This is the oesophagus of a 50 year old man complaining of intermittent dysphagia.

What is the likely diagnosis?
a) eosinophilic oesophagitis
b) oesophageal candiasis
c) achalasia
d) diffuse oesophageal spasm
e) oesophageal atony
In the video clip you will note "trachealisation" of the oesophagus with longitudinal furrows (in the clip this is seen best at the squamo-columnar junction) and white spots, due to eosinophilic micro-abscesses. The patient had a history of "allergies to seafood" and had to be careful not too eat too fast due to the risk of vomiting.  However, in the last 10 years he had only required endoscopic removal of a food bolus on three occasions.  Whilst children often present with abdominal pain, vomiting and dysphagia, adults may also complain of dysphagia, heartburn, or a food bolus impaction. Patients frequently have a history of allergy (asthma, hay fever, eczema) and there is a male predominance (70% in adults). Diagnosist criteria includes (1) clinical symptoms of esophageal dysfunction and (2) >15 eosinophils/HPF. As eosinophils may also be found in reflux oesophagitis, a lack of responsiveness to high-dose PPI is added to the list above by some authorities.  Althought the above criteria may seem straight forward, there are challenges such as standardisation of oesophageal biopsy protocols, uncertainty in how eosinophil counts are determined and variability in the size of microscope high-power fields.  My personal preference is to take 2 biopsies at three levels in the oesophagus. Histologically, the eosinophilic infiltrate can be surprisingly patchy. In addition, there may be eosinophilic micro-abscesses (easily seen as white spots in the video). In addition, up to one third of patients have a dense fibrotic response in the submucosa.  This may be evident as "easy ripping" of the surface mucosa at intubation or after dilatation.   The condition is thought to be due to aberrant responses to environmental antigens (mainly food).  There appears to be an antigen-driven hypersensitivity reaction characterized by a mixed IgE-dependent/delayed-type reaction at least partially attributable to IL-13 (a TH2 cytokine).

Treatment options include elimination diets or elemental diet as well as repeated short courses of topical fluticasone propionate. Those who believe in an overlap between reflux disease and eosinophilic oesophagitis would also advocate PPI therapy.  Strictures can be dilated and there is evidence that symptoms then improve for at least a year. There have also been some reports of improvement with the leukotriene receptor antagonist montelukast and promising early results with mepolizumab, a monoclonal antibody against interleukin-5.

This was found in an elderly patient investigated for an iron deficiency anaemia.

What is the most likely diagnosis? a) HP associated gastritis
b) Gastric CMV infection
c) Cameron's ulceration
d) Gastric amyloidosis
e) Linitus plastica These are Cameron ulcers first described by Cameron and Higgins in 1986 (presumably in a post-mortem series?).  They are typically seen in at the site of large hiatal hernias. Although most may well be asymptomatic, the classical association is with iron deficiency anemia.  The exact pathogenesis is not known but the location suggests that it has something to do with the "to-and-fro" motion of the stomach through the diaphragmal hiatus. There is no particular association with HP infection. For these reasons, this type of ulceration may not improve with PPI therapy.  Nevertheless, empirical treatment would be with PPI therapy and (if indicated) iron supplementation. In a study by Moskovitz et.al (Am J Gastroenterol 1992;87:622–626), 64% of cases healed with H2 blockers and presumably PPI therapy is even more effective. Surgical repair, although curative, is not always possible as most patients are elderly.

An Italian lady presents with abdominal pain immediately after returning from holiday in Calabria. At gastroscopy this worm is found attached to the gastric wall.


What is the most likely diagnosis? a) Hook worm
b) Anisakis
c) Roundworm
d) Threadworm
e) Tapeworm
Antonello TreccaThis is an Anisakis simplex worm which can infect humans who eat undercooked, fresh fish caught in the sea.  The infection is most common in areas of the world where fish is consumed raw or undercooked.  Freezing the fish will also kill the worm and for this reason some countries insist that all fish must be frozen at -20°C or below for at least seven days before consumption. Soon after ingestion, the patient develops abdominal pain.  If the larvae pass into the small bowel, a severe eosinophilic granulomatous response may develop mimicking Crohn's disease.  Less commonly, patients present with urticaria and anaphylaxis. The diagnosis is endoscopic as the 2 cm worms can be seen attached to the mucosa.  Ascaris worms (roundworms) are much longer (12-40cm), hookworms are smaller (2-3mm) and are attached to the duodenal and/or small bowel mucosa - not the gastric mucosa. Threadworms are commony found in the colon and tapeworms are huge. Surprisingly, there is no medical treatment although there have been reports of response to albendazole.  If found at endoscopy, the worm should be physically removed by biopsy. In marine mammals, eggs are excreted in the faeces.  The eggs are eaten by crustaceans which in turn are eated by fish or squid.  In these, the worm burrows through the wall of the gut and encysts outside of the viscera. The infected fish or squid is then eaten by a marine mammal (whale, seal, or dolphin).  Humans are of course "dead end hosts" but as marine mammals have similar mucosa to that of humans, the worm can feed from the human gut. 

This shows the distal oesophagus of a 65 year old man on a Barrett's surveillance programme. He is entirely asymptomatic, although samples taken 12 and 6 month earlier had found high-grade dysplasia within the Barrett's. He has now been referred for a second opinion. His oesophagus is examined with high-definition, narrow band imaging, indigo carmine dye spray and autofluorescence.

What is the correct management?
a) Take another round of samples
b) Remove the unstable areas identified with the trimodality imaging
c) Ablate the Barrett's using photodynamic imaging
d) Ablate the Barrett's using radiofrequency ablation
e) Refer the patient for an oesophagectomy
Even with white light you can see that the mucosa is nodular.  This is the appearance of dysplastic Barrett’s, remarkable similar to the appearance of dysplastic ulcerative colitis.  The indigo-carmine image emphasizes this further.  Towards the end of the video clip you see the auto-fluoroscopic images.  It takes some getting used to and the technology is plagued by a lot of false positives.  Nevertheless, the video indicate that there are multiple areas of unstable Barrett's (pink patches).  However, a more important issue than all of this is the area of stricturing.  The presence of a stricture is suspicious for invasive Barrett's cancer.  Even if the strictured area only contains high-grade dysplasia, it would be difficult to treat endoscopically at the stiff ablation catheters will not be able to treat the sharp angulation.  For these reasons, the correct answer is E - refer for surgical resection!  Analysis of the resected oesophagus confirmed an invasive cancer (sm2 invasion) but all lymphnodes were negati ve for cancer. I have reservations about our ability to accurately identify pathes of Barrett's dysplasia by endoscopic means.  On careful reading of most studies, you will conclude that random biopsies are still needed.  Taking "mapped samples" whereby you carefully record the level   (in cm from the mouth) and the laterial location (in a clock face fashion) is a good way to confirm precisely where the dysplasia is located.  Naturally, if there is a small, visible nodule, this would not be needed.

This 8mm lesion was found in the mid-oesophagus of a 60 year old woman.

This lesion was found in the mid-oesophagus.  Which of the following options is correct? a) this is a benign lesion and as samples are unlikely to be diagnostic no biopsies are needed
b) this lesion is benign and biopsies will confirm this
c) this lesion is indeterminate and should be extensively sampled
d) this lesion is likely to be an early malignancy which can be removed endoscopically
e) this lesion is likely to be cancer which can not be removed endoscopically This depressed (ulcer-like) lesion is very suspicious.  Benign pathology such as granular cell tumours and leiomyomas will not be ulcerated.  The two most likely causes are: 1) primary squamous cell carcinoma or 2) a metastases. “Technically” it is possible to remove this lesion endoscopically.  In fact, I was able to easily remove it using the “Cook Duette” kit.   However, as I rather suspected, although the cancer was superficial there was invasion of the lymphatic channels and a positive deep margin.  Subsequently, some of my peers thought that I should not have removed the lesion.  They argued that instead I should have made an endoscopic diagnosis that the lesion was too advanced for endoscopic resection.  Unfortunately, the Multidisciplinary Cancer Teams in England always want precise histological confirmation before deciding on treatment.  For example, even if a colonic lesion looks like a cancer on abdominal CT, endoscopy and biopsies are still required.  The reason for this is that to subject a patient to potentially hazardous surgery to treat benign disease or to remove lesions which can be removed endoscopically is now becoming unacceptable practise.  In removing this lesion endoscopically (using the Duette kit), I probably subjected the patient to a less than 1:200 risk of a perforation and less than 1:50 risk of late bleeding.  I believe that these are acceptable risks to achieve a precise diagnosis.  After all, analysis of the EMR specimen provided more information than could have been obtained from surface biopsies, EUS and CT.  However, I agree that it would NOT have been acceptable to attempt to remove this lesion by ESD.  This technique subjects the patient to a 1:20 risk of perforation which could lead to wide dissemination of cancerous cells throughout the chest.

What is the diagnosis?

This is the sigmoid colon in a 55 year old man undergoing colonoscopy to investigate a recent change in bowel habit.  On the video you can see that the mucosa is covered with what appears to be numerous polyps.  What is the diagnosis? The clue is that when the nodules are pricked with a needle, they seem to deflate somewhat! 

The patient has pneumatosis intestinalis.  There are two forms of the disease: Primary idiopathic pneumatosis intestinalis (which is said to constitute some 15 - 20% of cases although I don’t think that there is reliable data on this) and secondary to conditions such as Chronic Obstructive Pulmonary Disease or end-stage ischaemic colitis, often associated with gas in the portal system.  In infants, the condition is associated with necrotizing enterocolitis. In the benign idiopathic condition, the patient is asymptomatic and the multiple thin-walled cysts, about 1 cm in size, develop in the colon or small intestine.  The gas pockets are then incidental findings at endoscopy or X-ray.   In the published literature many conditions have been linked with the condition.  However, many of these associations probably reflect the numerous indications for the endoscopy at which the pneumatosis was detected, (rather than being a cause for the condition). I have only seen the primary condition myself on two occasions.  In both cases, the cysts were confined to the colon (more on the left than the right side) and the patient symptoms could not be explained by the pneumatosis.  Treatment of the pneumatosis therefore did not seem to be relevant. There are several aetiological theories for the condition:
Mechanical: Luminal gas entering the through a defect in the mucosa. 
Pulmonary: increased intra-thoracic pressure allows air to dissect along vessels in the mediastinum and retroperitoneum eventually ending up in the wall of the intestines.  However gas has not been found outside of the intestinal wall in the condition.
Bacterial: The presence of an unknown gas producing bacterial species which creates small pockets of gas within the intestinal wall.  Treatment with high flow oxygen and metronidazole have been described although the real question is – apart from patients moribund from an ischaemic colitis, when does the condition require treatment?

This ulcer was found in the sigmoid colon of a 35 yr old man with ulcerative colitis

He was admitted with a week history of worsening bloody diarrhoea and high fever. He had been diagnosed with ulcerative colitis 10 years earlier. His medication on admission included prednisolone 20 mg per day, mesalazine 400 mg three times a day, and azathioprine 50 mg twice a day. On examination he is pyrexial with a temperature of 38.5°C and is tender in the left iliac fossa.
Hb 13.5 g/dl            Na 140 mmol/l
WBC 3.2 x 10 9        K  4.0 mmol/l
Plat 300 x 109            Urea 7.5 mmol/l
ESR 18 mm/h        Creat 52 µmol/l
CRP 120                Glucose 5.2 mmol/l
Bili 15 µmol/l
ALP 990 iu/l
AST 375 iu/l
Albumin 34 g/l
The leucopenia progressed 48 hours after admission (white cell count 2.6 x 109/l, neutrophils 2.0 x 109/l). Blood, stool, and urine cultures are all negative. In view of the abnormal liver function tests and history of ulcerative colitis, an MRCP is performed to investigate the possibility of primary sclerosing cholangitis, but this is normal.
The patient is started on empirical intravenous cefuroxime and metronidazole 72 hours after admission. However, the swinging pyrexia does not respond to the antibiotics. A colonoscopy is carried out (see video file) and biopsy specimens are taken (photograph).  What is the diagnosis?
a) ulcerative colitis
b) Crohn’s disease
c) CMV colitis
d) Ischaemic colitis
e) HIV associated colitis A diagnosis of cytomegalovirus colitis induced immunosuppression was made on the basis of the classical appearance of ‘‘owl’s eye’’ inclusion bodies. The patient was started on IV ganciclovir (5 mg/kg) on day 16 of the admission and became apyrexial 48 hours later. Patients with inflammatory bowel disease (particularly ulcerative colitis) are predisposed to developing an acute exacerbation secondary to cytomegalovirus disease.  In most cases patients are receiving some immune modulatory therapy before the onset of symptoms. As in this case, CMV can give rise to a self limiting hepatitis.  Mortality rates for patients with cytomegalovirus enterocolitis have been quoted as high as 70%. 

This EMR specimen has been pinned up up-side-down . The flat polyp had just been resected from the sigmoid colon. As the polyp is being mounted on a polystyrene board, the patient begins to complain of abdominal discomfort. Can you tell why?

 

To the inexperienced, a therapeutic colonic perforation is easily missed.  The enclosed video demonstrated the classical appearance of a perforation.   In most cases, the patient remains comfortable for a period before vague abdominal pains develop.  The patient is different to the usual wind-like low abdominal discomfort, relieved by passing wind which most patients suffer for a few hours after their examination.  The pain of a perforation is continuous, rapidly worsening and occasionally pleuritic.  It is my own practice, after colonic polypectomy, to explain to the patient what type of discomfort is to be expected and what are the symptoms of a delayed perforation or significant late bleeding.  Naturally, all patients are also given a copy of their colonoscopy report, a written information leaflet about their polypectomy or EMR procedure together with 24hour contact numbers (including my own mobile number).  After all resections, I scrutinise the specimen for the tell-tale disk of muscle propria which indicates that the resection has been too deep.  In this image, you can clearly see a paler disc-like area in the centre of the up-side-down specimen.   The video shows the corresponding EMR procedure.  The lesion is a serrated adenoma which hasn’t lifted very well due to previous sampling.  For this reason, I have selected a very stiff snare, pressing it firmly onto the lesion.  The cutting was taking rather too long and, anticipating a perforation, I had the clips ready.  In the video clip you can see how I turn the EMR specimen over in the colon to confirm the presence of a muscle propria disk.  The defect is closed endoscopically, the patient was commenced on intravenous antibiotics and admitted for overnight observation.  Provided that the patient remains completely well and apyrexial, I can be persuaded to discharge younger patients the following day.  However, when to discharge depends on your local policies, the opinion of your colorectal surgeons and the age and comorbidities of your patient.  In my care, these patients never have an abdominal X-ray or CT scan.  The reason for this is that the scan will usually show either local air or air in the peritoneum.  As the defect has been clipped shut, the decision to operate is now clinical rather than radiological.   Should the patient deteriorate, surgical repair should be carried out as early as possible and preferably within a few hours.  When patients develop “perforation-like” pain within the endoscopy unit following an EMR, I will usually re-intubate and close the complete mucosal defect with clips.   Surprisingly, the pain very quickly settles after the perforation is closed endoscopically.  However, patients who develop pain after discharge from the endoscopy unit, usually have an abdominal CT to distinguish a perforation from a “post-polypectomy syndrome”.  If there is radiological evidence of a full thickness perforation, immediate surgery is scheduled.

Case details:

This is a 55 year old lady undergoing flexible sigmoidoscopy to investigate her recurrent rectal prolapse.  This polypoid lesion was found on retroverting the endoscope in the rectum.  On the right is a close-up of the lesion.  Four biopsies were taken from the lesion without difficulty.

What is the most likely histology?

  1. Fibro-epithelial polyp
  2. Inflammatory polyp
  3. Adenoma
  4. Adenocarcinoma
  5. Squamous cell carcinoma

Explanation 1:

The polyp is angry red, with normal small round crypts and a couple of superficial erosions (appear as specs of white mucus).  Histology confirmed that this was an inflammatory polyp.  You may ask how to correctly differentiate between this and a fibro-epithelial polyp?  It’s simple – the patient will cry out aloud when you sink the teeth of your biopsy forceps into any rectal lesion covered with squamous mucosa.

Explanation 2:

Rectal Prolapse

Frequency: The peak incidence is within the fourth and seventh decades of life.  Up to 90% of patients are women.  Although multiple pregnancies are often implicated in the aetiology, 35% of patients are nulliparous.

Aetiology: The aetiology of rectal prolapse is unknown, but it is often associated with long-standing constipation. Other predisposing conditions include chronic straining during defecation, pregnancy, previous surgery, and neurological disease. The pathophysiology of rectal prolapse is also not completely understood or agreed upon.

The are two main theories:

  1. The rectal prolapse is a sliding hernia through a defect in the pelvic fascia.
  2. The rectal prolapse starts as a circumferential internal intussusception of the rectum beginning 6-8 cm proximal to the anal verge. With time and straining, this progresses to full-thickness rectal prolapse, although some patients never progress beyond this stage.

Certain anatomic features found during surgery for rectal prolapse are common to most patients. These features include a patulous or weak anal sphincter with levator diastasis, deep anterior Douglas cul-de-sac, poor posterior rectal fixation with a long rectal mesentery, and redundant rectosigmoid. Whether these anatomic features are the cause or result of the prolapsing rectum is not known.  The condition is often concurrent with pelvic floor descent and prolapse of other pelvic floor organs such as the uterus or the bladder. 

Clinical details: Three different clinical entities are often combined and called rectal prolapse: full-thickness rectal prolapse, mucosal prolapse, and internal prolapse (internal intussusception). Treatment of these 3 entities differs.  Full-thickness rectal prolapse is the most commonly recognized type and is defined as protrusion of the full thickness of the rectal wall through the anus.  In mucosal prolapse, only the rectal mucosa (not the entire wall) protrudes from the anus.  Internal intussusception may be a full thickness or a partial rectal wall disorder, but the prolapsed tissue does not pass beyond the anal canal and does not pass out of the anus.

Patients with rectal prolapse report a mass protruding through the anus. Initially, the mass protrudes from the anus only after a bowel movement and usually retracts when the patient stands up. As the disease process progresses, the mass protrude more often, especially with straining and Valsalva manoeuvres such as sneezing or coughing. Finally, the rectum prolapses with daily activities such as walking and may progress to continual prolapse.  In addition to a protruding mass from the anus, patients often report faecal incontinence.

Patients with mucosal prolapse have similar problems but not to the same degree.  Mucosal prolapse typically exhibits radial folds of the protruding mucosa instead of the concentric rings of a full rectal prolapse.

Case details:

This is a 55 year old lady undergoing flexible sigmoidoscopy to investigate her recurrent rectal prolapse.  This polypoid lesion was found on retroverting the endoscope in the rectum.  On the right is a close-up of the lesion.  Four biopsies were taken from the lesion without difficulty. What is the most likely histology?
  1. Fibro-epithelial polyp
  2. Inflammatory polyp
  3. Adenoma
  4. Adenocarcinoma
  5. Squamous cell carcinoma

Explanation 1:

The polyp is angry red, with normal small round crypts and a couple of superficial erosions (appear as specs of white mucus).  Histology confirmed that this was an inflammatory polyp.  You may ask how to correctly differentiate between this and a fibro-epithelial polyp?  It’s simple – the patient will cry out aloud when you sink the teeth of your biopsy forceps into any rectal lesion covered with squamous mucosa.

Explanation 2:

Rectal Prolapse

Frequency: The peak incidence is within the fourth and seventh decades of life.  Up to 90% of patients are women.  Although multiple pregnancies are often implicated in the aetiology, 35% of patients are nulliparous. Aetiology: The aetiology of rectal prolapse is unknown, but it is often associated with long-standing constipation. Other predisposing conditions include chronic straining during defecation, pregnancy, previous surgery, and neurological disease. The pathophysiology of rectal prolapse is also not completely understood or agreed upon. The are two main theories:
  1. The rectal prolapse is a sliding hernia through a defect in the pelvic fascia.
  2. The rectal prolapse starts as a circumferential internal intussusception of the rectum beginning 6-8 cm proximal to the anal verge. With time and straining, this progresses to full-thickness rectal prolapse, although some patients never progress beyond this stage.
Certain anatomic features found during surgery for rectal prolapse are common to most patients. These features include a patulous or weak anal sphincter with levator diastasis, deep anterior Douglas cul-de-sac, poor posterior rectal fixation with a long rectal mesentery, and redundant rectosigmoid. Whether these anatomic features are the cause or result of the prolapsing rectum is not known.  The condition is often concurrent with pelvic floor descent and prolapse of other pelvic floor organs such as the uterus or the bladder.  Clinical details: Three different clinical entities are often combined and called rectal prolapse: full-thickness rectal prolapse, mucosal prolapse, and internal prolapse (internal intussusception). Treatment of these 3 entities differs.  Full-thickness rectal prolapse is the most commonly recognized type and is defined as protrusion of the full thickness of the rectal wall through the anus.  In mucosal prolapse, only the rectal mucosa (not the entire wall) protrudes from the anus.  Internal intussusception may be a full thickness or a partial rectal wall disorder, but the prolapsed tissue does not pass beyond the anal canal and does not pass out of the anus. Patients with rectal prolapse report a mass protruding through the anus. Initially, the mass protrudes from the anus only after a bowel movement and usually retracts when the patient stands up. As the disease process progresses, the mass protrude more often, especially with straining and Valsalva manoeuvres such as sneezing or coughing. Finally, the rectum prolapses with daily activities such as walking and may progress to continual prolapse.  In addition to a protruding mass from the anus, patients often report faecal incontinence. Patients with mucosal prolapse have similar problems but not to the same degree.  Mucosal prolapse typically exhibits radial folds of the protruding mucosa instead of the concentric rings of a full rectal prolapse.

This nodule was found in the oesophagus of a 30 year old lady undergoing gastroscopy because of dyspepsia.

This is the typical appearance of a granular cell tumour (Abrikossoff’s tumour) first described in 1926 by Abrikossoff, a Russian pathologist. These lesions can probably develop anywhere.  Outside of the GI tract, the lesions are most commonly found at the tongue, head and neck.  Within the GI tract, the oesophagus is the most common location. Lesions are rarely multiple (about 10% of cases).  Granular cell tumours are said to be more common in women and black people.  The endoscopic features are typical as is histology showing cells containing an eosinophilic (pink/red) cytoplasm resembling granules which stain positive with the Periodic Acid Schiff stain.  Immunohistochemical staining will be positive for S-100 protein and neuron-specific enolase (NSE).  Stains for desmin, vimentin, and cytokeratin will be negative.  Surface biopsies may fail to confirm the diagnosis as these nodules are submucosal.  The differential diagnosis is that of a small lipoma or a leiomyoma (which are much more common in the oesophagus than in the stomach).   When detected these lesions are usually referred for endoscopic resection. Unfortunately, as the lesions arise from the deep submucosa, the local recurrence is high (30-40%).  For this reason I now treat the raw EMR surface, immediately post-resection with APC set at low power (35W).  A tiny proportion of these lesions (said to be less than 2%) are said to be malignant overall.  However, the published literature is confusing as GI and non-GI lesions are frequently combined.  Perhaps the only compelling reason for removing these from the oesophagus, is that it is cheaper than surveillance!  In the attached video, a small granular cell tumour is removed using the Olympus EMR cap.  In a sedated and compliant European patient, you will probably have a “therapeutic window” of 20 minutes.  Longer procedures are better carried out under general anaesthesia.  All my oesophageal EMR’s are now carried out using the “Duette banding kit”.  With this equipment, the resections can be carried out without the need for a submucosal injection in less than 3 minutes.  Is used to have some trouble with bleeding and now set my ERBE diathermy unit on “level 4” (setting the device at a maximum proportion of coagulation in the blend mix when I press the yellow pedal) and “endocut” 120W.  The Endocut setting senses the amount of power required and will only give you what is needed to cut the polyp up to a maximum ceiling power which you have selected (if more power is needed the device alarms).  However, I know of some who carry out oesophageal EMR’s using “pure coagulation” (blue pedal).  
<1 2 3 4