Abc of the upper gastrointestinal

ABC OF THEUPPER GASTROINTESTINAL TRACT

ABC OF THEUPPER GASTROINTESTINAL TRACT

Edited by

ROBERT PH LOGANSenior Lecturer, Division of Gastroenterology,

University Hospital, Nottingham

ADAM HARRISConsultant Physician and Gastroenterologist, Kent and Sussex Hospital, Tunbridge Wells

JJ MISIEWICZHonorary Consultant Physician and Honorary Joint Director,

Department of Gastroenterology and Nutrition, Central Middlesex Hospital, London

JH BARONHonorary Professorial Lecturer,

Mount Sinai School of Medicine, New York, USA

© BMJ Books 2002BMJ Books is an imprint of the BMJ Publishing Group

All rights reserved. No part of this publication may be reproduced,stored in a retrieval system, or transmitted, in any form or by anymeans, electronic, mechanical, photocopying, recording and/orotherwise, without the prior written permission of the publishers.

First published in 2002by BMJ Books, BMA House, Tavistock Square,

London WC1H 9JR

www.bmjbooks.com

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

ISBN 0-7279-1266-6

Typeset by BMJ Electronic ProductionPrinted and bound in Spain by GraphyCems, Navarra

Cover image depicts Helicobacter pylori bacterium. Coloured transmission electron micrograph (TEM) of a

section through a Helicobacter pylori bacterium. With permission from Dr Linda Stannard, UCT/Science Photo Library

v

ContentsContributors vi

Foreword ROY POUNDER vii

1 Implications of dyspepsia for the NHS 1RICHARD LOGAN and BRENDAN DELANEY

2 Oesophagus: Heartburn 4JOHN DE CAESTECKER

3 Oesophagus: Atypical chest pain and motility disorders 8JOHN BENNETT

4 Dysphagia 12WILLIAM OWEN

5 Epidemiology and diagnosis of Helicobacter pylori infection 16ROBERT PH LOGAN and MARJORIE M WALKER

6 Pathophysiology of duodenal and gastric ulcer and gastric cancer 19JOHN CALAM and JH BARON

7 Management of Helicobacter pylori infection 22ADAM HARRIS and JJ MISIEWICZ

8 Indigestion and non-steroidal anti-inflammatory drugs 26JM SEAGER and CJ HAWKEY

9 Upper gastrointestinal haemorrhage 30HELEN J DALLAL and KR PALMER

10 Indigestion: When is it functional? 33NICHOLAS J TALLEY, NGHI PHUNG and JAMSHID S KALANTAR

11 Upper abdominal pain: Gall bladder 37CD JOHNSON

12 Cancer of the stomach and pancreas 41MATTHEW J BOWLES and IRVING S BENJAMIN

13 Anorexia, nausea, vomiting, and pain 45RC SPILLER

Index 49

vi

ContributorsJH BaronHonorary Professorial Lecturer, Mount Sinai School ofMedicine, New York, USA

Irving S BenjaminProfessor of Surgery, Guy’s, King’s and St Thomas’s School ofMedicine, Kings College, London

John BennettTreasurer, Royal College of Physicians, London

Matthew J BowlesConsultant Liver Transplant and General Surgeon, King’sCollege Hospital, London

John de CaesteckerConsultant Gastroenterologist, Glenfield Hospital NHS Trust,Leicester

John CalamProfessor of Gastroenterology, Imperial College School ofMedicine, London

Helen J DallalSpecialist Registrar, Royal Aberdeen Infirmary, Aberdeen

Brendan DelaneyReader in Primary Care and General Practice, University ofBirmingham, Birmingham

Adam HarrisConsultant Physician and Gastroenterologist, Kent and SussexHospital, Tunbridge Wells

CJ HawkeyProfessor of Gastroenterology, University Hospital, Nottingham

CD JohnsonReader in Surgery, University Surgical Unit, SouthamptonGeneral Hospital, Southampton

Jamshid S KalantarStaff Specialist in Gastroenterology, Department of Medicine,University of Sydney, Nepean Hospital, Penrith, Australia

Richard LoganProfessor of Clinical Epidemiology, University Hospital,Nottingham

Robert PH LoganSenior Lecturer, Division of Gastroenterology, UniversityHospital, Nottingham

JJ MisiewiczHonorary Consultant Physician and Honorary Joint Director ofthe Department of Gastroenterology and Nutrition, CentralMiddlesex Hospital, London

William OwenConsultant Surgeon, Guy’s and St Thomas’s Hospital Trust,London

KR PalmerConsultant Gastroenterologist, Western General Hospital,Edinburgh

Nghi PhungGastroenterologist, Department of Medicine, University ofSydney, Nepean Hospital, Penrith, Australia

RC SpillerProfessor of Gastroenterology, University of Nottingham

JM SeagerResearch Pharmacist, Division of Gastroenterology, UniversityHospital, Nottingham

Nicholas J TalleyProfessor of Medicine, Department of Medicine, University ofSydney, Nepean Hospital, Penrith, Australia

Marjorie M WalkerSenior Lecturer, Department of Histopathology, ImperialCollege, London

Robert Logan, Adam Harris, George Misiewicz and Hugh Baron, the Editors of ABC of the Upper Gastrointestinal Tract, are all closelyassociated with Sir Francis Avery Jones and the Central Middlesex Hospital. Thirty years ago this month I started with Sir Francis ashis last Medical Registrar, and that is why I have been invited to write this Foreword.

Sir Francis was famous for saving the lives of pre-war patients with bleeding peptic ulcers by offering food and fluid, rather thanstarvation. With Sir Richard Doll, he introduced controlled clinical trials for peptic ulcer disease (indeed for all diseases, in duecourse), showing that carbenoxolone was better than placebo for gastric ulcer healing. He relied upon barium meals to diagnoseulcers and cancer.

How would Sir Francis recognise this excellent collection of 13 articles about the upper gastrointestinal tract? Well, a lot haschanged because of the three factors that damage this part of the human body – acid, Helicobacter pylori and non-steroidal anti-inflammatory drugs. Our ability to image this part of the body is another big development. Finally, having thought that we’d madesurgeons redundant, they have staged a comeback.

For 25 years we have been able to decrease acid secretion moderately with histamine H2 receptor antagonists, and for 15 yearswe have had powerful control using proton pump inhibitors. We thought that this could deal with everything acidic, albeit with the need for indefinite maintenance treatment. But in 1981 Warren and Marshall identified, with typical Australian confidence, H pylori. Eradication was the new excitement – mostly curing duodenal ulceration, but leaving a “healthy” stomach with plentiful acidsecretion, and the liability for acid reflux. So Sir Francis would be missing peptic ulceration, and would be more concerned aboutreflux, Barrett’s oesophagus, and oesophageal cancer. Our ageing population welcomed the pain relief and mobility achieved bynon-steroidal anti-inflammatory drugs – only to be troubled by peptic ulcers that were often silent until they haemorrhaged orperforated. The best of times, the worst of times …

The improvements in our ability to investigate this area have been breathtaking: endoscopy, ultrasound, retrograde cholan-giopancreatography, computerised tomography, magnetic resonance cholangiopancreatography, endoscopic ultrasound, 13C-ureabreath testing, oesophageal manometry, and pH profiles. Organic disease can be excluded by exhaustive investigations, andfunctional disorders are the new challenge.

Finally, the development of minimally invasive surgery leads the surgical renaissance. No longer gastric resections for chronicpeptic ulcer disease – now laparoscopic cholecystectomy for gallstones identified by ultrasound, or fundoplication for the acid refluxthat now damages the oesophagus.

This explains why I welcome this collection of articles about the upper gastrointestinal tract – it is an amazing story of rapidadvance and excitement, of genuine medical and surgical achievement, with real improvement in the quality of life for many of our patients. The Editors and authors of ABC of the Upper Gastrointestinal Tract have created a book that would have intrigued Sir Francis.

Roy PounderRoyal Free Hospital, London

vii

Foreword

1 Implications of dyspepsia for the NHSRichard Logan, Brendan Delaney

There is no precise definition of dyspepsia. It can be definedpragmatically as upper abdominal or retrosternal pain, with orwithout other symptoms thought to be arising from the uppergastrointestinal tract—which is the approach that has beengenerally adopted by epidemiological studies.

It has been suggested that dyspeptic symptoms can becategorised as ulcer-like, reflux-like, and dysmotility-like as aguide to the underlying cause. These groups, however, overlapconsiderably, with mixed patterns being common. Symptompatterns are not strong predictors of underlying disease.Recently it has been proposed that if heartburn or acidregurgitation are the dominant symptoms then these aresufficiently accurate predictors of gastro-oesophageal reflux tomake a safe and accurate diagnosis (see next article). Fewer thana fifth of sufferers have this symptom pattern, and the predictiveaccuracy needs confirmation.

PrevalenceDyspepsia is common: in a recent UK survey 40% of adultsreported having had one or more dyspeptic symptoms in theprevious year, and about a half described these as beingmoderate to severe. Of this group, more than half were takingdrugs for dyspepsia (40% of which were prescribed) and 22%had seen their general practitioner about dyspepsia in theprevious year. Thus, 9% of all those interviewed reportedconsulting their doctor about dyspepsia in the previous year.

Most dyspeptic patients have no clinically significantabnormalities on investigation. Up to 20% may have endoscopicreflux oesophagitis, and 15-20% may have peptic ulcer disease,including duodenitis. A declining proportion, currently around2%, will have a gastric or oesophageal cancer, with other“alarm” symptoms such as dysphagia or weight loss usuallybeing present.

Whether dyspepsia is becoming more common is unclear,but general practice consultations for non-ulcer dyspepsia havebeen increasing. In contrast, morbidity and mortality resultingfrom peptic ulcer disease is declining; mortality fromoesophageal cancer has now overtaken mortality from gastriccancer, which has declined steeply over the past 50 years.

Cost to the NHSThe management of these patients has a considerable impacton the NHS. At any one time up to 4% of the population arethought to be taking prescribed drugs for dyspepsia. In the pastfew years the costs of these prescriptions have risendramatically and now account for over 10% of drugexpenditure in primary care (471m in 1999 in England andWales), although this may now have peaked.

Investigation is also costly. The number of uppergastroscopies performed each year in the United Kingdom hasbeen steadily increasing and was thought to be over 450 000 in1996, a little over one endoscopy for every 100 adults in Englandand Wales. Endoscopy has been estimated to cost £80-£450 perprocedure depending on the hospital (NHS Reference costs1998). In addition, assessment of dyspepsia and abdominal painis one of the commonest reasons for referral to hospital.

Box 1.1 “Alarm” symptoms in patients with dyspepsiasuggesting possibility of malignant diseasex Anaemiax Loss of weightx Anorexiax Recent onset of progressive symptoms ( < 3 months)x Melaena or haematemesisx Dysphagia

Table 1.1 Patients consulting with upper gastrointestinaldisorders in general practitioner morbidity surveys

No of patients(per 10 000/year)

Condition (ICD code) 1981-2 1991-2 ChangeNon-ulcer dyspepsia: 178 330 85%Oesophagus (530) 24 103Gastritis or duodenitis (535) 27 74Other disorders (536) 127 153

Peptic ulcer (531-534) 57 52 − 9%All disorders 720 866 20%

Figure 1.1 Dyspepsia is not new and has been known throughout history(Indigestion by Cruickshank (1792-1872))

Year

Annu

al m

orta

lity/

mill

ion

popu

latio

n

1950 1960 1970 1980 1990 19990

100

200

300

400

500MenWomenMen

Gastric cancer

Oesophageal cancerWomen

Figure 1.2 Mortality from gastric and oesophagealcancer in England and Wales 1950-99

1

To these costs one might also add the costs of managingcomplications such as gastrointestinal bleeding from pepticulceration and oesophagitis.

Managing dyspepsia in primary careWhy do patients consult with dyspepsia?According to the health belief model, the decision to consult ageneral practitioner is determined by a person’s perception ofthe likelihood of serious disease and the potential for cure.

Five factors are thought to influence whether a patientconsults a doctor: how the patient perceives the problem, howthe patient’s peers perceive the problem, the availability ofmedical care, the availability of non-medical treatments, andwhether the patient can afford treatment. Other triggers—suchas an interpersonal crisis; interference with personalrelationships, work, or physical functioning; sanctioning; andsetting of external time criteria—are required to force amedicalisation of the symptoms before they are perceived asillness.

Qualitative studies have shown that patients with dyspepsiaare concerned with finding causal relations between life eventsand their symptoms. “Stomach disease” is most commonlylinked to stress and worry. No difference has been foundbetween people who consult a doctor about their dyspepsia andthose who do not in the frequency or subjective severity of theirsymptoms, but those who do consult have significantly more lifeevents. People consulting with dyspepsia are more likely tobelieve that their symptoms are due to serious illness, heartdisease, ulcers, or cancer in particular. The challenge for generalpractitioners is to maximise detection of serious and treatableconditions while minimising the cost and adverse effects ofinvestigation.

Strategies for managing dyspepsia in primary care

Initial empiric treatment with antacids or anti-secretory drugsA period of empiric treatment with antacids or H2 receptorantagonists has been the traditional strategy for managingpatients with dyspepsia first consulting their doctor. Thisstrategy recognises that most patients’ symptoms are episodicand remit spontaneously and that the risk of peptic ulcerbleeding or perforation is extremely low. It also assumes thatearly diagnosis in the rare patient with malignant disease andno alarm symptoms has little effect on outcome.

However, this strategy has several major limitations. It takeslittle account of the fact that many dyspeptic people consulttheir doctor only after months, if not years, of symptoms andself medication. It also fails to meet the expectations of patientswho are increasingly informed and often expect referral forinvestigation or prescription of more powerful drugs. It offerslittle reassurance to those patients who believe that theirsymptoms are due to more serious disease. In controlled trialsempiric treatment results in the lowest scores for patientsatisfaction.

Investigate all patients by early endoscopyIt has been argued that, as most dyspeptic patients end uphaving some investigation anyway, earlier investigation mayprove more cost effective. However, cost effectiveness modellinghas shown that if the incidence of peptic ulcer disease indyspeptic patients is less than 10%, it would take more than fiveyears for the costs of investigation to be recouped in savingsfrom fewer prescriptions. Results of several large trials based inprimary care of the cost effectiveness of this approach areexpected soon.

Box 1.2 Why do patients consult withdyspepsia?Major factorsx Patient’s perception of the problemx Patient’s peers’ perception of the problemx Availability of medical carex Availability of non-medical treatmentsx Cost to the patient

Minor factorsx Interference with work, personal life, or physical

functioningx Medicalisation of symptoms

Box 1.3 Causes of dyspepsiaFindings from endoscopy in 2659 patients aged> 40 years consulting their doctor for dyspepsia forthe first time:Hiatus hernia or oesophagitis 19%Peptic ulcer:Gastric 6%Duodenal 10%

Duodenitis 4%Gastric or oesophageal cancer 3%Normal (including gastritis only) 59%

Year

Annu

al c

ost (

£m)

1992 1993 1994 1995 1996 1997 19980

100

200

300

400

500

1991

Proton pump inhibitorsH2 receptor antagonistsTotal

Figure 1.3 Cost of prescriptions for ulcer healing drugsin England and Wales

Perc

enta

ge o

f sam

ple

Symptom

0

5

10

15

20

25

30

Epiga

stric

pain

Regurg

itatio

n

Acid ta

ste

Fulln

ess

Heartb

urn

Figure 1.4 Symptoms of dyspepsia reported by UKadults over the previous year

ABC of the Upper Gastrointestinal Tract

2

A recent Cochrane review has shown that initial endoscopymay be associated with a 15% relative reduction in symptomscompared with empiric acid suppression. A primary care trial inpatients aged over 50 years has shown that this small reductionin symptoms might be cost effective if the unit cost ofendoscopy could be kept to £100.

Only investigate a proportion of dyspeptic patientsThe rationale behind this proposal is to increase the diagnosticyield of endoscopy without missing any serious causes in thosenot investigated. Criteria involving age (related to gastric cancerrisk), consumption of non-steroidal anti-inflammatory drugs,and “alarm” symptoms have been proposed.

More recently, non-invasive tests for Helicobacter pylori havebeen suggested: lack of H pylori infection is a good predictor ofthe absence of peptic ulcer or gastric cancer. A retrospective studyexamined the effect of screening for H pylori by serology beforeendoscopy in patients aged under 45 and not takingnon-steroidal anti-inflammatory drugs. It found that positiveserology was highly predictive of endoscopic abnormalities, andendoscopy workload would have been reduced by 23% if only thepatients seropositive for H pylori had been investigated. However,recent prospective trials have suggested that in general practicesnear patient serology tests may lack sufficient accuracy and thatthe strategy might even increase endoscopic workload.

Test for H pylori and treatUp to 15% (and higher in some areas such as Glasgow) ofdyspeptic patients infected with H pylori may have peptic ulcerdisease, so there is an argument for treating such patients,without first proving the presence of an ulcer by endoscopy.This may have the added advantage of preventing peptic ulcersor gastric cancer that might develop in the future as well astreating current disease.

However, the successful eradication of H pylori requirescombination therapy and carries a risk of increasing bacterialresistance. As the prevalence of H pylori infection in thedeveloped world is steadily declining, treating patients withouttesting for infection is unlikely to be cost effective.

In contrast, decision analysis models suggest that testing forH pylori and treating those who test positive (“test and treat”) islikely to be one of the most cost effective approaches in primarycare. Three randomised controlled trials have shown thattesting for and treating H pylori was as effective as endoscopybased management in controlling symptoms in patientsreferred to secondary care, but cost less because many fewerendoscopies were performed. However, the long term effects oferadicating H pylori in patients without peptic ulcer disease areuncertain, and the effects on resource use compared withempiric acid suppression are unknown.

Role of hospitalsThe appropriate management of dyspeptic patients in primarycare relies on a combination of making a diagnosis todetermine the most effective treatment and reducing theuncertainty experienced by both doctors and patients. In termsof potentially effective treatment two groups of patients shouldbe actively sought— those with peptic ulcer disease, inducedeither by H pylori infection or non-steroidal anti-inflammatorydrugs, and those with early gastric cancer.

Endoscopy is not the only investigation that may be used tosupport management: tests for H pylori infection, particularlythe highly accurate 13C-urea breath tests and stool antigen tests,may be increasingly important in primary care. Specialists havean important role in coordinating local endoscopy services andin treating patients with difficult to manage dyspepsia.

Clinical summaryx Patients aged under 55 presenting with dyspepsia for the first time

should be managed initially with a short course of antacid orantisecretory drugs (based on consensus guidelines)

x Patients still symptomatic should be tested for H pylori infection (bylaboratory serum enzyme linked immunosorbent assay (ELISA),stool antigen test, or 13C-urea breath test) and, if positive,investigated by endoscopy or given eradication treatment

x Treatment for H pylori infection without testing is notrecommended as most patients treated will not be infected (cohortstudy)

x Patients with gastric and duodenal ulceration, newly diagnosed orstill requiring treatment, should receive H pylori eradicationtreatment (meta-analysis of randomised trials)

x Unless a major complication has occurred (such as bleeding), cureshould be taken as relief of symptoms without proceeding to abreath test (prospective cohort study)

x Patients with persistent symptoms should have a 13C-urea breathtest to confirm H pylori eradication (consensus guidelines)

x Treatment of asymptomatic H pylori infection is not recommended

Indigestion is reproduced with permission of the Wellcome Library. Thetable of patients consulting with upper gastrointestinal disorders is derivedfrom the Office of National StatisticsMorbidity Statistics from GeneralPractice—Fourth national study 1991-92. The graph of costs of drugprescriptions is based on data from the Prescription Pricing Authority. Thetable of causes of dyspepsia is adapted from Hallisey et al, BMJ1990;301:513-5. The graph of prevalence of symptoms of dyspepsia isadapted from Penston and Pounder, Aliment Pharmacol Ther 1996;10:83-9.The electron micrograph of H pylori is reproduced with permission ofDr Linda Stannard and Science Photo Library.

Figure 1.5 Transmission electron micrograph ofHelicobacter pylori

Implications of dyspepsia for the NHS

3

2 Oesophagus: HeartburnJohn de Caestecker

Gastro-oesophageal reflux disease (GORD) is defined assymptoms or mucosal damage (oesophagitis) resulting from theexposure of the distal oesophagus to refluxed gastric contents.However, the symptoms of reflux oesophagitis do not equatewith mucosal damage, and patients with endoscopic evidence ofoesophagitis do not necessarily have the worst symptoms.

In primary care GORD is therefore best thought of in termsof symptoms: symptom control is the aim of most managementstrategies, and indeed typical symptoms can guide doctors tothe correct diagnosis. Since frequency and intensity ofsymptoms are poorly predictive of the severity of mucosaldamage, with the converse also applying, endoscopy may be lessuseful than commonly perceived. A variety of other tests areavailable to diagnose and assess the severity of disease ifsymptoms are atypical and results of endoscopy normal.

Nevertheless, oesophagitis resulting from GORD hasbecome the commonest single diagnosis resulting fromendoscopy carried out for dyspepsia, although whether thisrepresents a true increase in prevalence or simply reflects achange in referral practice is unclear. There is little doubt that aspectrum of severity of disease exists, with most affected peoplenever consulting a doctor and only a minority with unremittingsymptoms or complications from the disease receiving attentionfrom hospital specialists. Consequently, treatment of patientspresenting in general practice may not be best guided by theoutcome of most clinical trials, which have recruited patientsfrom those referred to hospital.

Terminology and aetiologyOesophagitis refers to endoscopic or histological evidence of

an acute inflammatory process in the oesophagus. Only about60% of patients in whom GORD is eventually diagnosed haveendoscopic evidence of oesophagitis. Some evidence suggeststhat among patients in the community or those with atypicalpresenting symptoms the proportion with oesophagitis may beeven lower.

Hiatus hernia is present when gastric mucosal folds areobserved more than 2-3 cm above the diaphragm by endoscopyor barium radiology and is found in about 30% of people agedover 50 years. However, most patients with an hiatus hernia donot have GORD, but about 90% of patients with markedoesophagitis have hiatus hernia. Thus, hiatus hernia may notresult in GORD but can contribute to the disease. Hiatus herniaitself rarely gives rise to symptoms, although a large hernia mayundergo torsion (volvulus) to cause acute epigastric orretrosternal pain with vomiting.

Both oesophageal and gastric factors affect the occurrenceof reflux. The critical factor is lower oesophageal sphincterincompetence: most reflux occurs during transient relaxation ofthe lower oesophageal sphincter resulting from failed swallows(swallows not followed by a normal oesophageal peristalticwave) and gastric distension (mostly after meals). Recentevidence has indicated that the diaphragmatic crural fibressurrounding the oesophageal hiatus act as an external sphincterin concert with the intrinsic lower oesophageal sphincter.Failure of this crural mechanism may allow a hiatus hernia tooccur. The hernial sac may additionally provide a sump ofgastric contents available for reflux once the lower oesophagealsphincter relaxes. Oesophageal acid clearance depends both on

Classic symptoms

• Heartburn or pharyngeal "acid" regurgitation. Typically after meals, on bending, and lying flat

Atypical symptoms

• Dental enamel erosion

• Ear, nose and throat symptomsDysphonia, hoarseness and sometimes

globus hystericus. Pathologicalfindings include posterior laryngitis,

"contact ulcer" and subglottic stenosis

• Vomiting

• Respiratory symptomsRecurrent aspiration pneumonia

Asthma, especially nocturnal,postprandial and "difficult to

control asthma"Idiopathic" cough

• Chest painMay resemble angina and

appear on physical exertionOften unusual, atypical and

non-specific

• Epigastric pain

Figure 2.1 Presentation of gastro-oesophageal reflux disease

Figure 2.2 Four grades of endoscopic oesophagitis. Top left: Grade 1 (singleerosion with a sentinel fold of gastric mucosa). Top right: Grade 2. Bottomleft: Grade 3. Bottom right: Grade 4 (stricture)

4

swallowed saliva and intact lower oesophageal peristalsis, whichis impaired in about 30% of patients with GORD. Gastric acidproduction is usually normal in GORD, while delayed gastricemptying occurs in about 40%.

Duodenogastro-oesophageal reflux of bile may play asubsidiary role to that of gastric acid and pepsin in patients withan intact stomach and has been implicated in the pathogenesisof Barrett’s oesophagus and its sequelae.

Odynophagia, pain on swallowing, should be distinguishedfrom dysphagia, difficulty in swallowing (see next article).

Clinical features and presentationThere is a spectrum of clinical presentation, ranging fromsymptoms alone to complications resulting from mucosaldamage. Up to 40% of patients seen in hospital in whom refluxis eventually diagnosed have symptoms other than classicheartburn or pharyngeal acid regurgitation, including a varietyof respiratory and pharyngeal symptoms.

Natural course of GORDThe condition is characteristically chronic and relapsing: infollow up studies at least two thirds of patients continue to takedrugs continuously or intermittently for reflux symptoms for upto 10 years. Symptoms disappear in less than a fifth of thosetaking no drugs, and in the short term endoscopic evidence ofoesophagitis may come and go independently of symptoms.There is no evidence that patients inevitably go on to developsevere erosive oesophagitis, Barrett’s oesophagus, or stricture.

Symptomatic relapse after discontinuing treatment iscommon and is chiefly dependent on initial severity ofoesophagitis. In studies with large proportions of patientshaving initial severe oesophagitis, relapse rates of up to 80% atsix months have been reported.

InvestigationGORD is a disease of different facets. No single investigation iscapable of infallibly diagnosing the condition nor of assuring itsrelevance to symptoms in an individual case.

Choice of clinical investigation depends on presentation, thepatient’s age, presence or absence of “alarm” symptoms (such asdysphagia or weight loss), and the question to be answered.Thus, if the question is to decide if mucosal injury is present,endoscopy or barium radiology is most appropriate, whereas anacid perfusion test or 24 hour oesophageal pH monitoring withobservation of symptom-reflux association is the most usefultest in deciding if symptoms are due to oesophageal acidexposure.

Despite these considerations, in practice endoscopy is themost commonly performed initial investigation, combininginspection of the oesophagus with that of the stomach andduodenum to exclude other causes of dyspepsia. In up to twothirds of patients, however, the results of endoscopy are normal,particularly if patients are taking proton pump inhibitors orH2 antagonists at the time of investigation.

Therefore, in young patients with longstanding typicalsymptoms (heartburn or pharyngeal acid regurgitation aftermeals and with postural changes) no investigation is necessary.Atypical symptoms, dysphagia, or presentation with shortduration of symptoms in patients aged over 55 years usuallyrequire investigation. There are no a priori grounds fordiagnosis or treatment of Helicobacter pylori infection in mostpatients, since there is no evidence at present of an association.

Impaired clearance mechanisms:SalivaDistal oesophageal motility

Impaired "external sphincter"of diaphragmatic crural fibres

Acid and pepsin(the main noxious

factors, productionnot usually increased)

Impaired "intrinsic" loweroesophageal sphincter

Hiatus hernia

Delayed gastricemptying

Duodenogastricreflux

Figure 2.3 Main pathophysiological mechanisms in gastro-oesophagealreflux disease

Endoscopicoesophagitis

Pathological reflux(24 hour pHmonitoring)

Symptomatic reflux(history, acid perfusion test,

24 hour pH monitoring)

"Classic" GORD(symptoms, abnormal reflux,

oesophagitis)

Figure 2.4 Overlap between symptoms, endoscopicevidence of damage, and physiological findings in refluxoesophagitis

Time (hours)

Oeso

phag

eal p

H

10:00 14:00 18:00 22:00 02:00 06:00Day 2Day 1

M

M = Meal = Heartburn

M M

Sleep period

10:000

2

3

4

5

6

7

8

1

Figure 2.5 24 Hour recording of oesophageal pH in patient withgastro-oesophageal reflux disease but normal endoscopic appearance. Noteclose association between symptoms and reflux and large amount of daytimeand night time reflux. (Reflux=intraoesophageal pH <4)

Oesophagus: Heartburn

5

Indeed, there is some evidence that eradication of infection, ifpresent, may actually make acid suppression with proton pumpinhibitors more difficult in GORD.

TreatmentAimsFor most patients, the aim is acceptable symptom control usingthe least treatment necessary to achieve this. Therefore, ifsymptom control is the aim, endoscopy to assess healing ofoesophagitis is unnecessary. Indeed, it is now known that, atleast for patients treated with proton pump inhibitors, absenceof symptoms on treatment equates with healing of oesophagitis.

For those with complications, such as stricture or bleedingfrom oesophagitis, the aim will be long term healing ofoesophagitis.

Patients with Barrett’s oesophagus have a risk of between1 in 50 and 1 in 200 of developing adenocarcinoma of theoesophagus. Many gastroenterologists therefore recommendyearly or biennial endoscopic screening with multiple biopsiesto detect dysplasia. Patients with severe dysplasia often have anundetected early cancer and so are offered oesophagectomy.Surveillance of patients with Barrett’s oesophagus to detectsevere dysplasia or early cancer is controversial, partly becauseits benefits have not been established by well designedrandomised controlled trials. Clearly a surveillance policy isinappropriate in elderly patients who are unfit for surgery.Endoscopic ablation of the abnormal columnar mucosa inBarrett’s oesophagus by photodynamic laser or thermalmethods looks promising and may become standard treatment.It must be combined with high doses of proton pump inhibitorsor antireflux surgery to prevent continuing acid reflux.

General measuresPatients should be advised to lose weight if overweight. There isno formal evidence to support this assertion, but success(though rarely achieved) may result in improved symptomcontrol. Raising the head of the bed on 15 cm wooden blockshas been shown in a controlled trial to improve symptoms andhealing of oesophagitis. There is little evidence that avoidanceof specific foods has much effect on the course of the disease,but many patients have already identified and stopped eatingfoods that produce symptoms before consulting their doctor.Other potentially damaging drug treatment should also bereviewed.

While the benefits associated with these general measuresmay be unproved, they allow patients to be involved withdecision making and may help them avoid over-medicalisingtheir condition.

Antacids and alginatesAntacids are effective for short term relief of symptoms.Although their efficacy is difficult to confirm in controlled trials,many sufferers, particularly those who do not consult a doctor,rely on self medication with antacids. Alginates work by forminga floating viscous raft on top of the gastric contents thatprovides a physical barrier to prevent reflux. To maximise thiseffect, they are therefore best taken after meals, otherwise theyrapidly empty from the stomach and thus give only transientrelief of symptoms by virtue of their antacid content.

Acid suppression therapyThe two major classes of agent available are the H2 receptorantagonists and the proton pump inhibitors. There is littledoubt that proton pump inhibitors are more rapidly and

Box 2.1 Investigations for gastro-oesophageal reflux diseaseBarium radiologyStrengthsx Good for structural lesions (such as stricture) and detection of

motor abnormalitiesWeaknessesx Poor for assessing minor mucosal abnormalitiesx Detection of “free” reflux of barium is insensitive (40%) though

specific (90%) for GORD

EndoscopyStrengthsx Best for assessment of oesophagitisx May detect other potential causes of symptomsx Allows for biopsyWeaknessesx Detection of oesophagitis is insensitive for GORD ( < 60%)

24 hour oesophageal pH monitoringStrengthsx Patients monitored at home, where symptoms are most likely to

occurx Can quantify reflux and measure association with symptomsx Useful for atypical presentations if other tests are normal and to

confirm adequate acid suppression if poor response to treatmentx New probes can detect bile refluxWeaknessesx Expertise required for interpretation, so not available in all centresx Time consumingx Normal result does not exclude GORD

Box 2.2 General measures for managing gastro-oesophagealreflux diseasex Assess use of drugs—Avoid non-steroidal anti-inflammatory drugs

(associated with benign strictures) and if possible avoid or reducedose of nitrates, calcium channel antagonists, and anticholinergenics

x Avoid or reduce smoking—However, this may be counterproductiveas it often results in weight gain

x Avoid large evening meals and food or drink within 3 hours of bedtime—Have smaller meals spread through day

x Avoid or reduce fats or chocolatex Avoid or reduce any food or drink that is causing a problem—

Problems have been found with citrus fruits, fruit juices, tea andcoffee, peppermint, onions, garlic, cinnamon, cloves

Box 2.3 Complications of reflux oesophagitisx Often present without preceding or other

symptoms of reflux, especially in elderly patientsx Haematemesis or melaenax Iron deficiency anaemiax Dysphagia due to benign stricture, or cancer in

Barrett’s oesophagus

Figure 2.6 Endoscopic view of cancer in aBarrett’s oesophagus


6

completely effective for both relieving symptoms and healingoesophagitis, regardless of disease severity. Because of this, acost effectiveness argument has been made in favour of protonpump inhibitors as first choice treatments in all cases. However,the data on which these calculations have been made havegenerally come from hospital based clinical trials and may notbe applicable to general practice.

Many patients in primary care may achieve good and lastingsymptom relief from short intermittent courses of H2 receptorantagonists at standard doses (such as ranitidine 150 mg twicedaily or cimetidine 400 mg twice daily). For patients with severeor refractory oesophagitis, particularly those with complicationssuch as stricture, proton pump inhibitors are the drugs ofchoice. The optimal daily dose for most patients is omeprazole20 mg , lansoprazole 30 mg, pantoprazole 40 mg, orrabeprazole 20 mg, but higher doses may give additionalclinical benefit in patients with resistant oesophagitis. For mostpatients, there is no clinical advantage in choosing one protonpump inhibitor over another.

Motility modifying drugsThese include metoclopramide and domperidone. Althoughboth relieve symptoms of heartburn to a degree similar to H2

receptor antagonists, they do not heal oesophagitis. In addition,metoclopramide has a relatively high incidence of side effectson the central nervous system. However, these drugs may beuseful, particularly in patients with other dyspeptic symptomssuch as nausea or early satiety.

Maintenance treatmentOnly proton pump inhibitors, at standard or half standarddoses, have been shown to be effective agents for maintenanceof remission in those who require it. Indications formaintenance treatment includex Severe oesophagitis, especially presenting with complications(such as stricture, bleeding, peptic ulcers)x Barrett’s oesophagus (although there is no evidence thatcontinuous treatment prevents evolution to cancer)x Symptoms (typical or atypical) relapsing as soon as treatment isstopped.

SurgeryLaparoscopic anti-reflux surgery seems to be as successful asconventional surgery in controlling reflux in the short termwithout the disadvantages of a long hospital stay orconvalescence. It has become an increasingly popular option forpatients requiring long term medical treatment. The resultsfrom a randomised controlled trial comparing surgery withmaintenance drug treatment are awaited.

Box 2.4 Factors to consider when choosing a proton pumpinhibitorx Efficacy of drugs similar at equivalent doses, but lansoprazole 30 mg

may be equivalent to omeprazole 40 mg dailyx Pantoprazole and lansoprazole have the highest bioavailability and

thus most predictable effectsx Clinically important drug interactions are rarex Pantoprazole has least induction of cytochrome P450 and thus least

potential for interaction with metabolism of other drugs (such aswarfarin, theopyllines, benzodiazepines, phenytoin)

x Side effects rare and minor with all proton pump inhibitors

Box 2.5 Treatment strategy for gastro-oesophageal refluxdiseaseMild GORD(No or mild oesophagitis, symptoms mild or moderate)“Step up” strategyx Start with general measures (antacids and alginates)x If ineffective try H2 receptor antagonist at standard dose for 4-6

weeksx Reserve proton pump inhibitors for those with poor symptom

reliefx Suitable for many patients in general practicex Endoscopic monitoring unnecessaryx “Step down” strategy an alternative—Start with proton pump

inhibitors and reduce to minimal effective treatment

Severe GORD(Severe oesophagitis possibly with bleeding, strictures, or peptic ulcerswithin a Barrett’s oesophagus)“Step down” strategyx Start with proton pump inhibitors for 8 weeksx Reduce to maintenance treatment with half dose if possiblex If patient presented with bleeding or peptic ulcer endoscopic

confirmation of healing is needed. May require high dose of protonpump inhibitor (such as omeprazole 40 mg daily)

Atypical symptoms(Such as respiratory symptoms, ear, nose, and throat symptoms)x Prolonged treatment with proton pump inhibitors, often at high

dose (such as omeprazole 40 mg daily)x Confirmation of GORD desirable by pH monitoringx Confirm adequate acid suppression by pH monitoring if no

response

Box 2.6 Indications for surgical treatmentx Poor response to medical treatment

Failure to suppress acid reflux should be confirmedx Persisting “volume reflux”

Regurgitation of gastric contents without heartburn on maximalmedical treatmentOccurs especially at nightRisk of aspiration

x Difficult benign strictures requiring frequent dilation despitemedical treatment

x Patient’s choice among those requiring long term maintenancetreatment, including younger patients with Barrett’s oesophagus

Further readingx National Institute of Clinical Excellence.Guidance on the use of protonpump inhibitors in the treatment of dyspepsia. London: NICE, 2000

x Dent J, Brun J, Fendrick AM, Fennerty MB, Janssens J, Kahrilas PJ,et al. An evidence-based appraisal of reflux disease management—the Genval workshop report.Gut 1999;44(suppl 2):S1-16

x Lundell L, ed.Guidelines for the management of symptomaticgastro-oesophageal reflux disease. London: Science Press, 1998

Oesophagus: Heartburn

7

3 Oesophagus: Atypical chest pain and motilitydisordersJohn Bennett

Chest pain has been a diagnostic problem for centuries.Heberden’s superb clinical description of angina pectoris (andother chest pains) still stands, but today’s doctors may be nobetter at diagnosing cardiac pain accurately than theirpredecessors 60 years ago. Even when a doctor is sure that thepain is non-cardiac, the more inquiring patient often wants amore specific label.

Much of the interest in this subject has arisen from theextensive investigation of patients concerned that theirsymptoms could be cardiac in origin. Given the associated andinevitable selection bias, there is little objective evidence onwhich to base practice, but the oesophagus is undoubtedly oneof the organs that can generate problematic chest pain.

This article describes the oesophageal disorders responsible,and ways to diagnose them. Psychological factors are oftenimportant in patients with chest pain, so that common sense,understanding a patient and his or her problem, and goodcommunication are usually more important than diagnostictests and powerful drugs.

What does oesophageal pain feel like?Oesophageal pain has many patterns: it is often described asburning, sometimes as gripping, and it can also be pressing,boring, or stabbing. Usually in the anterior chest, it tends to befelt mainly in the throat or epigastrium and sometimes radiatesto the neck, back, or upper arms—all of which may equallyapply to cardiac pain. The commonest patterns of cardiac andoesophageal pains are quite different and well recognised, butperhaps 20% of each are much harder to feel confident about.

Mechanisms of oesophageal painDiscomfort or pain from the oesophagus may arise fromirritant stimuli to the mucosa or from mechanical effects on themuscular wall, each with different sets of receptors.

Mucosal stimulationAtypical chest pain arising from oesophageal mucosal irritationcan be imitated by infusing hydrochloric acid into theoesophagus. This causes discomfort in most people eventually,but, for it to be truly acid induced, it must resolve rapidly whenperfusion ceases (Bernstein test).

When the level of exposure of the lower oesophagus to acidis measured by a pH probe, two things can be seen. Firstly,atypical chest pain is usually associated with a drop in pH, butmany episodes of acid reflux are painless. Secondly, there maybe little relation between the amount of acid reflux and theseverity and frequency of atypical chest pain. This parallels thelongstanding observation that some patients with markedoesophagitis have little or no heartburn.

Other factors must therefore be involved in the thresholdfor experiencing pain, and these are discussed below.

Mechanical changesDistension of the oesophagus, by swallowing a large food bolus

or experimentally by inflating a balloon in its lumen, will causepain in most people. However, individuals with recurrent

Figure 3.1 The close proximity of the heart and oesophagus mean thatdistinguishing oesophageal from cardiac pain is often difficult

28%

37% 60%100%

12%49%35%

14%

10%

Pain radiation Oesophagealpain

Cardiacpain

23%100%

33%18%

29% 33%

5%

Figure 3.2 Frequency of pain experienced in differentanatomical sites by patients with cardiac pain andoesophageal pain

Time (hours)

0

HB = Heartburn S = Smoking

1 2 3 4 50

0

Oeso

phag

eal p

H

02468

10

2468

10

2468

10

HB

S

HB

S

HB

S

HB HB HB HB

S S S

HB HB HB HB

S S S

Figure 3.3 Intraoesophageal pH tracing showing frequent episodes of acidreflux, some of them associated with heartburn

8

spontaneous chest pain experience pain at much lower inflationvolumes, suggesting that they have a “lowered threshold.”Others may have induced distension because they have a highthreshold for belching.

Altered motility of the oesophagus (sometimes loosely referredto as “spasm”) can be a cause of chest pain. The rare conditionof diffuse oesophageal spasm (seen radiologically as a“corkscrew oesophagus”) is associated with pain, and so isachalasia. Powerful, prolonged contractions can be induced byinjection of edrophonium and may cause simultaneous pain,but similar contractions may also be seen in patients withoutchest pain. When pressures in the oesophagus are monitoredcontinuously for 24 hours, a few patients with recurrent chestpain can be seen to have pain episodes associated with variousabnormalities of oesophageal contractions, but this issurprisingly uncommon.

Pain threshold—psychological factorsPatients with other sorts of painful syndromes, such as irritablebowel syndrome or fibromyalgia, also feel pain induced byballoon distension of the oesophagus more readily than peoplewithout pain syndromes.

These observations led to the concept of “altered visceralreceptor sensitivity.” When such individuals are given standardpsychological tests many are found to have greater anxiety,depression, somatisation, neuroticism, or even panic disorderscores than control subjects, and some studies have shownimprovement in pain with the use of antidepressants oranxiolytics.

Moreover, some abnormalities of oesophageal motility—including “non-specific motility disorder” and “nutcrackeroesophagus”—can be induced by the unconscious gulping andhyperventilation performed by some anxious individuals.Unfortunately, well intentioned but misguided medicalinterventions aimed at excluding cardiac disease may worsenthis by raising patients’ anxiety and medical dependency.

Managing possible oesophageal painThis can be a difficult problem, complicated by the fact thatboth oesophageal problems and ischaemic heart disease arerelatively common in middle age and can coexist. Diagnosis ofone does not exclude the other. However, first ensure that thepatient is not taking sumatriptan for migraine; this drug causeschest pain in 3-5% of those taking it.

Atypical chest pain associated with heartburnPatients with atypical chest pain and reflux symptoms need noinvestigation provided their symptoms respond to advice ondiet and weight, stopping use of non-steroidalanti-inflammatory drugs, or to antacids or antisecretory drugs(see previous article). Endoscopy is desirable for anyone withalarm symptoms or a poor response to treatment.

“Pain like angina”If the pain has elements that suggest a cardiac origin—such asan association with exercise, radiation to neck, jaw, or arms, orthe presence of cardiac risk factors (family history, smoking,hypertension, obesity, hypercholesterolaemia, etc)—then theheart must first be investigated with resting and exerciseelectrocardiograms. If these are normal then the often difficultchoice must be made between more invasive cardiac tests (suchas coronary angiography, thallium scanning, etc) oroesophageal tests; specialist referral may be indicated, with thechoice between cardiologist or gastroenterologist determinedby the nature of the pain.

Box 3.1 Primary oesophageal motility disordersAchalasiax Absent distal peristalsisx Abnormal relaxation of lower oesophageal sphincter

Diffuse oesophageal spasmx Simultaneous contractionsx Intermittent peristalsis

Hypertensive (“nutcracker”) oesophagusx Increased contraction amplitude (mean > 180 mm Hg)x Normal peristalsis

Ineffective oesophageal motilityx Contractions of low amplitude or failed and non-transmitted

Sumatriptan causes chestpain in up to 5% ofmigraineurs

Figure 3.4 Barium swallow radiographs showing achalasia (left) and diffuseoesophageal spasm (“corkscrew oesophagus”) (right)

Cardiacopinion

first

>3 risk factors forischaemic heart disease

2 risk factors forischaemic heart disease

1 risk factor forischaemic heart disease

Pain relieved byglyceryl trinitrateand other nitrates

Pain completely abolishedby antacids or proton

pump inhibitors

History of reflux

Pain always occurswith episodes of reflux

Gastro-oesophagealopinion

first

Figure 3.5 Choices for proceeding with patients with chest pain and normalelectrocardiograms

Oesophagus: Atypical chest pain and motility disorders

9

All tests create anxiety in patients and their family, and canlead to medical dependence. This is especially so when tests are“on the heart.” Studies show that, even when oesophageal testsare positive, patients and their family doctors often continue tobelieve that their pain originates from the heart.

Before referring any patient to a specialist, a generalpractitioner should have a clear idea of what may be achievedand explain the risks and benefits to the patient. A period ofsympathetic and supportive observation may often be the bestinitial approach. This will not always be successful, and somepatients can absorb substantial health resources in visits todoctors, use of drugs, and urgent “false alarm” admissions tocoronary care units. Resident junior doctors wishing to consideran oesophageal cause for atypical chest pain should adopt thefollowing approach.

Investigating the oesophagus as a cause of atypical chest painSome patterns of symptoms make it likely that chest pain isoesophageal in origin: an association with meals, dysphagia,relief by antacids, associated acid reflux, and a history ofheartburn.

If reflux is suggested then an empiric approach may beappropriate. For patients with severe pain or dysphagia,endoscopy is the usual initial investigation (to look foroesophagitis, cancer, or other gastric disease), but a bariumswallow (recorded on video by an experienced radiologist wellbriefed about the symptoms) will give more information aboutmotility problems such as diffuse spasm or achalasia.

A “therapeutic trial” with high dose proton pump inhibitors(double healing doses) is an alternative approach with theattraction of simplicity, and if the pain disappears promptly andcompletely it is highly likely that reflux is the cause of the pain.However, it is important to remember that a misleading“placebo response” is common, particularly in anxious patients,and it is always an error to continue such drugs if thesymptomatic response is only partial.

If endoscopy or barium swallow fail to confirm the diagnosis24 hour pH monitoring is necessary to detect any abnormalacid reflux as the most likely treatable cause of “spasm.” If thereis no abnormal reflux, and the pain continues as a significantproblem, further tests are available to detect an oesophagealorigin, although the treatment is far from effective.

Oesophageal manometryStationary—This may show subtle motility abnormalities not

revealed by radiology, but all too often it shows only“non-specific” motility disorders, which are often clinicallyirrelevant.

Ambulatory—This is relatively simple for patients but involvescomplex technology in obtaining the records and analysingthem. It can, however, be useful in patients with intractableproblems, revealing abnormalities such as intermittent spasmand their association with episodes of pain, especially in patientswith intermittent dysphagia.

Oesophageal provocation testsPharmacological—Edrophonium can be injected during

stationary oesophageal manometry to provoke abnormalcontractions and pain that patients can compare with theirspontaneous symptoms. However, abnormal contractions canoccur in normal individuals, and patients may anticipatesymptoms.

Chemical—Acid perfusion of the oesophagus can similarlyreveal recognisable pain.

Mechanical—Balloon distension may show both an increasedsensitivity to pain and that the pain is indeed oesophageal.

Endoscopy or video barium swallow, or both

Possible oesophageal spasm

Oesophageal pH monitoring

Stationary manometry

Provocation tests: Acid perfusion Edrophonium Balloon distension

Ambulatory manometryor pH monitoring

Figure 3.6 Progression of tests for investigating theoesophagus as a cause of chest pain

Figure 3.7 Stationary oesophageal manometry is an outpatient proceduretaking about 30 minutes to perform. A polyvinyl tube about 4 mm indiameter is passed pernasally after topical anaesthesia. The tube consists ofseveral microcapillary tubes connected to external pressure transducers,which pass signals to a computer for real time display. Measurements ofdistal oseophageal pressure are taken during swallows of water while thetube is slowly withdrawn from the stomach

0.9% NaCl0.1N HCl

Figure 3.8 The acid perfusion test for chemicalprovocation of the oesophagus


10

TreatmentThe most important aspect of treating oesophageal pain is asympathetic appraisal of patients’ problems. Are they mainlyworried about their heart or does their pain interfere with life insome other way, or is their family most concerned?

Any reflux element is readily treated by the range of safeantisecretory drugs of different levels of power; “titratingupwards” is the best way to use them. If response seemsinexplicably inadequate, pH monitoring while taking the drugcan help.

Psychological problems need considerate management withboth drugs and psychotherapy as appropriate. Tricyclicantidepressants (such as amitriptyline 25-75 mg at night) maybe particularly useful because of their known value in functionalsomatic pain disorders such as irritable bowel syndrome.

When a true spasm disorder is demonstrated bymanometry, physical treatment may have a place. Achalasia(and sometimes diffuse spasm) may be helped by judiciousballoon dilatation of the gastro-oesophageal sphincter, but thiscarries the risk of perforation or subsequent gastro-oesophagealreflux. There are reports (all uncontrolled) of use of longoesophagomyotomy, but this is an invasive approach for arelatively benign disorder.

Nitrates and calcium channel blocking drugs can be tried—taken sublingually for episodes of pain or orally and regularlyas a prophylactic. Separating useful effects from placebo ones isnot easy, and these drugs have not been proved in randomisedclinical trials. They have two disadvantages, however. One is thatpatients, reassured they do not have angina, may have theirconfidence undermined when they read the datasheet andrealise they are taking antianginal drugs. The other is that anylatent reflux may be accentuated.

Anticholinergic drugs (the traditional “antispasmodics”)have no useful effect, and it is illogical to prescribemetoclopramide or other prokinetic agents.

Box 3.2 Treatments for oesophageal spasmRefluxx Antisecretory drugs

Anxiety or depressionx Explanation and reassurancex Psychotherapyx Anxiolyticsx Tricyclic antidepressants

Achalasiax Balloon dilatationx Cardiomyotomyx Botulinum toxin

Other spasm disordersx Nitritesx Calcium channel blockersx Possibly balloon dilatation

The cross sectional view of a human torso is adapted from the VisibleHuman Project. The list of primary oesophageal motility disorders isadapted from Richter JE. Lancet 2001;358:823-8.

Oesophagus: Atypical chest pain and motility disorders

11

4 DysphagiaWilliam Owen

Dysphagia is a distressing symptom and indicates a delay in thepassage of solids or liquids from the mouth to the stomach. It issometimes difficult to understand why patients may take so longto consult their doctor for advice, but whenever patients presenta diagnosis should be urgently sought. Dysphagia should bedistinguished from odynophagia, which is discomfort or painon swallowing hot or cold liquids and, occasionally, alcohol.

In clinical practice it is useful to separate those causes thatpredominantly affect the pharynx and proximal oesophagus(high dysphagia) and those mostly affecting the oesophagealbody and oesophagogastric junction (low dysphagia).

Management of high dysphagiaThere may be pointers in a patient’s history to a neuromuscularcause of high dysphagia such as neurological disease or atendency for spillage into the trachea when the patient eats,producing coughing or choking. A patient may find it easier toswallow solids or semisolids rather than liquids and may alsocomplain of nasal regurgitation of food. High dysphagia mustbe differentiated from globus hystericus (the feeling of having alump in the throat without any true dysphagia). Globus is acommon symptom, and when the patient is examined noabnormality is usually found. It is believed to be a functionaldisorder but is sometimes associated with gastro-oesophagealreflux.

Generally, radiology is often more rewarding thanendoscopy in clarifying the cause of high dysphagia.Cineradiography with liquid barium and bread soaked inbarium may give valuable functional as well as anatomicalinformation about the pharynx and cricopharyngeal segment.Most problems are related to failure of pharyngeal contractionor to cricopharyngeal relaxation, or a combination of both.

For patients with a pouch or a cricopharyngeal bar, surgicalmyotomy of the overactive cricopharyngeus with, if appropriate,pouch excision is the treatment of choice in most cases.

Other causes of upper dysphagia are more difficult tomanage. In patients recovering from a stroke who need feeding,a fine bore, soft feeding tube can be passed down underradiological guidance. Passage of a large soft oesophagealbougie under light sedation can occasionally alleviate thesymptoms, but if the problem is chronic and disabling then apercutaneous endoscopic gastrostomy—in which a gastrostomytube is passed into the stomach via a percutaneous abdominalroute under the guidance of an endoscopist—can be considered.

Management of low dysphagiaThe main concern with low dysphagia is that a patient may beharbouring a malignancy. The patient’s history may give cluesto this, such as a short duration of symptoms (less than fourmonths), a progressive differential dysphagia affecting solidsmore than liquids, or considerable weight loss. If, however, theproblem has existed for several years, equal difficulty isexperienced with solids and liquids, and there is no weight loss,then achalasia is a more likely cause.

A patient’s history may be misleading, however. Localisationof the point of dysphagia can be poor and patients with anobstructing carcinoma of the cardia occasionally localise thepoint of obstruction to the throat. Patients with a reflux stricture

Box 4.1 Causes of high dysphagia affectingpharynx and cricopharyngeusNeurologicalx Strokex Parkinson’s diseasex Cranial nerve palsy or bulbar palsy (such as

multiple sclerosis, motor neurone disease)

Anatomical or muscularx Myasthenia gravisx Oropharyngeal malignancy (uncommon)x Cricopharyngeal spasmx Pharyngeal pouch

Box 4.2 Causes of low (oesophageal)dysphagiax Carcinoma of oesophagus or oesophagogastric

junction (cardia)x Reflux disease with or without stricturex Motility disturbance of oesophagus (such as

achalasia, scleroderma, or diffuse oesophagealspasm)

Figure 4.1 The use of bougies to remedy dysphagia caused by oesophagealstricture has been a standard treatment for centuries. (Reproduced fromRobert Hooper’s The Anatomist’s Vade-Mecum 1805)

Figure 4.2 Barium swallow showingcricopharyngeal spasm (arrow)

12

may have no history suggestive of gastro-oesophageal reflux(the so called silent refluxers), and Barrett’s oesophagus is oftencharacterised by diminished oesophageal sensitivity and lack ofpain. Patients with achalasia may complain of chest pain andminor dysphagia only, and the condition may sometimes mimicgastro-oesophageal reflux.

Endoscopy is usually the best way to determine the cause ofdysphagia because of its high diagnostic accuracy and theopportunity to take biopsies or to proceed to dilatation, ifappropriate, at the same time.

Procedure after endoscopyEndoscopy will usually reveal a benign peptic stricture, anobvious tumour, or no abnormality.

Management of peptic stricturePeptic stricture is usually due to gastro-oesophageal reflux, butdrugs such as non-steroidal anti-inflammatory drugs, potassiumsupplements, or alendronic acid are occasional causes. Thedifferential diagnosis also includes caustic strictures afteringestion of corrosive chemicals, fungal strictures, andpostoperative strictures.

The diagnosis should be confirmed by biopsy or cytology,and the stricture then dilated. Often only one dilatation isneeded, but re-dilatation over the next few months may beneeded. Proton pump inhibitors reduce the rate ofre-stricturing, and thus the need for re-dilatation, and aretherefore recommended long term.

Occasionally, a patient with a seemingly benign stricturefails to respond to dilatation, and the dysphagia rapidly recurseven after repeated procedures. In such cases malignancyshould be suspected, and this may become obvious only afterrepeat biopsy or brush cytology.

The risk of rupture after oesophageal dilatation is about0.5%. Should a perforation occur the patient is managedconservatively with a strict “Nil by mouth” regimen, intravenousantibiotics, nasogastric drainage, and proton pump inhibitors. Afollow up radiological examination a few days later nearlyalways shows satisfactory healing of the perforation. If thepatient becomes severely shocked after rupture, however,surgery may be needed to treat this life threatening condition.

Barrett’s oesophagusBarrett’s oesophagus (columnar lined oesophagus) arises as aresult of a longstanding acid (and bile) reflux. The oesophagealepithelium undergoes metaplasia so that it changes from theusual pale squamous epithelium to a pink columnar epitheliumthat resembles gastric mucosa.

The risk of developing an adenocarcinoma of theoesophagus is increased about 80-fold with this condition, andcancer is heralded by the development of high grade dysplasiain the oesophagus. In the absence of high grade dysplasia orfrank carcinoma the finding of Barrett’s oesophagus does notpose an enormous risk to the patient. Controversy still exists asto whether regular surveillance is justified because the risk ofdeveloping oesophageal carcinoma is about 1 per 50-200 yearsof patient follow up. The risk of developing carcinoma is greaterif there is a long segment ( > 7 cm) of Barrett’s oesophagus.

Provided there is no significant dysplasia it is appropriate toreassure the patient and manage conservatively in the usualway, probably with endoscopic surveillance every one to twoyears. There is no evidence that long term treatment with aproton pump inhibitor or antireflux surgery either causesregression of Barrett’s oesophagus or abolishes the risk ofdeveloping cancer. Barrett’s oesophagus may also cause

Endoscopy

Tumour

Oesophageal or gastric tumour

Biopsy and brush cytology

Patient fit Patient unfit

Operable Inoperable

Patient fit

Achalasia

Reflux relateddysmotility

Primary motilitydisturbance Reflux treatment

Balloon dilatation Calcium channelblockers (such as

nifedipine)If it fails

MyotomyIf they fail

Diffuse spasmResection Palliation

Stricture

Benign fibrousstricture

Biopsy andbrush cytology

Dilate if possible

Maintenance treatment

with protonpump inhibitor

Normal

Laboratory tests

Ambulatory pH studies andoesophageal manometry and possibly

oesophageal provocation tests

Consider expanding metalstent or

Laser treatment

Spiral computedtomography and endoscopic

ultrasonography

Figure 4.3 Algorithm for endoscopic management of dysphagia

Figure 4.4 Endoscopic appearance of benignoesophageal stricture with associated severe reflux

Figure 4.5 Endoscopic appearance of Schatzki’s ring(constriction of the lower oesophagus)

Dysphagia

13

dysphagia because of an associated benign stricture, whichusually responds to simple dilatation and medical management.

Management of a tumourIf endoscopy reveals a tumour a decision has to be madewhether treatment is potentially curative or merely palliative.

Potentially curative treatmentThe diagnosis should be confirmed by biopsy or brush cytology.It is essential that the patient should undergo rigorous stagingbefore a resection is considered. An upper age limit of 75 yearsseems sensible, although fitness is clearly important and rulesare made only for guidance. Major cardiovascular or respiratoryproblems militate against operating. Spiral computedtomography and endoscopic ultrasonography (if available) givevaluable information about the size of the primary tumour aswell as the presence of any metastases. Palliative resection isusually contraindicated if a liver metastasis is present because ofthe poor results and high mortality.

The operation usually entails an oesophagogastrectomy, asmost tumours now are near the oesophagogastric junction. Thehospital mortality in specialised tertiary centres well versed inthe technique is in the order of 5%, but the five year survival isstill only about 25%. Several centres have reported that use ofpreoperative adjuvant chemotherapy gives a modestimprovement in survival at three years. The current trend is thatall cases of adenocarcinoma of the lower oesophagus that havebreached the oesophageal wall (T3 and T4) are to be givendownstaging chemotherapy for three or six cycles and then bereassessed to determine the operative status. The expectation isthat this will result in an improved long-term survival althoughdefinite data on this are not yet available.

Digestive function after an oesophagectomy is surprisinglygood, and patients should return to a normal diet within a fewweeks of leaving hospital. Patients may occasionally developdysphagia due to an anastomotic stricture; this is amenable toendoscopic dilatation, but tumour recurrence must be excluded.

If the biopsy shows a squamous cell tumour, which is usuallyfound in the middle or upper oesophagus, external beamradiotherapy may be a good alternative to surgery in olderpatients. Intraluminal radiotherapy (brachytherapy) is anattractive alternative and a faster mode of treatment.

Palliative treatmentA tumour may be considered unresectable because of doubtsabout a patient’s general fitness, because computed tomographyor endoscopic ultrasonography shows that it involves adjacentvital structures such as the aorta or trachea, or becausemetastases are present.

The restoration of swallowing is the prime objective in suchpatients, and the recent development of self expanding metalstents, which can expand up to 20 mm in diameter, has radicallychanged clinical practice. The stents are a great advance overthe old tubes and can be inserted either endoscopically orradiologically. Possible complications include stent migrationand growth of the tumour through the mesh. However, stentdesign is evolving rapidly, such as partly covered stents toprevent ingrowth of tumour and improved shape to minimisemigration. Bleeding and perforation are rare complications.

Laser recanalisation of oesophageal cancer also offers aneffective way of improving patients’ ability to swallow. This mayhave to be repeated at regular intervals, but it is particularlysuitable for high tumours, where stents are uncomfortable, andfor protuberant growths at the oesophagogastric junction.Alcohol injection into the tumour in an attempt to shrink it isused in some centres.

Box 4.3 Problems associated withoesophageal stentsx Migrationx Perforationx Tumour overgrowthx Obstruction

x Painx Bleedingx Reflux or aspiration

Figure 4.6 Endoscopic appearance of carcinoma of theoesophagus

Tumour

Oesophagus

Excision Anastomosis

Diaphragm

Stomach

Figure 4.7 Oesophagogastrectomy of an oesophageal tumour, involvingexcision of the distal part of the oesophagus and the proximal stomachfollowed by anastomosis

Figure 4.8 Endoscopic view of an inserted oesophagealstent


14

Dysphagia with normal endoscopic appearanceIf endoscopy shows no obvious abnormality a diagnosis of aprimary motility disturbance of the oesophagus such asachalasia should be considered.

AchalasiaAchalasia may be missed at endoscopy, and radiology is positivein only about 70% of cases. Oesophageal manometry willconfirm the diagnosis of achalasia by showing the absence ofoesophageal peristalsis and a non-relaxing lower oesophagealsphincter.

Pneumatic balloon dilatation of the lower oesophagealsphincter will often relieve the dysphagia, and if dilatations arerepeated (up to three) the success rate can be up to 77%.Patients should be warned about the risk of rupture of theoesophagus ( < 5%), which can be detected on a post-dilatationradiological study. If the patient fails to respond to balloondilatation, then a Heller’s myotomy of the lower oesophagealsphincter is recommended by either laparoscopic or opentechnique. This will improve the swallowing in most patientsalthough there is a small risk (10-15%) of inducingpostoperative gastro-oesophageal reflux.

Other causes of dysphagiaLaboratory investigations may reveal other causes of dysphagiain the presence of a normal endoscopy.

Reflux related dysmotility—These patients should be managedwith the standard treatment for acid reflux.

Diffuse oesophageal spasm—This is a uncommon disorder ofoesophageal motility characterised by repetitive non-peristalticwaves affecting the body of the stomach (see previous article).

Monilial (candidal) oesophagitis can be a rare complication oftreatment with antibiotics or inhaled corticosteroids or mayoccur in immunosuppressed patients.

AIDS related oesophagitis can be caused by cytomegalovirus,herpesvirus, or fungal disease.

ConclusionIn patients presenting with dysphagia, the cause is usually anobstruction. The differential diagnosis lies betweenoesophagogastric malignancy, reflux disease with a stricture, ora motility disturbance. The duration of the problem is probablythe most valuable clinical guide to the diagnosis, and endoscopyis the most rewarding investigation.The image of an inserted stent is reproduced with permission ofJeremy Livingstone.

Time (seconds)

Pres

sure

(mm

Hg)

0

050

1

23

4

56

PharynxUpper

sphincter

Lowersphincter

Oeso

phag

us

050

050

050

050

50

Swallow Swallow

Figure 4.9 Normal recording at oesophageal manometry (horizontal arrowsindicate progression of normal oesophageal peristalsis)

Pres

sure

(mm

Hg)

0

050

050

050

050

050

50Swallow Swallow

Time (seconds)

1

23

4

5

6Pharynx

Uppersphincter

Lowersphincter

Oeso

phag

us

Figure 4.10 Oesophageal manometry in achalasia of the cardia, showinglack of peristalsis and a non-relaxing lower oesophageal sphincter

1

2

3

4

5

6

PharynxUpper

sphincter

Lowersphincter

Oeso

phag

us

Pres

sure

(mm

Hg)

0

050

050

050

050

050

50

Time (seconds)

Swallow Swallow Swallow Swallow

Figure 4.11 Oesophageal manometry in diffuse oesophageal spasm, showingsynchronous multipeaked high pressure waves

Dysphagia

15

5 Epidemiology and diagnosis of Helicobacter pyloriinfectionRobert P H Logan, Marjorie M Walker

Helicobacter pylori is a small, curved, highly motile, Gramnegative bacillus that colonises only the mucus layer of thehuman stomach. Since its discovery in 1984, it has beenrecognised as the principal cause of peptic ulcer disease and asthe main risk factor for the development of gastric cancer.However, most infected people ( > 70%) are asymptomatic. Wetherefore need to discover how infection is acquired, why ulcersor cancer occur in so few of those infected, and how thissubgroup can be identified and treated.

Epidemiology of H pylori infectionH pylori is one of the commonest bacterial pathogens inhumans. The prevalence of infection varies but is falling in mostdeveloped countries. Seropositivity increases with age and lowsocioeconomic status. Retrospective seroepidemiological studieshave shown a cohort effect consistent with the hypothesis thatinfection is mainly acquired in early childhood. Until recently,however, it has been difficult to assess accurately the incidence(or route) of infection because of the inaccuracy and cost ofdetecting (non-invasively) H pylori in young children. Primaryacquisition in adults, or reinfection after successful eradication,does occur but is less common, with an annual incidence of0.3-0.7% in developed countries and 6-14% in developingcountries.

How H pylori is usually acquired and its route oftransmission are unknown. Since humans are the only knownreservoir of infection, it is likely that in developed countriesH pylori is picked up from siblings, other children, or parents,predominantly via the gastro-oral route. In developingcountries faecal-oral transmission may also occur. Various riskfactors are associated with H pylori infection, but the extent towhich these are simply markers of childhood socioeconomicdeprivation is unclear.H pylori infection is an occupationalhazard for gastroenterologists and is associated withperforming endoscopy.

Why is H pylori a chronic infection?Although H pylori initially induces an acute inflammatorygastritis, this immunological response by the host is generallynot sufficient to clear infection, which persists for life. Inaddition, infection with one strain of H pylori does not protectagainst subsequent co-infection with a different strain. Infectionwith multiple strains is quite common and occurs morefrequently in developing countries. Polyclonal infection allowsDNA to be exchanged between different strains, which couldpromote the spread of genes encoding important virulencefactors or resistance to antibiotics.

H pylori is not a new bacteria species and, by virtue of itsurease enzyme and other products, has become extremely welladapted to its unique niche within the gastric mucus. It also hasconsiderable genetic heterogeneity (no two strains are identical),and studies have suggested that this diversity may allow eachstrain to become uniquely adapted to each host to an extentthat, for some subjects, it may be considered as a commensalbacteria.

Figure 5.1 Coloured scanning electron micrograph of H pylori on surfaceof gastric cell

Age (years)

Prev

alen

ce (%

)

00

20

40

60

80

100

10 20 30 40 50 60 70 80

Developing countriesDeveloped countries

Figure 5.2 Prevalence of H pylori infection by age indeveloping and developed countries

Figure 5.3 Circular display of the H pylori genome. Therecent sequencing of the whole genome of two separateH pylori strains will help in developing new treatmentsand emphasises the importance of H pylori as a humanpathogen

16

Peptic ulcer and gastric cancerAbout 15% of infected individuals will develop peptic ulcer(duodenal or gastric) or gastric cancer as a long termconsequence of infection. The outcome of infection dependsmainly on the severity and topography of histological gastritis,which may be determined by the age at which infection isacquired. Infection in infancy is thought to lead to pangastritis,whereas acquisition in later childhood may lead to apredominantly antral gastritis only.

With antral gastritis there is loss of regulatory feedback (butwith an intact and undamaged acid secreting gastric corpus),and the high acid load reaching the duodenum leads to thedevelopment of duodenal gastric metaplasia. The islands ofgastric metaplasia are subsequently colonised by H pylori,leading to duodenitis and a high risk of duodenal ulcer.

In contrast, pangastritis, with an inflamed corpus, isassociated with the loss of acid secreting cells, which leads to anincreased risk of gastric ulcer and gastric cancer—similar to thatseen with autoimmune gastritis in pernicious anaemia.

DiagnosisInvasive testsH pylori can be detected at endoscopy by histology, culture, orurease tests, each with inherent advantages and disadvantages.All these biopsy based methods for detecting H pylori are liableto sampling error because infection is patchy. Up to 14% ofinfected patients do not have antral infection but have H pylorielsewhere in the stomach, especially if they have gastric atrophy,intestinal metaplasia, or bile reflux. In addition, after partiallyeffective eradication treatment, low levels of infection can easilybe missed by endoscopic biopsy, leading to overestimates of theefficacy of eradication treatment and reinfection rates. Protonpump inhibitors affect the pattern of H pylori colonisation of thestomach and compromise the accuracy of antral biopsy.Consensus guidelines therefore recommend that multiplebiopsies are taken from the antrum and corpus for histologyand for one other method (either culture or urease testing).

Histology—Although H pylorimay be recognised on sectionsstained with haematoxylin and eosin alone, supplementary stains(such as Giemsa, Genta, Gimenez, Warthin-Starry silver, Creosylviolet) are needed to detect low levels of infection and to show thecharacteristic morphology of H pylori. An important advantage ofhistology is that, in addition to the historical record provided,sections from biopsies (or even additional sections) can beexamined at any time, and that gastritis, atrophy, or intestinalmetaplasia can also be assessed. Biopsy specimens from otherparts of the stomach can be retained in formalin to be processedonly if antral histology is inconclusive.

Culture—Microbiological isolation is the theoretical goldstandard for identifying any bacterial infection, but culture ofH pylori can be unreliable. Risks of overgrowth orcontamination make it the least sensitive method of detection,and it is the least readily available test for use with endoscopy.Although only a few centres routinely offer microbiologicalisolation of H pylori, the prevalence of multiresistant strainsmakes it increasing likely that culture and antibiotic sensitivitytesting may become a prerequisite for patients with persistentinfection after initial or repeated treatment failure.

Urease tests are quick and simple tests for detecting H pyloriinfection but indicate only the presence or absence of infection.The CLO test and cheaper “home made” urease tests are ofsimilar sensitivity and specificity. However, the sensitivity ofurease tests is often higher than that of other biopsy basedmethods because the entire biopsy specimen is placed in the

Figure 5.4 Histology of gastricmucosa. Top left: normal antralmucosa, with sparse infiltrate oflymphocytes in lamina propria. Topright: active gastritis with neutrophilsinfiltrating epithelium and markedinfiltrate of lymphocytes in laminapropria. Bottom: atrophy of antralmucosa with loss of specialisedglands near muscularis mucosa

Figure 5.5 Culture of H pylori tested for antibioticsensitivity with an E strip that is impregnated with ascale of increasing concentrations of metronidazole

Figure 5.6 Positive and negative results of CLO test forH pylori. The urease of H pylori hydrolyses urea torelease ammonia, which is detected colorimetrically

Epidemiology and diagnosis of Helicobacter pylori infection

17

media, thereby avoiding the additional sampling or processingerror associated with histology or culture. The sensitivity ofbiopsy urease tests seems to be much lower (∼60%) in patientswith upper gastrointestinal bleeding, but this can be improvedby placing multiple biopsy samples into the same test vial.

Non-invasive tests

SerologyH pylori infection elicits a local mucosal and a systemic antibodyresponse. Circulating IgG antibodies to H pylori can be detectedby enzyme linked immunosorbent assay (ELISA) antibody orlatex agglutination tests. These tests are generally simple,reproducible, inexpensive, and can be done on stored samples.They have been used widely in epidemiological studies,including retrospective studies to determine the prevalence orincidence of infection.

Individuals vary considerably in their antibody responses toH pylori antigens, and no single antigen is recognised by serafrom all subjects. The accuracy of serological tests thereforedepends on the antigens used in the test, making it essentialthat H pylori ELISA is locally validated. In elderly people withlifelong infection, underlying atrophic gastritis has beenassociated with false negative results. Consumption ofnon-steroidal anti-inflammatory drugs has also been reportedto affect the accuracy of ELISAs.

Antibody titres fall only slowly after successful eradication,so serology cannot be used to determine H pylori eradication orto measure reinfection rates. Although titres of IgM antibodiesto H pylori fall with age, there are no reliable IgM assays toindicate recent acquisition, which, since this is usuallyasymptomatic, makes it difficult to identify cases of primaryinfection and thus establish possible routes of transmission.

An important advantage of serological methods over othertests for H pylori infection has been the development of simplefinger prick tests that use a fixed, solid phase assay to detect thepresence of H pylori immunoglobulins. These “near patienttests” (NPT) can be performed in primary care and are muchsimpler than the 13C-urea breath test, which is the only othertest for H pylori that is currently used as a NPT. However, theaccuracy of the serological NPT is lower than that reported forstandard ELISA tests using the same antigen preparations.These tests are often used to reassure patients, but to date nostudies have compared the accuracy, cost effectiveness, andreassurance value of the 13C-urea breath test with the serologicalNPT in primary care.

Urea breath testNon-invasive detection of H pylori by the 13C-urea breath test isbased on the principle that a solution of urea labelled withcarbon-13 will be rapidly hydrolysed by the urease enzyme ofH pylori. The resulting CO2 is absorbed across the gastricmucosa and hence, via the systemic circulation, excreted as13CO2 in the expired breath. The 13C-urea breath test detectscurrent infection and is not radioactive. It can be used as ascreening test for H pylori, to assess eradication and to detectinfection in children. The similar but radioactive 14C-urea breathtest cannot be performed in primary care.

Faecal antigen testIn the stool antigen test a simple sandwich ELISA is used todetect the presence of H pylori antigens shed in the faeces.Studies have reported sensitivities and specificities similar tothose of the 13C-urea breath test ( > 90%), and the technique hasthe potential to be developed as a near patient test. The mainadvantage of the test, however, is in large scale epidemiologicalstudies of acquisition of H pylori in children.

Table 5.1 Comparative accuracy, availability, and costs oftests for H pylori infection

Test Sensitivity Specificity Availability CostInvasiveHistology 88-95% 90-95% + + + + ££££Culture 80-90% 95-100% + + £££Urease test 90-95% 90-95% + + + + £-££Non-invasive13C-UBT 90-95% 90-95% + + + + £££14C-UBT 86-95% 86-95% + + + ££Serology:ELISA 80-95% 80-95% + + + £NPT 60-90% 70-85% + + + + ££

Stool antigen 90-95% 90-95% + + ££

UBT= urea breath test. NPT = near patient test.

Further readingx Graham DY. Helicobacter pylori infection in the pathogenesis of

duodenal ulcer disease and gastric cancer: a model.Gastroenterology1997;113:1983-91

x Misiewicz JJ, Harris A. Clinician’s manual on Helicobacter pylori.London: Science Press, 1995

The electron micrograph of H pylori is reproduced with permission ofJuergen Berger (Max-Planck Institute) and Science Photo Library.

1 2

3 4

Figure 5.7 Principle of the faecal antigen test. Polyclonalantibody to H pylori is adsorbed to microwells (1).Diluted patient samples are added to the wells, and anyH pylori in the faecal sample is bound to the adsorbedantibody (2). A second H pylori antibody conjugated toperoxidase is added and binds to H pylori (3). Afterunbound material is washed off, a substrate is addedthat reacts with bound peroxidase enzyme to produce ayellow colour (4), the intensity of which can bemeasured to estimate H pylori levels


18

6 Pathophysiology of duodenal and gastric ulcer andgastric cancerJohn Calam, J H Baron

Duodenal and gastric ulcers and gastric cancer are commonand serious diseases but occur in only a minority of peopleinfected with Helicobacter pylori. Mass eradication of H pylori isimpractical because of the cost and the danger of generatingantibiotic resistance, so we need to know how to targetprophylaxis. Knowledge of the mechanisms that lead to ulcerformation or to gastric cancer in the presence of H pyloriinfection is therefore valuable.

Various factors affect the outcome of H pylori infection,including the host response and particularly the extent andseverity of gastric inflammation and thus the amount of acidsecreted by parietal cells.H pylori can elevate acid secretion inpeople who develop duodenal ulcers, decrease acid throughgastric atrophy in those who develop gastric ulcers or cancer,and leave acid secretion largely unchanged in those who do notdevelop these diseases.

Regulation of gastric acid secretionSeveral specialised cells in the gastric mucosa contribute to thecontrol of acid secretion. G cells in the gastric antrum releasethe hormone gastrin. Gastrin acts on the enterochromaffin-likecells in the gastric corpus to release histamine, which stimulatesparietal cells to secrete acid. Gastrin also stimulates parietal cellsdirectly and promotes growth of enterochromaffin-like andparietal cells.

Histamine H2 receptor antagonists act by blocking the effectof histamine on parietal cells. Proton pump inhibitors act byinhibiting the enzyme in parietal cells that catalyses acidproduction for release into the gastric lumen. G cells,enterochromaffin-like cells, and parietal cells are all regulatedby release of the inhibitory peptide somatostatin fromsomatostatin cells, which are distributed throughout thestomach. The effect of H pylori infection on acid secretiondepends on which part of the stomach is most inflamedbecause this determines which of these cells are affected most.

H pylori related acid secretionHypersecretion in duodenal ulcer diseaseBefore the discovery of H pylori it was known that patients withduodenal ulcers secrete about twice as much acid as controlsbecause they have twice as many parietal cells. Patients withgastric ulcer and those with functional dyspepsia have normalacid output and parietal cell count. Thus there was goodevidence that acid played a major role in ulcer formation.Duodenal ulcers did not occur in achlorhydric people or inthose secreting < 15 mmol/h of acid. Duodenal ulcers can behealed, but not cured, by pharmacological suppression of acidsecretion below this threshold.

Areas of gastric metaplasia in the duodenum can becolonised by H pylori, causing inflammation (duodenitis) andleading to further damage of the mucosa. The extent of gastricmetaplasia is related to the amount of acid entering theduodenum—lowest in patients with pernicious anaemia whosecrete no acid and highest in patients with acid hypersecretiondue to gastrin-secreting tumours (Zollinger-Ellison syndrome).

Box 6.1 Causes of duodenal ulcerx Helicobacter pylori antral gastritisx Non-steroidal anti-inflammatory drugs

Rare causesx Crohn’s diseasex Hypergastrinaemia

IdiopathicGastrinoma

x Hyperparathyroidism

H pylori negative

H pylori positive

Functionaldyspepsia

Duodenalulcers

Gastriculcers

Gastriccancer

MALTlymphoma

Non-steroidalanti-inflammatory

drugs

Figure 6.1 Relation of H pylori infection to upper gastrointestinal conditions

Somatostatin

Acid

Gastrin

Histamine

Food

P

ECL

G

S

_

+

+

++

+

Figure 6.2 Autoregulation of acid secretion. Food stimulates release ofgastrin from antral G cells (G). Gastrin stimulates enterochromaffin-like cells(ECL) to release histamine, which stimulates parietal cells (P) in the gastriccorpus to secrete acid. Acid stimulates release of somatostatin fromsomatostatin cells (S) in the antrum, inhibiting further gastrin release

19

Acid hypersecretion in duodenal ulcer disease is virtually alwaysdue to H pylori infection because secretion returns to normalafter the infection is eradicated. The predominantly antralgastritis in duodenal ulcer disease leads to acid hypersecretionby suppressing somatostatin cells and increasing gastrin releasefrom the G cells in the gastric antrum.

Hyposecretion in patients at risk of gastric cancerH pylori infection predisposes to distal gastric cancer, butpatients who develop this complication have diminished acidsecretion. Low acid secretion in gastric cancer was, untilrecently, thought to be predominantly due to gastric corpusgastritis, the associated gastric atrophy leading to loss of parietalcells. However,H pylori associated acid hyposecretion can inpart be reversed by eradicating H pylori, suggesting thathyposecretion is due to inflammation rather than to permanentloss of cells.

H pylori associated acid hyposecretion may also be due toincomplete recovery from the loss of acid secretion that occurswith acute infection or may be in part genetically determinedbecause it is more common in the first degree relatives ofpatients with gastric cancer.

Low acid secretion predisposes to gastric cancer by severalmechanisms, including impaired absorption of vitamin C andovergrowth of salivary and intestinal bacteria within thestomach. By contrast, proximal gastric cancer (at thegastro-oesophageal junction) is associated with normal acidoutput. The rising incidence of this type of gastric cancer mayreflect the decreasing prevalence of H pylori infection inWestern societies.

Relation between distribution of gastritis and acid secretionThus distribution of H pylori gastritis determines acid secretionand the clinical outcome of H pylori infection, be that duodenalulcer, gastric ulcer, gastric cancer, or asymptomatic infection.Positive feedback may perpetuate the different patterns ofgastritis; for example, suppression of acid with a proton pumpinhibitor diminishes antral gastritis but allows H pylori tocolonise the corpus, which then becomes inflamed. This showsthat acid secretion normally protects the corpus from H pyloriinfection. This effect has several important consequences:x High acid secretion in people with duodenal ulcers may be selfperpetuating because it restricts gastritis to the antrum, leaving ahealthy corpus to continue secreting acidx Low acid secretion may be self perpetuating because itincreases corpus gastritis, which further depresses acid secretionx Proton pump inhibitors may be more effective in patients withH pylori infection than in those without because they promotecorpus gastritis, which further inhibits acid secretion.

Aspects of the environment, bacterium, or host that affectacid output or the severity of corpus gastritis might steer aperson infected with H pylori to a state of high acid secretion(predominantly antral gastritis) or to low acid secretion(predominantly corpus gastritis). This model is attractivebecause it allows studies of gastric physiology to be integratedwith other equally important determinants of disease outcome.

Other factors that might affect gastricphysiology and disease outcomeThe pathogenic importance of H pylori depends on theinteraction between bacterial virulence, the host, and theenvironment.

Primary infectionof H pylori

H pylori

Later childhood

Inflammation

Gastric cancer

UlcerationAtrophicgastritis

Intestinalmetaplasia

Duodenalgastric

metaplasia

Inflammation

H pyloriinfection Inflammation

Acid secretionincreased then decreased

Increasedacidsecretion

Increased gastrinproduction

Increased gastrinproductionDecreased somatostatin

production

Most peopleasymptomatic

Infancy

Dietary factors:high salt

low vitaminC and E

Acid hypersecretionAcid hyposecretion

Inflammation

Decreased somatostatinproduction

H pylori Acid

Excessacid

Figure 6.3 With acid hyposecretion (left), the main effect of H pylori gastritisaffecting the gastric body is to suppress parietal cells, leading to low acidsecretion, which is associated with gastric cancer. With acid hypersecretion(right), antral H pylori gastritis increases acid secretion by suppressingsomatostatin and elevating gastrin release, increasing the risk of duodenalulceration. Orange areas indicate extent and location of gastritis

Figure 6.4 Intestinal metaplasia ofantral mucosa. Inset shows largegoblet cells packed with mucin(shown by Alcian blue staining)

Figure 6.5 Duodenum with gastricmetaplasia and mild inflammation.Inset shows H pylori adhering tosurface epithelial cells


20

HostStudies of identical and non-identical twins have shown thathost factors are important in determining the outcome ofinfection. Duodenal ulcer is twice as common in those who areblood group O non-secretors. Studies in the mouse model ofHelicobacter infection have shown that different strains of micedeveloped either severe gastritis or hardly any gastritis wheninfected with the same strain of Helicobacter. The genesresponsible for the different outcomes are not known, butpreliminary evidence suggests that a variety of genes involved inthe inflammatory response affect the likelihood of H pyloriinfection progressing to duodenal ulcer disease.

BacteriumIn contrast, some investigators believe that H pylori is mainlyresponsible for disease because gastric mucosal inflammation isalways present and fully resolves only when infection issuccessfully treated. Most strains of H pylori can be divided intotwo distinct phenotypes based on the presence or absence of avacuolating toxin (Vac A toxin) and the products of the cagpathogenicity island (cagPI), a large chromosomal region thatencodes virulence genes and is similar to that found in otherenteric pathogens such as Escherichia coli and Salmonella typhi.People infected with strains of H pylori with the cagPI havemore severe mucosal damage and are more likely to haveduodenal ulcers or gastric cancer.

However, research has not so far identified H pylori genesthat predispose to either duodenal ulcer or gastric cancer.Furthermore, in developing countries, where H pylori infectsmost of the population, cagPI strains of H pylori are present inalmost all infected people but only a few develop clinicaldisease.

Identifying these and other bacterial virulence factorsassociated with more severe disease may allow screening tests tobe developed. These may then permit identification of patientsinfected with “bad” bacteria so that eradication treatment can betargeted to them.

EnvironmentGastric cancer is epidemiologically linked with diets high in saltand low in fresh fruit. Salt may change acid secretion because itsuppresses parietal cells, and salty diets cause gastric atrophy.Conversely, the antioxidant vitamins in fresh fruit might protectspecialised gastric cells from the reactive oxygen speciesreleased by inflammatory cells. A diet high in salt and lackingantioxidant vitamins might thus promote low acid secretionwith corpus gastritis. Cigarette smoking strongly predisposes toboth duodenal ulcer and gastric cancer.

Time and geographical trendsThe factors described above might explain some geographicaldifferences and changes with time in the prevalence of thedifferent upper gastrointestinal diseases. For example, a highprevalence of H pylori infection plus a traditional salty dietmight explain the high prevalence of gastric cancer in Japanand China. Rates of acid secretion have risen recently in Japan,perhaps because of a fall in the prevalence of H pylori or someWesternisation of the diet. In the United Kingdom thereplacement of salt with refrigeration to preserve food mighthave accounted for the rise in the prevalence of duodenal ulcerdisease in the middle of this century, as the gastric corpusbecame healthier and acid secretion higher.

Box 6.2 Summary pointsx Both duodenal ulcer and gastric ulcer are

essentially gastric ulcersx They occur in gastric mucosa in the stomach or

in gastric metaplasia mucosa in the duodenumx The mucosa may be attacked by

Secretagogues (excess gastrin, histamine, orcalcium producing excess of acid)Bacteria (H pylori)Drugs (non-steroidal anti-inflammatory drugs)

Further readingx Calam J. Clinicians guide to Helicobacter pylori. London: Chapman

and Hall, 1996x El-Omar EM, Penman ID, Ardill JES, Chittajallu RS, Howie C,

McColl KEL. Helicobacter pylori infection and abnormalities ofacid secretion in patients with duodenal ulcer disease.Gastroenterology 1995;109:681-91

x Harris AW, Grummett PA, Misiewicz JJ, Baron JH. Eradication ofHelicobacter pylori in patients with duodenal ulcers lowers basaland peak acid outputs in response to gastrin releasing peptide andpentagastrin.Gut 1996;38:663-7

x Logan RPH, Walker MM, Misiewicz JJ, Gummett PA, Karim QN,Baron JH. Changes in the intragastric distribution of Helicobacterpylori during treatment with omeprazole.Gut 1995;36:12-6

x Saponin P, Hyvarinen H, Psoralea M. H pylori corpus gastritis—relation to acid output. J Physiol Pharmacol 1996;47:151-9

Figure 6.6 Transmission electron micrograph of duodenal gastric metaplasiawith H pylori attached to the apical surface (arrows)

Pathophysiology of duodenal and gastric ulcer and gastric cancer

21

7 Management of Helicobacter pylori infectionAdam Harris, J J Misiewicz

This article discusses the current management of Helicobacterpylori infection in patients with dyspepsia with or withoutendoscopic abnormalities. We take an evidence based approachwhen possible and consider recent guidelines from national andinternational bodies pertaining to primary and secondary care.

Duodenal ulcer diseaseIn patients who are not taking non-steroidal anti-inflammatorydrugs (NSAIDs) duodenal ulcer will be due to H pylori infectionin 95% of cases, and eradication treatment can be prescribedwithout testing for H pylori. If there is any doubt about thediagnosis, such as a possible ulcer crater on a barium meal,endoscopic confirmation of duodenal ulcer and H pyloriinfection should be sought before prescribing treatment.

H pylori eradication treatment, if successful, will be effectivein curing the ulcer diathesis regardless of whether a patient isseen at the initial presentation of the disease or at a recurrence.Patients taking long term (maintenance) treatment with H2

receptor antagonists or proton pump inhibitors should also beoffered H pylori eradication treatment regardless of whetherthey are free of symptoms or still experiencing indigestion. Inmost cases eradication of H pylori cures the duodenal ulcerdisease, and maintenance treatment can be stopped.

After eradication treatmentUncomplicated duodenal ulcers heal quickly and completelyafter eradication of H pylori. Further antisecretory treatment,repeat endoscopy, or formal assessment of eradication is notnecessary, and one can await the clinical outcome.

Recurrent symptoms indicate either eradication failure orthe presence of some other disease. Subsequent managementwill not be clear unless the outcome of eradication treatment isknown, and this is best assessed by a 13C-urea breath testperformed more than four weeks after the antimicrobialtreatment. Recurrent symptoms after documented H pylorieradication are often due to gastro-oesophageal reflux disease,the symptoms of which may be misattributed to duodenal ulcer.

Complicated duodenal ulcerComplicated duodenal ulcers (such as bleeding or perforated)are associated with appreciable morbidity and mortality,especially in elderly people. Therefore, in patients withcomplicated duodenal ulcers, eradication of H pylori andcomplete epithelialisation of the ulcer crater need to beconfirmed by the 13C-urea breath test and endoscopy, afterwhich maintenance antisecretory treatment can be stopped.The prevalence of H pylori infection in patients withcomplicated duodenal ulcer may be lower than in those withsimple duodenal ulcer, and H pylori status should therefore beassessed before prescribing eradication treatment.

Duodenal ulcers recur in about 5% of patients initiallyinfected with H pylori even after eradication and in the absenceof reinfection or use of NSAIDs. Duodenal ulcers are also foundoccasionally in people not infected with H pylori. After exclusionof surreptitious use of ulcerogenic drugs and the rarer causes ofduodenal ulcer, such patients need long term maintenancetreatment with antisecretory drugs.

Box 7.1 Causes of duodenal ulcerCommon causesx H pylori infectionx Non-steroidal anti-inflammatory drugs

Rare causesx Zollinger-Ellison syndromex Hypercalcaemiax Granulomatous diseases (Crohn’s disease, sarcoidosis)x Neoplasia (carcinoma, lymphoma, leiomyoma, leiomyosarcoma)x Infections (tuberculosis, syphilis, herpes simplex, cytomegalovirus)x Ectopic pancreatic tissue

pH

2

Gastricjuice

Adherentlayer ofmucus

H pylori

Epithelium

Gastricgland

Blood

5

6.8

7.4

Figure 7.1 Microanatomy of gastric mucosa indicating the pH gradient

Known duodenal ulcer or ulcer found at endoscopy or barium meal

1 week of triple therapy for H pylori eradication

Second lineeradication treatment

Urea breath test >4 weeksafter treatment

Urea breath test >4 weeks after treatment or re-investigate if symptoms recur

Assume infection with H pylori

H pylori negative

H pylori negative

H pylori positive

H pylori positive

No furthertreatment

Maintenanceantisecretory

treatment

Patient stillsymptomatic

Patientasymptomatic

No furthertreatment

Consider: False negative result Gastro-oesophageal reflux disease Functional dyspepsia H pylori negative duodenal ulcer Gall stones

Figure 7.2 Management plan for uncomplicated duodenal ulcer in patientsnot taking NSAIDs

22

Gastric ulcerDiagnosisThe main difference in the management of gastric ulcers fromthat of duodenal ulcers is the need to exclude malignancy in anapparently benign gastric ulcer. Endoscopy is mandatory, withtargeted biopsies of the ulcer rim and base. About eight weeksafter treatment is started, endoscopy should be repeated toconfirm healing, obtain further biopsies from the original ulcersite, and if clinically indicated ascertain H pylori infection status.

TreatmentAs with duodenal ulcer, eradication of H pylori leads to healingof gastric ulcer and markedly decreases the incidence of relapse.Eradication of H pylori also seems to reduce the complicationsassociated with gastric ulcer, but the supporting evidence is lessstrong than for duodenal ulcer. Maintenance treatment withantisecretory drugs should therefore be started after successfuleradication of H pylori in those patients with gastric ulcer whohave a history of haemorrhage or perforation until completehealing of the ulcer is confirmed at endoscopy.

Ulcers associated with H pylori and NSAIDsMost gastric ulcers associated with H pylori infection or with useof NSAIDs occur in elderly women. Despite several studies, noclearly defined guidelines have emerged. NSAIDs and H pyloriseem to be independent risk factors for increased risk ofgastrointestinal bleeding. If a patient infected with H pylori hasulceration then H pylori should be eradicated before treatmentwith NSAIDs is started. There is no evidence that H pylorieradication relieves NSAID induced dyspepsia.

Gastro-oesophageal reflux diseaseThe interaction between H pylori, gastro-oesophageal refluxdisease (GORD), and treatment with antisecretory drugs isextremely complex and highly contentious. Epidemiologicalstudies have shown that the prevalence of H pylori infection isno higher in patients with GORD than in healthy controlsmatched for age and sex. Indeed,H pylori infection may be lesscommon in patients with GORD, particularly those with moresevere (erosive) disease, suggesting that the bacterium may havea protective role, perhaps by producing corpus gastritis andthus decreasing the output of acid. Moreover, proton pumpinhibitors used to treat GORD seem to be more effective atsuppressing acid and healing oesophagitis in the presence ofH pylori. After eradication of the bacterium, patients withGORD may require higher doses and longer duration of protonpump inhibitor treatment.

However, patients with GORD and H pylori infection whoneed prolonged treatment with standard doses of proton pumpinhibitors may be at increased risk of developing atrophicgastritis. It is well recognised that chronic atrophic pangastritisis associated with increased risk of proximal gastricadenocarcinoma. During profound acid suppression withproton pump inhibitors,H pylori infection spreads from theantrum to the gastric body and fundus and causes a chronicactive pangastritis that, with time, may progress to atrophicgastritis. The actual lifetime risk of subsequent gastric cancer isunknown and needs to be evaluated against the potentiallydetrimental effects of eradicating H pylori infection in patientswith GORD. Further studies are needed before thesecontradictory considerations can be resolved.

Box 7.2 Causes of gastric ulcerx H pylori infectionx Non-steroidal anti-inflammatory drugsx Neoplasia (carcinoma, lymphoma, leiomyosarcoma)x Stressx Crohn’s diseasex Infections (herpes simplex, cytomegalovirus)

Gastric ulcer found after barium meal

1 week of triple therapy for H pylori eradication

Endoscopy: Multiple targeted biopsies to exclude malignancy and test for H pylori

Benign and H pylori positive

Urea breath test >4 weeks after treatment

Benign and H pylori negative

8 weeks of antisecretory treatment

Repeat endoscopy

Repeat endoscopy >4 weeks after treatment

Repeat biopsies and 4 weeks of antisecretory treatment

Ulcer healed andH pylori positive

Ulcer healed Ulcer not healedSecond line

eradication treatment

H pylori positive

No further treatment

Ulcer healed Ulcer not healed

Repeat biopsiesNo further treatment

Maintenanceantisecretory

treatment

Ulcer healed andH pylori negative



H pylori negative

Figure 7.3 Management plan for gastric ulcer

Figure 7.4 Benign gastric ulcer (arrow) in upper part ofstomach

Proton pump inhibitors + H pylori infection increases atrophic gastritis Proved?

Barrett's oesophagus and GORD are risk factors for oesophageal cancer Important?

Important?

H pylori eradication reduces efficacy of proton pump inhibitors and H2 receptor antagonists

H pylori pangastritis may protect against Barrett's oesophagus and GORD

True?H pylori may cause reflux

Relevant?

Figure 7.5 Interactions between H pylori, GORD, and antisecretory drugs

Management of Helicobacter pylori infection

23

Functional dyspepsiaIn the absence of NSAID treatment, about 60% of youngpatients ( < 45 years old) with dyspepsia have functionaldyspepsia, about 25% have GORD, and 15% have peptic ulcerdisease. Although the evidence unequivocally supports H pylorieradication in peptic ulcer disease, the role of H pylori infunctional dyspepsia and the evidence to support its treatmentare much less clear.

Asymptomatic H pylori infectionThis presentation is becoming more common because of theincreasing use of commercial, non-invasive tests for H pylori. Apositive test often causes concern about the risk of developingstomach cancer; but we don’t know when H pylori has to beeradicated to prevent the progression to cancer, and there is noevidence yet that eradication of H pylori decreases this risk.

H pylori eradication treatmentThe aim of treating H pylori is to eradicate the organism from thestomach. Eradication is defined as negative tests for the bacteriumfour weeks or longer after treatment has finished. Prematureassessments may give false negative results because of temporaryclearance or suppression of H pylori. The best test to confirmeradication is the 13C-urea breath test. The recently describedstool antigen test may be an alternative in future. “Near patienttests” or laboratory based blood serology tests are not suitablebecause antibody titres take at least six months to decrease.

Treatment of H pylori is difficult because of the rapiddevelopment of resistance to antibacterial drugs, especially tonitroimidazoles, which occurs more commonly in women andpatients from developing countries because of previoustreatment for gynaecological infections or infective diarrhoeas.Resistance to clarithromycin may occur after failed treatment orafter use of this drug for other indications such as respiratorytract infections. Resistance to antibiotics other thannitroimidazoles can also develop but is less common

Low dose triple therapy—The most overall effective H pylorieradication regimens reported to date combine a proton pumpinhibitor with two of the following—amoxicillin, clarithromycin,or a nitroimidazole—for a week. There are few side effects (thecommonest being nausea, diarrhoea, and taste disturbance).Results from large randomised controlled trials have shownH pylori eradication in about 90% of patients.

Ranitidine-bismuth-citrate has been developed specifically fortreating H pylori infection. It retains both the antisecretory andantibacterial properties of the parent compounds but achievesacceptable eradication rates only when used as an alternative to aproton pump inhibitor in combination with clarithromycin andeither metronidazole or amoxicillin for a week.

Quadruple therapy—Classic bismuth based triple therapy ismore effective when coprescribed with a proton pump inhibitor(80-90% H pylori eradication). Efficacy is highly dependent oncompliance with the complicated regimen, and there arenumerous side effects. It is best reserved for use by hospitalspecialists to treat patients in whom triple therapy has failed.

What to tell patients?There has been much discussion of H pylori in the media, andmany patients are aware of its ulcerogenic and carcinogenicpotential and may request antibacterial treatment if they arefound to be infected. Eradication treatment is of proved benefit

Box 7.3 Risk factors for nitroimidazole resistance in H pylorix Previous use of nitroimidazoles, such as for gynaecological

infections, infective diarrhoeasx Failed eradication of H pylori with treatment regimen containing a

nitroimidazolex Urban or inner city areasx Patients born in developing countries

Table 7.1 Low dose triple therapy for H pylori eradication

Treatment

Proton pump inhibitortwice daily

Proton pump inhibitortwice daily

Amoxicillin1 g twice daily

Clarithromycin250 mg twice daily

Clarithromycin500 mg twice daily

Metronidazole400 mg twice daily

Duration 1 weekSide effects Nausea, diarrhoea, taste disturbancesEradication 90% 90% in MSS

75% in MRS

MSS =metronidazole sensitive strain of H pylori. MRS =metronidazole resistantstrain of H pylori

Table 7.2 Quadruple therapy for H pylori eradication

Treatment

Proton pump inhibitoronce or twice daily

Colloidal bismuth citrate120 mg four times daily

Tetracycline500 mg four times daily

Metronidazole400 mg four times daily

Duration 1 weekSide effects Commonly nausea, diarrhoea, taste disturbancesEradication > 75% in MRS

> 90% in MSS

MRS=metronidazole resistant strain of H pylori. MSS =metronidazole sensitivestrain of H pylori

PatientSmoking

Poor compliancewith treatment

TreatmentSide effects

H pyloriAntimicrobial

resistance

Figure 7.6 Possible reasons for failure of H pylorieradication


24

only in patients with duodenal or gastric ulcer associated withH pylori infection. At present there is no evidence to suggestthat screening and treating patients without risk factors willprevent gastric cancer. The risk of transmission to partners islow in adults, and treatment of the entire family is notwarranted.

Counselling patientsWhatever treatment is chosen, patients need careful counselling.The reasons for embarking on the treatment and theimportance of compliance despite possible side effects need tobe emphasised, and the possible side effects must be carefullydiscussed. The need for good compliance needs specialattention, as it is crucial to the success of treatment.

First line treatment—In areas with a low prevalence ( < 30%)of metronidazole resistant strains of H pylori one week of lowdose triple therapy consisting of a proton pump inhibitor,metronidazole, and clarithromycin is currently recommended.Patients’ compliance with treatment is likely to be good becauseof twice daily dosing and few side effects. If metronidazoleresistance is likely a proton pump inhibitor in combination withamoxicillin and clarithromycin given for one week is preferable.

Second line treatment—After a proved failure with a treatmentcontaining metronidazole, a patient is likely to be colonised by aresistant strain of H pylori. In this case a proton pump inhibitorshould be given in combination with amoxicillin andclarithromycin for a week, with about 90% success. If H pylorieradication is unsuccessful after a treatment containingclarithromycin and the patient is likely to harbour ametronidazole resistant strain of H pylori, then eitheromeprazole in combination with amoxicillin and metronidazoleor quadruple therapy are the only logical options, with roughly75% success.

SummaryDespite a vast amount of research, the only evidence basedindications for eradication of H pylori are for patients withduodenal ulcer or gastric ulcer who are not taking NSAIDs andfor patients with the extremely rare MALT lymphoma. Lowdose triple therapy given for one week will cure most patients oftheir infection: failures are due to bacterial resistance or poorcompliance. The importance of H pylori in NSAID associatedulceration is uncertain. Although H pylori is strongly associatedwith gastric cancer, there is no proof that eradication treatmentdecreases an individual’s risk of that disease.

Table 7.3 Indications for H pylori eradication treatment

DiagnosisEvidence basedindication

Duodenal ulcers not due to NSAIDs YesGastric ulcers not due to NSAIDs YesDuodenal or gastric ulcers due to NSAIDs NoFunctional dyspepsia Unknown or noGastro-oesophageal reflux disease Unknown or noGastric cancer Unknown or noMALT lymphoma Yes

NSAID = non-steroidal anti-inflammatory drug

The endoscopic image of benign gastric ulcer is reproduced with permission ofGastrolab Image Gallery.

H pylori eradication successful?

Is patient likely to have a metronidazole resistant strain of H pylori

No

Proton pump inhibitor twice dailyClarithromycin 250 mg twice dailyMetronidazole 400 mg twice daily

Yes

Proton pump inhibitor twice dailyClarithromycin 500 mg twice daily

Amoxicillin 1 g twice daily

Omeprazole 400 mg once dailyAmoxicillin 500 mg thrice daily

Metronidazole 400 mg thrice dailyor Quadruple therapy

Yes No


Figure 7.7 Choosing a treatment regimen for H pylori eradication

Management of Helicobacter pylori infection

25

8 Indigestion and non-steroidal anti-inflammatorydrugsJ M Seager, C J Hawkey

Non-steroidal anti-inflammatory drugs (NSAIDs) are usuallythought to pose a dilemma for doctors wishing to prescribethem. Their anti-inflammatory and analgesic properties haveled to their widespread use for rheumatoid and (much morecommonly) other conditions often regarded as more trivial.However they are ulcerogenic to the stomach and duodenumand lead to a threefold to 10-fold increase in ulcercomplications, hospitalisation, and death from ulcer disease.1

In fact, the dilemma is more complex than whetherpotentially life threatening drugs should be used to manageconditions that are uncomfortable but not in themselves lifethreatening. There is growing evidence that NSAIDs have otherincidental benefits. The only study to investigate overall lifeexpectancy with drug use found non-significant trends towardsenhanced rather than reduced life expectancy. Aspirin hasbenefits in preventing cardiovascular disease and probablycancer that seem to far outweigh the hazards of gastrointestinalulceration. Limited evidence suggests that these benefits may beshared by other NSAIDs.

NSAID useAbout 24 million prescriptions a year are written for NSAIDs inthe United Kingdom. Half of these are given to patients overthe age of 60. At any one time about 15% of elderly people aretaking an NSAID. Average prescribing rates are calculated to be426 scripts per 1000 population per year.

Less than 10% of NSAIDs used in the community seem tobe for rheumatoid arthritis, and less than half for any form ofarthritis. They are widely used for acute soft tissue injury andmore chronically for undiagnosed pains in the back andelsewhere. Chronic use is more common in elderly thanyounger patients, and prophylactic use of aspirin, mostly inrelatively low doses, for cardiovascular events is increasing,amounting to 9.5% of a relatively elderly population.

Gastrointestinal toxicity of NSAIDsNSAIDs are important in causing both (non-ulcer) dyspepsiaand ulcers (often silent and presenting with a complication).The unreliability of dyspepsia as a pointer to ulcerationunderlies many of the problems of managing patients takingNSAIDs.

Within 90 minutes of taking 300 mg or 600 mg of aspirin,nearly everyone develops acute injury consisting ofintramucosal petechiae and erosions. Non-aspirin NSAIDscause less florid acute injury, but endoscopic studies show thatabout 20% of those taking non-aspirin NSAIDs or aspirin atanti-inflammatory doses chronically have a gastric or duodenalulcer. Many patients who start NSAIDs will not be able tocontinue because of drug associated dyspepsia.

Ulcers probably form and heal spontaneously in mostNSAID users and usually cause little harm. However, aboutonce in every 50-100 patient years, ulcer bleeding orperforation develops that requires hospitalisation.2 As aconsequence, probably at least 1200 patients die each year inthe United Kingdom.1

Figure 8.1 Three ulcers (one bleeding) in the gastricantrum caused by NSAIDs. Such ulcer complications areestimated to cause up to 16 500 deaths each year in theUnited States and 2000 deaths a year in Britain

Age (years)

No o

f hos

pita

lisat

ions

/100

0 pe

rson

yea

rs

15 25 35 45 55 65 75 >850

5

10

15

20

25Male NSAID usersFemale NSAID usersMale non-usersFemale non-users

Figure 8.2 Hospitalisations due to complicationsassociated with NSAID use are a problem in elderlypatients

Figure 8.3 In 1938 Douthwaite and Lintott provided the first endoscopicevidence that aspirin caused gastric mucosal damage. Images show gastricantrum before (left) and after (right) administration of aspirin (reproducedfrom Douthwaite AH, Lintott JAM. Lancet 1938;ii:1222-5)

26

Who is at particular risk?Risk factors for gastroduodenal ulcer complications are nowfairly well defined. Most patients with NSAID associated ulcercomplications are elderly. This is because elderly people have ahigher background prevalence of ulcer problems, are morelikely to receive NSAIDs, and are probably more sensitive tothem. A history of ulcers (whether or not associated withNSAIDs) is a further risk factor.

A meta-analysis of recent studies shows that ibuprofen(<1200 mg/day) is associated with a lower level of risk thanother NSAIDs, whereas others such as azapropazone andpiroxicam are associated with a higher risk.3 These differencesprobably relate at least in part to effective dose, since doses ofibuprofen greater than 1200 mg carry risks similar to those withother NSAIDs.3 Among NSAIDs in general, risks rise steadilywith dose.3

As would be expected, the risk of ulcer bleeding is muchhigher if patients taking NSAIDs also receive warfarin.Interestingly, use of corticosteroids has been shown fairlyconsistently to magnify the risk of ulcer complications, to theextent that ulcers associated with NSAIDs and steroid mayaccount for the probably incorrect belief that corticosteroidswere themselves ulcerogenic.

Managing patients taking NSAIDsManagement of patients who need to take NSAIDs should bebased more on assessment of risk than on clinical, laboratory, orendoscopic investigation. NSAID use should be avoided inpatients with two or three of the risk factors for ulcercomplications, or if this is not possible they should receiveprophylactic treatment.

Although development of dyspepsia soon after starting anNSAID may preclude its use and can sometimes lead todiscovery of a previously silent ulcer, NSAID associateddyspepsia is generally a poor guide to the presence of an ulcer.Development of anaemia or new onset dyspepsia can identifyNSAID associated ulcers, but upper gastrointestinalinvestigation is often negative and reliance on these signs willmiss most ulcers, which most commonly present withcomplications on a relatively silent background.

Available drugs to treat ulcersPreclinical studies in animals and humans suggest that twocomponents contribute to development of NSAID associatedulcers. Firstly, inhibition of prostaglandin synthesis, byimpairing mucosal defences, leads to erosive breach of theepithelial barrier. Secondly, acid peptic attack deepens this intofrank ulceration, and pH is also an important determinant ofpassive NSAID absorption and trapping in the mucosa.Preventive treatment aimed at either mucosal defence or acidattack is available.

Misoprostol is a stable analogue of prostaglandin E1. Severalstudies show that it prevents acute gastric injury by a widevariety of agents including NSAIDs. In doses of 400-800 �gdaily misoprostol prevents gastric (and probably duodenal)ulcers. A large study has shown it to reduce the incidence ofhospitalisation for NSAID associated gastrointestinalcomplications.2 Unfortunately, as a prostaglandin, it causesdiarrhoea, abdominal cramps, and reflux at the doses necessaryto protect against NSAID associated ulcers.

Acid suppressing drugs—Normal doses of H2 antagonists haverelatively little affect on acute aspirin and NSAID associatedinjury in animals and in humans. High doses of H2 antagonists

Box 8.1 Risk factors for gastrointestinal complicationsoccurring with NSAIDsPatient related factorsx Age > 60 yearsx History of ulcer disease

Drug related factorsx Use of relatively toxic NSAIDx High dose of NSAID (or two NSAIDs used concurrently)x Concurrent use of anticoagulantx Concurrent use of corticosteroid

Uncertain or possible risk factorsx Duration of NSAID treatmentx Femalex Underlying rheumatic diseasex Cardiovascular diseasex Helicobacter pylori infectionx Smokingx Alcohol consumption

Table 8.1 Comparative toxicity of NSAIDs forgastrointestinal complications*

DrugNo ofstudies

Pooled relativerisk (95% CI)†

P value(heterogenicity)

Ibuprofen NA 1.0 NAFenoprofen 2 1.6 (1.0 to 2.5) 0.310Aspirin 6 1.6 (1.3 to 2.0) 0.685Diclofenac 8 1.8 (1.4 to 2.3) 0.778Sulindac 5 2.1 (1.6 to 2.7) 0.685Diflunisal 2 2.2 (1.2 to 4.1) 0.351Naproxen 10 2.2 (1.7 to 2.9) 0.131Indometacin 11 2.4 (1.9 to 3.1) 0.488Tolmetin 2 3.0 (1.8 to 4.9) 0.298Piroxicam 10 3.8 (2.7 to 5.2) 0.087Ketoprofen 7 4.2 (2.7 to 6.4) 0.258Azapropazone 2 9.2 (4.0 to 21.0) 0.832*Adapted from Henry et al.3

†Relative to ibuprofen.

Cyclo-oxygenase enzymes (COX-1 and COX-2)

NSAIDs

Thromboxane A2 Prostaglandins

Arthritis relievedHaemostasis impaired andmucosal protection lost

Figure 8.4 Inhibition of cyclo-oxygenase enzymes byNSAIDs relieves inflammation and pain but alsoremoves mucosal protection of gastric epithelium

Indigestion and non-steroidal anti-inflammatory drugs

27

and normal doses of proton pump inhibitors are protective,and long term studies have shown them to prevent both gastricand duodenal ulcers. There have been no randomised studies ofthe effect of acid suppression on NSAID associatedcomplications, although indirect evidence from epidemiology isencouraging.

Choice of treatmentPatients with NSAID associated ulcersIf patients present with ulcers NSAIDs should be stopped ifpossible since they retard healing. For patients who need tocontinue taking NSAIDs, large comparative studies have shownthat omeprazole 20 mg daily results in faster healing of gastricand duodenal ulcers than ranitidine 150 mg twice daily4 ormisoprostol 200 �g four times daily and is better tolerated thanmisoprostol.

Subsequent prevention of relapseStudies have shown that, once healing is achieved, NSAIDassociated ulcer relapse can be retarded by use of omeprazole,misoprostol, or high dose famotidine.4 These comparativestudies—based on preventing the development of ulcers,multiple erosions, or moderate to severe dyspepsia—have shownoverall higher efficacy for omeprazole 20 mg daily thanmisoprostol 200 �g twice daily or ranitidine 150 mg twice daily.4

In these studies omeprazole protected against ulcers, bothgastric and particularly duodenal, and erosions. Misoprostol wasassociated with the same rate of duodenal ulcer formation asplacebo but was particularly effective in preventing multipleerosions. In these studies the site of the initial lesion was astrong predictor of the site of subsequent relapse.

NSAID users without ulcersMany studies have shown that misoprostol can inhibit ulcerdevelopment in such patients, as can famotidine 40 mg twicedaily and omeprazole. These drugs have not been compared forrelative effectiveness in this group of patients.

Practical prescribingPatients presenting with gastric or duodenal ulcers who need tocontinue taking NSAIDs should be treated with omeprazole20 mg daily or another proton pump inhibitor until the ulcerheals. Although omeprazole is the only proton pump inhibitorto have been studied in large scale trials, its benefits areprobably a class effect. Patients with multiple erosions insteadmay be better served by misoprostol.

Overall, subsequent maintenance treatment is likely to bemore effective and better tolerated with a proton pumpinhibitor than misoprostol. Recognition that the site and natureof the original lesion is a strong predictor of the site and natureof relapse can aid management.

For patients who present with duodenal ulcers a protonpump inhibitor is an appropriate maintenance treatment. Forpatients with multiple erosions misoprostol is appropriate iftolerated. On current data there is little to choose betweenproton pump inhibitors and misoprostol with regard to efficacyin preventing gastric ulcers, but proton pump inhibitors arebetter tolerated.

Role of Helicobacter pylori eradicationThis is an extremely controversial topic. One study, of patientsstarting NSAIDs (naproxen) for the first time without a historyof dyspepsia or ulceration, showed that eradication of H pylori

Box 8.2 Side effects of NSAIDsx Dyspepsiax Oesophagitisx Oesophageal stricturesx Gastric and duodenal petechiae, erosions,

ulceration, bleeding, and perforationx Type C gastritisx Small and large bowel ulceration, bleeding, and

perforationx Exacerbation of colitis

Patient taking NSAID

High*

Endoscopy

Low Ulcer absentUlcer present

NegativePositive

H2 receptor antagonistfor 6-8 weeksStop NSAID ifpossible, otherwisegive standard doseof proton pumpinhibitor long term

* Age >60 years; Concurrent use ofcorticosteroids or anticoagulants; Previousulcer; More than one or high dose of NSAIDused; Systemic disease

H pylori eradicationtreatmentStop NSAID ifpossible, otherwisegive standard doseof proton pumpinhibitor long term

Refer togastrointestinal

specialist

Test for H pyloriinfection

Dyspepsia persists

Stop NSAID if possible orDecrease dose or

change class of drug

Standard dose of proton pumpinhibitor once daily or

Misoprostol 200 µg four times daily

Dyspepsia Haematemesis, malaena, anaemia

If gastric ulcer proved, repeat endoscopyafter 6-8 weeks' treatment to confirm healing

Risk of gastro-duodenal ulceration? Urgent endoscopy

Dyspepsia persists

Endoscopy

Figure 8.5 Algorithm for managing gastrointestinal side effects of NSAIDs

Figure 8.6 Chemical (type C) gastritisshowing oedema and mild chronicinflammation of lamina propria andvertical smooth muscle fibres and slightfoveolar hyperplasia. Contrast with Hpylori induced gastritis, which has amarked neutrophil infiltrate in the laminapropria and deeper gastric glands


28

substantially reduced the rate of gastric ulcer formation at twomonths. In another study, of patients who had previously hadulcers or moderate to severe dyspepsia,H pylori eradication didnot influence outcome at six months. Finally, althoughomeprazole is effective prophylaxis in patients without H pyloriinfection, it is more effective in those who remain infected.4

Most doctors, feeling uncomfortable about persistentH pylori infection, eradicate it in NSAID users. In our view thisis irrational if these patients are at sufficiently high risk to becandidates for co-treatment with acid suppressing drugs.

The futureA new generation of less toxic NSAIDs is probably imminent.Specific inhibitors of the inducible cyclo-oxygenase-2 enzyme(COX-2), which probably leave protective gastric prostaglandinsuninhibited, are now available. Meloxicam, another recentNSAID, is well tolerated. Whether this is because it is a partiallyselective COX-2 inhibitor or because tested doses are low isuncertain. Other developments include NSAIDs that donatecytoprotective nitric oxide to the gastric mucosa.5

ConclusionsCurrently, there is a wide range of views about what is anappropriate level of NSAID prescribing and no simple, allembracing resolution. In patients with risk factors forgastrointestinal complications the side effects of NSAIDs mayoutweigh all benefits. For such patients, omeprazole andmisoprostol can provide effective protection, and the choice isbetween patients’ generally poor tolerance of misoprostol andthe higher costs of omeprazole. Finally, a growing concern isthe correct management of low dose aspirin used forcardiovascular protection, and no patient studies havespecifically investigated this.

1 Hawkey CJ. Non steroidal anti-inflammatory drugs and peptic ulcers:facts and figures multiply, but do they add up? BMJ 1990;300:278-84.

2 Silverstein FE, Graham DY, Senior JR, Davies HW, Struthers BJ, BittmanRM, et al. Misoprostol reduces serious gastrointestinal complications inpatients with rheumatoid arthritis receiving nonsteroidalanti-inflammatory drugs. Ann Intern Med 1995;123:241-9.

3 Henry D, Lim LL-Y, Garcia Rodriguez LA, Perez Gutthann S, Carson JL,Griffin M, et al. Variability in risk of gastrointestinal complications withindividual non-steroidal anti-inflammatory drugs: results of acollaborative meta-analysis. BMJ 1996;312:1563-6.

4 Yeomans ND, Tulassay Z, Juhasz L, Racz I, Howard JM, van Rensburg CJ,et al, for the ASTRONAUT Study Group. A comparison of omeprazoleand ranitidine for treating and preventing ulcers associated with nonsteroidal anti inflammatory drugs.N Engl J Med 1998;338:719-26.

5 Hawkey CJ. Future treatments for arthritis—new NSAIDs, NO-NSAIDsor no NSAIDs? Gastroenterology 1995;109:614-6.

Box 8.3 NSAID protection strategiesx Use lowest possible doses of NSAIDsx Use safer NSAIDs

Low toxicity NSAIDsCOX-2 inhibitors

x Use NSAID prophylaxisProton pump inhibitorsMisoprostil

COX-2COX-1Coxibs

Thromboxane A2

Prostaglandins

Arthritis relievedProtection andhaemostasis maintained

Prostaglandins

Figure 8.7 Rofecoxib and celecoxib, selective COX-2inhibitors, can relieve pain and inflammation withoutrisking gastric ulceration

Indigestion and non-steroidal anti-inflammatory drugs

29

9 Upper gastrointestinal haemorrhageHelen J Dallal, K R Palmer

Acute upper gastrointestinal haemorrhage accounts for about2500 hospital admissions each year in the United Kingdom.The annual incidence varies from 47 to 116 per 100 000 of thepopulation and is higher in socioeconomically deprived areas.

Although hospital mortality has not improved over 50 yearsand remains at about 10%, older patients who have advancedcardiovascular, respiratory, or cerebrovascular disease that putsthem at increased risk of death now comprise a much higherproportion of cases. Many patients’ bleeding is associated withuse of non-steroidal anti-inflammatory drugs, but there is noevidence that prognosis is worse in patients who are takingthese drugs than in those who are not.

Presentation of bleedingAll patients who develop acute gastrointestinal bleeding needurgent assessment. Almost all should be admitted as anemergency to hospital. Only a small minority of young, fitpatients who have self limiting bleeding can be managed asoutpatients, but even those need urgent investigation. Patientswho present with haematemesis tend to have more severebleeding than those who present with melaena alone.

At the initial assessment it is important to define factors thathave prognostic importance. The main factors predicting deathinclude increasing age, comorbidity, and endoscopic findings.Mortality is extremely low in patients under 40 years old butthereafter increases steeply with advancing age. Patients whohave severe comorbidity—particularly renal insufficiency,hepatic failure, or disseminated malignancy—have a poorprognosis. Hospital admission may be precipitated bygastrointestinal bleeding in many of these patients, and death isoften due to disease progression rather than to bleeding.

A risk assessment score has been developed based on theoutcome of 4185 patients with acute gastrointestinal bleedingadmitted to hospitals in England.1 A series of independent riskfactors were scored, and the total score accurately predictsoutcome. Patients who score 2 or less have a mortality of 0.1%and a rebleeding rate of 4.3%, but a score in excess of 8 isassociated with a 41% mortality and rebleeding rate of 42.1%.

Patients who develop acute upper gastrointestinalhaemorrhage after hospitalisation for other serious illness havea much worse prognosis than those who are admitted becauseof bleeding, with a mortality of about 30%. Endoscopic findingsof active, spurting haemorrhage; a non-bleeding blood vesselvisible within an ulcer; and red spots on large varices areassociated with risk of further bleeding. The absence of theseendoscopic stigmata indicates little chance of rebleeding andearly discharge from hospital.

Causes of bleedingThe commonest cause of upper gastrointestinal haemorrhage ispeptic ulcer. A history of proved ulcer or ulcer-like dyspepsia isabsent in about 20% of cases. In these patients consumption ofaspirin or non-steroidal anti-inflammatory drugs is common.Infection with Helicobacter pylori is less prevalent in bleedingulcers than in uncomplicated ulcers. Severe ulcer bleeding isdue to erosion of the artery by the ulcer, and the severity ofbleeding depends on the size of the ulcer and the size of thearterial defect. Bleeding from a defect greater than 1 mm in

Box 9.1 Risk factors for death after hospital admission foracute upper gastrointestinal haemorrhagex Advanced agex Shock on admission (pulse rate > 100 beats/min; systolic blood

pressure < 100 mm Hg)x Comorbidity (particularly hepatic or renal failure and disseminated

cancer)x Diagnosis (worst prognosis for advanced upper gastrointestinal

malignancy)x Endoscopic findings (active, spurting haemorrhage from peptic

ulcer; non-bleeding, visible blood vessel; large varices with redspots)

x Rebleeding (increases mortality 10-fold)

Figures 9.1, 9.2., 9.3 Endoscopicstigmata associated with high risk offurther gastrointestinal bleeding. Topleft: an active, spurting haemorrhagefrom a peptic ulcer is associated withan 80% risk of continuing bleedingor rebleeding in shocked patients.Top right: a non-bleeding, visiblevessel represents either apseudoaneurysm of an eroded arteryor a closely adherent clot, and 50%of such patients rebleed in hospital.Left: large varices with red spots arealso strongly associated with bleeding

Peptic ulcer

VaricesMallory-Weiss tearErosive disease

Neoplasm

Other

No obvious cause

Oesophagitis

40%

5%10%5%6%

6%

24%

4%

Figure 9.4 Causes of acute upper gastrointestinalhaemorrhage

30

diameter is unlikely to stop spontaneously and does notrespond to endoscopic treatment. Large ulcers arising from theposterior part of the duodenal cap can erode thegastroduodenal artery and provoke brisk bleeding.

Bleeding from gastric erosions, oesophagitis, or vascularmalformations usually stops spontaneously and is not usuallylife threatening. Mallory-Weiss tears are a consequence ofretching, and most patients have a history of alcohol misuse,have features of other gastrointestinal disease such as pepticulcer or gastroenteritis, or have non-gastrointestinal causes ofvomiting. Bleeding usually stops spontaneously, althoughendoscopic haemostatic treatment is sometimes required.

Bleeding from upper gastrointestinal malignancy is notusually severe, and the prognosis is dictated by the stage of thedisease. Patients with extensive upper gastrointestinal cancerhave a dismal prognosis, but death is not usually a consequenceof gastrointestinal haemorrhage but of disease progression.

In any patient with acute gastrointestinal bleeding liverdisease should be considered because it requires specificmanagement. Oesophageal varices account for a smallproportion of cases but have a disproportionate impact onmedical resources. Bleeding is often severe, and other featuresof liver failure—such as fluid retention, hepatic encephalopathy,renal failure, and sepsis—often develop after the bleed. About athird of patients will die, and prognosis is related to the severityof the underlying liver disease rather than the size of varicealhaemorrhage.

Aortoduodenal fistula must be considered in patients whodevelop profuse bleeding and have undergone aortic aneurysmsurgery.

Management of bleedingPrimary care managementInitial resuscitation of shocked patients can be started beforehospital admission. Intravenous access should be obtained, andinfusion of a crystalloid started. Oxygen should be given.Management in primary care is limited, and the priority is toarrange early admission to hospital and to support associatedcomorbidity, such as that of angina or chest disease.

Hospital managementResuscitationThe first priority is to support the circulation rather than toidentify the source of bleeding. Endoscopy is undertaken onceresuscitation has been achieved. At least one large bore cannulais inserted into a substantial vein. When the pulse rate is morethan 100 beats/min or the systolic blood pressure falls below100 mm Hg, infusion with a crystalloid such as normal saline isstarted. The rate of infusion depends on the severity of shock. Arecent meta-analysis showed that crystalloids (such as normalsaline) should be used rather than colloids (such as dextrans). Ifblood transfusion is required the aim is to maintain ahaemoglobin concentration of 100 g/l. In patients withsuspected liver disease the use of normal saline should beavoided because of the risk of precipitating ascites.

EndoscopyAfter resuscitation, endoscopy is undertaken. In most cases thisis done electively on the next available routine list but within 24hours of admission. Only a minority of profusely bleedingpatients need “out of hours” emergency endoscopy.

On-call endoscopists must be experienced and be able toapply a range of endoscopic treatments. Endoscopy is necessaryto define the cause of bleeding, provide prognostic information,and to apply haemostatic treatment.

Box 9.2 Alert for features of liver diseasex Ascites x Splenomegalyx Jaundice x Fluid retentionx Alcohol misuse

Box 9.3 Blood tests on admission to hospitalHaemoglobin concentration—May be normal during the acute stagesuntil haemodilution occurs

Urea and electrolyte concentrations—Elevated blood urea suggests severebleeding

Cross match for transfusion—Two units of blood are sufficient unlessbleeding is extreme. If transfusion not needed urgently group theblood and save the serum

Liver function testsProthrombin time

Figure 9.5 Gross ascites and distended abdominal veins in advancedcirrhosis

Acute upper gastrointestinal haemorrhage

Resuscitation

Endoscopy

Varices

Bleedingcontinuesor recurs

Bleedingcontrolled

Sclerotherapy orbanding plusintravenousterlipressin

Conservativemanagement

and earlydischarge

Consider:angiography,colonoscopy,

operativeenteroscopy

Peptic ulcer (vascularmalformation

Mallory-Weiss tear)

Endoscopic injection,heat applicationor combination

Major SRH

No obvious cause

No majorSRH

Minorbleed

Majorbleed

Bleeding controlledBleeding continues

or recurs

Surgery

Reassessendoscopically

at 24 hours

TIPSS orsurgical referral

1. Banding programme2. Assess and manage underlying liver disease3. Consider propanolol

Figure 9.6 Algorithm for diagnosis and management of uppergastrointestinal bleeding (SRH=stigmata of recent haemorrhage,TIPPS=transjugular intrahepatic portosystemic shunt)

Upper gastrointestinal haemorrhage

31

Diagnosis—It is difficult to prove that diagnostic endoscopyimproves outcome, but it is clearly important to define a precisediagnosis in order to plan treatment.

Prognosis—Endoscopic stigmata are extremely useful indefining risk of further bleeding.

Treatment—A range of endoscopic treatments can beadministered to patients showing major endoscopic stigmata.Pharmacological, endoscopic, radiological, and surgicaltreatments are used.

TreatmentsDrugs

Non-variceal haemorrhage—There is increasing evidence tosupport the use of intravenous omeprazole, which in clinicaltrials reduces the risk of rebleeding and the need for surgicaloperation. Patients infected with H pylori should undergoeradication treatment after haemostasis has been achieved inorder to prevent further ulcer complications. One study hasshown that tranexamic acid reduces transfusion requirements inpatients presenting with non-variceal haemorrhage.

Variceal haemorrhage—Vasoactive drugs (such as terlipressin)reduce bleeding rates but have little impact on survival. Ifbleeding continues despite this treatment, a modifiedSengstaken-Blakemore (Minnesota) tube is inserted. It must beremembered that both vasoactive drugs and the Minnesota tubeare temporising measures used to control active bleeding untildefinitive endoscopic, surgical, or radiological measures aretaken. When varices have been obliterated portal pressure isreduced with propanolol at a dose to decrease the pulse rate by20%. This diminishes the risk of subsequent rebleeding.

Endoscopic treatmentNon-variceal bleeding—A range of endoscopic haemostatic

approaches are available. Each has a similar efficacy, but there isevidence that an injection combined with a thermal method isbest. Endoscopic treatment fails in about 20% of patients withbleeding ulcers, most often those with large, actively bleedingposterior duodenal ulcers. Endoscopist and surgeon must worktogether to identify and treat these patients at an early stage.

Varices—When active variceal bleeding is seen at endoscopy,intravariceal injection of a sclerosant (such as 5% ethanolamine,1% polidoconal, or sodium tetradecyl sulphate) is attempted. Analternative approach is oesophageal band ligation. Bandingobliterates varices more efficiently and has few complications,but it may be more difficult to perform in a patient with activebleeding.

SurgerySurgery is the best way of stopping active ulcer bleeding and

preventing rebleeding, but it carries high morbidity andmortality. It is now reserved for cases in which endoscopictreatment has failed. Specific protocols defining indications fora surgical operation are necessary.

Surgical treatment for acute variceal bleeding, includingoesophageal transection with devascularisation and porto-cavalshunt, is rarely done because of unacceptable mortality. It hasbeen replaced by TIPPS (transjugular intrahepaticportosystemic shunt). However, both interventions mayprecipitate encephalopathy.

1 Rockall TA, Logan RF, Devlin HB, Northfield TC. Risk assessment afteracute upper gastrointestinal haemorrhage.Gut 1996;38:316-21.

Box 9.4 Endoscopic treatment for non-variceal bleedingThermalx Heater probe x Multipolar electrocoagulation

Injectionx Adrenaline (1:10000 to x Sclerosants (ethanolamine,

1:100000) 1% polidoconal)x Alcohol (98%) x Procoagulants (thrombin, fibrin glue)

Mechanicalx Clips x Sutures x Staples

Combination

Box 9.5 Indications for surgical operation for peptic ulcerbleedingx Active bleeding unresponsive to endoscopic haemostasis

Profuse bleeding preventing endoscopic visualisation and treatmentBleeding continues despite application of endoscopic treatment

x Endoscopically proved rebleeding despite technically successfulendoscopic treatmentPatients at low risk of death, after two unsuccessful attempts atendoscopic treatmentHigh risk patients, after one failure of endoscopic haemostasis

Figure 9.7 Minnesota tube

Figure 9.8 Endoscopic treatment of varices. Intravariceal injection ofsclerosant (left) and band ligation of oesophageal varices (right)


32

10 Indigestion: When is it functional?Nicholas J Talley, Nghi Phung, Jamshid S Kalantar

Patients often complain of indigestion, but what do they mean?Indigestion is an old English word that means lack of adequatedigestion, but patients and doctors interpret this in differentways. Many patients mean heartburn or acid regurgitation, theclassic symptoms of gastro-oesophageal reflux disease. Somedescribe belching, abdominal rumblings, or even bad breath asindigestion. Others mean pain localised to the epigastrium or anon-painful discomfort in the upper abdomen which may bedescribed as fullness, bloating, or an inability to finish a normalmeal (early satiety). Dyspepsia is best restricted to mean pain ordiscomfort centred in the upper abdomen.

There are many causes of dyspepsia, but at least two thirdsof patients have no structural or biochemical explanation fortheir symptoms. It has been suggested that dyspepsia can besubdivided based on groups (or clusters) of symptoms.However, subgroups have not proved to be of value inidentifying the underlying cause of dyspepsia and overlapconsiderably. Some patients report having troublesome burpingassociated with abdominal bloating or discomfort that istransiently relieved by bringing up the wind. These patientshave aerophagy, and repeated swallowing of air may be obviousduring the consultation.

Causes of dyspepsiaHistory taking is key to identifying the likely cause of dyspepsia.

Gastro-oesophageal reflux diseaseIt is important and practical to distinguish gastro-oesophagealreflux disease (GORD) from dyspepsia. Frequent heartburn is acardinal symptom of GORD; acid reflux causes a retrosternal orepigastric burning feeling that characteristically radiates uptowards the throat, is relieved transiently by antacids, and isprecipitated by a meal or by lying down.

Up to 60% of people with upper gastrointestinal symptomsreport both heartburn and epigastric pain or discomfort. Thisoverlap can be confusing, but it is not the presence of asymptom but its predominance that is most helpful clinically.For example, if the main complaint is a burning epigastric painthat radiates up towards the throat then this is highly predictiveof GORD (as can be objectively demonstrated by abnormalresults from 24 hour oesophageal pH monitoring).

Reflux oesophagitis can be detected at endoscopy, but overhalf of patients with true GORD will not have evidence ofmucosal breaks (erosions). Oesophageal erythema or thepresence of a hiatal hernia are unreliable signs that cannot beused to determine if a patient has reflux disease. Although somepatients are unable to adequately describe their symptoms ordecide which is their predominant complaint, if a detailedhistory is taken a clinical diagnosis of GORD can be made inmost cases, including those in whom endoscopy is normal.

Peptic ulcersMany textbooks continue to propagate the myth that symptomscan accurately identify peptic ulcer disease. Unfortunately,classic ulcer symptoms (such as postprandial epigastric pain ornight pain) often occur in patients with functional dyspepsia,and many patients with an ulcer have atypical complaints.

Box 10.1 Major structural causes ofdyspepsiax Chronic peptic ulcer (duodenal or gastric)x Gastro-oesophageal reflux disease ( > 50% have

no oesophagitis)x Gastric or oesophageal adenocarcinoma (rare

but of concern for patient and doctor)

Box 10.2 Uncommon causes of upper abdominal pain ordiscomfort that may be confused with dyspepsiax Aerophagy (repetitive belching from air swallowing)x Biliary colic from gall stonesx Abdominal wall pain (a clinical clue is localised tenderness on

palpation not reduced by tensing the abdominal wall muscles)x Chronic pancreatitis (episodic dull steady upper abdominal pain

that may be aggravated by meals and radiate through to the back)x Malignancy (such as of pancreas or colon)x Mesenteric vascular insufficiency (postprandial pain, weight loss,

and a fear of eating)x Anginax Metabolic disease (such as diabetes, renal failure, hypercalcaemia)

Box 10.3 Conditions to be recognised from a patient’s historySymptomatic gastro-oesophageal reflux diseasex Burning retrosternal or epigastric pain or discomfort radiating

upwards towards the throat and relieved, albeit transiently, byantacids

x Regurgitation of acid

Irritable bowel syndromex Abdominal pain plus an erratic disturbance of defecation linked to

the pain (such as pain relief with defecation, looser or harder stoolswith pain onset, or more frequent or less frequent stools with painonset)

Biliary tract diseasex Biliary-type pain

Peptic ulcerx Classic ulcer symptoms do not distinguish peptic ulcer disease from

functional dyspepsia

1%Dysmotility-likeUlcer-like

Symptomatic gastro-oesophagealreflux disease (reflux-like)

9%

9%3%

18%

16%

27%

17%Non-specific

Figure 10.1 Overlap of subgroups of dyspepsia basedon symptoms in patients with documented functionaldyspepsia

33

Endoscopy remains the test of choice to rule out chronicpeptic ulceration, but its presence can now be inferred byindirect testing.Helicobacter pylori causes 90% of duodenal ulcersand 70% of gastric ulcers; aspirin and non-steroidalanti-inflammatory drugs (NSAIDs) account for most of theremainder. Patients who are not infected with H pylori and nottaking NSAIDs have a very low probability of ulcer disease.

Gastric cancerFear of gastric cancer is one of the main reasons why patientswith dyspepsia present to their general practitioner. Gastriccancer is found in less than 2% of all cases referred forendoscopy. Early gastric cancer comprises only 10% of cancercases, but it is important to diagnose because it is curable and60-90% of patients initially present with dyspepsia.

However, the risk of gastric cancer is extremely low inpatients under the age of 55 years presenting with the newonset of dyspepsia in most Western countries including Britain.Furthermore, “alarm” symptoms such as weight loss, dysphagia,or anaemia help to identify those who need to be investigatedin order to exclude malignancy, although between 15% and50% of dyspeptic patients with gastric cancer do not have thesesymptoms. Endoscopic evaluation is therefore recommended inolder patients presenting with new symptoms and in all patientswith alarm symptoms.

Gall stonesUltrasonography will detect gall stones in a minority of patientswith apparently unexplained dyspepsia. However, gall stonesare common and often incidental in the absence of biliarysymptoms. Biliary colic is characteristically severe, episodic, andconstant (rather than colicky) pain in the epigastrium or rightupper quadrant typically lasting one to several hours. This canusually be easily distinguished from the pain or discomfort offunctional dyspepsia.

While many patients with gall stones also complain ofbloating, nausea, and other vague upper abdominal symptoms,these complaints are just as common in patients without gallstones. Moreover, cholecystectomy does not reliably result inlong term relief of any of these vague complaints and cannot berecommended. Cholecystectomy in a patient with non-biliarytype pain is likely to result in the patient at a later date beinglabelled as having the post-cholecystectomy syndrome.

Functional dyspepsiaMost commonly, either no abnormalities or irrelevantabnormalities (such as gastric erythema or a few gastricerosions) are found at endoscopy; these patients are labelled ashaving functional (or non-ulcer) dyspepsia. As antisecretorydrugs may result in healing of ulcers or oesophagitis (andhence lead to a misdiagnosis of functional dyspepsia), thesedrugs are best not started before endoscopy if possible.

Functional dyspepsiaPathogenesisThe pathogenesis of functional dyspepsia remains uncertain.H pylori gastritis is detected in about half of patients withfunctional dyspepsia, but it is also common in otherwiseasymptomatic people. The question of whether this infectioncauses symptoms in patients without ulcer disease has beencontroversial. There is no evidence that specific symptomsidentify those with H pylori infection. Acid secretion is usuallynormal in patients with functional dyspepsia, except perhaps ina subset infected with H pylori.

Box 10.4 Alarm symptomsx Anorexiax Loss of weightx Anaemia due to iron deficiencyx Recent onset of persistent symptomsx Melaena, haematemesisx Dysphagia

0

20

40

60

80

Prev

alen

ce o

f sym

ptom

s (%

)

Weight

loss

Persist

ent v

omitin

g

Dysph

agia

Anaem

ia

Haemate

mesis

or mela

ena

Palpab

le mas

s

Family

histo

ry

Surgery

Metasta

ses

Perfora

tion

Figure 10.2 Prevalence of “alarm” symptoms in patientswith gastric cancer

Figure 10.3 Antral erythema and erosions inpatient with functional dyspepsia

Practice pointIf possible, H2 receptor antagonists andproton pump inhibitors should not bestarted before endoscopy (or should bestopped at least two weeks beforeendoscopy)

Other conditionsPsychological distress

ChemicalAcid dysregulation ormucosal sensitivityNSAIDBile

MechanicalAltered gastric andduodenal sensation(such as to balloon

distension)

FunctionalDelayed gastric

emptying BacterialH pylori

Figure 10.4 Pathogenesis of functional dyspepsia


34

In functional dyspepsia gastric (and duodenal) sensation isdisturbed (the “irritable stomach”), and in about half of patientsdistension induces symptoms at lower pressures or volumesthan it does in healthy people. Delayed gastric emptying can bedetected in a quarter to a half of patients with functionaldyspepsia. In addition, a subset of patients have alteredintragastric distribution of food, which reflects abnormalproximal gastric relaxation (a “stiff” fundus). There is anincreased probability of detecting gastric motor abnormalitiesin women and possibly in those with severe postprandialfullness or severe vomiting.

There is controversy as to whether functional dyspepsia is a“forme fruste” of the irritable bowel syndrome, and bothconditions may overlap. About a third of patients withfunctional dyspepsia have an erratic disturbance of defecationclosely linked to their pain, and probably truly have irritablebowel syndrome. There is also evidence of gut hypersensitivityin both functional dyspepsia and the irritable bowel syndrome.

Smoking and alcohol do not seem to be important infunctional dyspepsia, but coffee ingestion has been linked toexacerbation of symptoms. Some patients with functionaldyspepsia suffer from an anxiety disorder or depression, butwhether this is cause or effect remains unclear.

Identification and management

Investigation versus testing for H pyloriThe devil is in the detail, and a careful detailed appraisal of apatient’s history with a judicious approach to testing isnecessary. Older patients or those with alarm features warrantprompt referral for endoscopy and further investigations asrequired. Testing for H pylori infection will help in guidingmanagement in the remainder; between 20% and 60% of thosewith H pylori infection will have peptic ulcer disease.

For patients with H pylori infection, one course of action isto refer them for endoscopy to determine who has peptic ulcerdisease or functional dyspepsia (the two main considerations)and plan treatment accordingly (the “test and endoscope”strategy). Alternatively, a reasonable course of action is to treatinfected patients with appropriate antibiotics and observe theclinical course (the “test and treat” strategy). Although treatmentof infection may not cure functional dyspepsia (see below), itwill usually eliminate the peptic ulcer diathesis and hence willoften relieve the symptoms. Moreover, recent trials suggest that“test and treat” is a safe and cost effective strategy that results ina long term outcome similar to that with a strategy of promptendoscopy. Hence, “test and treat” has been gaining widespreadacceptance.

Principles of managementReassurance and explanation remain the key elements inmanaging documented or suspected functional dyspepsia.Patients should be advised that this is a real condition and thattheir symptoms are not imaginary. Furthermore, they should beadvised that the condition never leads to cancer or otherserious disease. Patients’ fears should be identified andaddressed. Modification of diet (such as avoiding foods thatprovoke symptoms and adopting a low fat diet because high fatfoods may impair gastric emptying) and stopping medicationscan be helpful. Antacids are no better than placebo infunctional dyspepsia, but notably the placebo response rangesbetween 20% and 60%.

Initial treatmentIn patients with a diagnosis of functional dyspepsia, a shortterm therapeutic trial for four weeks with an acid suppressant is

Box 10.5 Initial management of functional dyspepsiax Make a positive clinical diagnosisx Minimise investigations and don’t repeat tests without good reasonx Determine the patient’s agenda and identify psychosocial stressorsx Find out why the patient has presented nowx Provide education and reassurance

Box 10.6 Drugs that are most likely to cause dyspepsiax Non-steroidal anti-inflammatory

drugsx Digoxinx Antibiotics (macrolides,

metronidazole)x Corticosteroids, oestrogens

x Iron, potassium chloridex Levodopax Theophyllinex Quinidinex Niacin, gemfibrozilx Colchicine

Study

Blum et alKoelz et alMcColl et alTalley et alTalley et alMiva et alMalfertheiner et alBruley des Varannes et alFroehlich et al

Total

Test for heterogeneity Q=7.09, df=8, P=0.53

Treatment

119/16467/89

121/154101/13381/15033/48

269/46074/12931/74

Placebo

No of patients still dyspeptic

Risk ratio (95% CI)

130/16473/92

143/154111/14272/14328/37

143/21486/12434/70

Risk ratio (95% CI)

0.92 (0.81 to 1.03)0.95 (0.81 to 1.11)0.85 (0.77 to 0.93)0.97 (0.85 to 1.11)1.07 (0.86 to 1.34)0.91 (0.70 to 1.18)0.88 (0.77 to 0.99)0.83 (0.68 to 1.00)0.86 (0.60 to 1.24)

0.91 (0.60 to 0.96)P=0.0002

0.75 1.0 1.51.25

Figure 10.5 Results of systematic review comparing H pylori eradicationtreatment for non-ulcer dyspepsia with placebo

Patient with dyspepsia

Long history Short history

Careful, detailed history• "Have you ever had this or similar symptoms"

Is it peptic ulcer?• Test for H pylori

Referral orempiric H pylori

eradication treatment

"Alarm" features?

No

Positive

Yes

Goodresponse

Poorresponse

Negative

Is it reflux?• Clinical features• Response to protonpump inhibitors

Key management points• Fear of gastric cancer• Gall stones are common and often asymptomatic

Protonpump inhibitors

Is itfunctional?

Careful empirictreatment orinvestigation

Urgentreferral

Figure 10.6 Algorithm for management of functional dyspepsia

Ulcer-like(predominantlyepigastric pain)

Functionaldyspepsia Non-specific

Dysmotility-like(predominantly

discomfort rather thanpain, such as

fullness or bloating)

Empirictreatment

Prokinetic agentAntisecretory agent Prokinetic agent

Figure 10.7 Empiric treatment for functional dyspepsia

Indigestion: When is it functional?

35

worth while. Symptom subgroups may be of some help inpredicting a patient’s response to treatment if a predominantsymptom can be identified, but relying on clusters of symptomsis generally not useful.

H2 receptor blockers are, at best, of only modest efficacy infunctional dyspepsia. While some trials suggest that they aresuperior to placebo, other trials have not shown any additionalbenefit. Proton pump inhibitors are more efficacious thanplacebo in functional dyspepsia (by about 10-20%), but not inthe subgroup of patients with dysmotility-like dyspepsia.

If a patient fails to respond to an antisecretory drug afterfour weeks, it is reasonable to consider increasing the dose orswitching to an alternative or a prokinetic drug. If the patienthas failed to respond after eight weeks, then it is reasonable torefer the case to a specialist for further evaluation.

Eradicating H pylori infection cures functional dyspepsia inonly a minority of cases. A meta-analysis has suggested a smalltherapeutic gain over 12 months follow up (15 needed to betreated to cure one case). Whether sucralfate or bismuth isbetter than placebo in treating functional dyspepsia is unclear,but it is unlikely; misoprostol may actually aggravate symptomsand cause diarrhoea in some patients.

Long term managementFunctional dyspepsia is generally a relapsing and remittingcondition. Treatment should not be prolonged, and frequentdrug holidays should be prescribed. In patients with symptomsthat are difficult to control a trial of an antispasmodic orantidepressant may be useful, but specialist referral to confirmthe diagnosis and exclude rare causes of dyspepsia should firstbe considered. Some patients will benefit from behaviouraltherapy or psychotherapy.

AerophagyAir swallowing is often extremely resistant to treatment.Options include avoidance of chewing gum, aerated drinks, andsmoking; use of anti-gas agents (such as activated dimeticone orcharcoal); and relaxation therapy. However, no treatment is ofproved benefit, and anti-gas agents are no better than placebo.

The graph of prevalence of alarm symptoms in patients with gastric canceris adapted from Gillen D, McColl KE. Am J Gastroenterol 1999;94:75-9. Theforest plot comparing H pylori eradication treatment with placebo isreproduced from Moayyedi P, et al. BMJ 2000;321:659-64.

Box 10.7 Treatment for functional dyspepsiaInitial treatmentx Antisecretory drug (H2 receptor blocker, proton pump inhibitor)orx Prokinetic drug (domperidone) if antisecretory treatment failsx Switch treatment if first drug type fails

Resistant cases (failed initial treatment)x H pylori eradicationx Sucralfate or bismuthx Antispasmodic agent (such as mebeverine)x Antidepressant (such as selective serotonin reuptake inhibitor or

tricyclic drug)x Behavioural therapy or psychotherapyx No treatment is proved to be of benefit in these patients

Box 10.8 When to consider referring adyspeptic patient to a specialistx If prompt investigation is required (such as

recent onset of alarm symptoms)x Severe painx Failure of symptoms to resolve or substantially

improve after appropriate treatmentx Progressive symptoms


36

11 Upper abdominal pain: Gall bladderC D Johnson

Gall stones are common but often do not give rise tosymptoms. Pain arising from the gall bladder may be typical ofbiliary colic, but a wide variety of atypical presentations canmake the diagnosis challenging. After a period of uncertainty inthe 1980s, when operative techniques were challenged by drugtreatment and lithotripsy, it is now widely accepted thatsymptomatic gallbladder stones should be treated bylaparoscopic cholecystectomy. Clinical judgment and localexpertise will greatly influence the management of bile ductstones, particularly if cholecystectomy is also required.

Epidemiology of gall stonesIn the United Kingdom about 8% of the population aged over40 years have gall stones, which rises to over 20% in those agedover 60. Fortunately, 90% of these stones remain asymptomatic,but cholecystectomy is the most commonly performedabdominal procedure.

The incidence of gall stones varies widely, being greatlyinfluenced by dietary intake, particularly of fat. For example, inSaudi Arabia gallstone disease was virtually unheard of 50 yearsago, but, with increasing affluence and a Western type diet, gallstones are now as common there as in many Western countries.Genetic factors also contribute. The native Indian populationsof Chile and Peru are highly susceptible, with a close to 100%lifetime risk of gall stones in their female population. Severalrisk factors have been identified, which relate to the two majorstone types, cholesterol stones and pigment stones.

PathogenesisGall stones form when the solubility of bilirubin or cholesterolis exceeded. Pigment stones arise in the gall bladder when therehas been increased bilirubin production from breakdown ofhaemoglobin. Mixed stones contain both bilirubin andcholesterol and may be calcified. Precipitated bilirubin mayform a nidus for subsequent cholesterol deposition.

Secondary pigment stones form in the bile duct as aconsequence of obstruction or by accumulation around a smallprimary stone. These stones are associated with bacterialinfection and arise by bacterial deconjugation of thebilirubin-glucuronide complex.

Cholesterol stones arise because of an imbalance in themechanisms maintaining cholesterol in solution. Cholesterol isa hydrophobic molecule and is dispersed in micelles by thecombined action of bile salts and lecithin. The risk ofprecipitation is directly related to cholesterol concentration andinversely to the concentrations of bile salts and lecithin, givingrise to a triangular coordinate. Increased cholesterol excretionis largely of dietary origin but may also result from changes insteroid metabolism associated with pregnancy, oralcontraceptives, and obesity.

Bile salts are retrieved from the gut by the terminal ileum,and this enterohepatic circulation is essential for maintenanceof the bile salt pool. The endogenous synthesis of bile salt is ratelimited at a level much lower than its normal daily excretion bythe liver. Many gastrointestinal diseases affect bile saltmetabolism—in particular, Crohn’s disease and surgicalresection of the terminal ileum predispose people to gall stones.

Asymptomatic gall stones are common and require notreatmentTypical symptoms include biliary colic—right upperquadrant pain, radiating to the back, and lasting less than12 hoursSymptomatic gall stones are usually treated bylaparoscopic cholecystectomy

Box 11.1 Risk factors for gall stonesCholesterol stonesx Obesityx High fat dietx Oestrogens (female, pregnancy, oral contraception)x Hereditaryx Loss of bile salts (Crohn’s disease, terminal ileal resection)x Impaired gall bladder emptying (such as truncal vagotomy, type 1

diabetes, octreotide, parenteral nutrition, and starvation or rapidvoluntary weight loss)

Pigment stonesx Haemolytic diseasex Biliary stasisx Biliary infection

Figure 11.1 Mixed gall stone with bilirubin nucleusand attached clear cholesterol crystals

100

Perc

enta

ge ch

oles

tero

l

80

60

40

20

00

0

20

40

60

80

100

2040

Cholesterolcrystals

Cholesterolsolution

6080100

Percentage lecithin

Percentage bile salt

Figure 11.2 Triangular coordinates relating solubility ofcholesterol with concentrations of cholesterol, bile salts,and lecithin

37

Impaired gallbladder emptying predisposes to gall stones byincreasing the time that material stays in the gall bladder,allowing excessive crystal growth. In addition, the dilating andflushing effect of fresh hepatic bile is lost when the gall bladdercontracts poorly.

Symptoms associated with gall stonesBiliary colic is usually felt as a severe gripping or gnawing painin the right upper quadrant. It may radiate to the epigastrium,or around the lower ribs, or directly through to the back. It maybe referred to the lower pole of the scapula or the right lowerribs posteriorly. However, many variations on this pattern havebeen described, including retrosternal pain and abdominal painonly in the epigastrium or on the left side. Such symptoms, inthe presence of gallbladder stones, merit consideration ofcholecystectomy.

There may be difficulty when symptoms are less clear. In ayear about 25% of the adult population consults a generalpractitioner for dyspeptic symptoms. As nearly 8% of theseindividuals will have asymptomatic gall stones, many patientswith dyspeptic symptoms are given the label “gallstonedyspepsia.” A pattern of symptoms supposedly associated withgall stones has been described, but several careful studies ofpatients before and after cholecystectomy have failed to showany clear association with either a good or poor outcome. Sinceasymptomatic gall stones and dyspepsia are so common in thegeneral populations, they often coexist. Dyspeptic symptomsmay be too readily attributed to the presence of gall stones,leading to inappropriate and ineffective surgery. Notsurprisingly, therefore, symptoms may persist in up to 20% ofpatients after cholecystectomy.

ComplicationsGallbladder stones may be complicated by acute cholecystitis,mucocele, or empyema. These are difficult to distinguishclinically; a patient may present with an episode of acutecholecystitis that fails to resolve and at operation is found tohave an empyema or a mucocele. In addition to symptoms ofbiliary colic, such patients have pain that is constant and lastsfor more than 12 hours; they also have tenderness over the gallbladder, which may be palpable, and may have a fever andleucocytosis.

Complications of bile duct stones include obstructivejaundice and acute pancreatitis.

Any patient with suspected complications of eithergallbladder or bile duct stones should be referred for urgentspecialist assessment and may well require immediateadmission to hospital.

DiagnosisUltrasonography has replaced cholecystography as thediagnostic test for gall stones. About 95% of gallbladder stoneswill be detected by ultrasonography, which is cheap, quick, andharmless. If strong clinical suspicion of gall stones exist, andultrasonography does not show stones, the test should berepeated. Other diagnostic tests are less sensitive and are rarelyindicated.

ManagementThe management of gall stones depends on their position,either in the gall bladder or bile duct.

Box 11.2 Symptoms associated with gall stonesBiliary colicx Right subcostal or epigastric pain radiating to back or lower pole of

scapula lasting for 20 minutes to 6 hoursx Associated with vomiting brought on by (any) foodx May disturb sleep

Acute cholecystitisx Severe pain and tenderness in right subcostal region for > 12 hoursx Fever and leucocytosis

Obstructive jaundice with or without pain

Box 11.3 Symptoms of dyspepsia not associated with gallstonesx Repeated belchingx Inability to finish normal mealsx Fluid regurgitationx Nausea

x Fullness after normal mealsx Abdominal distension (bloating)x Epigastric or retrosternal burningx Vomiting (without biliary colic)

Storage and concentrationof bile salts in gall bladderbetween meals

Liver

Gall bladder

Small intestine

Largeintestine

Bile salts aid dispersion,digestion, and absorption of fat

Bile salts reabsorbed by specialisedmucosa in terminal ileum

Bile saltscarried in

portal venousblood to liver

Figure 11.3 Enterohepatic circulation of bile salts. Eachmolecule circulates at least once for each meal

Figure 11.4 Top: Ultrasound image of gall bladder withdark area (a) representing gall bladder and multiplewhite echoes (b) representing stones. Bottom: The gallbladder after cholecystectomy with multiple small stones


38

Gallbladder stonesThe management of gallbladder stones is now relativelystraightforward. Asymptomatic gallstones require nointervention as the risks of any procedure outweigh thepotential benefits.

In the 1980s dissatisfaction with the outcome of opencholecystectomy led to several alternative therapies such asextracorporeal shock wave lithotripsy and bile salt dissolutiontherapy. These treatments were restricted in their applicabilityand have been almost completely superseded by laparoscopiccholecystectomy. This procedure offers a more rapid recoveryand return to work, much less abdominal scarring, and at leastas good long term relief of symptoms as open cholecystectomy.In specialist hands almost all uncomplicated gallbladder stonescan be dealt with laparoscopically, with minimal risk of injury tothe bile duct.

Complicated gallbladder stones—If complications arise (such asacute cholecystitis, mucocele, or empyema) cholecystectomy isperformed. This will usually be by the laparoscopic approach,but up to 10% of operations have to be converted tolaparotomy. In some elderly patients with acute presentation,percutaneous cholecystostomy is performed under ultrasoundcontrol to relieve the infection and avoid the morbidity of anemergency operation. Subsequently, the stones may beextracted percutaneously, leaving the gall bladder in place, or, ifappropriate, cholecystectomy is performed.

Stones in the bile ductStones may migrate from the gall bladder into the bile duct.Cholangiography is performed before, during, or immediatelyafter cholecystectomy to demonstrate the presence or absenceof bile duct stones in patients with known risk factors. Acholangiogram may be obtained endoscopically, at operation,or by means of magnetic resonance imaging. Uncomplicatedbile duct stones should be removed when they are detected,because of the high risk of complications such as acutepancreatitis, obstructive jaundice, or cholangitis if they are leftin situ.

The management of asymptomatic duct stones iscontroversial. The traditional approach of opencholecystectomy with exploration of the common bile ductoffers simplicity and widespread applicability and avoidsexposing the patient to the risk of procedure relatedpancreatitis. Alternatives are laparoscopic cholecystectomy withendoscopic sphincterotomy and stone extraction either beforeor after cholecystectomy, or else laparoscopic exploration of thecommon bile duct. Currently practice varies according to theexpertise available locally.

Acalculous biliary painThe symptoms of biliary colic are characteristic but may occurin the absence of gall stones. In such cases a specialist mustdecide whether an operation to remove the gall bladder isappropriate, in the belief that symptoms are due to microscopiccrystals (microlithiasis) or to a structural abnormality of thecystic duct.

Occasionally, biliary colic seems to be associated with a highpressure sphincter of Oddi, and symptoms may resolve afterendoscopic sphincterotomy. Alternative explanations for socalled acalculous biliary pain include irritable bowel syndromewith upper gastrointestinal manifestations (see previous article).Chronic pancreatitis must also be carefully excluded. Anydecision to carry out a cholecystectomy for this conditionshould be made by a hepatobiliary specialist.

Box 11.4 Options for treatment of symptomatic gallbladderdiseaseLaparoscopic cholecystectomy—Safe in specialist hands, rapid recovery,permanently effective, current gold standard

Open cholecystectomy—Traditional, painful, prolonged recovery, scar

Alternative therapiesExtracorporeal shock wave lithotripsy—ComplexBile salt dissolution—ExpensivePercutaneous cholecyslithotomy—Leaves abnormal gall bladder in situ,high recurrence rate, suitable only for a few selected patients

Box 11.5 Risk factors suggesting presence of bile duct stonesat cholecystectomy for symptomatic gallbladder stonesx Common bile duct dilated ( > 6 mm on ultrasound)x Recent abnormal levels of liver enzymes or bilirubinx History of acute pancreatitisx History of obstructive jaundice

Chest wall

Rib

Catheterpassedthroughliver intogall bladder

Inflamed, thickwalled gall bladder

Gall stone impactedin cystic duct

Lung

Liver

Diaphragm

Figure 11.5 Percutaneous cholecystostomy for acutecholecystitis. Percutaneous drainage relieves the acutephase, allowing subsequent stone extraction via the draintrack or cholecystectomy when inflammation has resolved

Figure 11.6 Cholangiography forduct stones. Top left: Endoscopicretrograde cholangiogram showingtwo stones in the bile duct (arrows).Top right: Operative cholangiogram(no stones). Bottom: Magneticresonance cholangiogram obtainedby 3-D reconstruction from a singlebreath hold acquisition

Upper abdominal pain: Gall bladder

39

Gallbladder cancerGallbladder cancer is rare and usually asymptomatic until at anadvanced stage. Usually it is associated with gall stones and maybe discovered incidentally at operation. Suspicious features in apatient with biliary symptoms include weight loss, anaemia,persistent vomiting, and a palpable mass in the right upperquadrant. Such patients require urgent investigation. Theprognosis is good if the disease is diagnosed at an early stage,but complete resection is often not possible because of theadvanced stage at presentation.

The diagram of triangular coordinates relating cholesterol solubility withbile salts and lecithin is adapted from Admirand WH, Small DM. J ClinInvest 1968;47:1043-52. The magnetic resonance cholangiogram waskindly provided by Dr C N Hacking.


40

12 Cancer of the stomach and pancreasMatthew J Bowles, Irving S Benjamin

Cancers of the stomach and the pancreas share similarly poorprognoses. However, long term survival is possible if patientspresent at an early stage. In England and Wales carcinoma ofthe stomach and pancreas cause about 7% and 4% of all cancerdeaths respectively. In women they are the fourth and fifth mostcommon causes of cancer death; in men their respectiverankings are third equal (with colonic cancer) and seventh.

The incidence of distal gastric carcinoma has fallen in theWest, probably because of decreasing rates of infection withHelicobacter pylori, but it remains one of the main causes ofdeath from malignancy worldwide. The incidence of proximalgastric cancer seems to be rising. These two gastric cancersdepend on the distribution and severity of H pylori gastritis, asdiscussed in the earlier chapter on the pathophysiology ofduodenal and gastric ulcers and gastric cancer.1

Cancer of the stomachGastric adenocarcinoma is rare below the age of 40 years, andits incidence peaks at about 60 years of age. Men are affectedtwice as often as women. Chronic atrophic pangastritisassociated with H pylori infection is one of the most importantrisk factors for distal gastric cancer.

Clinical presentationSymptoms may not occur until local disease is advanced.Patients may have symptoms and signs related to secondaryspread (principally to the liver) and to the general effects ofadvanced malignancy, such as weight loss, anorexia, or nausea.Epigastric pain is present in about 80% of patients and may besimilar to that from a benign gastric ulcer. If caused byobstruction of the gastric lumen, it is relieved by vomiting.Carcinoma of the gastric cardia may cause dysphagia.

Constant abdominal pain, and particularly back pain, aresinister symptoms implying local invasion by tumour. Chronicor acute bleeding from the tumour may occur, with consequentsymptoms. There is often little to be found on examination, butthere may be a palpable epigastric mass. The classic Troisier’ssign (left supraclavicular lymph node enlargement) is rare.

Investigations and stagingEndoscopy and barium meal are the principal investigations.Endoscopy allows direct visualisation and biopsy of thecarcinoma. Differentiation between benign and malignantgastric ulcers at endoscopy can be difficult, and several biopsiesare therefore taken (ideally six) from all parts of the ulcer.Diagnostic accuracy approaches 100% if 10 samples are taken.A benign gastric ulcer is probably not a premalignantcondition.

A barium study gives a better impression of the anatomy ofthe tumour and the degree of obstruction. It is also helpful fordiagnosis of linitis plastica, which may be missed at gastroscopy.In the presence of dysphagia it is important to request a bariumswallow and meal rather than a barium meal alone.

Endoscopy and barium studies are complementary. If thefirst investigation is negative in a patient with sinister symptomsthe other test is indicated. If a diagnosis of benign ulceration ismade it is essential to repeat the endoscopy and biopsies after

Box 12.1 Risk factors for gastric cancerx H pylori infection and atrophic gastritisx Pernicious anaemiax Adenomatous gastric polypsx Partial gastrectomyx Abnormalities in E-cadherin genex Family history of gastric cancer

Box 12.2 Signs and symptoms of gastric cancerSymptomsx Pain

EpigastricBack (advanced)

x Anorexiax Vomitingx Dysphagiax Iron deficiency anaemiax Haematemesis or melaenax Weight loss

Signsx Cachexia, weight loss, anaemiax Epigastric massx Hepatomegalyx Palpable left supraclavicularnode (Troisier’s sign)

Figure 12.1 Endoscopic appearance of gastriccarcinoma on the lesser curve of the stomach

Figure 12.2 Barium meal showing large obstructingcarcinoma of the body of the stomach

41

four to eight weeks of medical treatment to confirm ulcerhealing and the benign nature of the lesion.

Staging of the disease by computed tomography of thethorax and abdomen, and sometimes by laparoscopy orendoscopic ultrasonography, is appropriate only in thosepatients who are proceeding to surgery.

Differential diagnosisOnce a gastroscopy or barium study has been performed, thereare usually few problems with the diagnosis of gastriccarcinoma. The difficulty lies in deciding which patients needurgent investigation of their presenting symptoms. A goodinitial symptomatic response to acid suppression does notexclude malignancy. Guidelines from the British Society ofGastroenterology for the investigation of dyspepsia suggest thatall patients aged over 45 years should undergo endoscopy,whereas those under 45 need endoscopy only if they havesymptoms or signs that raise suspicion of malignancy.

TreatmentCurative treatmentThe decision to perform a gastrectomy depends on the patient’sgeneral state of health and nutrition and the preoperativestaging of the cancer. If there is no evidence of local invasion orof metastatic spread, resection is offered as a potential cure.Overall perioperative mortality is about 2%. Long term survivaldepends principally on the extent of lymph node metastases.

Chemotherapy may have an increasingly important role toplay in treating gastric carcinoma. Recent emphasis has been onpreoperative chemotherapy in order to “downstage” thetumour. There seems to be little place for radiotherapy in thetreatment of gastric carcinoma at present.

Palliative treatmentPatients with distal obstructing tumours may benefit from asubtotal gastrectomy or gastrojejunostomy despite the presenceof metastases. Stenting of tumours of the gastric cardia relievesdysphagia. Other treatments include endoscopic laser therapyfor unresectable obstruction or bleeding lesions. Bloodtransfusion may be appropriate for symptomatic anaemia. Themanagement of pain from gastric carcinoma follows establishedpalliative care practice. Coeliac plexus nerve blocks may beeffective. As with any malignant condition, the management ofsymptoms is multidisciplinary and is often led by palliative careand hospice based teams.

PrognosisThe disease is incurable in about half of patients atpresentation. With regional lymph node metastases, five yearsurvival after gastrectomy is about 10%. In those with onlyperigastric lymph node involvement survival rises to 30%, andin those with gastric carcinoma confined to the stomach fiveyear survival is about 70%. Only 10% of patients with hepaticmetastases survive a year.

Early gastric cancerEarly gastric cancer is a carcinoma diagnosed before it haspenetrated the full thickness of the stomach wall ormetastasised, but this accounts for less than 5% of gastriccarcinomata in the West. In Japan, where the incidence ofgastric carcinoma is much higher (about 10%), populationscreening detects a far greater proportion of asymptomaticearly gastric cancer. With aggressive surgery, five year survivalrates of 90% have been reported from Japan. It is unclear,however, whether these differences in survival are due to earlydetection, differences in the disease or its pathologicaldefinition, or operative technique.

Figure 12.3 Light micrograph of human stomach cancer. Most of the cellsseen here are cancerous, having large, irregular shapes and multiple nuclei

Oesophagus

Liver

Gall bladder

Oesophago-jejunalanastomosis

Jejunum

Roux loopof jejunum

DuodenumEntero-

enterostomy

Duodenalstump

Stomach

Pancreas

Figure 12.4 Total gastrectomy for treatment of gastric cancer (left) andsubsequent reconstruction by Roux-en-Y anastomosis (right)

Figure 12.5 Early gastric cancer. Top left: endoscopic appearance of cancerbefore dye spraying. Top right: the same lesion after spraying with 0.2%indigo carmine dye. Bottom left: lesion outlined by burn marks beforeexcision. Bottom right: mucosal defect after removal of the lesion with 1 cmmargin (blue colour is due to indigo carmine dye)


42

Cancer of the pancreasThe incidence of pancreatic cancer is about 10 per 100 000population in Western Europe. The incidence rises steadily withage, and the disease is slightly more common in men than inwomen. Alcohol, chronic pancreatitis, diabetes, and coffee donot predispose to pancreatic cancer.

Pathological featuresThe commonest pancreatic neoplasm is ductaladenocarcinoma. Most cancers arise in the head, neck, oruncinate process of the pancreas and may compress thecommon bile duct. Less than a third occur in the body and tailof the pancreas.

Periampullary malignancies may arise from the pancreas,the distal common bile duct, the ampulla of Vater, or theduodenum. Pancreatic carcinoma accounts for up to 90% of thisgroup, but the rest are important—periampullary tumourspresent early because they obstruct the common bile duct andcause jaundice when they are small, so they have betterprognoses than pancreatic carcinoma.

Clinical presentationThe classic presentation is painless, progressive, obstructivejaundice. Most patients also have epigastric discomfort or dullback pain. A large carcinoma of the head of the pancreas mayobstruct the gastric outlet. Symptoms from a carcinoma of thebody or tail of the pancreas are usually more vague, and thetumour is often locally advanced by the time of diagnosis.

Steatorrhoea may sometimes occur as a result of pancreaticduct obstruction and may be difficult to differentiate from thepale stool of obstructive jaundice. There are also the generaleffects of malignant disease.

The patient is usually jaundiced and may be anaemic orcachectic. There may be an epigastric mass or an irregular,enlarged liver because of metastases. Courvoisier’s law statesthat, in the presence of jaundice, a palpable gall bladder isunlikely to be due to gall stones. This is because stones usuallyresult in a fibrotic gall bladder, which will not distend in thepresence of obstruction of the common bile duct.

Investigations and stagingSerum biochemistry will confirm jaundice and also give someinformation about its cause: alkaline phosphatase and�-glutamyltransferase tend to be predominantly raised inobstructive jaundice. Disproportionate elevation of theaminotransferases (transaminases) leads to suspicion ofhepatocellular involvement. Tumour markers may be of value indiagnosis: carcinoembryonic antigen (the marker associatedwith colonic carcinoma) is elevated in up to 85% of cases.Raised serum levels of CA 19.9 are associated with carcinomaof the pancreas but also with obstruction of the common bileduct from any cause. Lack of tumour markers should not delayinvestigation of jaundiced patients; their main use is inmonitoring response to treatment and disease progression.

Ultrasonography is the initial investigation for patients withjaundice. A dilated common bile duct or intrahepatic ductsdifferentiate obstructive (posthepatic) jaundice from prehepaticand hepatic jaundice. Liver metastases are easily detected.

Endoscopic retrograde cholangiopancreatography visualisesthe common bile and pancreatic duct, and carcinoma of thehead of the pancreas produces a characteristic malignantstricture of the lower end of the common bile duct. Brushingscan be taken for cytological analysis, and the stricture may bedilated and stented to re-establish bile drainage into the

Box 12.3 Risk factors for pancreatic cancerx Smokingx Partial gastrectomyx Dietary fatx Family history of pancreatic cancer

Box 12.4 Signs and symptoms of pancreatic cancerSymptomsx Obstructive jaundice—dark urine,

pale stools, pruritusx Pain

Back (common)Epigastric

x Vomitingx Weight lossx Anorexiax Haematemesis or melaena (late)

Signsx Jaundicex Cachexia, anaemiax Epigastric mass (late)x Palpable gall bladder(Courvoisier’s sign)

Figure 12.6 Computed tomogram showing dilated intrahepatic ducts causedby an obstructing lesion of the lower end of the common bile duct

Figure 12.7 Endoscopic retrogradecholangiopancreatography showing lower common bileduct stricture (endoscope has been withdrawn)

Cancer of the stomach and pancreas

43

duodenum. The main complication is acute pancreatitis,especially if therapeutic procedures are performed.

Computed tomography further assesses the primary tumourand detects lymph node involvement and hepatic or pulmonarymetastases. If a mass is seen a fine needle aspirate can be takenunder tomographic or ultrasound guidance for cytology, whichhas a sensitivity (a positive result when tumour is present) ofabout 70%. A core biopsy for histology can also be obtained.

Differential diagnosisAnicteric patients with pancreatic carcinoma are usually initiallyinvestigated for their pain by gastroscopy or ultrasonography.Unless good views of the pancreas are obtained by the latter,computed tomography is required for the diagnosis.

Chronic pancreatitis may have a similar presentation, butthere is usually a history of alcohol misuse. However, the twoconditions may be radiologically indistinguishable, and fineneedle aspiration cytology or histological assessment is needed.For prognosis, it is important to distinguish malignantperiampullary lesions from tumours of the head of the pancreas.

TreatmentSurgery provides the only realistic hope of long term survival,but it is of value only if the primary tumour is no more than afew centimetres in diameter and is free of major blood vesselsand if there is no metastatic spread. Unfortunately, few patientsmeet these criteria.

Suitable patients undergo Whipple’s procedure. The head ofthe pancreas, the distal common bile duct, the gall bladder, andthe duodenum and distal stomach are excised. Reconstructioninvolves anastomosis of the pancreatic duct, the commonhepatic duct, and the distal stomach to a loop of jejunum.Perioperative mortality is now less than 5% in experiencedhands, and complication rates have decreased, but Whipple’sprocedure remains a formidable operation, and patients mustbe fit in order to be suitable. A modification allows preservationof the distal stomach and pylorus, which may have long termnutritional benefits.

Distal pancreatectomy may be suitable for carcinoma of thebody or tail, but few patients are suitable. Total pancreatectomyand extended vascular resections are rarely advocated.

Postoperative chemotherapy has been shown to be of somebenefit after pancreatic resection, and there is currently muchinterest in the role of new chemotherapeutic agents in pancreaticcancer. Postoperative radiotherapy has proved ineffective.

Palliative treatmentJaundice is palliated by stenting the stricture at the lower end ofthe common bile duct; this has superseded operative palliation.Some 15-20% of patients develop duodenal obstruction, whichcan be relieved by laparoscopic gastrojejunostomy. There is noindication for prophylactic gastrojejunostomy, because mostpatients die of their disease before duodenal obstructionbecomes a problem.

Palliation of pain and of other symptoms is best managedby a hospice based multidisciplinary palliative care team.Coeliac plexus block is often extremely valuable.

PrognosisThe prognosis of unresectable pancreatic carcinoma is poor,with few patients surviving longer than a year from diagnosis.Five year survival after resection for pancreatic carcinoma hassteadily improved and is now 10-20% in major centres. Thisrises to about 50% for resection of periampullary tumours.

1 Calam J, Baron JH. ABC of the upper gastrointestinal tract: Pathophysiology ofduodenal and gastric ulcer and gastric cancer. BMJ 2001;323:980-2.

The light micrograph of gastric cancer cells is reproduced with permissionof Science Photo Library/Parviz M Pour.

Figure 12.8 Fine needle aspiration of a pancreatic massunder computed tomographic guidance

Oesophagus

LiverStomach

Gall bladder

Commonbile duct Pancreas

Jejunum

Pancreaticduct

Duodenum

Hepatico-jejunostomy

Gastro-jejunostomy

Pancreatico-jejunostomy

Figure 12.9 In Whipple’s procedure for pancreatic cancer the head of thepancreas, distal common bile duct, gall bladder, duodenum, and distalstomach are excised (left). Reconstruction involves anastomosis of thepancreatic duct, common hepatic duct, and distal stomach to a loop ofjejunum (right)

Figure 12.10 Radiogram of stent placed to relieveduodenal obstruction caused by carcinoma of thepancreas


44

13 Anorexia, nausea, vomiting, and painR C Spiller

With the steady decline in diseases associated with Helicobacterpylori infection, the commonest diagnosis, both in generalpractice and hospital outpatients, is increasingly likely to befunctional dyspepsia, a condition that is ill understood and forwhich management is poorly defined.

Prevalence of symptomsAlthough upper abdominal and epigastric pain is extremelycommon and hence a poor discriminator of disease, only 2-8%of the general population experience anorexia, nausea, andvomiting, and so these are much more likely to indicate disease.Thankfully, many of these events turn out to be short lived, andonly about 25% of those affected consult their generalpractitioner, but this still accounts for 1-2% of all consultationsin general practice.

Two thirds of patients give the severity of their symptoms asa reason for consulting, but a similar proportion consultbecause of fear of serious disease, a factor that must beconsidered when planning management. The challenge is toreliably sift out and satisfactorily reassure the 40% withfunctional disease without missing those with more seriouspathology.

PathophysiologyAnorexia, nausea, and vomiting with pain can all be regardedteleologically as protective reflexes whereby the body preventsthe entry of toxins into the body. They also reduce the passageof chyme through diseased parts of the upper gut, therebyminimising further pain.

There are many possible organic causes, but, because thereis an important central component, these behaviour patternscan be learnt and may be anticipatory. Thus, patients about toreceive chemotherapy may vomit at the sight of the drugs,which they have previously associated with vomiting. Anxietyand depression can also be associated with alterations intaste—with associated anorexia, nausea, and weight loss—through neural pathways as yet poorly defined.

PharmacologyToxins and hypertonic saline induce vomiting by stimulatingafferent serotonergic nerves in the vagus that connect with thechemoreceptive trigger zone in the floor of the fourth ventricleof the brain. These afferent nerves can also respond to acid,amino acids, and fatty acids. 5-HT3 receptor antagonists act onthe vagal afferents to reduce nausea and emesis. Thechemoreceptive trigger zone also responds to bloodbornestimuli such as apomorphine, resulting in vomiting. Dopamine2 receptor blockers act here to inhibit emesis and the subjectivesense of nausea that precedes it. Excessive distension of the gutwill induce pain via serosal stretch receptors whose outputpasses via sympathetic neurones to the central nervous system,while ulcers cause acid related pain mostly via vagal afferents.

Differential diagnosisThe commonest cause of anorexia, nausea, vomiting, and painis duodenal ulcer disease, followed closely by functional

Table 13.1 Prevalence of symptoms ingeneral population

Symptom Frequency PrevalenceHeartburn > 1/month 24%

> 1/week 13%Upper abdominal pain >1/year 26%

> 6/year 16%Acid regurgitation > 1/month 11%

> 1/week 7%Upper abdominal painlasting > 2 hours

4%

Nausea > 1/month 8%> 1/week 3%

Vomiting > 1/month 2%Anorexia 4%Weight loss > 3 kg 3%

Table 13.2 Gastrointestinal symptoms in dyspepsia relateddiseases

Symptom

% of patients with specified symptom

FD Oes DU GU IBS GS ARD GCaAnorexia 40 35 47 56 35 29 55 64*Nausea 39 17 34 39 32 28 37 48*Vomiting 24 22 34 34 11* 23 59* 49GI bleed 12 14 26* 23 5 7 32 34*Heartburn 20 64* 32 23 12 19 25 22Weight loss 23 20 26 34 16* 32 33 72Psychotropicdrug use

46* 35 26 20 38 31 18 9*

FD= functional dyspepsia, Oes = oesophagitis and reflux without oesophagitis,DU = duodenal ulcer, GU= gastric ulcer, IBS = irritable bowel syndrome,GS = gallstone disease, ARD= alcohol related dyspepsia, GCa = gastric cancer.*Significant difference from other diseases

Figure 13.1 Taking an emetic by Isaac Cruickshank(1757-1810)

45

dyspepsia and irritable bowel syndrome. Gastric ulcer,gastro-oesophageal reflux, gastric cancer, and gall stonesaccount for 5-10% each, and rarer diseases such as diverticulardisease, small intestinal Crohn’s disease, colon cancer, andpancreatitis make up the rest.

Pregnancy can, for a short while, be a mysterious cause ofnausea and vomiting. Likewise, hepatitis during its prodrome,can be misleading, but the appearance of jaundice makes allclear. Even rarer are various metabolic diseases usuallydiagnosed in different contexts—such as diabetic ketoacidosis,renal tubular acidosis, and adrenocortical insufficiency—whichmay all present with anorexia, nausea, vomiting, and obscureabdominal pain.

The possibility of drug induced nausea should always beconsidered, especially with non-steroidal anti-inflammatorydrugs, opiates, antibiotics, hormone preparations, andchemotherapeutic agents.

Limitations of gastrointestinal symptomsMany studies have shown that diagnosis from individualgastrointestinal symptoms alone is difficult. Obviously differentdiseases do have a different spectrum of symptoms. Thuspatients with gastric cancer frequently complain of profoundanorexia, weight loss, and nausea while those with alcoholrelated dyspepsia typically vomit and retch especially first thingin the morning. Other differences that stand out include thefrequent use of psychotropic drugs in functional dyspepsia andthe high incidence of heartburn in oesophageal disease.

However, even seemingly serious indicators such assubstantial weight loss ( > 3 kg), anorexia, and nausea aresurprisingly common in functional dyspepsia, which, being somuch commoner than gastric cancer, accounts for far morecases. Thus, most symptoms are neither specific nor sensitive forany particular condition. Even the classic features of pepticulcer such as relief of pain by antacids or food and nocturnalpain, though commoner in peptic ulcer (65%,75%, and 75%respectively) are sufficiently common in functional dyspepsia(60%, 40%, and 43% respectively) to be unhelpful in differentialdiagnosis.

One exception is perhaps biliary colic. This typically occursin attacks with long periods of freedom from pain, which whenit comes is located in the right upper quadrant and radiates tothe tip of the shoulder. When pain comes on in the lateevening, lasts over two hours, and is associated with sweatingand vomiting then biliary colic is likely.

Demographic cluesAlthough symptoms are neither sensitive nor specific, includingdemographic details will improve matters. The ratio of organicto functional disease steadily increases with increasing age, sincemalignancy is very rare under the age of 45 (none in most caseseries), while the incidence of peptic ulcer increases linearlywith age because of the increasing incidence of H pyloriinfection. Smoking is also associated with an increased risk ofpeptic ulcer and gastric cancer. Sex may also be a usefulindicator: men are about twice as likely as women to haveduodenal ulcer or gastric cancer, whereas women have a50-60% increased risk of having irritable bowel syndrome andgallstone disease. Surprisingly, time taken off work fromfunctional dyspepsia is as great or greater than it is fromorganic disease.

Non-gastointestinal featuresFortunately, distinguishing duodenal ulcer and functionaldyspepsia, the two most likely causes, can be made easier byincluding more information about the patient. A history of

Box 13.1 Metabolic causes of anorexia,nausea, or vomitingx Diabetic ketoacidosisx Renal tubular acidosisx Hypercalcaemiax Adrenocortical insufficiencyx Other rare causes of acidosis or alkalosis

Anxiety stress

Chemoreceptive trigger zone

Reflux oesophagitisHiatus hernia

Druginducedgastritis

HormonesDrugsAcidosis

Gastric ulcer

Duodenal ulcer

Gall stonesPancreatic pseudocyst

Small bowel obstructionCrohn's diseaseColon cancer

Figure 13.2 Possible reasons for anorexia, nausea, andvomiting with pain

Gall stones Duodenalulcer

Gastric cancer

Gastriculcer

Miscellaneousorganic

Irritable bowelsyndrome

Functional dyspepsia

Gastro-oesophageal

reflux

Figure 13.3 Causes of anorexia, nausea, vomiting, andgastrointestinal pain

0

20

40

60

80

Age (years)

% o

rgan

ic d

isea

se

<25 26-40 >40

Figure 13.4 Likelihood of gastrointestinal symptomsbeing due to organic disease in different age groups


46

peptic ulcer is a strong predictor of further ulceration, whilepatients with functional dyspepsia score higher on depressionand anxiety and tend to exhibit “somatisation.”

Somatisation is characterised by recurrent multipleunrelated somatic complaints. They can be recognised byfrequent visits to the doctor for many non-gastrointestinaldisorders over the previous six months. These patients are moredissatisfied with their health care and perceive their health aspoor. Exhaustion, time off work, palpitations, chest pain,breathing difficulties, and musculoskeletal symptoms are allmore common in patients with functional dyspepsia.

H pylori statusH pylori infection, which can be reliably assessed from a ureabreath test, is present in 90-95% of patients with peptic ulcerbut in only 20-30% of those with functional dyspepsia. IfH pylori status is combined with the psychosocial assessmentsmentioned above, 95% of patients with peptic ulcer and 80% ofthose with functional dyspepsia can be accurately identified.

Physical examinationMost examinations will be normal, but the presence of anabdominal mass or succussion splash suggesting obstruction ofthe gastric outlet is ominous and indicates the need for urgentreferral, as does evidence of small bowel obstruction withabdominal distension and hyperactive bowel sounds.

ManagementAs always, this depends on a careful history and knowledge of apatient’s medical and psychosocial background, together withhis or her age.

Patients aged under 45Given that malignant disease is very rare under the age of 45years, it is reasonable to manage younger patients at initialpresentation with a trial of either a prokinetic drug for nauseaor an antisecretory agent for pain, together with appropriateadvice about lifestyle.

If symptoms don’t rapidly subside, however, they should befurther investigated. Cost benefit analysis suggests that, althoughinvestigation is initially expensive, it increases patientsatisfaction and is cheaper over the long term ( > 2 years) byreducing prescription costs and reconsultation rates. Testing forH pylori should be the first step, since a negative result allowsone to exclude peptic ulcer with 90% confidence, provided thatuse of non-steroidal anti-inflammatory drugs is excluded.

Lifestyle adjustments such as weight reduction andavoidance of foods that aggravate symptoms will produceimprovements in about half of patients, especially those withrecent weight gain and heartburn.

A good response to a therapeutic trial of proton pumpinhibitors supports a diagnosis of gastro-oesophageal reflux.Patients who fail to respond to treatment at this stage are likelyto have functional dyspepsia.

Patients aged over 45Since 66% of older patients are likely to have organic pathology,it is probably reasonable to refer all of them for furtherinvestigation, usually endoscopy. Testing for H pylori status isunlikely to be helpful in this group because infection is socommon and hence non-specific.

After endoscopyEndoscopy will give a specific diagnosis in 50-75% of cases. Ifsymptoms occur in discrete attacks then ultrasound

Box 13.2 Key points in differential diagnosisof anorexia, nausea, and vomiting withabdominal painx Peptic ulcer is the commonest single causex > 90% of patients with duodenal ulcer are

infected with H pylorix Functional dyspepsia, gastro-oesophageal reflux,

and irritable bowel syndrome account for abouthalf of cases

x Functional dyspepsia is characterised by frequentvisits to doctor for non-gastrointestinalconditions, ingestion of psychotropic drugs, andnegative H pylori status

x All patients aged > 45 years with new symptomslasting more than four weeks should be referredfor investigation

Box 13.3 Key points in management ofanorexia, nausea, and vomiting withabdominal painPatients aged <45 yearsx Persistent symptoms after a short trial of

antisecretory or prokinetic drugs should beinvestigated by testing for H pylori infection

x Early investigation reduces prescription costs andrepeat visits

x A negative test is as efficient as endoscopy inreassuring young patients that they do not haveserious pathology

Patients aged >45 yearsx Two thirds of patients will have organic

pathologyx Early endoscopy is recommendedx Positive tests for H pylori infection are sensitive

but not specific for peptic ulcer and hence oflittle value

Anorexia, nausea, or vomiting with abdominal pain

Test for H pylori infection

Age < 45 years Age > 45 years

H pylori eradicationtreatment

Treat as reflux orfunctional dyspepsia

Positive Negative

No Yes

Functional features?

Improved Not improved

Discuss patient's concerns and fears

Refer for endoscopy and possibly further opinion and tests

Improved Not improved Improved Not improved

Figure 13.5 Algorithm for management of anorexia,nausea, vomiting, and pain

Anorexia, nausea, vomiting, and pain

47

investigation of the gall bladder should be considered.Otherwise, persistent symptoms in the absence of somatisationdisorder warrant further investigation with barium followthrough and computed tomography of the abdomen. If thesefail to show any abnormality then functional dyspepsia is likely.

Functional dyspepsiaThe most demanding, difficult patient who is never satisfiedwith your efforts is the most likely to have functional dyspepsia.Making the effort to uncover the underlying psychopathologymay save many fruitless and elaborate tests. Active treatment ofovert psychiatric disease or cognitive behavioural therapy withmore limited aims may be appropriate for some patients.

Dyspepsia without obvious abnormality or somatisationThis should comprise a relatively small subgroup of patients forwhom further investigations are indicated. Since endoscopy isrelatively insensitive in diagnosing reflux, 24 hour oesophagealpH monitoring may be helpful. If pH monitoring is notavailable a therapeutic trial of proton pump inhibitors is areasonable alternative and may produce a positive response inup to half of patients.

Other possibilities include delayed gastric emptying, whichis seen almost exclusively in women. However, results fromgastric emptying studies do not correlate well with symptoms soit is probably reasonable to give women with unexplainednausea and vomiting a trial of a prokinetic drug such ascisapride or metoclopramide without first undertaking suchstudies. Failing this, domperidone should be tried; this centrallyacting dopamine blocker is inexpensive, well tolerated, andeffective against nausea of central origin.

Box 13.4 Clinical summaryx Diagnosing the causes of anorexia, nausea, and vomiting depends

on careful assessment of all the relevant features of patients’medical and drug histories together with physical examination

x Most patients are understandably anxious about the possibility ofserious disease and will require some investigation

x Although duodenal ulcer is currently the commonest cause, it isdeclining and practitioners should be aware of the high incidenceof functional dyspepsia in this setting

x By looking carefully for the signs of functional dyspepsia, doctorscan make a positive diagnosis rather than one of exclusion afterexhaustive and fruitless tests

Taking an emetic is reproduced with permission of the Wellcome Trust.


48

49

abdominal pain (upper)gall bladder 37–40nausea and vomiting with 45–8

achalasia 9, 11dysphagia caused by 12–13, 14, 15

acid perfusion test, in GORD 5, 10acid reflux

Barrett’s oesophagus and 13dyspepsia and 1, 33dysphagia and 15upper abdominal pain 33see also gastro-oesophageal reflux

“acid” regurgitation, in GORD 4acid secretion, gastric 19–20acid suppression therapy

in NSAID-associated ulcers 27–8in oesophagitis 6–7

acute cholecystitis 37acute upper gastrointestinal haemorrhage 30–2adenocarcinoma

dyspepsia in 33gastric 41–2

chronic atrophic pangastritis and 23dyspepsia and 33

oesophageal, risk of 6pancreatic 43

adrenaline injection, for non-variceal haemorrhage 32

aerophagy (air swallowing) 33, 36age

gastrointestinal haemorrhage and 30NSAID-associated ulcers 26–7, 27

AIDS related oesophagitis, dysphagia in 15air swallowing (aerophagy) 33, 36“alarm” symptoms

gastric cancer 34in GORD 5, 6

alcohol-related disordersdyspepsia 46gastrointestinal haemorrhage 31

alendronic acid, peptic stricture and 13alginates, for oesophagitis 6amoxicillin, in low dose triple therapy 24, 25angina, upper abdominal pain 8, 9–10, 33anorexia 45–8

in gastric cancer 41antacids

for dyspepsia 2for oesophagitis 6

anticoagulants, and NSAID-associated ulcers 27antidepressants, for functional dyspepsia 36antioxidant vitamins (lack of), dietary factor in

H pylori infection 21

antisecretory drugsfor dyspepsia 2for functional dyspepsia 36 for gastric ulcer 23interactions between H pylori and GORD 23

antispasmodic agents, for functional dyspepsia 36antral gastritis 17, 20anxiety

achalasia 9, 11dyspepsia and nausea and vomiting 46, 47

aortoduodenal fistula, haemorrhage and 31aspirin

gastrointestinal toxicity 26, 27haemorrhage and 30

atrophic gastritis 23autoimmune gastritis 17azapropazone, gastrointestinal toxicity 27

bacterial infection see Helicobacter pyloriballoon dilatation, gastro-oesophageal

sphincter 11barium radiology

in chest pain 10in dysphagia 12in gastric cancer 41in GORD 5, 6in nausea and vomiting 48

Barrett’s oesophagus 5, 6, 7dysphagia caused by 13–14oesophageal cancer risk 23

behavioural therapy, for functional dyspepsia 36, 48

Bernstein test 8beta-adrenoceptor blocker, in variceal

haemorrhage 32bile duct stones 37

management 39bile reflux 5

and Barrett’s oesophagus 13bile salt dissolution therapy 39bile salts, enterohepatic circulation 37, 38biliary colic 33, 37

characteristics of 34, 38, 38, 46biliary tract disease 33bismuth-based therapy

for functional dyspepsia 36for H pylori eradication 24

bloating, in dyspepsia 33, 34blood group, and H pylori infection 21botulinum toxin, for achalasia 11bougies, for dysphagia 12bowel ulceration, side-effect of NSAIDs 28brachytherapy, for oesophageal tumours 14

IndexWhere not already indexed with a text reference, page numbers in bold refer to illustrations; those in italic to tabulated or boxedmaterial

Index

50

breath test, C-ureaH pylori diagnosis 3, 18H pylori eradication and 22

bulbar palsy, dysphagia caused by 12

CA 19.9, tumour marker 43cagPI (pathogenicity island) strain, H pylori phenotype 21calcium channel blockers, for oesophageal spasm 11cancer see adenocarcinoma; carcinoma; gallbladder cancer;

gastric cancer; gastrointestinal cancer; oesophageal tumours; pancreatic cancer

candida (monilial) oesophagitis, dysphagia in 15carcinoembryonic antigen 43carcinoma

duodenal ulcer caused by 22oesophageal 14

Barrett’s oesophagus and 13dysphagia caused by 12–13

cardia, carcinoma of 12cardiac pain, as upper abdominal pain 8, 9–10, 33cardiomyotomy, for achalasia 11celecoxib, COX-2 inhibitor 29chemotherapy

in gastric cancer 42in oesophageal tumours 14in pancreatic cancer 44

chest paincardiac 8, 9–10, 33in GORD 4oesophageal 8–11

cholangiography 39cholecystectomy, laparoscopic 37, 39cholecystitis, acute 37cholecystostomy 39cholesterol stones 37chronic atrophic pangastritis, and gastric

cancer 23, 41chronic pancreatitis, differential diagnosis 43cigarette smoking 21, 24, 27, 46cimetidine, H2 receptor antagonist 7cineradiography, in dysphagia 12cisapride, prokinetic drug 48clarithromycin

in low dose triple therapy 24, 24, 25resistance to 24

CLO urease test, H pylori diagnosis 17–18colitis, side-effect of NSAIDs 28computed tomography

in nausea and vomiting 48in pancreatic cancer 42–3

“corkscrew oesophagus” 9corpus gastritis 20corrosive chemical ingestion, peptic stricture and 13corticosteroids, and NSAID-associated ulcers 27Courvoisier’s law 43COX-2 (cyclo-oxygenase-2 enzyme) inhibitors, new

generation NSAID 29cranial nerve palsy, dysphagia caused by 12cricopharyngeal spasm, dysphagia caused by 12Crohn’s disease 19, 22

gallstones and 37nausea and vomiting 46

culture, H pylori diagnosis 17, 18C-urea breath test

H pylori diagnosis 3, 18H pylori eradication and 22

cyclo-oxygenase-2 enzyme (COX-2) inhibitors, NSAIDs 29cytomegalovirus infection 22

dental enamel erosion, in GORD 4depression and anxiety

dyspepsia and nausea and vomiting 46, 47oesophageal pain 9, 11

diabetesnausea and vomiting 46upper abdominal pain 33

diclofenac, gastrointestinal toxicity 27dietary factors, in H pylori infection 21dietary fat

gallstones and 37pancreatic cancer risk 43

dietary modification, in dyspepsia 35diflunisal, gastrointestinal toxicity 27dilatation, gastro-oesophageal sphincter 11diverticular disease, nausea and vomiting 46domperidone

for functional dyspepsia 36for heartburn 7for nausea and vomiting 48

dopamine blocker, in nausea and vomiting 48drugs

dyspepsia caused by 35nausea and vomiting caused by 46see also NSAIDs

drug treatmentsfor gastrointestinal haemorrhage 32for NSAID-associated ulcers 27–8

duodenal ulcer 22gender and 46H pylori eradication and 25nausea and vomiting with 45–6pathophysiology 19–21side-effect of NSAIDs 28

duodenogastric reflux 5dyspepsia

“alarm” symptoms 1alcohol-related 46causes of 33–4functional see functional dyspepsiaguidelines for investigation 42implications for the NHS 1–3management in primary care 2–3NSAID-associated 28–9symptoms 45

dysphagia 12–15“alarm” symptoms 1, 5, 6

ear, nose and throat symptoms, in GORD 4ectopic pancreatic tissue 22electrocoagulation, gastrointestinal haemorrhage 32ELISA (enzyme-linked immunosorbent assay)

antibody test 18endoscopic retrograde cholangiopancreatography 43endoscopy

chest pain and 10cost/benefit of 1–3dyspepsia and 35dysphagia and 13–14gastric cancer and 41, 42gastric ulcer and 23gastrointestinal haemorrhage 30, 31–2GORD and 5, 6

H pylori diagnosis 17nausea and vomiting and 47–8

enterochromaffin-like cells, in gastric acid secretion 19epigastric pain 33

gastric cancer and 41nausea and vomiting with 45–8

ethanolamine, sclerosant intravariceal injection 32extracorporeal shock wave lithotripsy 39

faecal antigen test, H pylori diagnosis 18famotidine, H2 receptor antagonist 28fenoprofen, gastrointestinal toxicity 27functional dyspepsia 24, 34–6

acid secretion 19H pylori eradication and 25nausea and vomiting in 45–6, 46–7, 48

fungal strictures, peptic 13

gallbladdercancer of 40impaired emptying and gallstone formation 37, 38upper abdominal pain 37–40

gall stonesgender and 46management 39nausea and vomiting 46pain in 33, 34risk factors 37

gastrectomy 42pancreatic cancer risk 43

gastric acid secretion, regulation 19–20gastric cancer

adenocarcinoma 41–2chronic atrophic pangastritis and 23dyspepsia in 33

“alarm” symptoms 34detection 3dyspepsia in 1, 34gender and 46H pylori and 16, 17, 25nausea and vomiting 46pathophysiology 19–21prognosis 42risk factors 41signs and symptoms 41, 46treatment 42

gastric dysmotility, dyspepsia in 33, 35, 36gastric ulcer 23

bleeding risk 23H pylori eradication 25nausea and vomiting 46pathophysiology 19–21side-effect of NSAIDs 28

gastrin, in gastric acid secretion 19gastrin-secreting tumours 19, 22gastritis

acid secretion and 20atrophic 23chemical 28extent and location 20H pylori and 16, 17

gastroduodenal artery, erosion of 31gastrointestinal cancer, haemorrhage and 31gastrointestinal haemorrhage

acute upper 30–2dyspepsia and 45

Index

51

management 31–2gastro-oesophageal reflux

dysphagia caused by 12–13nausea and vomiting 46, 47predictors of 1see also acid reflux

gastro-oesophageal reflux disease (GORD) 4–7, 22, 23dyspepsia in 33H pylori and 23, 25

gastro-oesophageal sphincter, balloon dilatation 11gastrostomy, for dysphagia 12G cells, in gastric acid secretion 19globus hystericus 12GORD see gastro-oesophageal reflux diseasegranulomatous diseases, duodenal ulcer caused by 22

H2 receptor antagonistsfor duodenal ulcer 22for dyspepsia 2for functional dyspepsia 36, 36for NSAID-associated ulcers 28for oesophagitis 6–7

haemorrhage see gastrointestinal haemorrhageheartburn

in dyspepsia 1, 45in GORD 4–7, 33, 46

Helicobacter pyloriacid secretion and 19–21asymptomatic 24diagnosis 16–18

histology 17, 18serology 3, 18stool antigen tests 3, 18

dietary factors in 21dyspepsia and 3, 34epidemiology 16–18gastric cancer and 41GORD and 5–6, 23host factors 21management 22–5

antisecretory drugs and 23eradication 24

indications for 25in NSAID-associated ulcers 28–9

nausea and vomiting 46–7peptic ulcers and 23, 34phenotypes 21

Helicobacter pylori gastritis, in functional dyspepsia 34, 36herpes simplex infection 22hiatus hernia 46

oesophagitis and 4–5pain in 33

histamine, in gastric acid secretion 19hypercalcaemia

duodenal ulcer caused by 22upper abdominal pain 33

hyperparathyroidism, duodenal ulcer and 19hypertensive (“nutcracker”) oesophagus 9

ibuprofen, gastrointestinal toxicity 27indigestion

definition 33NSAIDs and 26–9

indometacin, gastrointestinal toxicity 27irritable bowel syndrome

biliary pain in 39

irritable bowel syndrome – Continueddyspepsia in 33, 35gender and 46nausea and vomiting with 45–6

jaundicegall stones and 38, 39nausea and vomiting 46pancreatic cancer and 43

ketoprofen, gastrointestinal toxicity 27

lansoprazole, proton pump inhibitor 7laparoscopic anti-reflux surgery 7laparoscopic cholecystectomy 37, 39lecithin, in gallstone formation 37leiomyoma 22leiomyosarcoma 22lithotripsy, extracorporeal shock wave 39liver disease, gastrointestinal haemorrhage and 31low dose triple therapy, H pylori eradication 24, 25lymph node metastases, in gastric cancer 42lymphoma 22

malignant disease see adenocarcinoma; carcinoma; gallbladdercancer; gastric cancer; gastrointestinal cancer; oesophagealtumours; pancreatic cancer

Mallory-Weiss tears, gastrointestinal haemorrhage and 31MALT lymphoma 25manometry, in achalasia 15mebeverine, antispasmodic agent 36meloxicam, new generation NSAID 29mesenteric vascular insufficiency, upper abdominal pain 33metastases, in pancreatic cancer 42–3metoclopramide

for heartburn 7for nausea and vomiting 48

metronidazole, in low dose triple therapy 24, 25migraine, sumatriptan cause of chest pain 9Minnesota (Sengstaken-Blakemore) tube 32misoprostol, prostaglandin E1 analogue 27, 29

for functional dyspepsia 36monilial (candida) oesophagitis, dysphagia in 15motility disorders

gastric, dyspepsia in 33, 35, 36oesophageal 8–11achalasia 15

motility modifying drugs, for oesophagitis 7mucosal irritation, oesophageal 8myasthenia gravis, dysphagia caused by 12

naproxengastrointestinal toxicity 27H pylori eradication and 28–9

nausea 45–8in gastric cancer 41

neoplasia, duodenal ulcer caused by 22NHS, cost of dyspepsia 1–3nitrites, for oesophageal spasm 11nitroimidazole antibiotics, resistance to 24non-steroidal anti-inflammatory drugs see NSAIDSnon-variceal haemorrhage, treatment 32NSAIDs

gastrointestinal toxicity 26–7, 28drug-induced nausea and vomiting 46drug-induced ulcers

Index

52

drug treatments for 27–8duodenal 19, 22, 34gastric 23, 34H pylori eradication 25

gastrointestinal haemorrhage 30–1indigestion 26–9peptic stricture 13

use of 26management of 27

“nutcracker oesophagus” 9

obesity, gallstones and 37 odynophagia 12oesophageal band ligation 32oesophageal manometry 10oesophageal pH monitoring see pH monitoring

(oesophageal)oesophageal provocation tests 10–11oesophageal tumours 14

adenocarcinoma, dyspepsia in 33cancer, dyspepsia in 1carcinoma, dysphagia caused by 12–13

oesophageal varices, haemorrhage and 31, 32oesophagitis

endoscopic findings 4gastrointestinal haemorrhage and 31in GORD 4–7reflux 1, 33side-effect of NSAIDs 28

oesophagogastrectomy 14oesophagus

chest pain 8–11“corkscrew oesophagus” 9dilatation for peptic stricture 13distention of 8–9heartburn and 4–7, 46motility disorders 8–11

dysphagia caused by 12“nutcracker oesophagus” 9rupture after dilatation 13, 15spasm 9, 11

dysphagia in 15stricture 13

side-effect of NSAIDs 28see also gastro-oesophageal reflux

omeprazole, proton pump inhibitor 7, 28, 29in triple therapy 25

oral contraceptive use, gallstones and 37orophangeal malignancy, dysphagia caused by 12

painbiliary colic 37, 38chest

differential diagnosis 8–11in GORD 4

gastric cancer 41nocturnal 46retrosternal 1

in hernia torsion 4upper abdominal 33

cardiac 8, 9–10, 33nausea and vomiting with 45–8

pain controlin gastric cancer 42in pancreatic cancer 44

pain threshold, psychological factors 9

palliative treatmentin gastric cancer 42in pancreatic cancer 44

pancreatectomy procedures 44pancreatic cancer 41, 43–4

prognosis 44treatment 44

pancreatic pseudocyst 46pancreatitis

chronic, differential diagnosis 43nausea and vomiting 46upper abdominal pain 33

pangastritis 17Barrett’s oesophagus and GORD and 23chronic atrophic, and gastric cancer 23, 41

pantoprazole, proton pump inhibitor 7parietal cells, in gastric acid secretion 19, 20Parkinson’s disease, dysphagia caused by 12patient counselling, in H pylori infection 24–5peptic stricture

dysphagia caused by 14postoperative 13

peptic ulcer 1–3dyspepsia in 1–3, 33–4haemorrhage and 30–1H pylori and 16, 17nausea and vomiting 46–7see also duodenal ulcer; gastric ulcer

periampullary tumours 43pernicious anaemia 19, 41pharyngeal pouch, dysphagia caused by 12pH monitoring (oesophageal) 5, 6

in acid reflx 8, 10in GORD 33in nausea and vomiting 48

pigment stones 37piroxicam, gastrointestinal toxicity 27polidoconal, sclerosant intravariceal injection 32potassium supplements, peptic stricture and 13pregnancy

gallstones and 37nausea and vomiting 46

prokinetic drugsfor functional dyspepsia 36for nausea and vomiting 48

propranolol, beta-adrenoceptor blocker, in variceal haemorrhage 32

prostaglandin E1 analogue 27proton pump inhibitors

action of 19costs of 2for chest pain 10for duodenal ulcer 22for functional dyspepsia 36for nausea and vomiting 47for NSAID-associated ulcers 28for oesophagitis 6–7for peptic stricture 13in low dose triple therapy 24, 25interactions between H pylori and GORD 6, 23

psychological factorsin achalasia 9, 11in chest pain 8, 9, 11in dyspepsia 35, 36

psychotherapyfor functional dyspepsia 36

Index

53

for nausea and vomiting in functional dyspepsia 48

psychotropic drugs, in functional dyspepsia 45, 46

quadruple therapy, H pylori eradication 24, 25

radiology, in dysphagia 12radiotherapy, for oesophageal tumours 14ranitidine, H2 receptor antagonist 7ranitidine-bismuth-citrate, H pylori eradication 24reflux, see also gastro-oesophageal refluxreflux oesophagitis, pain in 33reflux related dysmotility, dysphagia in 15renal disorders

nausea and vomiting 46upper abdominal pain 33

respiratory symptoms, in GORD 4resuscitation, haemorrhage and 31rofecoxib, COX-2 inhibitor 29rupture, after oesophageal dilatation 13, 15

salt, dietary factor in H pylori infection 21sarcoidosis 22Schatzki’s ring 13sclerosants, intravariceal injection 32selective seretonin reuptake inhibitor 36Sengstaken-Blakemore (Minnesota) tube 32serology, H pylori diagnosis 3, 18smoking 21, 24, 27, 46sodium tetradecyl sulphate, sclerosant intravariceal

injection 32somatisation, dyspepsia and nausea and vomiting 46, 47spasm, oesophageal 9, 11, 15sphincter of Oddi, high pressure 39stents, oesophageal 14stomach

cancer of 41–2oesophagogastrectomy 14see also gastric

stool antigen tests, H pylori diagnosis 3, 18strictures

oesophageal 4, 7peptic 13, 14

stroke, dysphagia caused by 12sucralfate, for functional dyspepsia 36sulindac, gastrointestinal toxicity 27sumatriptan, chest pain caused by 9syphilis 22

terlipressin, vasoactive drug 32tolmetin, gastrointestinal toxicity 27transjugular intrahepatic portosystemic

shunt (TIPPS) 32tricyclic antidepressants

for achalasia 9, 11for functional dyspepsia 36

Troisier’s sign 41tuberculosis 22tumour markers, in pancreatic cancer 43tumours see adenocarcinoma; carcinoma; gallbladder

cancer; gastric cancer; gastrointestinal cancer; oesophageal tumours; pancreatic cancer

tumours, duodenal ulcer caused by 22

ulcerNSAID-associated 27–8, 28–9see also duodenal ulcer; gastric ulcer; peptic ulcer

ultrasonographygall stone test 38in jaundice 43

upper abdominal paingall bladder 37–40nausea and vomiting with 45–8

urea breath test, H pylori diagnosis 18urease tests, H pylori diagnosis 17–18

vacuolating toxin (Vac A toxin), in H pyloriphenotype 21

Index

54

variceal haemorrhage, treatment 32vasoactive drugs, gastrointestinal haemorrhage 32vomiting 45–8

in GORD 4in hernia torsion 4

weight loss, in dyspepsia 45, 46Whipple’s procedure 44

Zollinger-Ellison syndrome, gastrin-secreting tumour 19, 22

Date post:	16-Aug-2015
Category:	Health & Medicine
Upload:	scu-hospital
View:	29 times
Download:	1 times

Abc of the upper gastrointestinal

Health & Medicine