Group 1 (n = 54) and in Group 2 (n = 5) or after the intrathoracic 360 fundoplication in Group 3 (n = 66). Data were analysed using DeMeester s criteria [14] in terms of percentage of time that the oesophageal ph was <4 during the total, upright and supine periods of recording (95th percentile of normal volunteers: 4.5, 8.4 and 3.5%, respectively). Intraoperative findings and surgical technique Antireflux reoperations were performed by laparotomy (n = 91), laparoscopy (n = 1) or left thoracotomy (n = 13). Three were performed as emergency operations for acute strangulation of the fundoplication at the hiatal level with (n = 2) or without (n =1) Figure 1: Diagram showing the different groups and subgroups studied.

3 Y. Deswysen et al. / European Journal of Cardio-Thoracic Surgery 3 gastric perforation. The thoracic approach was used whenever oesophageal shortening was suspected from the preoperative barium swallow study. The GE junction was considered surgically irreducible whenever it could not be replaced in the abdomen without undue tension, despite extended surgical mobilization of the oesophageal body up to the aortic arch. The actual anatomy of the residual fundoplication was assessed by the senior author with regard to its completeness or partial breakdown and to its location around the stomach or around the oesophagus. For this, the residual fundoplication was carefully freed from any adhering tissue and progressively taken down. Intrathoracic fundoplication [15, 16] for Group 3 was performed by left thoracotomy according to a previously described surgical technique [17, 18]. The most critical steps of the operation include further enlargement of the hiatal sling by a 3-cm radial incision of the diaphragm to prevent any strangulation of the intrathoracic stomach at the hiatal level, and anchoring of the intrathoracic fundoplication to the hiatal sling using numerous sutures placed so as to prevent any tiring out of the gastric wall during diaphragmatic movements, especially when the patient is coughing after the operation. Figure 2: Prevalence of defective symptoms in the six groups or subgroups studied. THORACIC Data analysis The clinical charts of all patients were carefully reviewed for the presence of defective (i.e. heartburn and regurgitation) or obstructive symptoms (i.e. dysphagia, chest pain, gastric necrosis or perforation) before the antireflux reoperation in Groups 1 and 2 by the first author and after the intrathoracic fundoplication in Group 3 by the second author. Fifty-eight patients belonging to Groups 1 and 2 were interviewed by the first author with a view to uncovering any previous abdominal wall hernia repair and any history of sudden or repetitive increase in abdominal pressure following the initial antireflux operation. Statistical analysis Categorical variables were analysed using Fisher s exact and χ 2 tests. Mann Whitney U- and Kruskal Wallis rank-sum tests were used for continuous variables. A P-value of <0.05 was considered statistically significant (Statistica 5.0, StatSoft, Inc., Oklahoma, USA). RESULTS Symptoms The prevalence of defective symptoms significantly increased (P < ) from 0% (n = 0) in Group 3 to 3.7% (n = 1) in Subgroup 1A, to 84.4% (n = 38) in Subgroup 1B, to 86.7% (n = 13) in Subgroup 1C, to 100% (n = 8) in Subgroup 1D and to 100% (n = 10) in Group 2 (Fig. 2). As shown in Fig. 3, the prevalence of obstructive symptoms was significantly higher (P < ) in Subgroup 1A (100%; n = 27) than in Group 3 (5.9%; n = 5). In Group 1, it was significantly higher (P < ) in Subgroup 1A (100%; n = 27) than in Subgroups 1B (57.8%; n = 26), 1C (60.0%; n = 9) and 1D (25.0%; n = 2). There was no significant difference in terms of either defective or obstructive symptoms between the four subgroups Figure 3: Prevalence of obstructive symptoms in Groups 3 and the four subgroups of Group 1. belonging to Group 1 depending on whether they had experienced early or late failure (Table 1). Upper gastrointestinal examinations The prevalence of erosive oesophagitis at upper GI endoscopy was significantly lower (P = ) in Subgroup 1A (18.5%) than in Subgroups 1B (78.0%), 1C (66.7%) and 1D (62.5%), and in Group 2 (50.0%). As shown in Fig. 4, oesophageal acid exposure at 24-h oesophageal ph monitoring expressed as the percentage of total time that oesophageal ph < 4 (mean ± standard error of the mean, {SEM}) was significantly (P < ) lower in Group 3 (0.50 ± 0.10) and Subgroup 1A (2.63 ± 1.27) than in Subgroups 1B (8.09 ± 1.45), 1C (10.18 ± 2.39), 1D (7.42 ± 2.40) and in Group 2 (9.20 ± 4.37). Intraoperative findings: early vs late failures As shown in Fig. 5, perigastric fundoplication was as common (P = 0.90) in patients who had experienced early failure (54.8%) as in those with late failure (56.3%). There was a substantial trend

4 4 Y. Deswysen et al. / European Journal of Cardio-Thoracic Surgery Table 1: Symptoms in Group 1 patients: early vs late failure Defective symptoms Obstructive symptoms Early a (n = 31) Late a (n = 64) Early a (n = 31) Late a (n = 64) Subgroup 1A 0/11 (0.0%) 1/16 (6.3%) 11/11 (100%) 16/16 (100%) Subgroup 1B 15/16 (93.8%) 23/29 (79.3%) 10/16 (62.5%) 16/29 (55.2%) Subgroup 1C 2/3 (66.6%) 11/12 (91.7%) 3/3 (100%) 6/12 (50.0%) Subgroup 1D 1/1 (100%) 7/7 (100%) 0/1 (0.0%) 2/7 (28.6%) a All P-values not significant, ranging from to 1. early failure (29%) than in those with late failure (6.3%). All 13 patients had their residual fundoplication found to be located around the stomach rather than around the oesophagus at reoperation. Abdominal pressure and weakness Figure 4: Oesophageal acid exposure at 24-h oesophageal ph monitoring in the six groups or subgroups studied. The dotted line represents the 95th percentile of Demeester s reference population. Of the 58 patients belonging to Groups 1 and 2 for whom the information was available, 39 (67.2%) had a history of sudden or repetitive increases in abdominal pressure due to carriage of heavy loads (n = 25), forceful vomiting (n = 4), pregnancy (n = 2), prostatism (n = 2) or strain while awaking from anaesthesia (n = 6). Twenty-one of these 58 (36.2%) patients were operated on for an abdominal wall hernia that was incisional, inguinal, umbilical, epigastric and diaphragmatic in 15, 1, 2, 2 and 1 patients, respectively. DISCUSSION Figure 5: Prevalence of perigastric fundoplication (left), partial breakdown of the fundoplication (middle) and short oesophagus (right) in early (blue columns) and late (blank columns) failures in Group 1. (P = ) towards a higher prevalence of breakdown of the antireflux repair in late failure patients (29.7%) compared with those with early failure (12.9%). Short oesophagus, a condition that was present in 13 of the 95 Group 1 patients (13.7%), was significantly more common (P = ) in patients who experienced This single-centre retrospective study identifies three critical conditions for an antireflux fundoplication to function properly in the chest in order for a patient to be asymptomatic [18, 19]. First, the fundic wrap must be complete, covering 360 of the oesophageal circumference. Secondly, it must be located proximal to the GE junction, around the lower oesophagus. Thirdly, the hiatal sling must be large enough to give a wide berth to the fundoplication. These three conditions are met when the fundoplication is primarily constructed in the chest around a short oesophagus (Fig. 6A) or when an abdominal fundoplication herniates into the chest while remaining both complete and perioesophageal and when spontaneous disruption of the crura closure is also complete (Fig. 6B). Should one of these three conditions be lacking, the patient is bound to get defective or obstructive symptoms and to seek further medical consultation. Initial placement or subsequent downward migration of the fundoplication around the stomach as well as its spontaneous breakdown, and above all, the combination of the two, generate defective symptoms witnessing a loss in efficacy against GE reflux. The study shows indeed that neither a perigastric nor a partial fundoplication can function in the chest, as shown by pathological oesophageal acid exposure at 24-h oesophageal ph monitoring: 90% of the patients studied had defective symptoms requiring antireflux reoperation. In this respect, Toupet [20], 50 years ago, interestingly emphasized the need for anchoring the partial fundoplication to the diaphragm. He

5 Y. Deswysen et al. / European Journal of Cardio-Thoracic Surgery 5 described, presuming that the latter, should it herniate into the chest, could no longer function properly. The fact that a 360 fundoplication may still be effective against reflux when it herniates into the chest at follow-up is a strong argument favouring the performance of a complete rather than a partial abdominal fundoplication in routine antireflux surgery. The study shows that there are two conditions for the onset of obstructive symptoms once an abdominal fundoplication has Figure 6: (A) Radiograph showing the absence of any strangulation effect in a Group 3 patient who underwent a primary intrathoracic 360 perioesophageal fundoplication including enlargement of the hiatus by radial incision of the diaphragm for short oesophagus. The dotted line represents the diaphragm. (B) Radiograph showing the herniation of an abdominal perioesophageal fundoplication made under tension in a patient in whom the presence of a short oesophagus had not been detected prior to initial antireflux surgery (arrow: GE junction; dotted line: diaphragm). This patient had spontaneously acquired an intrathoracic perioesophageal fundoplication similar to the one usually performed transthoracically for short oesophagus. He was totally asymptomatic due to spontaneous complete breakdown of the crura closure with no strangulation effect. Figure 7: (A) Radiograph showing strangulation (arrow: diaphragmatic pinching phenomenon) of the herniated fundoplication because of incomplete breakdown of the crura closure in a Group 1 patient. (B) Radiograph showing an initially abdominal fundoplication that had herniated into the chest and was found to be located around the stomach at reoperation in a patient in whom oesophageal shortening had not been detected prior to initial antireflux surgery. Note the gastric slipknot phenomenon (arrow). The dotted line represents the diaphragm. herniated into the chest. First, incomplete spontaneous disruption of the crura closure (Fig. 7A) creates a strangulation effect on the gastric wall, usually responsible for dysphagia or chest pain, sometimes leading to life-threatening complications such as gastric bleeding from collar ulcers, or even necrosis of the intrathoracic portion of the stomach and subsequent gastric perforation into the lower mediastinum. This pinching effect at the hiatal level, although it accounted for obstructive symptoms in all studied subgroups, is less common when the fundoplication has both partially broken down and slipped down around the stomach. This observation probably reflects a smaller amount of gastric tissue at the hiatal level in these patients in comparison with those with a complete fundoplication herniating into the chest. Secondly, downward migration of a still complete fundoplication around the stomach as part of the herniation process generates a slipknot effect on the gastric wall (Fig. 7B), so that the proximal part of the stomach cannot expand enough to accomodate ingested food and the patient subsequently suffers from either dysphagia or lower retrosternal pain during meals. In this respect, perigastric location of the wrap in early failure patients reflects either its initial misplacement or its early downward migration along the stomach, whereas in late failure patients, it may be assumed that the initially perioesophageal wrap with time gradually migrated along the stomach. There are three conditions identified in the study that predispose to the herniation of an abdominal fundoplication into the chest. First, the existence of a short oesophagus precludes performing a tension-free fundoplication proximal to the GE junction from the abdomen. Should this be attempted, the wrap would be either placed around the subcardiac area of the stomach [21], leaving some secreting gastric tissue proximally [22] or made at all cost, i.e. under undue tension around the oesophagus. In the latter instance, the short oesophagus is predisposed to come back to its natural location in the lower mediastinum. The GE junction is either pushed transhiatally by any increase in abdominal pressure or attracted upward by the negative pressure that exists in the chest [23, 24]. The higher prevalence of a short oesophagus observed in the early failure patients of this study in comparison with the late failure patients probably reflects the fact that thoracic herniation of a GE junction that has been forcefully reduced into the abdomen usually occurs very early following antireflux operation. All this confirms that an oesophagus that never existed or that lost its space in the abdomen is best approached from the chest as far as a perioesophageal fundoplication is concerned. Secondly, many GERD patients suffer from diffuse weakness of all walls of the abdominal cavity. Indeed, one-third of the patients in the present study had a history of abdominal wall-repairing surgery including surgical repair of an incisional, umbilical, epigastric, diaphragmatic or inguinal hernia. In this respect, we have to bear in mind that the diaphragm with all its orifices is nothing else than the roof of the abdomen. Furthermore, unlike the other parts of the abdominal wall, the upper aspect of this roof is exposed to a negative-pressure environment, i.e. the lower intrathoracic pressure, which certainly adds to the intrinsic weakness of the tissues. Thirdly, two-thirds of the 58 patients who had been interviewed for the purposes of the study reported a history of sudden or repetitive increases in intra-abdominal pressure due to the carriage of heavy loads, forceful vomiting, pregnancy or strain while awaking from anaesthesia just after the previous antireflux operation. This emphasizes the fact that long-term outcomes of an antireflux operation depend not only on the quality of the surgical THORACIC

6 6 Y. Deswysen et al. / European Journal of Cardio-Thoracic Surgery procedure, but also on the patient s behaviour at follow-up [11]. This observation also questions the indication of an antireflux operation in GERD patients likely to carry heavy loads in their daily life. In any event, proper performance of an antireflux fundoplication only creates an effective barrier against reflux. Surgery neither remedies the pre-existing tissular weakness nor suppresses the risk of high-pressure episodes in the abdomen, two conditions that are also at the origin of a hiatal hernia in most non-operated GERD patients. The higher prevalence in our recruitment of late failures (Group 1: 64 of 95) compared with early failures (Group 1: 31 of 95) together with the trend for a higher prevalence of breakdown of the fundoplication in late failure patients than in those with early failure both confirm the well-known observation that the results of an antireflux operation deteriorate with time, probably reflecting the cumulative effects of the mechanical constraints exerted upon the hiatal area over the years. In conclusion, symptomatic failure of an antireflux operation reflects a complex anatomical situation in the hiatal area in which several abnormalities coexist. One of these, i.e. the herniation of a 360 fundoplication from the abdomen into the chest, does not pose a real problem unless disruption of the hiatal sling is incomplete, and the wrap concommittantly breaks down or slides along the stomach, because neither a perigastric nor a partial fundoplication functions properly in the chest. Neither GERD patients carrying heavy loads on a regular basis nor those having a short oesophagus are good candidates for an abdominal antireflux fundoplication. ACKNOWLEDGEMENTS The authors are grateful to Claire Craddock-de Burbure for her help revising the paper. Conflict of interest: none declared. REFERENCES [1] Collard JM, Verstraete L, Otte JB, Fiasse R, Goncette L, Kestens PJ. Clinical, radiological and functional results of remedial antireflux operations. Int Surg 1993;78: [2] Collard JM, Romagnoli R, Kestens PJ. Reoperations for unsatisfactory outcome after laparoscopic antireflux surgery. Dis Esophagus 1996;9: [3] Collard JM. Failures of laparoscopic antireflux surgery. In: Stein H, Fuchs KH, Bonavina L (eds). Functional Foregut Disorders. Germany: Johan Ambrosius Barth Verlag, 1998, [4] Del Genio GM, Collard JM. Acute complications of anti-reflux surgery. In: Ferguson MK, Fennerty MB (eds). Managing Failed Anti-Reflux Therapy. London, UK: Springer Verlag, 2006, [5] Collard JM, Romagnoli R, Otte JB. Are there specific complications related to intrathoracic Nissen fundoplication? What is the role of vagal injury, devascularization of the fundus and constriction of blood flow in acute complications following intrathoracic Nissen fundoplication? In: Giuli R, Galmiche JP, Jamieson GG, Scarpignato C (eds). The Esophagogastric Junction. Paris, France: John Libbey, 1998, [6] Collard JM, Verstraete L, Romagnoli R. What are the consequences of herniation of the repair into the chest? Why are they different from the primary intrathoracic Nissen? In: Giuli R, Galmiche JP, Jamieson GG, Scarpignato C (eds). The Esophagogastric Junction. Paris, France: John Libbey, 1998, [7] Siewert JR, Stein HJ, Feussner H. Reoperations after failed antireflux procedures. Ann Chir Gynecol 1995;89: [8] Braghetto I, Csendes A, Korn O, Burdiles P, Valladares H, Cortés C et al. Anatomical deformities after laparoscopic antireflux surgery. Int Surg 2004;89: [9] Watson DL, Jamieson GG, Mitchell PC, Devitte PG, Britten-Jones PG. Stenosis of the esophageal hiatus following laparoscopic fundoplication. Arch Surg 1995;130: [10] Mattioli S, Lugaresi ML, Costantini M, Del Genio A, Di Martino N, Fei L et al. The short esophagus: intraoperative assessment of esophageal length. J Thorac Cardiovasc Surg 2008;136: [11] Collard JM, De Gheldere C, Dekock M, Otte JB, Kestens PJ. Laparoscopic antireflux surgery. What is real progress? Ann Surg 1994;220: [12] Pringot J, Ponette E. Radiological examination of the esophagus. In: Vantrappen G, Hellemans J (eds). Diseases of the Esophagus. Berlin, Germany: Springer Verlag, 1974, [13] Collard JM, Goncette L. Reflux-related short esophagus: radiological diagnosis and surgical treatment. Acta Endosc 2008;38: [14] DeMeester TR. Prolonged oesophageal ph-monitoring. In: Read NW (ed). Gastro-intestinal Motility: Which Test? Petersfield, UK: Wrighston Biomedical Publishing Ltd, 1989, [15] Kirschner M. Ein Neues Verfahren der Oesophagoplastik. Arch Klin Chir 1920;114: [16] Nissen R, Rossetti M. Ergebnisse und Schlussfolgerungen. In: Nissen R, Rossetti M (eds). Die Behandlung von Hiatushernien und Refluxösophagitis mit Gastropexie und Fundoplicatio. Stuttgart, Germany: Georg Thieme Verlag, 1959, [17] Nissen R. Transthorakal Fundusraffung zur Beeinflussung besonderer Formen von Refluxoesophagitis. Langenbecks Arch Klin Chir 1960;293: [18] Collard JM, De Koninck XJ, Otte JB, Fiasse RH, Kestens PJ. Intrathoracic Nissen fundoplication: long-term clinical and ph-monitoring evaluation. Ann Thorac Surg 1991;51:34 8. [19] Volonte F, Collard J-M, Gutschow C, Goncette L, Strignano P. Intrathoracic periesophageal fundoplication for short esophagus: a twenty-year experience. Ann Thorac Surg 2007;83: [20] Toupet A. Technique d oesophago-gastroplastie avec phréno-gastropexie appliquée dans la cure radicale des hernies hiatales et comme complément de l opération de HeIIer dans les cardiospasmes. Mém Acad Chir 1963;11 13: [21] Maillet P, Beaulieux J, Dumurgier C. A propos du traitement des brachyoesophages acquis. Chirurgie 1976;102: [22] Horvath KD, Swanstrom LL, Jobe BA. The short esophagus: pathophysiology, incidence, presentation, and treatment in the era of laparoscopic antireflux surgery. Ann Surg 2000;232: [23] Meredith J, Kon N, Poole G, Pennell T. Physiologic effects of intrathoracic placement of Nissen fundoplication. Curr Surg 1985;42:32 4. [24] Bemelman WA, Verburg J, Brummelkamp WH, Klopper PJ. A physical model of the intrathoracic stomach. Am J Physiol 1988;254:G