o 3','3'q. ),
IMPROVING OUTCOMES FOLLOWING SURGERY FOR GASTRO.OESOPHAGEAL REFLUX DISEASE . LAPAROSCOPIC ANTIREFLUX SURGERY
Thesis submitted in
November,1997
for the degree of Doctor of Medicine
in the University of Adelaide
by
David Ian'Watson M.B.,B.S. (Adel) 1984 F.R.A.C.S. (Gen Surg) 1992
The work described in this thesis was performed within the Department of Surgery at the
University of Adelaide, and the Royal Adelaide Centre for Endoscopic Surgery at the
Royal Adelaide Hospital 11
This work is dedicated to my wife Claire
and my children James, Edward and Timotþ.
I am forever indebted to them for their ongoing patience, considerable sacrifices, and unending support which has enabled this work to be completed. 111
SHORT TABLE OF CONTENTS
1 OUTCOME FOLLOWING LAPAROSCOPIC ANTIREFLUX SURGERY 2 1.1 Laparoscopic antireflux surgery - literature review 3 l-2 Outcome of prospective studies assessing outcomes following laparoscopic antireflux surgery 31
2 LABORATORY RESEARCH 122 2.I Comparison of anterior, posterior and total fundoplication using a viscera model t23 2-2 Effîcacy of anterior, posterior and total fundoplication in a porcine model r32
3 CONTROLLED TRIALS ASSESSING STRATEGIES TO IMPROVE OUTCOMES FOLLOWING SURGERY FOR GASTRO.OESOPHAGEAL REFLUX 146 3.1 Prospective randomised trial of laparoscopic versus open Nissen fundoplication for gastro-oesophageal reflux disease 147 3-2 Prospective double blind randomised trial of laparoscopic Nissen fundoplication with division and without division of short gastric vessels 166 3.3 Prospective double blind randomised trial of laparoscopic anterior partial versus Nissen fundoplication 191
4 CONCLUSIONS 214
5 BIBLIOGRAPHY 227
6 APPENDIX 256 1V
TABLE OF CONTENTS
ABSTRACT x DECLARATION xiii ACKNOWLEDGEMENTS xiv PREFACE xvi
ANTIREFLUX 1 OUTCOME FOLLOWING LAPAROSCOPIC ., SURGERY 1.1 LAPAROSCOPIC ANTIREFLUX SURGERY . LITERATURE REVIE\ry 3
1.1.1 OPEN NISSEN FUNDOPLICATION 3 T.1.2 RANDOMISED TRIALS FOR OPEN FUNDOPLICATION 5 L.1.2-I Nissen versus partial fundoplication 6 Thor and Silander 6 |.I.2.IJ 'Walker I.1.2.1.2 et al 7 I.1.2.I.3 Lundell et al 8 I-1.2.1.4 DeMeester et al 9 1.I.2.2 Other trials 10 I.1.2.2.L Hill et al 10 1.1 2.2.2 Janssen et al 10 1.1 2.2.3 Luostarinen et al 11 1.1 2.2.4 Spechler et al 11
1.1.3 INDICATIONS FOR SURGERY 11
T.1.4 APPROPRIATE SITUATIONS FOR THE APPLICATION OF A
1.1.5 EARLY RESULTS AND COMPLICATIONS FOLLOWING LAPAROSC OPIC FUNDOPLICATION 13 1.1 .5.1 Pneumothorax t6 1.1 .5.2 Vascular injury L6 1.1 .5.3 Paraoesophageal hiatus herniation t7 1.1 .5.4 Dysphagia t7 1.1 .5.5 Pulmonaryembolism 18 1.1 .5.6 Perforation of the gastrointestinal tract 18 1.1 .5.7 Mortality 19
1.1.6 LAPAROSCOPIC VERSUS OPEN SURGERY 19 1. 1.6.1 Clinical outcome comparisons 20 1. 1.6.2 Cost comparisons 2t 1. 1.6-3 Randomised comparisons 22
T.1.7 RE-OPERATTVE SURGERY FOLLOWING LAPAROSCOPIC FUNDOPLICATION 22 1.1.8 RANDOMISED TRIALS FOR LAPAROSCOPIC SURGERY 23 v
1.1.9 CONTROVERSIES AND DISAGREEMENTS 24 1.1 .9.1 Laparoscopic partial fundoplication 25 1.1 .9.2 Laparoscopic division of the short gastric vessels 25 1.1 .9.3 Selective versus routine hiatal repair 27 1.1 .9.4 The tailored approach to laparoscopic fundoplication 27 1.1 .9.5 The short oesophagus 28
1.1.10 FUTURE DIRECTIONS 29
1.1.11 RESEARCH OVERVIEW AND AIMS 29
1.2 OUTCOME OF PROSPECTIVE STUDIES ASSESSING OUTCOMES FOLLOWING LAPAROSCOPIC ANTIREFLUX SURGERY 31 T.2.T AIMS 31 T.2.2 GENERALMETHODOLOGY 31 I.2.2.I Preoperative assessment 31 1.2.2.2 Surgical technique for laparoscopic Nissen fundoplication 32 1.2.2.3 Post-operative management 36 I.2-2.4 Data management and follow-up protocol 36 1.2.3 INITIAL OUTCOME ASSESSMENT OF PROSPECTTVE CASE SERIES 37 1.2.3.1 Initial outcomes - the first 70 patients 37 1.2.3.1.I Patients and methods 37 1.2.3.I.2 Results 38 1.2.3.1.3 Discussion 43 1.2.3.2 Experience with 230laparoscopic antireflux operations and resulting changing surgical strategies 45 L.2.3.2-I Patients and methods 45 1.2.3.2.2 Results 46 I.2.3.2.3 Discussion 50 t.2.4 UNIQUE COMPLICATIONS OF LAPAROSCOPIC FUNDOPLICATION 52 I.2.4.1 Pneumothorax 52 I.2.4.1.I Case reports 53 I.2.4.1.2 Discussion 54 1.2.4.2 Para-oesophageal hiatus hernia 56 I.2.4.2-I Patients and methods 56 1.2.4.2.2 Results 56 I.2.4.2.3 Discussion 58 L.2.4.3 Stenosis of the oesophageal hiatus 60 1.2.4.3.I Report of cases 60 I.2.4.3.2 Discussion 63 T.2.5 THE LEARNING CURVE FOR LAPAROSCOPIC FUNDOPLICATION 65 1.2.5-l Defining the learning curve for laparoscopic fundoplication 65 I.2.5.1.1 Methods 65 1.2.5.1.2 Results 67 1.2.5.I.3 Discussion 76 1.2.5.2 Experience with laparoscopic surgery for gastro-oesophageal reflux beyond the leaming curve 78 1.2.5-2.I Patients and methods 79 L.2.5.2.2 Results 79 I.2.5.2.3 Discussion 82 V1
T.2.6 SHOULD THE CHOICE OF PROCEDURE FOR REFLUX BE INFLUENCED BY PREOPERATIVE PATHOLOGY? 83 1.2.6.1 Preoperative endoscopic grading of oesophagitis versus outcome 83 1.2.6.1.I Patients and methods 84 1.2.6.I.2 Results 86 I.2.6.1.3 Discussion 89 I.2.6.2 Oesophageal motility parameters before and after laparoscopic Nissen fundoplication 9l 1.2.6.2.1 Methods 92 I.2 -6.2.I.1 Patient selection 92 l -2.6.2.1.2 Oesophageal Manometry 92 I.2.6.2.1.3 Data analysis 93 1.2.6.2.1.4 Clinical outcome scores 95 1.2.6.2.1.5 Statistical analysis 95 I.2.6.2.2 Results 95 1.2.6.2.3 Discussion 99 1.2.6.3 The outcome of laparoscopic Nissen fundoplication in patients with disordered pre-operative peristalsis r02 I.2.6.3.I Patients and methods t02 I.2.6.3.1.1 Manometry 102 I.2.6.3.1.2 Patients 103 1.2.6.3.1.3 Follow-uP 103 1.2.6.3.2 Results 103 I.2.6.3.3 Discussion 109 1.2.6.4 Illness behaviour versus outcome following laparoscopic antireflux surgery 110 I.2.6.4.1 Methods 111 I.2.6.4.I J Illness Behaviour Questionnaire 111 1.2.6.4.1.2 Clinical Follow-uP 113 1.2.6.4.1.3 Statistical Analysis 113 7.2.6.4.2 Results IT3 I.2.6.4.3 Discussion 118
1.2.7 CONCLUSIONS t19
2 LABORATORY RESEARCH 122 2.1 coMPARrsoN oF ANTERTOR, POSTERTOR AND TOTAL FUNDOPLICATION USING A VISCERA MODEL 123 2.1.1 ArM 123 2.1.2 MATERIALS AND METHODS 123 2.1..2.1 Baseline study t26 2.I.2.2 Parameters following oesophageal myotomy t26 2-1.2.3 Parameters following anterior hemi-fundoplication t26 2.1,.2.4 Parameters following posterior hemi-fundoplication 126 2.1.2.5 Parameters following total fundoplication r27 2.I.2.6 Statistics t2l 2.T.3 RESULTS r2'7 2.1.4 DISCUSSION t29 v11
2.2 EFFTCACY OF ANTERTOR, POSTERTOR AND TOTAL FUNDOPLICATION IN A PORCINE MODEL 132 2.2.1 AIM t32 2.2.2 MATERIALS AND METHODS r32 2.2-2.I General Operative Procedure t33 2.2.2.2 Anterior fundoPlication 135 2.2-2.3 Posterior fundoplication 136 2.2.2.4 Total fundoplication 136 2.2.2.5 Postoperative care 136 2.2.2.6 Postoperative studies r37 2.2.2.7 Statistics r37 2.2.2.8 Ethics 138 2.2.3 RESULTS 138 2.2.4 DISCUSSION 140
3 CONTROLLED TRIALS ASSESSING STRATEGIES TO IMPROVE OUTCOMES FOLLOWING SURGERY FOR GASTRO- OESOPHAGEAL REFLUX L46 3.1 PROSPECTIVE RANDOMISED TRIAL OF LAPAROSCOPIC VERSUS OPEN NISSEN FUNDOPLICATION FOR GASTRO. OESOPTIAGEAL REFLUX DISEASE 147 3.1.1 AIMS 147
3 1 .2 METHODS t47 3.1 .2.I Participant assignment r47 3.L .2.2 Patient selection and pre-operative assessment 148 3.1 .2.3 Operating technique 148 3.1 .2.4 Post-operative care 1,49 3.r .2.5 Masking t49 3.1 .2.6 Clinical follow-up r49 3.L.2.7 Objective follow-up 151 3-1.2.1 I Oesophageal manometrY 151 3.1.2.7.2 Ambulatory 24hr pH monitoring r52 3.1.2.7.3 Upper gastrointestinal endoscopy r52 3.I.2.8 Statistical analysis r52 3.I.2.9 Ethical approval 153 3.7.3 RESULTS t53 3.1.3.1 Pre-operative assessment r53 3.1 .3.2 Surgery 158 3.1 .3.3 Early hospital outcomes 158 3.1 .3.4 Late hospital outcomes 160 3.1 .3.5 Clinical outcome assessment 160 3.r 3.6 Objective post-operative investigations 161
3.1.4 DISCUSSION t62 v111
3.2 PROSPECTIVE DOUBLE BLIND RANDOMISED TRIAL OF LAPAROSCOPIC NISSEN FUNDOPLICATION WITH DIVISION AND WITHOUT DIVISION OF SHORT GASTRIC VESSELS 166 3.2.1 AIM r66 3.2.2 METHODS 166 3.2.2.I Participant assignment r66 3-2.2.2 Patienf selection and preoperative investigation t67 3.2.2.3 Operating technique r67 3.2.2-4 Post-operative care 168 3.2.2.5 Masking 168 3.2.2.6 Clinical follow-up r69 3.2.2.7 Objective follow-uP r7l 3.2.2.7.I Oesophageal ManometrY t7r 3.2.2.7.2 Ambulatory 24hout pH monitoring r72 3.2.2.7.3 Radionuclide Oesophageal Emptying Study t72 3.2.2.7.4 Barium Swallow examination 772 3.2.2.8 Statistical analysis t72 3.2.2.9 Ethical approval r73 3.2.3 RESULTS r74 3.2.3.1 Preoperative assessment 175 3.2.3.2 Surgery r82 3.2.3.3 Early hospital outcomes 183 3.2.3.4 One to six month post-operative clinical outcome 185 3.2.3.5 Objective post-operative investigations 186 3.2.4 DISCUSSION r87 3.3 PROSPECTIVE DOUBLE BLIND RANDOMISED TRIAL OF LAPAROSCOPIC ANTERIOR PARTIAL VERSUS NISSEN FUNDOPLICATION 191 3.3.1 AIMS 191 3.3.2 METHODS 191 3.3.2.1 Participant assignment 191 3 -3.2.2 Patient selection and preoperative investigation 191 3.3.2.3 Operating technique t92 3.3.2.4 Post-operative care t93 3.3.2.5 Masking 193 3.3.2.6 Clinical follow-up 193 3.3.2.7 Objective follow-uP 194 3.3.2.7.1 Oesophageal ManometrY t94 3.3.2.7.2 Ambulatory 24hour pH monitoring 194 3.3.2.7.3 Radionuclide Oesophageal Emptying Study 194 3.3-2.7.4 Upper Gastrointestinal Endoscopy 194 3.3.2.8 Statistical analysis 195 3.3.2.9 Ethical approval 196
3.3.3 RESULTS 196 3.3 3.1 Preoperative assessment r97 3.3 3.2 Surgery 204 3.3 3.3 Early hospital outcomes 206 3.3 3.4 One to six month post-operative clinical outcome 207 3.3 3.5 Objective post-operative investigations 209
3.3.4 DISCUSSION 2t0 1X
4 CONCLUSIONS 2L4 4.I IS LAPAROSCOPIC ANTIREFLUX SURGERY ADVANTAGEOUS? 215
4.1.1 Outcome of case series studies 215 4.1.2 Prospective randomised trials of laparoscopic versus open Nissen fundoplication 216 4.2 SHORT GASTRIC VESSELS 218 4.3 NISSEN VERSUS PARTIAL FUNDOPLICATION 220 4.3.1 Uncontrolledstudies 220 4.3-Z Experimentalmodels 220 4.3.3 A role for patients with oesophageal dysmotility 22I 4.3.4 Prospective randomised trials of Nissen versus partial fundoplication 22r 4.4 IMPLICATIONS 223 4.5 FUTURE DIRECTIONS 223
5 BIBLIOGRAPHY 227
6 APPENDIX 256 X
ABSTRACT
The advent of laparoscopic surgical techniques has refocussed the medical community's interest in surgery for gastro-oesophageal reflux disease. This has led to a significant increase in the operative caseload in centres interested in surgery for this problem, thereby providing a renewed opportunity for the evaluation of variations to standard surgical techniques which might improve the outcome for patients undergoing antireflux surgery.
Recent published reports highlight promising short term outcomes following laparoscopic antireflux operations, with equivalent control of reflux symptoms demonstrated at short term follow-up, and reduced overall perioperative morbidity. However, long term outcomes following the laparoscopic approach remain unreported, and some unique complications have arisen following laparoscopic antireflux procedures. To reduce the risk of post-operative dysphagia, some surgeons advise either the routine or selective application of a partial fundoplication technique. Although the results of uncontrolled case series suggest that this approach is advantageous, the results of 4 previous prospective randomised trials of Nissen fundoplication versus posterior partial fundoplication, 3 performed by open surgical techniques, did not demonstrate any reduction in the incidence of dysphagia following surgery. The anterior partial fundoplication technique has not been evaluated in a prospective randomised trial.
Initial experience at the Royal Adelaide Hospital identified an increased risk of several complications following laparoscopic Nissen fundoplication. These include; pneumothorax, para-oesophageal hiatus herniation, and hiatal stenosis. As a result of this, a laparoscopic technique was developed during which diathermy was used sparingly or not at all during hiatal dissection, the oesophageal hiatus was routinely repaired posteriorly, and a short loose 3600 fundoplication was calibrated around a large intra- oesophageal bougie. A learning curve, comprising each individuals first 20 procedures, x1
and the institutions first 50 procedures, was also identified. Outcomes improved once each surgeons'experience was beyond this phase.
To improve the outcome of surgery for reflux disease it was hypothesised that the modification of standard open antireflux surgery techniques to a laparoscopic approach, the routine division of the short gastric vessels to mobilise the fundus fully during Nissen fundoplication, and the performance of a partial (anterior) fundoplication would be advantageous. Laboratory research using both a bench top model, and a laparoscopic approach in a live porcine model, compared the efficacy of the Nissen procedure, with both anterior and posterior paftial fundoplication techniques. The results of these studies confirmed that both partial fundoplication techniques restored an adequate antireflux barrier.
A prospective randomised trial of laparoscopic versus open Nissen fundoplication confirmed that the application of laparoscopic techniques reduces the overall morbidity of surgery for reflux disease, and significantly shortens the post-operative hospital stay, and the overall length of convalescence following surgery. However, the duration of the operating time for antireflux surgery was lengthened. A second randomised trial of division versus no division of the short gastric vessels during laparoscopic Nissen fundoplication, demonstrated no reduction in the incidence of dysphagia or any other adverse outcome following division of the short gastric vessels. Vessel division, however resulted in longer overall operating times, and significant postoperative bleeding in one patient. A third randomised trial compared laparoscopic Nissen fundoplication with anterior partial fundoplication. Whilst initial follow-up 1 and 3 months following surgery revealed no advantages for the partial fundoplication technique, significant advantages for the partial fundoplication technique were identified 6 months following surgery.
Specifically, the incidence and severity of dysphagia for solid food was reduced by the partial fundoplication technique, as was the incidence of other adverse outcomes including inability to belch and increased passage of flatus. Overall patient satisfaction was better following the anterior partial fundoplication technique. x11
The studies described in this thesis support the routine application of laparoscopic
techniques to antireflux surgery, but not the routine division of the short gastric vessels
during Nissen fundoplication. The use of an anterior partial fundoplication was
advantageous, although follow-up remains short term. Longer term follow-up will be
needed before it can be recommended that patients requiring surgery for gastro-
oeosphageal reflux disease should routinely undergo an anterior partial fundoplication in
preference to the Nissen technique. Nevertheless, with time this approach may prove to
be the most appropriate surgical operation for gastro-oesophageal reflux disease.
xlv
ACKNOWLEDGEMENTS
I am indebted to many people for their assistance with the studies described in this thesis,
and I am grateful for their co-operation and help.
Professor Glyn Jamieson's supervision, support and wise counsel has been invaluable for
the production of this work. Without access to his patients many of the clinical studies
could not have been undertaken.
Associate Professor Peter Devitt, Mr Robert Britten-Jones, Mr Philip Game and Mr
Randolph Williams also contributed patients to the studies described in sections 1,3.2 and
3.3, and Mr Christopher Stoddard, Mr Malcolm Reed and Professor Alan Johnson
contributed patients to the study described in section 3.1.
Dr Philip Mitchell, Dr George Mathew, Dr Robert Baigrie, Dr Mark Richardson and Dr
Nick Davies, all assisted with clinical data collection at the Royal Adelaide Hospital, and
contributed to the studies reported in sections 1 and 3. Dr John Treacy and Dr Ralph
Gourlay assisted with follow-up data collection at the Royal Hallamshire Hospital
(section 3.1).
Dr Greg Pike was the independent investigator who helped collect follow-up data for
studies 3.2 and 3.3. He also performed the oesophageal emptying studies in these
sections. His willing assistance was invaluable.
Ms Tanya Ellis assisted with the collection of prospective audit data for patients
undergoing laparoscopic antireflux surgery at the Royal Adelaide Hospital. She also
entered data into the prospective computerised data bases maintained by the Royal
Adelaide Centre for Endoscopic Surgery. XV
Ms Jenny Myers and Mrs Nicky Hanna performed the manometric and pH studies in sections I,3.2 and 3.3. Mrs Jenny Globe performed the manometric and pH studies in section 3.1.
Mrs Nicky Hanna, Dr George Mathew and Dr Greg Pike provided assistance with the experiments in sections 2.1 and2.2. Mrs Hanna provided valuable assistance with equipment maintenance, animal handling, and the follow-up studies in section 2-2-
Financial assistance for these studies was provided by grants from the Royal Australasian
College of Surgeons Research Foundation (sections 2.2 and3-2),the University of
Adelaide Faculty of Medicine (section 3.3) and the South Australian Health Commission
(priming grant for the Royal Adelaide Centre for Endoscopic Surgery).
Approval for clinical studies was provided by the Human Ethics Committees of the Royal
Adelaide Hospital (sections I,3.2 and 3.3), and the Royal Hallamshire Hospital,
Sheffield, UK (section 3.1). Approval for the animal experiments (section 2.2) was
provided by the Animal Ethics Committees of the Institute of Medical and Veterinary
Science and the University of Adelaide. xv1
PREFACE
Sections I.2,2.1,2.2 and 3.2 of this thesis have either been published or accepted for
publication as:
Jamieson GG, Watson DI, Britten-Jones R, Mitchell PC & Anvari M. Laparoscopic
Nissen Fundoplicatíon. Annals of Surgery 1994;220: 137-145-
Watson DI, Jamieson GG, Devitt PG, Mathew G, Britten-Jones R, Game PA & Williams
RS. Changing strategies in the perfolmance of laparoscopic Nissen fundoplication as a
result of experience with 230 operations. Surgical Endoscopy 1995;9:96I-966-
Watson DI, Mitchell P, Game PA and Jamieson GG. Pneumothorax during laparoscopic
dissection of the oesophageal hiatus. Aust NZ J Surg 1996;66:7lI-712.
Watson DI, Jamieson GG, Devitt PG, Mitchell PC & Game PA. Para-oesophageal hiatus
hernia: an important complication of laparoscopic Nissen fundoplication. British Journal
of Surgery 1995; 82 521-523.
Watson DI, Jamieson GG, Mitchell PC, Devitt PG & Britten-Jones R. Stenosis of the
oesophageal hiatus following laparoscopic fundoplication. Archives of Surgery 1995;
130: 1014-1016.
'Watson DI, Baigrie RJ & Jamieson GG. A learning curve for laparoscopic
fundoplication: Definable, avoidable, or just a waste of time? Ann Surg 1996;224: 198-
203. xv11
Watson DI, Jamieson GG, Baigrie RJ, Mathew G, Game PA, Devitt PG & Britten-Jones
R. Laparoscopic surgery for gastro-oesophageal reflux: Beyond the learning curve. Br
J Surs 1996; 83: 1284-1281 .
Watson DI, Foreman D, Devitt PG & Jamieson GG. Preoperative grading of oesophagitis versus outcome following laparoscopic Nissen fundoplication. The
American J ournal of Gas tro ente rolo gy 1991 ; 92: 222-225.
Mathew G,'Watson DI, Myers JC, Holloway RH & Jamieson GG. Oesophageal motility parameters before and afær laparoscopic Nissen fundoplication. Br J Surg 1997;67:
289-292.
Baigrie RJ, Watson DI, Myers JC, Jamieson GG. The outcome of laparoscopic Nissen
fundoplication in patients with disordered pre-operative peristalsis. Gut 1997;40: 381-
385.
Watson DI, Chan ASL, Myers JC & Jamieson GG. Illness behaviour influences the
outcome of laparoscopic antireflux surgery. The Journal of the American College of
Surgeons 1997 ; 184: 44-48.
'Watson DI, Mathew G, Pike GK, Baigrie RJ & Jamieson GG. Efficacy of anterior,
posterior and total fundoplication in a pig model. Br J Sørg (1998) - In Press'
Watson DI, Mathew G, Pike GK & Jamieson GG. Comparison of anterior, posterior and
total fundoplication using a viscera model. Diseases of the Esophagus 1991;10: 110- rl4. xv111
Watson DI, Pike GK, Baigrie RJ, Mathew G, Devitt PG, Britten-Jones R & Jamieson GG.
Prospective double blind randomised trial of laparoscopic Nissen tïndoplication with division and without division of short gastric vessels. Annals of Surgery 1997;226: 642-
652
Copies of the front pages of these papers are included in the appendix. xlx 2
sEcfloM 1
OUTCOME FOLLOWING LAPAROSCOPIC ANTIREFLUX SURGERY 3
7.7 LAPAROSCOPIC ANTIREFLUXSURGERY LITERATURE REVIEW
No currently practising general surgeon would be unaware of the recent and rapid development of minimal access surgical techniques, largely made possible by the development of new equipment and technology for the operating theatre. In a short time period laparoscopic cholecystectomy has become the standard treatment for symptomatic gallstone disease in the developed rrs.l¿128,189,224. Procedures such as laparoscopic inguinal hernia repair and laparoscopic colonic resection, however, have failed to achieve similar acceptance230,232. These procedures have been criticised because of their potential to cause new and different complications to those seen following open approachesl2T ,I51. The current status of laparoscopic antireflux surgery may lie somewhere between these positions, although it is rapidly gaining acceptance as the preferred surgical treatment for severe gastro-oesophageal reflux disease, and may with time become the standard surgical approach-
Since laparoscopic Nissen fundoplication was first described in 1991 by both Geagea64
and Dallemagne et al39,laparoscopic antireflux surgery has been embraced
enthusiastically in many centres11,33,34,55,143,168,177 . Nevertheless, it remains a
procedure under close scrutiny, until satisfactory long term outcomes have been
established. Whilst long term outcomes will not be available for many more years, the
early results of laparoscopic antireflux procedures have now been reported by many
groups3,16 ,I9,24,30,37 ,59,65,6J ,69-J2,77 ,80,90,139,142,194'20I'206'217 '219'228.
1.1.1 OPEN NISSEN FUNDOPLICATION
Fundoplication was first described by Nissen in 1956158 ¡s[s'ù/ing the serendipitous
discovery that a fundal patch, used to reinforce an oesophageal suture line, also corrected
gastro-oesophageal reflux. Whilst achieving good control of pathological reflux in the
majority of patients186, an incidence of adverse sequelae has led to the subsequent 4 modification of Nissen's original technique. Shortening the fundoplication length to 1 to
2 cm, dividing the short gastric vessels to achieve tull mobilisation of the gastric fundus, calibration of a loose fundoplication over a large intra-oesophageal bougie, and modifi.cation of the complete fundoplication to one of a number of partial fundoplication variants, have all been advoc a¡s643,46,49,76,205,211. Few of these modifications, however, have been evaluated within the context of prospective randomised
¡i¡s45,75,179,120,125,209 .
Following four decades of experience with open antireflux procedures, long Ûerm outcomes following open surgery have been well described. The post-operative hospital stay following open surgery ranges from 7 to 14 days4s,zII, and short term morbidity is acceptable 43 ,7 5 . Long term success is achieved in the maj ority of patients43 '124 .
Rossetti reported that87.5Vo of patients followed for more than 10 years following a 3600 fundoplication without short gastric vessel division, were free of reflux symptoms and adverse sequelas186. Similarly, DeMeester repofted agl%o success rate for the Nissen fundoplication procedure at an average follow-up of 45 months43. Other studies report long term success rates of 85 to 907o following open antireflux surgeryg3,llg'123'211.
RecentlyLuostarineç¡s¡s,¡I24 reported thaf.7l7o of aseriesof 25patientswhounderwent
the combination of Nissen fundoplication and pyloroplasty remained free of reflux 20
years or longer after the original procedule.
Adverse outcomes following open Nissen fundoplication include persistent dysphagia and
the gas bloat syndrome43,75,93. Although less common in late follow-up, many patients
report troublesome early dysphagial19. The assessment of dysphagia symptoms,
however, can be difficult because variable methods of clinical assessment and scoring are
used in the various reported studies. There is also the potential for the introduction of
assessment bias by clinician investigators, pafticularly if follow-up is performed by the
operating surgeon. Despite these difficulties, it is likely that the incidence of persistent
long term troublesome dysphagia is less than 57o following Nissen
fundoplica¡iç¡¡43,7 5,93,1I9,119,123. Uncontrolled reports suggest that the performance 5 of a partial fundoplication variant may reduce this inciden ce2II. However, this proposal has not been supported by the results of published randomised tfiah45,93,1'19'120'209
(see below). If laparoscopic antireflux surgery is to achieve uniform acceptance by the medical community, then the short term surgical morbidity traditionally associated with antireflux surgery must be reduced by this new method of surgical access, and long term outcomes must equal those reported following open surgery.
Tabte 1.1.L Expected outcomes fotlowing open Nissen fundoplication
Short term
Complication rate 1g-29Eo724
Incidental splenectomY 73o7o 43,75,124
Mortality .g-1o7o43,50
Length of hospital stay I to 14 days45,ZlI
Long term
Control of reflux at 5 Years+ 29gEo43,50,I19
Good or Excellent long term outcome 9gEo43,I19,I23,186
Long term troublesome dysPhagia a5Eo43,50
1..I.2 RANDOMISED TRIALS FOR OPEN FUNDOPLICATION
Despite the large number of publications describing fundoplication outcomes in the era of
open surgery, few randomised trials have ever been conducted (Table 1.I.2). Three trials
have compared a Nissen fundoplication with a postedor partial
fundoplica¡isrJl9,120,203,209, and one has compared a Nissen fundoplication with a
Hill and a Belsey procedurel19. None of these studies describe any significant increase
in the likelihood or severity of dysphagia following Nissen fundoplication, compared to a
posterior partial fundoplication procedure. HoweveL, with the exception of the study
reported by Lundell et al119 which enrolled 137 patients, all of the studies assessing the
partial fundoplication variants enrolled small numbers of patients, which were inadequate
to allow one to draw statistically valid conclusions. Other small studies have compared 6
Nissen fundoplication with the Angelchik prosthesis and the Ligamentum Teres cardiopexy, demonstrating advantages for Nissen fundoplica1isn75,9l, and a recent larger study has compared medication with surgical therapyl95
Table 1.1.2 Randomised trials investigating open antireflux surgery
Tríat No Patíents entered
Nissen vs posterior partial fundoplica¡ie1l19' 120'124'209 137 ,33, 52
Nissen vs Hill vs Toupet4s 45
Nissen vs Angelchik prosthesisTs 61
Nissen vs Ligamentum Teres cardiopexy9l 20
Nissen vs medicati6l195 247
Nissen with vs without division of short gastric vessels 125'126 25
l.l.2.l Nissen versus partial fundoplication
1.1.2.1.1 Thor and 5¡¡oo¿rr" 124
This study enrolled 31 patients, who were randomised to undergo either a Nissen
fundoplication (12 patients) or a Toupet posterior 1800 to 2gg0 partial fundoplication (19
patients). The Nissen wrap was 4 cm in length, and it was calibrated over a 40 Fr bougie.
The oesophageal hiatus was not repaired, the hepatic branch of the vagus nerve was
divided, and the short gastric vessels were not divided. Follow-up was for 5 years. A
good or excellent outcome was achievedinBl12 of the Nissen group and 18/19 of the
Toupet group. However, because of the small number of patients enrolled, this
difference did not reach statistical significance (P=0.06, Fisher's exact test; calculated
from data reported by the authors in their publication).
The incidence of post-operative complications was similar, as was the control of reflux at
5 years. Four of the 12 Nissen patients experienced persistent dysphagia, compared to 2 7 of the 19 Toupet patients (P=0.18, Fisher's exact test; calculated from author's data).
The resting lower oesophageal sphincter pressure measured by post-operative manometry was higher following Nissen fundoplication (19.5 mmHg vs 16.9). Three of the patients who underwent Nissen fundoplication underwent further surgery for dysphagia. In each instance this was for the development of a'slipped Nissen'. No reoperations were required in the Toupet group.
This study can be criticised for failing to enrol an adequate number of patients to have any chance of finding statistically significant differences between the clinical outcomes following the two procedures. It could also be argued that the Nissen fundoplication technique used would no longer be acceptable, as current practice suggests that the fundoplication should be shortened to 1 to 2 cm, and a larger calibrating bougie is now routine for most surgeons43,9O. Also reported in this trial was a very high incidence of reoperation for the 'slipped Nissen' phenomenon. The 257o reoperation rate for this problem is far in excess of the low rate reported in other 5¡u¿is543'119't25'209, suggesting that the authors technique for Nissen fundoplication may have been defective.
1.1.2.1.2 Walker t¡ o¡209
This study compared a Nissen fundoplication performed with selective division of the short gastric vessels, with a 3gg0 posterior paltial fundoplication (Lind). As only 26 patients were enrolled in each group, the same statistical criticisms which applied to the study from Thor and Siland erI24 can also be made of this study. The Nissen fundoplication was 3 cm in length and it was calibrated over a 40 Fr bougie. New
dysphagia vvas seen post-operatively at 6 weeks in 8 of the Nissen group and 6 of the Lind
group (P=0.75, Fisher's exact test; calculated from authors data). Persisænt late
dysphagia was more likely following posûerior partial fundoplication (4 patients) than
Nissen fundoplication (2 patients, P=0.67). The incidence of early and late gas bloat
syndromes was identical, as was the rate of post-operative complications. No advantages
for the posterior partial fundoplication technique were demonstrated by this study. 8
1.1.2.1.3 Lundell et øl lI9'120
This study ,was reporte d tn 2 separate publication s120,I20 - The first report described 6 month post-operative outcomes in 71 patients (38 Nissen fundoplication without dividing short gastric vessels vs 33 Toupet partial fundoplications)120. Resting lower oesophageal sphincter pressure was higher in the Nissen group (20 mmHg vs 14). No differences were seen for any clinical outcome except for dysphagia at 3 months. At six months the incidence of dysphagia was identical. Dysphagia was experienced \n 15 (39Vo) of the
Nissen group and3 (97o) of the Toupet group at 3 months (P=0.005). At 6 months the incidence of dysphagia had fallen to 4 (I\Vo) and2 (6Vo) tespectively (P=0.68)-
Hearrburn was well controlled in31 (91%) of the Nissen patients and 31 (94Vo) of the
Touper parients. Belching was said to be notmalin34 (89Vo) and 30 (9IV") respectively.
The incidence of gas bloat symptoms and flatulence following the two procedures were also similar. The initial results of this trial demonstrated no important outcome differences at 6 months.
A subsequent report described outcomes following 3 to 5 years follow-up in 137 patients
(65 Nissen,72Toupeg119. Three pationts underwent splenectomy during their initial
Toupet fundoplication, versus no patients in the Nissen group. Dysphagia at 5 years was
more likely following partial than Nissen fundoplication (l6Vo versus 10%, P=NS),
although in all instances the symptom was repofted to be mild. Flatulence was commoner
after Nissen fundoplication at 2 and 3 years but not at other earlier or later time intervals.
Recurrence of reflux occurred in 67o of the Toupet group, and 5Vo of the Nissen group.
Reoperation was more common in the Nissen group than in the Toupet group, although
this did not reach statistical significance (5i65 vs lll I: P=0.10, Fisher's exact test). One
patient in the Toupet group underwent further surgery for severe gas bloat symptoms.
Five of the Nissen group underwent reoperation for a post-operative paraoesophageal
hiatus hernia. Significantly, hiatal repair was only performed infrequently in this trial,
and in only one of the five patients who developed a postoperative hernia. The authors
concluded that Toupet partial fundoplication performed better due to the lower
reoperation rate, and the lower incidence of late flatulence. However, the dysphagia rate 9 was not improved following partial fundoplication, and it is possible that the reoperation rate could have been greatly reduced in the Nissen group if the hiatus had been routinely repaired at the original procedure, suggesting that the two procedures probably produce similar long term outcomes.
1.1.2.1.4 DeMeester et ø145
This trial randomised 45 patients to undergo one of 3 operations (Nissen fundoplication
15, Hill repair 15, and Belsey repair 15). Follow-up to 6 months was reported. As with the study by Thor and Silanderl24,this study was statistically inadequate as it enrolled only a small number of patients. The dysphagia rate 6 months after surgery wa.s similar for all 3 procedures. Reflux recurred early in 1 patient following the Hill procedure and in 2 following a Belsey procedure. No patients in the Nissen group developed recurrent reflux. Patients remained in hospital for 20 days following the transthoracic Belsey repair, and for 12 days following the other 2 plocedures. Due to the small number of patients enrolled in this study, it is diffîcult to draw any firm conclusions about the relative advantages of the different approaches.
If the outcomes of all 4 randomised trials are considered, it is apparent that no
improvement in the dysphagia rate following posterior partial fundoplication has been
demonstrated, and the overall outcome for the procedures compared appears to be similar.
Only the trial reported by Lundell et all19 enrolled a statistically credible number of
patients. Nevertheless, the manometrically recorded lower oesophageal sphincter
pressure was slightly higher following Nissen fundoplication, although this did not have
any clinical implications. As the studies have only effectively compared a posterior
partial fundoplication with the Nissen procedure, however, these conclusions cannot be
extrapolated to the anterior partial fundoplication, and it is possible, but certainly not
proven, that this modification may result in outcome improvemen15211. 10
1.L.2.2 Other trials
Other small studies have compared Nissen 1'undoplication with the Angelchik prosthesisT5, and Nissen fundoplication with the Ligamentum Teres repair9l. Whilst these studies also only enrolled a small number of patients, because of the magnitude of the differences identified, they did reveal distinct advantages for the Nissen procedure, in terms of the rates of surgical revision and the incidence of recurrent reflux.
1.1.2.2.1 Hill et at75
This trial randomised 61 patients to undergo either a Nissen fundoplication without division of the short gastric vessels (31) or the placement of an Angelchik antireflux prosthesis (30). Follow-up was over a7 yeat period, with a good long term result obtained in I7122 from the Angelchik group and20125 of the Nissen group. Two of the
Angelchik prostheses were removed for persistent dysphagia, and one more because of postoperative infection. Five patients also had pelsistent dysphagia following placement of the Angelchik prosthesis. No Nissen fundoplications required surgical revision. One patient required endoscopic dilatation for dysphagia following Nissen fundoplication.
Whilst long term outcomes were similar, the Angelchik prosthesis was associated with a higher likelihood of surgical revision.
1.1.2.2.2 Janssen et øl9I
Janssen et al randomised 20 patients to undergo either Nissen fundoplication (10 patients)
or Ligamentum Teres cardiopexy (10). Although both procedures effectively corrected
reflux for the first 3 months following surgery, by 12 months, 6 of the 10 patients who
underwent the Ligamentum Teres repair required further surgery for recurrent reflux,
compared to only one in the Nissen group (P= 0.05). The results of post-operative
ambulatory pH monitoring were very poor in the patients undergoing the Ligamentum
Teres repair, with a mean acid exposure time of 24.0Vo versus 3.ïVo Tor the Nissen
procedure (P< 0.05). Despite the small number of patients in this trial, the results of the
Ligamentum Teres repair were so poor that continued use of this procedure by either
laparoscopic or open techniques cannot bejustified. 11
1.1.2.2.3 Luostarínen et øl 125'126
Luostarinen et al recently reported the results of a small trial of division versus no division of the short gastric vessels during open Nissen fundoplication. Twenty five patients were entered into this trial, which was repofted after the commencement of a larger study during laparoscopic Nissen fundoplication at the Royal Adelaide Hospital
(described in section 3.2). No significant outcome differences were demonstrated. The results of this study are discussed in detail in sections 3.2 and 4.
1.1.2.2.4 Spechler t¡ o¡ I95
Spechler ¿195 describe the only study to compare operative with non-operative therapy "¡ for reflux. In this study, performed in veterans hospitals in the USA, 247 patients (243 men and 4 women) were randomised to one of three groups; continuous medical therapy with a H2 blocker, medical therapy for symptoms only, or open Nissen fundoplication.
Forty patients withdrew from the trial following randomisation, 24 of these from the surgery group. 176 were followed for at least 1 year, and 106 for 2 years. Seven patients' symptoms persisted on medical therapy to the extent that they were reallocated to a surgical procedure. Overall patient satisfaction was highest in the surgical group at both the 1 and 2year follow-up intervals. However, neither the surgical approach, nor the
medical treatment trialed in this study, would now be regarded as optimal management
for reflux disease in the new era of laparoscopic surgery and proton pump inhibitor
medications.
1.1.3 INDICATIONS FOR SURGERY
The advent of the laparoscopic approach to antireflux surgery now offers a potentially
more acceptable surgical option for patients with severe gastro-oesophageal reflux.
Because patients anticipate reduced post-operative pain and a more rapid recovery, they
perceive that surgery is now a more acceptable treatment option. This has led to many
symptomatic patients, who previously did wish to undergo open antireflux surgery, to T2 now request a laparoscopic fundoplication8O,106, 1s¿¿ing to many centres rapidly gaining large experiences with laparoscopic antireflux surgery3'29'59"70'77 '80'194'217 '228
An expanded role for surgical management has therefore been proposed, with some surgeons now claiming that earlier surgical correction of pathological reflux may prevent progression to the complications of Barrett's oesophagus or stricture formation2g'101'106,
It should be recognised, however, that this opinion is speculative and has not yet been supported by research outcomes. The fact that successful results cannot always be guaranteed following any variant of antireflux surgery, should mandate an ongoing policy of careful patient selection. The indications for surgely have not changed with the advent 0'7 7'90'228 of lap aro sc opic fundoplic ation7 .
Abnormal gastro-oesophageal reflux disease should be demonstrated preoperatively either by ambulatory 24 hour pH monitoring, or endoscopically by the presence of oesophagitis at endoscopic examin v¡i6¡¡32,2l0. As well patients should have either ongoing reflux symptoms despite adequate medical management, pteferably with a proton pump inhibiting agent, or they should have expressed a desire to avoid lifelong medication-
Indications for surgical management remain unchanged in the laparoscopis s¡v10'77 '90.
I.I.4 APPROPRIATE SITUATIONS FOR THE APPLICATION OF A LAPAROSCOPIC APPROACH
In the early experience of many surgeons, difficulties were encountered with laparoscopic
antireflux surgery in patients who had a hypertrophied left lobe of the liver, in patients
with a large para-oesophageal hiatus hernia, in the morbidly obese, and in patients who
had undergone previous upper abdominal surgical procedures3,19,80,90. However,
recent reports suggest that many of these problems can be overcome with improved
experience and the availability of specific pieces of equipme¡70'206'217 - Many
surgeons have now confirmed the feasibility of repair of giant paraoesophageal hernia's
associated with intra-thoracic migration of the stomach2l '21 '35'53'66'160'I'78'185'187.
Improved liver retraction devices, such as the Nathanson Liver Retractor (Cook Australia, 13
Brisbane, Australia) now facilitate excellent and stable exposure of the oesophageal hiatus, even in the presence of an enlarged 1iver70. Adhesions secondary to previous upper abdominal surgery often do not present significant difficulties48,77,142,219, except in patients who have undergone previous proximal gastric or hiatal surgery25,II4,2l9,220- In the latter scenario, dense adhesions involving the hiatal region may be difficult to safely dissect laparoscop ica[y25.
1.1.5 EARLY RESULTS AND COMPLICATIONS FOLLOWING
LAPAROS COPIC FUNDOPLICATION
Initial reports of small case series of laparoscopic Nissen fundoplication with follow-up of
3 months or less first appeared in the published literature in 1991 and 199239'55'64.
Although these studies confirmed the technical feasibility of laparoscopic antireflux surgery, the lack of adequate follow-up data and the small patient numbers precluded any adequate assessment of the merits of the procedures described. The first large study was published by Cuschieri et al in 1993, who repofied promising results from a multicentre series of 116 patients34. Further large single centre experiences describing series of more than 100 patients have been published si¡çe3,4,16,29,37,59,70,77,80,90,142,201,206,219,228,with follow-up intervals of 2 to 3 years described in some later 5¡u6is529,80'206'2ll -
The clinical outcomes from the larger published series are summarised in Table 1.1.3.
Mean or median operating times vary from 30 to 185 minutes in these studies. Some of the reported variation in operating times may be due differences in laparoscopic technique, as well as possible increased operation times in some centres involved in
teaching surgical ¡¡ai¡¡s677 ,213. Reported complication rates rango from 2 to
26Eo70,71,77 ,zIJ , and surgical revision has been needed in a small group of patients in
most series. Variation in these rates may be influenced by the effect of the institutional
learning curve, technical factors associated with the choice of surgical technique, and
different criteria used for the recognition and classification of complications in different
reports. For example dysphagia rates may depend on the choice of clinical measure used I4 to determine post-operative dysphagia, and who applies the measure chosen. The operating surgeon may elicit a different response to an independent investigator.
Table 1.1.3 Reported outcomes following laparoscopic fundoplication
Author Procedure No of patients Conversion Operating Time Complication Recurrenlon- performed rate (Vo) (min) rate (7o) going reflux (7o)
Cuschieri35 LNF- 716 0.9 150 (mean) 13 8.6
Jamieson90 LNF- 155 12.3 120 (median) 93* 1.9 waßon217 LNF- 174 9.2 80 (median) 2.3 0.7 y¿""t¡s228 LNF+ 132 3.3 118 (mean) 7.5 1.1
Dallemagne3T LNF+ 368 7.4 107 (mean) 4.0 3.2 * HinderTT LNF+ 198 3.0 150 (median) 11.6 8.0 Cadierel9 LNF+ 80 3.8 150 (median) 8.8 0.0 HallerbackTl LNF- ó0 15.0 110 (median) 10.0 2.0 Fontauma¡d59 LNF- 148 0.0 90 (median) 72.5 0.0 ¡4"çsrnan142 PPF 283 0.5 82 (mean) 3.5 0.5 GotteyTo LNFI 200 4.5 149 (median) 8.0 0.5 Anvari3,4 LNF- 195 2.0 68 (mean) 8.9 0.6 1¡un6180,206 LNFI 300 1.5 185 (mean) rt.7 0;t
Catteyz2 LNF+ 100 2.0 101(mean) 8.0 6.5 çor¡s¡29,30 LNF- /PPF 231 2.2 30 (mean) 12.0 6.0 '1.4 Medina145 LNF+ 74 139 (mean) 4J* 6.7 Bittne.11 LNF- 35 14.3 110 (mean) 25.7 3.8 p¡¡.¡"t177 LNFI 70 2.9 186 (mean) 4.3 0.0
O'ReiUY159 PPF 100 0.0 ? 4.0 0.0 ¡4o*n¡"¡152 PPF 51 2.0 140 (median) 9.8 4.1
gastric Key: LNF+ = laparoscopic Nissen fundoplication with division of short. vessels gastric vessels LNF- = taparoscopic Nissen fundoplication without rlivision of short short gashic vessels LNF1 = úparoscópic Nissen fundoplication with selective division of PPF = laparoscopic posterior partial fundoplication * - post-operative death
Laparoscopic Nissen fundoplication is reported to control reflux symptoms in 91 to I00Vo
of patients followed for up to two years206,217, results which mirror previous experience
with open anti-reflux surgery43'93'123'186' Post-operative hospital stays have been
short in all published reports, with mean/median stays ranging from 2 to 5
days70,77,80,142,217. Overall results from these initial case series suggest that 15 laparoscopic antireflux surgery is effective, and that it does result in an overall reduction in the short term morbidity associated with surgery for reflux. However, several complications unique to the laparoscopic approach have now been described (Table
1.1.4).
Table 1.1.4 unique or common complications following laparoscopic antireflux surgery
a pneumothorax 1 0,9 8, 1 37,182,225
a Pneumomediastinum 1 63,I9 6
a Pulmonary embolism9 0'I 53'I7 2
a Injury to major vessels7
a Paraoesophageal hiatus hernia92' I 53,220
a Hiatal 5¡enssis221
a Mesenteric thrombosis I 45,I 46
Oesophageal perforati on26'l I7'I 53]88/0I
a Gastric perforationgO' 1 53' 1 8 8
o Duodenal perforationTT
a Bowel perforation26
a Cardiac laceration5T
a Pleuroperic¿¡di1i5208 16
1.1.5.1 Pneumothorax
Left sided pneumothorax may occur intraoperatively following an injury to the lett pleural membrane caused by high retro-oesophageal dissection10,98,137,182,225- The risk of this may be up to 2Vo, although it is possible that careful dissection technique and experience may reduce its likelihood. Its occunence does not mandate the placement of a chest drain as the CO2 gas in the pleural cavity is rapidly resorbed at the completion of the procedure. Pneumomediastinum is also descri6"¿163,196 although it is rarely a clinical problem.
1.1.5.2 Vascular injurY
Vascular injury to the inferior vena cava, the left hepatic vein, the abdominal aorta, and the inferior phrenic vessels have all been reported7. It is possible that this complication is associated with a combination of aberrant anatomy, inexperience, and the excessive use of monopolar diathermy cautery dissection. Intraoperative bleeding more commonly follows inadvertent laceration of the left lobe of the liver by a laparoscopic liver retractor or other instrument, and haemorrhage from poorly secured short gastric vessels during fundal mobilisation. A rare complication is cardiac tamponade, which was reported by
Firoozman d et aI57 , due to laceration of the right ventricle by a liver retractor.
The overall risk of perioperative haemorrhage associated with antireflux surgery, however, may well be reduced by the laparoscopic approach77,90- The risk of
splenectomy due to inadvertent splenic injury during Nissen fundoplication is
dramatically reduced by laparoscopic appro aç¡77 ,219. Whilst splenectomy rates of 1 to
3qo43,75,124 have been reported in association with open fundoplication, splenectomy is
a.Íare event following laparoscopic fundal mobilisation. It is likely that the more precise
dissection and better anatomical exposure afforded by the laparoscopic approach
minimises the likelihood of inadvertent splenic trauma- 71
1.1.5.3 Paraoesophageal hiatus herniation
Although para-oesophageal hiatus herniation may occasionally present in the late follow- up period following open fundoplication 1 19, most large case series of the laparoscopic procedure report an incidence of para-oesophageal herniation, particularly in the immediate postoperative period24,70,77 ,90,92,220. The incidence of this has ranged up toTVo in publishedrepots24,220. This problem may be more likely following laparoscopic surgery, due to the tendency to extend laparoscopic oesophageal dissection further in to the thorax than during open surgery20l, the increased risk of breaching the left pleural membrane10,98,137,182,225, and the effect of reduced post-operative pain.
Reduced pain may permit more abdominal force to be transmitted to the hiatal area during coughing, vomiting or other forms of early exertion, thereby pushing the stomach into the
¡¡o¡a*220. It is possible that routine hiatal repair may reduce the risk of this problem7O,220. ¡¡ is also possible that some of the apparent variation in this incidence of the problem may reflect differences in the steps taken by different investigators to actively look for this complication. The incidence following open fundoplication may also be greater than previously recognite¿ 1 19.
1.1.5.4 Dysphagia
Early severe dysphagia requiring surgical revision has been reported in a number of
,"¡"124,25,70,114,153,22I. Conversion of a Nissen fundoplication to a posterior partial fundoplication has been performed for troublesome dysphagia by both open and laparoscopic techniques, usually with success48,70,1l4,I53. Over tightening of the oesophageal hiatus, another potential cause of early dysphagia, may be difficult to assess intraoperatively because ofreduced tactile feedback during laparoscopic surgery. This can be corrected by early laparoscopic re-intervention to release one or two hiatal sutures
Late narrowing of the oesophageal hiatus due to post-operative scarring, even in patients not undergoing hiatal repair, has also been reported in the Royal Adelaide Hospital
experience22l gee 1.2.4.3). It is possible that this complication is due to an
idiosyncratic response to diathermy dissection which occurs in certain patients. 18
Correction requires widening of the diaphragmatic hiatus by either open surgical or laparoscopic techniques.
1.1.5.5 Pulmonary embolism
Pulmonary embolism has been reported following laparoscopic antireflux surgery9O,l53,l72. This is likely to result from a combination of primarily mechanical factors. The combination of head up tilt of the operating table, intra-abdominal insufflation of gas under pressure, and elevation of the legs of patients in stirrups, may greatly reduce venous flow in leg veins97. Coagulation changes associated with surgery are not altered by the application of laparoscopic techniqueslT2. To prevent this problem it is mandatory that anti-thromboembolism prophylaxis is routinely applied-
1.1.5.6 Perforation of the gastrointestinal tract
Perforation of the oesophagus, stomach, duodenum and bowel have all been described following laparoscopic antireflu* surge.y26,'17,90,717,153,188,201. 1y¡i1st perforation of the small bowel and the colon may be related to laparoscopic access in general, oesophageal and gastric perforations are specific risks oflaparoscopic antireflux surgery, with an incidence of approxim ately lVo leported by most ,"¡626'77 '90'Ll7 '153'188'201'
Perforation of the back wall of the oesophagus can occur when dissecting the posterior aspect of the oesophagus with laparoscopic instrume¡618,188. The anterior oesophageal wa1l is at risk when a bougie is passed to calibrate the tightness of the Nissen
*tap117,188. Gastric perforation is usually an avulsion injury of the gastric cardia due to rough handling by the surgical assistant, or the use of inappropriate grasping instruments to retract the stomac¡90,188. These injuries can be repaired by sutures, applied either
laparoscopically or by an open technique. Awareness of the potential for these injuries
may reduce the likelihood of their o.curr"nce18. 19
1.1.5.7 Mortality
Three deaths have been described following laparoscopic antiretlux procedure77,l45'146-
Hinder et al77 describe a patient who died secondary to peritonitis and sepsis from a duodenal perforation, presumably due to an instrumental injury occurring outside of the restricted field of view provided by the laparoscope. Mitc¡"1146 described an instance of thrombosis of both the superior mesenteric artery and the coeliac axis. This rare complication has also been seen following laparoscopic cholecystectomyl92, suggesting that it may be related to the use of positive pressur€ insufflation54. R further case of infarction of the liver following laparoscopic fundoplication was reported by Medina et u¡I45.
I.I.6 LAPAROSCOPIC VERSUS OPEN SURGERY
Although laparoscopic approaches may reduce shoft term surgical morbidity, it is unreasonable to expect better long term outcomes than those following equivalent open procedures. It is even possible that technique modifications introduced to facilitate various laparoscopic approaches could result in poorer long term outcomes. Until the results of long term studies are available the true outcome of laparoscopic antireflux surgery, and its status compared to open antireflux Surgery remains unknown.
Few direct comparisons of clinical outcome, surgical morbidity and cost effectiveness have been attempted between early series of laparoscopic Nissen fundoplication and open surgical experience (historical or otherwise) within single institutionsz3,60,82,110,170,181,183, and of the cost comparisons attempted, all are within the context of the American health sysþm81'82,110,183. Given the constraints of the non randomised research methodology for almost all of these studies, and the different
cost structure of American health care, a fair comparison of laparoscopic and open
techniques is difficult. All published studies report that laparoscopic antireflux surgery
reduces hospital treatment costs and early surgical morbidity23,60,82'110,170,181,183'
Similar advantages were identified by non-randomised studies which compared
laparoscopic inguinal hernia repair, appendicectomy and cholecystectomy with their 20 equivalent open procedures13,104,224. However, these advantages have not all been
I 32'134'75 4 supported by prospective randomised trials -
1.1.6.1 Clinical outcome comparisons
The non-randomised comparisons between open and laparoscopic fundoplication are summarised in Table 1.1.5. Although laparoscopic surgery generally required more operating time than the equivalent open surgical procedure, one groop23 reported that the laparoscopic approach was actually quicker. The incidence of post-operative complications was reduced, the length of post-operative hospital stay was shortened by 3 to 7 days, and patients returned to full physical function between 6 and 27 days quicker following laparoscopic antireflux surgery. Overall efficacy of reflux control was similar.
Table L.L.5 Comparison of laparoscopic and open fundoplication: non- randomised studies
Author Number of op thne Hospitul stay complícatíons Tilne off work Good' outcome procedures (míns) (¿l"yÐ (E ) (¡l"yÐ (E')
Cnampautt23 Lap 32 t25 5.4 ? 2t 9l Open 29 t40 8.9 ? 38 9l Frantzides60 Lap 36 186 1.8 2.8 ? 97 Open 29 168 9.2 ll.2 ? 100 RicnardslS3 Lap 12 tt3 1.5 11 t6 ? Open 160 100 6.1 37.5 43 ? RaturerlSl ,| Lap 74 \73 4.0 24 ? ,| ,| Open 72 769 8.1 30 p"¡"¡*82,170
Lap 34 22r 5.8 ? ? 94 ,| Open 47 165 8.8 ? 95 2I
L.1.6.2 Cost comparisons
Four non-randomised studies have compared the cost of laparoscopic antiretlux surgery with an open procedure, all within the Amedcan health care setting81,82,110,183. ¡s cost comparisons have been reported from Europe or Australasia. The outcome of these comparisons is summarised in Table 1.1.6. All of these studies demonstrated that laparoscopic Nissen fundoplication is cost effective within the American health financing system. A cost saving of between US$555 and US$8906 per case, was achieved by the laparoscopic approach, due to reduced post-operative hospital costs. Operating theatre
increased by laparoscopic surgery. The study reported by Laycock ¿110 costs were "¡ may be criticised as it compared a laparoscopic Nissen fundoplication with a Belsey repair through an open thoracotomy incision.
It is difficult to translate the potential savings identified in these studies into savings in the
Australian health care system. The analyses may in fact underestimate the potential benefits of laparoscopic fundoplication as they did not take into account the potential community savings which might be achieved by the shortened post-hospital convalescent period. If this is factored in, then any reduction in the time patients spend away from productive work due to the laparoscopic approach will lead to important community savings. However, it remains to be seen whether major cost savings can be demonstrated by prospective randomised trials. 22
Table 1.1.6 comparison of the cost of laparoscopic and open fundoplication (all figures in US$)
Author Number of operating roorn Bed cost TotøI hospital cost patìenß cost p¡.¡u¡¿s183
Lap 72 $2324 $2007 $4331
Open 160 $1287 51329 $8616
IncarboneS2
Lap 11 $16730 $9904 $26,634
Open 9 $14283 $12906 927,189 IglesiasSl Lap 25 $3253 $7848 $11101
Open 34 $2724 $11s62 $14286 Laycockllo Lap 20 $6077 $9316 $15393
Open 20 $5292 $19007 $2A299
1.1.6.3 Randomised comParisons
Three randomised controlled trials which compare a laparoscopic Nissen fundoplication with its open surgical equivalent have been recently reported6l'108'216, two in abstract format only. One of these is described in detail in section 3.1, and a full discussion of the results of the other trials is included in section 3.I.4. The studies from Franzen et al6I and Laine ¿108 were both reported recently, after the commencement of the "¡ randomised trial described in this thesis216. Nevertheless, the results of all of these early trials confirm advantages for the laparoscopic approach, albeit less dramatic than the
advantages expected from the results of non-randomised studies.
I.I.7 RE.OPERATIVE SURGERY FOLLOWING LAPAROSCOPIC FUNDOPLICATION
Most large series of laparoscopic anti-reflux surgery describe an incidence of adverse
outcomes requiring subsequent surgical revision24,'|0,77,114,206. The incidence varies
with the length of post-operative follow-up and the criteria used for reporting. An
incidence of 2Vo to 6Vo at short term follow-up is usual, with many of the reoperations 23 required to correct unique complications of the laparoscopic approach26 ,92,117 ,I53,153,188,20I,220,221. Early identification of these problems may facilitate early laparoscopic repair213. However, later identification, beyond the first post-operative week often necessitates a more difficult open surgical procedure, due to adhesion formation and fibrosis rendering the laparoscopic option in many instances imPossible 25,2I3'22I -
Laparoscopic revision procedures performed more than 3 months after the original fundoplication have been also been descrilo"¿48,213. However, the expertise and experience needed to perform these technically demanding operations is not widely available. On the other hand, laparoscopic revision within a few days of the original procedure for complications such as post-opelative paraoesophageal herniation or early severe dysphagia are technically less demanding, and may not add greatly to the overall morbidity of the laparoscopic approachTT ,2I3. To facilitate this, many surgeons recommend routine early post-operative barium swallow x-rays to enable early identification and laparoscopic correction.
1.1.8 RANDOMISED TRIALS FOR LAPAROSCOPIC SURGERY
Apart from the 3 randomised trials discussed in 1.1.6.3, which have compared laparoscopic with open Nissen fundoplication, only three other randomised trials examining laparoscopic antireflux surgery have been reported. Two of these studies compared different techniques for dividing the short gastric vessels during laparoscopic
Nisen fundoplication. Laycock et 4111 randomised 20 patients to have these vessels divided between metal clips or by the Ultracision Ultrasonic Shears (Ultracision,
Smithfield, Rhode Island, USA). A time saving was demonstrated for the Ultrasonic
Shears. Swanstrom and Pennin gs20Z randomised 31 patients in a similar study which
also demonstrated a time saving of 10 minutes for the Ultrasonic Shear technique.
Neither of these studies enrolled a large group of patients, or attempted to assess outcome
differences between different antireflux procedures. 24
The third study was recently reported by Laws ¿109. This trial randomised 39 patients "¡ to undergo either a laparoscopic Nissen fundoplication or a laparoscopic Toupet fundoplication. No significant short term outcomes differences were demonstrated between the two procedures. The study, could be cliticised for not enrolling an adequate number of patients to have a reasonable chance of demonstrating any significant differences. At present there is a paucity of data from randomised trials available to enable surgeons to determine the most appropriate technique for the performance of antireflux surgery.
1.1.9 CONTROVERSIES AND DISAGREEMENTS
The advent of laparoscopic antireflux surgery has focused debate between protagonists of various antireflux surgical techniques. Whilst the Nissen fundoplication remains the technique advocated by the majority of surgeons, a number of variants of posterior and anterior partial fundoplication as well as other techniques have been advocated to reduce the likelihood of postoperative dysphagia and gas bloat symptomss,29,105,129,14I,156,166,212. Some surgeons advocate selective use of partial fundoplication variants for patient with oesophageal dysmotility62'80'101'115.
Others advocate routine application of a particular partial fundoplication
variantZ9,l05,2I2.
Table 1.1.7 Controversies for laparoscopic antireflux surgery 'When . should one apply a partial fundoplication technique?
. Which partial fundoplication technique is appropriate?
. Should the short gastric vessels be divided during laparoscopic Nissen
fundoplication?
. Should the hiatus be repaired selectively or routinely?
. Should the choice of procedure be tailored to the preoperative findings?
. How common is the short oesophagus and how should it be dealt with? 25
1.1.9.1 Laparoscopic partial fundoplication
The results of the 3 trials which compared posterior partial t'undoplication techniques with the Nissen technique performed by open surgical techniques have been discussed previously (section I.I.2.I). These studies demonstrated no significant outcome differences. There is no reason to believe that the results of these studies should not be considered when one is assessing the relative merits of different laparoscopic techniques.
No prospective randomised comparisons have been reported which compare the laparoscopic variants of the posterior partial fundoplication procedures with the Nissen procedure, and a randomised trial assessing an anteúor partial fundoplication technique performed either laparoscopically or by an open method has never been performed.
Early case series of patients who have undergone either a posterior or an anterior partial fundoplication, describe good or excellent early ¡s5u1¡59'29 '152'159'166'212- Reported post-operative dysphagia rates are low, 29,14I,152,166, although Watson et al reported
that2of their first 26 patients to undergo an anterior 1200 fundoplication experienced food bolus obstruction. It is necessary to be cautious when comparing different case series, which may be assessed post-operatively by different personnel who have almost certainly used different methods of symptom assessment. Before the routine use of laparoscopic partial fundoplications can be advocated, benefits should be demonstrated by randomised trials which confirm that the incidence of post-operative dysphagia or other
adverse outcomes is reduced and that long term efficacy is comparable.
1.1.9.2 Laparoscopic division of the short gastric vessels
Following the advent of laparoscopic Nissen fundoplication, argument has intensified
about whether the short gastric vessels should be divided to achieve full fundal
mobilisation, or whether an anterior wall complete fundoplication achieved without short
gastric vessel division can achieve an equally low rate of postoperative dysphagia and
other adverse outcome s4,38,J9 - Because good results have been published following
series of both open and laparoscopic Nissen fundoplication performed with and without
division of the short gastric vessels4,43,77,90,186, it has been difficult to resolve from the 26 evidence of uncontrolled case series whether division of these vessels is necessary. This problem is compounded by the lack of standardisation of assessment criteria tbr dysphagia, which renders comparison between different published reports difficult.
The original studies which popularised division of the short gastric vessels described non- randomised comparisons of patients undergoing open surgery43,49'50. 1s study by
DeMeester et aI43 made 2 concurrent changes to the surgical technique used (division of the short gastric vessels, and shorlening of the wrap), making assessment of the impact of short gastric vessel division difficult. Only a small number of patients in both
DeMeester's43 study and the study performed by Donahue50 underwent post-operative oesophageal manometry. DeMeester43 actually claimed that shortening the fundoplication length resulted in a reduction in the incidence of persistent dysphagia, and that division of the short gastric vessels resulæd in improved relaxation of the lower oesophageal sphincter. The study of manometdc versus clinical outcome described in section l-2.6.2 suggests that this outcome may be only weakly associated with persistent dysphagia.
Several surgeons have compared their laparoscopic experience with a Nissen fundoplication performed without dividing the short gastric vessels, with their subsequent experience with dividing short gastric u"rrs1.38'79. In these studies, laparoscopic
Nissen fundoplication was commenced without dividing these vessels. With subsequent improving laparoscopic experience, these surgeons began to divide the vessels. They then compared their experience with the two techniques. Analysis of the outcomes suggested that dysphagia was more common in patients in whom the short gastric vessels were left
¡1¿ç¡38,79. From this it was concluded that the short gastric vessels should be divided to minimise the incidence of post-operative dysphagia. However, this analysis fails to take
the potential for a learning curve bias into ¿sç6un¡213. An alternative hypothesis is that
the dysphagiaratefell as surgeons gained experience in laparoscopic antireflux surgery,
i.e. as the leaming curve was overcome, rather than because the short gastric vessels were
routinely divided. 21
Until recently no prospective randomised comparisons had been made to assess this issue trial from Luostarinen u1I25,126 which examined the issue of The small randomised "¡ short gastric vessel division during open surgery in25 patients, demonstrated no significant outcome differences. Hence it is possible that the most important technical step necessary for the avoidance of post-operativo dysphagia following Nissen fundoplication is to ensure a short loose wrap, irrespective of whether the short gastric vessels are divided or not.
1.1.9.3 Selective versus routine hiatal repair
Whilst routine hiatal repair was not a major issue during open fundoplication, this manoeuvre has become more important during the laparoscopic approach. The high incidence of up to l\Vo of early para-oesophageal hiatal herniation reported in patients who did not undergo laparoscopic hiatal rcpairz20 versus the lower incidence of 2 to 3Vo reported in patients undergoing repair70 suggest that this step should be routine. The overall incidence of this problem in the acute postoperative period, even in patients who underwent hiatal repair, may still be greater than that seen following open surgery. On the downside the laparoscopic surgeon may also have difficulty judging the degree of hiatal narrowing performed laparoscopically, with over-tightening of the hiatus, which can result in severe early dysph agia, a real possibility. No controlled trials have examined this necessity of this manoeuvrc. However, in the light of reported experience, it would seem sensible to perform some form of hiatal repair during laparoscopic anti- reflux surgery.
L.1.9.4 The tailored approach to laparoscopic fundoplication
Recently a number of surgeons have advocated 'tailoring' the specific choice of antireflux
procedure performed according to various pre-operatively assessed parameters62,101,115-
Kauer and Littlel15 hane suggested that the laparoscopic approach should only "1¿101 be used for patients with simple oesophagitis, and not for patients with the complications
of peptic stricture formation or Barrett's oesophagus. Both groups of authors recommend 28 that patients with complicated disease and those with potential oesophageal shortening should undergo an open transthoracic procedule. However, these criteria for a'tailored' approach are not supported by the majority of surgeons performing laparoscopic antireflux surgery, who have successfully applied laparoscopic procedures in patients with all grades of oesopha gitis3,70,2I5.
An alternative 'tailored' approach advocates modifying the type of fundoplication according to preoperative manometdc outcomes62,80. Advocates of this approach would perform a partial rather than a Nissen fundoplication in patients with preoperative motility disturbances. However, there is no evidence from controlled trials to support this proposal. Other studies suggest that a loose Nissen fundoplication may produce equally good outcomes in patients with poor motility8.
1.1.9.5 The short oesoPhagus
The incidence of oesophageal shortening found at antireflux surgery is controversial.
Most surgeons report a very low incidence of shortened oesophagus during laparoscopic anti-reflux surgery, with the incidence of conversion to an open procedure because of this problem being less than I%90. Although this suggests that the problem is infrequent, some surgeons argue that the ability of surgeons to determine adequate oesophageal length at the time of laparoscopic antireflux surgery is limited, and that the problem is in fact more common than is usually appreciated101,200. Whether this is correct is uncertain. Because of this problem Kauer u1101 advocate an open transthoracic "1 approach to perform a Collis gastroplasty, in patients with a hiatal hernia length of more
than 5 cm. Swanstrom .1u1200 recently described a method for the laparoscopic
construction a Collis gastroplasty. This offers a potential solution for laparoscopic
surgeons who encounter patients with a shortened oesophagus- 29
1.1.10 FUTURE DIRECTIONS
The long term viability of laparoscopic anti-reflux surgery depends on the ability of surgeons to demonstrate reduced surgical morbidity compared to the equivalent open surgical technique, as well as excellent long term clinical outcomes. Long term outcomes will not be available for some years. Research is needed to confirm the potential advantages of the laparoscopic approach, and to determine the optimal surgical technique(s) for patients presenting to sufgeons for antireflux Surgery.
1.1.11 RESEARCH OVERVIEW AND AIMS
The studies reported in this thesis were designed to answer the following questions;
a Does the laparoscopic approach offer important advantages over open surgical
techniques for patients undergoing sufgery for gastro-oesophageal reflux
disease?
a Should the short gastric vessels be divided during the laparoscopic Nissen
procedure?
a Do partial fundoplications effectively prevent reflux?
Does a partial fundoplication have advantages over the Nissen technique?
a Which surgical procedure should be performed in patients with reflux? 30 31
7.2 OUTCOME OF PROSPECTIVE STUDIES ASSESS'NG OUTCOMES FOLLOWING LAPAROSCOHC ANTIREFLUX SURGERY
I.z.L AIMS
To assess prospectively the initial experience with laparoscopic antireflux surgery at the
Royal Adelaide Hospital.
To identify adverse outcomes specific to the laparoscopic approach
To develop strategies which may improve the overall outcome following antireflux surgery, which might then be suitable for testing in fulther clinical and laboratory studies .
I.2.2 GENERAL METHODOLOGY
1.2.2.1 Preoperative assessment
All patients undergoing laparoscopic antireflux surgery had gastro-oesophageal reflux disease documented by either assessment with flexible endoscopy and/or 24hout ambulatory pH monitoring. Endoscopic evidence of ulcerative oesophagitis, or pH monitoring evidence of abnormal acid exposure (pH < 4 for 77o or more of the examination, or between 4 and 77o with good symptom comelation) was required to confirm this. In addition oesophageal manometry was performed routinely to assess
oesophageal peristalsis and lower oesophageal sphincter function, and to exclude gross
motility disorders (e.g. achalasia). Whilst preoperative barium contrast examinations
were not performed routinely, they were arranged when necessary to clarify anatomy e.g.
large hiatal hernia. 32
All patients underwent treatment with either all2rcceptor antagonist or a proton pump inhibitor for at least three months before deciding for surgery, and were generally ret'erred
for surgery either because of unsatisfactory symptom control on medical therapy, or
because of an expressed wish to avoid the taking of tablets lifelong.
1.2.2.2 Surgical technique for laparoscopic Nissen fundoplication
The operative technique evolved during the initial 150 cases, but with increasing
experience and the later availability of improved equipment it became relatively standard
later in the experience analysed. The patient is positioned on the operating table in the
lithotomy position with the table tilted 20 to 30o head up, and the surgeon sits or stands
between the patient's legs, with an assistant on the patient's left and the scrub nurse on the
patient's right. A Veress needle is inserted in the mid clavicular line immediately below
the left costal margin and the abdomen is insufflated with CO2 gas to a maximum
pressure of 12 mmHg. Five ports are placed as shown in figure 1.2.I. A L2 mm port (A)
is introduced just to the left of the midline, mid-way from the xiphisternum to the
umbilicus, for the laparoscope. Most procedures are completed using the 0o laparoscope,
although an angled laparoscope (3Qo or 45o) is used if adequate vision is not obtained.
The 0o laparoscope alone was used in more than907o of the procedures. High placement
of the camera port (A) midway between the umbilicus and the xiphoid facilitates the view
obtained through the 0o scope. 33
Figure 1.2.L Port placement for laparoscopic antireflux surgery
a o D o E B O o A C o
Additional ports are placed under vision. 5 mm ports are placed in the midclavicular line
just below the right costal margin (E) for insertion of a grasping forceps for dissection,
and in the epigastrium just below the xiphistemum (D) for a grasping forceps used as a
liver retractor. Since 1995 this has been replaced by a Nathanson Liver Retractor (Cook
Australia, Queensland) which is introduced through a 5 mm wound and held in place by a
fixed retraction system. A 5 mm port is placed in the left flank in the left anterior axillary
line (C) for the passage of an atraumatic grasper for the cardia. The fifth port is a 12 mm
port, placed in the left midclavicular line immediately below the costal margin (B). This
is the principal operating port for dissecting and suturing and it is also used for the
laparoscope if any difficulty is encountered in seeing the left side of the oesophagus. The
surgeon operates through ports B and E (Figure 1.2.1)- The assistant, standing on the
patient's left, controls the laparoscope and grasps the cardia of the stomach (via C) to
facilitate oesophageal dissection. 34
Dissection of the oesophagus is often achieved without using any diathermy current. It is commenced by opening the transparent windows of peritoneum in the lesser omentum above and below the hepatic branch of the anterior vagus nerve which can nearly always be clearly seen coursing to the right. This manoeuvre is useful as it helps orientate the surgeon and it reveals from right to left the caudate lobe, the right pillar of the diaphragmatic hiatus and the oesophageal hiatus with contained oesophagus. The hepatic branch of the vagus nerve is routinely preserved. The peritoneum to the left of the right pillar is opened by bluntly dissecting in a "groove" between the oesophagus and the right pillar. This is defined by passing a grasping instrument from right to left across the right pillar. The cardia is retracted in the direction of the left iliac fossa by the assistant, to facilitate this. The space between the right pillar of the oesophageal hiatus and the oesophagus is dissected vertically using blunt dissection, scissors or occasionally a hook diathermy dissector. If dissection is within the conect plane, then cautery is usually not necessary. The aim is to clean the right pillar of tissue and this concurrently begins the mobilisation of the oesophagus. Similarly, with the cardia pulled towards the right iliac fossa, the peritoneum over the left pillar ofthe hiatus is visualised and dissected from the edge of the left pillar. Exposure of the left pillar posteriorly as far as possible greatly facilitates the subsequent passage of an instrument behind the oesophagus.
Dissection is next undertaken in the trough between the oesophagus and the left pillar of the hiatus and then in the trough between the oesophagus and the right pillar of the hiatus.
The posterior vagus nerve is usually seen during this latter dissection and it usually lies with the right posterior aspect of the oesophagus. It is usually dissected away from the oesophagus and not included in the wrap. Dissection is continued around the distal oesophagus posteriorly, eventually opening a "window" behind the oesophagus.
During the dissection of the oesophageal hiatus and the oesophagus, because of the angle which the dissecting instruments tend to take, there is a tendency to dissect proximally
into the mediastinum. The left pleura is often seen if this occurs, where it is easily
perforated, leading to a pneumothorax (see section 1.2.4.I). Further oesophageal 35 dissection is facilitated by the passage of a nylon tape around the oesophagus to lift it forward. A tape is passed through poft B, passed behind the oesophagus and brought back through the same port. The port is then removed, the tape grasped by an artery forceps extracorporeally and the port is then replaced alongside the tape. A grasper can be used to pick up the two limbs of the tape in front of the oesophagus - and traction is exerted in whichever direction facilitates the procedure. This manoeuvre is particularly useful for facilitating suturing of the hiatal pillars behind the oesophagus. A large
"window" is dissected out behind the oesophagus to allow easy passage of the stomach.
This window is developed between the posterior vagus nerve and the oesophagus in front.
Hiatal repair is performed posteriorly using interrupted 2/0 Prolene or Novafil sutures before carrying out the fundoplication. Hiatal repair was performed infrequently during the initial 2 years of experience at the Royal Adelaide Hospital. However, when subsequent experience demonstrated problems with post-operative paraoesophageal herniation (see section 1.2.4.2), practice was modified to include routine posterior hiatal repalr.
A grasper is next passed behind the oesophagus via port E. The anterior wall of the fundus approximately 5 cm from the cardia and midway between the cardia and greater curve is picked up by a grasper held by the first assistant and passed to the jaws of the grasper behind the oesophagus and is then pulled around behind the oesophagus. The wrapped stomach is then approximated to the anterior wall of the stomach in front of the oesophagus, forming a loose 360o wrap. This can be achieved with or without division of the short gastric vessels. If the wrap seems tight then the stomach can be manipulated until a looser piece of stomach is selected. A 52 French gauge bougie or larger is placed in the oesophagus before suturing. Non-absorbable sutures are used for the
fundoplication. Three interrupted sutures are placed, using 2i0 Prolene or Novafil, and yno¡229, knots are tied extracorporeally using a modified Weston achieving an
approximately 2 cm long fundoplication. The bougie is then removed. 36
1.2.2.3 Post-operative management
Early in the series a nasogastric tube was placed perioperatively and then removed on the first postoperative day. Subsequently this was omitted for most patients. Oral fluid intake was commenced on the evening of surgery or the first post-operative day, and soft solid foods were allowed24 hours later. A barium swallow examination was performed routinely on the second post-operative day to check for early para-oesophageal herniation of the stomach, or other problems amenable to eally laparoscopic correction. Patients usually left hospital between two and four days after surgery-
1.2.2.4 Data management and follow-up protocol
Details of each laparoscopic or attempted laparoscopic antireflux procedure performed by surgeons from the Royal Adelaide Hospital were collected prospectively and stored in a computerised data base (FileMaker Pro version 3.0, Claris Corporation, Santa Clara,
California, USA). Information about the pre-operative assessment and management, surgical procedure and immediate post-operative outcome for each patient was sourced from the operating surgeon, and from reports of the outcome of pre-operative investigations.
Postoperative clinical follow-up was obtained by the application of a standardised questionnaire administered by a non-clinical scientific officer 3 and 12 months following surgery, and then annually thereafter. Symptomatic dysphagia was assessed using two visual analogue scales from 0 to 10. These were; severity of dysphagia when swallowing liquids (no dysphagia = 0, total dysphagia = 10), and severity of dysphagia when swallowing solid foods (no dysphagia = 0, total dysphagia = 10). The presence or absence of symptoms suggesting gas bloat was also determined by asking patients whether they ever experienced discomfort due to upper abdominal distension or a sensation of bloating of the upper abdomen, including occasional or mild bloat symptoms.
The ability to burp, the ability to relieve abdominal distension, and the patients willingness to undergo the same procedure again under similar preoperative circumstances, were also determined. Overall satisfaction with the surgical outcome was 37 determined using a 0 to 10 visual analogue scale (completely satisfied = 10, totally unsatisfied = 0). Details of adverse outcomes such as hospital readmission, complications, or surgical revision were recorded.
Oesophageal manometry and Z4hour pH monitoring has been performed when possible on all patients between 3 and 6 months following surgery. Barium meal examination was also performed after 3 months from 1992 to 1995. Endoscopy was only been performed when clinically indicated until 1995. Later it was incotporated into the follow-up protocol for the study described in section 3.3.
L.2.3 INITIAL OUTCOME ASSESSMENT OF PROSPECTWE CASE SERIES
1.2.3.1 Initial outcomes'the first 70 patients
(Published previously as; Jamieson GG, Watson DI, Britten-Jones R, Mitchell PC &
Anvari M. Laparoscopic Nissen Fundoplication. Annals of Surgery 1994;220l. I37-
14s90)
1.2.3.1.1 Patíents and methods
Between September 1991 and February 1993,70 patients (39 male and 31 female) who had undergone a laparoscopic approach for a Nissen fundoplication were prospectively evaluated. Age ranged from 19 to 91 years (median 45) and weight ranged from 52 to
l23kg (median 75 kg). Twenty four patients had undergone previous open abdominal surgery (8 upper, 16 lower). One patient had undergone transthoracic repair of a hiatal hernia as an infant 40 years earlier. In the same time frame 15 patients had an open fundoplication either because the procedure was thought to be contraindicated laparoscopically, or because of patient preference'
All patients had symptomatic gastro-esophageal reflux disease documented by endoscopy
(69 patients) and/or oesophageal manometry and 24hov ambulatory pH monitoring (57 38 patients). Preoperative barium meal examination was performed in 51 patients. A hiatus
hernia was present in 36 patients, but was greater than 5 cm in length in only six- Two
patients presented with large para-oesophageal hiatal herniae associated with gastric
volvulus, one requiring preliminary nasogastric tube decompression. The duration of
reflux symptoms ranged up to 20 years (median 5 yrs). All patients had undergone
treatment with H2 receptor antagonists of omepfazole for at least thlee months and were
referred for surgery, either because of unsatisfactoty symptom control or because of an
expressed wish to avoid the taking of tablets lifelong.
1.2.3.1.2 Results
Of 70 laparoscopic Nissen fundoplications attempted, 60 were completed
laparoscopically. Two of the ten not completed were earþ in the series (cases 1 and 5).
The first was abandoned because of an inability to pull the stomach around behind the
oesophagus probably due to inadequate dissection of the posterior "window" so that there
was not enough room behind the oesophagus for passage of the stomach. The second was
a patient with a large sliding hiatus hernia. Failure to realise at this early stage the
necessity of traction on the cardia, meant that the oesophagus and proximal stomach kept
retracting into the chest. The resulting intrathoracic dissection via the abdomen was
difficult and instruments could not be passed behind the oesophagus to open a posterior
"window". The third patient was the 20th in the series. This patient was morbidly obese
which meant the instruments proved too short and the procedure was abandoned at an
early stage. The thirtieth patient had a scleroderma oesophagus shown manometrically
preoperatively, and although the procedure was completed, the surgeon felt that the wrap
was too tight. The patient was therefore opened and a loose wrap was constructed. This
was one of only two patients in this early series in whom short gastric arteries were
divided.
The 31st patient also had an adynamic oesophagus and once again the surgeon was not
happy that a loose wrap could be constructed, so after 2ll2 hours operating time the
procedure was converted to an open operation. The 36th patient proved to have a lot of 39 peri-oesophagitis and a large peri-hiatal fat pad. The decision to open this patient was made early, and the operation proved difficult as an open procedure. The 45th patient was
similar to the 36th. The 53rd patient had a large hiatus hemia with an intrathoracic
stomach, and shortened oesophagus. The oesophagus could not be mobilised
laparoscopically. The 62nd patient had extensive peri-oesophagitis and the 64th patient
had extensive adhesions from previous upper abdominal surgery. Both patients were
opened after short trial dissections. All of these patients were relieved of reflux
symptoms.
Operating time ranged from 45 to 290 minutes (median 120 minutes). The time taken to
complete the procedure fell as experience improved, with the median time taken for the
latter 35 procedures being 95 minutes. Intracorpoleal knot tying was used in 9 of the first
12 procedures, whereas extracorporeal knot tying was used exclusively for the last 50
operations completed laparoscopically. This lead to a large reduction in operating time.
After problems early in the series in two patients with significant hiatus hernia, early
patient selection precluded large herniae. However, with developing experience, ten later
patients with a hiatus hernia >5 cms in size underwent laparoscopic fundoplication with
hiatal repair.
Postoperative hospital stay for completed laparoscopic fundoplications ranged from 2 to 8
days (median 4 days). All patients were able to return to normal activity within 2 weeks
of laparoscopic surgery. Postoperative complications occurred in 12 patients (Table
l.Z.3.I). Three patients (in all of whom the operations were completed open) were
readmitted within three weeks of their discharge with chest pain due to pulmonary
embolism. All did well following anticoagulation. Another patient was readmitted with
pneumonia two weeks after surgery. One patient suffered a left pneumothorax. One
patient experienced respiratory distress due to aspiration following extubation at the
conclusion of his operation. Following two days of assisted ventilation, he made a good
recovery and was discharged on the 8th post-operative day. One patient developed
cervical and mediastinal surgical emphysema due to CO2 gas tracking up into the
mediastinum and neck. The insufflation pressute had been increased from 15 to 18 40 mmHg during surgery. Severe chest pain ensued post-operatively, with narcotic analgesia
being required for 2 days and discharge was delayed to 7 days after surgery. The
maximum insufflation pfessure used was reduced to 12 mmHg following this
complication. One patient was readmitted with vomiting three days after discharge. A
barium meal showed an intra-abdominal gastric volvulus. The patient settled down with
nasogastric aspiration and without the need for any other treatment.
Table 1.2.3.1 Complications
Pulmonary embolus 3
Pneumonia 1 Pneumothorax I
Aspiration 1
Mediastinal and cervical surgical emphysema - 1
Gastric dilatation/obstruction 2 Reoperation 4
- rolling hiatus hernia 1
- recurrent reflux 1
- gastric obstruction/misplaced fundoplication 1
- dysphagia 1
- cholecYstectomY 2
Six patients underwent open surgery within 6 months of their laparoscopic fundoplication.
Two were in the first six cases of the series. The first patient (number 3) represented with
dysphagia due to an incarcerated rolling hiatus hernia requiring open reduction and hiatal
narrowing. His fundoplication, was intact. He remains free of reflux symptoms at 3 years
follow-up. The second patient (number 6) experienced recurrent reflux one week after
laparoscopic fundoplication. Subsequent barium meal showed that the wrap had
disrupted. When this wrap was originally formed, the large oesophageal bougie had to be
removed as the stomach kept retracting behind the oesophagus with the bougie in-situ, 4I preventing suturing. This suggested that the wrap was constructed under tension.
Unfortunately, open revisional surgery was unsuccessful, as he developed further
symptoms 4 months later, with erosive oesophagitis evident at endoscopy, despite an
apparently intact fundoplication.
The third patient was the 23rdinthe series. He was 91 years old and originally presented
with a large hiatus hemia and an organo-axial gastric volvulus. After initial
decompression with a nasogastric tube, he underwent laparoscopic reduction of the large
hiatus hernia, Nissen fundoplication and hiatal repair. A tube gastrostomy through a4 cm
muscle splitting incision was performed also as an anterior Sastropexy. His initial post-
operative progress was good and he was discharged at day 6. However, 2 days later he
developed acute abdominal pain. Gastrograffin studies revealed a large obstructed
stomach. At laparotomy the proximal stomach was dilated and the distal stomach
collapsed. It was apparent that the fundoplication had been formed using the midbody of
stomach and the gastrostomy tube had been placed distally in the posterior wall of the
antrum. This combination rotated and compressed the midbody of stomach resulting in
an iatrogenic obstruction. His fundoplication was revised and he made a subsequent
satisfactory recovery. At open operation the hiatal repair was found to be sound and the
stomach was entirely intra-abdominal.
The 35th patient had poor oesophageal motility demonstrated by preoperative
oesophageal manometry, in association with marked reflux. Consequently, at
laparoscopic surgery a very loose fundoplication was constructed. Postoperative barium
meal examination demonstrated a technically excellent fundoplication, and the reflux
symptoms resolved. However, the patient developed marked dysphagia for solids and
liquids, necessitating open revisional surgery 7 weeks post-operatively, converting the
3600 fundoplication to a partial wrap with resolution of the dysphagia. There were also
two patients who continued to experience some symptoms after their fundoplication-
Both were subsequently shown to have gall stones and both underwent laparoscopic
cholecystectomy with resolution of their symptoms- 42
Early troublesome dysphagia for solids was experienced by 12 patients. It resolved in I patients by 3 months. Two patients had mild dysphagia for solids at follow-up of 3 and 4 months, and two patients had dysphagiaat 8 and 10 months follow-up which was troublesome enough for them to state they would not have had the operation had they known the outcome.
The outcome of the initial experience is summarised in Table 1.2.3.2. Sixty-nine out of
70 patients operated on for reflux were free of reflux symptoms at early follow-up, with
56 patients having undergone laparoscopic surgery alone. One patient continued to have symptoms of gastro-esophageal reflux. Clinical follow-up for this study ranged from three to eighteen months (mean 8). Postoperative endoscopy was performed in 11 patients, revealing an intact fundoplication and absence of oesophagitis in all instances.
Oesophageal manometry was performed in 25 patients, revealing an adequate high pressure zone in all patients with a rise in mean lower oesophageal sphincter pressure in these patients from 11.6 mmHg preoperatively to 21.3 mmHg post-operatively- Of the
32 barium meal examinations performed, a satisfactory fundoplication was present in all asymptomatic patients.
Table 1.2.3.2 Outcome
Complete symptomatic relief of reflux 69
- laparoscoPy onlY 54
- immediate laparotomY 10
- delayed laParotomY 5
Ongoing symptoms of reflux 1
- delayed laparotomy 1
Dy sphagia ( trouble some ) 2 43
One unexpected finding revealed itself on post-operative barium studies and endoscopy.
In eight patients out of 32 assessed, the wrap appeared to have been fashioned from the body of the stomach rather than the fundus. This was seen in its grossest form in the patient with the large para-esophageal hemia, and in the patient who represented with a gastric volvulus, who have been described above. In six other patients, however, all of whom achieved excellent clinical results, the stomach was converted into a bilobed shape.
1.2.3.1.3 Díscussion
The operation performed in the initial series is essentially the same as originally described by Nissenl58, e*"ept for a shortening of the fundoplication to 2 cm and the use of a 52
French gauge bougie in the oesophagus when constructing the fundoplication. It is identical to the procedure which was performed open, at the Royal Adelaide Hospital since 1990. Surgical technique was changed to an anterior wall fundoplication at this time, when surgeons became aware of the excellent results achieved by Johansson et al.
Their study had greater than 5 year follow up and the patients were studied in a careful and objective rnanne.93. The change to this technique was not related to the advent of laparoscopic fundoplication.
The fundoplication performed achieves the goals stated by DeMees¡s1198, i.e. a short, loose circumferential fundoplication, calibrated around alarge intra-oesophageal bougie, although short gastric vessels were not divided. Dallemagne39 described division of these vessels in some his initial patients, although this was not found to be necessary in any except two of the patients undergoing surgery at the Royal Adelaide Hospital.
Geagea64 also found it unnecessary to divide short gastric vessels when constructing a loose Nissen fundoplication laparoscopically'
This initial series demonstrated the feasibility of performing a Nissen Fundoplication laparoscopically, although follow-up was shoft. Technique underwent some minor modifications with experience, enabling considerable improvements in operating time and widening of patient selection. It was found that it is essential that a large "window" is 44 dissected behind the oesophagus to enable a tension free fundoplication to be constructed, and to facilitate posterior hiatal repair. Also, unless the stomach can be readily brought around behind the oesophagus without tension, so that it does not retract when a large bougie is placed through the distal oesophagus, then the fundoplication will probably be too tight and a different point on the anterior wall of the stomach should be sought to ensure a loose wrap.
Whilst follow-up did not exceeded 18 months in this initial series, the clinical results were promising. Nevertheless, six of the seventy patients required a second operation - an incidence of reoperation not much different from the much criticised reoperation rate of the Angelchik prostherirl3l. Also, three of the patients who were eventually opened had greatly prolonged operations and subsequently represented with pulmonary embolism.
This outcome should not be forgotten, as the head up position, increased intra-abdominal pressure, and the use of stirrups creates a scenario where venous stasis in the calf veins is likely.
Also, it was not easy to judge exactly which part of the proximal stomach was being used for the fundoplication. This resulted in a257o incidence of "bilobed stomach" in the first
32 patients who underwent contrast radiology following the procedure. This was symptomatic on at least one occasion (needing reoperation). This occurs if the body of
the stomach is used for the wrap, rather than the fundus. Taking care to avoid this by
"walking" up the stomach with laparoscopic grasping instruments to ascertain the location
of the apex of the fundus before selecting the piece of fundus for the fundoplication
reduces the likelihood of the problem-
Laparoscopic Nissen fundoplication was shown to be an effective antireflux operation in
the short term in this initial study. In uncomplicated and straight forward cases it had
similar advantages to laparoscopic cholecystectomy, achieving a reduction in hospital stay
and less postoperative pain compared to historical experience with the traditional
procedure. 45
L.2.3.2 Experience with 230laparoscopic antireflux operations and resulting changing surgical strategies
(Published previously as; Watson DI, Jamieson GG, Devitt PG, Mathew G, Britten-
Jones R, Game PA & Williams RS. Changing strategies in the performance of laparoscopic Nissen fundoplication as a result of experience with 230 operations.
Surgic al Endos copy 1995; g' 961-966279¡
1.2.3.2.1 Patíents and methods
Between September 1991 and January 1995,230 patients underwent an attempted laparoscopic Nissen fundoplication (I42 male, 88 female). All patients presenting for primary antireflux surgery were offered this approach, regardless of any perceived difficulties due to preoperative pathology. Age ranged from 19 to 91 years (median 46) and weight from 37 to I23 kg (median 78). 85 patients (377") had undergone previous open abdominal surgery (15 upper, 55 lower, 15 upper and lower). Three patients had undergone earlier transthoracic repair of a hiatal helnia without fundoplication.
All patients were symptomatic and had gastro-oesophageal reflux proven by ulcerative oesophagitis on endoscopy (155 patients) or 24 hr ambulatory pH monitoring (198 patients), prior to undergoing surgery. Barium meal examination was performed pre- operatively in 151 patients and preoperative oesophageal manometry was performed routinely. A sliding hiatus hernia was demonstrated by barium meal or endoscopy in
108 patients (477o). A large para-oesophageal hernia containing the majority of the stomach was present preoperatively in seven, and was the primary indication for surgery in five. 46
1.2.3.2.2 Results
Of the 230laparoscopic Nissen fundoplications attempted,2OT (90%) were completed using the laparoscopic technique. All patients converted to open surgery were included in this analysis. The indications for conversion to open surgery are summarised in Table
1.2-3.3. Conversion became less common with increasing experience. Thirteen of the first 100 patients compared to only 5 of the last 100 required open surgery to complete the procedure. Early dissection difficulties due to large hiatal hernia and obesity were usually overcome with later experience. Liver hypertrophy limited exposure in a small group of patients, and was difficult to overcome as the Nathanson liver retractor was not available for these cases.
Table 1.2.3.3 Conversion to open fundoplication
Obesity 7 (3.07o)
Large hiatus hernia 5 (2.2Vo) Adhesions 4 (1.77o)
Hypertrophied left lobe of liver 3 (I.3Vo)
Difficulty ensuring loose fundoplication 2 (0.9Vo)
Perioesophagitis 2 (0.gEo)
Posterior hiatal repair was performed in 112 patients (497o) and was carried out routinely in the last 84 . Short gastric vessels were divided in 58 (25%) patients. Operating time ranged from 30 to 260 minutes (median 95), but with increasing experience the time required to complete the procedure dropped . The median operating time for the first 50 procedures was 120 minutes (range 45 to 260 minutes) compared to 80 minutes for the last 50 (range 30 to 210 minutes), despite the routine performance of posterior crural repair and division of short gastric vessels in 507o of the latter group. Post-operative hospital stay ranged from 1 to 19 days (median 3 days). Post-operative follow-up ranged up to 40 months (median 16). The median time for return to normal physical activities or employment was three weeks following surgery. 47
Post-operative complications occurred in 20 patients (Table L.2.3.4)- One patient died following the development of thrombosis of the supedor mesenteric and coeliac ur¡"ri.r146. This was the only death in the series. The four pulmonary emboli occurred early in the experience, in patients whose operations were prolonged due to conversion to open fundoplication. Pneumotho rax225 was associated with the passage of instruments behind the oesophagus, and into the thorax. This was not associated with any intra- operative cardiorespiratory compromise. Gastric perforation occurred twice, due to an inexperienced assistant using overly vigorous retraction of the cardia with a grasping instrument. In both instances it was recognised at the time and repaired laparoscopically with no further consequences. Reoperation was required in23 patients, 14 within the first three months after their fundoplication (Table 1.2.3-5), and 9 at alater date (Table
l-2.3.6)- The incidence of subsequent surgical revision declined with increased experience and modifications to technique.
Table 1.2.3.4 Post-operative complications (30 days) Pneumothorax 4 (t.7Eo) Pulmonary embolus 4 (1.18o)
Acute paraoesophageal herniation 3 (t.3Eo)
Gastric obstruction 3 (t.3Vo)
Respiratory infection 2 (0.97o)
Gastric perforation 2 (0.97o)
Haemorrhage 1 (0.47")
Mesenteric thrombosis (died) 1 (0.48o) 48
Table 1.2.3.5 Early reoperation (within 3 months)
Paraoesophageal Hernia 5 ('2:ZVo) Dysphagia 4 (1.77o)
Gastric obstruction 2 (0.9Vo)
Recurrent reflux 1 (0.47o)
Bleeding 1 (0.48")
Mesenteric Thrombosis 1 (0.48o)
Table 1.2.3.6 Late reoperation
Paraoesophageal hernia 5 (2.2Vo)
Dysphagia 3 (l.3Vo)
Recurrent reflux 1 (0.48o)
Para-oesophageal herniation was a major problem. This is discussed in detail in section
1.2.4.2. Of the 10 patients who underwent further surgery for it, posterior hiatal narrowing had been performed initially in only 2. Re-operation for paraoesophageal herniation was performed using an open technique in 7 patients, and laparoscopically in 3
Of the laparoscopic procedures, two were performed within 4 days of the original procedure and one was a delayed repair five months later.
Postfundoplication dysphagia was significant in 32 (líVo) out of 208 patients reviewed three months following surgery. In most patients this resolved spontaneously or following endoscopic dilatation (7 patients), but seven patients required reoperation for persistent severe dysphagia. Neither division of short gastric vessels nor the operator's
experience appeared to influence the likelihood of this problem. Spontaneous
improvement in the symptoms of dysphagia continued to occur for up to 12 months, with
only 9 out of 143 (6Vo) patients reviewed at this stage continuing to experience this
problem. Four (2 early, 2late) were found to have stenosis of the oesophageal hiatus due 49 to post-operative scaffing. This is discussed in section I.2.4.3. In two of the remaining three patients the fundoplication was revised to a partial wrap. One other patient with excessive dysphagia six days following surgery, underwent laparoscopic reassessment-
A tight fundoplication was discovered which was leadily revised laparoscopically by the removal and replacement of a single suture.
post-operative barium meal examinations were performed on 126 patients, demonstrating
the appearance of a satisfactory fundoplication in 120 (95Vo). Distortion of the gastric
body due to creation of the fundoplication using the mid to lower body rather than the
fundus of the stomach as discussed previously was evident on 13 early examinations.
Eleven of these patients were asymptomatic. In two patients however, gastric obstruction
resulted, requiring later revision. With awareness of the potential for this problem, and
the institution of steps to guarantee the correct use of the fundus, later barium meal studies
demonstrated the abolition of the problem. Oesophageal manometry was performed
three months following surgery in 90 patients, revealing an adequate high pressure zone
in all but one patient, with a rise in lower oesophageal sphincter pressure from a median 5
mm Hg (range 0 to 27) to 19.5 mm Hg (range 0 to 50) following surgefy.
The overall clinical outcome is summarised in Table 1.2.3-7 . Compleæ relief of reflux
symptoms was achievedin226patients (98Vo), although only 203 (887o of the overall
series) underwent a single operation. The majority of patients were very pleased with the
outcome of their surgery. When asked 3 months after surgery to rate on a visual
analogue scale their satisfaction with the outcome (0 = dissatisfied, 10 = satisfied),867o
rated the operation from 7 to 10. When patients were also asked the question; "Would
you have this operation again, knowing what you now know about it?" 887o of patients
answered yes. Interestingly 16 out of 23 patients who underwent surgical revision and 13
of the 20 patients who developed a complication also answered yes. 50
Tabte 1.2.3.7 Overall outcome
Relief of symptoms with single operation 203 (88.3Vo)
Second operation required 23 (IÙVo)
Recurrent reflux symptoms 3 (t.3Vo)
Operative mortalitY 1 (0.4Vo)
1.2.3.2.3 Discussion
This study examined the short to medium term outcome of laparoscopic fundoplication in a larger series of patients. The results confirmed that the majority of patients had complete relief of reflux symptoms. Nearly 907o of patients achieved a good outcome, defined by patient scored satisfaction and an expressed willingness to undergo laparoscopic fundoplication again in the same circumstances. These outcomes are similar to those achieved following open fundoplication43,93,186. Up to l\Vo ofpatients however, may have been disadvantaged by the laparoscopic approach due to the significant early reoperation rate. The majority of the reoperations were within 3 months of the original procedure and it is possible that these may not have been needed if these patients had undergone open surgery. Comparable experiences with open antireflux surgery do not report similar early problems with hiatal stenosis, para-oesophageal herniation and gastric obstruction. In contrast, the rate ofre-operation for postoperative bleeding and recurrent reflux was low. Three procedures wero successfully undertaken laparoscopically in the first post-operative week (2 for herniation, 1 for dysphagia).
These patients were not greatly disadvantaged and did not experience major delays in their recovery. However, the other patients required open surgery which was often technically difficult.
The 6Vo incidence of dysphagia 12 months following surgery was comparable with previously reported experience following open fundoplication43,93. Some surgeons have advocated the routine use of a partial fundoplication to reduce the incidence of this problem. However, this claim has never been objectively substantiated, with none of the 51 randomised trials published of posterior partial versus Nissen fundoplication,
demonstrating any reduction in the incidence of dysphagia3l'Il9'120'124'209 '
No patient underwent splenectomy in this series, and only one significant post-operative
bleeding complication occurred in the 58 patients undergoing short gastric vessel division
All comparable sized series of open Nissen fundoplication report an incidence of
splenectomy in at least a small number of patients43,204. The magnified view and the
more accurate dissection enabled by the laparoscopic technique, as well as the ability to
dissect the short gastric vessels close to the gastric wall, without the access difficulties
encountered through an upper midline incision, may explain the absence of splenectomy.
Because of problems encountered in this experience, strategies were developed to
minimise the likelihood of early and late complications in patients undergoing
laparoscopic fundoplication. Routine posterior hiatal repair was instituted to minimise
the risk of para-oesophageal herniation. The amount of diathermy used for dissection of
the oesophageal hiatus ,was also greatly reduced. In many instances it is possible to fully
dissect the distal oesophagus using a combination of scissors and blunt dissectors without
the use of diathermy at all.
The issue of division of short gastric vessels has been controversi¿43,186. The
procedure is simpler if these vessels are not divided. Because of the controversy and the
lack of earlier controlled trials, a trial of division versus no division of short gastric
vessels during laparoscopic Nissen fundoplication was initiated (section 3.2). 52
1.2.4 UNIQUE COMPLICATIONS OF LAPAROSCOPIC FUNDOPLICATION
1.2.4.1 Pneumothorax 'Watson (Published previously as; DI, Mitchell P, Game PA and Jamieson GG-
Pneumothorax during laparoscopic dissection of the oesophageal hiatus. A.N.Z. J Surg
1996;66:711-72225¡
The proximity of the left pleura to the distal thoracic oesophagus, places it at risk of penetration during laparoscopic mobilisation of the distal oesophagus. This can occur during any stage of the surgical dissection, but is particularly likely when dissecting or passing instruments behind the distal oesophagus. It was not described in any of the initial published series of laparoscopic fundoplication6'35'39'55'64'227 '228, although three published letters did mentio¡¡ i¡10,182,218. pu¡i¡g the initial experience of 190 laparoscopic Nissen fundoplications at the Royal Adelaide Hospital it was encountered five times, suggesting that it is a significant risk during laparoscopic hiatal dissection.
1.2.4.1.1 Case reporß
Cøse no. 7:
A39 year old man underwent a laparoscopic Nissen fundoplication. During dissection behind the oesophagus, the left pleura was visibly breached. The anaesthetist was informed, but noticed no change in the patient's haemodynamic state, oxygen saturation measured by pulse oximetry, or airway resistance to ventilation. The surgical procedure continued uneventfully. A chest radiograph obtained in the recovery room revealed a
507o pneumothorax. An intercostal drain was placed and removed on the first post- operative day. The patient made an uneventful recovery, being discharged on the third post-operative day.
Case No.2:
A 91 year old man with a very large rolling hiatus hernia and gastric volvulus underwent laparoscopic fundoplication, crural repair and anterior gastropexy. Insufflation distended 53 the large hiatal hernia sac, providing excellent exposule within the sac. As the oesophagus was being dissected, however, the left pleura was breached, with resultant collapse of the sac into the abdomen due to gas entering the left pleural space. The view via the laparoscope became difficult, although no cardio-pulmonary compromise resulted
The procedure was completed laparoscopically. Post-operative chest x-rays failed to demonstrate the pneumothorax, and no intercostal drain was used.
Case No.3:
A37 year old man underwent elective laparoscopic fundoplication. Three years earlier he had been involved in a severe motor vehicle accident, suffering multiple rib fractures, right haemothorax and left pneumothorax. Following this he experienced increasingly severe symptoms of gastro-oesophageal reflux, confirmed by 24 hour pH monitoring.
Shortly afær passing a grasper behind the oesophagus during laparoscopic oesophageal mobilisation, the anaesthetist noticed a fall in arterial oxygen saturation, measured by pulse oximetry. A pneumothorax was suspected, and the abdomen was temporarily deflated, whilst the oxygen saturation returned to normal. The abdomen was then reinflated and the procedure continued uneventfully, with the exception of one further episode requiring furlher temporary deflation. An urgent post-operative chest X-ray confirmed a left pneumothorax, and an intercostal drain was inserted for 48 hours. The patient made an uncomplicated recovery and was discharged on the third post-operative day.
Case No.4:
A 45 year old man underwent elective laparoscopic fundoplication and cholecystectomy.
During dissection behind the oesophagus, the left pleura was inadvertently opened widely, enabling the laparoscope to enter and visualise the left pleural cavity. There were no cardiorespiratory effects associated with this. The anaesthetist, however, noted increased resistance to ventilation, which was relieved by decreasing the inflation pressure from 14 to 10 mm Hg. The surgery was completed uneventfully, and the post-operative chest X- 54 ray revealed no pneumothorax. The patient required no intercostal drain, and made an
uneventful recovery.
Case No.5:
During dissection of the posterior oesophagus for laparoscopic fundoplication in a 66 yeat
old woman, the left pleura was visibly breached, with the left lung clearly seen through
the laparoscope. The anaesthetist was informed, but noticed no change in
cardiorespiratory status, or any alteration in resistance to ventilation. The procedure
progressed uneventfully. Post-operative chest X-ray revealed no pneumothorax, and again
the patient made an uneventful recovery.
1.2.4.1.2 Díscussíon
These five cases confirm the risk of pneumothorax due to damage to the left pleura during
laparoscopic oesophageal mobilisation. In the initial experience at the Royal Adelaide
Hospital this risk was calculated to be 3Vo. It is possible, however, that the risk is under
estimated. As carbon dioxide gas is rapidly resorbed from the peritoneal and pleural
cavities due to its high water solubility, and routine post-operative chest x-rays in the
recovery room have not been usual, small pneumothoraces may rapidly disappear and not
be detected. This is illustrated by the second, fourth and fifth cases, where despite clear
evidence of a breach in the left pleura, no gas in the pleural cavity could be demonstrated
on post-operative chest X-raY.
During other laparoscopic procedures, pneumothorax is rarc56,23I. Isolated reports
during laparoscopic cholecystectomy have been published20,40,63,23I. It is thought that
most of these cases are related to developmental or anatomic diaphragmatic defects149,
although previous surgery at the oesophageal hiatus56 and diaphragmatic traumiT4
during laparoscopic cholecystectomy have also been implicated-
Although four of the patients reported suffered no cardiorespiratory compromise, one did,
with resultant difficulty completing the procedute laparoscopically. Reid et a1É82 55 reported in a letter, a case associated with significant cardiorespiratory compromise
during hiatal dissection, requiring cessation of the operation, deflation of the abdomen,
chest x-ray and placement of a chest drain before completing the procedure. However,
this action is probably unnecessary in the majority of cases with no more than close
observation by the anaesthetist usually required, whilst the surgical procedure continues-
Intercostal drainage is probably rarely required. Awareness of the risk of damage to the
pleura during laparoscopic oesophageal mobilisation is essential, so that steps can be
taken to avoid the problem, and thereby minimise its risk. 56
1.2.4.2 Para-oesophageal hiatus hernia 'Watson (Published previously as; DI, Jamieson GG, Devitt PG, Mitchelt PC & Game
PA. Para-oesophageal hiatus hernia: an important complication of laparoscopic Nissen fundoplication. British Journal of Surgery 1995; 82:52I-523220¡
Post-operative para-oesophageal herniation has been a significant problem, especially during the early experience with laparoscopic Nissen fundoplication at the Royal
Adelaide Hospital described previously. It has also been reported following laparoscopic antireflux surgery performed in other centreslT, and it is of particular importance as it often requires further coffective surgery.
1.2.4.2.1 Pøtíents and methods
Patients developing post-operative para-oesophageal herniation following 253 laparoscopic Nissen fundoplications for gastro-oesophageal reflux performed by surgeons from the Royal Adelaide Hospital were reviewed. Some were identified by the routine follow-up protocol, during which post-operative barium contrast examinations were performed three months after surgery in 81 patients and during the first three post- operative days in 65 patients. Other patients presented with chronic symptoms which were investigated by chest X-Ray, Barium meal or endoscopy, and others presented with acute symptoms, all with the resulting diagnosis of para-oesophageal herniation.
1.2.4.2.2 Results
Seventeen patients with post-operative para-oesophageal herniation (9 male, 8 female) were identified from the overall group of 253 patients undergoing laparoscopic Nissen fundoplication between September 1991 and October 7994. The herniae occurred randomly throughout this experience and were not associated with any particular operator
Age ranged fuom23 to 73 years (median 46) and weight from 50 to 95 kg (median 79).
These demographic factors were identical to the overall patient group. Nine of the \7 patients had a pre-operative sliding hiatus hernia demonstrated by endoscopy or barium meal, of which eight were less than 5 cm in length. At the time of original surgery only 3 57 patients (l}Eo) underwent hiatal repair, two anteriorly, and one posteriorly , all with a
single 2/0 Prolene suture. This compares with 120 patients (477o) undergoing hiatal
repair out of 253 inthe overall experience over the same time period (P<0.05, Fisher's
exact test).
Ten patients developed symptoms attributable to a para-oesophageal hernia from one day
to 21 months following surgery (median one month). Three developed acute pain and
severe dysphagia within 24 hours of surgery, two following periods of prolonged early
vomiting. Another developed symptoms after performing strenuous lifting within two
weeks of the initial surgery. Seven patients were asymptomatic. The hernia was initially
diagnosed by Barium meal (13 patients), endoscopy (1), or chest X-Ray (3). Two
patients had undergone previous barium meal studies at one and three months post-
operatively, with no hernia demonstrable. Hernia size was less than 2 cm in maximum
dimension in three patients, 2 to 5 cm in five patients and greater than 5 cm in nine
patients. All herniae contained stomach, although one also contained colon and small
bowel.
Ten patients underwent further surgery between 36 hours and two years after initial
laparoscopic fundoplication (median three months). Of these, the fundoplication was
revised in six, and posterior hiatal repair was performed in all ten. Dense adhesions
between the fundoplication, stomach and diaphragm caused difficulties in dissection in
seven out of ten. Only in three patients revised within 2 to 4 days due to an acute
diagnosis, were adhesions not a problem. Two of the acute reoperations were performed
laparoscopically, and the other by an open technique. In both patients revised
laparoscopically, it was evident that the left pleura had been breached before the second
operation, the left lung being visible through the hiatus after reduction of the stomach in
one, and a postoperative pneumothorax occurring before revision in the other. A
percutaneous endoscopic gastrostomy was placed under laparoscopic vision at the second
operation in one patient undergoing laparoscopic revision, functioning as an anterior 58 gastropexy. In the other undergoing laparoscopic revision, the fundoplication was reconstructed, as the original sutures had pulled out.
In those patients undergoing delayed surgery, a hernial sac was present in two, and the
stomach was densely adherent to surrounding structures, including left lung in five.
Following surgical revision, patients remained in hospital from 5 to 18 days (median 7).
Nine patients experienced no post-operative complications, whilst one patient developed a
large pleural effusion delaying discharge. Long telm relief of symptoms was achieved in
all patients, with follow-up ranging up to two years.
1.2.4.2.3 Díscussiott
This study highlights the risk of post-operative paraoesophageal herniation following
laparoscopic Nissen fundoplication. Whilst this can occuf following open
fundoplica¡ign113,123,165,207 , the risk may be greater following laparoscopic
fundoplication. This risk was not associated with patient weight or any individual
operator. There are, however, several factors which may make this complication more likely.
Early in the Royal Adelaide Hospital experience with the laparoscopic procedure, the
oesophageal hiatus was repaired infrequently as this was consistent with earlier "open"
experience, where in over 300 patients hiatal repair had only been performed in
approximately 20Vo of patients, with good results and an incidence of paraoesophageal
hernia of less than2Vo (unpublished data). Although these results almost certainly under
represent the incidence following open surgery, as follow-up is available for only 7 5Vo of-
patients, there were no early para-oesophageal herniae in the open cases, unlike the early
laparoscopic series where the results demonstrate a signifîcant association between failure
to narrow the diaphragmatic hiatus and post-operative para-oesophageal herniation- A
similar correlation between omission of hiatal repair and para-oesophageal herniation has
1 65. been reported following open fundoplication in .¡i1¿tsn 59
There are several possible explanations for why the incidence of para-oesophageal hernia is apparently higher after laparoscopic fundoplication. First, patients undergoing
laparoscopic procedures experience significantly less post-operative pain, and are able to
return to normal activity much earlier. This may allow them embark on activities which
raise intra-abdominal pressure which can forcefully project the stomach into the thoracic
cavity, before the hiatal area is obliterated by scar tissue during the normal healing
process. One of the patients in this series developed symptoms acutely after performing
heavy physical activity two weeks after surgery, and two had prolonged and violent
vomiting in the early post-operative period before developing an acute hernia.
Second, during laparoscopic dissection of the oesophagus it is easy to dissect higher into
the hiatus than during open fundoplication and the area behind the oesophagus is probably
opened up to a greater degree than with open fundoplication. Third, as a contributing
factor in the chain of events leading to herniation, is injury to the left pleura, which is at
risk when dissecting behind the oesoph agw225. In many of the patients undergoing
revision, at surgery the fundus of stomach was densely adherent to the left lung, and both
patients who underwent acute laparoscopic revision were noted to have had a pleural
breach. A pneumothorax may develop during surgery due to left pleural injury, but not be
noticed by either the anaesthetist or surgeon. As this rapidly resolves because of the use
of carbon dioxide gas, it does not cause a clinical pneumothorax and yet it is possible that
the fundus of the stomach may move into the left pleural cavity because of negative
intrathoracic pressure, which might be accentuated by postoperative retching or vomiting.
What can be done to reduce the likelihood of this complication? Because of this early
experience, posterior hiatal repair is now performed whenever possible, even in the
absence of a pre-operative hiatus hernia. This complication can still occur despite repair
of the hiatusSl, but it is believed to be less likely if an effective repair is undertaken.
Barium contrast radiology is also performed on all patients within 48 hours of surgery so
that if a problem is evident, laparoscopic repair can be performed at an early stage, as
demonstrated by two patients later in this group of patients. Delayed open revision has 60 been difficult in all instances and if early laparoscopic reduction and gastropexy is possible then it is preferable.
Although paraoesophageal hernia occurs after open fundoplication it seems likely that
laparoscopic fundoplication may predispose patients to this complication. Nevertheless,
routine narrowing of the oesophageal hiatus and awareness of the potential for this
problem may reduce the incidence or at least allow early laparoscopic repair to be
undertaken.
L.2.4.3 Stenosis of the oesophageal hiatus 'Watson (published previously as; DI, Jamieson GG, Mitchell PC, Devitt PG & Britten-
Jones R. Stenosis of the oesophageal hiatus following laparoscopic fundoplication.
Archives of Surgery 1995; 130: 1014-1916221¡
Troublesome dysphagia has been reportedgO following laparoscopic Nissen
fundoplication. It is usually thought to be due to either a pre-operative oesophageal
motility disturbance or the construction of a tight fundoplication. However, among the
initial250 procedures performed at the Royal Adelaide Hospital, three patients required
re-operation for postoperative dysphagia because of stenosis of the oesophageal hiatus.
1.2.4.3.1 Report of cases
Case No 7
A 36 year old man with gastro-oesophageal reflux disease and recurrent duodenal ulcer
ulceration, underwent elective laparoscopic Nissen fundoplication and highly selective
vagotomy. Pre-operative oesophageal manometry revealed normal oesophageal body
motility and a hypotensive lower oesophageal sphincter (9 mm Hg). The surgery
progressed uneventfully. The vagal branches and associated vessels on the lesser curye
gastric wall were dissected and divided between clips. The vagal nerve trunks and
associated tissue were then swept off the oesophagus from below, with small vessels 6I again divided between clips or cauterised. To complete the procedure a 3600 loose
Nissen fundoplication was constructed over a 52Fr bougie without division of the short gastric vessels. Because the approach to the oesophagus during highly selective vagotomy involves the dissection of the oesophageal wall from below, the diaphragmatic crura were not formally dissected and crural repair was not undertaken. Post-operatively the patient made an initially good recovery with minimal immediate dysphagia, and was discharged 3 days later.
He returned after one week describing severe and worsening dysphagia commencing 4 to
5 days following fundoplication. Endoscopy was performed, demonstrating an intact but apparently tight fundoplication. Dilatation with an 18 mm diameter bougie was performed and followed by symptomatic improvement. However, the dysphagia again worsened and two further dilatations failed to achieve long term relief. Barium meal X- ray confirmed oesophageal narrowing at the level of the upper border of the infradiaphragmatic fundoplication. Post-operative oesophageal manometry performed2 days before surgical revision again demonstrated normal oesophageal body motility and a lower oesophageal sphincter resting pressure of 13 mm Hg.
Further surgery was performed 14 weeks after the original procedure, by an open technique. By this time the patient had lost 6 kg from his original weight. Whilst the fundoplication appeared to be correctly constructed and not tight, immediately above it the diaphragmatic hiatus had stenosed to form a tight stricture compressing the lower oesophagus. This was widened by dividing the diaphragm anterior to the oesophagus, and post-operatively there was immediate relief of dysphagia. The patient was well and free of dysphagia 6 months later.
Case No.2
A 36 year old woman with symptomatic gastro-oesophageal reflux underwent a
technically easy laparoscopic Nissen fundoplication without division of the short gastric
vessels. A loose fundoplication was constructed around a 52Ft bougie. The 62
oesophagus and diaphragmatic hiatus were dissected with hook cautery, and because no hiatus hernia was present, hiatal repair was not undertaken. Pre-operative manometry
had demonstrated normal oesophageal motility and a lower oesophageal sphincter resting
pressure of 10 mmHg. Her early progress was uneventful and she was discharged two
days later experiencing minimal dysphagia.
Two weeks after surgery she returned with progressive, severe dysphagia commencing in
the second post-operative week. Endoscopy was performed suggesting a tight
fundoplication, which was dilated with an 18 mm diameter bougie. This failed on three
occasions to achieve prolonged relief of symptoms. Subsequent barium meal X-ray
revealed oesophageal obstruction at the level of the fundoplication, and post-operative
manometry, revealed the lower oesophageal sphincter resting pressure to be 13 mmHg.
Further surgery was performed 7 months following the original procedure. Stenosis of
the oesophageal hiatus immediately above a well constructed Nissen fundoplication was
demonstrated with the hiatus narrowed to 5 mm diameter and failing to admit the tip of a
finger. Contrasting this, a finger could be slid between the fundoplication and the
oesophagus whilst a52Fr bougie was located in the oesophagus, confirming the
looseness of the wrap. The stenosis was relieved by anterior division of the diaphragm,
achieving complete and continued relief of symptoms at 11 months follow-up. The
patient had previously lost 10 kg in weight following the initial fundoplication, but
gradually regained this after revision.
Case No 3
A 57 year old woman underwent an apparently uneventful laparoscopic Nissen
fundoplication for gastro-oesophageal reflux, during which the short gastric vessels were
not divided. The fundoplication was calibrated over a 52Fr bougie. Pre-operative
manometry revealed the lower oesophageal sphincter resting pressure to be 3 mm Hg, and
pre-operative endoscopy demonstrated a2 cm sliding hiatus hernia. At the time of
surgery a posterior hiatal repair was performed using a single 2/0 Prolene suture. Her 63 post-operative recovery was uneventful and she was discharged 3 days later experiencing minimal dysphagia.
During the 2nd week after surgery she also developed severe dysphagia. Endoscopy
suggested narrowing at the level of the fundoplication and dilatation with an 18 mm
bougie was performed. Whilst this provided sholt term symptomatic improvement, her
symptoms rapidty returned. A further endoscopic dilatation failed to produce longer
lasúng benefits, and she progressed to surgical revision 2 months following the original
procedure. Oesophageal manometry was not petformed before revision in this patient.
The fundoplication was again demonstrated to be loose and well constructed by the easy
passage of a finger between the fundoplication and the oesophagus whilst a 52Fr bougie
was sited in the oesophagus, and the dysphagia was found to be due to hiatal stenosis to a
diameter of approximately 5 mm. Anterior division of the hiatus relieved the stenosis
with relief of symptoms when reviewed 6 months following surgery.
1.2.4.3.2 Díscussion A minority of patients following laparoscopic fundoplication develop persistent
dysphagiagO, usually due to a tight fundoplication or a pre-existing oesophageal motility
disorder. Patients whose lower oesophageal sphincter mechanism fails to relax fully at
post-operative oesophageal manometry may be at greater risk of this symptom43. The
three patients described, all developed dysphagia because of the mechanical problem of
scarring and stenosis of the oesophageal hiatus. All three described short periods of good
swallowing in the first post-operative week, before the progressive onset of severe
dysphagia. All patients were unresponsive to oesophageal dilatation, requiring further
surgery to widen the narrowed hiatus. The operative findings, confirmed by post-
operative oesophageal manometry in the first two cases, proved that a tightly constructed
fundoplication was not responsible for the dysphagia experienced. The onset of
symptoms approximately one week after surgery, after an initial period with minimal
dysphagia in all cases, is consistent with the finding of stenosis. 64
This complication following fundoplication has not been reported following open
approaches to fundoplication. The precise mechanism remains unclear. All three
procedures involved different approaches to the hiatus and dissection of the lower
oesophagus. In the fundoplications the pillars of the diaphragmatic hiatus were well
defined to enhance oesophageal dissection. During highly selective vagotomy combined
with fundoplication the pillars were not seen as the tissue surrounding the oesophagus was
swept up during the inferior approach. Metal clips were used extensively in the
fundoplication with highly selective vagotomy but were not used during either
fundoplication as the short gastric vessels were not divided. Suture repair to narrow the
oesophageal hiatus was performed in only one of the three cases-
Apart from the laparoscopic approach, the only factor common to all cases was the use of
diathermy during oesophageal dissection. Whilst on review of operative videotapes this
did not appear to be excessive, this complication does not occur following open surgery
during which the hiatus is usually dissected by a combination of scissors and blunt finger
dissection. The zone of injury created by electrocautery during laparoscopic surgery may
be greater than anticipated at the time. This factor has been implicated in late stenosis of
the common bile duct following laparoscopic cholecystec tomy233. It has recently been
demonstrated that following monopolar diathermy application to blood vessels, a zone of
damage extends up to 10 mm from the point of the diathermy contact2.
Because it is possible to dissect the oesophageal hiatus laparoscopically using a
predominantly blunt technique with nontoothed grasping instruments, the surgical
technique for laparoscopic fundoplication at the Royal Adelaide Hospital was modified to
either eliminate or minimise diathermy use. 65
I.2.5 THE LEARNING CURVE FOR LAPAROSCOPIC FUNDOPLICATION
1.2.5.1 Defrning the learning curve for Iaparoscopic fundoplication
(Published previously as; Watson DI, Baigrie RJ & Jamieson GG. A learning curve for laparoscopic fundoplication: Definable, avoidable, or just a waste of time? Ann Surg
1996;224: I98-2ß2I3\)
Initial experience with laparoscopic Nissen fundoplication confirmed that whilst overall outcomes for antireflux surgery appear to be improved by the laparoscopic approach, there are also complications which are unique to the laparoscopic procedure; oesophageal perforation34,228,acute para-oesophageal herniationl6,220 stenosis of the oesophageal lliv¡tß22I and pneum6¡¡s¡Ð(182. As this experience comprised procedures performed by a greater number of surgeons and surgeons-in-training than that reported from most centres, a component of the incidence of adverse outcomes may have been associated with learning difficulties, which may be less apparent in reports from individuals or smaller groups of surgeons. There has also been no agreement about what constituæs adequate supervised experience before the solo performance of complex laparoscopic procedures, and what is the true length of the learning curve for antireflux surgery. This study sought to determine whether a learning curve can be defined, whether it is clinically significant, and whether any of the adverse outcomes associated with it can be avoided by the careful supervision of new surgeons.
1.2.5.1.1 Methods
From September 1991 to July 1995, 11 surgeons or surgeons-in-training undertook 280 laparoscopic fundoplications at the Royal Adelaide Hospital. Follow-up ranged up to 46 months (median 18) for this analysis. Each surgeon's individual experience commenced at a different stage in the overall institutional experience. Data for these procedures was analysed to determine the learning curve for laparoscopic fundoplication. 66
Information was collected for each of the following; patient weight, complications, requirement for surgical revision, operating time, perfolmance of hiatal repair and division of short gastric vessels at operation, intra-operative conversion from laparoscopic to open fundoplication, post-operative hospital stay, and three clinical outcome measures assessed 3 months after surgery (presence of solid food dysphagia, patient satisfaction score, and the expressed willingness of each patient to undergo surgery again under identical preoperative circumstances). Complications occurring within 30 days of surgery were included in the analysis, as was any requirement for further surgery related to the primary procedure at any stage after the initial surgery.
Solid food dysphagia was determined using the visual analogue scale (0 = no dysphagia, to 10 = severe dysphagia). For this study, a score from 4 to 10 inclusive was determined represent moderate to severe dysphagia for solid food. The satisfaction score was also determined by a visual analogue scale (0 = totally unsatisfied, 10 = totally satisfied). A score of 7 or greater was accepted as evidence of patient satisfaction with the surgical outcome.
This information was then analysed in three ways; a) The overall institutional learning curve was determined by comparing subgroups of patients, defined by their chronological order within the overall institutional experience; procedures1to20,2lto50,51 to100, 101 to150, 151to200,201to250,
and25l to 280. b) Individual learning experiences were assessed for each surgeon by determining each patient's chronological position in each individual's experience. Operations 1 to 5 for each surgeon were combined into one group, as were operations 6 to 10, Il to 20,2I to 40, 41 to 60, and 61 and over respectively. c) Because surgeons commenced laparoscopic fundoplication at different times
during the institutional experience, the outcomes for 'early' and'late starters'were compared. The first two surgeons performing laparoscopic fundoplication commenced
clinical procedures following preliminary development work in the animal laboratory, 67 whereas the last six surgeons were able to assist a number of procedures before being
proctored by surgeons who had by then attained a large clinical experience. The tirst two
surgeons were defined as'early Starters', whereas the last six surgeons, who all
commenced clinical procedures 18 months or more after the 'early starters', were defined
as 'late starters'. The experience of 3 other surgeons commencing fundoplication at an
intermediate stage was omitted from this part of the analysis. Groups of patients
undergoing procedures 1 to 10 and 11 to 20 for both groups of surgeons were compared,
to assess the effect of supervision during a surgeon's early laparoscopic experience.
1.2.5.1.2 Results
The analysis of the overall institutional experience is summarised in Table 1.2.5-1. The
complication rate (figure 1.2.5-1r),early and late re-operation rates (figure t.2-5.2), and the
necessity for conversion to open surgery, were all greater during the first 50 procedures,
particularly during the first 20 procedures performed. All parameters stabilised beyond
this initial experience. Although a lower complication rate was seen only in patients 101
to 150, all other outcome measures in this group were similar to the other patient cohorts
beyond the first 50 patients, suggesting that this low complication rate is simply a
statistical variation (i.e. 1/50 versus 5150,P=0.20, Fisher's exact test). Median patient
weight remained stable throughout the overall experience suggesting that surgical
diffrculty was unchanged throughout the study. Operating time (figure l-2.5.3) declined
steadily as overall experience improved, with the most significant improvement occurring
after the initial 20 procedures, despite a greater number of patients undergoing hiatal
repair and division of short gastric vessels later in the study. The 3 month clinical
outcomes of dysphagia, patient satisfaction score and willingness to undergo surgery
again, did not improve with greater overall experience. 68
Table 1.2.5.1 overall experience with laparoscopic fundoplication determined by chronological order for all cases (all figures median)
Patient Number 1-20 2r-50 51-100 101-150 151-200 20r-250 2s1-280
Complication 3OVo lTVo 87o 2Vo L2Vo 70% lOVo rate
Reoperation 25Vo I3Vo IjVo 6Vo l2Vo 6Vo 3Vo rate
Conversion 2OVo 17Vo l\Vo l2Vo 70Vo 47o 7Vo to open
Weight (kg) 72 78 75 78 80 76 84
Operating 185 110 90 90 85 77 80 time (min)
Hiatal repair 5vo 33Vo ZOVo 3lVo 90Vo 9OVo 93Vo
Short gastrics 2OVo 7Vo \Vo 27Vo 5I% 48Vo 30Vo divided
Hospital 5 4 4 J .J J J stay (days)
Dysphagia 227o l4Vo 33Vo 17% 23Vo 35Vo nla 3 months
Satisfaction 88Vo 837o 89% 90Vo 88Vo 9lVo nla score 7-10
Would have 94Vo 79Vo 95Vo 90Vo 88Vo 94% nla again 69
Figure 1.2.5.1 complication rate versus overall experience.
40
30 s o (ú
Ê 20 .9 (ú e CL E o o 10
0 1-20 21-50 51-100 101-150 151-200 201-250 251-2AO
Number in unit's exPerience 70
Figure 1.2.5.2 Re-operation rate versus overall experience.
30
s 20 o (ú c .9 (ú
(L0) 0 10 o É.
0 1-20 21-50 51-100 101-150 151-200 201-250 251-280
Number in unit's experience 7I
Figure 1.2.5.3 Operating time versus overall experience.
200
.EÊ,
o E
tt) c 100 G o CL o c Ît.g o =
0 1-20 21-50 51-100 101-150 151-200 201-250 251-280
Number in unit's experience
The outcome of fundoplication analysed according to each surgeon's individual
experience is summarised in Table 1.2.5.2. The complication rate (figure I.2.5-4) and re-
operation rates (figure I.2.5.5) were higher for patients undergoing surgery during a
surgeon's first 5 procedures, declining to approximately I}Vo beyond the initial 5 cases
(figure I.2.5.4). A similar trend for surgical revision (figure 1.2.5-5) was seen, with the
highest rate occurring during a surgeon's first 5 cases, and declining slowly thereafter. In
contrast to the overall experience, the rate ofconversion to open fundoplication did not
improve with greaær individual experience. Operating time (figure l-2.5.6) declined
steadily, although not as dramatically as the decline seen with the analysis of the overall
experience. The performance of hiatal repair was more likely later in both the overall
and individual experiences, and division of short gastric vessels was more commonly
performed later within the context of a prospectively randomised trial. Hiatal repair
became a routine in an attempt to avoid the complication of para-esophageal herniation. 72
Table I-2.5.2 demonstrates no significant correlation between either routine hiatal repair or division of short gastric vessels and the outcome measures described. The 3 month outcomes of dysphagia, satisfaction score, and willingness to undergo surgery also failed to correlate with operative experience.
Table 1.2.5.2 Combined experience with laparoscopic fundoplication determined by chronological order for each surgeon (atl figures median)
Pøtíent Number 1-5 6-10 11-20 21'40 41-60 6I+ Number of 41 32 48 10 51 37 patients
Complication 24Vo 9Vo IjVo 47o 67o 13Vo rate
Reoperation lTVo l37o I07o 8.57o \Vo 5Vo rate
Conversion rc% l37o l4%o 97o 6Vo l67o to open
Weight (kg) 79 77 79 72 83 80
Operating t20 93 95 85 80 85 time (min) Hiatal repair 397o 29Vo 357o 607o 76% 977o
Short gastrics l5Vo I3Vo 27Vo 22Vo 447o 537o divided õ Hospital 4 4 3 J 3 3 stay (days)
Dysphagia I47o 327o 24Vo 277o 27Vo 19Vo 3 months
Satisfaction 907o 83Vo 90Vo 937o 857o 84Vo score 7-10
Would have 90Vo 967o 86Vo 9I7o 96Vo 89Vo again 73
Figure 1.2.5.4 Complication rate versus surgeon's experience.
30
s 20 o IU
Ê o G .9 -Gr 10 E o o
0 1-5 6-10 11-20 21-40 41 -60 61 +
Number in surgeon¡s exPerience 14
Figure L.2.5.5 Re-operation rate versus surgeon's experience.
20
* o G
Ê 10 .9 G c) ct o o E
0 1-5 6-10 11 -20 21 -40 41 -60 61+
Number in surgeon's exPerience 75
Figure 1.2.5.6 Operating time rate versus surgeonts experience.
150
Ê 'E
100 a) E
E) = (ú o ct o 50 tr .g!t o =
0 1-5 6-10 11-20 21 -40 41 -60 61+
Number in surgeon's exPerience
The experience of the 'early' and 'late starters' is compared in Table I-2-5-3. The parameters examined for the 'late starters' first 10 cases were similar to the 'early starters' second 10 procedures, and better than their first 10. 'Late starters' had lower complication, re-operation and conversion rates, as well as much quicker initial operating times, when compared to 'early starters' at comparable levels of experience. Clinical outcomes were similar at3 months, and were unrelated to either experience or surgeon.
Hiatal repair and short gastric vessel division were performed more commonly by 'late starters' due to alterations in surgical technique introduced later in the overall experience.
This increased rate of short gastric vessel division (within the context of a prospective randomised trial) may account for the higher median operating time of 120 minutes for
the 'late starters' second 10 operations. 76
Table I.2.5.3 Experience with laparoscopic fundoplication performed by 'early' and
'late starters' (all figures median) 'Early Starters' 'Late Stnrters' 1-10 11-20 1-10 11'20 Number of 20 20 30 10 patients
Complication 25Vo l57o I77o lOVo rate
Reoperation 2OVo lÙVo IjVo l07o rate
Conversion 207o 20Vo 7Vo 30Vo to open
Weight (kg) 76 76 81 78
Operating 185 95 100 t20 time (min)
Hiatal repair 5Vo 307o 63Vo 40Vo
Short gastrics 207o I07o 2IVo 507o divided - Hospital 5 4 J 2 stay (days)
Dysphagia 33Vo 247o 23Vo 20Vo 3 months
Satisfaction 94Vo 847o 93Vo 89Vo score 7-10
Would have 947o 847o 1007o 897o again
1.2.5.1.3 Díscussion
A surgeon's early experience with laparoscopic procedures is usually associated with prolonged operating times and technical difficulties due to the need to adapt to ne'w surgical instruments and an altered method of vision. This early leaming experience may be associated with a higher rate of intra-operative and post-operative complications, as well as a greater likelihood of conversion from laparoscopic to open surgery. Various 17 factors may influence the length of a learning curve and its consequences to patients.
Individual learning experiences may be shortened by supervision by experienced surgeons, and shared institutional experience may reduce the learning time for the individual surgeons ¡nuo1u"¿191. It is now well accepted that the learning experience with laparoscopic cholecystectomy has been associated with an increased incidence of common bile duct rnjuryl9O. This risk appears to be highest within the first 5 to 15 cases 8'1'69'235 of each individual's experienc e7 -
The results described in this study confirm an institutional leaming curve for laparoscopic fundoplication of up to 50 procedures. The problems associated with this were particularly significant within the first 20 cases. By examining individual experiences, it is apparent that problems were most likely during the first 5 procedures performed by individual surgeons. Complication, reoperation and conversion rates declined to approximately I\Vo during the next 15 procedures, with little further improvement beyond this point. The comparison of procedurcs performed by 'early' and 'late' starters confirms that proctoring can make a difference. The complication, re-operation and conversion rates were all significantly less in the first 10 cases petformed by 'late starters', and operating times were reduced by 85 minutes, despite a higher rate of posterior hiatal repair. These outcomes suggest that the problems associated with the introduction of laparoscopic fundoplication can be reduced by experienced supervision during a surgeon's early cases.
Table I.2.5J highlights two changes in operative technique which may impact on adverse outcomes following laparoscopic fundoplication. A superficial analysis may associate improved outcomes with division of short gastric vessels and routine posterior hiatal repair. Whilst routine repair should reduce the incidence of early post-operative para- esophageal herniation22O, this problem contributed only a proportion of the postoperative
morbidity analysed, with repair not impacting greatly on the total adverse outcomes seen
later in this experience. Para-esophageal herniation continued to occur sporadically in
the second half of the series, despite hiatal closure. Later in the series, short gastric t8 vessels were divided, usually within a prospective randomised trial. A full analysis of this trial is discussed in section 3.2. The data in this study does not demonstrate that division had any influence on the parameters measured, with the exception, perhaps, of operating time for 'late starters' (Table L.2.5.3).
These results suggest that the first 20 operations at the Royal Adelaide Hospital and the first 5 procedures for each individual surgeon were associated with a definable learning experience. The risk of adverse outcomes did not stabilise until either 50 procedures had been performed by the institution, or up to 20 by individual surgeons. The comparison of
'early'and 'late starters' confirms improved early outcomes with the provision of experienced supervision. The improved times in the early procedures performed by 'late starters' suggests that proctoring can save a considerable amount of time, and that the learning problems associated with laparoscopic fundoplication can be minimised by experienced supervision of a surgeon's approximately first 20 procedures.
1.2.5.2 Experience with laparoscopic surgery for gastro-oesophageal reflux beyond the learning curve
(Published previously as; Watson DI, Jamieson GG, Baigrie RJ, Mathew G, Game PA,
Devitt PG & Britten-Jones R. Laparoscopic surgery for gastro-oesophageal reflux:
Beyond the learning curve. Br J Surg 1996:83: 1284-1297217¡
To determine whether results of laparoscopic surgery for reflux improved with improved
experience, overall outcome was re-evaluated aftet excluding all patients operated on by
surgeons during their learning curve (as defined in section 1.2.5.1). This allowed the
analysis of the outcome of surgery in a group of patients whose surgery was performed by
experienced surgeons only. 79
1.2.5.2.1 Patíents ønd methods
From September 1991 to July 1995, 320 patients underwent a laparoscopic Nissen fundoplication by one of 12 surgeons and trainees at the Royal Adelaide Hospital. The surgeon's learning curve was defined as the first 20 procedures performed by any individual surgeon. These patients were excluded, thereby defining a group of patients whose surgery was performed by "experignced" surgeons. Patients who underwent another laparoscopic procedure at the same time as their fundoplication were also excluded from this analysis.
1.2.5.2.2 Results
Of the 320 patients who underwent the laparoscopic Nissen procedure, 140 operations were performed by surgeons who had completed 20 or fewer procedures. A further 4 patients had a cholecystectomy, and2 a highly selective vagotomy at the time of laparoscopic fundoplication. Thus, 174 patients wele identified who had undergone surgery by five "experienced" surgeons. These surgeons had each performed 102,64, 51,
39 and 27 procedures in their individual experience. There were 103 men and 71 women. Age ranged from 15 to 82 years (median 48), and weight from 51 to 120 kg
(median 80). A sliding hiatus hernia was demonstrated pre-operatively in 88 patients
(SIVo). Fifty four patients (3lVo) had undergone previous open abdominal surgery of whom 22 (I3Vo) had upper abdominal surgery.
One hundred and sixty procedures were completed laparoscopically, whilst 16 (97") required intra-operative,conversion to open surgery due to difficulties with the laparoscopic approach (Table 1.2.5.4). Operating time ranged from 30 to 210 minute.s
(median 80). Short gastric vessels were divided in 61 patients (35Vo), and the hiatus was narrowed with sutures in 116 (67Vo) patients. The post-operative hospital stay ranged from 1 to 19 days (median 3). 80
Table 1.2.5.4 Conversion to open surgery
Obesity 5
Large hiatus hernia 4
Adhesions 3
Perioesophagitis 2
Liver hypertrophy 1
Bleeding 1
Complications occurred in 9 patients (5Vo) within 30 days of surgery (Table 1.2.5.5). No respiratory morbidity was seen. The problem of severe dysphagia was managed successfully by endoscopic dilatation in two patients, endoscopic retrieval of a food bolus in one, and re-operation in two. The food bolus obstruction was associated with early consumption of large pieces of unchewed meat, despite specific instructions on avoidance. Early re-operation within 30 days of the original procedure was necessary in
4 patients (2.3Vo); once for acute paraoesophageal herniation, once for bleeding from a short gastric vessel, and twice for dysphagia (one of these was achieved laparoscopically on the second postoperative day by the removal and replacement of a single "ttghf' suture). A laær surgical procedure was required for another 6 patients (3.4Vo), between 2 and 11 months after the original surgery; 2 for paraoesophageal herniation (one revised laparoscopically at 6 months), 3 for stricturing of the diaphragmatic hiatus resulting in dysphagia despite a demonstrably loose fundoplication, and one for gastric obstruction due to a poorly constructed fundoplication.
Table L.2.5.5 Post-operative complications
Dysphagia 5
Paraoesophagealherniation 1
Deep vein thrombosis 1
Bleeding short gastric vessel 1
Pneumothorax 1 81
Barium swallow examination 3 to 6 months after surgery in 113 (65Vo) patients confirmed an intact fundoplication in aIl (l00Vo) patients examined. However, delayed emptying of barium through the fundoplication was seen in7 (67o) patients, a finding which correlated with postoperative dysphagia. This was associated with either a presumed tight fundoplication (4 patients) or narrowing of the oesophageal hiatus (3 patients). A small paraoesophageal hernia was noted in a further 3 patients. This was symptomatic in two, requiring late surgical revision, whilst the other patient remains symptom free.
Post-operative oesophageal motility studies confirmed correction of acid reflux in all
(I00Vo) 75 patients examined post-operatively, and a rise in resting lower oesophageal sphincær pressure from a median 7 mmHg (I to 44) preoperatively, to 22 mmIJg (4 to 70) post-operatively. The clinical outcome in the subgroup of patients who had post- operative testing was similar to the remaining patients who did not undergo these tests.
Clinical follow-up of greater than three months was available for I44 (83Vo) patients. At three months, all but one patient was free from all reflux symptoms (99.37o). However,
29 patients (20Vo) described dysphagia for solid foods. Using a visual analogue scale from 0 (torally unsatisfied) to 10 (totally satisfied), I3I (917") patients rated their surgical
outcome 7 or better. A score of 7 to 10 approximated grades I and2 of a modified
Visick scale used in other studies (see sections3.I,3.2 and 3.3). I32 (9280) patients
expressed willingness to undergo the same procedure again under identical pre-operative
circumstances.
Of the 85 (49Eo) patients who werc followed for more than 12 months, all but one (99Vo)
were free of reflux symptoms. Nine (10.6Vo) reported some dysphagia for solid food and
77 (917") rated the operation 7 or better on the visual analogue scale of satisfaction.
Thirty two (18%) of the patients had also been interviewed 2 yearc after surgery. Thirty 82 one (977o) were free of reflux symptoms, 2 (67o) continued to experience dysphagia for some solid foods, and 29 (9I8") were satisfied with their clinical outcome.
1.2.5.2.3 Díscussíott
In comparison with earlier experience at the Royal Adelaide Hospital90 '2I9 air¡6 elsewhere34,77,¡aps¡oscopic fundoplication by experienced surgeons is associated with an improved outcome. The median operating time of 80 minutes, compared favourably with all comparable publications reporting more than 50 procedures34,77,90,219'228
(Table I.2.5.6), where median operating times ranged from 110 to 150 minutes.
Table 1.2.5.6 Early outcomes following laparoscopic fundoplication
Author No of patients Conversion Operating Time Complication Follow'up Symptomatic Rate (7o) (min) Rate (7o) (months) Failure (Øo)
Cuschieri35 116 0.9 150 (mean) 13 13 (median) 8.6
Jamieson9O 155 72.3 120 (median) 9'7 9 (mean) 1.9
Vy'eerß228 ß2 3.3 118 (mean) 7 '5 3 (median) 1.1
HinderTT 198 3.0 150 (meclian) lr.6 12 (median) 8.0
Cadierel9 80 3.8 150 (median) 8.8 3 (median) 0.0
HallerbackTl 60 15.0 110 (median) 10.0 6 (approx) 2.0
']he 57o complication rate in this analysis improves on the rate of I}Vo seen in the earlier
experience (section I.2.Z.Z¡2I9 and the rate of tp to l3%o reported by
s¡hs¡s19,34,7I,77,90,219,228 (Table I.2.5.6). Of importance was a reduction in the
incidence of post-operative para-oesophageal herniation and the incidence of venous
thromboembolism in this group of patients. The unusual problem of hiatal stenosis was
still seen in this adjusted series, although the necessity for revision of the fundoplication
for persistent dysphagia was uncommon (2 patients). The early re-operation rate of
2-3Vo, as well as the overall re-operation rate, declined fi'om the rate seen earlier (section
L.2.3.2¡219. Although the complication and re-operation rates compared favourably 83 with previously published experience of open Nissen fundoplic¿¡isn43'112'I23'I95, ¡¡s early reoperation rate may still be higher for the laparoscopic procedure. It is unusual for reoperation to be required within 12 months of open fundoplication. On the other hand, the follow-up of patients who had laparoscopic fundoplication at the Royal Adelaide
Hospital has been much more intensive, and the follow-up was obtained by a non-clinical scientific officer rather than the operating surgeon, or any other member of the hospital clinical staff.
Late follow-up of 85 patients reviewed 12 months following surgery, and32 patients reviewed at 2 years confirmed the durability of lapatoscopic fundoplication as a treatment for symptomatic gastro-oesophageal reflux; very few patients developed further reflux symptoms. The incidence of solid food dysphagia has declined with longer follow-up, and patient satisfaction remained at9l7o one and two years after surgery.
I.2.6 SHOULD THE CHOICE OF PROCEDURE FOR REFLUX BE INFLUENCED BY PREOPERATIYE PATHOLOGY?
l.2.6.l Preoperative endoscopic grading of oesophagitis versus outcome 'Watson (Published previously as; DI, Foreman D, Devitt PG & Jamieson GG.
Preoperative grading of oesophagitis versus outcome following laparoscopic Nissen
fundoplication. The American Journctl of Ga,stroenterology 1997;92:222-225215¡
As well as argument over technical aspects of the laparoscopic procedure, disagreement
has arisen about patient selection for the laparoscopic approach3'77 'I0I'Ill'768'2I9.
Whilst several reports describe good outcomes following the laparoscopic Nissen
procedure when applied to an unselected patient populati on3,77 ,219, others have argued
that a more selective approach should be followedlOl'115'168'170. It has been
recommended that the laparoscopic approach should not be used in patients with
endoscopic evidence of complicated gastro-oesophageal reflux disease, i.e. peptic 84 stricture formation or Barrett's oesophagus, due to an associated incidence of shortened oesophagus, and potential technical difficulties with the laparoscopic approach101,115,168,170. It has sometimes even been suggested that surgical treatment is inappropriate for patients who do not have ulcerative oesophagitis evident at pre- operative endoscopy examination. This study was performed to determine whether the presence of unequivocal oesophagitis at pre-operative endoscopy, and its severity, would influence the clinical outcome following a laparoscopic Nissen fundoplication procedure in patients with objectively proven gastro-oesophageal reflux disease.
1.2.6.1.1 Patíents ønd methods
Patients undergoing surgery between October 1991 and December 1995 were selected for this study if they met the following inclusion criteria: 1) Objective evidence of gastro-oesophageal reflux disease at either pre- operative endoscopic examination (ulcerative oesophagitis present), or 24 hour
ambulatory pH monitoring (pH<4 for more thanTVo of the monitoring period, or pH<4 between 47o andTVo of the monitoring period with a 507o or better symptom correlation
with reflux events).
2) Adequaûe details of the pre-operative endoscopy were available to enable
oesophagitis to be accurately graded.
3) The patients underwent a 3600 Nissen fundoplication. 4) The patients were assessed clinically at least 3 months following surgery
by an independent investigator.
Patients were excluded from analysis if they had not undergone adequate clinical follow-
up, had undergone a partial fundoplication, or the preoperative endoscopy details were
insufficient to allow accurate grading of oesophagitis.
Oesophagitis was initially graded according to the Savary-Miller classifiç¿1i6n161. This
grading system was then reduced to 3 groups according to Table 1.2.6.7. Group 1
comprised patients with no evidence of ulceration at pre-operative endoscopy. Group 2 85 comprised patients with uncomplicated erosive or ulcerative oesophagitis, and Group 3 comprised patients with complicated disease, i.e. stricture tbrmation or columnar lined lower oesophagus (Barrett's oesophagus). Stricture formation was regarded as present when the endoscopist either had difficulty passing the endoscope through a region of narrowing or had to undertake a dilatation before passing the endoscope. Columnar lined lower oesophagus was regarded as present if any area of columnar mucosa extended more than 3 cm above the anatomic gastro-oesophageal junction. For patients who underwent more than one endoscopic examination before surgery, the highest grade was used in this study, to minimise the influence on endoscopic findings of pre-operative medical treatment with acid suppression.
Tabte1,.2.6.l Oesophagitisgrading
(I to IV according to Savary Miller)
Group I
Grade 0 normal mucosa without lesions, although erythema may be present
Group 2
Grade I one or more isolated lesions with erythema
Grade II multiple confluent lesions which are not circumferential
Grade trI circumferential confl uent lesions
Group 3
Grade IV chronic complicated lesions (deep ulcers, strictures, Barrett's oesophagus
The following post-operative clinical parameters were assessed using visual analogue
scales:
1) Heartburn (0= no heartburn, 10= severe heartburn).
2) Dysphagia for solid foods (0= no dysphagia, 10= severe dysphagia). 3) Overall satisfaction with surgical outcome (0= not satisfied, 10= entirely satisfied)
Patients were also asked if they would elect to undergo surgery again if they had the
opportunity to again make a decision about treatment. Data was analysed using a one- 86 way analysis of variance (ANOVA) to compare the 3 patient groups. A P value of less than 0.05 was considered to be significant.
1.2.6.1.2 Results
231patients undergoing laparoscopic anti-reflux procedures between October 1991 and
December 1995 met the entry criteria for this study. All patients were offered a laparoscopic fundoplication, irrespective of preoperative oesophagitis grading, preoperative hiatal hernia size, and likely adhesions from previous open upper abdominal surgery. Patients who underwent a laparoscopic partial fundoplication (some as part of a trial and some as a result of surgeon preference) and patients who had not yet undergone appropriate clinical follow-up due to recent surgery were excluded from analysis. Ninety one patients were female and 140 male. Age ranged from 15 to 80 years (median 47) and follow-up ranged from 3 to 36 months (median 12 months).
Oesophagitis grading is summarised in Table L.2.6-2. Fifty nine patients had no evidence
of endoscopic oesophagitis (group 1), 148 had uncomplicated oesophagitis (group 2) and
24hadBarrett's oesophagus or an oesophageal stricture (group 3). Of the 24 patients in
group 3, 5 had a peptic stricture alone, 12had biopsy proven Barrett's oesophagus (3 cm
or greater in length), and 7 had both a peptic stricture and Barrett's oesophagus. The sex
ratio, age distribution and follow-up intervals were similar for all 3 groups. 87
Table 1.2.6.2 Grading of study patients
Oesophagítís Grøde Number of patíents
Group 1 59
Grade 0 59
Group 2 148
Grade I 7I
Grade II 53
Grade III 24
Group 3 24
Grade IV 24
Eighteen (7.TVo) procedures required conversion to an open operation because of difficulties completing the procedure laparoscopically. Two (3.4Vo) patients in group 1,
18 (7.87o) in group 2 and3 (12.57o) in group 3 were converted (P= 0.15, Chi-squared test). Significantly, no operation was converted to an open procedure because of a
shortened oesophagus associated with either Barrett's oesophagus or stricture formation.
Only one procedure in the larger experience was converted due to a shortened
oesophagus. This patient had a giant mixed sliding/paraoesophageal hernia, and no
endoscopic evidence of oesophagitis.
Twenty (8-6V") patients underwent subsequent surgical revision between 2 days and 3
years after the original operative procedure. Nine of these were performed within 4
weeks of the first operation, 5 laparoscopically and 4 by an open technique. Eleven were
performed at a later stage, including one laparoscopically. The reasons for revision were
paraoesophageal herniation (10), fundoplication revision for dysphagia (6), widening of
the oesophageal hiatus (3), and gastric obstruction (1). The likelihood of revision was
unrelated to oesophagitis grading. Only one patient with complicated oesophagitis
required a second procedure. This entailed the laparoscopic repair of an acute 88 paraoesophageal hernia on the fourth postoperative day. The patient's subsequent clinical outcome has been excellent.
The clinical outcomes for each group are summarised in Table I.2.6.3. No significant differences in heartburn, dysphagia or satisfaction scores were found between the groups.
Absolute relief of heartburn, defined by a score of 0 on the 0 to 10 visual analogue scale for heartburn, was achieved in80Vo of patients in group I,737o in group 2 and967o in group 3. This analysis excludes a significant proportion of patients who were highly satisfied with their outcome, but scored I or 2 out of 10 for heartburn post-operatively.
A further analysis of the complicated oesophagitis group revealed no outcome differences between patients with Barrett's oesophagus and patients with a reflux related stricture prior to surgery.
Tabte 1.2.6.3 Post-operative clinical outcome versus pre-operative oesophagitis
grade
All figures are mean of 0 to 10 visual analogue scores (95Vo confidence intervals) Group L GrouP 2 GrouP 3 Heartburn 0.7 (0.2,1.3) 1.0 (0'6, 1.3) 0.08 (-0.09,0-26)
P=0.12
Dysphagia for solids 2.4 (1.7,3.2) 2-2 (7.7,2-7) L-4 (0.5,2-3)
P=0.33
Satisfaction 8.2 (7.6,8.9) 8.3 (7-8, 8.7) 9.1 (8.5' 9'8)
P=0.26
Overall satisfaction with the surgical outcome was high in all groups with only 8.4Vo of
patients in group I, L3-5Vo in group 2 and 4.27o in group 3 indicating that they would not
elect to undergo surgical treatment if able to make their treatment choice again.
Unsatisfactory results were mainly due to post-operative dysphagia, excessive flatulence 89 or the development of paraoesophageal hemiation. Short gastric vessel division did not influence any clinical outcome.
1.2.6.1.3 Díscussion
Whilst most laparoscopic surgeons would not hesitate to offer laparoscopic fundoplication to patients with ulcerative oesophagitis without complications (Savary
Miller grades I , II and III), controversy has arisen over the choice of surgical procedure for patients with Barrett's oesophagus, peptic stricture formation and possible oesophageal shortening101,115,168,170. Others have questioned the suitability of
surgery for patients without proven endoscopic oesophagitis, even if reflux disease is
objectively proven by pH monitoring.
Several reports have advocated a 'tailored approach' to the selection of patients and
surgical approaches for antireflux surgery62,10I,115,168,170, with many surgeons
advocating modification of the Nissen procedure to a partial fundoplication particularly in
patients with poor peristalsis demonstrated by preoperative oesophageal
manometry62,2ll. Peters and DeMeester have suggested that the selection of patients
for laparoscopic fundoplication should be restricted to patients with normal oesophageal
body motility, reduced lower oesophageal sphincter tone and uncomplicated endoscopic
oesophagitis101,168,170. They recommend that patients with a peptic stricture or
Barrett's oesophagus should undergo an open transthoracic Nissen, Belsey or Collis-
Belsey procedure rather than a laparoscopic Nissen procedure. It is suggested that the
transthoracic approach allows greater mobilisation of the oesophagus and the opportunity
for an oesophageal lengthening procedure in these patients who may be more likely to
have an associated motility disturbance or a shortened oesophagus-
Similarly, Little 115 utgo.r that the laparoscopic approach should be reserved for patients
with either minimal or no endoscopic evidence of oesophagitis. He suggests that an open
thoracic approach should be applied in patients with motility disturbances, stricture
formation and "advanced oesophagitis". Patients with "moderate, but not erosive 90 oesophagitis" are recommended to undergo an open abdominal procedure. However, while the idea of such a tailored approach may seem a rational and even attractive concept, no scientific data has been produced to sustain such advocacy.
While it is agreed that a shortened oesophagus reprcsents a considerable problem for surgeons wishing to perform a laparoscopic anti-reflux procedure, it was not encountered in any patients with complicated oesophagitis. The laparoscopic procedures performed in the group of patients with complicated oesophagitis were not any more technically demanding than the procedures performed for the other patients. Because the results of a laparoscopic Nissen fundoplication in patients with Banett's oesophagus and stricture formation in this study were as good, if not better than the overall study group, it is believed that it is appropriate to offer the laparoscopic approach to all patients, irrespective of their pre-operative oesophagitis grade.
Some surgeons and physicians may also have difficulty with the surgical treatment of patients without objective evidence of oesophagitis before surgery. The absence of endoscopic oesophagitis is regarded by some primary physicians as a contraindication to surgery due to a lack of evidence of reflux disease. However, the outcome of this study confirms that patients without oesophagitis, but with reflux disease documented by pH monitoring also have good relief of reflux symptoms following laparoscopic Nissen fundoplication.
Whilst endoscopj'is a helpful diagnostic tool, which confirms the pre-operative diagnosis of gastro-oesophageal reflux disease in the majority of patients, oesophagitis grading in this study did not predict post-operative outcomes following a laparoscopic Nissen fundoplication. Therefore, it is appropriate to undertake laparoscopic antireflux surgery in all patients with reflux, without regard for the degree of oesophageal damage indicated endoscopically. 97
1.2.6.2 Oesophageal motility parameters before and after laparoscopic Nissen fundoplication
(Published previously as; Mathew G, Watson DI, Myers JC, Holloway RH & Jamieson
GG. Oesophageal motility parameters before and after laparoscopic Nissen fundoplication. Br J Surg 1997; 67: 289-292136¡
Whilst recently published American Gastroenterology Association consensus guidelines do not support routine preoperative manometric assessment before antireflux sotg"ry1, oesophageal manometry is performed in many centres for the routine pre-operative investigation of patients undergoing surgery for gastro-oesophageal t"¡o*167. As well as documenting lower oesophageal sphincter pressure, it is the most useful method of assessment of peristaltic activity and is used by some surgeons to justify modification to
standard surgical approaches if patients have motility disorders. Previous studies have
demonstrated that the Nissen fundoplication results in an increase in the manometrically
determined pressures recorded in the lower oesophageal sphinctef zone following
surgery43,45,68,164. Both the resting lower oesophageal sphincter pressure (LOSP) and
the residual pressure following swallow induced sphincter relaxation (residual relaxation
pressure) are increased following Nissen fundoplica¡iett103.
It has also been shown recently that when the upper sphincær fails to open normally, a
ramp in hypopharyngeal pressure can be measured in the hypopharynx immediately
preceding the pharyngeal peristaltic pressure waue28. A similar pressure ramp has been
demonstrated in the distal oesophagus in an experimental model of oesophageal
obstruction at the gastro-oesophageal junction148. This pressure correlates with
resistance to the passage of a swallowed bolus across the sphincter, but its clinical
significance has not been detemined and its importance for the assessment of post-
fundoplication dysphagia is unknown.
Manometric changes following fundoplication performed by open
techniques43,45,68,l03,l64 have been well documented, although findings have not been 92 uniform. Whilst two recently published small series have reported changes in oesophageal motility before and after laparoscopic fundoplis¿1i6¡90,138, no detailed reports of changes in motility following large series of laparoscopic Nissen fundoplication have been published. This study was performed to quantitate the effect of a laparoscopic
Nissen fundoplication on oesophageal motility, including 'ramp'pressure, and to corelate this with independentþ measured symptom scores in order to determine whether or not changes in oesophageal motility predict adverse clinical outcomes.
1.2.6.2.1 Methods 1-2.6.2.1.1 Patient selection
Patients were selected for this study from a group of 350 who underwent laparoscopic
antireflux surgery between October 1991 and December 1995 at the Royal Adelaide
Hospital, and were included if they had undergone pre-operative oesophageal manometry,
laparoscopic Nissen fundoplication, and post-operative manometry in the University of
Adelaide Department of Surgery. This enabled subsequent independent review of the
original manometry recordings, and the determination of ramp pressures. Patients were
excluded if they had not undergone independent clinical assessment more than 3 months
following surgery. 103 patients (56 male, 47 female; age20 to 79 years, median 54
years) satisfied all of the inclusion criteria. This group did not differ significantly from
the total group in regard to age, sex, weight, operating time, and the clinical outcome
scores used in this study. Manometry was perfotmed between 3 and 6 months following
surgery (mean 5.5 months, confidence intervals 4.7,6-4)-
1.2.6.2.1.2 OesophagealManometry
Patients were fasted for 6 hours before each study and all medications affecting
oesophageal motility were ceased 2 days earlier if necessary, except proton pump
inhibitors which were ceased 5 days before motility studies. Oesophageal manometry
was performed using an 8-lumen water petfused catheter incotporating a sleeve sensor
(Dentsleeve, Adelaide), with signals recorded on a polygraph chart recorder (Grass
Instrument Company, Model TD, Massachusetts, USA). The lower oesophageal 93 sphincter (or post-fundoplication high pressure zone) was located by the station pull- through technique and the centre of the sleeve was positioned over the lower oesophageal sphincter. Each lumen of the catheter was connected in series with a pressure transducer
(Stratham P23lD,Gould Inc, Oxnard, California) and was constantly perfused with degassed distilled water at 0.6 mVmin by a low compliance pneumohydraulic pump
(Amdorfer Medical Specialities, Greendale, Wisconsin).
Basal lower oesophageal sphincter pressure was measured over a 5 minute period using a sleeve sensor, followed by measurement of the amplitude and propagation of primary peristalsis, and residual relaxation pressurc of the lower oesophageal sphincter during 10 swallows of 5 ml water boluses. Swallows were always more than 30 seconds apart.
1.2.6.2.1.3 Data analysis
Basal lower oesophageal sphincter (LOSP), referenced to intragastric pressure, was
determined from the recordings obtained during the initial 5 minute resting period.
Residual relaxation lower oesophageal sphincter or gastro-oesophageal junction pressure
was measured by averaging the minimum relaxation pressure recorded during each of the
10 water swallows.
Whilst the lower oesophageal ramp pressure was not measured routinely during the initial
manometry study, this was determined subsequently by an independent assessor who
reviewed the original manometric recordings. The ramp pressure represented a rise in
intraluminal oesophageal pressure, occurring just before swallow induced peristaltic
contractions. It was identified immediately proximal to the lower oesophageal sphincter
pre-operatively or proximal to the gastro-oesophageal high pressure zone post-
operatively. This pressure was determined by measuring the difference between end
expiratory basal oesophageal pressure and the maximum pressure of the pressure ramp.
While ramp pressure can be identified and measured from conventional manometric
recordings which use a polygraph chart recorder paper speed of 100 mm/min, it is more
clearly defined when using a faster speed of 5 mm/sec (figure 1.2.6.I). Ramp pressure 94 was sought using the faster speed in this study. Individual ramp pressures were measured for each of the 10 water swallows used to determine oesophageal peristalsis and residual relaxation pfessure, and a mean ramp pfessure calculated for each patient.
Figure 1.2.6.L Distal oesophageal manometric recording in which ramp pressure is readity visualised using a chart speed of 5 mm/sec.
A = amplitude of ramp pressure
B = amplitude of distal oesophageal peristaltic contraction
B
A
Peristalsis was also assessed during each of the water swallows. Propagation was
considered to be successful when a peristaltic wave progressed the entire length of the
oesophagus. Successful propagation of 7 or more of the l0 water swallows was used to
define normal oesophageal peristalsis.
Post-operative manometry was performed 3 to 6 months following surgery as part of the
follow-up protocol for patients undergoing laparoscopic Nissen fundoplication. 95
1.2.6.2.1.4 Clinicaloutcome scores
Clinical outcome following surgery was determined from the standardised clinical questionnaire administered within one month of the post-operative manometric assessment (i.e. 3 to 6 months after surgery). Symptomatic dysphagia was assessed using visual analogue scales for dysphagia when swallowing liquids (no dysphagia = 0, total dysphagia - 10), and when swallowing solid foods (no dysphagia = 0, total dysphagia= 10). The presence or absence of symptoms suggesting gas bloat was also determined by asking patients whether they ever experienced discomfort due to upper abdominal distension or a sensation of bloating of the upper abdomen, including occasional or mild bloat symptoms. Overall satisfaction with the surgical outcome was also determined using a visual analogue scale (completely satisfied = 10, totally unsatisfied = 0).
1.2.6.2.1.5 Statistical analysis
All data are expressed as mean wíth95Vo confidence intervals. Changes in manometric
outcomes before and after fundoplication, and comparisons of different patient groups
were assessed using the Mann Whitney U-æst. Spearman's correlation coefficient was
used to determine correlations between clinical and post-operative manometry outcomes'
1.2.6.2.2 Results
The effects of laparoscopic Nissen fundoplication on oesophageal body and lower
oesophageal sphincter function are summarised in Table 1.2.6.4. Surgery resulted in
significant rises in resting lower oesophageal sphincter pressure, residual relaxation
pressure, and ramp pressure. Fundoplication had variable effects on peristaltic success
Fifteen patients propagated less than 7\Vo of wet swallows pre-operatively. Of these 8
had propag atedT\Vo of more wet swallows at post-operative manometry. On the other
hand, the peristalsis of 13 other patients changed from more thanT}Vo of swallows
propagated pre-operatively to less than 7}Vo post-operatively. 96
Table L.2.6.4 Changes in oesophageal motility following laparoscopic Nissen fundoplication
all figures mean (95Vo confidence intervals), all pressure values in mmHg Preoperative PostoPerative
Vo primary peristalsis 90Vo (85Vo,947o) 89Vo (847o,937o) P=0.38
LOS resting pressure 8.5 (6.8, 10.9) 2I.5 (19.5,23.5) P<0.0001
Residuat relaxation 1.2 (0.78,I.6) 10.8 (9.5, 12.0) P<0.0001 pressure
Ramp pressure 10.5 (9.9, 11.0) 20-5 (19-3,21.6) P<0'0001
Correlations between ramp pressure and resting lower oesophageal sphincter pressure and
residual relaxation pressure were assessed before and after fundoplication. A weak but
statistically significant correlation was found between pre-operative residual relaxation
pressure and ramp pressure (r=0.20, P=0.049). No significant correlation was found
between pre-operative resting lower oesophageal sphincter pressure and ramp pressure,
and between post-operative ramp pressure and either residual relaxation or resting lower
sphincter pressures.
Correlations between post-operative clinical outcomes and post-operative manometry
findings are summarised in Table 1.2.6.5. No correlation between clinical outcome and
either ramp pressure or peristaltic success was found. A weak but statistically significant
correlation was seen between residual relaxation pressure and dysphagia for liquids (r=
0.20). No significant correlations were evident between residual relaxation pressure and
other clinical outcomes, or solid food dysphagia and any clinical outcome. 97
Table 1.2.6.5 Correlation betweenpost-operative oesophageal manometry findings and clinical outcomes
all figures Spearman's correlation coefficients (r) Solid dysphagia Liquid dysphagia Satisfaction
LOS resting pressure 0.12 0.18 -0'08
Residual relaxation 0.18 0.20* 0.02 pressure
Ramp pressure -0.03 -0-02 -0'008
7o pimary peristalsis -0.I2 0'03 O'I2
* P=0.049, Þ0.05 for all other values
Dysphagia versus post-operative manometric findings was also examined by defining a
large patient group with either mild or no dysphagia (visual analogue score= 0 to 4), and a
smaller group with moderate or severe dysphagia (score= 5 to 10) for both solid and
liquid substances. This analysis is summarised in Table I.2.6.6- LOS resting pressure
was not significantly elevated in patients with moderate to severe dysphagia both for
liquids and solids. Residual lower oesophageal sphincter relaxation pressure was greater
in patients with more severe dysphagia for liquids (P=0.044), but not in patients with
dysphagia for solids. Ramp pressure was not increased significantly in patients with
more severe dysphagia. 98
Table 1.2.6.6 Rel ationship between oesoph ageal manometric findings and dysphagia
all figures mean (95Vo confrdence intervals), all pressure values in mmHg
Dysphøgía
None or míld Moderate or Severe Dysphagia for Solids n 88 15
LOS resting pressure 2l.I (r8.8,23.2) 24.0 (19.9,28.2) P=0.14
Residual relaxation 10.6 (9.2,72.0) tt.6 (8.4,14.9) P=0.49 pressure
Ramp pressure 20.4 (79.2,21.6) 2I.I (16.9,25.4) P=0.98
Dysphagia for Liquids n 82 2t
LOS resting pressure 20.8 (18.5, 23.1) 24.3 (r9.8,28.7) P=0.I2
Residual relaxation 10.1 (8.8, 11.5) t3.2 (r0.2, 16.r) P=0.044 pressure
Ramp pressure 20.6 (I9.3,2I.9) 19.8 (16.8, 22'8) P=0.65
Lower oesophageal sphincter pressure, residual relaxation pressure and ramp pressure
were all similar in highly satisfied patients (satisfaction score 7 to 10), when compared to
less satisfied patients (score 0 to 6). Manometric outcomes did not correlate with
symptoms of gas bloat (Table I.2.6.7).
Table 1.2.6.7 Post-operative oesophageal manometry and symptoms of gas bloat
all figures mean (957o confidence intervals), all pressure values in mmHg
Symptoms present SYmPtoms absent n=64 n= 39
LOS resting pressure 22.2 (19.6,24.7) 20-4 (I7 -2,23-7) P=0'32
LOS nadir pressure 11.4 (9.7, 13.0) 9-7 (1-9,lI-5) P=0'31
Ramp pressure 20.4 (18.9,22.0) 20-5 (79.0,22.0) P=0.69 99
1.2.6.2.3 Díscussion
This study documents changes in oesophageal manometry in 103 patients undergoing a laparoscopic Nissen fundoplication. Similar increases in resting lower oesophageal sphincter pressure, residual relaxation pressure and ramp pressure have been reported by other researchers following Nissen fundoplication performed using open surgical techniques 43,45,68'164 -
Two recent studies of small groups of patients have detailed changes in manometric findings before and after laparoscopic fundoplication. In a study involving 15 patients
who underwent a laparoscopic Nissen fundoplication, McAnena et ¿1138 demonstrated an
increase in resting lower oesophageal sphincter pressure from 5 to 11 mmHg, but failed to
report either residual relaxation pressures or ramp pressure. In their study, manometry
was performed with a triple lumen perfused catheter, equipped with side hole sensors only 'Watson rather than a perfused sleeve sensor. A second study reported by s¡ v1212
examined outcomes following a laparoscopic anterior partial fundoplication in 26
patients. Using a solid state manometry catheter, they found that mean resting lower
oesophageal sphincter pressure increased from 8.4 to 11.5 mmHg. This smaller pressure
rise compared to those following Nissen fundoplication perhaps reflects the nature of the
partial fundoplication performed.
A ramp in hypopharyngeal pressure was described recently by Cook et al in patients
undergoing upper oesophageal sphincter manometry for Zenker's diverticulum2S. This
ramp is thought to be a measure of sphincter resistance, and as such to correlate with
dysphagia. An experimental study using a feline -o6s1148 demonstrated that a similar
pressure ramp can also be measured proximal to the lower oesophageal sphincter, and that
its magnitude correlates with the degree of obstruction determined by the pressure in a
cuff around the lower oesophageal sphincter. On this basis, it is believed that the ramp in
the distal intra-oesophageal pressure during primary peristalsis triggered by swallowing
liquids reflects resistance to flow across the lower oesophageal sphincter. The increase in 100 ramp pressure following fundoplication is consistent with this notion and the known effect of fundoplication on lower oesophageal sphincter ¡u¡s1isrì43,138. It might be expected, therefore, that changes in lower oesophageal ramp pressure following Nissen fundoplication would correlate with dysphagia, due to the extrinsic compression caused by the circumferential fundal wrap. However, whilst this study demonstrates a significant rise in the ramp pressure following a lapatoscopic Nissen fundoplication, it did not correlate with dysphagia, suggesting that other factors such as integrity of primary peristalsis are more important. Measurement of ramp pressure therefore appears to be of no value for the postoperative assessment of patients following antireflux surgery.
Earlier studies in the era of open surgery have suggested that antireflux surgery can result
in improvement in postoperative oesophageal peristaltic activity45. Whilst this study did
demonstrate improvement in oesophageal primary peristalsis in 8 out of 15 patients with
abnormal preoperative motility, this was counter-balanced by reduced peristalsis in 13 out
of 88 patients with initially normal motility. Although rates of peristaltic success used to
define normal peristalsis vary among different institutions, and range from 70 to 907o,
using the definition of peristaltic success applied in this study100,122, these results
suggest that there has been no overall change in oesophageal peristalsis following
laparoscopic Nissen fundoplication, and do not support the concept that antireflux surgery
improves defective peristalsis. It is acknowledged, however, that it may be necessary to
study patients over a much longer follow-up time to be more sure of this statement. It
may even be possible that fundoplication could impair peristalsis if it generates sufficient
resistance to the passage of a swallowed bolus, especially in patients with low amplitude
peristalsis pre-operatively. The experimental feline study reported by Mittal et al
demonstrated impaired peristalsis in cats swallowing larger bolus volumes and in cats
with greater distal oesophageal outflow obstructionl43-
It is widely held that the adverse outcomes of dysphagia and gas bloat syndrome
following antireflux surgery are associated with the construction of a tight fundoplication,
and that elevated lower oesophageal sphincter pressures measured by post-operative ,l'¿
101 oesophageal manometry will confirm or pfedict these problems43'46. Few studies objectively examined this belief43. It has been found in a small group of patients previously studied, no relationship between resting LOSP and dyspha gia87 . On the other hand in a more intensively studied group at the Royal Adelaide Hospital, where many patients were selectively studied because of dysphagia, there was a relatively good correlation between resting LOSP, residual relaxation pressure, ramp pressure and dysphagiaSS.
Of the manometric variables measured in the current study, only post-operative residual
lower oesophageal sphincær relaxation pressure conelated with any adverse surgical
outcome. Dysphagia for liquids was more likely in patients with higher post-operative
residual relaxation pressures. This suggests that it is likely to be a better measure of the
degree of obstruction caused by a fundoplication, than other measured manometric
parameters. Nevertheless, the correlation coefficient of 0.20, whilst statistically
significant, does not support a strong clinical correlation.
The lack of any clinically significant correlation between manometric outcomes and the
clinical outcomes of dysphagia or satisfaction with surgery in the cuffent study was
unexpected, particularly as other studies have suggested that elevated residual relaxation
pressure may be associated with dysphagia or other adverse outcomes following
fundoplica¡isrþ3,I21. However, these studies have usually retrospectively assessed
small groups of patients. The specific relationship between dysphagia and lower
oesophageal sphincter pressure, has not been prospectively studied in a large patient
group, using blinded clinical and manomeÍic assessments. Nevertheless, it is possible
that the number of patients with severe dysphagia is still too small for the measured
increase in lower sphincær resting pressure to reach significance. However, the findings
suggest that postoperative oesophageal manometry is of limited value in the assessment of
patients undergoing antireflux surgery. r02
1.2.6.3 The outcome of laparoscopic Nissen fundoplication in patients with disordered pre-operative peristalsis
(Published previously as; Baigrie RJ, Watson DI, Myers JC, Jamieson GG. The outcome of laparoscopic Nissen fundoplication in patients with disordered pre-operative peristalsis. Gut 1997 :¿O: 381-3858)
The use of a 3600 wrap remains controversial in patients with disordered peristalsis, some
surgeons preferring a partial or modified wrap to avoid post-operative
dysphagia46,62,l57,l80. Dysphagia occurs in most patients immediately after
fundoplication and is present at 3 month s in I0-40Vo42,219. By one year this has
reduced to less than 10Vo39,219. This study examined the incidence of symptoms,
including dysphagia, and patient satisfaction after laparoscopic fundoplication in patients
with disordered peristalsis.
r.2.6.3.1 Patíents and methods
1.2.6.3.1.1 Manometry
Oesophageal manometry was carried out using the same equipment and an identical
protocol to that described in section I.2.6.2.1.2. The amplitude and success of primary
peristalsis \ilas measured during at least 10 swallows of 5 ml water boluses at 30 second
intervals. Peristaltic success was defined as a wet swallow propagating the oesophageal
body by primary peristalsis, and peristaltic failure as a failure to produce a contraction
amplitude >10 mmHg in two or more recording sites. Manometric data were analysed for
the occurrence of failed primary peristalsis and/or hypotensive (sometimes called feeble)
peristalsis in the distal oesophagus. Abnormal primary peristalsis was defined as less
than 50Vo of water swallows, progressing through the whole oesophagus. A value
between 50 and 707o was regarded as equivocal primary peristalsis. A mean contraction
amplitude of less than25 mm Hg in the distal oesophagus was regarded as abnormal.
These criteria have been described previouslyl00'I22. 103
1.2.6.3.1.2 Patients
Thirty one patients with disordered peristalsis who underwent a laparoscopic Nissen fundoplication were identified and were categorised in four groups.
Equivocal Primnry Peristalsis; primary peristalsis >507o and
Abnormnl Primary Peristalsis; primary peristalsis <50Vo of swallows propagated,
maximum contraction pressure >25 mmHg.
Abnormnl Maximal Contraction Pressure; primary peristalsis >50Vo of swallows
propagated, maximum contraction pressure <25 mmHg-
Abnormnl Primnry Peristalsis and Maximnl Contraction Pressurer primary peristalsis
<50Vo of swallows propagated and maximum contraction pressure <25 mm Hg.
1.2.6.3.1.3 Follow-up
Preoperative assessment and post operative clinical follow-up utilising the standardised
questionnaire 3, 12 and24 months after surgery was assessed. Visual analogue scales for
heartburn (0 = no heartburn, 10 = severe heartburn), solid food dysphagia (0 = no
dysphagia, 10 = severe dysphagia), and patient satisfaction with the procedure (0 =
completely dissatisfied, 10 = completely satisfied) were analysed. A dysphagia score >3
was regarded as troublesome dysphagia. The patient's willingness to undergo the same
procedure again under similar pre-operative circumstances was also determined-
Between 3 and 6 months after surgery, patients were encouraged to undergo further
oesophageal manometry examination.
1.2.6.3.2 Resulß
Eighteen male and 13 female patients, with a median age of 45 (range = 33-81) were
studied. No patient was lost to clinical follow-up, with 1I (367") patients followed for
more than two years at the time of data analysis, 19 (61Vo) patients for 12 months and 1
(3Vo) patients for 6 months. Post operative manometry was performed on 17 (557"). r04
Equivocal Primnry P eristalsis ( n= 8 )
At one to two years six patients had satisfaction scores of at least 9 (Table 1.2.6.8), and
would choose to undergo the procedure again. Four patients had a pre-operative
dysphagia score >3, but by 12 months only one of these still had a score >3. One patient
with a satisfaction score of 6, developed abdominal pain 2 months after the operation and
was found to have retroperitoneal fibrosis requiring ureterolysis. She would still choose to
undergo the operation again. The last patient had a satisfaction score of zeto with a
reflux score of 10. While it seems probable that the wrap has failed, the patient, who had
significant psychiatric problems, was not prepared to be reinvestigated. No other patient
was troubled by bloating or reflux at one year. Two patients underwent post-operative
manometry with a rise in primary peristalsis in one patient and a fall in the other (Table
1.2.6.8 and Figure 1.2.6.2). 105
Table 1.2.6.8 Pre- and post-operative manometry data, and symptom scores for all30 patients. s and solids) is recorded. ntraction pressure in the distal oesophagusi LOS = nadir pressure.
Pafient group PP MCP LOS Nadir DysPhngia Satisfactian postop preoP Posto? Preo? Postoø postop postop preop postop Equivocal PP 60 100 t9 5 0 0 10 onty (n=8) 60 40 32 13 5 8 6 60 0 3 9 60 0 0 10 60 7 I 10 ó0 4 0 10 50 0 J 0 50 5 0 10
Abnormal PP 40 70 26 3 0 3 9 only (n=4) 30 10 0 9 30 8 0 9 10 100 25 -2 5 3 10
Abnormal MCP 2l 23 19 19 10 0 8 only (n=13) 20 31 9 -4.5 6 4 8 15 t2 10 7 0 0 10 22 t2 10 7 0 0 10 10 23 M 30 7 0 10 19 53 20 t3 0 J 10
24 7L 10 8 8 2 9
19 34 t6 t2 2 I 9
18 11 4 -3.5 5 2 4
15 58 15 6 0 2 10 20 4 0 10 22 0 0 10
Abnormal PP 40 100 2l 60 50 18 10 0 10 & MCP (n=6) 30 40 25 120 15 1 0 0 10 0 10 0 0 10 0 t5 5 J 10 0 0 0 J 10 20 20 10 0 10 106
Figure 1.2.6.2 Maximum contraction pressures in the distal oesophagus, measured preoperatively and at least 3 months post-operatively, in patients with abnormal pre-operative maximum contraction pressures
Pre-operative (z) Post-operat¡ve (%) 100 100
80 80
60 60
40 40
20 20
Abnormnl Primary P eristalsis (n=4 ).
All patients in this group had an excellent symptomatic outcome at 12 to 24 months, with satisfaction scores of 9 or 10 and no reflux (Table 1.2.6.8). All would undergo the operation again given similar pre-operative circumstances. Three of the four patients had a pre-operative dysphagia score >3, but no patient complained of dysphagia, bloating or reflux one year after the operation.
Post-operative manometry was available on two patients whose primary peristalsis had recovered in the first patient from 407o pre-operatively to 707o, and in the second patient
I07o pre-operatively to I007o (Table I.2.6.8 and figure l-2.6-2)- t07
Abnormnl M aximum C ontr action P re s sur e ( n= I 3 ).
Eleven patients who were 12 months post operation and one at 6 months, were virtually
symptom free, with satisfaction scores of at least 8 (Table 1.2.6.8). AII 12 would choose
to undergo the operation again, and no patient had residual reflux or was troubled by
bloating. Seven of the group had a pre-operative dysphagia score >3, but by 12 months
only one of these still had a score >3. Post-operative manometry was carried out on 11
patients in this group with seven patients showing a higher than pre-operative value, two
patients a lower value and two remaining unchanged (Table I.2.6.8 and figure 1.2.6.3)-
Figure 1.2.6.3 Primary peristalsis amplitude, measured preoperatively and at
Ieast 3 months post-operatively, in patients with equivocal or abnormal pre'
operative primary peristalsis
Pre-operative (mmHg) Post-operative (mmHg) 80
60 60
40 40
20 20
0 108
One patient had a satisfaction score of 4 and would not choose to undergo the procedure again. She denied dysphagia or reflux symptoms but had become intolerant of a variety of foods and developed tremors and palpitations which she attributed to the operation. Her operation had entailed division of short gastric vessels and hiatal repair, and a barium meal and repeat endoscopy nine months later revealed normal post-fundoplication appearances. Post-operative manometry and pH studies showed evidence of a very loose wrap , low contraction pressure and equivocal reflux (Table 1.2.6.8). Post-operatively, oesophageal emptying was 53 seconds - within the notmal range. The patient has been assessed by a cardiologist and a neurologist and her symptoms are thought to be anxiety related.
Abnormnl Primary Peristalsis and Maximum ContrcLctiott Pressure (n=6).
This group included three patients with an adynamic oesophagus. Two year clinical follow-up data were available for four patients, including those with an adynamic oesophagus. At 3 and 12 months, satisfaction scores were all 10, with no reflux. Three of the six patients had a pre-operative dysphagia score >3, bú 12 months after the operation no patient complained of dysphagia, bloating or reflux. Post-operative manometry in two patients, showed an improvement in both primary peristalsis and contraction amplitudes
(Table 1.2.6.8 and figures 1.2.6.2 and 1.2.6.3). The three patients with an adynamic oesophagus declined post-operative manometry.
Analysis of the whole group shows median post-operative values of 16 mmHg (range 4 to
50) for basal lower oesophageal sphincter pressul'e, and 6 mmHg (range -4.5 to 22-5) for nadir lower oesophageal sphincter pressure. Twenty-eight (907o) patients had satisfaction scores of at least 8 and all would undergo surgery again. Fifteen (48Vo) patients had dysphagia scores of >3 preoperatively but only two (67o) patients had persistence of this symptom at one year. 109
1.2.6.3.3 Díscussí.on
This study describes the first prospectively assessed series of patients with disordered peristalsis undergoing laparoscopic fundoplication. Even in the era of open fundoplication, there was a paucity of prospective data and most published work was retrospective. Kahrilas et al99 defined "peristaltic dysfunction" as either failed propagated peristalsis, or peristaltic sequences with feeble contractions. They suggested that between 25 and 50Vo ofpatients with peptic oesophagitis suffer from peristaltic dysfunction. Although experience at the Royal Adelaide Hospital suggests a lower prevalence, the true incidence is probably masked by differences in study populations and manometric selection criæria.
The importance of detecting a severe motility disturbance pre-operatively in patients referred for anti-reflux surgely patients, is a widely held belief on the grounds that a total fundoplication will be followed by obsttuctive symptoms. Recently DeMeester's group reported successful outcomes from open Nissen fundoplications in over 50 patients with mild non-specific motility disorders197. Although none of their patients had post- operative manometry, symptom outcome was excellent. However, their patients all had considerably milder disorders of peristalsis than those patients reported in this study, and their manometric selection criteria for a Nissen fundoplication would have excluded almost all the patients with disordered primary peristalsis in this study.
Whilst manometric outcome studies should yield reproducible results, there is evidence, particularly from 24 hour ambulatory manomotry studies, that outcomes can vary with time in individual patients. It is difficult, however, to persuade patients to submit to more than one manometric examination following successful antireflux surgery to guarantee the reproducibility of the outcome data in this study. Nevertheless, the data presented do suggest that disordered peristalsis (excluding achalasia) is not a contraindication to a 3600 loose laparoscopic fundoplication, with the incidence of dysphagia being improved by
surgery. Pre-operative dysphagia is increasingly being recognised as a common symptom in gastro-oesophageal reflux disease12 with 30-40 7o of patíents in the overall experience 110 at the Royal Adelaide Hospital reporting this symptom (see sections 3.2 and 3.3). The
487o incidence of pre-operative dysphagia in the patients in this study retlects their preoperative peristaltic dysfunction but this fell to 67o, one year after operation. Two patients had an imperfect clinical outcome, but neither have dysphagia as a significant symptom. The findings of this study suggest that a loose 3600 fundoplication, either with or without division of short gastric vessels, is appropriate treatment for gastro oesophageal reflux even when the patient has disordered oesophageal peristalsis.
1.2.6.4 Illness behaviour versus outcome following laparoscopic antireflux surgery
(Published previously as; Watson DI, Chan ASL, Myers JC & Jamieson GG. Illness behaviour influences the outcome of laparoscopic antireflux surgery. The Journal of the
American College of Surgeons 1997;194' 44-49214¡
It has been demonstrated by Pilowsky et aIl16 that illness behaviour has the potential to influence the clinical outcome for patients undergoing coronary artery by-pass surgery.
Illness behaviour determines the manner in which symptoms are perceived by patients, and acted upon144. Other studiesl5 have suggested that psychosocial factors can exert a long-term influence on recovery and rehabilitation following this type of surgery. Further sn¡¿is595,96 huue demonstrated psychological and social differences in patients with gastro-oesophageal reflux, and that these differences may alter the perception of acid reflux. The influence of illness behaviour on the outcome of antireflux surgery, however, remains largely unexplored. This study sought to evaluate the influence of pre and postoperative illness behaviour on clinical outcome in patients who underwent a laparosc opic Nissen fundoplication. 111
1.2.6.4.1 Methods
Seventy seven patients, from an overall gfoup of 365 patients who underwent a
laparoscopic Nissen fundoplication at the Royal Adelaide Hospital between September
1991 and October 1995, completed an illness behaviour questionnaire (IBQ) before
surgery. Only patients undergoing preoperative oesophageal motility studies performed
by one particular research technician were requested to complete the preoperative
questionnaire. Demographic and clinical factors for this subgroup were compared with
the overall surgical population to ensure that the sample was representative.
In January 1996, all patients who had completed the IBQ before laparoscopic Nissen
fundoplication, and who had been followed post-operatively for at least 3 months were
asked to complete an identical questionnaire. This was sent by mail, with an
accompanying letter providing instructions for completion of the IBQ. The results of the
IBQ were then compared to clinical outcomes obtained independently by a scientist who
utilised a standardised follow-up protocol. The post-operative outcomes of patient
satisfaction, heartburn and dysphagia for solid food were examined.
1.2.6.4.1-l lllness Behaviour Questionnaire
The Illness Behaviour Questionnaire, designed and validated by Pilowsky and
Spencs174,175, is used to provide an assessment of psychosocial status. It consists of 62
yes-no items which elicit various dimensions of illness behaviour. Patient responses from
the IBQ were assessed using a validated scoring method. This generates seven first order
and three second order factors. For each factor, a score on an ordinal scale is obtained,
with higher scores usually representing less desirable psychosocial states or illness
behaviour. t12
First Order Factors
The first order factors obtained from the IBQ can be categorised as being either general (4 scales) or illness-related (3 scales). The general factors are: a) Affective inhibition (range 0-5). This is characterised by difficulty conveying
personal feelings, especially negative ones, to others. b) Affective disturbance (range 0-5). This refers to patients' levels of anxiety and/or
sadness c) Denial (range 0-5). This is concerned with the extent to which patients are willing
to acknowledge the existence of problems other than illness. d) Irritability (range 0-6). This provides an assessment of patients' levels of anger
and inter-personal friction.
The illness-related psychosocial factors are:
e) General hypochondriasis (range 0-9). This refers to the level of phobic concern
about one's health. It is characterised by a level of arousal or anxiety, which is, to
some extent, recognised by patients as being excessive- Ð Disease conviction (range 0-6). This is characterised by the belief that physical
disease exists, symptom pre-occupation, and rejection of doctors' reassurance.
g) Psychological vs somatic perception of illness (range 0-5). A high score indicates
that the patient feels somehow responsible for his illness, and perceives a need for
psychiatric rather than medical treatment'
Second Order Factors
From patients' responses to the IBQ, the Whitely Index of HypochondriasislT3 and two
other second order factors can be determined. These factors enable an appreciation of
more global aspects of illness behaviour. h) Whitetey Index of Hypochondriasis (range 0-14). A high score indicates a
hypochondriacal response to illness. Ð Affective state (range 0-20). This factor is the sum of the scores for the following first order factors: general hypochondriasis, affective disturbance and irritability. 113 j) Disease affirmation (range 0- 11). This factor is derived from the scores for the
following first order factors: disease conviction and psychological vs somatic
perception.
1.2.6.4.1.2 ClinicalFollow-uP
Clinical outcomes following surgery were determined from the application of the
standardised follow-up questionnaire. For this study, the most recent outcomes following
laparoscopic Nissen fundoplication were scored. The following visual analogue scales
were used; satisfaction with the surgical outcome (completely satisfied = 10, totally
unsatisfied = 0), severity of heartburn (completely relieved of heartburn = 0, severe heartburn - 10), and the severity of dysphagia when swallowing solid foods (no
dysphagia = 0, total dysphagia = 10).
1.2.6.4.1.3 StatisticalAnalysis
Spearman's correlation coefficient was used to assess correlations between individual
clinical outcomes and illness behaviour scores. Changes in illness behaviour scores
before and after laparoscopic fundoplication were assessed using the Wilcoxon rank sum
test. Fisher's exact test for categorical data and an unpaired Student's t-test for
continuous variables were used to compare the demography of the study sample with the
overall patient population undergoing laparoscopic antireflux suf gery.
1.2.6.4.2 Results
The77 patients completing a pre-operative IBQ comprised 57 men and20 women, with a
mean age of 47 years (95Vo confidence intervals [CI] 44,50), a mean weight of 79.0 kg
(CI's 75.7, 82.3), and a mean duration of reflux symptoms of 9.6 years (CI's 7.3, I2.0).
The study group's demography and clinical outcome scores were representative of the
overall population of patients undergoing laparoscopic antireflux surgery, with the
exception of a higher proportion of male patients (Table l-2-6-9). lt4
Table 1.2.6.9 Comparison of sample group (n=77) with other patients undergoing lap aroscopic fundoplication (n=262).
All figures are means (95Vo conftdence intelvals). Sample group Other patients P value
Age (years) 46.8 (43.6, 50.0) 47 -4 (45-7,49-l) 0'49 Male:Female 0.74 (0.63,0.83) 0.58 (0-52' 0'64) 0'011
Weight (kgs) 79.0 (7 5.7 , 82.3) 79.9 (77 .9, 82.0) 0'65
Duration symptoms (yrs) 9.6 (7.3,12.0) 8.0 (7.0' 9.0) 0'54
Postop satisfaction score 8.42 (7 .87 ,8.96) 8-26 (7 -80,8.72) 0'87
Postop heartburn score 0.66 (0.28, 1.04) 0.65 (0.41,0.89) 0'69
Postop dysphagia score 1.80 (1.20, 2.40) l-80 (1.37 ,2-23) 0'75
Patients completed the IBQ an average 2.1 months (CI's 1.5, 2.8) before laparoscopic
surgery. Of the original TT patients,53 (69Vo) completed a further questionnaire at a
mean 2.1 years (CI's 1.8, 2.5) posroperatively. As clinical follow-up was obtained
separately to the IBQ, the average follow-up time was 1.8 years (CI's 1.6, 2.0) following
surgery. Six of the 24 patients not completing a postoperative IBQ were lost to routine
clinical follow-up. Five of these had participated in the standardised clinical follow-up
protocol 3 months after surgery, whilst one had been followed for 2 years. A further
patient died2 years after fundoplication from metastatic colon cancer. Seventeen patients
continue to be followed clinically but did not complete the post-operative questionnaire.
The demography and clinical outcomes for the patient group completing the post-
operative IBQ were similar to the smaller group not completing the questionnaire for; age
sex, preoperative weight, preoperative symptom duration, and postoperative heartburn and
dysphagia scores. However, the group completing a postoperative IBQ had a higher mean
satisfaction score following surgery (9.0; CI's 8.6, 9.4) than the other group (7'2; CI's 5.8,
8.7; P= 0.03) 115
Influence of laparoscopic fundoplication on illness behaviour 'l'he Table 1.2.6.10 summarises the IBQ scores before and after surgery. scores for
disease conviction and disease affirmation were significantly lower, following
laparoscopic antireflux surgery. No other illness behaviour characteristics changed
following surgery.
Table 1.2.6.L0 Illness behaviour factors before and after laparoscopic antireflux
surgery.
All figures are means (95Vo confidence intervals). Factor Preoperative score Postoperative score P value
General hypochondriasis 1.17 (0.80, 1.54) 1.20 (0.ó6, 1.75) 0.77
Disease conviction 2.43 (2.08,2.19) r.92 (1.49,2.34) 0.0014
Psychological vs somatic 1.46 (L.27, l'66) t.72 (1.43,2.00) 0.13 perception of illness
Affective inhibition 2.74 (2.34,3'13) 2.82 (2.36,3.29) 0.91
Affective disturbance 2.01 (7.59,2.44) 2.U (1.49,2.59) 0.52
Denial 3.44 (3.11, 3.78) 3.14 (2.71,3.57) 0.49 Irritability 1.48 (1.14, 1'84) t.57 (1.11,2.o4) 0.75
Whiteley index 3.56 (2.90,4.27) 294 (2.08,3.79) 0.062
Affective state 4.73 (3.75,5.7I) 5.02 (3.70,6.35) 0.55
Disease affirmation 5.93 (5.48, 6.38) 5.07 (4.50,5.63) 0.0002
Influence of pre-operative illness behaviour on ntrgical outcome
Table I.2.6.11 summarises the correlation between the preoperative illness behaviour
factors and the assessed clinical outcomes. Poor satisfaction was more likely in patients
with high preoperative scores for general hypochondriasis (P=0.0004), affective
disturbance (P=0.0014), the'Whiteley index (P=0.0013), and affective state (P=0.0004).
No signifîcant differences were demonstrated for the other psychosocial factors- No
statistically significant differences were seen between any of the IBQ scores and either the
postoperative heartburn or solid food dysphagia scores. 116
Table 1.2.6.11 Pre-operatively assessed illness behaviour versus post'operative clinical outcome.
All figures are Spearman's correlation coefficient (r) and 957o confidence intervals.
* = P<0.05 Factor Satisfaction score Dysphagia score Heartburn score
General hypochondri asis -0.42/" (_0.60, _0.20) 0.21 (-0.033,0.43) 0.18 (-0.066,0.40)
Disease conviction -0.14 (-0.37,0.10) -0.13 (-0.36,0.12) 0.17 (-0.071,0.40)
Psychological vs somatic 0.14 (-0.11,0.37) -0.098 (-0.33,0.15) -0.15 (-0.38,0.091) perception of illness
Affective inhibition -0.12 (-0.36,O.12) 0.22 (-0.019,0.44) -o.23 (-O.45,0.01)
Affective disturbance -o3g'ß (-0.58, -0.15) o.oo9 (-0.24,0.25) 0.19 (-0.053,0.42)
Denial -0.014 (-0.26,O.23) -0.15 (-0.38,0.098) 0.0098 (-o.23,o.25)
Irritability -0.12 (-0.35,O.12) -0.024 (-0.26,0.22) o.Mz (-0.20, 0.28)
Whiteley index -0.38* (_0.59, _0.15) 0.19 (-0.062,0.42) 0.22 (-O.O30,0.44)
Affective state -0.44+ G0.62,-0.20) 0.0098 (-0.25,0.26) 0.19 (-0.069,0.42)
Disease affirmation -0.17 (-0.40,0.084) 0.084 (-0.32,0.17) 0.19 (-0.053,0.42)
Correlation between post-operative illness behaviour and surgical outcome
Table 1.2.6.I2 summarises correlations between the illness behaviour factors determined post-operatively and the assessed clinical outcomes. Only the 53 patients who completed the IBQ post-operatively were included in this analysis. rn
Table 1.2.6.12 Post-operatively assessed illness behaviour versus post'operative clinical outcome.
All figures are Spearman's correlation coefficient (r) and 957o confidence intervals.
* = P<0.05 Factor Satisfaction score Dysphagia score Heartburn score
General hypochondriasi s -035* (-0.58, -0.065) 0.065 (-0.23,0.35) 0.062 (-0.23,0.35)
Disease conviction -034* (-0.58, -0.054) 0.13 (-0.17,0.40) o.4g+ (0.23,0.68)
Psychological vs somatic -0.045 (-0.34,0.26) -0.12 (-0.40,0.19) -0.30* (-0.55,0.004) perception of illness
Affective inhibition -0.42* (-0.ó3, -0.15) 0.20 (-0.087, 0.46) 0.16 (-0.13,0.42)
Affective disturbance -0.56* çO.73, -0.32) 0.16 (-0.14,0.43) o.27 (-0.023,O.52)
Denial 0.22 (-0.068,0.48) -0.029 (-0.31, 0.26) -0.21 (-0.46, 0.080)
Irritability o.23 (-0.49,0.075) 0.03t (-0.27,0.32) 0.13 (-0.17,0.41)
V/hiteley index -035* (-0.59, -0,064) 0.085 (-0.21 ,0.37) 0.25 (-0.048,0.50)
Affective state -0.49* (-0.69, -0.21) 0.083 (-0.23,0.0.38) 0.18 (-0.13,0.46)
Disease affirmation -0.15 (-0.43,0.77) 0.070 (-o.24, 0.36) 0.42* (0.13,0.64)
Poor satisfaction was more likely in patients with high post-operative scores for general hypochondriasis (P=0.015), disease conviction (P=0.018), affective inhibition (P=0.0027), affective disrurbance (P<0.0001), the Whiteley index (P=0.015), and affective state
(P=0.0009). No statistically significant correlations were seen between any of the IBQ scores and postoperative solid food dysphagia. The perception of heafiburn in the post- operative period did, however, correlate with high scores for disease conviction
(p=0.0005) and disease affirmation (P=0.0045). A negative correlation was associated with psychological versus somatic perception of illness (P=0.046), which indicates that patients perceived any postoperative heartburn as a physical rather than psychological problem. 118
1.2.6.4.3 Díscussion
Previous studies have demonstrated the importance of psychosocial t'actors in an individual's perception of symptoms of gastro-oesophageal reflux disease, confîrming that physical factors alone fail to account for all features of this 6its¿5s41,94. Othet studies have revealed that psychosocial factors can affect a patient's perception of the outcome of certain surgical procedureslT6. 1¡is study reveals that psychosocial factors and illness behaviour also influence the clinical outcome following laparoscopic antireflux surgery.
Whilst the illness behaviour questionnaire applied in this study was originally designed to quantify abnormal illness behaviour, its application has provided interesting insights into the post-operative outcomes of anti-reflux surgery. The results suggest that patients exhibiting abnormal preoperative affective or hypochondriacal illness behaviour scores were less likely to be satisfied with their outcome following laparoscopic antireflux surgery. However, preoperative illness behaviour did not predict the therapeutic outcome of relief of heartburn, or the adverse outcome of solid food dysphagia. The results of the postoperative scores were similar to those seen preoperatively, with the addition of an association between disease conviction and affirmation and the presence of any postoperative heartburn symptoms. It should be realised, however, that all patients in the study group had significant relief of heartburn following surgery. The complaint of dysphagia was not associated with any illness behaviour factor.
Perhaps not surprisingly, illness behaviour scores were stable following surgery, with the exception of disease conviction and disease affirmation. The scores for these two scales improved, in keeping with the overall improvement in reflux related symptomatology which should follow laparoscopic fundoplication. These results are similar to the findings
of Pilowsk y et aIlT 6 which compared the clinical outcome of coronary artÊ,ry by-pass
surgery, with the illness behaviour scores before and after surgery.
It is acknowledged that only 69Vo of the original study group completed the post-operative
IBQ, and that this limits one's ability to draw some conclusions from this study. Bias 119 could be introduced if mean scores for the individual pre and post-operative IBQ scales are compared with clinical outcomes. However, it is believed that the analysis of data with correlation coefficients reduces the likelihood of this problem impacting on the study results.
There are several possible explanations for the influence of illness behaviour on the outcome of laparoscopic fundoplication. One possibility is that these factors have the ability to induce changes in physiology, possibly mediated by neuropsychoimmunological or psychoendocrinological mechunl**r199. This is probably unlikely. Alternatively, illness behaviour may exert an effect by altering the extent to which patients perceive and report the severity of their symptoms102.
This study has demonstrated in an objective way what surgeons see in their clinical practice every day; i.e. with a seemingly similar outcome as determined by objective measures, some patients are happier with their result than others. It highlights the need to be aware of psychological factors which may adversely influence the outcome of laparoscopic fundoplication, as well as the inability of surgical therapy to significantly alter many of the psychosocial indicators of health.
L.2.7 CONCLUSIONS
The prospective analysis of outcome following laparoscopic Nissen fundoplication
described in this section has confirmed the feasibility of the laparoscopic approach for the
surgical correction of reflux disease. However, initial experience revealed that this
procedure may be associated with complications, and that conversion from the
laparoscopic to an open approach is required in technically difficult situations.
Nevertheless, some advantages are apparent, with splenectomy rrare event following
laparoscopic fundoplication and respiratory morbidity unusual. The specific
complications of pneumothorax, early and late postoperative para-oesophageal hiatus
herniation, and stenosis of the oesophageal hiatus were identified. Whether the overall t20 incidence of complications is reduced by the laparoscopic approach cannot be determined from this analysis.
A learning curve for laparoscopic fundoplication was found to be up to 20 procedures for each individual surgeon, and 50 procedures for the institutional experience. During this experience the incidence of complications, surgical revision, and the conversion rate were all higher. Operating times were prolonged. However, results did improve with experience. This has confirmed the need to eliminate the leaming curve to determine the true outcome following laparoscopic surgery, and to eliminate its influence before embarking on clinical trials.
Perhaps surprisingly, these results suggest that the various preoperative parameters proposed by some surgeons to be important for tailoring surgery to enable the risk of adverse clinical outcomes to be reduced, failed to significantly influence outcome.
Preoperative oesophagitis grading and manometric findings do not appear to be important outcome predictors. The only criæria measurable before surgery which influenced outcome was illness behaviour.
The studies in sections 2 and 3 of this thesis investigate the questions which have not been answered by this analysis:
Is the laparoscopic approach to reflux surgery really better?
Can outcomes be improved by dividing the short gastric vessels, or by performing a partial fundoplic ation? 12T t22
sEcfloN 2
LABORATORY RESEARCH 123
2.7 COMPARISON OF ANTERIOR, POSTERIOR AND TOTAL FUNDOPLICATION USING A VISCERA MODEL
(Published previously as; Watson DI, Mathew G, Pike GK & Jamieson GG.
Comparison of anterior, posterior and total fundoplication using a viscera model.
Diseases of the Esophagus 1997;10: 110-11a223¡
z.I.L AIM
To determine the contribution which mechanical factors make to different fundoplication
variants' ability to restore an anti reflux banier.
2.I.2 MATERIALS AND METHODS
Ten fresh specimens of the stomach and oesophagus from the domestic white pig were
sourced from a local abattoir. Each stomach was emptied of its contents and washed
through an anterior gastrotomy, which was subsequently closed water tight with sutures.
The diaphragmatic hiatal structures and all other supporting attachments to the Sastro-
oesophageal junction, were dissected away from the junction, before laying the organs in
an anatomically correct position on a flat tray. The oesophagus was fixed proximally to
the edge of the tray so that it lay straight , loosely without tension, whilst the gastro-
oesophageal junction and stomach remained untethered.
Each set of viscera underwent 5 consecutive studies. To prevent any potential for
repeated gastric distension stretching the gastric wall and thereby biasing gastric volume
outcomes, the order in which the different types of fundoplication were performed was
varied during the study. For each study the following parameters were measured once:
. intraluminal pressure at the gastro-oesophageal junction t24
. integrity of the gastro-oesophageal junction / lower oesophageal sphincter zone
during gastric distension
. intra-gastric pressure at which the sphincter zone became incompetent to intra-
gastric fluid
. volume of intragastric fluid infused.
A side hole comprising 30 to 40Vo of the oesophageal wall circumference was made in the oesophagus, 10 cm above the gastro-oesophageal junction. This was used for the passage of a manometry catheter, and to allow visual assessment of any leakage of intragastric fluid. A single channel fine bore water perfused manometry catheter was passed into the gastric lumen through a small incision in the anterior gastric wall (Figure
2.1.1). A purse string suture was used to create a water tight seal around the catheter,
which rwas connected to a pressure transducer (Stratham P23lD, Gould Inc, Oxnard,
California) to continuously measure intragastric pressure. The catheter was constantþ
perfused with degassed distilled water at 0.5 mVmin by a low compliance
pneumohydraulic pump (Arndorfer Medical Specialities, Greendale,'Wisconsin), and the
signals were recorded on a polygraph chaft recorder (Grass Instrument Company, Model
7D, Massachusetts, USA). A24Fr Foley catheter was also passed through a hole in the
anterior gastric wall, the balloon inflated, and a purse string suture used to seal around it.
This tube was used for the instillation of fluid into the stomach.
An 8-lumen water perfused catheter incorporating a sleeve sensor (Dent Sleeve,
Adelaide) was passed through the oesophageal side hole and then distally into the
stomach. The catheter was connected in series with a pressure transducer (Stratham
P23lD), constantly perfused with degassed distilled water at 0.5 mVmin by a low
compliance pneumohydraulic pump (Arndorfer Medical Specialities), and connected to a
polygraph recorder. The'lower oesophageal sphincter'pressure was determined by
gradually withdrawing the catheter and noting pressure changes as each sensor passed
through the gastro-oesophageal junction, and the distal oesophagus. Competence of the
gastro-oesophageal junction was determined by infusing water coloured with a blue dye
into the gastric lumen through the Foley catheter. Infusion continued until blue liquid 125 appeared through the side hole in the oesophagus, or began to leak through suture holes in the gastric wall, or the gastric wall tension increased to the extent that damage to the gastric wall was thought to be possible if the experiment continued. The volume of intragastric fluid infused and the concurrent intragastric pressure were recorded. All surgery during the following five study protocols was performed by a surgeon experienced in anti reflux surgery (the author).
A
B
c
Figure 2.1.1 Preparation of porcine oesophagogastric specimen
A = oesoPhageal manometry catheter
B = Foley catheter for gastric infusion C - gastric manometry catheter 126
2.1.2.1 Baseline study
The baseline parameters described were assessed betbre any operative procedure.
2.1.2.2 Parameters following oesophageal myotomy
A distal oesophageal myotomy was performed and extended at least 1 cm onto the wall of the stomach, to completely divide the lower oesophageal sphincter. The lower oesophageal sphincter pressure was then re-measured and its competence reassessed.
The myotomy was assessed to be adequate if the sphincter zone was totally incompetent to the infusion of fluid into the stomach. If this was not achieved, then the myotomy was extended and the study repeated.
2.I.2.3 Parametersfollowinganteriorhemi-fundoplication
An anterior fundoplication was created by suturing the antero-medial gastric fundus to the
right side of the distal oesophagus. Two sutures separated by a distance of 1 to 1.5 cm
were placed in the oesophageal wall at the 9 o'clock position (with the oesophagus
visualised in cross section from below). This fundoplication was not supported by any
surrounding hiatal structures. In particular, no sutures were placed to any hiatal muscle,
and no attempt was made to stabilise the anterior fundoplication in any way apart from
the sutures described. The intra-oesophageal pressure generated by the fundoplication
was measured and competence of the fundoplication assessed.
2.L.2.4 Parameters following posterior hemi-fundoplication
A posterior fundoplication was created by suturing the medial fundus to the oesophageal
wall with 2 rows of sutures placed at the 10 o'clock and 2 o'clock positions on the
oesophageal wall. A 1 cm strip of bare anterior oesophagus including the myotomised
section remained visible. Four sutures were used, 2 on each side, and spaced 1 to 1.5 cm
apart. The intra-oesophageal pressure generated by the fundoplication was measured and
competence of the fundoplication assessed. 127
2.1.2.5 Parameters following total fundoplication
A 1 to 1.5 cm long total fundoplication secure d 6y 2 sutures incorporating the 2 pieces of stomach, but not the oesophagus, was fashioned. This fundoplication was fashioned loosely, so that the surgeon's finger could easily slide between the oesophagus and the wrap. The intra-oesophageal pressure generated by the fundoplication was re-measured
and competence of the fundoplication reassessed. During each phase of the study, care was taken not to over distend and thereby damage the gastric wall, when assessing the
competence of each variant of fundoplication.
2.1.2.6 Statistics
Sets of data were analysed using a one way analysis of variance (ANOVA), which
compared results following anterior, posterior and total fundoplication. All data is
represented as median (range).
2.I.3 RESULTS
Results for the study are summarised in Table 2.1.1. The gastro-oesophageal junction
was competent to gastric distension in 3 sets of viscera prior to oesophageal myotomy.
Subsequent myotomy resulted in complete abolition of any high pressure zone at the
gastro-oesophageal junction, and incompetence to the infusion of small volumes of
intragastric fluid in all instances. r28
Table 2.1.1 Study outcomes
LOSP Volwne ínfused Intragastric pressure Competent GOJ (mmHg) (mmHg)
Baseline 10 (0-20) 400 (180-840) 4 (0-14) 5lt0 (507")
Post myotomy 0 (0-3) r3s (30-240) 0 (0-10) 0lt0 (oEo)
Anterior fu ndoplication 11 (4-16) 930 (360-1200) 8 (2-20) 6110 (60v")
Posterior fundoplication 13.s (5-20) 840 (300-1200) 7 (2-16) 6lt0 (607")
Total fundoplication t2 (6-3s) 1200 ('720-7920) 15 (10-40) t0ll0 (100E")
LOSP = intraluminal pressure measured at gastro-oesophageal junction Competent GOJ = no leakage of intragastric fluid through gastro-oesophageal junction
All types of fundoplication generated a significant rise in the intraluminal pressure
measured within the oesophagus (P<0.0001). This pressure change was independent of
the type of fundoplication performed (P=0.45, anterior vs posterior vs total). Six anterior
fundoplications, 6 posterior, and all 10 total fundoplications remained fully competent to
intra-gastric infusion of between 700 to 1920 mls of water. These volumes varied due to
variability of gastric volume between different specimens. In studies where the gastro-
oesophageal junction yielded to infusion of fluid, the volume ranged from 300 to 840
mls. In each instance this volume was greatly in excess of that infused during the initial
postmyotomy study with competence significantly improved by both anterior and
posterior fundoplication.
Following total fundoplication the volume infused and the maximum intragastric pressure
achieved were significantly greater than following either anterior or posterior
fundoplication (P=0.011 and P=0.0051 respectively), due to the greater intrinsic
competence of this type of fundoplication. All total fundoplications were competent to
gastric distension, irrespective of the infusion volume. In each study gastric infusion
was stopped when the stomach was thought to be either at risk of rupturing, or fluid
began leaking through suture holes. Competence was retained by all total
fundoplications despite substantial intra- gastric pressure. r29
Visual assessment of the patterns of gastric distension during the int-usion of water into the stomach, confîrmed that the portion of stomach adjacent to the oesophagus which comprised the fundoplication, always distended first. As it expanded it compressed and flattened the adjacent oesophagus to restore competence to the gastro-oesophageal junction. In doing this the fundoplication appeared to act as a flutter valve. Following
anterior fundoplication, the compression was generated entirely by the portion of stomach lying in front of the oesophagus, whereas following both posterior and total fundoplication distension occurred predominantly in the portion of fundus lying behind
the oesophagus. As this expanded it lifted and angulated the oesophagus forwards.
2.I.4 DISCUSSION
Whitst it is not intended to extrapolate the results obtained using a bench top viscera
model directly to the clinical situation, this study does provide information about certain
mechanical functions of different types of fundoplication. Because of the absence of
muscle tone and supporting hiatal structures, only intrinsic mechanical factors were
examined by this model. However, preliminary oesophageal myotomy was necessary to
eliminate residual resistance at the gastro-oesophageal junction, to ensure free reflux in
all instances. Earlier bench top studies have examined the mechanisms of lower
oesophageal sphincter competen ce47 ,I79 and confirmed the efficacy of the Nissen
fundoplication14. In 1971 Butterfield14, using human autopsy specimens, confirmed the
intrinsic competence of the Nissen, Belsey, Hill and Thal repairs in intact gastro-
oesophageal specimens, as well as specimens in which the gastro-oesophageal junction
had been excised. The current study examined a different spectrum of antireflux
procedures, used oesophageal myotomy rather than sphincter excision to ensure reflux,
and included oesophageal manometric measurements.
Previous clinical and experimental studies using anaesthetised animals have attempted to
understand the mechanism of action of antireflux operations. These studies have 130 demonstrated that the total fundoplication results in a rise in resting pressure at the cardio-oesophageal junction121,135, and a decreased ability of the lower oesophageal sphincter to relax with swallowingl2l. Accelerated gastric emptying following fundoplication was demonstrated by Maddern130, suggesting that other factors contribute to the antireflux mechanism. The finding by Matikainenl33 that as intragastric pressure increases, the fundoplication tightens, increasing intraluminal lower oesophageal pressure, is consistent with the observations of the effect of gastric distension in viscera in this study. This supports the concept of a fundoplication acting as a flutter valve, explaining why a total fundoplication can still be effective even if constructed loosely, even if it does not lead to a rise in resting lower oesophageal sphincter pressure.
Other publications have suggested that several concurrent manoeuvres are essential for a successful antireflux operation. These include establishing a length of intra-abdominal oesophagu s52,211, hiutul rcpair52,I47, creation of a fundoplis¿1isrì179 and accentuation of the angle 6¡ 1¡¡193. It is also claimed that gastric wall muscle tone and motor function is important for fundoplication function. Indeed Iwai et al85 concluded that gastrin stimulation following fundoplication in a dog model resulted in a significantly higher lower oesophageal sphincter pressure, suggesting that a fundoplication creates a new muscle sphincter substitute which reacts to stimuli such as gastrin. However, by eliminating the contribution of gastric wall muscle tone, and the support of adjacent structures such as the hiatal muscle sling, this study confirms that mechanical factors alone, acting by either a flutter valve effect or by an effect of the LaPlace law, can still act quite effectively to prevent reflux associated with gastric distension. This is further supported by the observation that the 'lower oesophageal sphincær pressure' generated by each type of fundoplication in the bench top viscera is similar to that reported in post- operative human 5tu¿ie543,209,2I1,212.
The total fundoplication was totally competent in all studies. Both the anterior and posterior fundoplications also restored competence to the gastro-oesophageal junction, although four anterior and four posterior fundoplications eventually allowed reflux of 131 fluid following the instillation of large volumes under pressure. It is important, however, to note that the anterior fundoplication was not supported by sutures to any adjacent hiatal structures, as is usually advocated by proponents of this type of antireflux
procedure2l1. Despite this the anterior fundoplication still provided a competent
antireflux valve. It should be noted the volumes infused into the stomach in this study
exceeded normal physiological stresses, and the degree of competence achieved by all 3
variants of fundoplication used here would almost certainly be adequate to prevent reflux
in the clinical situation. The total fundoplication plovides an effective but hyper-
competent anti-reflux barrier. The inability of any total fundoplications to yield to
intragastric distension, despite their loose construction may explain why patients often
are unable to belch or vomit following this procedute43. This hyper-competence
probably contributes to the potential adverse outcomes of dysphagia and gas bloat.
This study suggests that mechanical factors could be major contributors to the ability of a
fundoplication to restore gastro-oesophageal competence. Anterior, posterior and total
fundoplications are all effective. However, it is possible that the partial fundoplications,
by not creating hypercompetence, may reduce the risk of dysphagia and bloat. t32
2.2 EFFICACY OF ANTERIOR, POSTERIOR AND TOTAL FUNDOPLICATION IN A PORCINE MODEL
'Watson (Accepted for publication as; DI, Mathew G, Pike GK, Baigrie RJ & Jamieson
GG. Efficacy of anterior, posterior and total fundoplication in a pig model. Br J Surg
(199S) -1nPtsss222¡.
2.2.1 ArM
To compare, using a porcine model, the adequacy of the antireflux mechanism produced
by the Nissen fundoplication against that produced by two partial fundoplication variants
(anterior and posterior), with all operations performed laparoscopically.
By comparing the results of this study with outcomes assessed in the pig viscera study
(Section 2.1), information was sought which will generate a better understanding of how
mechanical and other proposed antireflux mechanisms contribute to the production of an
antireflux barrier following fundoplication.
2.2.2 MATERIALS AND METHODS
Fifteen domestic white pigs underwent laparoscopic antireflux surgery and post-operative
evaluation according to the study protocol summarised in Figure 2.2-1. The pigs were
randomly allocated to each study group. Oesophageal myotomy and fundoplication were
performed laparoscopically, followed by gastrostomy and oesophagostomy placement.
Manometric evaluation was undertaken as detailed subsequently. Five pigs underwent
an anterior, 5 a posterior and 5 a total fundoplication. All operations were performed by
a surgeon experienced in laparoscopic anti-reflux surgery. Subsequent oesophageal
manometry and fundoplication 'yield' studies were perfotmed 2 weeks after surgery in all
animals. Several animals also had 4 and 6 week post-operative studies. 133
Figure2.2.l Study plan
Week 0 Oesoph agostomy Plac ement
Oesophageal Manometry
Oesophageal Myotomy
Oesophageal Manometry
Anterior Posterior Nissen Fundoplication Fundoplication Fundoplication
Gastrostomy placement
Oesophageal Manometry
Week 2 Manometry and Yield studies
Week 4 Manometry and Yield studies
Week 6 Manometry and Yield studies
2.2.2.L General Operative Procedure
General anaesthesia was commenced following induction with intramuscular ketamine and an endotracheal tube was positioned for inhalational halothane anaesthesia. Access to the oesophageal lumen was achieved by the preliminary placement of an oesophagostomy cannula, with exposure of the cervical oesophagus obtained through a 134 left neck incision. A purpose built 1 cm diameter T-shaped plastic cannula with 2 cm
long arms was used. The arms were sited within the lumen of the cervical oesophagus,
and the oesophageal wall defect was closed around the cannula with a purse string suture.
The intra-oesophageal portion of the cannula formed an open half circle when viewed in
cross-section, achieving direct access to the oesophageal lumen, without obstructing the
passage of swallowed liquids or solids.
Preliminary oesophageal manometry was performed next to determine the baseline lower
oesophageal sphincter pressure. An 8-lumen water perfused manometry catheter
incorporating a sleeve sensor (Dent Sleeve, Adelaide, South Australia) was passed
through the oesophagostomy and then distally into the stomach. The catheter was
connected in series with a pressure transducer (Stratham P23lD, Gould Inc, Oxnard,
California), which was constantly perfused with degassed distilled water at 0.5 mUmin by
a low compliance pneumohydraulic pump (Arndorfer Medical Specialities, Greendale,
Wisconsin), and it was connected to a polygraph recorder (Grass Instrument Company,
Model7D, Massachusetts, USA). The position of the lower oesophageal sphincter was
deærmined by gradually withdrawing the catheter and noting pressure changes as each
sensor passed through the gastro-oesophagealjunction, and the distal oesophagus. The
catheter was then positioned so that the mid-point of the sleeve sensor was sited at the
midpoint of the high pressure zone. Resting lower oesophageal sphincter pressure was
measured over a 5 minute period using the sleeve sensor. To eliminate the effect of
intra-abdominal pressure, the basal end expiration intragastric pressure was subtracted
from that of the lower sphincter pressure to determine the resting lower oesophageal
sphincter pressure.
Laparoscopy commenced with insertion of a Veress needle into the abdomen
approximately 2 cm above the umbilicus, to facilitate induction of a carbon dioxide
pneumoperitoneum. Four laparoscopic ports were placed; an 11 mm port at the site of
Veress needle insertion, an 11 mm port in the left upper quadrant, and 5 mm ports in the
epigastrium and left flank. A purpose designed 4 mm diameter rigid curved retracting 135 instrument was pushed through the abdominal wall of the right upper quadrant, to facilitate liver retraction and to elevate the oesophagus following initial mobilisation.
The oesophageal hiatus was dissected with laparoscopic scissors, exposing both hiatal
pillars and the oesophageal wall. Exposure was facilitated by tilting the operating table
450 laterally to the right, a manoeuvre which often eliminated the need for retraction of
the pig's liver to display the oesophageal hiatus. Diathermy dissection was not used at
any stage of the procedure.
A distal oesophageal myotomy was performed next using laparoscopic scissors. The
myotomy vi/as extended approximately 1 cm onto the anterior wall of the stomach and 5
cm proximally, to achieve complete division of the lower oesophageal sphincter. The
pneumoperitoneum was then released and the lower oesophageal sphincter pressure was
re-evaluated manometrically, to confirm that the lower oesophageal sphincter had been
completely ablated. This both mimicked the common situation in humans where lower
oesophageal sphincter pressure is low and also made it likely that the subsequent
fundoplication was responsible for any antireflux banier tested during post-operative
studies. The pneumoperitoneum was then re-established and the hiatus was repaired
posteriorly using I or 2 interrupted 2/0 polypropylene sutures.
2.2.2.2 Anterior fundoplication
An anterior fundoplication was fashioned by suturing a portion of the antero-medial
gastric fundus loosely to the right lateral wall of the distal oesophagus and concurrentþ
to the posterior hiatal repair or right hiatal pillar. This was achieved by the placement of
4 or 5 interrupted 2i0 polypropylene sutures (Prolene). The repair anchored a length of
oesophagus within the abdomen, accentuated the angle of His, and fashioned a loose
anterior 1 800 fundoplication. 136
2.2.2.3 Posterior fundoplication
To create a posterior fundoplication, a piece of the antero-medial fundal wall was drawn loosely behind the oesophagus and sutured to the antero-medial oesophageal wall with three2l0 polypropylene sutures. A second row of three2l0 polypropylene sutures secured the anterior fundus to the antero-lateral oesophageal wall. This fashioned a2700 posterior fundoplication, leaving an approximately 1 cm wide strip of uncovered anterior oesophagus which included the myotomised section.
2.2.2.4 Total fundoplication
The Nissen fundoplication entailed a 1.5 to 2 cm long circumferential fundoplication secured by 3 sutures which each incorporated seromuscular sutures placed through the 2 pieces of the stomach, as well as the oesophagus. This fundoplication was fashioned loosely from the anterior wall of the fundus, and it was checked by passing an instrument between the oesophagus and the fundoplication at the completion of the procedure to ensure that the fundoplication was not tight.
The pneumoperitoneum \¡/as then released and the lower oesophageal sphincter pressure was re-evaluated manometrically. The left upper quadrant port wound incision was exænded to fashion a 4 cm long muscle splitting incision. The anterior gastric wall was delivered through this wound, and a 3 cm long, 1.5 cm diameter flanged gastrostomy cannula was placed through an anterior gastrotomy. The tube was secured using a 2 layered purse string suture, and the abdominal wall wound was closed around the cannula. This provided access to the gastric lumen for post-operative studies.
2.2.2.5 Postoperative care
Narcotic analgesia was administered for the fftst24 hours following surgery.
Intravenous fluids were maintained for the same time period. Oral fluids were given on the first post-operative morning, and the pigs commenced a soft diet the next day. r37
2.2.2.6 Postoperative studies
Post-operative studies were scheduled to take place at2,4 and 6 weeks after surgery.
Sedation with ketamine was necessary to prevent discomfort from passage of the oesophageal manometry catheter, and also to prevent discomfort due to gastric distension.
The manometry catheter was passed distally through the oesophagostomy cannula, and the fundoplication was located by the station pull through technique. The resting pressure generated by the fundoplication was then determined using the sleeve sensor in the same fashion as in the preoperative studies.
The catheter was then withdrawn and replaced by an 18 Fr nasogastric tube. The tip of this tube was sited 5 cm above the manometrically determined proximal point of the high pressure zone. A24Fr Foley catheter was passed into the gastric lumen through the gastrostomy cannula, and the catheter balloon was inflated to seal the gastrostomy.
Competence of the gastro-oesophageal junction was assessed by infusing water coloured with blue food dye into the gastric lumen through the Foley catheter. At the same time, the oesophageal tube was aspirated by a low pressure suction device. Infusion continued until either blue liquid appeared in the oesophagostomy aspirate, or the intra-gastric volume reached 2000 mls. The integrity of the gastro-oesophageal junction / lower oesophageal sphincter zone during progressive gastric distension was therefore deærmined and the volume of intragastric fluid infused was recorded. No tube or measuring device was placed across the gastro-oesophageal junction during this phase of the study.
2.2.2.7 Statistics
All data is presented as medians with 95Vo confr.dence intervals in brackets. Data were entered onto a computerised spread sheet program (Microsoft Excel Version 5.0) and analysed using a commercially available statistical package (InStat version 2.01,
GraphPad Software, San Diego California). Data sets were analysed using one way analysis of variance (ANOVA) to compare outcomes following anterior, posterior and total fundoplication. A P value of less than 0.05 was accepted as statistically significant. 138
2.2.2.8 Ethics
Animals were treated in accordance with the 'Australian Code of Practice for the Care and Use of Animals for Scientific Purposes', and ethical approval for this study was granted by the Animal Ethics Committees of the University of Adelaide and the Institute of Medical and Veterinary Science, Adelaide, South Australia.
2.2.3 RESULTS
Fifteen domestic white pigs underwent operation and subsequent post-operative studies.
All pigs underwent studies at 2 weeks post-operatively. Ten pigs were studied at all of the planned postoperative intervals during the six week study period. The remaining 5 pigs were killed by barbiturate overdose between 2 and 5 weeks following surgery to avoid suffering from problems they developed from either the gastric or oesophageal cannulae. Three of these pigs underwent a Nissen fundoplication, and2 underwent a posterior partial fundoplication. None of the problems encountered in these 5 animals was associated with the type of fundoplication performed. No difficulties with the intake of either oral fluids or solids were encountered in any animal.
The duration of surgery and the associated intraoperative studies were similar for each fundoplication variant (all 120 minutes median, P=0.93). The preoperative weights were similar for each group (anterior and Nissen fundoplication groups 20kg, posterior fundoplication group 22kg, P=0.90), as were the weights at the conclusion of the postoperative studies (anterior and Nissen 25.4kg, posterior 26.6,P=0.65). All animals resumed oral liquid and solid intakes at similar time intervals, irrespective of the fundoplication type, with liquids tolerated within 24 hours of surgery, and solids resumed a day later.
The manometrically measured pressures recorded in the lower oesophageal sphincter region are summarised in Table 2.2.I. Before oesophageal myotomy, a pressure ranging r39
from 4 to 50 mmHg was measured at the gastro-oesophageal junction. In all instances,
the subsequent distal oesophageal myotomy abolished this high pressure zone.
Subsequently, each fundoplication variant generated a significant rise in the intraluminal
pressure measured at the gastro-oesophageal junction. The post-fundoplication
intraoperative measurement was lowest in the anterior fundoplication group, although this
difference was not significant. The pressures recorded 2 weeks following surgery for
each fundoplication variant were similar. Similarly, when the lower oesophageal
sphincter pressures recorded during the last study performed on each animal were
analysed, the median pressures were similar , although a trend towards lower pressure
following Nissen fundoplication was seen. Transient lower oesophageal sphincter
relaxation events were not identified during any of the post-operative studies performed.
Table 2.2.1'lower oesophageal sphincter' pressure outcomes (mmHg)
Tíme of study Anteríor Posteríor Total (Nissen) P Preop 10 (-3, 30) 19 (6,33) 25 (1, 54) 0.33
Post myotomy 0 (0,0) 0 (0, 0) 0 (0, 0)
Postop 10 (8, 19) 14 (lI, 43) 14 (I3,2I) 0.067
Week 2 12 (9,I3) 14 (4,16) 10 (4, 16) 0.17 Final study 14 (9,20) 13 (7,22) 8 (2, 15) 0.13
All figures median (95Vo conftdence intervals)
Two weeks after surgery, all of the anterior and posterior fundoplications, and 4 of the 5
total fundoplications remained fully competent to the intra-gastric infusion of up to 2000
mls of water. The Nissen fundoplication which yielded to gastric distension became
incompetent following the infusion of 360 mls of fluid. At 4 weeks all of the Nissen (2
pigs) and posterior (3 pigs) fundoplications tested were fully competent. One the five
anterior fundoplications became incompetent following the infusion of 1,300 mls. The
fundoplication of 2 pigs undergoing anterior and one pig posterior fundoplication yielded
to the infusion of liquid 6 weeks after surgery, with the volumes infused at the point of
fundoplication yield ranging from 1200 to 1800 mls. 140
2.2.4 DISCUSSION
Both short and long term outcomes following Nissen fundoplication have been reported exûensively in the era of open surgery and have confirmed the status of the Nissen fundoplication as the gold standard for anti-reflux surgery. Long term success rates of approximately 90Vo can be expected following this procedrr"43, although most reports following both open and laparoscopic surgery confirm a small but significant incidence of troublesome post-operative dysphagia43,90,119. Also reported is an incidence of other adverse outcomes including gas bloat syndrome, an inability to belch or vomit, and recurrent gastro-oesophageal reflux.
Partial fundoplication has been proposed both for routine anti-reflux surgery and for patients with manometrically proven oesophageal dysmotility. Uncontrolled studies of patients undergoing posterior or anterior partial fundoplication during the laparoscopic and open eras have suggested that the incidence of dysphagia and other adverse outcomes may be reduced with equivalent control of reflux symptoms compared to total fundoplica¡isî9,2II. However, improvements in outcome have not been confirmed by prospective randomised trials119,209. One uncontrolled study suggests that higher resting lower oesophageal sphincter pressures occur following Nissen fundoplication compared to partial fundoplication9, and within the subset of patients undergoing Nissen fundoplication it has been postulated that adverse outcomes are more likely in patients with higher post-operative lower oesophageal sphincter ptes.u.".43. However, the fact that the anterior partial fundoplication is associated with a lower resting lower oesophageal sphincter pressure, yet still controls reflux, suggests that reflux control is not dependent on the measured post-operative resting sphincter pressure ulot-t"211.
Ismail and BancewiczS+ assessed the efficacy of the Nissen fundoplication in humans by measuring 'yield' pressure. These studies wele performed by siting an endoscope in the gastric lumen for gastric insufflation of air, whilst measuring intra-gastric pressure with a r4t manometry catheter. When the fundoplication was seen endoscopically to open and relax, the concurrent intra-gastric pressure was recorded. This pressure was designated the fundoplication 'yield' pressure. The authors suggest that the ability of a fundoplication to resist abnormal reflux is reflected by this yield pressure which depends on the fundoplication remaining competent during increasing gastric distension. The effect of the endoscope straddling the lower oesophagus, on fundoplication competence, was not measured.
The results of this study apply a different method for the assessment of fundoplication yield, assessing the influence of different fundoplication types using a model in which the lower oesophageal sphincter was completely ablated before performing an antireflux procedure. The intra-operative lower oesophageal sphincter pressures measured immediately after construction of the fundoplication were greater following both total and posterior fundoplication when compared to anterior partial fundoplication, although the failure to reach significance may reflect the numbers involved. However, these results are similar to those reported in previous open surgical randomised trials in humans.
These differences disappeared over the passage of several weeks. This study demonstrates that following gastric distension, the restoration of competence at the gastro-oesophageal junction over the shofi term is independent of fundoplication type.
These results are similar to the previous study using abattoir sourced porcine viscera
(section 2.D223. Resting lower oesophageal sphincter pressures in the ex vivo viscera model after various types of fundoplication varied from 11 to 13.5 mmHg and competence at the gastro-oesophageal junction was restored in most instances. The resting lower oesophageal sphincter prcssures were of a similar magnitude to those measured in the current study, suggesting that fundoplication induced manometric increases in the lower oesophageal sphincter zoîe may be due, at least in part, to passive mechanical factors. t42
A variety of mechanisms have been proposed to explain how antireflux surgery works.
These include changes in various mechanical parameters acting at the gastro-oesophageal junction due to alterations in anatomy following fundoplica¡ien116, as well as effects due
to hormonal stimulation of the intrinsic gastric muscle within the portion of stomach used
to construct a fundoplicationS5, and acceleration of gastric emptying130. These effects
may all lead to a reduction in the prevalence of transient lower oesophageal sphincter
relaxations, a phenomenon which has been demonstrated following Nissen
fundoplicationS3. The absence of transient lower oesophageal sphincter relaxations
85,116 in the current study also suggests that these are abolished by an effective
fundoplication.
When the lower sphincter is competent there appear to be two possible mechanisms by
which reflux can occur, associated with gastric distension. The first, and currently
believed to be the most important mechanism, is via reflex sphincter relaxation, initiated
from stretch fibres in the wall of the fundus83,86,234. This is the so called belch reflex.
The second mechanism is effacement of the lower sphincter brought about by distension
of the fundus. Effacement leads to shortening of the sphincter making it easier to force it
open as pressure increases within the stomac¡44,47,17I; i.e. the yield pressure of the
sphincter falls. However, there are practical difficulties in confirming this mechanism, as
it is difficult to make accurate length measurements of a very short sphincter, particularly
when its length is changing dynamically.
It is clear that either a partial or total fundoplication has the potential to greatly modify
both of these mechanisms. They do this by preventing distension of the stomach in the
region of the lower sphincter. The abolition of lower sphincter relaxationsS3, and the
increased yield pressures of the sphinctetS4 huue both been reported in humans after
fundoplication. However, neither of these mechanisms can be operative in ex vivo
models, and yet both types of fundoplication prevent reflux associated with gastric
distension in these This suggests that physical factors associated with ^o6¡.ls184,223. fundoplication may play as strong, or even a stronger role in reflux prevention, than 143 alterations in the physiology of the area. When the stomach in ex vivo preparations is filled with fluid, the fundoplication becomes distended before the rest of the t¡o uç¡223, with this distension impinging directly on the lower oesophagus. This also explains why such fundoplications seem to be effective when placed in the "¡-51150.
Restoration of an adequate length of intra-abdominal oesophagus, which maintains the gastro-oesophageal junction within the abdomen is thought by many to be impotant|J .
Both the Nissen fundoplication and partial fundoplications achievs¡¡ß43,2II. Straining may transmit the raised intra-abdominal pressure to the intra-abdominal oesophageal segment, thereby acting to push the oesophageal walls together to prcvent reflux. A hiatus hernia prevents this, resulting in less resistance to reflux of the gastric contents4T.
Also proposed is a mucosal flap valve or flutter valve116, which results from accentuation of the angle His. Butterfieldl4, usittg human autopsy specimens, demonstrated that the Nissen, Belsey, Hill and Thal repairs all restored competence to the gastro-oesophageal junction. When viewing these repairs from within, they all produced a flap valve across the opening of the oesophagus, with its free edge lying anteriorly in the Belsey and Thal repairs, and posteriorly in the Nissen and Hill repairs. The similarities in the post-fundoplication lower oesophageal sphincter pressures and the ability of each type of fundoplication to restore competence at the gastro-oesophageal junction seen in the current study and the previous bench top study (section 2-¡223 supports the proposal that each type of fundoplication creates a flap valve at the gastro- oesophageal junction, and that this is compressed against the junction by elevations in intra-abdominal and intragastric pressure, thereby preventing reflux. This mechanism is independent of intrinsic gastric wall muscle tone, hormonal stimulation, gastric emptying, hiatal repair and the position of the fundoplication within the abdomen or the thorax.
Nevertheless the current study is in some ways not a physiological model. The stresses placed on the fundoplication during the yield studies celtainly exceed the physiological situations which occur in humans, and the preliminary cardiomyotomy totally ablated any residual lower oesophageal sphincter tone. t44
This study does confirm that all three variants of fundoplication restore adequate competence to the gastro-oesophageal junction and result in similar elevations in post- operative lower oesophageal sphincter pressure. This provides suppoft for the uncontrolled clinical outcome studies following laparoscopic fundoplication which describe good results following each type of fundoplication, and for the previous randomised trials which demonstrated little difference between posterior partial fundoplication and total fundoplication179,209. As this study demonstrates that the laparoscopic anterior partial fundoplication provides an effective antireflux barrier, and uncontrolled studies suggest that it is clinically effective, further assessment of this procedure was commenced within a prospective randomised trial of laparoscopic anterior partial fundoplication versus total fundoplication (section 3.2). 145 r46
sFcfloM 3
CONTROLLED TRIALS ASSESSING STRATEGIES TO IMPROVE OUTCOMES FOLLOWING SURGERY FOR GASTRO.OESOPHAGEAL REFLUX r47
3.7 PROSPECTIVE RANDOMISED TRIAL OF LAPAROSCOPIC VERSUS OPEN N'SSEru FUNDOPLICATION FOR GASTRO- OESOPHAGEAL REFLUX DISEASE
3.1.1 ArMS
To determine whether the laparoscopic approach reduces the overall morbidity associated with surgery for gastro-oesophageal reflux disease.
To determine whether laparoscopic Nissen fundoplication achieves equal long term
control of reflux symptoms with an acceptably low risk of adverse outcomes.
3.1.2 METHODS
3.1.2.1 Participant assignment
Patients undergoing Nissen fundoplication for gastro-oesophageal reflux disease at the
Royal Hallamshire Hospital, Sheffield, England were randomised to undergo either
Nissen fundoplication performed using a laparoscopic technique or Nissen fundoplication
performed through open laparotomy access. Randomisation was performed by the
operating theatre nursing staff opening one of 100 sealed opaque envelopes on the day
before surgery, to enable the preparation of appropriate operating room instrumentation.
The envelopes were prepared before the study commenced by a departmental secretary.
To minimise the risk of selection bias patients, anaesthetic and surgical staff were not
aware of the randomisation outcome until after general anaesthesia commenced.
Informed consent for the study and randomisation method was obtained from all
participating patients. 148
3.1.2.2 Patient selection and pre-operative assessment
Patients with proven gastro-oesophageal reflux disease presenting for primary anti-reflux surgery were considered for entry into this study. Patients were not entered into the study if they had a documented oesophageal motility disorder which was assessed by the treating surgeon to preclude a Nissen fundoplication, if they required a concurrent
additional abdominal procedure at the same time as fundoplication (e.g. cholecystectomy), or if they had undergone previous oesophageal or gastric surgery. All patients underwent pre-operative assessment with oesophageal manometry to assess
oesophageal peristalsis and lower oesophageal sphincter pressure. Twenty four hour
ambulatory pH monitoring and upper gastrointestinal endoscopy were used to confirm the
presence of reflux disease and to assess the degree of oesophageal mucosal damage.
3.1.2.3 Operating technique
Patients undergoing anti-reflux surgery irrespective of the method of surgical access,
underwent Nissen fundoplication using a technique standardised before the trial
commence 6227 . This was similar to the technique described in section I.2.2.2. Initially
both hiatal pillars were dissected, the oesophagus mobilised, and posterior hiatal repair
was performed using 2/0 Prolene sutures. The anterior wall of the gastric fundus was
then mobilised behind the oesophagus, and a 1.5 to 2 cm long loose 3600 fundoplication
constructed whilst a 52Ft gauge bougie was positioned within the oesophageal lumen to
assist in calibration of the wrap. Three or 4 interrupted2l0 Prolene sutures were used to
secure a loose Nissen fundoplication. The short gastric vessels were not divided in either
study group
Access for open surgery was obtained through an upper midline laparotomy incision. At
completion, this was closed with 1 nylon sutures using a mass closure technique. Access
for the laparoscopic procedure was provided by three 10 mm and two 5 mm laparoscopic
ports. Sutures were secured using extra corporeally tied'Weston knots in the
laparoscopic group229, and surgical square knots in the open group. Stringent attempts 149 were made to ensure that both the laparoscopic and open Nissen fundoplication techniques were the same, except for the surgical access method.
All surgery was performed by one of 3 surgeons. One of these (the author) had had earlier experience at the Royal Adelaide Hospital, acting as either the surgeon or assistant in 45 laparoscopic Nissen fundoplications. Before commencing the trial, a further 20 patients underwent laparoscopic Nissen fundoplication to enable the other participating
surgeons to become adept at the laparoscopic approach.
3.1.2.4 Post-operative care
Post-operative care was standardised, to ensure that both trial groups had equal
opportunities for rapid recovery and early hospital discharge. Nasogastric tubes were not
used in either study group. Morphine analgesia was provided to both groups through
patient controlled analgesia machines. All patients wore offered oral fluids on the first
post-operative day and a soft diet on the second day. Whilst discharge from hospital was
allowed in both groups on the third post-operative day, patients were encouraged to
choose when they were discharged from hospital. All patients, irrespective of the
operative technique, were informed pre-operatively that they would be able to leave
hospital 3 days after the surgical procedure. Patients were instructed to avoid bread and
lumpy solid foods for the first 3 to 4 weeks following surgery, and then to subsequently
increase the consistency of their diet in a stepwise fashion.
3.1.2.5 Masking
Due to the overt differences between the surgical access wounds, no attempts were made
to conceal the operation type from individual patients, or from the clinician performing
clinical follow-up assessments.
3.1.2.6 Clinical follow-up
Patients were interviewed personally by a surgical registrar who applied a structured
questionnaire pre-operatively, and then post-operatively 1,3,6, and 12 months following 150 surgery. Longer term follow-up is to be sought by yearly face to face interview. Patient weight was measured on each occasion, and the presence or absence of the following
symptoms was sought; heartburn, epigastric pain, regurgitation, dysphagia, odynophagia,
early satiety, gas bloat, anorexia, nausea, vomiting, nocturnal coughing and wheezing.
Analgesic consumption during the post-operative inpatient stay was recorded, and pain
scores were assessed on the first post-operative day using a visual analogue score (0= no
pain, 5= severe pain). The time interval between surgery and the commencement of oral
fluids and solids was determined, and any complications occurring in the post-operative
period were recorded. The overall outcome following surgery was also assessed using
two different scoring systems. Patients were asked to rank the outcome of their surgery
using a modified Visick grading system (Table 3.1.1) and an outcome assessment score
(Table 3.1.2).
Table 3.1.1 Modified Visick grading
1 No symptoms
2 Mild symptoms easily controlled by simple care such as avoiding certain foods or
small meals etc.
3 Moderate symptoms not controlled by simple care but not
interfering with social or economic life
4 Moderate symptoms interfering with social or economic life
5 Symptoms as bad or worse than pre-operatively
Table 3.1.2 Outcome assessment
Excellent Complete recovery
Good Major improvement with minor problems
Fair Major improvement with still significant problems or
adverse effects
Poor Minor or no improvement or deterioration 151
3.1.2.7 Objective follow-up
Objective investigation with oesophageal manometry,24 hr ambulatory pH monitoring
and upper gastrointestinal endoscopy was pelformed 3 to 4 months after surgery. These
tests sought to assess lower oesophageal sphincter zone pressures, efficacy of reflux
control, post-operative anatomy, and healing of oesophagitis.
3.1.2.7.1 Oesophageal mønometry
Patients were fasted for 6 hours before each study and all medications affecting
oesophageal motility were ceased 3 days earlier if necessary. Proton pump inhibitors
were ceased 2 weeks before testing. Oesophageal manometry was performed using a 6
lumen water perfused catheter incorporating sidehole sensors. A Dent sleeve was not
incorporated into this catheter. Each lumen of the catheter was connected in series with a
pressure transducer and was constantly perfused with degassed distilled water at 0.5
mUminute by a low compliance pneumohydraulic pump. Signals were recorded on a
Polygraph chart recorder (Grass Instrument Company, Massachusetts, USA). The lower
oesophageal sphincter (or post fundoplication high pressure zone) was located by the
station pull through technique, and the resting lower oesophageal sphincter pressure
determined. The amplitude and propagation of primary oesophageal peristalsis and
residual relaxation pressure of the lower oesophageal sphincter zone was determined
during 10 swallows of 5 ml water boluses. Normal propagation of peristalsis was defined
as complete propagation from the proximal to distal oesophagus of 8 or more of the 10
water swallows.
Whilst it is recognised that the pressures measured at the gastro-oesophageal junction
following fundoplication result from a complex new anatomical arrangement with a
significant contribution provided by the extrinsic pressure generated by the
fundoplication, the resting and residual relaxation pressures measured following surgery
are referred to as post-operative lower oesophageal sphincter pressures to simplify the
wording used in this thesis. r52
3.1.2.7.2 Ambuløtory 24 hr pH monítorùtg
Studies commenced immediately after oesophageal manometry. A glass pH probe
(Radiometer, Copenhagen, Denmark) was positioned 5 cm above the lower oesophageal sphincær zone as measured by oesophageal manometry, and then connected to a
Digitrapper (Synectics Medical, Stockholm, Sweden). Patients were encouraged to continue with normal activities for 24 hours. The probe was then removed and data was
analysed by computer. The percentage of the study duration during which the
oesophageal pH was less than 4, and the total number of reflux episodes were recorded.
3.1.2.7.3 Upper gastroíntestinal endoscopy
Upper gastrointestinal endoscopic examination was performed using a fibreoptic
endoscope (Olympus, Japan). The presence or absence of visible oesophagitis was
determined and scored using the Savary Miller grading system (see Table 1.2.6.1). The
appearance of the fundoplication, the presence of any concurrent para-oesophageal hiatus
hernia, and any anatomical distortion (e.g. bi-lobe stomach9o¡ was also determined.
3.L.2.8 Statistical analysis
The trial was designed to evaluate the following clinical outcomes; operating time,
length of post-operative stay, incidence of complications, time taken until full physical
activity was resumed, post-operative reflux control and incidence of dysphagia. Before
commencing the study, preliminary statistical calculations determined that 84 patients (42
in each group) would be needed to demonstrate a207o difference in any outcome measure
at a significance level of P<0.05 and power of 90Vo.
All analyses were performed on an intention to treat basis, with all patients remaining in
their initial allocated group for subsequent analysis of data. This report describes
outcomes in patients followed for a minimum 3 month period and the results of early
post-operative objective tests. This time period enables the determination of differences
between the peri-operative variables and short term outcomes. Longer term outcomes, 153 which are necessary for the determination of the efficacy of reflux control, will be published once follow-up matures.
All data was entered onto a computerised data base (Filemaker Pro Version 3.0, Claris
Corporation, Santa Clara, California) and was analysed using a commercially available
statistical package (Instat Version 2.01 Graph Pad Software San Diego California).
Fisher's Exact Test was used to determine the significance of 2 x 2 contingency tables, the
Chi-squared test to determine the significance of larger contingency tables, and the two
tailed Mann Whitney U-test to determine the significance of non-parametric data sets.
Statistical significance was accepted when the P value was less than 0.05. Unless stated
elsewhere, all data is reported as the percentage of total patients in each group, or as the
median (range).
3.1.2.9 Ethical approval
The protocol for this study was approved by the Medical Research Ethics Committee of
the Royal Hallamshire Hospital. The study was conducted in accordance with the World
Medical Association Declaration of Helsinki (revised 1989).
3.1.3 RESULTS
Between May 1993 and May 1994,42 patients undergoing a Nissen fundoplication were
entered into this study. Twenty patients were randomised to undergo Nissen
fundoplication utilising laparoscopic access and 22 to the same procedure by the open
technique. All patients were assessed by personal interview I,3,6 and L2 months
following their surgery. No patient withdrew from the study or was lost to follow-up.
3.1.3.1 Pre-operative assessment
Both groups were similar for age, sex, height, weight, incidence of previous abdominal
surgery, duration of symptoms and medications consumed in the pre-operative period
(Table 3.1.3). Analysis of presence or absence of pre-operative symptoms (Table 3.1.4) 154 revealed no significant differences between the two groups, with the exception of a slightly lower proportion of the laparoscopic group reporting pre-operative heartburn.
Pre-operative Visick scores were identical (Table 3.1.5). Pre-operatively, a significant proportion of patients in each group experienced dysphagia for solid food to some extent,
with 42Vo of the laparoscopic and 3I7o of the open group reporting this symptom before
surgery was performed.
Table 3.1.3 Preoperative parameters
(all figures expressed as median (range))
Laparoscopic Open P value
Number of patients 20 22
Age (years) 50.5 (29-1r) 43 (26-69) 0.4r
Sex 9M; 1lF 15M;7F 0.21
Weight (kgs) 69 (43-t02) 75.3 (44-9s) 0.22
Previous abdominal surgery
upper 24Vo I3Vo 0.66
lower 42Vo 377o 0.72
Duration symptoms (mths) 60 (9-360) 60 (8-240) 0.39 Table 3.1.4 Summary of pre and post-operative symptoms
Pre-operatíve 7 month postop 3 months postop 6 months postop 12 months postop Inparoscopic Open Inparoscopíc Open Inparoscopic Open Laparoscopic Open Laparoscopic Open Heartburn 80Vo* IffiVo* 57o 4.5Vo 57o ïVo 07o 07o 67o 07o Epigastric Pain 807o 957o l57a 07o lÙVa ÙVo 6Vo 1Vo 6% ïVo Regurgitation 80Vo 867o 5Vo 4.5Vo 07o 07o ÙVo ÙVo 6Vo ÙVo Dysphagia lumpy solids 42Va 3I7o 407o 4IVo l0Vo 27Vo 6Vo lVo 6Vo 7Vo
soft solids I0Vo ÙVo ÙVo 0Vo 6Vo 0Vo 07o 0Vo Odynophagia 29Vo 25Vo 29Vo* 07o* 0Vo 7Vo 67o 7Vo l2Vo 07o Early satiety 59Vo 697o 4lVo 7I7o lSVa 36Vo I2Vo l47o 6Vo 0Vo Gas bloat 597o 637o 247o 23Vo 67o 2IVo 6Vo 74Vo 0Vo 0Vo Anorexia 47Vo 257o 297o 297o 187o l4Vo 18Vo 7Vo I2Vo 17o Nausea 47Vo 387o l87o l4Vo 07o 7Vo ÙVo ÙVo 0Vo 07o Vomiting ISVo 50Vo l27o 7Vo 67o J7o ÙVo 0Vo 6Vo 0Vo Nocturnal cough l2%o 257o 67o l4Vo 0Vo 7Vo 07o 07o 0Vo 07o
Nocturnal wheeze 0lo 6Vo ÙVo I4Vo 07o ÙVo 07o 07o 07o ÙVo
No significant dffirences demonstratedbetween trial groups (ie P>0.05 at allfollow-up intervals), * except heartburn assessed preoper&tively (P=0.04i Fisher's exact test), and odynophagia assessed at I month (P=0.048 Fisher's emct test) Table 3.1.5 Outcome and Visick grading
Pre-operøtive 7 monthpostnp 3 months postop 6 months postop 72 months postop Laparoscopic Open Laparoscopic Open Laparoscopic Open Laparoscopic Open Laparoscopic Open Outcome Excellent nla nla 26Vo 27% 58Vo 417o 7ïVo 64Vo 657o 64.Vo Good nJa nla 587o 64Vo 37Vo 55Vo 20Vo 32Vo 25Vo 367o Fair nla nla r1.70 9Vo 5Vo 07a l07o 57o l0Vo 07a Poor nla nla 57o 07o 0Vo 5Vo ÙVa ïVo 07o ÙVa Visick grade
1 0Vo ÙVo 53Vo 27Vo 79Vo 597o 80Vo 68Vo 7ÙVo 737o
2 07o 0Vo 327o 68Vo IIVo 367o I5Vo 277o 257o 27Vo
3 35Vo 4lVo 5Vo jVo 5Vo 07o 57o ÙVo 5Va 0Vc
4 657o 59Vo IIVo 57o 57a 5Vo 0Vo 57o 07o 0Vc 5 07o 07o 07o 0Vo 07o ÙVo 0Vo O7o ÙVo ÙVo n/a = not applicable. No tests for significance between groups at comparable follow-up intervals were significant (ie P>0.05 at allfollow-up interttals)- r57
The endoscopic oesophagitis grade visualised before surgery was similar for the two groups. 40Vo of the laparoscopic group and 297o of the open group had evidence of complicated reflux disease (Barrett's oesophagus or stricture formation, P=0.52). A hiatus hernia was visualised pre-operatively in 80Vo of the laparoscopic group and 57Vo of the open group (median length 5 cm in each group). Pre-operative oesophageal manometry outcomes were similar (Table 3.1.6). No significant differences in resting lower oesophageal sphincær pressure, lower oesophageal sphincter residual relaxation
pressure and peristaltic success were seen. The measured mean percentage exposure to
an acid pH of less than 4 during pre-operative 24 hour pH monitoring was 7 .5Vo in the
laparoscopic group and9-5Vo in the open group.
Table 3.1.6 Oesophageal manometry results
(figures expressed median (range))
Laparoscopic Open P value
Preoperative
LOS resting pressure 11 (0-17) 7 (0-20) 0.09
LOS residual pressure 0 (0-0) 0 (0-0) 1.0
No. propagated swallows 10 (1-10) 10 (5-10) 0.87
Normal peristalsis propagation 89Vo 82Vo 0.60
Post-operative
LOS resting pressure 17 (8-31) 23 (8-47) 0.t4
LOS residual pressure 2 (0-8) 3 (0-I2) 0.38
No. propagated swallows 9 (4-10) 10 (3-10) 0.92
Normal peristalsis propagation 80Vo 807o 1.0
AII pressures measured in mmHg
Swallowíng assessedfrom l0 wet swallows 158
3.L.3.2 Surgery
Data was analysed on an intention to treat basis, with patients whose procedure was
converted from a laparoscopic to open procedure remaining in the laparoscopic group for
subsequent analysis. 42Vo of patients undergoing laparoscopic surgery had undergone
previous lower abdominal surgery, compaled to 37% of the open group, and247o of the
laparoscopic group versus 13% of the open group had undergone an earlier open upper
abdominal operation. Adhesions rwere present in a smaller proportion of both groups,
l3%o laparoscopic versus l97o open. Possibly due to the ballooning effect of
pneumoperitoneum, a hiatus hernia was visualised at surgery in87%o of the laparoscopic
group compared to only 257o of the open group (P=0.001).
Median operating time was 79 minutes (range 47 to I54) in the group undergoing
laparoscopic surgery versus 43 minutes (22 to 78, P<0.0001) in the open group. Median
operating theatre time was 109 minutes (range 67 to 185) in the laparoscopic group,
versus 65 minutes (43 to 105, P<0.0001) in the open group. The operating room set-up
time was slightly longer in the laparoscopic group (27.5 versus 21 minutes, P=0.041).
Four of the laparoscopic procedures were converted to an open operation to enable the
procedure to be completed safely. This was necessary because of obesity in 2 patients,
intra-operative bleeding from a lacerated inferior phrenic vein in one, and dense peri-
oesophagitis which prohibited safe laparoscopic oesophageal mobilisation in another.
Surgeons were asked to rate the difficulty of the operation using a 0 to 10 visual analogue
scale. The assessed difficulty rating was similar for both operations (5 versus 5). The
proportion of laparoscopic and open procedures performed by each surgeon was similar.
The length of the upper abdominal incision used for open fundoplication varied from 13
to26 cm (median 18.3)
3.1.3.3 Early hospital outcomes
While the pain scores measured for each group on the first post-operative day were
similar (median laparoscopic = 0, open = l),85Vo of patients in the laparoscopic group 159 had either no pain, or minimal pain, compared to 657o in the open group (P=0.16). Post- operative morphine consumption delivered by either a patient controlled analgesia machine or subcutaneous injection, was significantly greater in the open surgical group
(median I27 mgversus 25 mg P<0.001).
Patients in both groups resumed oral fluid and solid intakes at the same time interval following surgery (fluids on the first postoperative day, and solids on the second day; median). The median length of post-operative stay was 3 days for the laparoscopic group
(mean 3.55 days, range 2 to 8), compared to 4 days for the open group (mean 4.68, runge
3 to 11; P=0.0048). Median total hospital stay was 4 days in the laparoscopic group compared to 5 days in the open group (mean 4.55 versus 5.68; P=0.0048). All patients were free of heartburn and regurgitation symptoms at the time of discharge.
Complications occurring within 30 days of surgery were less common in the laparoscopic group. One patient (5Vo) developed both respiratory basal atelectasis, and urinary retention requiring catheterisation for 24 hours, in the early post-operative period. This patient's procedure had been converted from a laparoscopic to open operation. In comparison, nine patients (4I7o; P=0.0098) in the open group developed an early post- operative complication. Most of these were minor and did not delay post-operative hospital discharge. Three patients developed pulmonary infection or basal atelectasis requiring antibiotic therapy, 3 required urinary catheterisation for 24hour periods for acute urinary retention, 2 developed an intravenous cannula site infection, 1 developed a minor wound infection, and 2 developed a pyrexia of unknown origin (temperature in excess of 38.50C intermittently for more than 24 hours). None of these complications prolonged the postoperative hospital stay beyond the fifth post-operative stay. One patient, however, was given a non-steroidal anti-inflammatory medication for post- operative pain relief following open fundoplication. This patient developed acute 'Whilst, haemorrhage from a duodenal ulcer on the fourth post-operative day. pre- operative endoscopy two weeks earlier, had visualised a normal duodenum, repeat endoscopy demonstrated thrombus in the base of a posteriorly sited duodenal ulcer. 160
Although, the problem was successfully managed non-operatively, a total post-operative stay of 11 days resulted from this problem.
3.1.3.4 Late hospital outcomes
Normal physical activity was resumed between one and 8 weeks following laparoscopic
Nissen fundoplication (median 2 weeks) compared to 3 to 22 weeks following open
Nissen fundoplication (median 8 weeks, P<0.001).
Four patients (207o) in the laparoscopic group were readmitted to hospital in the 12 month follow-up period following surgery. Readmission occurred for acute appendicitis 19 weeks following fundoplication in one patient, for the repair of a port site hernia 12 months following surgery in a second patient, and for acute dysphagia due to a possible food bolus obstruction 18 weeks following surgery in another. The latter problem resolved without either surgical or endoscopic intervention within 24 hours of its onset.
A further patient underwent revision of the original laparoscopic fundophcation 30 weeks following the initial surgory. This was required for post-operative dysphagia. The
Nissen fundoplication was revised to a posterior partial fundoplication, using a conventional open approach. Outpatient endoscopic dilatation was required once in another patient (5V"). This patient, however, had undergone multiple endoscopic dilatations for a peptic stricture before antireflux surgery was performed.
No patient in the open group required either readmission or reoperation during the initial
12 month follow-up period. One patient (5Vo), however, did undergo endoscopic dilatation for post-operative dysphagia. This patient described no dysphagia at pre- operative assessment.
3.1.3.5 Clinical outcome assessment
The symptom assessment one, 3, 6 and 12 months following surgery is summarised in
Table 3.I.4. All patients in the open group and all but one in the laparoscopic group were completely free of reflux symptoms 12 months following fundoplication. The incidence 161 of dysphagia for both lumpy and soft solid foods declined with progressive follow-up, with no significant differences seen between the groups at any follow-up interval. Whilst the incidence was high (40 and 4I7o) one month after surgery, it fell to 6 and 77o by 12 months. Residual dysphagia at 12 months was only experienced for lumpy solid foods.
Soft solids and liquids were swallowed without difficulty.
Gas bloat symptoms occurred in approximately a quarter of patients in each group one month following surgery. However, the incidence of this symptom was less than that seen pre-operatively. No patients experienced bloat symptoms at twelve months follow- up.
No significant differences in the overall outcome assessments were seen at any post- operative follow-up interval (Table 3.1.5). In both groups, a good or excellent outcome was reported by 90 to 95Vo of patients. Visick grading improved significantly following surgery. Whilst 537o of patients were graded Visick 1 one month following laparoscopic fundoplication versus 27Vo in the open group, this difference disappeared as follow-up progressed. By 12 months at least 707o of patients in each group were scored as Visick
1. 95 to L00Eo of patients were scored as Visick 1 or Visick 2 atthe 6 and 12 month follow-up intervals.
3.1.3.6 Objective post-operative investigations
Upper gastroinûestinal endoscopy was performed in 19 (95%) of the laparoscopic group and2l (95E") of the open group. There was no evidence of persistent ulcerative oesophagitis in any patient. Six patients in each group had persistent Barrett's oesophagus at post-operative endoscopy, and one patient in the laparoscopic group required dilatation for a persistent peptic stricture. Post-operative appearances of either para-oesophageal hiatus herniation or movement of the fundoplication into the thorax were not seen in any patient in this study. In all cases the fundoplication was intact at the time of post-operative endoscopy. One patient in the laparoscopic group did exhibit the endoscopic appearance of a bilobed stomach9O. This patient subsequently underwent t62 surgical revision for post-operative dysphagia, with the anatomical effor corrected at the time of this procedure.
Twenty four hour pH monitoring was obtained 3 months following surgery in95Vo of the laparoscopic group and 867o of the open group. The percentage time the pH was less than 4 ranged from 0 to 5.5Vo (median 0.307o, moan 0.86Vo) in the laparoscopic group, compared to 0 to 5.0Vo (medían0.50Vo, mean I.l7o,P=0.27) in the open group. The median number of reflux episodes occurring in the laparoscopic group was 3, compared to
8 for the open group (P=0.027).
Post-operative oesophageal manometry outcomes are summarised in Table 3.I.6. The
resting lower oesophageal sphincter pressure was 17 mmHg in the laparoscopic group,
compared to 23 mmHg in the open group. This difference was not statistically
significant (P=0.14). Lower oesophageal sphincær residual relaxation pressure was
similar (2 versus 3 mmHg, P=0.38). 80Vo of patients in both groups propagated 8 or
more of their 10 wet swallows post-operatively.
3.1.4 DISCUSSION
Initial published series of laparoscopic anti-reflux surgical procedures suggested that the
laparoscopic approach is associated with reduced ear'ly post-operative morbidity, and that
it may reduce the overall morbidity associated with surgery for gastro-oesophageal reflux
¿irs¿se19,37 ,70,77 ,90,206,219. Early attempts to compare the outcome of laparoscopic
antireflux surgery with historical experiences with open fundoplication also appear to
confirm that the laparoscopic approach dramatically reduces the length of post-operative
hospital stay and the incidence of early post-operative complications, and that it is equally
efficacious in the early control of reflux symptoms170,181. Other studies have compared
the cost of the laparoscopic approach with open surgery, suggesting that the shortened
post-operative hospital stay results in significant cost savings, despite longer operating f63 dmes82,110. However, these comparisons have only assessed costs in the American health care system.
It is now apparent that whilst early reports of uncontrolled series of patients undergoing other laparoscopic procedures, such as inguinal hernia repair, appendicectomy and laparoscopic cholecystectomy, suggested that the laparoscopic approach resulted in significant reductions in peri-operative morbidity and hospital stay58'73'189, the subsequently reported results of prospective randomised trials have often not reproduced the same degree of predicted benefitl32,134,155. Trials comparing laparoscopic versus open inguinal hernia repair and laparoscopic versus open appendicectomy have not demonstrated significant advantages for the laparoscopic approach132,155 - Similarly, trials comparing laparoscopic cholecystectomy with a mini-laparotomy approach have revealed that the reduction in post-operative hospital stay, if any, is much less than most surgeons would have predicted following the initial assessment of uncontrolled
."¡ss134,140. Because the initial evaluation process for laparoscopic anti-reflux surgery has mirrored that followed earlier for other laparoscopic procedures, it is essential to critically evaluate this new surgical approach.
Conceptually, laparoscopic anti-reflux surgery makes sense. Unlike cholecystectomy which has a minimally invasive open altemative or mini-laparotomy approach134,I40, there are no currently available minimally invasive methods for open anti-reflux surgery
Whilst the early results of this trial suggest advantages for laparoscopic anti-reflux surgery, these were not as great as many surgeons might have expected. The median post-operative hospital stay of 4 days following open fundoplication was significantly shorter than the 7 to9 days reported in non-randomised studies82,l10,181. It is likety that the length of the post-operative hospital stay is significantly influenced by information provided to patients before surgery, and expectations of their attending medical and nursing staff. Overall post-operative stay in this study was reduced by a mean of 1.1 days, suggesting that improvements in historical practices for open anti- reflux surgery may also achieve significant cost savings from earlier post-operative r64 discharge. In this study all patients were infolmed that they would be able to leave hospital at an early stage following surgely irrespective of the surgical technique.
The procedures compared in this study were standardised so that the variable investigated was the method of surgical access. The median procedure time of 79 minutes for laparoscopic Nissen fundoplication, was shorter than that reported in most earlier published experience s77 ,90, suggesting that many of the initial learning difficulties which can arise during the laparoscopic approach had been overcome before commencing the trial. Nevertheless, the comparable open Nissen procedure was quicker to perform
(median 43 minutes). The 36 minute advantage for open surgery is similar to that reported in non-randomised s1u¿ies82,110'170,181. Surprisingly, the operating room set up time was increased by only 6 minutes by the laparoscopic approach, despite the increased complexity of the equipment required.
Benefits were demonstrated for the laparoscopic approach in the early post-operative period. Post-operative analgesic requirements and pain scores were reduced. The incidence of early post-operative complications was also reduced significantly by the laparoscopic approach. Whilst late re-admission was more likely, 3 of the 4 patients requiring readmission following laparoscopic Nissen fundoplication were admitted for either an unrelated problem or a minor problem which was rapidly resolved. It is diffîcult to know whether the one patient who required revision of the initial fundoplication 7 months after laparoscopic Nissen fundoplication, reflects a higher early surgical revision rate following this procedure, or whether this event was a random occurrence. Non-randomised studies have suggested that there is an important incidence of early surgical revision due to specific laparoscopy related problems24,90,220,22I.
Further follow-up, and other trials will be required to evaluate this further.
The most dramatic benefit of laparoscopic Nissen fundoplication was the shortening of the convalescent period from 8 to 2 weeks. This will allow significant community cost savings if it results in reduced time off work for employed patients presenting for anti- 165 reflux surgery. The clinical and objective outcomes assessed post-operatively confirm that the laparoscopic procedure is as effective as open Nissen fundoplication for the control of reflux symptoms in the initial 12 months following surgery.
In the light of the results reported in section L.2.5.1, which suggest that the learning curve for laparoscopic fundoplication comprises up to 20 patients per surgeon213, it might be suggested that the learning curve was not fully eliminated before this study commenced.
V/hilst one surgeon had been involved in over 40 procedures before the trial started, the other surgeons had a shared pre-trial experience of only 20 cases. Whilst, this may have contributed to a higher conversion rate to the open procedure than that reported by other groups19,70,77,90,206, alarge group of patients presenting for anti-reflux surgery at the
Royal Hallamshire Hospital had complicated reflux disease (i.e.. Barrett's oesophagus or stricture formation). Performing laparoscopic antireflux surgery for these patients may be technically more difficult, a factor which can contribute to the conversion rate.
The results of this study confirm that whilst laparoscopic Nissen fundoplication takes longer than open fundoplication to perform, it does reduce the duration of the post- operative hospital stay, convalescence time, and the overall complication rate associated with anti-reflux surgery. Clinical outcomes 12 months following surgery are unrelated to the surgical access method. t66
3.2 PROSPECTIVE DOUBLE BLIND RANDOMISED TRIAL OF LAPAROSCOPIC N'SSE V FUNDOPLICATION WITH DIVISION AND WITHOUT DIVIS'ON OF SHORT GASTRIC VESSETS
(Published previously as; Watson DI, Pike GK, Baigrie RJ, Mathew G, Devitt PG,
Britten-Jones R & Jamieson GG. Prospective double blind randomised trial of laparoscopic Nissen fundoplication with division and without division of short gastric vessels. Annals of Surgery 1997; 226: 642-652226¡
3.2.1 ArM
To determine whether division of the short gastric vessels to achieve full mobilisation of the gastric fundus is necessary to reduce the risk of post-operative dysphagia and other adverse sequelae.
3.2.2 METHODS
3.2.2.L Participant assignment
Patients undergoing laparoscopic Nissen fundoplication for gastro-oesophageal reflux disease at the Royal Adelaide Hospital were randomised to undergo fundoplication with or without division of the short gastric vessels. Informed consent was obtained from all participants, and randomisation occurred in the operating theatre, after the commencement of general anaesthesia, by opening one of 120 previously sealed opaque envelopes.
Preparation of the envelopes was undertaken before the study commenced by a research officer not directly involved in the trial, and envelopes were selected by a departmental secretary, at a surgeon's request. r67
3.2.2.2 Patientselectionandpreoperativeinvestigation
All patients with proven gastro-oesophageal reflux disease, who presented for primary antireflux surgery by the laparoscopic technique were considered for entry into this trial.
Patients were excluded from consideration only if they had an oesophageal motility disorder which precluded a 3600 fundoplication, or required a concurrent abdominal procedure at the same time as fundoplication (e.g. cholecystectomy), or had undergone previous antireflux surgery. All patients underwent preoperative investigation with oesophageal manometry and endoscopy. Preoperative manometric testing also included an acid reflux provocation test and Bernstein test. 24hr pH monitoring was performed routinely for patients who did not have unequivocal reflux disease demonstrated by preliminary endoscopic and manometric studies. The majority of patients also underwent pre-operative barium meal X-ray examination.
3.2.2.3 Operating technique
Laparoscopic Nissen fundoplication was performed using the surgical technique described in section 1.2.2.290,219. In brief, this consisted of dissection of the hiatal pillars, followed by full oesophageal mobilisation, and routine posterior hiatal repair. If short gastric vessels were to be divided, this step was performed next. Vessels were dissected, secured by metal clips, and divided. Division usually commenced at the level of the inferior pole of the spleen, and progressed superiorly along the greater curvature of the stomach until the left pillar of the hiatus was seen. Division of short gastric vessels was thought to be adequate if the superior part of the gastric fundus could be brought loosely around the oesophagus for construction of the fundoplication.
If short gastric vessels were not divided, the anterior wall of the gastric fundus was pulled behind the oesophagus for construction of the fundoplication. Occasionally, when the short gastric vessels had not been divided, the first piece of fundus selected appeared tight. By repositioning instruments and grasping an adjacent piece of fundus, a much looser wrap could be constructed. Care was taken to ensure that the completed fundoplication was not tight by having a 52Fr bougie within the abdominal oesophagus. 168
Three or four 2/0 Prolene interrupted sutures were used to secure the wrap which was 1.5 to 2 cm in length. If the laparoscopic procedure was converted to an open procedure due to intra-operative difficulties, short gastric vessels were still divided or not divided according to the randomisation schedule, with the patient remaining in the trial.
3.2.2.4 Post-operative care
Nasogastric tubes were not used in any patients. Patients were allowed oral fluids post- operatively on the evening ofthe day ofsurgery, and soft solid food the next day.
Discharge from hospital was encouraged after the second post-operative day. Patients were instructed to avoid bread and lumpy foods for the first 3 to 4 weeks following surgery, and then to gradually increase the consistency of their diet. A barium meal examination was usually obtained on the second post-operative day, to detect any problems amenable to early laparoscopic reintervention (e.g. acute paraoesophageal hernia, tight fundoplication or hiatus).
3.2.2.5 Masking
Whether or not the short gastric vessels had been divided during laparoscopic fundoplication was concealed from all patients. As patients had no direct access to casenotes or trial records, and both laparoscopic procedures utilised identical operative wounds, all remained unaware of the exact procedure for the duration of the trial follow- up period. Whilst operating surgeons were aware of the sxact procedure performed, all follow-up was obtained by a scientific officer who was 'blinded' to the randomisation of each patient. As he was not involved in the initial surgery, he remained unaware of the allocated group for each patient throughout the follow-up period. Participant data was entered into a computerised data base by another research assistant who was not involved in direct patient follow-up. Final data analysis was performed independently by both the scientific officer, and the author. r69
3.2.2.6 Clinical follow-up
Patients were interviewed pre-operatively and then 1, 3 and 6 months after surgery by a scientific officer, using a structured questionnaire. Whilst longer ærm follow-up will be sought, it is not available for reporting in this thesis. The presence or absence of each of the following symptoms was sought; heartburn, epigastric pain, regurgitation, dysphagia for lumpy solids, soft solids and liquids, odynophagia, early satiety, inability to belch, epigastric bloating, anorexia, nausea, vomiting, noctumal coughing and wheezing. The ability to relieve bloating and whether a normal diet was being consumed was also determined. Heartburn was also scored using a visual analogue scale (0 = no heartburn,
10 = severe heartburn).
Dysphagia was scored by several methods. Visual analogue scales (0 = no dysphagia, 10
= total dysphagia) were independently applied for solids and liquids, as well as a previously validated score36 (0 = no dysphagia, 45 = severe dysphagia) which combines information about diffîculty swallowing 9 types of liquids and solids (Table 3.2.1). This latter score was reversed from that originally described so that the numerical score increased with the severity of dysphagia. Overall outcome was determined using 3 further scales. Patients ranked the outcome of surgery using a modified Visick grading
(Table 3-2.2), and were asked to score the outcome as excellent, good, fair, or poor (Table
3-2.3). An overall assessment of satisfaction with the operative outcome was determined by a further visual analogue scale (0 = dissatisfied, 10 = satisfied). r70
Table 3.2.1 Dysphagia score 1 Water
2 Milk (or thin soup)
3 Custard (or yoghurt or pureed fruit) 4 Jelly
5 Scrambled egg (or baked beans or mashed potato)
6 Baked fish (or steamed potato or cooked carrot) 7 Bread (or pastries) 8 Apple (or raw carrot)
9 Steak (or potk or lamb chop)
The presence of any dysphagia for each liquid or solid substance listed above is first determined and scored; dysphagiaalways= l point, sometimes =ll2point,never= 0points. A scorefrom0 (no dysphagia) to45
(severe dysphagia) is then determined by multiplying the score for each substance by the adjacent line number, and then summing all nine lines.
Table3.2.2 Modified Visick grading
1 No symptoms
2 Mild symptoms easily controlled by simple care such as avoiding certain
foods or small meals etc.
3 Moderate symptoms not controlled by simple care but not interfering with
social or economic life
4 Moderate symptoms interfering with social or economic life
5 Symptoms as bad or worse than pre-operatively t7r
Table 3.2.3 Outcome assessment
Excellent Complete recovery
Good Major improvement with minor problems
Fair Major improvement with still significant problems or adverse effects
Poor Minor or no improvement or deterioration
3.2.2.7 Objective follow-up
Objective investigation with oesophageal manometry,24 hour pH monitoring, barium meal examination, and a radionuclide oesophageal emptying study were performed 3 to 4 months following surgery. Investigation sought to assess lower oesophageal sphincter function, control of reflux, post-surgical anatomy, and the presence of any post-surgical oesophageal obstruction due to a tight wrap or any other cause.
3.2.2.7.1 Oesophageal MønometrY
Patients were fasted for 6 hours before each study and all medications affecting oesophageal motility ceased 3 days earlier if necessary. Oesophageal manometry was performed using an 8-lumen water perfused catheter incorporating a sleeve sensor (Dent
Sleeve, Adelaide) with signals recorded on a polygraph chart recorder (Grass Instrument
Company, Model TD, Massachusetts, USA). The lower oesophageal sphincter (or post- fundoplication high pressure zone) was located by the station pull through technique and the centre of the sleeve was positioned at the central point of the lower oesophageal sphincter. Each lumen of the catheter was connected in series with a pressure transducer
(Stratham P23lD, Gould Inc, Oxnard, California) and was constantly perfused with degassed distilled water at 0.5 mVmin by a low compliance pneumohydraulic pump
(Amdorfer Medical Specialities, Greendale,'Wisconsin). The resting lower oesophageal sphincær pressure was measured over a 5 minute period followed by measurement of the amplitude and propagation of primary peristalsis, and residual relaxation pressure of the lower oesophageal sphincter, during 10 swallows of 5 ml water boluses. 172
3.2.2.7.2 Ambul.atory 24 hour pH monitoríng
Medications affecting gastric acid production were ceased 3 days earlier (Proton pump inhibitors ceased 5 days earlier). A glass pH probe (Radiometer, Copenhagen, Denmark) was positioned 5 cm above the lower oesophageal sphincter measured by oesophageal manometry, and connected to a Digitrapper (Synectics Medical, Stockholm, Sweden).
The patient was encouraged to continue with normal activities for 24 hours. The study was analysed for the percentage of time during which pH was less than 4 and for the correlation between reflux symptoms and measured reflux events.
3.2.2.7.3 RadíonuclideOesophagealEmptyírtgStudy
This test measured oesophageal emptying of 3 swallows of a solid meal of cooked ground beef containing 10 to 12 MBq of 99m technetium sulphur colloid dispersed in egg white.
Oesophageal emptying was measured as the average time taken for 957o of each of three
10g solid boluses to clear from the oesophagus. The normal emptying time for this test is between 7 and 93 seconds.
3.2.2.7.4 BaríumSwallow examinøtíon
Swallowed radio-opaque barium contrast was used to image the distal oesophagus, fundoplication and stomach. Imaging determined any gross delay in oesophageal emptying, the site of the fundoplication (abdominal versus thoracic), the presence or absence of any paraoesophageal herniation, and any abnormal distortion of gastric anatomy. Prone oblique views were obtained specifically to examine for paraoesophageal herniation.
3.2.2.8 Statistical analysis
The primary clinical outcomes which the trial was designed to evaluate were post- operative dysphagia and control of reflux symptoms. Before the commencement of the trial it was determined that 84 patients (42 \n each group) would be needed to demonstrate a20Vo difference in these outcome measures, at a significance level of P<0.05, and power of 90Vo. To ensure that this was achieved, it was intended that 100 patients would be t73 recruited, allowing for an estimated 207o of all patients refusing the objective post- operative investigations. All analyses \ryere performed on an intention to treat basis, with all patients remaining in their initial allocated group for this analysis.
Before commencing the trial, it was intended to report the initial outcomes and results of post-operative testing once all patients had been followed for an initial 6 month period
(this study). This time period is thought to allow for the adequate assessment of any differences in the incidence of post-operative dysphagia between the two trial groups.
Medium to long term outcomes are more impoftant for determining the efficacy of reflux control, and will be described once follow-up has matured further.
Data was entered onto a computerised data base (Filemaker Pro version 3.0, , Claris
Corporation, Santa Clara, California) and analysed using a commercially available statistical package (InStat version 2.01, GraphPad Software, San Diego, California).
Fisher's Exact Test was used to determine the significance of 2 x 2 conttngency tables. A two tailed Mann-Whitney U-Test, was used to assess the significance of non-parametric sets of data, and an unpaired Student T-Test to detemine the significance where it was reasonable to assume a parametric distribution (height and weight). Statistical significance was accepted at a P value of less than 0.05. Unless otherwise stated all data is reported as the percentage of the total patients in each group, or as the mean (95Vo confidence intervals).
3.2.2.9 Ethical approval
The protocol for this study was approved by the Royal Adelaide Hospital Human
Research Ethics Committee, and the study was conducted in accordance with the World
Medical Association declaration of Helsinki (revised 1989), and the National Health and
Medical Research Council of Australia's guidelines on human experimentation. 174
3.2.3 RESULTS
From ly'ray |994to October 1995, 102 patients undergoing a laparoscopic 3600 Nissen fundoplication at the Royal Adelaide Hospital were entered into this trial. Fifty patients were randomised to undergo fundoplication without short gastric vessel division, and 52 to undergo division of these vessels. During the same period 38 further patients underwent a laparoscopic Nissen fundoplication performed by surgeons contributing patients to this study. Three of these patients were excluded because of the need to perform a concurrent abdominal procedure (cholecystectomy - 2, highly selective vagotomy - 1). The remaining 35 patients either refused entry into the trial because they had a preference for a specific procedure to be performed, or they were unwilling to participate in the follow-up protocol. Of the 102 patients entered 98 (96Vo) were available for follow-up one month after surgery,99 (97Vo) at 3 months, and IO0 (98Vo) at
6 months. Although prospectively collected follow-up data was not available for a small number of patients at the specific follow-up intervals, no patient elected to withdraw from the study. Missing data was due to an inability to contact patients at the specific follow- up intervals. Only two patients could not be contacted 6 months after surgery, one following a decision to emigrate to Greece. Figure 3.2.1 summarises the follow-up compliance within the trial. 175
Figure 3.2.1 Compliance with follow-up protocol
102 patients randomised
Vessels not divided Vessels divided 50 patients 52
1 month followup 1 month followup 48 patients (967o) 50 (96Eo)
3 month followup 3 month followup 48 patients (967o) 51 patients (987o)
6 month followup 6 month followup 49 patients (98Vo) 51 (987o)
Objective investigations Objective investigations Barium meal 3?patients (647o) Barium meal 30 patients (587o) patients (787o) Oesophageal emptying 4l patients (79Vo) 25 patients (507o Oesophageal manometry/pH 31 patients (607o)
3.2.3.1 Preoperative assessment
Both groups were similar for age, sex, height, weight, cigarette and alcohol consumption, incidence of previous abdominal surgery, duration of symptoms, and medications consumed before surgery (table 3.2.4). Analysis of the presence or absence of preoperative symptoms (table 3.2.5), as well as the assessment of heartburn using the visual analogue scale (table 3-2.6), revealed no significant differences. A significant proportion of patients in each group experienced preoperative dysphagia to some extent, with an incidence of. 43Vo and 527o when assessed using the dysphagia score (table 3.2-7)
Whilst different methods of scoring dysphagia elicited slightly different rates, there was no clinically or statistically significant difference between the two groups. Preoperative
Visick grading was also similar for each group (table 3.2.8) 176
Table3.2.4 Preoperativeparameters
(all figures expressed as mean (95Vo contidence intervals))
Vessels not divided Vessels divided P value
Number of patients 50 52
Age (years) 46.7 (42.7,50.8) 45.3 (41.8, 48.8) 0.50 Sex 31M; 19F 31M;21F 0.84
Height (cm) 170 (167,173) 112 (169,175) 0.39 V/eight ftgs) 83.5 (79.4,87.5) 84.5 (80.0, 88.9) 0.74 Cigarette smoker 20Vo l27o 0.38
Alcohol consumed 67Vo 667o 1.00
Previous abdominal surgery 38Vo 40Vo 0.84
Duration symptoms (yrs) 9.1 (6.7, 11.5) 8.2 (5.6, 10.8) 0.38
Preoperative medications
Omeprazole 68Vo 66Vo 0.83
H2 Blocker 78Vo 84Vo 0.44
Cisapride l87o 207o 0.80 Table 3.2.5 Summary of pre and post-operative symptoms
Pre-operøtive 7 month postop 3 months postop 6 months postop Not divided Divided Not divided Divided Not divided Divided Not divided Divided Heartburn 947o 88Va 8Vo 87o 4Vo 67o 6Vo l07o Epigastric Pain 57Vo 567o 38Vo 507o 3I7o 377o 29Va l87o Regurgitation 90Vo 887a 29Vo 26Vo l9%a t47o l4%o 67o Odynophagia 2'77o l6Vo lÙVo I4Vo 13% \Vo 4% 4Va Early satiety 37Vo 30Vo 857o 90Vo 63Vo 697o 4IVo 49,qo 357o 4l7o Epigastric bloat 59Vo 44Vo 507o 507o 33vo* 557o* Anorexia 47o 6Va 2lVo l6Vo ïVa 6Vo 8Va 2Vo Nausea 2lVo l67o 2l%o I6Vo 2Vo lÙVo l0% 2Vo Vomiting 22Vo 267o 47o l07o ÙVo ÙVo ZVo ÙVo Nocturnal cough 39Vo 407o l5%o l4%o \Vo l47o l2Vo 6Vo Nocturnal wheeze I67o 207o 8Vo ljVo 4Vo I2Vo 8Vo 47o Can relieve bloat 79Vo 797o 527o 507o 637o 597o 65Va 57Vo
Unable to belch 0% ÙVo 48Vo 42Va 46Vo 5I7o 387o 53Vo Eats normal diet 55Vo 787a 48Vo 48Vo 85% 84Vo 92Vo 88Va
No significant dffirences demonstrated between trial groups (ie P>0.05 at all follow-up intervals), * except epigastric bloat 3 months afier surgery @=0.043, Fisher's exact test)- t78
Table 3.2.6 Assessment of heartburn by visual analogue scale
(figures expressed as mean (95Vo confidence intervals))
Vessels not divided Vessels divided P value Preoperative 4.8 (3.7, 5.8) 3.7 (2.8,4.7) 0.18
1 month post-operative 0.21 (-0.10,0.51) 0.20 (-0.01, 0.41) 0.88
3 months post-operative 0.11 (-0.17,0.32) 0.47 (-0.06,0.98) 0.63
6 months post-operative 0.33 (-0.03,0.69) 0.48 (-0.04, 1.00) O.75 Table3.2.7 Dysphagia assessment
Pre-operafive I month postop 3 monÍhs postop 6 monfhs postop Not divided Divided Not divided Divided Not diviied Divided Not diviied Divided Dysphagia for
Lumpy solids 37Vo 36% 56% 56Vo 48Va 49% 33Va 29Va
Soft solids 4Vo 8Vo t7% l6Vo 8Vo t3% 2Vo 2Vo
Liquids 6Vo 2Va l5Vo r8% l3Vo tÙVo 6Vo 76Vo Visual analogue scale Solids 1.7 (0.9,2.5) 2.5 (t.5,3.4) 3.4 (2.6,4.2) 3.2 (2.4,4.0) 2.2 (7.4,3.0) 2.7 (1.3,2.7) 1.4 (0.8,2.0) 1.3 (0.7, 1.9) Liquids 0.74 (0.1, t.4) 0.78 (0.2, 1.3) 1.4 (0.6,2.3) 0.9 (0.4, 1.4) 0.60 (0.0ó, 1.2) 0.61 (0.2, 1.1) 0.39 (-0.07,0.9) 0.48 (0.1, 0.8) Dysphagia score Overall result 9.5 (5.5, 13.6) 7.6 (4.5, 10.8) 14.5 (10.8, 18.2) 13.5 (10.2, 16.8) 8.3 (5.1, 11.5) 7.9 (5.2,10.5) 4.8 (2.4,7.2) 4.6 Q.3,6.9) Scored 0 only 48Vo 57Vo 26Vo 27Vo 43Vo 39Vo 63Vo 6lVo
Figures are either percentage of total, or mean (95Vo confidence intervals) No tests for signiflcance between groups at comparable follow-up intewals were signifrcant (ie P>0.05 at allfollow-up intenals). Table 3.2.8 Outcome, satisfaction, and Visick grading
Pre-operatíve 7 month postop 3 months postop 6 months postop Not divided Divided Not divided Divided Not divided Divided Not divided Divided Outcome Excellent nla nla ITVo l4Vo 23Vo 197o 29Vo 257a Good nla nla 607o 60Vo 647a 67Vo 59Vo 69,qo Fair nla nla 237o 24Vo IIVo l2Vo l0Vo 4Vo Poor nla nla 0Vo 27o 2Vo 27a 2Vo 27o Visick grade
1 0Vo 0Vo llVo l2Vo 23Vo I5Vo 25Vo 23%
2 0Vo 0Vo 537o 46Vo 60Vo 7ÙVo 57Vo 697o a J 26Vo 297o 257a 30Vo IIVo 97o lÙVa 2Vo 4 74Vo l17a IlVo 8Vo 4Vo 4Vo 67o 4Vo
5 0Vo 07a 0Vo 4Vo 2Vo 27o 27o 2Vo Satisfaction score Mean score nla nla 8.3 7.9 8.7 8.5 8.5 8.6 95Vo CI's nla nla 1.6,8.9 7.2,8.6 8.2,9.2 7.8,9.r 7.8,9.2 8.0,9.2 n/a = not applicable. CI = confidence interval No tests for significance between groups at comparable fotlow-up intervals were significant (ie P>0-05 at allfollow-up intervals)- 181
Endoscopic grading of oesophagitis before surgery was similar, with twelve (2480) of the non-division group and 10 (lgEo) of the division group having complicated reflux disease, demonsffated by either Barrett's oesophagus or stricture formation (P=0.80). A hiatus hernia was seen pre-operatively in 24 (48Vo)of the non-division group versus 28 (54Vo) of the division group (P=0.67). Barium meal examination was performed pre-operatively in
30 patients who did not undergo short gastric division and32 who did. A hiatus hernia was demonstrated by 16 (53V") and 18 (567") examinations respectively.
Preoperative oesophageal manometry outcomes (table 3.2.9) were similar. No statistically significant differences were seen between the groups, although the mean resting lower oesophageal sphincter resting pressule (LOSP) was lower in the nondivision group (6.3 vs 8.3 mmHg, P=0.08). 24hour ambulatory pH monitoring was performed in
22 patientnot undergoing vessel division and24 undergoing division. The mean percentage exposure to an acid pH of less than 4 was I0.}Vo (6.47o,13.7Vo) and lO.3Vo
(6.I%, I4.57o) respectively (P=0.68). r82
Table 3.2.9 Oesophageal manometry results
(figures expressed as 7o of total, or mean (95Vo confídence intervals))
Vessels not divided Vessels divided P value
Preoperative
No. propagated swallows 9.1 (8.6, 9.6) 8.7 (7.9,9.4) O92
Peristaltic amplitude 66.8 (55.3,78.4) 62.8 (5I.1,74.5) 0.59
No. normal peristalsis 897o 82Vo 0.39
LOS resting pressure 6.3 (4.5, 8.1) 8.3 (6.2, 10.3) 0.08
LOS nadir pressure 0.52 (0.27,0.78) 1.06 (0.46, 1.66) 0.46
Resting LOSP < 10 '797o 66Vo 0-18
Post-operative
No. propagated swallows 7.8 (6.ó, 9.1) 8.2 (7.1,9.3) 0.44
Peristaltic amplitude 60.5 (45.2,75.1) 79.8 (61.7,97.9) 0.14
No. normal peristalsis 687o 727o 1-00
LOS resting pressure 24.5 (20.I,28.9) 20.9 (I7.1,24.7) 0.23
LOS nadir pressure I3.3 (9.4, I7 .3) 11.0 (8.4, 13.6) 0.46
Resting LOSP < 10 UVo I07o 0.25
AII pressures measured in mrnHg
Swallowing assessed from I0 wet swallows
3.2.3.2 Surgery
Surgery was performed by one of 7 surgeons. All patients not randomised to undergo division of the short gastric vessels had a loose fundoplication successfully fashioned without resorting to vessel division. One patient randomised to undergo division did not have the short gastric vessels divided, because intra-operative anaesthetic difficulties meant it was necessary to complete the procedure as rapidly as possible. This patient remained in the division group for subsequent analysis. One patient in the division group sustained an intra-operative perforation of the anterior gastric wall due to an injury from a grasping instrument. This was successfully repailed laparoscopically. 183
Between 3 and 8 short gastric vessels (median 5) were divided between metal clips when required. There was some variation between the number of vessels clipped by different surgeons, with some using electrocautery and others clips for smaller vessels. Bleeding sufficient to impair visibility was encountered during short gastric vessel division in 3 patients (6Vo). However, this was overcome in all instances, with no necessity for conversion to open surgery. However, the laparoscopic procedure in 4 (\Vo) of the patients in the division group was converted to open surgery during the stage of oesophageal dissection. The reasons for conversion were; obesity and liver hypertrophy
- 2 patients, large hiatus hernia - 1 patient, and the inability to manipulate an instrument safely behind the oesophagus due to perioesophagitis - 1 patient. All procedures in the non-division group were successt'ully completed laparoscopically.
Operating time varied from 35 to 170 minutes (mean 70.6, median 65, CI's 63.0, 78.1) when the vessels were not divided, versus 59 to 215 minutes (mean I07.9, median 105,
CI's 99.0, 116.8, P<0.0001) when the vessels were divided. The corresponding operating room times were 60 to 185 minutes (mean 95.2,median 91, CI's 87.5, 102.9) versus 75 to
245 (mean 132.7, median 135, CI's 121.7,743.8, P<0.0001). Operating surgeons were asked to rate the difficulty of the operative procedure using a scale from 1 to 10.
Procedures in which short gastric vessels were divided (mean score 6.0; CI's 5.5, 6.6) were perceived to be harder than when the vessels were not divided (score 4.7 , CI's 4-1,
5.4,P=0.O072)
3.2.3.3 Early hospital outcomes
The time periods between surgery and the commencement of oral fluids and solids, and the length of post-operative hospital stay were not altered by division of the short gastric vessels (table 3-2-10). The incidence of post-operative complications was also unaffected by vessel division (table 3.2.II). The majority of complications were minor, and did not impact on later outcomes. However, two patients in the group not undergoing vessel division required laparoscopic revision on the second and fourth post-operative days for 184 acute post-operative paraoesophageal herniation. In both instances a hernia was discovered at a routine barium meal examination on the second post-operative day, facilitating early diagnosis and laparoscopic repair. Three patients in the division group also required surgical revision during the follow-up period. One patient with severe dysphagia underwent re-exploration laparoscopically on the fifth post-operative day. The problem in this instance, tight closure of the diaphragmatic oesophageal hiatus despite calibration of the closure with a 52 Fr bougie, was rectified by removing the top hiatal suture. The second patient underwent a laparotomy 6 hours following the initial procedure for bleeding due to slippage of a clip previously placed across a divided short gastric vessel. The third patient had persistent sevele post-operative dysphagia due to fibrous stenosis of the oesophageal hiatus. Twelve weeks later the hiatus was widened at open surgery, relieving the swallowing difficulty. One patient in the non-division group and2 in the division group also required early flexible oesophagoscopy for disimpaction of a bolus food obstruction, due to inappropriate early consumption of large lumps of meat.
Table 3.2.10 Early hospital outcomes
(figures expressed as mean (957o confidence intervals))
Vessels not divided Vessels divided P value
Postop stay (days) 3.75 (3.16,4.34) 3.93 (3.35,4.52) 0.56
median 3 median 3
Days to oral fluids 1.16 (0.80, 1.51) I.31 (I.03,1.7I) 0.31
median 1 median 1
Days to solids 2.27 (I.82,2.73) 2.50 (2.16,2.84) 0.15
median 2 median2 185
Table 3.2.11 30 day complications
Vessels not divided Vessels divided P value
Urinary retention 10 Minor respiratory 02
Ileus (> 2 days) 32
Paraoesophageal hernia 20
Bleeding short gastric vessel 01 Tight hiatal repair 01 In hospital fall 10
Total number of patients 7 (l4Vo) 6 (rzEo) 0.77
Patients undergoing division of short gastric vessels took an average one and a half weeks longer to return to normal physical activity, possibly due to the conversion of 4 of the procedures to open operations (mean 4.6 weeks (3.4, 5.8) versus 6.3 weeks (4.6,7.9),
P=0.064).
3.2.3.4 One to six month post-operative clinical outcome
A detailed analysis of the outcome of the blinded standardised clinical assessment is summarised in tables 3.2.5,3.2.6,3.2.7 and3.2.8. No differences between the incidence of assessed symptoms in each group were seen at any stage of the initial six month follow-up period, with the exception of a higher incidence of epigastric bloating 3 months after surgery in patients who underwent division of short gastric vessels (table 3-2.5)-
The ability of patients to relieve symptoms of bloat, and their ability to belch was not altered by dividing short gastric vessels. The incidence and severity of heartburn assessed by the visual analogue scale was also identical (table 3.2.6). Outcomes were similar at all follow-up intervals, when assessed by both the visual analogue satisfaction score, outcome scale and modified Visick scale (table 3.2.8). Similarly, the incidence and severity of dysphagia assessed 1, 3 and 6 months after laparoscopic Nissen fundoplication was not altered by division of the short gastric vessels (table 3.2-7). No 186 trend towards improvement in the overall outcome following division of short gastric vessels could be demonstrated by careful analysis of the different symptom scores.
3.2.3.5 Objective post-operative investigations
Eighty (78Vo) patients underwent a post-operative oesophageal emptying study, 66 (65Eo) a barium swallow examination,56 (557o) oesophageal manometry, and 46 (45Vo) underwent post-operative 24 hour pH monitoring. The clinical outcomes in patients who underwent post-operative investigation were similar to the outcomes in those patients who declined investigation.
Barium swallow examination revealed a small asymptomatic paraoesophageal hernia in 4 patients (l2o/o), and delayed emptying of barium from the oesophagus in 2 (6Vo) patients who had undergone division of their short gastric vessels. None of these appearances were evident in patients who had not had the vessels divided, except in one patient with delayed oesophageal emptying. The fundoplication in all but one patient in the non- division group lay compleæly within the abdominal cavity, whereas in7 (ZIVo) of the division group the wrap partly straddled the diaphragm. No fundoplications in either group migrated fully into the thoracic cavity. All fundoplications appeared to be correctly constructed using gastric fundus, irrespective of operative technique.
Oesophageal manometry outcomes following surgery are summarised in table3-2.9.
Oesophageal body motility parameters were similar, whilst mean lower oesophageal sphincter resting and nadir pressures were 3.6 and2.3 mmHg higher respectively in the group of patients who did not undergo division of short gastric vessels, although this difference failed to reach statistical significance. 24hour pH monitoring demonstrated normalisation of acid exposure times in all but 3 patients (one in the non-division group
and 2 in the division group). All of these patients had minimally elevated acid exposure times, and none had any symptom of gastro-oesophageal reflux. 187
The mean oesophageal emptying time measured by the radionuclide method was 109 seconds (86, 132) in patients who did not undergo division of the short gastric vessels, compared to I27 seconds (103, 151, P=0.26) in those whose vessels were divided. 447o of patients undergoing vessel division had a normal emptying time compared to 497o of patients whose vessels were not divided (P=0.82).
3.2.4 DISCUSSION
Division of short gastric vessels during open Nissen fundoplication has been advocated in reports from both DeMeester43 andDonahue50. It has been suggested that this manoeuvre reduces the incidence of post-operative dysphagia, as well as enabling greater relaxation of the lower oesophageal sphincter region during swallowing. However, whilst persuasively argued, scientific data establishing this has been lacking. For instance, Donohue50 studied an uncontrolled series of 77 patients followed for an average of 4.I years, with objective manometric follow-up in only 19 (257o), and DeMeester43 studied a non-randomised series of 100 patients, of whom 36 underwent subsequent post- operative manometric assessment43. All modifications to Nissen's original operation have been introduced and advocated without supporting evidence from any controlled clinical trials. This has led to divergent opinions about whether modifications such as short gastric vessel division really do reduce the incidence of post-operative dysphagia3,38,43,50,65,186. Assessment outside prospective controlled trials often results in the comparison of different groups of patients who undergo surgery by different surgeons at different stages in their experience. This introduces the possibility of unintentional bias, and means that conclusions from such studies should be seen as hypotheses to be tested.
It should be recognised that good clinical outcomes have been reported following total fundoplication, both with and without short gastric vessel division, using both open and laparoscopic techniques3,43,50,65,77 ,186,219. Rossetti reported the 20 year outcomes of 875 patients who underwent fundoplication without division of short gastric 188 vessek186 . 8':..57o achieved a good or excellent outcome. At follow-up of 1 to 13 years
(mean 45 months) DeMeester reported thatgl%o of patients who underwent short gastric vessel division achieved a good or excellent outcome43.
Published outcomes of laparoscopic fundoplication are all short term. However, both
Anvari3 and Geagea65 have described substantial experience with Nissen fundoplication without vessel division, reporting dysphagia rates of 5.4Vo and 07o respectively. This compares with the initial experience of Hinder77 who has advocated routine division of the short gastric vessels, in which 23Vo oT 198 patients experienced dysphagia at early follow-up. Peters and DeMees¡e¡170 who also routinely divide these vessels, report a
9.4Vo incidence of dysphagia 3 months after surgery in their initial 34 patients undergoing laparoscopic Nissen fundoplication.
The difficuþ encountered when comparing these studies is that different procedures are performed by different surgeons, who may be at different stages in their experience.
These problems are compounded by the use of different patient selection criteria, and variation in post-operative assessment methodology. Personal follow-up obtained by an operating surgeon may elicit different outcomes to those obtained independently by a non-surgeon investigator. In earlier experience at the Royal Adelaide Hospital2l9 ¿g well as that of Anvari3 and Geagea65 nearly all patients were offered the laparoscopic
approach, whereas Peters and DeMeesþr168 have recentþ advocated a more selective
approach, excluding patients with Barrett's oesophagus and oesophageal stricture from
consideration. These selection differences are likely to impact on the incidence of post-
operative dysphagia.
The results of this double blind prospective randomised trial minimise the risk of bias
inherent in the non-randomised studies and retrospective reviews published previously.
A significantly higher rate of post-operative manometric and other objective
investigations was achieved compared to these non-randomised studies43,50. The
current trial demonstrated no significant differences between the study groups. Whilst 189 some minor differences in individual dysphagia scores are apparent, when all criteria used for dysphagia assessment are considered together, the overall results reveal no trend towards an improved outcome in either group, nor any signifi.cant difference in lower oesophageal sphincter pressure, oesophageal emptying time, or barium meal outcome. It might be argued that a trend of difference has been established between the groups in certain parameters measured at 6 months, favouring division of short gastric vessels; e.g. dysphagia for lumpy solids (337o versus 29Vo),lower oesophageal sphincter pressure
(24.5 versus 20.9), lower oesophageal sphincter nadir pressure (13.3 versus 11.0 mmHg), and that these figures may have reached significance in a larger study. Nevertheless, the other dysphagia scores and patient satisfaction were clearly comparable between the two groups, and oesophageal empfying was quicker in the non-division group. If there is any difl'erence, it is likely to be marginal. The overall outcome scores also were not influenced by division of the short gastric vessels. The construction of a loose wrap is of probably more importance for minimising the incidence of post-operative dysphagia, than whether short gastric vessels are divided or not.
The ability of patients to belch following surgery was not improved by dividing the short gastric vessels. Whilst approximately 40 to 50Vo of patients claimed they were unable to belch 6 months after surgery, this apparently high incidence reflects the supercompetent valve produced by the Nissen fundoplication. Other work from the Department of
Surgery at the Royal Adelaide Hospital suggests that effective belching is unlikely following the Nissen procedureS9. Patients who claim the ability to belch usually report oesophageal belching rather true gastro-oesophageal reflux of gas.
Whilst the overall proportion of patients who were scored Visick I or 2 for their clinical outcome following surgery was similar for this trial and the study reported in 3.1, the proportion scoring Visick grade 1 was significantly less than in the previous trial
(approximately 25Vo versus 70Vo). This is likely to be due the different follow-up methodology used for the two studies. Follow-up in the current study was obtained by a blinded non-clinical investigator, whereas in 3.1 it was obtained by a surgical registrar. 190
This highlights the possibility of different follow-up methodologies influencing measured clinical outcomes. This can be a signifîcant problem in non-randomised trials in which different investigators assess outcomes of different patient groups. However, in randomised trials where follow-up is obtained in a uniform fashion, a valid comparison of different surgical techniques can still be made.
Dividing short gastric vessels was associated with a prolongation of operating time by approximately 40 minutes, resulting in increased expense and technical difficulty. The one patient requiring early re-intervention for bleeding demonstrates the added potential for intra-operative and post-operative haemorrhage following division of these vessels.
Following short term clinical follow-up at 6 months and objective investigation 3 to 4 months after surgery, this trial has failed to show any reduction in the incidence or severity of dysphagia following division of short gastric vessels during laparoscopic
Nissen fundoplication.
Early correction of reflux symptoms was identical for the two groups, but longer term follow-up will be needed to assess the durability of the two operations and the incidence of recurrent reflux. For the present it is concluded that division of short gastric vessels during laparoscopic Nissen fundoplication is only indicated in the uncommon circumstance that a loose wrap of fundus cannot be constructed. 191
3.3 PROSPECTIVE DOUBLE BLIND RANDOMISED TRIAL OF LAPAROSCOPIC ANTERIOR PARTIAL VERSUS V FUNDOPLICATION 'V'SSE
3.3.1 AIMS
To determine whether the routine performance of a laparoscopic anterior partial fundoplication can reduce the incidence of post-operative dysphagia or other adverse outcomes following laparoscopic Nissen fundoplic ation.
3.3.2 METHODS
The protocol for this trial was similar to that in the study described in section 3.2
3.3.2.1 Participant assignment
Patients undergoing laparoscopic surgery for gastro-oesophageal reflux disease were randomised to undergo either a Nissen fundoplication or an anterior partial fundoplication using laparoscopic techniques. Informed consent was obtained from all participants, and randomisation occurred in the operating theatre, after the commencement of general anaesthesia, by opening one of 120 previously sealed opaque envelopes. Preparation of the envelopes ,was undertaken before the study by a research officer not directly involved with the trial, and envelopes were selected and opened by a departmental secretary, at the operating surgeon's request.
3.3.2.2 Patientselectionandpreoperativeinvestigation
All patients with proven gastro-oesophageal reflux disease, who presented for primary
antireflux surgery by the laparoscopic technique were considered for entry into this trial.
Patients were excluded only if they had an oesophageal motility disorder which precluded
a 3600 fundoplication, or required a concurrent abdominal procedure at the same time as fundoplication (eg cholecystectomy), or had undergone previous antireflux surgery. All 192 patients underwent preoperative investigation with oesophageal manometry and endoscopy. Preoperative manometric testing also included an acid reflux provocation test and Bernstein test. Twenty four hour pH monitoring was performed routinely for patients who did not have unequivocal reflux disease demonstrated by preliminary endoscopic and manometric studies. The majority of patients also underwent pre- operative barium meal X-ray examination.
3.3.2.3 Operating technique
Laparoscopic Nissen fundoplication was performed using the technique described in section 1.2.2.2. This comprised routine posterior hiatal repair and the performance of a loose Nissen fundoplication around a 52 Fr bougie. The short gastric vessels were not divided.
The initial steps for the laparoscopic anterior fundoplication were similar to the Nissen procedure, commencing with hiatal dissection and oesophageal mobilisation followed by posterior hiatal repair. The hepatic branch of the vagus nerve was routinely preserved.
A 1800 anterior partial fundoplication was fashioned by fixing the anterior wall of the fundus to the front of the anterior oesophagus and the diaphragmatic hiatus. This was achieved by suturing the fundus to the right lateral wall of the abdominal oesophagus and to the right hiatal pillar and the posterior hiatal repair, and the left lateral wall of the oesophagus to the left hiatal pillar using 5 or 6 2/0 Novafil or Prolene interrupted sutures.
Short gastric vessels were left intact. The anterior partial fundoplication accentuated the angle of His, repaired the hiatus, anchored a 3 to 5 cm length of oesophagus within the abdomen and fashioned a partial fundoplication which was sutured to the oesophagus and the hiatal ring.
If the laparoscopic procedure was converted to an open procedure due to intra-operative difficulties, the randomisation schedule was still followed, with the patient remaining in the trial for subsequent analysis. 793
3.3.2.4 Post-operative care
Nasogastric tubes were not used in any patients. Patients were allowed oral fluids post- operatively on the evening of the day of surgery, and soft solid food the next day.
Discharge from hospital was encouraged after the second post-operative day. Patients were instructed to avoid bread and lumpy foods for the first 3 to 4 weeks following surgery, and then to gradually increase the consistency of their diet. A barium meal examination was routinely obtained on the second post-operative day, to detect any problems amenable to early laparoscopic reintervention.
3.3.2.5 Masking
The extent of fundoplication performed laparoscopically (Nissen or anterior) was concealed trom all patients. As patients had no direct access to casenotes or trial records, and both laparoscopic procedures utilised identical operative wounds, all remained unaware of the exact procedure for the duration of the trial follow-up period. All follow- up was obtained by a scientific officer who was 'blinded' to the randomisation of each patient. He remained unaware of the allocated group for each patient throughout the follow-up period. Participant data was entered into a computerised data base by another research assistant who was not involved in direct patient follow-up.
3.3.2.6 Clinical follow-up
Patients were interviewed pre-operatively and then 1, 3 and 6 months after surgery using a structured questionnaire similar to the one used in the study described in section 3.2.
Longer term follow-up was not available for reporting in this thesis. The presence or absence of each of the following symptoms was sought; heartburn, epigastric pain, regurgitation, dysphagia for lumpy solids, soft solids and liquids, odynophagia, early satiety, inability to belch, epigastric bloating, ânorexia, nausea, vomiting, nocturnal coughing and wheezing. The ability to relieve bloating and whether a normal diet was being consumed was also determined. Heartburn was also scored using a visual analogue scale (0 = no heartburn, 10 = severe heartburn). 194
Dysphagia was scored by several methods. Visual analogue scales (0 = no dysphagia, 10
= total dysphagia) were independently applied for solids and liquids, as well as the previously validated score (0 = no dysphagia,45 = severe dysphagia) which combines information about difficulty swallowing 9 types of liquids and solids (see Table 3.2.I).
Overall outcome was determined using 3 further scales. Patients ranked the outcome of surgery using a modihed Visick grading (Table 3.2.2), and were asked to score the outcome as excellent , good, fair, or poor (Table 3.2.3). An overall assessment of satisfaction with the operative outcome was also scored by further visual analogue scale
(0 = dissatisfied, 10 = satisfied).
3.3.2.7 Objective follow-up
Objective investigation with oesophageal manomefty,24 hour pH monitoring, upper
gastrointestinal endoscopy, and a radionuclide oesophageal emptying study was performed 3 to 4 months following surgery. Investigation sought to assess lower
oesophageal sphincter function, control of reflux, post-surgical anatomy, healing of
oesophagitis, and the presence of any post-surgical oesophageal obstruction.
3.3.2.7.1 Oesophøgeal Manometry
The technique used was identical to that described in section 3.2.7.1.
3.3.2.7.2 Ambulatory 24 hour pH monítoriltg
The technique used was identical to that described in section 3.2.7 -2.
3.3.2.7.3 Rødíonuclíde OesophøgeølEmptyírtg Study
The technique used was identical to that described in section 3.2.7.3
3.3.2.7.4 Upper GøstroíntestinøI Endoscopy
Post-operative upper gastrointestinal endoscopy was performed in the endoscopy suite at
the Royal Adelaide Hospital by one of three surgeons who were unaware of which type of
fundoplication had been performed previously. The presence or absence of visible 195 oesophagitis was determined and scored using the Savary Miller grading system (see
Table I.2.6.I). The appearance and integrity of the fundoplication, and the presence of any concurrent para-oesophageal hiatus hernia was also determined.
3.3.2.8 Statistical analysis
The primary clinical outcomes which the trial was designed to evaluate were post- operative dysphagia and control of reflux symptoms. Before commencing the trial it was determined that 84 patients (42 in each group) would be needed to demonstrate a20Vo difference in these outcome measures, at a significance level of P<0.05, and power of
90Vo. To ensure that this was achieved, it was intended that at least 100 patients would be recruited, allowing for an estimated 20Vo of all patients refusing the objective post-
operative investigations. All analyses were performed on an intention to treat basis, with
all patients remaining in their initial allocated group for this analysis.
Before commencing the trial, it was intended to report the initial outcomes and results of
post-operative testing once patients had been followed for 6 months. This time period is
thought to allow for the adequate assessment of any differences in the incidence of post-
operative dysphagia between the two trial groups. Medium to long term outcomes are
more important for determining the efficacy of reflux control, and will be reported once
follow-up has matured.
Data were entered onto a computerised data base (Filemaker Pro version 3.0) and
analysed using a commercially available statistical package (InStat version 2-0I).
Fisher's Exact Test was used to determine the significance of 2 x 2 contingency tables. A
two tailed Mann-Whitney U-Test, was used to assess the significance of non-parametric
sets of data, and an unpaired Student T-Test to determine the significance of sets of data
where it was reasonable to assume a parametric distribution (height and weight).
Statistical significance was accepted at aP value of less than 0.05. Unless otherwise
stated all data is reported as the percentage of the total patients in each group, or as the
mean (957o confidence intervals). 196
3.3.2.9 Ethical approval
The protocol for this study was approved by the Royal Adelaide Hospital Human
Research Ethics Committee.
3.3.3 RESULTS
From December 1995 to April 1997,107 patients undergoing laparoscopic antireflux surgery were entered into the trial. Fifty three patients were randomised to undergo a
Nissen fundoplication, and 54 to undergo an anterior fundoplication. During the same period 70 further patients underwent a laparoscopic Nissen fundoplication performed by surgeons contributing patients to this study. Three of these patients were excluded because of the need to perform a concurrent laparoscopic cholecystectomy procedure.
Twenty patients were not entered as the procedure was to be performed by an inexperienced surgeon-in-training, 13 had undergone previous antireflux or gastric surgery, 8 had a giant hiatal hernia, rather than reflux as the primary indication for surgery, and 13 were from remote areas (more than 500 km away), restricting their ability to participate in the follow-up protocol. The remaining 16 patients either refused entry into the trial because they expressed a preference for a specific procedure to be performed, or they were unwilling to participate in the follow-up protocol. Of the 107 patients entered, 106 (99Vo) were available for follow-up one month after surgery, 104
(97Eo) at 3 months, and 95 (89V") at 6 months. Although prospectively collected follow- up details were not available for a small number of patients at the specific follow-up intervals, no patient elected to withdraw from the study. Missing data was due to an inability to contact patients at the specific follow-up intervals. Twelve patients underwent surgery within 6 months of analysis of data, and consequently have not yet contributed clinical data at the last follow-up interval. Figule 3.3.1 summarises the follow-up compliance within the trial. 197
Figure 3.3.1 Compliance with follow-up protocol
107 patients randomised intraoperatively
Anterior fundoplication Nissen fundoplication 54 patients 53 patients
1 month followup I month followup 54 patients (l00Vo) 52 patients (98Vo)
3 month followup 3 month followup 52 patients (96Vo) 52patients (98Vo)
6 month followup 6 month followup 47 patients (87Vo) 48 parients (9170)
Objective investigations Objective inves tigation s Endoscopy 25 patients (46Vo) Endoscopy 21 patients (40Vo) emptying 39 patients (72Vo) Oesophageal emptying 33 patients (62Vo) manometry/pH 32 patients (597o) Oesophageal manometry/pH 30 patients (57Vo)
3.3.3.1 Preoperative assessment
Both groups were similar for age, sex, height, weight, cigarette and alcohol consumption,
incidence of previous abdominal surgery, duration of symptoms, and medications
consumed before surgery flable 3.3.1). Analysis of the presence or absence of
preoperative symptoms (Table 3.3.2), as well as the assessment of heartburn using the
visual analogue scale (Table 3.3.3), revealed no significant differences. A significant
proportion of patients in each group experienced preoperative dysphagia to some extent,
with an incidence of 33Vo in the anterior fundoplication group and 367o in the Nissen
group when assessed using a score of 0 for the 0 to 45 dysphagia score (Table 3.3.4)-
Whilst different methods of scoring dysphagia elicited slightly different rates, there was
no clinically or statistically significant difference between the two groups. Preoperative
Visick grading was also similar for each group (Table 3.3.5). 198
Tabte3.3.1 Preoperativeparameters
(all figures expressed as mean (95Vo confidence intervals))
Anterior Nissen P value
Number of patients 54 53
Age (years) 44 (40,49) 46 (42, 50) 0.35 Sex 34male,20 female 36 male, 17 female 0.69 Height (cm) I72 (169,176) I7I (167,174) 0.54
Weight (kgs) 82 (77,86) 86 (81, 90) o.24
Cigarette smoker 3IVo l97o 0.30
Alcohol consumed 677o 657o 1.00
Previous abdominal surgery 4LVo 34Vo 0.55
Duration symptoms (yrs) 9.4 (6.9, 11.9) 9.5 (7.1, 11.9) 0.74
Preoperative medications
Omeprazole 92Vo 83Vo 0.15
H2 Blocker 85Vo 79Vo 0.46 Table 3.3.2 Summary of pre and post-operative symptoms
Pre-operøtíve 7 month postop 3 months postop 6 months postop Anterior Nissen Anterior Nissen Anterior Nissen Anterior Nissen Heartburn 947o 94Vo 77a 27o 67o 8Vo l37o lÙVo Epigastric Pain 69Vo 68Va 3l7o lTVo ll%a l9Vo 26Va 337o Regurgitation SlVo TlVo 9vo* Tvo* 6Va 4Vo 6Vo 2Vo Odynophagia 26Vo LlVo 9Va 6Vo 2% 2Va 27o 2Vo Early satiety 35Vo 32Vo 54Vo 48Vo 21Vo 27Vo 79Va I9Vo Epigastric bloat 527o 49Va 24Vo 23Va 2IVo lTVo I9Vo 2lVo Anorexia 19Va 9Vo lTvo** 4vo** IÙVo 2Va 2Vo 2Vo Nausea 3lVo 327o I17o lïVo 2Vo 67o 67o 2Vo Vomiting 22Vo 23Vo 77o 67o 2Vo 4Vo 2Vo 2Vo Nocturnal cough 377o 40Vo 97o 4Vo lïVo 2Vo I3Vo 67o Noctumal wheeze 177o lTVo 6Vo 07o 4Vo 4Vo l37o 47o Can relieve bloat 80Vo 827o 70Vo 72Vo 84Vo 887o 83Vo 687o Unable to belch 0Vo 07o 30Vo 35Vo rivo* 35vo* 17vo* 35vo* Eats normal diet 63Vo 68Vo 677a 67Vo 92Vo 887o 93Vo 857o Increased flatus nla nla 43Vo 29Vo 50Vo 447o 3ovo* 5ovo* Diarrhoea nla nla 67o 47o l27o 4Vo 9Vo 47o
No significant dffirences dennnstratedbetween trial groups (ie P>0.05 at allfollow-up intervals), ** * P= 0.06, P= 0.052 200
Table 3.3.3 Assessment of heartburn by visual analogue scale
(figures expressed as mean (957o confidence intervals))
Anterior Nissen P value Preoperative 4.3 (3.4,5.1) 4.1 (3.1,5.0) 0.69
1 month post-operative 0.5 (0.03, 1.0) 0.1 (-0.1, 0.3) 0.52
3 months post-operative 0.2 (-0.2,0.6) 0.3 (-0.1, 0.7) 0.74
6 months post-operative 0.4 (-0.04, 0.9) 0.2 (-0.04,0.4) 0.70 Table 3.3.4 Dysphagia assessment
Pre-operalive I monÍh postop 3 monlhs postop 6 monlhs postop Anterior Nissen Anterior Nissen Anterior Nissen Anterior Nissen Dysphagia for
Lumpy solids 19Vo 34Va 63Vo 65Vo 33Va 46Va 75Vo* 38Vo+
Soft solids 4Va llVo 7Vo 2Va 2Vo 2Vo 2Vo 2Vo
Liquids 4Vo llVo 7Vo l5Vo 6Vo 6Vo ÙVo 4Vo Visual analogue scale Solids 1.5 (0.8,215) 1.4 (0.7,2.1) 2.7 (2.0,3.4) 3.0 (2.2,3.8) 1.1 (0.6, 1.6) 1.6 (1.1,2.2) 0.5 (0.04, 1.0)T 1.1 (0.6, 1.6)T Liquids 0.5 (0.1,0.9) o.7 (0.2,1.1) 0.2 (0.01,0.4) 0.5 (0.2,0.8) 0.1 (-0.02,0.3) 0.1 (-0.02,0.3) 0.0 (0.0,0.0) 0.2 (0.05,0.51) Dysphagia score Overall result 6.7 (3.6,9.7) 5.6 (2.9,8.2) 9.7 (7.1,72.2) tt.3 (8.4,14.2) 4.2 (2.2,6.1) 5.6 (3.6,7.7) 1.9 (0.1, 3.7)** 4.2 (2.3,6.2)**
Scored 0 only 67Vo 64Vo 347o 33Vo 67Vo 54Vo 85Vo+ 63Vo*
Figures are either percentage of tutu\, or medn (95Vo confidence inîervals) + = P=0.02, ** = P=0.05 & f = P-0.07. No other tests Íor significance between groups were significant (ie P>0.05 at all follow-up intervals). Table 3.3.5 Outcome, satisfaction, and Visick grading
Pre-operatíve 7 month postop 3 months postop 6 months postop Anterior Nissen Anterior Nissen Anterior Nissen Anterior Nissen Outcome Excellent nla nla l5Vo l5Vo 527o 25Vo 5I7o 29Vo Good nJa nla 65Vo 69Vo 42Vo 63Vo 38Vo 54Vo Fair nla nla I6Vo I4Vo 4Vo 6Vo 9Vo l5Vo Poor nla nla 4% 2Vo 2Vo 67a 27o 2Vo Visick grade
1 0Vo jVo 97o l2Vo 40Vo 23Vo 5IVo 29Vo
2 0Vo 0Vo 6lVo 63Vo 54Vo 63Vo 32Vo 44Vo a J 3r% 23Vo l5Vo ITVo 27o 6Vo llTo 197o
4 69Vo 777o 5Vo 67o 2Vo 2% 4Vo 67o
5 0Vo 0Vo 47o 27o 2Vo 6Vo 27o 2Vo Satisfaction score Mean score nla nla 8.3 8.3 8.8 8.3 8.8 7.9 957o CI's nla nla 1.7,8.8 7.8, 8.9 8.2,9.3 7.7,9.0 8.3,9.3 7.r,8.6 n"/a = not applicable. CI = confrdence interval No tests for significance between groups at comparable follow-up intervals were signifrcant (ie P>0-05 at allfollow-up intervals), except outcome assessments at 3 and 6 months ( j months, P= 0.02, 6 months, P=0.04; Chi-squared test , Excellent vs Good vs Fair/Poor)- 203
Endoscopic grading of oesophagitis before surgery was similar, with 13 (24Vo) of the
Nissen group and 11 (217o) of the anterior fundoplication group having complicated reflux disease, demonstrated by either Barrett's oesophagus or stricture formation (P=
0.S1). A hiatus hernia was seen pre-operatively in 23 (43Vo) of the Nissen group versus
27 (50Vo) of the anterior fundoplication group (P= 0.55). Barium meal examination was
performed pre-operatively in 31 patients who underwent Nissen fundoplication and 28
who underwent anterior fundoplication. A hiatus hernia was demonstrated by 19 (6IEo)
and I7 690) examinations respectively.
Preoperative oesophageal manometry outcomes (Table 3.3.6) were similar. No
statistically significant differences were seen between the groups, although the mean
resting lower oesophageal sphincter resting pressure (LOSP) was lower in the anterior
fundoplication group (6.4 vs 8.6 mmHg, P= 0.36). 24hour ambulatory pH monitoring
was performedln34 patients undergoing Nissen fundoplication and 35 undergoing
anterior fundoplication. The mean percentage exposure to an acid pH of less than 4 was
12.6 Vo (9.0 Vo, 16.2 Vo) and I2.2Vo (9.17o, 15.3Vo) respectivelY (P= 0.84 ). 204
Table 3.3.6 Oesophageal manometry results
(figures expressed as 7o of total, or mean (95Vo confidence intervals))
Anterior Nissen P value
Preoperative
No. propagated swallows 8.9 (8.4,9.5) 8.6 (8.0, 9.2) 0.42
No. normal peristalsis 837o 82Vo 1.00
LOS resting pressure 6.4 (4.8,8.1) 8.6 (6.1, 11.1) 0.36
LOS nadir pressure 0.7 (0.2,I.2) 1.4 (0.6,2.2) 032
Resting LOSP < 10 l57o 67Vo 0.39
Post-operative
No. propagated swallows 8.2 (7.2,9.2) 8.7 (7.7,9.6) 0.55
No. normal peristalsis 75Vo 907o 0.19
LOS resting pressure I8.3 (14.2,22.4) 28.9 (23.3,34.5) 0.004
LOS nadir pressure 5.6, (3.4,7.8) 12.8 (9.9,15.7) 0.0002
Resting LOSP < 10 25Vo 37o 0.03
AII pressures measured in mmHg
Swallowing assessed from l0 wet swallows
3.3.3.2 Surgery
Surgery was performed by one of 7 surgeons. Two patients randomised to undergo
anterior fundoplication underwent a Nissen procedure, and one patient randomised to a
Nissen fundoplication, eventually underwent an anterior fundoplication. The patient who
initially underwent a Nissen fundoplication, underwent reoperation for dysphagia 23 days
later by an open surgical approach. The Nissen fundoplication was converted to an
anterior 1390 partial fundoplication. The oesophagus in one of the patients randomised to
undergo anterior partial fundoplication was perforated posteriorly during initial
laparoscopic dissection, because of dense peri-oesophagitis which resulted in difficuþ
defining the correct dissection plane. This procedure was converted to an open operation,
the oesophageal tear repaired, and the procedure completed as a Nissen fundoplication, to 205 ensure that the oesophageal repair was reinforced by the gastric wall. The other patient in the anterior fundoplication group developed an acute post-operative para-oesophageal hiatus hernia, requiring laparoscopic reoperation on the 3rd post-operative day. At this
procedure the fundoplication was reconstructed as a Nissen procedure, as the hiatal rim to
which the anterior fundoplication would usually be anchored, was torn and thought to be
inadequate to anchor the anterior fundoplication. All three of these patient remained in
the group to which they were originally randomised for subsequent analysis.
Minor intra-operative difficulties due to minor bleeding from the liver occurred during 4
Nissen and2 anterior fundoplications. Minor bleeding also occurred from the omentum
and from the hiatal rim during 2 further anterior fundoplication procedures. The left
pleural membrane was perforated during another anterior fundoplication, although this did
not cause any intra-operative or post-opslative problems.
The laparoscopic procedure in 1 (2 Vo) of the patients in the Nissen group was converted
to open surgery during the stage of oesophageal dissection. This was due to difficulty
dissecting the oesophagus within a very large hiatal hernia. Four (77o) of the anterior
fundoplication procedures were converted to an open procedure due to one ofeach ofthe
following problems; large hiatus hernia, obesity, oesophageal perforation (discussed
above), and bleeding from the liver. The last patient had a large fatty left liver lobe,
which split when the liver was retracted in the usual fashion.
Operating time varied from32 to 184 minutes (mean 62,median 58 , CI's 55,70) for
Nissen fundoplication, versus 35 to 144 minutes (mean 69 , median 60 , CI's 61,76 ,P=
0.10) for anterior fundoplication. The conesponding operating room times were 55 to
205 minutes (mean 88, median 78, CI's 79,97) versus 60 to 160 (mean 91, median 80,
CI's 84, 98, P= 0.I2). Operating surgeons werc asked to rate the difficulty of the
operative procedure using a scale from I to 10. Nissen procedures (mean score 4.7; CI's
4.I, 5-3) were perceived to be of equal difficulty to anterior partial fundoplication (score
4.6, CI's 4.0,5.2, P= 0.88) 206
3.3.3.3 Early hospital outcomes
The time periods between surgery and the commencement of oral fluids and solids, and the length of post-operative hospital stay were not altered by the type of fundoplication performed (Table 3.3.7). The incidence of post-operative complications was slightly higher in the anterior fundoplication group, although many of the complications seen were minor and did not impact significantly on these patients'recovery or later outcomes
(Table 3.3.8). However, two patients who underwent an anterior partial fundoplication required laparoscopic revision on the third post-operative day, one for acute post- operative paraoesophageal herniation, and one for severe post-operative dysphagia. The latter patient experienced dysphagia due to the posterior hiatal repair closing the hiatus too tightly. This problem was corrected laparoscopically by undoing the fundoplication, removing the highest hiatal repair suture, and then refashioning an anterior fundoplication. One patient (discussed earlier) underwent open revision of a Nissen fundoplication to an anterior partial fundoplication on the 23rd post-operative day due to significant early post-operative dysphagia. A further patient who underwent a Nissen fundoplication developed a left subphrenic collection on the third post-operative day.
This was drained percutaneously with radiological assistance. A barium contrast study demonstrated no communication with the gut lumen. 207
Table 3.3.7 Early hospital outcomes
(flrgures expressed as mean (95Vo confidence intervals))
Anterior Nissen P value
Postop stay (days) 3.3 (2.8,3.7) 3.1(2.9,3.4) 0.68
median 3 median 3
Days to oral fluids I.2 (0.9, I.4) 1.1 (1.0, 1.3) 0.80
median 1 median 1
Days to solids 2.0 (I.7,2.4) 2.0 (I.8,2.3) 0.50
median 2 median 2
Table 3.3.8 30 day complications
Anterior Nissen P value
Urinary retention 1 0
Minor respiratory 2 0
Pneumothorax 1 0
Paraoesophageal hernia 1 0
Dysphagia 1 1
Subphrenic collection 0 1
Total number of patients 6 (ll%o\ 2 (4Vo) 0.27
Patients in both groups resumed normal physical activity on average 3 weeks after fundoplication.
3.3.3.4 One to six month post-operative clinical outcome
A detailed analysis of the outcome of the blinded standardised clinical assessment is summarised in tables 3.3.2,3.3.3,3.3.4. and 3.3.5. No differences between the incidence of assessed symptoms in each group were seen at the one and three month follow-up 208 intervals, with the exception of the outcome assessment and Visick scores 3 months following surgery (Table 3.3.5). Patients who underwent an anterior fundoplication were
more likely to report an excellent outcome and a Visick 1 grading than the Nissen
fundoplication group. However, this was not due to an increased number of patients in
the Nissen group reporting a poor result, but due to more of the patients who underwent
an anterior fundoplication rating the operation 'excellent', and a Visick grading of 1.
Overall a similar number of patients in both groups rated their outcome good or excellent
or Visick I or 2.
At the six month follow-up interval a number of significant differences became apparent,
with the overall outcome favouring the anterior partial fundoplication. The incidence of
dysphagia tbr lumpy solid food, as well as the 0 to 45 dysphagia score were significantly
less following anterior fundoplication, and a greater proportion of patients reported an
excellent outcome following surgery (P= 0.04). Trends to improved outcomes in the
anterior fundoplication group were also seen for; belching ability (P= 0.06), frequency of
flatus passed per rectum (P=0.06), solid food visual analogue score (P= 0.07), satisfaction
score (P= 0.09), and willingness to undergo similar surgery again (P= 0.09).
Both fundoplication variants controlled reflux equally effectively, with the heartburn
scores similar at all follow-up intervals (Table 3.3.3) and the incidence of reflux
symptoms similar (Table 3.3.2). Two months following an initially successful anterior
fundoplication, one patient developed recurrent reflux disease, objectively proven by
endoscopy and24 hour pH monitoring. This patient was planned for surgical revision to
a Nissen fundoplication, but had not undergone further surgery by the six month follow-
up interval. Although two patients in the Nissen group described recuffent reflux
symptoms, neither had any demonstrable abnormality at either endoscopy examination or
24hour pH monitoring. Overall one patient developed recurrent reflux within 6 months
of an anterior fundoplication, and no patients following Nissen fundoplication. 209
Patients in the anterior fundoplication group were more likely to express a willingness to undergo their procedure again, should similar preoperative circumstances arise (Table
3.3.9).
Table 3.3.9 Willingness to undergo similar surgery again in similar
circumstances
Follow-up interval Anterior Nissen P value l month 89Vo 9jVo 1.00
3 months 967o 87Vo 0.16
6 months 96Vo 83Vo 0.09
3.3.3.5 Objective post-operative investigations
72 (67Eo) patients underwent a post-operative oesophageal emptying study, 46 (43Eo) upper gastrointestinal endoscopy examination, 62 (587") oesophageal manometry, and 48
(45Vo) underwent post-operattve 24 hour pH monitoring.
Upper gastrointestinal endoscopy examination revealed a small asymptomatic sliding hiatus hernia in 2 patients following Nissen fundoplication and one following anterior fundoplication. 19 (76Vo) of the anterior fundoplication patients had no endoscopic evidence of oesophagitis at follow-up endoscopy ,2 (87o) had persistent Barrett's oesophagus, | (47o) had a stricture requiring endoscopic dilation (also present preoperatively), and 3 (lzEo) had evidence of mild oesophagitis (Savary Miller grade 1).
In the Nissen fundoplication group, 19 (907") had no evidence of oesophagitis, L (5Vo) had persistent Barrett's oesophagus, and I (57o) had evidence of mild oesophagitis
(Savary Miller grade 1). All fundoplications appeared to be correctly constructed using gastric fundus, irrespective of operative technique.
Oesophageal manometry outcomes following surgery are summarised in Table 3.3-6.
Oesophageal body motility parameters were similar, whilst mean lower oesophageal 2t0 sphincter resting and residual relaxation pressures were significantly higher in the group of patients who underwent laparoscopic Nissen fundoplication. 24 hour pH monitoring demonstrated normalisation of acid exposure times in all but 6 (I3V") patients (3 in the
Nissen group and 3 in the anterior fundoplication group). All of but one of these patients was free of any Symptom of gastro-oesophageal reflux, and had no evidence of oesophagitis at endoscopy. The remaining patient, who had previously undergone an anterior fundoplication, had symptomatic and endoscopic evidence of recurrent reflux disease.
The mean oesophageal emptying time measured by the radionuclide method was 90 seconds (69, 110) in patients in the anterior fundoplication group, compared to 114 seconds (87, 142, P= 0.17) for the Nissen group. 6'7Vo of patients undergoing an anterior fundoplication had a normal emptying time compared to 55Vo of patients who underwent a Nissen fundoplication (P= 0.34).
3.3.4 DISCUSSION
The question of which fundoplication technique offels the best outcome for patienls undergoing surgery for gastro-oesophageal reflux disease has been 'Whilst, controversiv¡43,76,205,2II. uncontrolled studies have reported good results following both laparoscopic and open surgery for both Nissen, anterior and posterior fundoplica¡isn3,43,76,80,170,184,205,21I,212,217 ,these outcome studies have done little to resolve this controversy. A number of studies which compared the outcome of a partial fundoplication with historical expedence suggest advantages for the partial fundoplication variants141,166. However, the results reported may be subject to bias, due to the methodology used. The three randomised trials of Nissen versus posterior partial fundoplication from the open erallg,203,209, aswell as a recent laparoscopic study reported by Laws s¡ ¿1109, all failed to demonstrate a significant reduction in the incidence of dysphagia following the partial fundoplication technique. 211
An anterior partial fundoplication was chosen for comparison with the Nissen procedure in this randomised trial, because it had not been previously tested within a prospective randomised trial, and published experience with a similar technique suggested that it may have important advantages over the Nissen procedure. Experience reported by 'Watson46'212 suggests that the anterior partial fundoplication should result in a significant reduction in the incidence of post-operative dysphagia, improved ability to belch, and less risk of gas bloat symptoms following antireflux surgery. It is possible that the dysphagia sometimes seen following both Nissen and posterior fundoplication is due to the piece of fundus pulled posteriorly behind the oesophagus, which may angulate the oesophago-gastric junction forward, a phenomenon which does not occur following anterior partial fundoplication.
The results of this trial revealed no overall outcome differences between the two procedures at the 1 and 3 month follow-up intervals. However, at 6 months patients undergoing laparoscopic anterior partial fundoplication were less likely to experience dysphagia for solid food, were less likely to be troubled by excessive passage of flatus, and were more likely to be able to belch normally. Overall satisfaction and outcome scores were better following anterior fundoplication at 6 months, and patients were more likely to express a willingness to undergo the same procedure again, if confronted with similar preoperative circumstances. It is surprising that these differences were not apparent at the earlier follow-up intervals.
Endoscopic and ambulatory pH monitoring study outcomes were similar, with only one study patient, who underwent an anterior fundoplication, developing objectively proven recurrent reflux within the initial 6 month follow-up period. Whilst most endoscopies revealed complete healing of oesophagitis, slightly more patients had persistent Savary
Miller grade 1 oesophagitis following partial fundoplication (3 versus 1), and the mean acid exposure time at pH monitoring was slightly higher (2.5Vo versus 1.17o). This suggests that the anterior partial fundoplication may not be as effective for the prevention of reflux as the Nissen fundoplication, although within the trial all but one patient had 2t2 adequate control of pathological reflux. It may be that the Nissen fundoplication creates an overcompetent valve at the gastro-oesophageal junction, and that the anterior partial fundoplication restores the gastro-oesophageal junction to a more physiological state.
Nevertheless, despite the encouraging outcomes at 6 months following anterior partial fundoplication, this trial has not fully resolved which procedure will be the most appropriate in the long term. The anterior fundoplication's durability remains unproven by this study. If its durability is eventually shown to be as good as the Nissen fundoplication, then the results of this study will support the previously proposed advantages of less dysphagia, improved ability to belch and less adverse sequelae.
However, it is also possible that the long-term incidence of recurrent reflux following anterior fundoplication may be higher than that seen following Nissen fundoplication, in which case a risk versus benefit assessment will need to be made for each individual, balancing the risk of recurrent reflux against the risk of other adverse outcomes. Long term follow-up will be needed before this issue can be fully clarified. 2t3 214
SECTION 4
CONCLUSIONS 2t5
The studies described in this thesis have investigated whether the laparoscopic approach to antireflux surgery is better than the previous technique performed through an open abdominal incision, and whether procedural modifications, specifically division of short gastric vessels, and the performance of a partial fundoplication, improve subsequent clinical outcomes.
4.I IS LAPAROSCOPIC ANTIREFLUX SURGERY ADVANTAGEOUS?
4.1.L Outcome of case series studies
Published studies in the surgical literature suggest that short term clinical outcomes are improved following laparoscopic antireflu* surgery3,19,70,Jl,I43'143'228, although no studies have been reported which describe the follow-up of patients for more than 3 years.
The initial optimism which accompanied many early reports of virtually no morbidity following laparoscopic procedures3g,55,64,I42,l43, has been tempered by the recent outcomes of more rigorous s¡u¿iss3,24,70,J7 ,80,217 which do acknowledge that the laparoscopic approach is associated with some morbidity. Early reports were often associated with very limited follow-up, and post-operative assessment was usually made by the operating surgeon, rather than an independent investigator. The outcome of larger, independently assessed series, with longer follow-up has usually been associated with a limiæd, but significant number of adverse outcomes3,15,7J,80,90,195. Nevertheless, the overall morbidity of surgery appears to be improved when compared to historical experience, suggesting significant advantages for laparoscopic approaches 23,43,60,1 62,17 0,I81,I83 -
The prospective outcome analysis studies described in section 1.2 arc consistent with other published experience. Overall morbidity appears to be reduced, and the median hospital stay of 3 days is an improvement on historical experience. However, an incidence of re-operation of approximately I07o reported in the initial analysis (sections
I.2.2 and 1.2.3), in particular the incidence occurring during the first 3 months following surgery, is ofconcern. 216
Specific adverse outcomes which are unique to laparoscopic antireflux surgery were identified from this experienceT ,118,201,220,221,225. So*e of these may be due to changes in surgical techniques made to facilitate laparoscopic dissection. Diathermy dissection is implicated in the development of hiatal sls¡s5¡s221, dissection within the chest and the passage of instruments behind the oesophagus may result in left pneumothorax if the pleural membrane is damagedz25, v¡r¿ paradoxically acute para- oesophageal hiatus herniation may be more common because of the reduced postoperative pain experienced following the laparoscopic approachzzO. Routine hiatal repair may reduce the risk of this latter complication, although subsequent experience confirms that this risk has not been elimi¡¡v¡s670,2I7. While not seon during this analysis, other surgeons have described an incidence of oesophageal perforation due to either instrumental injury to the posterior wall of the oesophagus or a bougie perforating the anterior walI118,201. A11 of these adverse outcomes represent a downside for the laparoscopic approach which must be carefully weighed up against the risks of the open approach and the efficacy of non-operative measures. This justifies prospective randomised trials to determine whether the laparoscopic approach is really better than the standard open procedure.
4.1.2 Prospective randomised trials of laparoscopic versus open Nissen fundoplication
The results of the study in section 3.1 confirm that the overall morbidity of surgery for reflux is reduced by the laparoscopic approach. The overall incidence of complications was reduced and the hospital stay was shortened, although not by the magnitude which was predicted from earlier audit studies. Outcomes in the group of patients undergoing attempted laparoscopic surgery resembled the results from the early studies described in section 1.2, with a conversion rate of approximately 107o, a postoperative stay of 3 days, and acceptably low morbidity. Also the median operating time of 79 minutes was shorter than the initial learning experience at the Royal Adelaide Hospital, suggesting that many 2t7 of the early learning difficulties associated with this plocedure had been overcome in this particular study.
Surprisingly, however, the postoperative hospital stay was shortened only by a mean 1.1 days, due to the shorter hospital stay seen following open fundoplication within the trial, compared to traditional experience. This suggests that there is much scope to achieve significant reductions in hospital stay in patients undelgoing open antireflux surgery. The most significant advantage for the laparoscopic approach, howevet, was seen following discharge from hospital. The time taken for patients to return to notmal physical activity was decreased from 8 to 2 weeks, confirming significant community advantages for the laparoscopic approach.
The results and conclusions from this trial have recently been confirmed by two further prospective randomised trials comparing laparoscopic versus open Nissen fundoplication.
A study from Franzen et a161, published in abstract form, described similar advantages for laparoscopic fundoplication. They reported an initial analysis of 36 patients undergoing antireflux surgery. A one day reduction in hospital stay due to earlier discharge following open surgery was also seen in this study. Laine s¡ a1108 recently reported the outcome of
110 patients randomised to undergo laparoscopic or open Nissen fundoplication, the majority without division of the short gastric vessels. Hospital stay was halved from 6.4 to 3-2 days, and patients returned to work quicker (37 vs 15 days). Operating time was prolonged 31 minutes from 57 to 88 minutes, an amount which is similar to that reported in section 3.1. Control of reflux symptoms was similarly effective in both of these studies, although outcomes beyond 12 months postoperative are yet to be reported.
The results of all 3 of randomised studies confirm that laparoscopic antireflux surgery has important short term advantages over the equivalent open approach. Short term control of reflux is similar to that following open surgery. Nevertheless the longer term outcomes of all these trials are necessary before equivalent efficacy can be claimed with certainty. 2r8
4.2 SHORT GASTRIC VESSELS
Many surgeons have advocated routine division of the short gastric vessels during both open and laparoscopic Nissen fundoplication to achieve full fundal mobilisation, thereby guaranteeing a loose fundoplica¡isî38,43,49,50,80,170,202. '¡¡¡ is thought by some to be an essential step, which is necessary to reduce the likelihood of post-operative swallowing difficulties. This step became popular following the publication of a number of uncontrolled studies which compared experience with this manoeuvre with historical experience with a Nissen fundoplication performed without dividing the short gastric u..."1.43,49,50. A¡guably the most influential paper advocating routine division of the short gastric vessels was published by DeMeester et al in 198643 - In this study, surgical technique was modified on a 3 occasions in a series of 100 patients undergoing surgery over a 12 year period. Because the second modification involved both division of short gastric vessels and shortening of the fundoplication, the conclusions drawn about the impact of short gastric vessel division should be regarded as speculative. Other uncontrolled studies of Nissen fundoplication either with or without division of the short gastric vessels confuse the issue further, as both good and bad results have been reported whether these vessels were divided e¡ ns150,93,186.
In the new era of laparoscopic surgery, this issue has become even more contentious3,65,80,170,206. More intensive prospective follow-up of patients following the laparoscopic Nissen procedure has suggested that the incidence of post-operative dysphagia may have increased following the introduction of laparoscopic approaches24,80,90,Il0. Howerrer, the manner in which follow-up data has been collected following laparoscopic fundoplication (prospective and by an independent observer) is usually different to the previous patterns of data collection following open surgery (usually case note review). This means that conclusions drawn about dysphagia rates must be regarded as tentative at best, and therefore valuable primarily to enable hypotheses to be constructed which might be tested in randomised trials. 2r9
Further compounding the difficulty one faces when deciding whether to divide the short gastric vessels or not, is the potential problem of a learning curve bias which can lead to confusion when drawing conclusions from some recent laparoscopic series. The studies described in section 1.2.5 suggest that the risk of adverse outcomes is greater during a surgeon's early experience. Many surgeons who did not divide the short gastric vessels during their initial experience with laparoscopic Nissen fundoplication, reported a significant incidence of adverse outcomes following their initial cases24,34,38,80,90. 45 a consequence of this some modified their technique to incorporate routine division of the 'When short gastric vessels34,38,80. outcomes improved, it was assumed that this was due to division of the vessels. However, other surgeons reported good outcomes within larger series of patients whose vessels were left i¡¡¿s¡3,65. The experience at the Royal
Adelaide Hospital (section I.2.5.2) improved with experience rather than due to division of the short gastric vessels. This highlights the need to be careful when drawing conclusions from uncontrolled series due to the possible influence of both historical and learning curve biases.
The results of the double blind prospective randomised trial in section 3.2 demonstrated no advantage for patients who underwent division of the short gastric vessels. This study in fact has demonstrated that patients may be disadvantaged by having their vessels divided, due to the additional operating time required for this manoeuvre, and the risk of major haemorrhage during vessel division or due to bleeding from one of these vessels in the immediate post-operative period. These results are supported by a recently published trial of 25 patients who were randomised to undergo vessel division or to have their vessels left intact during open Nissen fundoplic a¡isr125,l26. While this study can be criticised for not enrolling a suffîcient number of patients to achieve statistically significant results, and the raîEe of symptom scores was less rigorous than the study
described in section 3.2, its results are in agreement with the outcomes described in this thesis. Specifically the incidence and severity of post-operative dysphagia 'was not improved by dividing the short gastric vessels, and the oesophageal emptying time for liquids was actually longer in the group of patients whose vessels were divided. In the 220 absence of any other randomised trials, it remains for the protagonists of routine division of the short gastric vessels to repeat these studies. Unless further evidence to the contrary is reported from other controlled trials, then it is appropriate to omit the step of division of the short gastric vessels when performing a laparoscopic Nissen fundoplication.
4.3 NISSEN VERSUS PARTIAL FUNDOPLICATION
4.3.1 Uncontrolled studies
The uncontrolled studies describing outcomes following the various forms of partial fundoplication usually describe good or excellent ¡s5¡1159,105,I4I'I52,I59,162,166,21I,212.
These studies suggest that the incidence of post-operative dysphagia is low, with at least four series claiming that dysphagia didn't occur beyond the early post-operative period following a laparoscopic partial fundoplication procedure105,142 ,I52,I59 . Further advantages are reported to be a lower incidence of gas bloat syndrome, improved ability to belch and a reduction in the incidence of various other adverse outcomes105,152,159. Ptopottents claim, indeed with some support from manometric studies, that the partial fundoplication variants create a more physiological situation following sutgery21l,2I2. However, as with the analysis of the short gastric division issue, there is a risk that significant assessment bias might be introduced when determining the potential advantages of the partial fundoplication techniques if experience is compared to that of other surgeons or to historical controls. Only the outcome of prospective randomised trials should be considered when determining the relative advantages of the various partial fundoplication techniques.
4.3.2 Experimental models
The studies described in section 2 confirm the short term efficacy of both the anterior and posterior partial fundoplication techniques. Perhaps surprisingly, the results of the porcine viscera study (section 2.1) wele similar to the results of the in vivo porcine
studies. This adds weight to the argument that all 3 techniques tested are appropriate for the correction of pathological reflux. It also suggests that all forms of fundoplication act
as a mechanical barrier to reflux, and that this mechanism is independent of any 22I contributions by muscle tone in the gastric wall. A simple flap valve mechanism created by the close apposition of the gastric fundus to the distal oesophagus appeared to restore the antireflux mechanism in these models.
4.3.3 A role for patients with oesophageal dysmotility
It is a widely held belief that patients with preoperative oesophageal dysmotility who undergo fundoplication, are at a greater risk of post-operative dysphagia following a
Nissen fundoplication, than following a partial fundoplica¡isr12,l66. Surgeons holding this opinion selectively perform a partial fundoplication for these patients. Unfortunately there is a paucity of good research to support this opinion. Series of patients with poor pre-operative motility who have undergone partial fundoplication are described with good outcomes101,166. However, these outcomes are difficult to compare with outcomes following Nissen fundoplication. The study in section I.2.6.3 described an overall excellent outcome following laparoscopic Nissen fundoplication without division of the short gastric vessels in patients with unequivocal poor preoperative motility. Similar results were reported by Bremner et aLI2 in a group of patients with lesser degrees of preoperative oesophageal dysmotility. The issue of selective application of partial fundoplication techniques for patients with pre-operative dysmotility remains unsupported by the current published studies. The results of randomised trials are needed to determine an appropriate role for the partial fundoplication procedures.
4.3.4 Prospective randomised trials of Nissen yersus partial fundoplication
Analysis of the previously reported results of 3 prospective randomised of Nissen fundoplication versus a posterior partial fundoplication during the open surgical era, confirms that there is no significant advantage for either technique. While the study by
Thor and Silander203 suggests that there may be some advantages for the posterior technique, the number of patients enrolled was small and the differences were not supported by the results from the larger studies reported by Lundell et all19,I20 an6
Walker s¡ a1209. 222
Laws u1109 recently reported the early outcomos fi'om a small randomised trial of "¡ laparoscopic Nissen fundoplication with short gastric vessel division versus posterior partial (Toupet) fundoplication. Thirty nine patients were randomised, 23 to undergo
Nissen fundoplication and 16 a posterior fundoplication. No advantages were demonstrated for either procedure, although the study could be criticised for not enrolling an adequate number of patients to achieve statistical significance.
The study described in section 3.3 is the first prospective randomised trial to compare a
Nissen fundoplication with an anterior pattial fundoplication technique. Both procedures were performed laparoscopically. As with the trials of posterior partial fundoplication, the early study outcomes did not demonstrate any advantages or disadvantages for the partial fundoplication technique. However, the analysis of the 6 month outcomes has demonstrated significant advantages for the anterior partial fundoplication. Dysphagia was less likely, and the incidence of other adverse outcomes was less at this follow-up interval. Overall patient satisfaction with the procedure was greater following anterior fundoplication, and control of pathological leflux was satisfactory. This study demonstrates for the first time in a prospective randomised trial advantages for a partial fundoplication technique. The way the anterior fundoplication is constructed is different to the posterior fundoplication variants, with the fundus sitting in front rather than behind the oesophagus. This may account for the advantages seen for this procedure, and the lack of advantages demonstrated in previous trials of Nissen versus posterior partial fundoplication. These advantages, however, remain short term until follow-up has matured further. Long term reflux control has not been demonstrated by this study.
However, if equally effective in the long term, the trial outcomes provide a strong case for the routine application of an anterior partial fundoplication in patients with gastro- oesophageal reflux disease requiring surgical management.
4.4 IMPLICATIONS
While the studies described provide useful information about antireflux surgery, the outcome analysis has been performed following short tetm follow-up only. Follow-up of 223 patients in the randomised trials ranged from 6 to 12 months. This is probably adequate to allow conclusions to be drawn about dysphagia rates, and the adverse outcomes.
However, firm conclusions about long term efficacy cannot be made until long term follow-up data is available. The short term control of reflux achieved by both laparoscopic Nissen and anterior fundoplication was equivalent to the outcomes achieved following open surgery. This was confirmed by the experimental studies, which demonstrated efficacy for all types of fundoplication tested. As the operations performed are identical to open procedures with known long term outcomes, except for the method of surgical access, it is likely that long term reflux control will be similar, particularly for the Nissen fundoplication.
The studies described have demonstrated that the laparoscopic approach reduces the overall morbidity associated with surgery for gastro-oesophageal reflux and that division of the short gastric vessels during laparoscopic Nissen fundoplication is usually unnecessary. Significantly an anterior partial fundoplication had significant short term advantages over the Nissen technique. An anterior fundoplication performed laparoscopically, may be the preferred surgical approach for the correction of gastro- oesophageal reflux disease, if it achieves adequate long term control of gastro- oesophageal reflux.
4.5 FUTURE DIRECTIONS
As the results of these studies have demonstrated short term advantages for modifications to the original fundoplication technique of Nissen, by performing an anterior hemi- fundoplication, it is important to confirm these findings within long term follow-up. This data will be collected and reported as it becomes available.
The development of significant post-operative dysphagia following fundoplication also resulted from a tight oesophageal hiatus in a number of patients described in these studies
(section I.2.4.3), including one patient who underwent an anterior fundoplication (section
3.3). This raises the possibility that a component of the dysphagia which follows 224 fundoplication, both partial and total, may be related to the method of hiatal closure.
Posterior hiatal closure often lifts the oesophagus antedorly, away from its usual anatomical position, and thereby angulates the distal oesophagus. This is most obvious when a large hiatal defect is present, and a long posterior hiatal closure is performed. To investigate this possibility, a double blind prospective landomised trial of posterior versus anterior hiatal closure during laparoscopic Nissen fundoplication has been commenced.
Anterior hiatal closure should not result in anterior oesophageal displacement. If displacement at the hiatus does contribute to post-operative dysphagia, then less dysphagia would be expected in patients undergoing anterior hiatal closure. On the other hand it remains to be seen whether anterior closure achieves as effective closure as the posterior technique, and thereby prevents migration of the fundoplication into the chest, or the occurrence of post-operative para-oesophageal hiatus herniation.
A further possible contributing factor promoting post-operative dysphagia is that fundoplication might result in disturbances to oesophageal motility in the immediate and early post-operative period. Whilst it is known that oesophageal obstruction can result in secondary peristaltic impairment, or even pseudo-achalasia with absent oesophageal body peristalsis and absent'lower sphincter' relaxation demonstrated manometrically, the overall changes in manometric outcome investigated in I.2.6.2 suggest that Nissen fundoplication is not associated with an overall deterioration in oesophageal body peristalsis. It is possible though, that oesophageal motility disturbances in the initial post- operative period could contribute to the early dysphagia experienced by most patients following fundoplication. This possibility is being investigated by a manometric study of post-operative motility in the immediate period following fundoplication.
Although the results of studies described in this thesis confirm significant advantages for the laparoscopic approach for antireflux surgery, the issue of when surgical intervention should be recommended has not been addressed. At present surgical management is reserved principally for patients who continue to experience symptoms despite management with a proton pump inhibitor, or for patients who do not wish to take 225 medication lifelong. The promotion of gastrinomas in experimental studies in rats, and the development of atrophic changes in the gastric mucosa in humans ingesting proton pump inhibitorsl07, raises the question as to whether medical or surgical therapy should be offered as the preferred treatment option for patients with severe reflux disease. The only randomised trial to address this issue compared the open approach to Nissen fundoplication with a H2 receptor antagonir1l95. Surgical therapy was shown to be superior to medication. However, neither treatment option would now be accepted as optimal treatment for patients with severe reflux. Only a prospective randomised trial of laparoscopic Nissen fundoplication versus proton pump inhibitor medication, can effectively determine which therapy offers the best long term control of reflux symptoms not controlled by H2 receptor antagonists. 226 227
sEcTtoN 5
BIBLIOGRAPHY 228
1 Policy and position statement: An American Gastroenterological Association medical position statement on the clinical use of esophageal manometry - Gastroenterol
1994;L07:1865-1884.
2 ArnaralJF. Laparoscopic fundoplication with ultrasonic energy. Min Inv Ther 1994;3
(Supplement 1) :25. (Abstract)
3 Anvari M, Allen C, Borm A. Laparoscopic Nissen fundoplication is a satisfactory
alternative to long-term omeprazole therapy. Br J Surg 1995;82:938-942.
4 Anvari M, Allen CJ. Prospective evaluation of dysphagia before and after laparoscopic
Nissen fundoplication without routine division of short gastrics. Surg Inpar Endosc
1996;6:424-429.
5 Aye RW, Hill LD, Kraemer SJM, Snopkowski P. Early results with the laparoscopic
Hill repair. Am J Surg 1994;L67:542-546.
6 Bagnato J. Laparoscopic Nissen fundoplication. ,Sørg Lapar Endosc 1992;2:188-190.
7 Baigrie RJ, Watson DI, Game PA, Jamieson GG. Vascular perils during laparoscopic
dissection of the oesophageal hiatus. Br J Surg 1997:84:556-557 .
8 Baigrie RJ, Watson DI, Myers JC, Jamieson GG. The outcome of laparoscopic Nissen
fundoplication in patients with disordered pre-operative peristalsis. Gut 1997;40:381-385.
9 Betl RCW, Hanna P, Powers B, Sabel J, Hruza D. Clinical and manometric results of
partial (Toupet) complete (Rosetti-Nissen) fundoplication. Endosc laparoscopic and ^Sørg 1996-10:724-728. 229
10 Biswas TK, Smith JA. Laparoscopic total fundoplication: Anaesthesia and complications. Anaesthesia and Intensive Care 1993;21:127 -128-
11 Bittner HB, Meyers WC, Brazer SR, Pappas TN. Laparoscopic Nissen fundoplication:
Operative results and shorl-term follow-up. Am J Surg 1994;167:193-200.
12 Bremner RM, DeMeester TR, Crookes PF. The effect of symptoms and nonspecific
motility abnormalities on outcomes of surgical therapy for gastroesophageal reflux
disease. J Thorac Cardiovasc Surg 1994.107:1244-1250.
13 Brooks DC. A prospective comparison of laparoscopic and tension-free open
herniorrhaphy . Ar ch S ur g I99 4;129 :3 6l -3 66.
14 Butterfield WC. Current hiatal hernia repairs: Similarities, mechanisms, and extended
indications - an autopsy study. Surgery I97I;69:910-916.
15 Byrne DG. The stability of illness behaviour after myocardial infarction- Int J
P sy chiat M e d 19 80 ;10 :23 -3 I.
16 Cadiere GB. La chirurgie anti-reflux: indication, principe et apport de la coeloiscopie.
Rev Med Brux 1994;15:25-30.
17 Cadierc GB, Bruyns J, Himpens J, Verroken R, Muls V, Panzer JM. Laparoscopic
Nissen fundoplication.,Sar6 Endos c 199 4;5:47 5. (Abstract)
18 Cadiere GB, Himpens J, Bruyns J. How to avoid esophageal perforation while
performing laparoscopic dissection of the hiatus. Surg Endosc 1995;9:450-452.
19 Cadiere GB, Houben JJ, Bruyns J, Himpens J, Panzer JM, Gelin M. Laparoscopic
Nissen fundoplication: technique and preliminary results. Br J Surg 1994;81:400-403. 230
20 Canoll BJ, Phillips EH, Daykhovsky L. Laparoscopic choledochoscopy: An effective approach to the common duct. ,I Inparoendosc Surg 7992:2:15-2I.
21 Casabella F, Sinanan M, Horgan S, Pelligrini CA. Systematic use of gastric fundoplication in laparoscopic repair of paraesophageal hernias. Am J Surg 1996;l7L:485-
489.
22Cattey RP, Henry LG, Bielefield MR. Laparoscopic Nissen fundoplication for gastroesophageal reflux disease: Clinical expedence and outcome in first 100 patients.
Surg Lapar Endosc 1996:6:430-433.
23 Champault G, Volter F, Rizk N, Boutelier P. Gastroesophageal reflux: Conventional surgical treatment versus laparoscopy. A prospective study of 61 cases - Surg Lapar
Endo s c 199 6;6:43 4- 440.
24Collard JM, de Gheldere CA, De Cock M, Otte JB, Kestens PJ. Laparoscoplc antireflux surgery. What is real progress? Ann Surg 1994;220:146-154.
25 Collard JM, Romagnoli R, Kestens PJ. Reoperation for unsatisfactory outcome after laparoscopic antireflux surgery. Dis Esoph 1996;9:56-62.
26 ColLet D, Cadiere GB. Conversions and complications of laparoscopic treatment of gastroesophageal reflux disease. Am J Surg 1995;169:622-626.
27 Congreve DP. Laparoscopic paraesophageal hemia repair. J Laparoendosc Surg
1992;2:45-48. 231
28 Cook IJ, Gabb M, Panagopoulos V, Jamieson GG, Dodds WJ, Dent J, Shearman DJ.
Pharyngeal (Zenker's) diverticulum is a disorder of upper esophageal sphincter opening.
Gas tr o enter ol 1992;103 :1229 - I23 5.
29 Coster DD, Bower WH, Wilson VT, Brebrick RT, Richardson GL. Laparoscopic
partial fundoplication vs laparoscopic Nissen-Rosetti fundoplication. ^Sørg Endosc 1997;ll:625-631.
30 Coster DD, Bower WH,'Wilson VT, Butler DA, Locker SC, Brebrick RT.
Laparoscopic Nissen fundoplication - A curative, safe, and cost-effective procedure for
complicated gastroesophageal reflux disease. Surg lnpar Endosc 1995;5: III-IL7 .
31 Crookes PF, DeMeester TR. Does Toupet fundoplication out-perform the Nissen
procedure as the operation of choice for gastro-esophageal reflux disease? Dis Esoph
1994;7:265-267.
32 Crookes PF, DeMeester TR. Complete and partial laparoscopic fundoplication for
gastroesophageal reflux disease. S ur g Endo s c 1997 ;ll:613 -614.
33 Cuschieri A. Laparoscopic antireflux surgery and repair of hiatal hernia. World J Surg
L996:17:40-45.
34 Cuschieri A, Hunter J, Wolfe B, Swanstrom LL, Hutson V/. Multicentre evaluation of
laparoscopic antireflux surgery. Preliminary report. Surg Endosc 1993:7:505-510.
35 Cuschieri A, Shimi S, Nathanson LK. Laparoscopic reduction, crural repair, and
fundoplication of large hiatal hernia. Am J Surg 1992,163:425-430.
36 Dakkak M, Bennett JR. A new dysphagia score with objective validation- J Clin
G as tr o entro I 1992;14 :99- 1 00. 232
37 Dallemagne B, Taziaux P, Weerts J, Jehaes C, Markiewicz S. Chirurgie laparoscopique du reflux gastro-oesophagien. Ann Chir 1995 ;49:30-36 -
38 Dallemagne B, Weerts JM, Jehaes C, Markiewicz S. Causes of failures of
laparoscopic antireflux operations.'Surg Endosc 1996; 10:305-3 10.
39 Dallemagne B, Weerts JM, Jehaes C, Markiewicz S, Lombard R. Laparoscopic Nissen
fundoplication: Preliminary report. Surg Lapar Endosc 199 1 ;1 : 138- 143.
40 Dashow L, Friedman I, Kempner R, Ruddick J, McSherry C. Initial experience with
laparoscopic cholecystectomy at the Beth Israel Medical Centre. Surg Gynecol Obstet
1992.175:25-30.
4lDay JP, McDonald JE, Bradley LA, Richter JE. Patients demographics, symptoms and
manometric data are poor predictors of gastroesophageal reflux disease. Gastroenterol
199 1 ;100:45.(Abstract)
42DeMeester TR. Surgical management of gastroesophageal reflux. In: Castell DO, Wu
WC, Ott DJ, eds. Gastroesophageal reflux disease. Pathogenesis, diagnosis, therapy.
New York: Futura Publishing Co, 1985;243-280.
43 DeMeester TR, Bonavina L, Albertucci M. Nissen fundoplication for
gastroesophageal reflux disease. Evaluation of primary repair in 100 consecutive patients.
Ann Sur g 1986.204:9 -20.
44 DeMeester TR, Ireland AP. Gastric pathology as an initiator and potentiator of
gastroesophageal reflux disease. Dis Esoph 1997;10:1-8. 233
45 DeMeester TR, Johnson LF, Kent AH. Evaluation of current operations for the prevention of gastroesophageal reflux. Ann Strg I974;180:5Il-525-
46 DeMeesûer TR, Sæin HJ. Minimizing the side effects of antireflux surgery. World J
Surg 1992;16:335-336.
47 DeMeester TR, Wernly JA, Bryant GH, Little AG, Skinner DB. Clinical and in vitro analysis of determinants of gastroesophageal competence. A study of the principles of antireflux surgery. Am J Surg 1979;137:39-46.
48 Depaula AL, Hashiba K, Bafutto M, Machado CA. Laparoscopic reoperations after
Endosc 1995;9:681-686. failed and complicated antireflux operations. ^Surg
49 Donahue PE, Bombeck CT. The modified Nissen fundoplication - reflux prevention without gas bloat. Chir gastroent I9ll;Ll:15-27 .
50 Donahue PE, Samelson S, Nyhus LM, Bombeck T. The floppy Nissen fundoplication.
Effective long-term control of pathological reflux. Arch Surg 1985;120:663-668.
51 Dunn J. Laparoscopic hiatus hernia repair. Proceedings of the ÃACS Annual Scientific
Con gr e s s I99 4;459.(Abstract)
52 Dunnington GL, DeMeester TR. Outcome effect of adherence to operative principles of Nissen fundoplication by multiple surgeons. Am J Surg 1993;166:654-659.
53 Edelman DS. Laparoscopic paraesophageal hernia repair with mesh. Surg Lapar
Endosc 1995;5:32-37. 234
54 Eleftheriadis E, Kotzampassi K, Botsios D, Tzartinoglou E, Farmakis H, Dadoukis J
Splanchnic ischemia during laparoscopic cholecystectomy. Surg Endosc 1996;10:324-
326.
55 Falk GL, Brancatisano RP, Hollinshead J, Moulton J. Laparoscopic fundoplication: A preliminary report of the technique and postoperative care. Aust NZ J Surg 1993;17:40-
45.
56 Farn J, Hammerman AM, Brunt LM. Intraoperative pneumothorax during laparoscopic cholecystectomy: A complication of prior transdiaphragmatic surgery. Szrg
Lapar Endo s c 1993 ;3 :2I9 -222.
57 Firoozmand E, Ritter M, Cohen R, Peters J. Ventricular laceration and cardiac
Nissen fundoplication. Lapar Endosc 1996;6:394- tamponade during laparoscopic ^Sørg 397.
58 Fitzgibbons RJ, Camps J, Comet DA, Nguyen NX, Litke BS, Annibali R, Salerno
GM. Laparoscopic inguinal herniorrhaphy. Results of a multicenter trial. Ann Surg
1995;221:3-13.
59 Fontaumard E, Espalieu , Boulez J. Laparoscopic Nissen-Rossetti fundoplication.
Surg Endosc 1995;9:8 69-873.
60 Frantzides CT, Carlson MA. Laparoscopic versus conventional fundoplication. ./
Laparoendos c Surg 1995 ;5:137 -743.
61 Franzen T, Anderberg B, Tibbling L, Johansson KE. A report from a randomized study of open and laparoscopic 3600 fundoplication. Surg Endosc 1996;10:582.(Abstract) 235
62 Fuchs KH, Heimbucher J, Freys SM, Theide A. Management of gastro-esophageal reflux disease 1995; Tailored concept of anti-reflux operations. Dis Esoph 1994;7:250'
254.
63 Gabbott DA, Dunkley AB, Roberts FL. Carbon dioxide pneumothorax occurring during laparoscopic cholecystectomy. Anae s the s ia 1992;37 :5 87 -5 8 8.
64 Geagea T. Laparoscopic Nissen's fundoplication: preliminary report on ten cases. ,Sørg
Endo s c 199 I ;5:17 0 - 17 3.
65 Geagea T. Laparoscopic Nissen-Rossetti fundoplication. Surg Endosc 1994;8:1080-
1084.
66 Gharagozloo F, Evans SRT, Attai DJ, Axelrad AM, Benjamin SB. Video-laparoscopic reduction of an intrathoracic stomach. Surg Lapar Endosc 1996;6:234-238.
67 Glise H, Hallerback B, Johansson B. Quality-of-life assessments in evaluation of laparoscopic Rosetti fundoplication. Sur g Endo s c I 995 ;9: 1 83- 1 89.
68 Goodall RJR, Temple JG. Effect of Nissen fundoplication on competence of the gastro- oesopha geal junction. G ut 1980 ;21 : 607 -6 1 3.
69 Gooszen HG,'Weidema'WF, Ringers J, Horbach JMLM, Masclee AAM, Lamers
CBI{W. Initial experience with laparoscopic fundoplication in the Netherlands and comparison with an established ûechnique (Belsey Mark'IItl). Scandinavian Journal of
G as tr o enter olo gy 1993 :200 (Suppl) : 24- 27.
70 Gotley DC, Smithers BM, Rhodes M, Menzies B, Branicki FJ, Nathanson L.
Laparoscopic Nissen fundoplication - 200 consecutive cases - Gut 1996;38:487-49I. 236
71 Hallerback B, Glise H, Johansson B, Radmark T. Laparoscopic Rossetti fundoplicati on. Sur g Endo s c 1 994 ;8 : I 477 - I 422.
T2IJar:.rs SC. Laparoscopic antireflux surgery. Am J Surg 1996;17l:482-484.
T3Heinzelmann M, Simmen HP, Cummins AS, Largiader F. Is laparoscopic appendectomy the new'gold standard'. Arch Surg 1995;130:782-785-
T4HerzogU, Kocher T, Ackermann C. Laparoscopic cholecystectomy - experiences and results with a new surgical technique. SchweizMedWochenschr 1992;122:659-662.
75 Hill ADK, Walsh TN, Bolger CM, Byrne PJ, Hennessy TPJ. Randomized controlled trial comparing Nissen fundoplication and the Angelchik prosthesis. Br J Surg
1994;81:72-74.
76 HillLD. An effective operation for hiatal hernia: an eight year appraisal. Ann Surg
1967;166:681-692.
77 lJinder RA, Filipi CJ, Wetscher G, Neary P, DeMeester TR, Perdikis G. Laparoscopic
Nissen fundoplication is an effective treatment for gastroesophageal reflux disease. Ann
Surg I99 4;220 :47 2- 483.
78 Huang SM, V/u CW, Hang HT, Liu M, King KL, Lui WY. Bile duct injury and bile leakage in laparoscopic cholecystectomy. Br J Surg 1993;80:1590-1592.
79 Hunter JG, Swanstrom L,'Waring JP. Dysphagia after laparoscopic antireflux surgery.
The impact of operative technique. Ann Surg 1996:224:5I-57 .
80 Hunter JG, Trus TL, Barum GD, Waring JP. A physiologic approach to laparoscopic fundoplication for gastroesophageal reflux disease. Ann Surg 1996;223:673-687. 237
81 Iglesias JL, Meier DE, Thompson WR. Cost analysis of laparoscopic and open fundoplication in children. Pediatríc Endosurgery & Innovative Techniques 1997;1:15- 2r.
32Incarbone R, Peters JH, Heimbucher J, Dvorak D, Bremner CG, DeMeester TR. A contemporaneous comparison of hospital charges for laparoscopic and open Nissen
Endos I99 5 :1 5 1 - 1 55. fundoplication. ^Sarg c ;9
33Ireland AC, Holloway RH, Toouli J, Dent J. Mechanisms underlying the antireflux action of fundoplication. Gut 1993 ;34:303-308.
34Ismail T, Bancewicz J, Barlow J. Yield pressure, anatomy of the cardia and gastro- oesophageal reflux. Br J Surg 1995;82:943-941.
85Iwai N, Kaneda H, Tsuto T, Hashimoto K, Yanagihara J, Majima S. Antireflux mechanism in Nissen fundoplication - an experimental study. Ipn J Surg 1984;14:258-
26t.
86 Jamieson GG. Mechanisms of gastro-oesophageal reflux. Aust NZ J Surg
1988;58:193-195.
87 Jamieson GG, Myers JC. The relationship between intra-operative manometry and clinical outcome in patients operated on for gastro-esophageal reflux disease. World J
Surg 1992;16:337-340.
88 Jamieson GG, Tew S, Gabb M, Holloway R, Ferguson S, Tew P. What causes dysphagia after fundoplication? Proceedings of the Sixth World Congress of the
International Society for Diseases of the Esophagus Milan, Italy 1995;154.(Abstract) 238
89 Jamieson GG, Tew S, Holloway R, Ferguson S. Can patients belch after anti-reflux surgery? Proceedings of the Sixth World Congress of the International Society for
Diseases of the Esophagus, Milan, Italy 1995;93.(Abstract)
90 Jamieson GG, Watson DI, Britten-Jones R, Mitchell PC, Anvari M. Laparoscopic
Nissen fundoplic a tion. Ann S ur I I99 4 ;220 :I37 - I 45 .
91 Janssen IM, Gouma DJ, Klementschitsch P, van der Heyde MN, Obertop H.
Prospective randomised comparison of teres cardiopexy and Nissen fundoplication in the surgical therapy of gastro-oesophageal reflux disease. Br J Surg 1993;80:875-878.
92 Johansson B, Glise H, Hallerback B. Thoracic herniation and intrathoracic gastric perforation after laparoscopic fundoplication. Sørg Endosc 1995;9:917-9I8.
93 Johansson J, Johnsson F, Joelsson B, Floren C-H, Walther B. Outcome 5 years after
3600 fundoplication for gastro-oesophageal reflux disease. Br J Surg 1993;80:46-49.
94 Johansson KE, Ask P, Boeryd B, Fransson SG, Tibbling L. Oesophagitis, signs of reflux and gastric acid secretion in patients with symptoms of gastro-oesophageal reflux disease. Scandinavian Journal of Gastroenterology 1986;21,:837 -847 .
95 Johnston BT, Lewis SA, Love AHG. Do heartburn sufferers have a specific personality profile? G a s t o e nt e r o I 7992 ;102 : A9 1 . (Abstract)
96 Johnston BT, Lewis SA, Love AHG. Psychological factors in gastro-oesophageal reflux disease. Gut 1995;36:481-482.
97 Jorgensen JO, Gilles RB, Lalak NJ, Hunt DR. Lower limb venous hemodynamics during laparoscopy: An animal study. Surg lttpar Endosc 1994;4:32-35. 239
98 Joris JL, Chiche J-D, Lamy ML. Pneumothorax during laparoscopic fundoplication:
Diagnosis and treatment with positive end-expiratory pressurc. Anesth Analg
1995;81:993-1000.
99 Kahrilas PJ, Dodds WJ, Hogan WJ. Effect of peristaltic dysfunction on esophageal volume clearance. Gastro enterol 1988;9 4:7 3-80.
100 Kahrilas PJ, Dodds WJ, Hogan'WJ, Kern M, Amdorfer RC, Reece A. Esophageal peristaltic dysfunction in peptic esophagitis. Gastroenterol I986;9L:897 -904.
101 Kauer WKH, Peters JH, DeMeester TR, Heimbucher J, keland AP, Bremner CG. A tailored approach to antireflux surgery. J Thorac Cardiovasc Surg I995;ll0:I4l-141.
102 Kellner R. Hypochondriasis and somatisation . JAMA 1987:'258l.2118-2122.
103 Kiroff GK, Maddern GJ, Jamieson GG. A study of factors responsible for the efficacy of fundoplication in the treatment of gastro-oesophageal reflux. Aust NZ J Surg
1984;54:109-112.
104 Kluiber RM, Hartsman B. Laparoscopic appendectomy. A comparison with open appendectomy. Dis Colon Rectum 1996;39: 1008-101 1.
105 Kraemer SJM, Aye R, Kozarek RA, Hill LD. Laparoscopic Hill repair. Gastrointest
Endosc 1994 40:155- 1 59.
106 Kraus MA. Establishing a laparoscopic antireflux program: A private practice experience . Surg Inpar Endosc 1996;6:108- 1 13. 240
107 Kuipers EJ, Lundell L, Klinkenberg-Knol EC, Havu N, Festen HPM, Liedman B.
Atrophic gastritis and Helicobacter pylori infection in patients with reflux esophagitis treated with omep r azole or fundoplic ation. N E J M 199 6 ;33 4 : l0 I8 - 1022.
108 Laine S, Rantala A, Gullichsen R, Ovaska J. Laparoscopic vs conventional Nissen fundoplication. A prospective randomized study. Surg Endosc I99l.ll:44I-444.
109 Laws HL, Clements RH, Swillies CM. A randomized, prospective comparison of the
Nissen versus the Toupet fundoplication for gastroesophageal reflux disease. Ann Surg
1997;225:641-654.
110 Laycock WS, Oddsdottir M, Franco A, Mansour K, Hunter JG. Laparoscopic Nissen fundoplication is less expensive than open Belsey Mark IY. Surg Endosc 1995:9:426-429.
111 Laycock'WS, Trus TL, Hunter JG. New technology for the division of short gastric vessels during laparoscopic Nissen fundoplication. Szrg Endosc 1996;10:71-73.
II2I-eonard HK, Crozio RF, Ellis FH. Reoperation for complications of the Nissen fundoplication.,I Thorac Cardiovas c Sur g I98I ;81 : 50-56.
113 læonardi HK, Crozier RE, Ellis FH. Reoperation for complications of the Nissen fundoplication. J Thorac Cardiovasc Surg 1981;14:50-56.
114 Lim JK, Moisidis E, Munro WS, Falk GL. Re-operation for failed anti-reflux surgery.
Aust NZ J Surg 1996;66:73I-133.
115 Little AG. Gastro-oesophageal reflux and oesophageal motility diseases; Who should perform antireflux surgery? Ann Chir Gynaecol 1995 ;84: 103- 105. 241
116 Little AG. Nissen fundoplication for gastroesophageal reflux disease: How does
Nissen fundoplication prevent reflux? Dis Esoph L996;9:247-250.
117 Lowham AS, Filipi CJ, Hinder RA, Swanstrom LL, Stalær K, dePaula A, Hunter JG,
Bublewicz TG, Haake K. Mechanisms of avoidance of esophageal perforation by anesthesia personnel during laparoscopic foregut surgery. Surg Endosc 199610979-982.
118 Lowham AS, Filipi CJ, Hinder RA, Swansttom LL, Stalter K, dePaula A, Hunter JG,
Buglewicz TG, Haake K. Mechanisms and avoidance of esophageal perforation by anesthesia personnel during laparoscopic foregut surgery. Surg Endosc 1996;10979-982.
119 Lundell L, Abrahamsson H, Ruth M, Rydberg L, Lonroth H, Olbe L. Long-term results of a prospective randomized comparison of total fundic wrap (Nissen-Rossetti) or semifundoplication (Toupet) for gastro-oesophageal reflux. Br J Surg 1996;83:830-835.
120 Lundell L, Abrahamsson H, Ruth M, Sandberg N, Olbe LC. Lower esophageal sphincter characteristics and esophageal acid exposure following partial or 3600 fundoplication: Results of a prospective, randomized clinical study. World J Surg
1991;15:1L5-12I.
121 Lundell L, Myers JC, Jamieson GG. The effect of antireflux operations on lower oesophageal sphincter tone and postprandial symptoms. Scandinavian Journal of
Gas tro ent er o lo gy 1993 ;28:7 25 -7 37.
I22Llundell LR, Myers JC, Jamieson GG. The influence of pre-operative oesophageal motor function on long term outcome of anti-reflux surgery. GuIIet 1993;3:50-55.
123 Luostarinen M. Nissen fundoplication for reflux esophagitis. Long-term clinical and endoscopic results in 109 of I21 consecutive patients. Ann Surg 1993;217:329-331. 242
124 Luostarinen M,Isolauri J, Laitinen J, Koskinen M, Keyrilainen O, Markkula H,
Lehtinen E, Uusitalo A. Fate of Nissen fundoplication after 20 years. A clinical, endoscopical, and functional analysis. Gut 1993',34: 10 1 5- 1020.
125 Luostarinen M, Koskinen M, Reinikainen, Karvonen J, Isolauri J. Two antireflux operations: Floppy versus standard Nissen fundoplication. Ann Med 1995;27:199-205-
126 Luostarinen MES, Koskinen MO, Isolauri JO. Effect of fundal mobilisation in
Nissen-Rossetti fundoplication on oesophageal transit and dysphagia. Eur J Surg
1996;162:37-42.
I27 Macintyre IMC, Miles WFA. Critical appraisal and current position of laparoscopic hernia repair. J R Coll Edinb 1995;40:331-336.
128 Macintyre IMC, Wilson RG. Laparoscopic cholecystectomy. Br J Surg 1993;80:552-
559.
129 MacPhee WM. The laparoscopic placement of the Angelchik anti-reflux prosthesis.
Min Inv Ther 1993;2:5-9.
130 Maddern GJ, Jamieson GG. Fundoplication enhances gastric emptying. Ann Surg
1985;210:296-299.
131 Maddern GJ, Myers JC, Mclntosh N, Bridgewater FG, Jamieson GG. The effect of the Angelchik prosthesis on esophageal and gastric function. Arch Surg 199I;1261418-
1422.
132 Maddern GJ, Rudkin G, Bessell JR, Devitt P, Ponte L. A comparison of laparoscopic and open hernia repair as a day surgical procedure. Surg Endosc 1994;8:1404-1408. 243
133 Maitkainen M, Kaukinen L. The mechanism of Nissen fundoplication. Acta Chir
S cand 1984; 150:653-655.
134 Majeed AW, Troy G, Nicholl JP, Reed MWR, Stoddard CJ, Peacock J, Johnson AG
Randomised, prospective, single-blind comparison of laparoscopic versus small-incision cholecystectomy. Innc et L996;347 :989 -99 4.
135 Martinez de Haro L, Parrilla Paricio P, Ortiz Escandell MA, Morales Cuenca G,
Videla Troncoso D, Cifuentes Tebar J, Garay Pelegrin V. Antireflux mechanism of
Nissen fundoplication . A manometric study. Scandinavian Journal of Gastroenterology
1992:27:4ll-420.
136 Mathew G, Watson DI, Myers JC, Holloway RH, Jamieson GG. Oesophageal motility parameters before and after laparoscopic Nissen fundoplication. Br J Surg
1997;67:289-292.
137 Matkinen M-T, Yli-Hankala A, Kansanaho M. Early detection of CO2 pneumothorax with continuous spirometry during laparoscopic fundoplication. Acta Anaesthesiol Scand
1995;39:4lI-413.
138 McAnena OJ, Wilson PD, Evans DF, Kadirkamanathan SS, Mannur KR, Wingate
DL. Physiological and symptomatic outcome after laparoscopic gastric fundoplication. Br
J Surg 1995;82:7 95-7 9l .
139 McAnnena OJ, Willson PD, Evans DF, Kadirkamanathan SS, Mannur KR, Wingate
DL. Physiological and symptomatic outcome after laparoscopic gastric fundoplication. Br
J Surg 1995;1995:7 95-797 . 244
140 McGinn FP, Miles AJG, Uglow M, Ozmen M, Terzi C, Humby M. Randomizedtrial of laparoscopic cholecystectomy and mini-cholecystectomy. Br J Surg 1995;82:1374-
1377.
141 McKernan JB. Laparoscopic repair of gastroeophageal reflux disease: Toupet partial fundoplication versus Nissen fundoplic ation.,Surg Endo s c 1 994 ; I : 8 5 1 -85 6.
l42McKernan JB, Champion JK. Laparoscopic antireflux surgery. Am Surgeon
1995;61:530-536.
143 McKernan JB, Laws HL. Laparoscopic Nissen fundoplication for the treatment of gastroesophageal reflux disease. Am Sur g eon 799 4:60 :87 -93.
I44Mechanic D. The concept of illness behaviour. J Chron Dis 1962;189:194.
145 Medina LT, Vientimilla R, V/illiams MD, Fenoglio E. Laparoscopic fundoplication. ,I
Lap aro endo s c S ur g 199 6 ;6:219 -226 -
146 Mitchell PC, Jamieson GG. Coeliac Axis and Mesenteric arterial thrombosrs following laparoscopic Nissen fundoplication. Aust NZ J Surg 1994;64:728-730.
147 Mittal RK, Rochester DF, McCallum RW. Electrical and mechanical activity in the human lower esophageal sphincter during diaphragmatic contraction. ,I Clin Invest
1988;81:1 182-1 189.
148 Mittal RV, Ren J, McCallum RW', Shaffer HA, Sluss J. Modulation of feline esophageal contractions by bolus volume and outflow obstruction. Am J Physiol
1990;258:G208-G215. 245
149 Miyamoto Y, Higuchi A, Kamitani K, Shakunaga K. Pneumothorax during laparoscopy. Masui 1992:,41:1 3 1 1 - 1 3 13.
150 Moghissi K. Intrathoracic fundoplication for reflux stricture associated with short oesophagu s. Thor ax 1 983 ;38:36-40.
151 Monson JRT, Hill ADK, DarziA. Laparoscopic colonic surgery. Br J Surg
1995;82:150-157.
152 Mosnier H, Leport J, Aubert A, Kianmanesh R, Sbai Idrissi MS, Guivarc'h M. 1^270
gastroesophageal degree laparoscopic posterior fundoplasty in the treatment of reflux. "I Am Coll Surg 1995;18l:220-224.
153 Munro'W, Brancatisano R, Adams IP, Falk GL. Complications of laparoscopic fundoplication: The first 100 patients. Surg Lapar Endosc L996;6:42I-423.
154 Mutter D, Vix M, Bui A, Evrard S, Tassetti V, Brcton JC, Marescaux J. Laparoscopy not recommended for routine appendectomy in men: Results of a prospective randomized study. Surgery 1996;120:7 l-7 4.
155 Mutter D, Vix M, Bui A, Evrard S, Tassetti V, Breton JF, Marescaux J. Laparoscopy not recommended for routine appendectomy in men: Results of a prospective randomized study. Surgery L996;L20:7 l-7 4.
156 Nathanson LK, Shimi S, Cuschieri A. Laparoscopic ligamentum teres (round ligament) cardiopexy. Br J Surg 199178947-95I.
157 Negre JB. Post-fundoplication symptoms; Do they restrict the success of the Nissen fundoplication? Ann S ur g 1983 ;198 : 698-700. 246
158 Nissen R. Eine einfache operation zur beeinflussung der refluxoesophagitis - Schweiz
M ed W o chens chr 1 956 ;86: 59 0-592.
159 O'Reilly MJ, Mullins SG, Saye'WB, Pinto SE, Falkner PT. Laparoscopic partial fundoplication: Analysis of 100 consecutive cases. J Laparoendosc Surg 7996;6:141-150
160 Oddsdottir M, Franco AL, Laycock'WS, Waring JP, Hunter JG. Laparoscopic repair of paraesophageal hernia. New access, old technique. Surg Endosc 1995;9:1 64-168.
161 Oltyo J-8, Fontolliet Ch, Lang BF. La nouvelle classification de Savary des oesophagites de reflux . Acta Endo s copíc a 1992:22:301 -320.
162 Otttgnon Y, Pelissier EP, Mantion G, Clement C, Birgen C, Deschamps JP, Caryon P,
Gillet M. Gastroesophageal reflux. Comparison of clinical, pH-metric and manometric results of Nissen's and of Toupet's procedures - Gastroenterol Clin BioI I994;L8:920-926.
163 Overdijk LE, Rademaker BM, Ringers J, Odoom JA. Laparoscopic fundoplication: A new technique with new complications? J CIin Anesth 7994;6:32I-323-
164 Papp JP. Determination of lower esophageal sphincter pressure in patients having a
Nissen or Belsey fundoplication- Am J Gastroenterol 1979;71:154-757.
165 Parikh D, Tam PKH. Results of fundoplication in a UK paediatric centre. Br J Surg
I99I;78:346-348.
166 Patti MG, De Bellis M, De Pinto M, Bhoyrul S, Tong J, Arcerito M, Mulvihill SJ,
Way LW. Partial fundoplication for gastroesophageal reflux. Surg Endosc 1997;ll:445-
448. 247
167 Peters JH, DeMeester TR. Gastroesophageal reflux. Surg CIin of Nth Am
1993;73:llI9-1144.
168 Peters JH, DeMeester TR. Early experience with laparoscopic Nissen fundoplication.
In: Szabo Z,Lewis JE, Fantini GA, eds. Surgical Technology International IV- San
Francisco: Universal Medical Press Inc, 1995;109-113.
169 Peters JH, Ellison C, Innes JT. Safety and efficacy of laparoscopic cholecystectomy.
Ann Sur g 199 | ;213:3-12.
170 Peters JH, Heimbucher J, Kauer WKH, Incarbone R, Bremner CG, DeMeester TR.
Clinical and physiological comparison of laparoscopic and open Nissen fundoplication. ,I
Am CoII Surg 1995:180:385-393.
171 Pettersson GB, Bombeck CT, Nyhus LM. The lower esophageal sphincter: mechanisms of opening and closure. Surgery 1980;88:307-3I4.
l72Pke GK, Bessell JR, Mathew G, Watson DI, Mitchell PC, Jamieson GG. Changes in fibrinogen levels in patients undergoing open and laparoscopic Nissen fundoplication.
Aust NZ J Surg 1996;66:94-96.
173 Pilowsky I. Dimensions of hypochondriasis. Br J Psychiat 1967;113:89-93.
174 Pilowsky I, Spence ND. Patterns of illness behaviour in patients with intractable pain.
J Psychosom Res I975;19:279-287.
175 Pilowsky I, Spence ND. Manualfor the lllness Behaviour Questionnaire. Adelaide:
University of Adelaide, 1983; 248
176 Pilowsky I, Spence ND, Waddy JL. Illness behaviour and coronary artery by-pass surgery. J Psychosom Res 1979;23:39-44.
177 Pitchq DE, Curet MJ, Martin DT, Castillo RR, Gerstenberger PD, Vogt D, Ztcker
KA. Successful management of severe gastroesophageal reflux disease with laparoscopic
Nissen fundoplication. Atn J Surg 1994;168:5 47 -554.
178 Pitcher DE, Curet MJ, Martin DT, Vogt DM, MasonJ,Zucker KA. Successful laparoscopic repair of paraesophageal hernia. Arch Sur g 1995 ;130:590-596.
179 Pope CE. A dynamic test of sphincter strength: Its application to the lower esophageal sphincter. Gastro enterol 19 67 :52:7 7 9 -7 86.
180 Pope CE. Esophageal motility - who needs it? Gastroenterol1918;74:1088-1094.
181 Rattner DW, Brooks DC. Patient satisfaction following laparoscopic and open
antireflux surgery. Ar ch S ur g I99 5 ;130 :289 -29 4.
182 Reid DB, Winning T, Bell G. Pneumothorax during laparoscopic dissection of the
diaphragmatic hiatus. Br J Surg 1993;80:670.
183 Richards KF, Fisher KS, Flores JH, Christensen BJ. Laparoscopic Nissen fundoplication: Cost, morbidity, and outcome compared with open surgery. Surg Lapar
Endo s c 199 6:6:1 40 - I 43.
184 Richardson WS, Trus TL, Thompson S, Hunter JG. Nissen and Toupet fundoplications effectively inhibit gastroesophageal reflux irrespective of natural anatomy
and function. Surg Endosc 7997;11,:26I-263. 249
185 Rosati R, Bona S, Fumagalli U, Chella B, Peracchia A. Laparoscopic treatment of paraesophageal and large mixed hiatal hernias. Surg Endosc 1996;10:429-43I-
186 Rossetti M, HellK. Fundoplication for the treatment of gastroesophageal reflux in hiatal hernia. World J Surg 1977:l:439-444.
187 Sartorelli KH, Rothenberg SS, Karrer FM, Lilly JR. Thoracoscopic repair of hiatal hernia following fundoplication: A new approach to an old problem. J Laparoendosc Surg
1996;6:S91-S93.
188 Schauer PR, Meyers WC, Eubanks S, Norem RF, Franklin M, Pappas TN.
Mechanisms of gastric and esophageal perforations during laparoscopic Nissen fundoplic ation. Ann S ur g 199 6 :223 :43 - 52.
189 Schirmer BD, Edge SB, Dix J, Hyser MJ, Hanks JB, Jones RS. Laparoscopic
cholecystectomy. Treatment of choice for symptomatic cholelithiasis. Ann Surg
1991;213:665-676.
190 Schlumpf R, Klotz HP, Wehrli H, Herzog U. A nation's experience in laparoscopic cholecystectomy. Surg Endosc 1994;8:35-4I.
191 Schnieder J, Koehler RH, Brams DM, Kleinschmidt J, Smith RS. A standardised
approach to teaching the laparoscopic Nissen fundoplication. Surg Endosc
I99 5 :9 :240. (Abstract)
192 Schon RT. Laparoscopic upper abdominal operations and mesenteric infarction. "I
Laparoendo s c Surg 1995 ;5:389-39 1.
193 Skinner DB. Pathophysiology of gastroesophageal reflux. Am Surg 1985;202:546- s56. 2s0
194 Snow LL, Weinstein LS, Hannon JK. Laparoscopic reconstruction of
gastroesophageal anatomy for the treatment of reflux disease. Surg Endosc 1995;9:774-
780.
195 Spechler SJ. Comparison of medical and surgical therapy for complicated
gastroesophageal reflux disease in veterans. NEJM 7992;326:786-792.
196 Stallard N. Pneumomediastinum during laparoscopic Nissen fundoplication.
Anae s the s i a 199 5 :50 :667 - 668.
197 Stein HJ, Bremner RM, Jamieson J, DeMeester TR. Effect of Nissen fundoplication
on esophageal motor function. Arch Surg 1992;127:7 88-791.
198 Stein HJ, DeMeester TR. Surgical management of esophageal disorders- Curuent
Opinion in Gastro enterolo gy 1992;8:613 -623.
199 Steptoe A. The links between stress and illness. J Psychosom Res 1991;35:633-644.
200 Swanstrom LL, Marcus DR, Galloway GQ.Laparoscopic Collis gastroplasty is the
treatment of choice for the shortened esophagus. Am J Surg 7996;17l:477-48I.
201 Swanstrom LL, Pennings JL. Safe laparoscopic dissection of the gastroesophageal junction. Am J Surg 1995;169:507-511.
202 Swanstrom LL, Pennings JL. Laparoscopic control of short gastric vessels. J Am Coll
S ur g I99 5 ;l8l:3 47 -3 5 L.
203 Thor KBA, Silander T. A long-ûerm randomized prospective trial of the Nissen procedure versus a modified Toupet technique. Ann Surg 1989;210:719-724. 25r
204 Toppino M, Morino M, Garrone C, Comba A. Disappointing long-term results of laparoscopic adjustable silicone gastric banding. Surg Endosc 1996;10:572.(Abstract)
205 Toupet A. Technique d'oesophago-gastroplastie avec phrenogastropexie appliquee dans la cure radicale des hernies hiatales et comme complement de I'operation d'heller dans les cardiospasmes. MedAcad Chir 1963:89:394.
206 Trus TL, Laycock'WS, Branum G, Waring JP, Mauren S, Hunter JG. Intermediate follow-up of laparoscopic antireflux surgery.Am J Surg 1996;17l:32-35-
207 Tunel WP, Smith EI, Carson JA. Gastroesophegeal reflux in Childhood. Ann Surg
1983;197:560-565.
208 Viste A, Horn A, Lund-Tonnessen S. Reactive pleuropericarditis following laparoscopic fundoplication. S ur g Lapar Endo s c 1997 ;7 :206-208.
209 Walker SJ, Holt S, Sanderson CJ, Stoddard CJ. Comparison of Nissen total and Lind partial transabdominal fundoplication in the treatment of gastro-oesophageal reflix. Br J
Surg 1992;79:470-414.
210 V/aring JP, Hunter JG, Oddsdottir M,'Wo J,Katz E. The preoperative evaluation of patients considered for laparoscopic antireflux surgery. Am J Gastroenterol1995;90:35-
38.
211 V/atson A, Jenkinson LR, Ball CS, Norris TL. A more physiological alternative to total fundoplication for the surgical correction of resistant gastro-oesophageal refhx. Br J
Surg I99I;78: 1088- 1094. 252
212 Watson A, Spychal RT, Brown MG, Peck N, Callender N. Laparoscopic
'physiological' antireflux procedure: preliminary results of a prospective symptomatic and objective study. Br J Surg 1995;82:651-656.
213 Watson DI, Baigrie RJ, Jamieson GG. A learning curve for laparoscopic fundoplication. Definable, avoidable, or a waste of time? Ann Surg 1.996;224:198-203
214 Watson DI, Chan ASL, Myers JC, Jamieson GG. Illness behaviour influences the outcome of laparoscopic antireflux surgery. J Am CoII Surg 1997;184:44-48-
215 Watson DI, Foreman D, Devitt PG, Jamieson GG. Preoperative grading of oesophagitis versus outcome following laparoscopic Nissen fundoplication. Am J
G as tr o e nt e r o I 1997 ;92:222-225 .
216 Watson DI, Gourlay R, Globe J, Reed MWR, Johnson AG, Stoddard CJ. Prospective randomised trial of laparoscopic versus open Nissen fundoplication. Gut 1994;35
(supplement 2): S 15.(Abstract)
217 Watson DI, Jamieson GG, Baigrie RJ, Mathew G, Devitt PG, Game PA.
Laparoscopic surgery for gastro-oesophageal reflux: beyond the learning curve. Br J Surg
1996;83:1284-1287.
218 V/atson DI, Jamieson GG, Britten-Jones R, Mitchell P, Game PA. Pneumothorax during laparoscopic dissection of the diaphragmatic hiatus. Br J Surg 1993;80: 1353-1354.
219 Watson DI, Jamieson GG, Devitt PG, Mathew G, Britten-Jones RE, Game PA,
Williams RS. Changing strategies in the performance of laparoscopic Nissen fundoplication as a result of experience with 230 operations. .Sørg Endosc 1995;9:961-
966. 253
220 Watson DI, Jamieson GG, Devitt PG, Mitchell PC, Game PA. Paraoesophageal hiatus hernia: an important complication of laparoscopic Nissen fundoplication- Br J Surg
1995;82:521-523.
221 Watson DI, Jamieson GG, Mitchell PC, Devitt PG, Britæn-Jones R. Stenosis of the esophageal hiatus following laparoscopic fundoplication. Arch Surg 1995;130:1014-1016.
222Watson DI, Mathew G, Pike GK, Baigrie RJ, Jamieson GG. Efficacy of anterior, posterior and total fundoplication in a pig model. Br J Surg I998;(in press)
223 Watson DI, Mathew G, Pike GK, Jamieson GG. Comparison of anterior, posterior and total fundoplication using a viscera model. Dis Esoph 1997;10:527 -530.
224Watson DI, Mathew G, Williams JAR. Impact of laparoscopic cholecystectomy in a major teaching hospital : clinical and hospital outcomes. Med J Aust 1995;163:527-530.
225 V/atson DI, Mitchell PC, Game PA, Jamieson GG. Pneumothorax during laparoscopic dissection of the oesophageal hiatus. Aust NZ J Surg 1996;667lI-712.
226 Watson DI, Pike GK, Baigrie RJ, Mathew M, Devitt PG, Britæn-Jones R, Jamieson
GG. Prospective double blind randomised trial of laparoscopic Nissen fundoplication with division and without division of short gastric vessels. Ann Surg 1997;226:642-652.
227 Watson DI, Reed MWR, Johnson AG, Stoddard CJ. Laparoscopic Fundoplication for
Gastro-oesophageal Reflux. Annals of the Royal College of Surgeons of England
1994;76:264-268.
228Weefts IM, Dallemagne B, Hamoir E, Demarche M, Markiewicz S, Jehaes C,
Lombard R, Demoulin JC, Etienne M, Ferron PE, Fontaine F, Gillard V, Delforge M. 254
Laparoscopic Nissen fundoplication: detailed analysis of I32 patients. Surg Lapar Endosc
1993;3:359-364.
229 Weston PV. A new clinch knot. Obstet Gynecol1991;78:144-747.
230'Wexner SD, Cohen SM, Ulrich A, Reissman P. Laparoscopic colorectal surgery - Are we being honest with our patients? . Dis Colon Rectum 1995;38:723-727.
231 Whiston RJ, Eggers KA, Morris RW, Stanatakis JD. Tension pneumothorax during laparoscopic cholecystectomy. Br J Surg 199l;78:1325-
232 Windsor JA, McCay H. Inguinal hernia repair by laparoscopic surgeons: Early experience and attitudes. Aust NZ J Surg 1995;65:470-474.
233 Woods MS, Traverso W, Kozarek RA. Characteristics of Biliary Tract Complications during laparoscopic Cholecystectomy: A multi-institutional study. Am J Surg
1994;167:21-34.
234Wyman JB, Dent J, Heddle R, Dodds WJ, Lewis I. The mechanism of belching. Aust
NZ J Med 1984;14:307.
235 Zucker KA, Bailey RW, Gadacz TR, Imbembo AL. Laparoscopic guided cholecystectomy. Am J Surg 199l;16l:36-44. 255 2s6
sEcTtoN 6
APPENDIX
Jamieson, G., Watson, D., Britten-Jones, R., Mitchell, P., and Anvari, M., (1994) Laparoscopic nissen fundoplication. Annals of Surgery, v. 220 (2), pp. 137-145.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1097/00000658-199408000-00004
Watson, D.I., Jamieson, G.G., Devitt, P.G., Matthew, G., Britten-Jones, R.E., Game, P.A., and Williams, R.S., (1995) Changing strategies in the performance of laparoscopic Nissen fundoplication as a result of experience with 230 operations. Surgical Endoscopy, v. 9 (9), pp. 961-966.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1007/BF00188451
Watson, D.I., Jamieson, G.G., Devitt, P.G., Mitchell, P.C., and Game, P.A., (1995) Paraoesophageal hiatus hernia: An important complication of laparoscopic Nissen fundoplication. British Journal of Surgery, v. 82 (4), pp. 521-523.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1002/bjs.1800820428
Watson, D.I., Jamieson, G.G., Mitchell, P.C., Devitt, P.G., and Britten-Jones, R., (1995) Stenosis of the esophageal hiatus following laparoscopic fundoplication. Archives of Surgery, v. 130 (9), pp. 1014-1016.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1001/archsurg.1995.01430090100029
Watson, D.I., Mitchell, P., Game, P.A., and Jamieson, G.G., (1996) Pneumothorax during laparoscopic mobilization of the oesophagus. Australian and New Zealand Journal of Surgery, v. 66 (10), pp. 711-712.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1111/j.1445-2197.1996.tb00723.x
Watson, D.I., Baigrie, R.J., and Jamieson, G.G., (1996) A learning curve for laparoscopic fundoplication: definable, avoidable, or a waste of time? Annals of Surgery, v. 224 (2), pp. 198-203.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1097/00000658-199608000-00013
Watson, D.I., Jamieson, G.G., Baigrie, R.J., Mathew, G., Devitt, P.G., Game, P.A., and Britten-Jones, R., (1996) Laparoscopic surgery for gastro-oesophageal reflux: beyond the learning curve. British Journal of Surgery, v. 83 (9), pp. 1284-1287.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1046/j.1365-2168.1996.02329.x
Watson, D.I., Foreman, D., Devitt, P.G., and Jamieson, G.G., (1997) Preoperative endoscopic grading of esophagitis versus outcome after laparoscopic Nissen fundoplication. American Journal of Gastroenterology, v. 92 (2), pp. 222-225.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
Baigrie, R.J., Watson, D.I., Myers, J.C., and Jamieson, G.G., (1997) Outcome of laparoscopic Nissen fundoplication in patients with disordered preoperative peristalsis. Gut, v. 40 (3), pp. 381-385.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1136/gut.40.3.381
Watson, D.I., Chan, A.S., Myers, J.C., and Jamieson, G.G., (1997) Illness behavior influences the outcome of laparoscopic antireflux surgery. Journal of the American College of Surgeons, v. 184 (1), pp. 44-48.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
Watson, D.I., Mathew, G., Pike, G.K., and Jamieson, G.G., (1997) Comparison of anterior, posterior and total fundoplication using a viscera model. Diseases of the Esophagus, v. 10 (2), pp. 110-114.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
Mathew, G., Watson, D.I., Myers, J.C., Holloway, R.H., and Jamieson, G.G., (1997) Oesophageal motility before and after laparoscopic Nissen fundoplication. British Journal of Surgery, v. 84 (10), pp. 1465-1469.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1111/j.1365-2168.1997.02812.x
Mathew, G., Watson, D.I., Myers, J.C., Holloway, R.H., Jamieson, G.G., (1997) Oesophageal motility before and after laparoscopic Nissen fundoplication. British Journal of Surgery, v. 84 (10), pp. 1465-1469, October 1997
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1111/j.1365-2168.1997.02812.x
Watson, D.I., Pike, G.K., Baigrie, R.J., Mathew, G., Devitt, P.G., Britten-Jones, R., and Jamieson, G.G., (1997) Prospective double-blind randomized trial of laparoscopic Nissen fundoplication with division and without division of short gastric vessels. Annals of Surgery, v. 226 (5), pp. 642-652.
NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.
It is also available online to authorised users at:
http://dx.doi.org/10.1097/00000658-199711000-00009