Investigation of the stomach and duodenum

Flexible endoscopy

Amongst all of the methods used to investigate and image the stomach and duodenum, flexible endoscopy is now the ‘gold standard’. The original gastroscopes were fibre-optic (Hirschowitz), but now most use a solid-state camera mounted at the instrument’s tip (Figs 51.5). The main advantage of the modern instruments is that they do not need the fragile fibre optic fibre bundle to transmit the image. In addition, as the output is via a monitor rather than an eyepiece, the other members of the endoscopy team see the image. This is useful when taking biopsies or performing interventional techniques, and also facilitates teaching and training.

Flexible endoscopy is more sensitive than conventional radiology in the assessment of the majority of gastroduodenal conditions. This is particularly the case with peptic ulceration, gastritis and duodenitis. In upper gastrointestinal bleeding endoscopy is far superior to any other investigation and in most circumstances is the only imaging required. Although in Japan double-contrast barium meals performed by very experienced radiologists are able to detect quite small gastric cancers, endoscopy is far superior in most centres and also allows biopsies to be taken.

Although fibre-optic endoscopy is a safe and commonly used investigation, it is important that all personnel under­taking this procedure are adequately trained and that resusci­tation facilities are always available. Although the morbidity and mortality associated with upper gastrointestinal endoscopy is extremely low, it is not without hazard. Careless and rough handling of the endoscope during intubation of a patient may result in perforations of the pharynx and oesophagus. Any other part of the upper gastrointestinal tract may also be perforated. An inadequately performed endoscopy is also dangerous as a serious condition may be overlooked. This is particularly the case in respect of early and curable gastric cancer, the appearances of which may often be extremely subtle and may be missed by inexpe­rienced endoscopists with tragic consequences. In general, a more experienced endoscopists will have a higher index of suspicion for any mucosal abnormalities and will take more biopsies.

Upper gastrointestinal endoscopy is normally carried out under sedation usually with incremental doses of diazepam or midazolam until the patient is adequately sedated. Midazolam is often preferred for its amnesic effect. Sedation is of particular concern in the case of gastrointestinal bleed­ing as it may have a more profound effect on the patient’s cardiovascular stability. It has now become the standard to use pulse oximetry to monitor patients during upper gastro intestinal endoscopy, and nasal oxygen is often also administered. Opiates are not usually necessary, although they are commonly used for endoscopic retrograde cholangio­pancreatography (ERCP). Buscopan is useful to abolish duodenal motility for examinations of the second and third parts of the duodenum. Examinations of this type ate best carried out using a side-viewing endoscope such as is used for

ERCP

Some patients are relatively resistant to sedation with benzodiazipenes, particularly those who are accustomed to alcoholic beverages. Increasing the dose of benzodiazipenes in these patients may not afford any useful sedation but merely make the patient more restless and confused. Such patients are sometimes better endoscoped fully awake using a local anesthetic throat spray and a narrow-gauge endoscope. Whatever the circumstances, it is important that resuscitation facilities are available including agents that reverse the effects of benzodiazipenes, such as flumazenil.

The technology associated with upper gastrointestinal endoscopy is continuing to advance. Instrumentation is now available in many centres which allows both endoscopy and endoluminal ultrasound to be performed simultaneously (see later). Intervention via the endoscope is also developing rapidly. A variety of haemostatic measures is used in the treatment of bleeding ulcers such as injection with various substances, diathermy, heater probes and lasers. These approaches appear to have utility in the treatment of bleed­ing ulcers, although good controlled trials in this area are not abundant. There is no good evidence that such interventional procedures at the moment work in patients who are bleeding from very large vessels, such as the gastroduodenal artery or splenic artery.

Contrast radiology

Upper gastrointestinal radiology is not used as much as in previous years as endoscopy is a more sensitive investigation for most gastric problems. There is, however, a number of circumstances where the barium meal is of great value and augments the value of endoscopy. These include large hiatus hernias of the rolling type and chronic gastric volvulus where it may be difficult for the endoscopists to determine exactly the anatomy or, indeed, negotiate the deformity to see the distal stomach. Linitus plastica may be missed by even relatively experienced endoscopists as the mucosal aspect of the stomach may not look particularly abnormal. This condition may be diagnosed more easily by using contrast radiology, although this is of limited value to the patient as the outlook is so poor.

Ultrasonography

Standard ultrasound imaging can be used to investigate the stomach, particularly in patients with neoplasia. Thickening of the gastric wall can be seen in malignancy, some assessment made of local invasion, and liver and peritoneal disease is often detected. However, used, conventionally, it is less sensitive than other modalities. By contrast, endoluminal ultrasound and laparoscopic ultrasound are probably the most sensitive techniques available in the preoperative staging of gastric cancer. In endoluminal ultrasound the transducer is usually attached to the distal tip of the instrument. However, devices have been developed which may be passed down the biopsy channel, albeit with poorer image quality. Five layers (Fig. 51.7) of the gastric wall may be identified on endoluminal ultrasound and the depth of invasion of a tumour can be assessed with exquisite accuracy (90 per cent accuracy for the ‘T’ component of the staging). Enlarged lymph nodes can also be identified and the technique’s accuracy in this situation is about 80 pet cent. Finally, it may be possible to identify liver metastases not seen on axial imaging. Laparoscopic ultrasound is also a very sensitive imaging modality, espe­cially when combined with the laparoscopy (see later). It is one of the most sensitive methods of detecting liver metastases from gastric cancer.

An additional use of ultrasound is in the assessment of gastric emptying. Swallowed contrast is utilised which is designed to be easily seen using an ultrasound transducer. The emptying of this contrast is then followed directly. The accuracy of the technique is similar to that of radioisotope gastric emptying studies (below).

Computerised tomography (CT) scanning and magnetic resonance imaging (MRI)

The resolution of the CT scanners is continuing to improve and this form of axial imaging is of increasing value in the investigation of the stomach, especially gastric malignancies (Fig. 51.8). The presence of gastric wall thickening associated with a carcinoma of any reasonable size can be easily detected by CT but the investigation lacks sensitivity in detecting smaller and curable lesions. It is much less accurate in ‘T’ staging than endoluminal ultrasound. Lymph node enlargement can be detected and, based on the size and shape of the nodes, it is possible to be reasonably accurate in detecting nodal involvement with tumour. However, as with all imaging techniques, it is limited. Microscopic tumour deposits in lymph nodes cannot be detected when the node is not enlarged and, by contrast, lymph nodes may undergo reactive enlargement but not contain tumour. These problems apply to all imaging techniques. The detection of small liver metastases is improving, although in general terms metastases from gastric cancer are less easy to detect using CT than those, for instance, from colorectal cancer. This is because metastases from gastric cancer may be of the same density as liver and may not handle the intravenous contrast any differently. MRI scanning does not at present offer any specific advantage in assessing the stomach, although it has a higher sensitivity for the detection of gastric cancer liver metastases than conventional CT imaging.

Laparoscopy

This technique is now well used in the assessment of patients with gastric cancer. Its particular value is in the detection of peritoneal disease that is difficult by any other technique, unless the patient has ascites or bulky intraperitoneal disease. Its main limitation is in the evaluation of posterior extension but other techniques are available to evaluate posterior invasion, especially CT and endoluminal ultrasound. Usually laparoscopy is combined with peritoneal cytology unless laparotomy follows immediately.

Gastric emptying studies

These are useful in the study of gastric dysmótility problems, particularly those that follow gastric surgery. The principle of the examination is that a radioisotope-labelled liquid and solid meal are ingested by the patient and the emptying of the stomach is followed on a gamma camera. This allows the

proportion of activity in the remaining in the stomach to be assessed numerically, and it is possible to follow liquid and solid gastric emptying independently (Fig. 51.9).

  Tests of gastric acid secretion and of pH monitoring

In the past tests of gastric acid secretion were frequently performed, particularly by surgeons. Recent advances, however, have made these tests virtually redundant except in the context of physiological and pharmacological studies. Part of the interest in these tests related to the higher levels of gastric acid secretion commonly found in patients with duodenal ulcer disease (see later).

Traditionally, basal and maximal acid output is measured. A nasogastric tube is passed into the stomach, the basal secretion collected over 1 hour and the acid output in millimoles calculated. To obtain the maximal acid output intramuscular injection of the gastrin analogue pentagastrin is given at a dose of 6 pg/kg body weight and the secretions are collected over the ensuing hour. The maximal acid output is calculated as the peak 15-minute collection multiplied by 4 or twice the peak 30-minute collection in two consecutive collections.

A wealth of data exists on the acid outputs in a whole range of populations. Ultimately, the peak acid output is related to the parietal cell mass; larger stomachs produce more acid. Although traditionally patients with duodenal ulcers do secrete more acid than normal, there is considerable overlap between normal and duodenal ulcer populations. Patients with gastric ulcers produce acid at approximately normal levels. In a patient with a gastrinoma the basal acid output is unusually high and there may be little response to pentagastrin, the parietal cell mass already being near maximally stimulated by the gastrin produced by the tumour. It is now known that other nongastrin peptide products also may result in the Zollinger—Ellison syndrome.

The insulin test was previously beloved of gastric surgeons, particularly those interested in the quality of vagotomy. The test was originally described by Hollander and involves the induction of hypoglycemia in a postvagotomy patient using an intravenous soluble dose of insulin of 0.2 units/kg body weight. Gastric acid secretion is then measured in the 2 hours following insulin injection. The theoretical basis of the test is that the hypoglycaemia stimulates the hypothalamic nuclei inducing a parasympathetic response. If the vagus is intact to any degree this is reflected in a rise in gastric acid secretion. Various criteria are used to determine whether a test is positive or not. Distinction is often made between an ‘early’ positive test, thought to mean that a substantial vagal innervation remains, and a ‘late’ positive test, the significance of which is less clear. Although in these tests they did correlate to some degree with ulcer relapse following vagotomy, their use in patient manage­ment was always limited as they are performed postoperatively and therefore could not influence patient management. In this respect the intraoperative Grassi test may have been a more useful method of controlling the quality of vagotomy. However, all of these tests are virtually redundant as vagotomy is now uncommonly performed on the elective situation.

24-hour intragastric pH monitoring

Studies of this type became very popular during the 198 Os, particularly in the investigation of new gastric antisecretory agents. The pH within the stomach is measured over a 24-hour period either by the passage of a nasogastric tube and regular aspiration or by placing a radiotelemetry capsule on a tether within the stomach and monitoring the pH with an externally worn aerial. The median daytime, night-time or 24-hour intragastric pH can thus easily be calculated and, by converting the pH to hydrogen ion concentration, the median intragastric acidity over various periods can be studied. This latter manoeuvre allows the percentage reduction in acidity produced by a pharmacological agent to be measured (the step of converting pH to hydrogen ion concentration is obviously vital as pH is a logarithmic scale). Although of great physiological and pharmacological interest, these tests have little clinical relevance.

Measurement of plasma gastrin

The measurement of plasma gastrin by radioimmunoassay is of use in the diagnosis of gastrinoma (Zollinger—Ellison syn­drome). In most assays the normal fasting gastrin level is about 50 ng/litre, but in gastrinomas very high levels, some­times many thousands of ng/litre, can be found. However, the other common cause of hypergastrinaemia is hypochlorhydria associated with gastric atrophy and very high gastrin levels are found in pernicious anaemia. Antral gastrin is released to excess as a result of the negative feedback loop. A condition is said to exist when there is autonomous production of gastrin by the antrum but this condition, if it exists, is of little clinical importance